From bugzilla-daemon at cs.uiuc.edu  Thu Jan  1 09:52:48 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 1 Jan 2009 09:52:48 -0600
Subject: [LLVMbugs] [Bug 3274] i1 promotion is not simplified by LegalizeDAG
In-Reply-To: <bug-3274-206@http.llvm.org/bugs/>
Message-ID: <200901011552.n01FqmAT032363@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3274


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Duncan Sands <baldrick at free.fr>  2009-01-01 09:52:47 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20081229/071590.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  2 02:41:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Jan 2009 02:41:21 -0600
Subject: [LLVMbugs] [Bug 3277] New: llvm-g++ front-end outputing code that
 is not folded by irbuilder?
Message-ID: <bug-3277-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3277

           Summary: llvm-g++ front-end outputing code that is not folded by
                    irbuilder?
           Product: tools
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: slow-compile
          Severity: normal
          Priority: P2
         Component: llvm-g++
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


On this testcase (Reduced from kc++):
class impl_abstract_phylum {
public:
    virtual int subphylum(int) const;
    virtual ~impl_abstract_phylum() { }
};
int impl_abstract_phylum::subphylum(int no) const {
  return 0;
}

llvm-g++ -O0 is producing:

define linkonce void
@_ZN20impl_abstract_phylumD1Ev(%struct.impl_abstract_phylum* %this) nounwind {
entry:
...
        br label %bb

bb:             ; preds = %entry
        %2 = trunc i32 0 to i8          ; <i8> [#uses=1]
        %toBool = icmp ne i8 %2, 0              ; <i1> [#uses=1]
        br i1 %toBool, label %bb1, label %bb2

bb1:            ; preds = %bb
        %3 = load %struct.impl_abstract_phylum** %this_addr, align 4    
        %4 = bitcast %struct.impl_abstract_phylum* %3 to i8*            ; <i8*>
[#uses=1]
        call void @_ZdlPv(i8* %4) nounwind
        br label %bb2


This is curious for two reasons: 1) we're emitting obviously dead code, which
slows down compilation at -O0.  2) IRbuilder should be folding the %2 truncate
instruction!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  2 08:23:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Jan 2009 08:23:35 -0600
Subject: [LLVMbugs] [Bug 3278] New: ParseAssemblyString does not add
 definitions to the passed module anymore
Message-ID: <bug-3278-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3278

           Summary: ParseAssemblyString does not add definitions to the
                    passed module anymore
           Product: new-bugs
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: cmdkeen at gmx.de
                CC: llvmbugs at cs.uiuc.edu


The behaviour of ParseAssemblyString changed between the LLVM 2.1 and 2.2
releases. Until 2.1 a call like this

Module* parsedModule = ParseAssemblyString( code, targetModule, &errorInfo );

added the globals, types and functions in code to targetModule. Since 2.2 this
does not happen anymore.

The documentation still states that targetModule is "A module to add the
assembly too.".

Also see this thread
http://lists.cs.uiuc.edu/pipermail/llvmdev/2008-December/018994.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  2 22:06:36 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Jan 2009 22:06:36 -0600
Subject: [LLVMbugs] [Bug 3210] missing CPUID entry for Core2 Quad
In-Reply-To: <bug-3210-206@http.llvm.org/bugs/>
Message-ID: <200901030406.n0346a1t012493@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3210


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Evan Cheng <evan.cheng at apple.com>  2009-01-02 22:06:35 ---
Ah, I didn't see your proposed patch until now. Anyway, it's fixed with 61555,
61556, and 61602. I wonder if we should add a new processor names for Atom and
Corei7.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan  3 02:40:13 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Jan 2009 02:40:13 -0600
Subject: [LLVMbugs] [Bug 3257] llvm-gcc rejects "+g" constraint
In-Reply-To: <bug-3257-206@http.llvm.org/bugs/>
Message-ID: <200901030840.n038eDCG022515@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3257


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |evan.cheng at apple.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Evan Cheng <evan.cheng at apple.com>  2009-01-03 02:40:13 ---
What does this inline asm mean? :-) Anyway, it's fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20081229/071672.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan  3 03:31:14 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Jan 2009 03:31:14 -0600
Subject: [LLVMbugs] [Bug 3279] New: Failed bootstrap: Unsupported asm: input
 constraint with a matching output constraint of incompatible type!
Message-ID: <bug-3279-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3279

           Summary: Failed bootstrap: Unsupported asm: input constraint with
                    a matching output constraint of incompatible type!
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


With SVN r61604 llvm-gcc bootstrap fails in libgomp:
$ /home/edwin/llvm-bootstrap/llvm-obj/./gcc/xgcc
-B/home/edwin/llvm-bootstrap/llvm-obj/./gcc/
-B/home/edwin/llvm-bootstrap/llvm-obj/../install/x86_64-unknown-linux-gnu/bin/
-B/home/edwin/llvm-bootstrap/llvm-obj/../install/x86_64-unknown-linux-gnu/lib/
-isystem
/home/edwin/llvm-bootstrap/llvm-obj/../install/x86_64-unknown-linux-gnu/include
-isystem
/home/edwin/llvm-bootstrap/llvm-obj/../install/x86_64-unknown-linux-gnu/sys-include
-DHAVE_CONFIG_H -I. -I../../../llvm-gcc4.2/libgomp -I.
-I../../../llvm-gcc4.2/libgomp/config/linux/x86
-I../../../llvm-gcc4.2/libgomp/config/linux
-I../../../llvm-gcc4.2/libgomp/config/posix -I../../../llvm-gcc4.2/libgomp
-Wall -Werror -pthread -ftls-model=initial-exec -O2 -g -O2 -MT lock.lo -MD -MP
-MF .deps/lock.Tpo -c ../../../llvm-gcc4.2/libgomp/config/linux/lock.c  -fPIC
-DPIC -o .libs/lock.o
Unsupported asm: input constraint with a matching output constraint of
incompatible type!

Bugpoint reduced testcase:
; ModuleID = 'bugpoint-reduced-simplified.bc'
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-unknown-linux-gnu"
        %struct.omp_nest_lock_t = type { i32, i32 }

@gomp_ialias_omp_init_lock = alias void (i32*)* @omp_init_lock          ; <void
(i32*)*> [#uses=0]
@gomp_ialias_omp_init_nest_lock = alias void (%struct.omp_nest_lock_t*)*
@omp_init_nest_lock            ; <void (%struct.omp_nest_lock_t*)*> [#uses=0]
@gomp_ialias_omp_destroy_lock = alias void (i32*)* @omp_destroy_lock           
; <void (i32*)*> [#uses=0]
@gomp_ialias_omp_destroy_nest_lock = alias void (%struct.omp_nest_lock_t*)*
@omp_destroy_nest_lock         ; <void (%struct.omp_nest_lock_t*)*> [#uses=0]
@gomp_ialias_omp_set_lock = alias void (i32*)* @omp_set_lock            ; <void
(i32*)*> [#uses=0]
@gomp_ialias_omp_set_nest_lock = alias void (%struct.omp_nest_lock_t*)*
@omp_set_nest_lock              ; <void (%struct.omp_nest_lock_t*)*> [#uses=0]
@gomp_ialias_omp_unset_lock = alias void (i32*)* @omp_unset_lock               
; <void (i32*)*> [#uses=0]
@gomp_ialias_omp_unset_nest_lock = alias void (%struct.omp_nest_lock_t*)*
@omp_unset_nest_lock          ; <void (%struct.omp_nest_lock_t*)*> [#uses=0]
@gomp_ialias_omp_test_lock = alias i32 (i32*)* @omp_test_lock           ; <i32
(i32*)*> [#uses=0]
@gomp_ialias_omp_test_nest_lock = alias i32 (%struct.omp_nest_lock_t*)*
@omp_test_nest_lock             ; <i32 (%struct.omp_nest_lock_t*)*> [#uses=0]

declare void @omp_init_lock(i32* nocapture) nounwind

declare void @omp_destroy_lock(i32* nocapture) nounwind readnone

declare void @omp_destroy_nest_lock(%struct.omp_nest_lock_t* nocapture)
nounwind readnone

declare i32 @omp_test_nest_lock(%struct.omp_nest_lock_t*) nounwind

declare void @omp_set_nest_lock(%struct.omp_nest_lock_t*) nounwind

define void @omp_unset_nest_lock(%struct.omp_nest_lock_t* %lock) nounwind {
entry:
        br i1 false, label %bb, label %return

bb:             ; preds = %entry
        %asmtmp.i = tail call i64 asm sideeffect "syscall",
"={ax},0,{di},{si},{dx},~{dirflag},~{fpsr},~{flags},~{memory},~{cx},~{r11}"(i32
202, i32* null, i32 1, i32 1) nounwind          ; <i64> [#uses=0]
        ret void

return:         ; preds = %entry
        ret void
}

declare void @omp_init_nest_lock(%struct.omp_nest_lock_t* nocapture) nounwind

declare i32 @omp_test_lock(i32*) nounwind

declare void @omp_unset_lock(i32*) nounwind

declare void @omp_set_lock(i32*) nounwind


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan  3 13:24:34 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Jan 2009 13:24:34 -0600
Subject: [LLVMbugs] [Bug 3250] ice: shiftAmt <= BitWidth && "Invalid shift
	amount"
In-Reply-To: <bug-3250-206@http.llvm.org/bugs/>
Message-ID: <200901031924.n03JOYSZ020204@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3250


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |gohman at apple.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Dan Gohman <gohman at apple.com>  2009-01-03 13:24:32 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20081229/071686.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan  4 14:45:22 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 Jan 2009 14:45:22 -0600
Subject: [LLVMbugs] [Bug 3278] ParseAssemblyString does not add definitions
 to the passed module anymore
In-Reply-To: <bug-3278-206@http.llvm.org/bugs/>
Message-ID: <200901042045.n04KjMj1009434@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3278


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-04 14:45:21 ---
Sorry for the delay on this, and thanks for filing the bugzilla!  This patch
should fix it, please verify:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20081229/071712.html

-Chris


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan  4 21:41:41 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 Jan 2009 21:41:41 -0600
Subject: [LLVMbugs] [Bug 3281] New: diagnostics in LLParser
Message-ID: <bug-3281-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3281

           Summary: diagnostics in LLParser
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM assembly language parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: gohman at apple.com
                CC: llvmbugs at cs.uiuc.edu


I ran a bunch of polygen-generated testcases through
the new AsmParser implementation. Attached are some
testcases that either trigger assertion failures
due to llvm-as not diagnosing invalid constructs,
some that are rejected by the old AsmParser and
accepted by the new AsmParser and look questionable,
and some that are rejected by the new AsmParser that
look questionable.

FWIW, the new AsmParser also correctly diagnoses many
errors that the old one missed :-).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan  5 02:27:04 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Jan 2009 02:27:04 -0600
Subject: [LLVMbugs] [Bug 3281] diagnostics in LLParser
In-Reply-To: <bug-3281-206@http.llvm.org/bugs/>
Message-ID: <200901050827.n058R4tn031919@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3281


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Chris Lattner <clattner at apple.com>  2009-01-05 02:27:04 ---
I fixed the rest in a series of patches.

It would be nice to remove deprecated crapola from your polygen grammar, such
as attributes on function types, vicmp/vfcmp etc.  See the FIXMEs in
LLParser.cpp for the deprecated stuff.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan  5 02:53:57 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Jan 2009 02:53:57 -0600
Subject: [LLVMbugs] [Bug 3279] Failed bootstrap: Unsupported asm: input
 constraint with a matching output constraint of incompatible type!
In-Reply-To: <bug-3279-206@http.llvm.org/bugs/>
Message-ID: <200901050853.n058rvl6006137@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3279


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Evan Cheng <evan.cheng at apple.com>  2009-01-05 02:53:56 ---
Doh. I accidentally inverted a comparison in 61604. This is fixed now. Sorry
about the breakge.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan  5 03:37:25 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Jan 2009 03:37:25 -0600
Subject: [LLVMbugs] [Bug 3282] New: New testsuite failure: test/Assembler/
 2005-05-05-OpaqueUndefValues.ll
Message-ID: <bug-3282-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3282

           Summary: New testsuite failure: test/Assembler/2005-05-05-
                    OpaqueUndefValues.ll
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


According to buildbot this fails since r61682.

FAIL:
/home/edwin/llvm-bootstrap/llvm/test/Assembler/2005-05-05-OpaqueUndefValues.ll
Failed with exit(1) at line 1
while running:  llvm-as <
/home/edwin/llvm-bootstrap/llvm/test/Assembler/2005-05-05-OpaqueUndefValues.ll
|  llvm-dis | llvm-as
BC<C0><DE>!^L^@^@ESC^@^@^@^A^P^@^@^P^@^@^@^G<81>#<91>A<C8>^DI^F^P29<92>^A<84>^L%^E^H^Y^^^D<8B>b<80>^@E^BB
<92>^KB^D^P2^T^XI
2D$H
<90>!#Dr<80><8C>^@!<86>
<8A>
d^L^_^@^@Q ^@^@^F^@^@^@ e<82><84>^@&@H^E        &@<C6>^EB^B^@^@^@^@^@^@llvm-as:
<stdin>:4:16: invalid type for undef constant
@x = global %t undef
               ^
llvm-dis: Invalid bitcode signaturesh: acroread: command not found

I don't know if the test is wrong, or asmparser is wrong.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan  5 11:40:05 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Jan 2009 11:40:05 -0600
Subject: [LLVMbugs] [Bug 3281] diagnostics in LLParser
In-Reply-To: <bug-3281-206@http.llvm.org/bugs/>
Message-ID: <200901051740.n05He5cn029692@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3281


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |




--- Comment #6 from Dan Gohman <gohman at apple.com>  2009-01-05 11:40:04 ---
Here are a few more polygen-generated testcases, tested
with the latest svn.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan  5 12:12:42 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Jan 2009 12:12:42 -0600
Subject: [LLVMbugs] [Bug 3282] New testsuite failure: test/Assembler/
 2005-05-05-OpaqueUndefValues.ll
In-Reply-To: <bug-3282-206@http.llvm.org/bugs/>
Message-ID: <200901051812.n05ICg4d030937@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3282


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-05 12:12:42 ---
Fixed, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090105/071773.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan  5 17:46:18 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Jan 2009 17:46:18 -0600
Subject: [LLVMbugs] [Bug 3235] instcombine crash on i128
In-Reply-To: <bug-3235-206@http.llvm.org/bugs/>
Message-ID: <200901052346.n05NkI1Q010549@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3235


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #7 from Chris Lattner <clattner at apple.com>  2009-01-05 17:46:18 ---
Fixed, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090105/071824.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan  6 05:21:25 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Jan 2009 05:21:25 -0600
Subject: [LLVMbugs] [Bug 3283] New: llc crash on random .bc file
Message-ID: <bug-3283-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3283

           Summary: llc crash on random .bc file
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


With SVN r61805 I run:
$ llvm-as <tst.ll >x.bc && llc x.bc -f

I get this:

llc: TargetData.cpp:468: unsigned char llvm::TargetData::getAlignment(const
llvm::Type*, bool) const: Assertion `Ty->isSized() && "
Cannot getTypeInfo() on a type that is unsized!"' failed.
0   llc       0x0000000000c2dd1f
1   libc.so.6 0x00007fd7d0afcf60
2   libc.so.6 0x00007fd7d0afced5 gsignal + 53
3   libc.so.6 0x00007fd7d0afe3f3 abort + 387
4   libc.so.6 0x00007fd7d0af5dc9 __assert_fail + 233
5   llc       0x0000000000b5ee5e llvm::TargetData::getAlignment(llvm::Type
const*, bool) const + 222
6   llc       0x0000000000718e76
7   llc       0x000000000071a4d9
8   llc       0x0000000000bc2ff6 llvm::FunctionPassManager::doFinalization() +
198
9   llc       0x0000000000500618 main + 4216
10  libc.so.6 0x00007fd7d0ae91a6 __libc_start_main + 230
11  llc       0x00000000004feae9 std::ios_base::Init::~Init() + 57
Aborted

This is either a bug in llvm-as (for accepting invalid bitcode), or a bug in
llc, this is the .ll file:
hidden thread_local addrspace(39) constant opaque undef


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan  6 05:56:52 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Jan 2009 05:56:52 -0600
Subject: [LLVMbugs] [Bug 3284] New: llvm-as crashing on 'fp128
	zeroinitializer'
Message-ID: <bug-3284-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3284

           Summary: llvm-as crashing on 'fp128 zeroinitializer'
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Using SVN r61805:

llvm-as: APFloat.cpp:2849: void llvm::APFloat::initFromAPInt(const
llvm::APInt&, bool): Assertion `0' failed.
0   llvm-as   0x000000000054a03f
1   libc.so.6 0x00007f39c38d0f60
2   libc.so.6 0x00007f39c38d0ed5 gsignal + 53
3   libc.so.6 0x00007f39c38d23f3 abort + 387
4   libc.so.6 0x00007f39c38c9dc9 __assert_fail + 233
5   llvm-as   0x000000000051b053 llvm::APFloat::APFloat(llvm::APInt const&,
bool) + 51
6   llvm-as   0x0000000000498634 llvm::Constant::getNullValue(llvm::Type
const*) + 500
7   llvm-as   0x000000000046b8f8
llvm::LLParser::ConvertGlobalValIDToValue(llvm::Type const*, llvm::ValID&,
llvm::Constant*&) + 360
8   llvm-as   0x0000000000472563 llvm::LLParser::ParseGlobalValue(llvm::Type
const*, llvm::Constant*&) + 211
9   llvm-as   0x0000000000472668 llvm::LLParser::ParseGlobal(std::string
const&, char const*, unsigned int, bool, unsigned int) + 248
10  llvm-as   0x0000000000479368 llvm::LLParser::ParseNamedGlobal() + 248
11  llvm-as   0x000000000047aa2d llvm::LLParser::ParseTopLevelEntities() + 189
12  llvm-as   0x000000000047aeb4 llvm::LLParser::Run() + 20
13  llvm-as   0x00000000004627b0 llvm::ParseAssemblyFile(std::string const&,
llvm::ParseError&) + 480
14  llvm-as   0x000000000044dc3a main + 154
15  libc.so.6 0x00007f39c38bd1a6 __libc_start_main + 230
16  llvm-as   0x000000000044d719 __gxx_personality_v0 + 433
Aborted

Testcase:
@o = linkonce protected constant fp128 zeroinitializer


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From yonggangluo at hotmail.com  Tue Jan  6 08:02:07 2009
From: yonggangluo at hotmail.com (=?gb2312?B?WW9uZ2dhbmcgTHVvICjC3tPCuNUp?=)
Date: Tue, 6 Jan 2009 22:02:07 +0800
Subject: [LLVMbugs] Cmake report errors on svn 61806
Message-ID: <BAY110-DS377E4A7109F744ECB4D54A9DE0@phx.gbl>

CMake Error at cmake/modules/GetTargetTriple.cmake:14 (message): 
Failed to execute D:/svn/llvm/autoconf/config.guess

Call Stack (most recent call first):

cmake/config-ix.cmake:59 (get_target_triple)

CMakeLists.txt:105 (include)







What's happened?



I am using TDM MinGW g++ 4.3.2 





and Perl 5.10

and sh version is 3.1.0



I am using CMake system to generate CodeBlocks project files.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20090106/827bbe1c/attachment.html 

From ofv at wanadoo.es  Tue Jan  6 11:55:43 2009
From: ofv at wanadoo.es (=?windows-1252?Q?=D3scar_Fuentes?=)
Date: Tue, 06 Jan 2009 18:55:43 +0100
Subject: [LLVMbugs] Cmake report errors on svn 61806
In-Reply-To: <BAY110-DS377E4A7109F744ECB4D54A9DE0@phx.gbl> ("Yonggang Luo
	=?gb2312?B?KMLe08IJuNUpIidz?= message of "Tue,
	6 Jan 2009 22:02:07 +0800")
References: <BAY110-DS377E4A7109F744ECB4D54A9DE0@phx.gbl>
Message-ID: <m3sknw1fq8.fsf@telefonica.net>



"Yonggang Luo"
<yonggangluo-PkbjNfxxIARBDgjK7y7TUQ at public.gmane.org> writes:

> CMake Error at cmake/modules/GetTargetTriple.cmake:14 (message): 
> Failed to execute D:/svn/llvm/autoconf/config.guess
>
> Call Stack (most recent call first):
>
> cmake/config-ix.cmake:59 (get_target_triple)
>
> CMakeLists.txt:105 (include)
>
> What's happened?
>
> I am using TDM MinGW g++ 4.3.2 
>
> and Perl 5.10
>
> and sh version is 3.1.0
>
> I am using CMake system to generate CodeBlocks project files.

First, please file a proper bug report with as much information as you
can. This mailing list is not for reporting bugs.

CodeBlocks is not tested.

The problem consists on GetTargetTriple.cmake using
autoconf/config.guess for all compiler/platforms except MSVC. Can your
shell execute autoconf/config.guess ? If the answer is yes, the problem
is elsewhere. If no, modify cmake/modules/GetTargetTriple.cmake for
returning a fixed answer just like it does for MSVC. Removing the
if/else altogether and leaving the 

set( ${var} "i686-pc-win32" PARENT_SCOPE )

should work. Of course, this is not acceptable as a general solution, so
you need to look on what is missing for supporting CodeBlocks.

HTH.

-- 
Oscar


From bugzilla-daemon at cs.uiuc.edu  Tue Jan  6 14:03:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Jan 2009 14:03:27 -0600
Subject: [LLVMbugs] [Bug 3286] New: Assertion failed: (0 && "Unknown stmt
 kind!") when calling Stmt->dumpAll(); 
Message-ID: <bug-3286-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3286

           Summary: Assertion failed: (0 && "Unknown stmt kind!") when
                    calling Stmt->dumpAll();
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: AST
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: alexei.svitkine at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2366)
 --> (http://llvm.org/bugs/attachment.cgi?id=2366)
code that triggers the assert

When running the attached code, an assert triggers in the StmtVisitor code.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan  6 14:46:00 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Jan 2009 14:46:00 -0600
Subject: [LLVMbugs] [Bug 3287] New: Vector of float with length 2 broken on
	x86
Message-ID: <bug-3287-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3287

           Summary: Vector of float with length 2 broken on x86
           Product: new-bugs
           Version: unspecified
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


Vectors of float on the x86 use the MM registers and not the MMX registers
which wreaks havoc.  Here is a simple test program which prints nan instead of
1.
If you comment out the (pointless) store it prints 1.

This is with llvm-2.4.

foo.ll:

define double @foo(<2 x float>*) {
        store <2 x float> < float 1.000000e+00, float 2.000000e+00 >, <2 x
float>* %0
        ret double 1.000000e+00
}


c.c:

double foo(float *);

main()
{
  float fs[2];
  printf("%g\n", foo(fs));
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan  6 14:53:55 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Jan 2009 14:53:55 -0600
Subject: [LLVMbugs] [Bug 3288] New: Multi-dimensional aggregate assert
 failure in SelectionDAGLowering
Message-ID: <bug-3288-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3288

           Summary: Multi-dimensional aggregate assert failure in
                    SelectionDAGLowering
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: markleone at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Inserting at certain offsets in a multi-dimensional aggregate tickles an
index-related assert-failure in SelectionDAGLowering:

        %i = insertvalue [2 x [2 x i32]] undef, [2 x i32] undef, 1

It doesn't matter whether the operands are undef, but the 1 index is
significant.

Assertion failed: (Begin + idx <= End), function operator[], file
llvm-trunk/include/llvm/ADT/SmallVector.h, line 127.

  llc  llvm::SmallVectorImpl<llvm::SDValue>::operator[](unsigned int) + 86
  llc  llvm::SelectionDAGLowering::visitInsertValue(llvm::InsertValueInst&) +
775
  llc  llvm::SelectionDAGLowering::visit(unsigned int, llvm::User&) + 1487
  llc  llvm::SelectionDAGLowering::visit(llvm::Instruction&) + 40
  llc  llvm::SelectionDAGISel::SelectBasicBlock(llvm::BasicBlock*,
llvm::ilist_iterator<llvm::Instruction>,
llvm::ilist_iterator<llvm::Instruction>) + 95
  llc  llvm::SelectionDAGISel::SelectAllBasicBlocks(llvm::Function&,
llvm::MachineFunction&, llvm::MachineModuleInfo*, llvm::TargetInstrInfo const&)
+ 2100
  llc  llvm::SelectionDAGISel::runOnFunction(llvm::Function&) + 852


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan  6 19:03:25 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Jan 2009 19:03:25 -0600
Subject: [LLVMbugs] [Bug 3286] Assertion failed: (0 && "Unknown stmt kind!")
 when calling Stmt- >dumpAll(); 
In-Reply-To: <bug-3286-206@http.llvm.org/bugs/>
Message-ID: <200901070103.n0713P4h026545@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3286


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-06 19:03:25 ---
ParseAST is deallocating the AST before you access it.  Please run your app
through debug malloc or valgrind.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan  6 19:51:57 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Jan 2009 19:51:57 -0600
Subject: [LLVMbugs] [Bug 3289] New: Assertion `LRegs.size() == 1 && "Can't
 handle this yet!"' failed.
Message-ID: <bug-3289-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3289

           Summary: Assertion `LRegs.size() == 1 && "Can't handle this
                    yet!"' failed.
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


Seen using r61781 on Ubuntu Hardy on x86.

Versions before and including 61492 are fine, later versions display this bug.

regehr at john-home:~/volatile/tmp110$ llvm-gcc -O2 small.c
cc1: ScheduleDAGRRList.cpp:713:
void<unnamed>::ScheduleDAGRRList::ListScheduleBottomUp(): Assertion
`LRegs.size() == 1 && "Can't handle this yet!"' failed.
small.c:18: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.

regehr at john-home:~/volatile/tmp110$ cat small.c

char safe_mul_func_uint8_t_u_u (char _ui1, char _ui2)
{
  return _ui1 * _ui2;
}

int g_54;

int func_66 (short p_68)
{
  int l_71 = 0x0ABB00F6L;
  if (func_33
      (1 ^ (safe_mod_ (safe_mul_func_uint8_t_u_u (p_68 * l_71, 1)), 1) ? 1
       : (g_54
          && 1) - safe_mod_ (safe_mul_func_uint8_t_u_u (p_68 * l_71,
                                                        func_40 (1))), 1))
    {
    }
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan  6 19:57:23 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Jan 2009 19:57:23 -0600
Subject: [LLVMbugs] [Bug 3290] New: scalarrepl doesn' t promote alloca in
 2008-03-24-BitField-And-Alloca.c
Message-ID: <bug-3290-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3290

           Summary: scalarrepl doesn't promote alloca in 2008-03-24-
                    BitField-And-Alloca.c
           Product: libraries
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


This is split off from bug 3232, where we learned that diabling gcc's inliner
caused a perf regression in 2008-03-24-BitField-And-Alloca.c because scalarrepl
isn't promoting the allocas anymore.

This in turn is because of how the struct is being passed by value to a
function: due to the ABI lowering stuff in the front-end, it is passed as an
i64, so we get this code, which confuses scalarrepl:

        %struct.Key = type { { i32, i32 } }
...
        %iospec = alloca %struct.Key            ; <%struct.Key*> [#uses=3]
...
        %2 = bitcast %struct.Key* %iospec to i64*               ; <i64*>
[#uses=1]
        store i64 %key_token2, i64* %2, align 8

Perhaps scalarrepl can handle this the same way it handles memcpy.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan  7 09:45:34 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 Jan 2009 09:45:34 -0600
Subject: [LLVMbugs] [Bug 3292] New: Assertion `i*2 < getNumOperands() &&
 "Invalid value number! "' failed.
Message-ID: <bug-3292-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3292

           Summary: Assertion `i*2 < getNumOperands() && "Invalid value
                    number!"' failed.
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


Seen using r61781 on Ubuntu Hardy on x86.

Versions before and including 61492 are fine, later versions display this bug.

regehr at john-home:~/volatile/tmp111$ llvm-gcc -O2 small.c
cc1: /home/regehr/llvm/include/llvm/Instructions.h:1928: llvm::Value*
llvm::PHINode::getIncomingValue(unsigned int) const: Assertion `i*2 <
getNumOperands() && "Invalid value number!"' failed.
small.c:36: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.

regehr at john-home:~/volatile/tmp111$ cat small.c

unsigned safe_sub_func_int64_t_s_s (int _si1, unsigned _si2)
{
  return ((_si1 ^ _si2) & (-_si2 ^ _si2)) < 0 ? : _si1 - _si2;
}

unsigned g_24;
unsigned g_30;
unsigned g_94;

unsigned func_25 (int p_27, unsigned p_29, unsigned p_31, unsigned p_33,
                  unsigned p_34)
{
  return g_24;
}

int func_40 (int p_42, int p_43)
{
  unsigned l_45 = 0;
  for (g_30 = 1; g_30; g_30 = g_30 - 1)
    {
      unsigned l_47 = 0;
      for (g_30 = 19; g_30; g_30 = safe_sub_func_int64_t_s_s (g_30, 1))
        {
          unsigned l_46 = 0;
          if (func_25 (l_45, l_46, l_46, l_47, 1))
            return 1;
        }
    }
  return g_30;
}

void func_15 (int p_16)
{
  unsigned l_20 = 0;
  func_40 (g_94, l_20);
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From yonggangluo at hotmail.com  Wed Jan  7 12:18:25 2009
From: yonggangluo at hotmail.com (=?gb2312?B?WW9uZ2dhbmcgTHVvICjC3tPCuNUp?=)
Date: Thu, 8 Jan 2009 02:18:25 +0800
Subject: [LLVMbugs] The rm command can't be running under cmd.exe
Message-ID: <BAY110-DS493D2C461AC9708FC5527A9DF0@phx.gbl>

-      COMMAND rm -f *${CMAKE_CXX_OUTPUT_EXTENSION}
+      COMMAND del /Q *${CMAKE_CXX_OUTPUT_EXTENSION}


because rm -f is can be running on -ix system. and msys and cygwin system.
but the problem is rm -f can't be running on windows system.

on windows operating system. only del /Q can be running?

and cmake have a solution to this problem . That's ${RM}


# The command to remove a file.
RM = D:\Tools\cmake\bin\cmake.exe -E remove -f


the cmake submit this variable to resolve this problem.
but i don't know how to use it.



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20090108/3a716343/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: AddPartiallyLinkedObject.cmake.diff
Type: application/octet-stream
Size: 696 bytes
Desc: not available
Url : http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20090108/3a716343/attachment.obj 

From ofv at wanadoo.es  Wed Jan  7 13:27:13 2009
From: ofv at wanadoo.es (=?windows-1252?Q?=D3scar_Fuentes?=)
Date: Wed, 07 Jan 2009 20:27:13 +0100
Subject: [LLVMbugs] The rm command can't be running under cmd.exe
References: <BAY110-DS493D2C461AC9708FC5527A9DF0@phx.gbl>
Message-ID: <m3k59629ym.fsf@telefonica.net>

"Yonggang Luo (??????)"
<yonggangluo at hotmail.com> writes:

> -      COMMAND rm -f *${CMAKE_CXX_OUTPUT_EXTENSION}
> +      COMMAND del /Q *${CMAKE_CXX_OUTPUT_EXTENSION}

On the future, please file bug reports at

http://www.llvm.org/bugs/

I've committed r61875, which  replaces rm with "cmake -R remove", but
I'm afraid that more work is needed for using the build on Windows/Mingw
with a non-posix environment.

[snip]

-- 
Oscar


From bugzilla-daemon at cs.uiuc.edu  Wed Jan  7 15:02:28 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 Jan 2009 15:02:28 -0600
Subject: [LLVMbugs] [Bug 3293] New: llvm-gcc: llvm-gcc/gcc/libgcc2.c:809:
 internal compiler error: Aborted
Message-ID: <bug-3293-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3293

           Summary: llvm-gcc: llvm-gcc/gcc/libgcc2.c:809: internal compiler
                    error: Aborted
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Keywords: compile-fail
          Severity: blocker
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: dknoto at next.com.pl
                CC: llvmbugs at cs.uiuc.edu


System:

Linux wenus 2.6.24.7_desktop-2 #1 SMP PREEMPT Sun Jun 22 18:23:12 CEST 2008
x86_64 Dual-Core_AMD_Opteron(tm)_Processor_2210 PLD Linux

GCC:

Reading specs from /usr/lib64/gcc/amd64-pld-linux/4.1.2/specs
Target: amd64-pld-linux
Configured with: ../configure --program-suffix=4 --prefix=/usr
--with-local-prefix=/usr/local --libdir=/usr/lib64 --libexecdir=/usr/lib64
--infodir=/usr/share/info --mandir=/usr/share/man
--x-libraries=/usr/X11R6/lib64 --enable-shared --enable-threads=posix
--enable-languages=c,c++ --enable-c99 --enable-long-long --enable-nls
--disable-werror --with-gnu-as --with-gnu-ld --with-demangler-in-ld
--with-system-zlib --with-slibdir=/lib64 --without-system-libunwind --without-x
--disable-multilib --with-gxx-include-dir=/usr/include/c++/4.1.2
--disable-libstdcxx-pch --enable-__cxa_atexit --enable-libstdcxx-allocator=new
amd64-pld-linux
Thread model: posix
gcc version 4.1.2 (PLD-Linux)

LLVM:     SVN r61858
LLVM-GCC: SVN r61862

Compilation:

$ ../llvm-gcc/configure --prefix=/usr/local/llvm/
--enable-languages=c,c++,fortran --enable-llvm=$PWD/../llvm --disable-bootstrap
--disable-multilib

$ make

Result:

[...]
make[4]: Leaving directory
`/home/users/dknoto/Oprogramowanie/Low.Level.Virtual.Machine/llvm-gcc-objects/gcc'
/home/users/dknoto/Oprogramowanie/Low.Level.Virtual.Machine/llvm-gcc-objects/./gcc/xgcc
-B/home/users/dknoto/Oprogramowanie/Low.Level.Virtual.Machine/llvm-gcc-objects/./gcc/
-B/usr/local/llvm//x86_64-unknown-linux-gnu/bin/
-B/usr/local/llvm//x86_64-unknown-linux-gnu/lib/ -isystem
/usr/local/llvm//x86_64-unknown-linux-gnu/include -isystem
/usr/local/llvm//x86_64-unknown-linux-gnu/sys-include  -O2  -O2 -mtune=opteron
-O2  -DIN_GCC    -W -Wall -Wwrite-strings -Wstrict-prototypes
-Wmissing-prototypes -Wold-style-definition  -isystem ./include  -fPIC -g
-DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED  -I. -I.
-I../../llvm-gcc/gcc -I../../llvm-gcc/gcc/. -I../../llvm-gcc/gcc/../include
-I../../llvm-gcc/gcc/../libcpp/include  -I../../llvm-gcc/gcc/../libdecnumber
-I../libdecnumber
-I/home/users/dknoto/Oprogramowanie/Low.Level.Virtual.Machine/llvm-gcc-objects/../llvm/include
-I/home/users/dknoto/Oprogramowanie/Low.Level.Virtual.Machine/llvm/include
-DL_popcountsi2 -fvisibility=hidden -DHIDE_EXPORTS -c
../../llvm-gcc/gcc/libgcc2.c -o libgcc/./_popcountsi2.o
cc1: StringMap.cpp:177: void
llvm::StringMapImpl::RemoveKey(llvm::StringMapEntryBase*): Assertion `V == V2
&& "Didn't find key?"' failed.
../../llvm-gcc/gcc/libgcc2.c:809: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.
make[3]: *** [libgcc/./_popcountsi2.o] Error 1
make[3]: Leaving directory
`/home/users/dknoto/Oprogramowanie/Low.Level.Virtual.Machine/llvm-gcc-objects/gcc'
make[2]: *** [stmp-multilib] Error 2
make[2]: Leaving directory
`/home/users/dknoto/Oprogramowanie/Low.Level.Virtual.Machine/llvm-gcc-objects/gcc'
make[1]: *** [all-gcc] Error 2
make[1]: Leaving directory
`/home/users/dknoto/Oprogramowanie/Low.Level.Virtual.Machine/llvm-gcc-objects'
make: *** [all] Error 2


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan  7 22:27:05 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 Jan 2009 22:27:05 -0600
Subject: [LLVMbugs] [Bug 3294] New: another math bug
Message-ID: <bug-3294-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3294

           Summary: another math bug
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


This looks a lot like PR 3275 but it is still wrong in r61912 on Ubuntu Hardy
on x86.

LLVM up to and including r59527 gets this right, since then it is wrong.

regehr at john-home:~/volatile/tmp112$ llvm-gcc -O0 small.c -o small
regehr at john-home:~/volatile/tmp112$ ./small
-1
regehr at john-home:~/volatile/tmp112$ llvm-gcc -O1 small.c -o small
regehr at john-home:~/volatile/tmp112$ ./small
10
regehr at john-home:~/volatile/tmp112$ cat small.c
#include <stdio.h>

unsigned g_9;

void func_1 (void);
void func_1 (void)
{
  for (g_9 = 11; g_9 < (unsigned)-8; g_9--)
    {
    }
}

int
main (void)
{
  func_1 ();
  printf ("%d\n", g_9);
  return 0;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan  7 23:43:34 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 Jan 2009 23:43:34 -0600
Subject: [LLVMbugs] [Bug 3295] New: loop-reduce / scalar-evolution crash
 expanding {0, +, 2, +, 2} addrec
Message-ID: <bug-3295-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3295

           Summary: loop-reduce / scalar-evolution crash expanding
                    {0,+,2,+,2} addrec
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicholas at mxc.ca
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2369)
 --> (http://llvm.org/bugs/attachment.cgi?id=2369)
testcase

opt -loop-reduce b.bc crashes trying to emit an addrec of {0,+,2,+,2} because
the SCEV for %indvar returns CouldNotCompute.

And that's correct. This "loop" has a i1 false for a branch condition, meaning
that scev won't analyze it. I'm not sure who's at fault here.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan  8 00:16:09 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 Jan 2009 00:16:09 -0600
Subject: [LLVMbugs] [Bug 3296] New: another math bug
Message-ID: <bug-3296-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3296

           Summary: another math bug
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


This is possibly the same bug as PR 3275 and/or PR 3294.  However, this one
first manifests in r61826.  This in on Ubuntu Hardy on x86.

Much of this bulky test input is irrelevant to the problem which involves only
the induction variable.  However, removing code from the body of the loop makes
the problem go away.

llvm-gcc is not valgrind-clean at -O2 for this testcase.

regehr at john-home:~/volatile/tmp113$ llvm-gcc -O1 small.c -o small
regehr at john-home:~/volatile/tmp113$ ./small
0
regehr at john-home:~/volatile/tmp113$ llvm-gcc -O2 small.c -o small
regehr at john-home:~/volatile/tmp113$ ./small
-20
regehr at john-home:~/volatile/tmp113$ cat small.c
#include <stdio.h>
#include <stdint.h>
#include <limits.h>


#define safe_add_macro_int64_t_s_s(si1,si2) \
                ((((((int64_t)(si1))>((int64_t)0)) &&
(((int64_t)(si2))>((int64_t)0)) && (((int64_t)(si1)) >
((INT64_MAX)-((int64_t)(si2))))) \
                  || ((((int64_t)(si1))<((int64_t)0)) &&
(((int64_t)(si2))<((int64_t)0)) && (((int64_t)(si1)) <
((INT64_MIN)-((int64_t)(si2)))))) \
                 ? ((int64_t)(si1)) \
                 : (((int64_t)(si1)) + ((int64_t)(si2))) \
                 ) 

static int64_t
safe_add_func_int64_t_s_s(int64_t _si1, int64_t _si2)
{
  return safe_add_macro_int64_t_s_s(_si1,_si2);
}

#define safe_lshift_macro_uint16_t_u_s(left,right) \
        (((((int)(right)) < ((uint16_t)0)) \
                         || (((int)(right)) >= sizeof(uint16_t)*CHAR_BIT) \
                         || (((uint16_t)(left)) > ((UINT16_MAX) >>
((int)(right))))) \
                        ? ((uint16_t)(left)) \
                        : (((uint16_t)(left)) << ((int)(right))))

static uint16_t
safe_lshift_func_uint16_t_u_s(uint16_t _left, int _right)
{
  return safe_lshift_macro_uint16_t_u_s(_left,_right);
}

#define safe_sub_macro_int8_t_s_s(si1,si2) \
                ((((((int8_t)(si1))^((int8_t)(si2))) \
                & (((((int8_t)(si1)) ^ ((((int8_t)(si1))^((int8_t)(si2))) \
                & (((int8_t)1) <<
(sizeof(int8_t)*CHAR_BIT-1))))-((int8_t)(si2)))^((int8_t)(si2)))) <
((int8_t)0)) \
                ? ((int8_t)(si1)) \
                : (((int8_t)(si1)) - ((int8_t)(si2))) \
                )

#define safe_mod_macro_int16_t_s_s(si1,si2) \
  (((((int16_t)(si2)) == ((int16_t)0)) || ((((int16_t)(si1)) == (INT16_MIN)) &&
(((int16_t)(si2)) == ((int16_t)-1)))) \
  ? ((int16_t)(si1)) \
  : (((int16_t)(si1)) % ((int16_t)(si2))))

#define safe_mul_macro_uint8_t_u_u(ui1,ui2) \
  (((uint8_t)(ui1)) * ((uint8_t)(ui2)))

static uint8_t
safe_mul_func_uint8_t_u_u (uint8_t _ui1, uint8_t _ui2)
{
  return safe_mul_macro_uint8_t_u_u(_ui1,_ui2);
}

uint32_t g_14;

void func_13 (uint32_t p_15);
void func_13 (uint32_t p_15)
{
  for (g_14 = -21; g_14 >= 12; g_14 = safe_add_func_int64_t_s_s (g_14, 1))
    {
      uint32_t l_17 = 1;
      uint32_t l_20;
      (safe_lshift_func_uint16_t_u_s
       ((safe_sub_macro_int8_t_s_s (p_15, 1)), 1))
        && (safe_mod_macro_int16_t_s_s (l_17, p_15)) ||
        safe_mul_func_uint8_t_u_u (l_20, 1);
    }
}

int main (void)
{
  func_13 (1);
  printf ("%d\n", g_14);
  return 0;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan  8 09:34:29 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 Jan 2009 09:34:29 -0600
Subject: [LLVMbugs] [Bug 3298] New: Assertion `New != this &&
 "this->replaceAllUsesWith(this) is NOT valid!"'
Message-ID: <bug-3298-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3298

           Summary: Assertion `New != this && "this-
                    >replaceAllUsesWith(this) is NOT valid!"'
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


This works up to and including r61492 on Ubuntu Hardy on x86, and breaks in
more recent versions.

regehr at john-home:~/volatile/tmp115$ llvm-gcc -O2 small.c
cc1: Value.cpp:317: void llvm::Value::replaceAllUsesWith(llvm::Value*):
Assertion `New != this && "this->replaceAllUsesWith(this) is NOT valid!"'
failed.
small.c:29: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.

regehr at john-home:~/volatile/tmp115$ cat small.c

char
safe_div_func_char_s_s (char _si1, char _si2)
{
  return 1 && _si2 ? : _si2;
}

int
safe_sub_func_int64_t_s_s (int _si1, int _si2)
{
  return 1 << _si2;
}

int
func (int p_79, int p_80)
{
  char l_82 = 1;
  for (p_80 = 0; p_80;)
    {
      for (p_79 = 1; p_79 <= 0; p_79 = safe_sub_func_int64_t_s_s (p_79, 0)) {
        l_82 = 0;
      }
      if (p_80) {
        for (; p_80; p_80 = 1) {
          safe_div_func_char_s_s (p_79, l_82);
        }
      }
    }
  return safe (l_82);
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan  8 11:17:03 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 Jan 2009 11:17:03 -0600
Subject: [LLVMbugs] [Bug 3299] New: LLVM optimizer can' t eliminate loop
 with floating point counter
Message-ID: <bug-3299-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3299

           Summary: LLVM optimizer can't eliminate loop with floating point
                    counter
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: syoyofujita at gmail.com
                CC: llvmbugs at cs.uiuc.edu


LLVM optimizer fails to eliminate following dead loop with floating point
counter.
(The exception is that i is incremented by 1.0f)

/* floop.c */
void
func() {
   float i;
   for (i = 0.0f; i < 1000.0f; i += 1.2f) {
   }
}

$ clang -emit-llvm-bc floop.c
$ opt -std-compile-opts floop.bc | llvm-dis

opt -std-compile-opts test.bc | llvm-dis
; ModuleID = '<stdin>'
target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128"
target triple = "i386-apple-darwin9"

define void @func(...) nounwind readnone {
entry:
        br label %for.inc

for.inc:                ; preds = %for.inc, %entry
        %i.0.reg2mem.0 = phi float [ 0.000000e+00, %entry ], [ %add, %for.inc
]; <float> [#uses=1]
        %add = add float %i.0.reg2mem.0, 0x3FF3333340000000             ;
<float> [#uses=2]
        %cmp = fcmp olt float %add, 1.000000e+03                ; <i1>
[#uses=1]
        br i1 %cmp, label %for.inc, label %return

return:         ; preds = %for.inc
        ret void
}

(loop wasn't eliminated after optimization)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan  8 11:50:07 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 Jan 2009 11:50:07 -0600
Subject: [LLVMbugs] [Bug 3301] New: x(x){} confuses clang.
Message-ID: <bug-3301-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3301

           Summary: x(x){} confuses clang.
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: AST
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bolzoni at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bagnara at cs.unipr.it


Small test case:
$ cat x.c 
x(x){ return x; }

$ clang x.c
f.c:1:1: error: redefinition of 'x' as different kind of symbol
x(x){ return x; }
^
f.c:1:3: note: previous definition is here
x(x){ return x; }
  ^
2 diagnostics generated.
$


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan  8 13:05:49 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 Jan 2009 13:05:49 -0600
Subject: [LLVMbugs] [Bug 3281] diagnostics in LLParser
In-Reply-To: <bug-3281-206@http.llvm.org/bugs/>
Message-ID: <200901081905.n08J5nAS030671@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3281


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #10 from Chris Lattner <clattner at apple.com>  2009-01-08 13:05:48 ---
Thanks Dan, fixed!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan  8 23:47:17 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 Jan 2009 23:47:17 -0600
Subject: [LLVMbugs] [Bug 3290] scalarrepl doesn' t promote alloca in
 2008-03-24-BitField-And-Alloca.c
In-Reply-To: <bug-3290-206@http.llvm.org/bugs/>
Message-ID: <200901090547.n095lHgA027515@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3290


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #8 from Chris Lattner <clattner at apple.com>  2009-01-08 23:47:17 ---
Part 3 is implemented here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090105/071949.html

This test is now passing, reenabled here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090105/071950.html

-Chris


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan  8 23:58:58 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 Jan 2009 23:58:58 -0600
Subject: [LLVMbugs] [Bug 3270] Strange function declaration brings confusion
	to clang.
In-Reply-To: <bug-3270-206@http.llvm.org/bugs/>
Message-ID: <200901090558.n095ww8L027911@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3270


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |WORKSFORME




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-08 23:58:57 ---
Seems fine to me, this compiles as expected:

struct u { int a; };
struct u f();
struct u *s, g();

void foo() { g(); }


The AST dump also looks right:


(CompoundStmt 0x2004a50 <bc.c:5:12, col:19>
  (CallExpr 0x2004a30 <col:14, col:16> 'struct u'
    (ImplicitCastExpr 0x2004a10 <col:14> 'struct u (*)()'
      (DeclRefExpr 0x20049d0 <col:14> 'struct u ()' FunctionDecl='g'
0x20048c0))))


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 00:08:53 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 00:08:53 -0600
Subject: [LLVMbugs] [Bug 3298] Assertion `New != this &&
 "this->replaceAllUsesWith(this) is NOT valid!"'
In-Reply-To: <bug-3298-206@http.llvm.org/bugs/>
Message-ID: <200901090608.n0968rCm028281@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3298


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Chris Lattner <clattner at apple.com>  2009-01-09 00:08:53 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090105/071953.html

Thanks John and DanspaceG!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 00:12:47 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 00:12:47 -0600
Subject: [LLVMbugs] [Bug 3302] New: yet another math bug??
Message-ID: <bug-3302-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3302

           Summary: yet another math bug??
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


This one first appears in r61493 on Ubuntu Hardy on x86.

Unlike 3294 and 3296, it does not involve the induction variable.

Also, unlike the others, this testcase is super simple, so should be an easy
one!

llvm-gcc is valgrind-clean at -O2 for this testcase.

regehr at john-home:~/volatile/tmp118$ llvm-gcc -O1 small.c -o small
regehr at john-home:~/volatile/tmp118$ ./small
11
regehr at john-home:~/volatile/tmp118$ llvm-gcc -O2 small.c -o small
regehr at john-home:~/volatile/tmp118$ ./small
1
regehr at john-home:~/volatile/tmp118$ cat small.c
#include <stdio.h>

static unsigned
safe_sub_func_unsigned_u_u (unsigned _ui1, unsigned _ui2)
{
  return _ui1 - _ui2;
}

unsigned g_57;

void func_1 (void);
void func_1 (void)
{
  unsigned l_65;
  for (l_65 = 10; l_65 < 19; l_65 = safe_sub_func_unsigned_u_u (l_65, 1))
    {
      g_57++;
    }
}

int
main (void)
{
  func_1 ();
  printf ("%d\n", g_57);
  return 0;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 02:02:22 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 02:02:22 -0600
Subject: [LLVMbugs] [Bug 2293] lli/llc crash with "Cannot promote to smaller
	type"
In-Reply-To: <bug-2293-206@http.llvm.org/bugs/>
Message-ID: <200901090802.n0982Mod031843@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2293


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #7 from Dan Gohman <gohman at apple.com>  2009-01-09 02:02:22 ---
LegalizeTypes is now enabled by default in svn. The given testcase
now compiles cleanly.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 02:16:59 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 02:16:59 -0600
Subject: [LLVMbugs] [Bug 2659] redundant movl generated on x86
In-Reply-To: <bug-2659-206@http.llvm.org/bugs/>
Message-ID: <200901090816.n098Gxm4032435@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2659


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Dan Gohman <gohman at apple.com>  2009-01-09 02:16:59 ---
This is fixed in r61847.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 03:08:43 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 03:08:43 -0600
Subject: [LLVMbugs] [Bug 3303] New: CMake building config files for Mingw32
Message-ID: <bug-3303-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3303

           Summary: CMake building config files for Mingw32
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Keywords: build-problem, portability
          Severity: major
          Priority: P2
         Component: llvm-config
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: yonggangluo at hotmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2372)
 --> (http://llvm.org/bugs/attachment.cgi?id=2372)
for llvm svn version

Because the config files are not write proper for Mingw32, so I modified these
files. And now it can work on Windows platform proper for mingw32-make, and can
be execute under cmd.exe, and don't need run a bash . But it's also need msys
now.
because it's can execute by cmd.exe, so it's can work with CodeBlocks -win32.

But I don't test it under ix-like system. So please test it before commit it.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 06:06:34 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 06:06:34 -0600
Subject: [LLVMbugs] [Bug 3304] New: Crash in RewriteStoreUserOfWholeAlloca
Message-ID: <bug-3304-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3304

           Summary: Crash in RewriteStoreUserOfWholeAlloca
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


$ opt -scalarrepl x.bc -disable-output
opt: lib/VMCore/Type.cpp:1023: static const llvm::IntegerType*
llvm::IntegerType::get(unsigned int): Assertion `NumBits >= MIN_INT_BITS &&
"bitwidth too small"' failed.
...
The bitwidth is zero!  Doubtless due to the empty struct type.  Testcase:

        %struct.c37304a__vrec = type { i8, %struct.c37304a__vrec___disc___XVN }
        %struct.c37304a__vrec___disc___XVN = type {
%struct.c37304a__vrec___disc___XVN___O }
        %struct.c37304a__vrec___disc___XVN___O = type {  }

define void @_ada_c37304a() {
entry:
        %v = alloca %struct.c37304a__vrec               ;
<%struct.c37304a__vrec*> [#uses=1]
        %0 = getelementptr %struct.c37304a__vrec* %v, i32 0, i32 0             
; <i8*> [#uses=1]
        store i8 8, i8* %0, align 1
        unreachable
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 09:15:43 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 09:15:43 -0600
Subject: [LLVMbugs] [Bug 3293] llvm-gcc: llvm-gcc/gcc/libgcc2.c:809:
 internal compiler error: Aborted
In-Reply-To: <bug-3293-206@http.llvm.org/bugs/>
Message-ID: <200901091515.n09FFhcE026293@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3293


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




--- Comment #3 from Duncan Sands <baldrick at free.fr>  2009-01-09 09:15:42 ---
> I wrote the gcc version in bug report, 4.1.2 (+ PLD patches).

So you did, d'oh!  The gcc 4.1.2 compiler is known to be
broken on many platforms, see
  http://llvm.org/docs/GettingStarted.html#brokengcc
The typical symptom is that people get the error you are
getting, so I'm pretty sure this is the cause.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 12:19:18 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 12:19:18 -0600
Subject: [LLVMbugs] [Bug 3304] Crash in RewriteStoreUserOfWholeAlloca
In-Reply-To: <bug-3304-206@http.llvm.org/bugs/>
Message-ID: <200901091819.n09IJI95032752@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3304


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-09 12:19:18 ---
Sorry, I'm being an idiot again.  Fixed here, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090105/071965.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 12:35:36 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 12:35:36 -0600
Subject: [LLVMbugs] [Bug 3305] New: clang emits wrong type warning:
 incompatible pointer types passing 'struct logfile *',
 expected '  struct logfile *'
Message-ID: <bug-3305-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3305

           Summary: clang emits wrong type warning: incompatible pointer
                    types passing 'struct logfile *', expected 'struct
                    logfile *'
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: FreeBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: rdivacky at freebsd.org
                CC: llvmbugs at cs.uiuc.edu


attached (gcc -E processed) test case emits this:

....
screen.c:6101:17: warning: incompatible pointer types passing 'struct logfile
*', expected 'struct logfile *'
      logfwrite(p->w_log, buf, strlen(buf));
                ^~~~~~~~

there are many cases of this strange warning, clang as of Jan 7th does not
exhibit this behaviour


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 14:44:45 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 14:44:45 -0600
Subject: [LLVMbugs] [Bug 3306] New: gcc compat: empty declarations are
 errors not warnings
Message-ID: <bug-3306-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3306

           Summary: gcc compat: empty declarations are errors not warnings
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: trivial
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


clang rejects this code which gcc accepts with a warning:
--
  __attribute__((visibility("default")));
--


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan  9 14:54:55 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Jan 2009 14:54:55 -0600
Subject: [LLVMbugs] [Bug 3307] New: gcc compat: clang rejects pointer/ int
 mismatch in conditional expression
Message-ID: <bug-3307-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3307

           Summary: gcc compat: clang rejects pointer/int mismatch in
                    conditional expression
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


clang rejects the following code which gcc accepts with a warning:-
--
ddunbar at lordcrumb:hard$ cat conditional-compat.c 
void f0(int a0, void *a1, int a2) {
  a0 ? a1 : a2;
}
ddunbar at lordcrumb:hard$ clang -fsyntax-only conditional-compat.c 
conditional-compat.c:2:6: error: incompatible operand types ('void *' and
'int')
  a0 ? a1 : a2;
     ^ ~~   ~~
1 diagnostic generated.
ddunbar at lordcrumb:hard$ gcc -fsyntax-only conditional-compat.c 
conditional-compat.c: In function 'f0':
conditional-compat.c:2: warning: pointer/integer type mismatch in conditional
expression
--


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 10 14:12:26 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 Jan 2009 14:12:26 -0600
Subject: [LLVMbugs] [Bug 3311] New: llvm-gcc bootstrap broken on x86-64 linux
Message-ID: <bug-3311-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3311

           Summary: llvm-gcc bootstrap broken on x86-64 linux
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


Fails during stage2, when running the compiler built with the compiler
built during stage1:

xgcc ... -g0 -finhibit-size-directive -fno-inline-functions -fno-exceptions
-fno-zero-initialized-in-bss -fno-toplevel-reorder  -fno-omit-frame-pointer
-fno-asynchronous-unwind-tables \
          -c ../../gcc-4.2.llvm/gcc/crtstuff.c -DCRT_BEGIN \
          -o crtbegin.o
In file included from /usr/include/string.h:423,
                 from ../../gcc-4.2.llvm/gcc/tsystem.h:103,
                 from ../../gcc-4.2.llvm/gcc/crtstuff.c:68:
/usr/include/bits/string2.h: In function ???__strpbrk_c2???:
/usr/include/bits/string2.h:1129: internal compiler error: Segmentation fault


Steps to reproduce: bootstrap llvm-gcc.  A non-bootstrap build is not enough.
I see this on two different x86-64 linux boxes.

Bisection shows it was broken by this commit:

------------------------------------------------------------------------
r61847 | evancheng | 2009-01-07 03:08:57 +0100 (Wed, 07 Jan 2009) | 10 lines

The coalescer does not coalesce a virtual register to a physical register if
any of the physical register's sub-register live intervals overlaps with the
virtual register. This is overly conservative. It prevents a extract_subreg
from being coalesced away:

v1024 = EDI  // not killed
      =
      = EDI

One possible solution is for the coalescer to examine the sub-register live
intervals in the same manner as the physical register. Another possibility is
to examine defs and uses (when needed) of sub-registers. Both solutions are too
expensive. For now, look for "shortvirtual intervals" and scan instructions to
look for conflict instead.

This is a small win on x86-64. e.g. It shaves 403.gcc by ~80 instructions.

------------------------------------------------------------------------


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 10 16:39:56 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 Jan 2009 16:39:56 -0600
Subject: [LLVMbugs] =?utf-8?q?=5BBug_3312=5D_New=3A__Recent_breakage_relat?=
 =?utf-8?b?ZWQgdG8g4oCYZW51bXPigJkgaW4gcjYyMDE2?=
Message-ID: <bug-3312-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3312

           Summary: Recent breakage related to ???enums??? in r62016
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: danchr at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2375)
 --> (http://llvm.org/bugs/attachment.cgi?id=2375)
getopt.i (preprocessed, darwin/ppc)

I was recently able to compile GNU Emacs using Clang, but now I get compiler
errors. Reverting r62016 fixed one of the files, the pre-processed output of
which is attached below.

This may be related to an ???enum??? being declared within a ???struct???.

2009-01-09  Douglas Gregor  <doug.gregor at gmail.com>

        * lib/Sema/SemaDecl.cpp:
        Don't bother setting NextDeclarator for EnumConstantDecls. It isn't
        used


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 11 09:13:25 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 Jan 2009 09:13:25 -0600
Subject: [LLVMbugs] [Bug 3313] New: LangRef never defines syntax for type
	symbol table
Message-ID: <bug-3313-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3313

           Summary: LangRef never defines syntax for type symbol table
           Product: Documentation
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: General docs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicholas at mxc.ca
                CC: llvmbugs at cs.uiuc.edu


This sort of code:

  %struct.foo = type { i32, i32 }

is never defined by the LangRef, though it's used in the example for
getelementptr. That should be in there somewhere. It should also make it clear
whether it applies to all Types or just StructType, and what the behaviour is
on multiple definitions of the same type (currently we silently collapse them
and pick one name).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 11 13:21:42 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 Jan 2009 13:21:42 -0600
Subject: [LLVMbugs] [Bug 3314] New: clang treats local array with constant
 var as size as VLA
Message-ID: <bug-3314-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3314

           Summary: clang treats local array with constant var as size as
                    VLA
           Product: clang
           Version: unspecified
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: AST
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: pelle at morth.org
                CC: llvmbugs at cs.uiuc.edu


Example:
void tests_const_var_array(void)
{
        const int sz = 10;
        int arr[sz];

        arr[0] = 1;
}

$ ccc -c bing.c 
bing.c
clang -emit-llvm-bc -x c -o bing.o bing.c
Assertion failed: (0 && "FIXME: Local VLAs not implemented yet"), function
EmitLocalBlockVarDecl, file
/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_ports_lang_llvm-devel/work/llvm-53722/tools/clang/lib/CodeGen/CGDecl.cpp,
line 162.

Expected result:
clang should realize sz is constant and treat this as a normal array, not a
VLA.


If nothing else, this should be considered when local VLAs are implemented.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 11 13:53:12 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 Jan 2009 13:53:12 -0600
Subject: [LLVMbugs] [Bug 3314] clang treats local array with constant var as
	size as VLA
In-Reply-To: <bug-3314-206@http.llvm.org/bugs/>
Message-ID: <200901111953.n0BJrCu5013263@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3314


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-11 13:53:11 ---
Clang is doing the right thing there.  Per the C definition, that is a VLA.  I
would much rather implement VLAs that add hacks to work around them being
missing.  Thanks for filing the bug though!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 11 14:26:13 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 Jan 2009 14:26:13 -0600
Subject: [LLVMbugs] [Bug 3315] New: Error on the attached file
Message-ID: <bug-3315-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3315

           Summary: Error on the attached file
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Target Description Classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: andersca at mac.com
                CC: llvmbugs at cs.uiuc.edu


The attached file causes the following error in llc

Cannot yet select: 0x180ecf4: ch = declare 0x180f024, 0x180ef14, 0x180ee04
0   llc               0x008704ae char const* std::find<char const*, char>(char
const*, char const*, char const&) + 98
1   llc               0x008709e3 llvm::sys::PrintStackTraceOnErrorSignal() +
593
2   libSystem.B.dylib 0x9655b2bb _sigtramp + 43
3   libSystem.B.dylib 0xffffffff _sigtramp + 1772768623
4   libSystem.B.dylib 0x965cf23a raise + 26
5   libSystem.B.dylib 0x965db679 abort + 73
6   llc               0x0032ba2d (anonymous
namespace)::X86ISelAddressMode::dump() + 1185
7   llc               0x00340291 bool
llvm::isa_impl_cl<llvm::SDNode*>::isa<llvm::AtomicSDNode>(llvm::SDNode*) +
13813
8   llc               0x0045710b llvm::cast_retty<llvm::JumpTableSDNode,
llvm::SDValue>::ret_type llvm::dyn_cast<llvm::JumpTableSDNode,
llvm::SDValue>(llvm::SDValue const&) + 1024401
9   llc               0x002baec7
llvm::createX86ISelDag(llvm::X86TargetMachine&, bool) + 16145
10  llc               0x0045b363 llvm::cast_retty<llvm::JumpTableSDNode,
llvm::SDValue>::ret_type llvm::dyn_cast<llvm::JumpTableSDNode,
llvm::SDValue>(llvm::SDValue const&) + 1041385
11  llc               0x002bd923
llvm::createX86ISelDag(llvm::X86TargetMachine&, bool) + 26989
12  llc               0x005811d8 llvm::SelectionDAGISel::CodeGenAndEmitDAG() +
2930
13  llc               0x00582e4f
llvm::SelectionDAGISel::SelectBasicBlock(llvm::BasicBlock*,
llvm::ilist_iterator<llvm::Instruction>,
llvm::ilist_iterator<llvm::Instruction>) + 575
14  llc               0x00583698
llvm::SelectionDAGISel::SelectAllBasicBlocks(llvm::Function&,
llvm::MachineFunction&, llvm::MachineModuleInfo*, llvm::TargetInstrInfo const&)
+ 2100
15  llc               0x00583a55
llvm::SelectionDAGISel::runOnFunction(llvm::Function&) + 873
16  llc               0x007f51bc
llvm::FPPassManager::runOnFunction(llvm::Function&) + 270
17  llc               0x007f5412
llvm::FunctionPassManagerImpl::run(llvm::Function&) + 116
18  llc               0x007f5598
llvm::FunctionPassManager::run(llvm::Function&) + 156
19  llc               0x000034c5 main + 2615
20  llc               0x00002236 start + 54


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 11 14:54:16 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 Jan 2009 14:54:16 -0600
Subject: [LLVMbugs] [Bug 3313] LangRef never defines syntax for type symbol
	table
In-Reply-To: <bug-3313-206@http.llvm.org/bugs/>
Message-ID: <200901112054.n0BKsG8f015375@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3313


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-11 14:54:15 ---
Here ya go:
http://llvm.org/docs/LangRef.html#namedtypes


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 11 21:21:43 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 Jan 2009 21:21:43 -0600
Subject: [LLVMbugs] [Bug 3244] ice: Don't know how to copy this physical
	register!
In-Reply-To: <bug-3244-206@http.llvm.org/bugs/>
Message-ID: <200901120321.n0C3Lh1t027925@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3244


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Evan Cheng <evan.cheng at apple.com>  2009-01-11 21:21:42 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090105/072022.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 09:02:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 09:02:20 -0600
Subject: [LLVMbugs] [Bug 3288] Multi-dimensional aggregate assert failure in
 SelectionDAGLowering
In-Reply-To: <bug-3288-206@http.llvm.org/bugs/>
Message-ID: <200901121502.n0CF2Kat029054@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3288


Mark Leone <markleone at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Mark Leone <markleone at gmail.com>  2009-01-12 09:02:16 ---
Fixed by Dan Gohman in r61828.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 11:53:32 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 11:53:32 -0600
Subject: [LLVMbugs] [Bug 3316] New: Constant folding for 'frem' is flaky
Message-ID: <bug-3316-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3316

           Summary: Constant folding for 'frem' is flaky
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: markleone at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Constant folding for 'frem' seems flaky for certain values.  For example, "frem
float 0.1, 1.0" returns -0.9 when constant folded:

define float @test1() {
entry:
    %y = frem float 0x3FB99999A0000000, 1.0
    ret float %y
}

@x = global float 0x3FB99999A0000000    ; float 0.1

define float @test2() {
    %x = load float* @x
    %y = frem float %x, 1.0
    ret float %y
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 12:07:06 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 12:07:06 -0600
Subject: [LLVMbugs] [Bug 3317] New: APint error during x86 lowering
Message-ID: <bug-3317-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3317

           Summary: APint error during x86 lowering
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicolas.geoffray at lip6.fr
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2381)
 --> (http://llvm.org/bugs/attachment.cgi?id=2381)
Bug test case

With the attached test case, I try to compile with llc -march=x86 (other
platforms work) but get the error:

llc: APInt.cpp:942: llvm::APInt& llvm::APInt::zext(uint32_t): Assertion `width
> BitWidth && "Invalid APInt ZeroExtend request"' failed.
0   llc       0x08a241ee
1             0xffffe420  + 0
2             0xffffe410  + 16
3   libc.so.6 0xb7cf4d90 gsignal + 80
4   libc.so.6 0xb7cf67f8 abort + 392
5   libc.so.6 0xb7cedcce __assert_fail + 238
6   llc       0x089e66c2 llvm::APInt::zext(unsigned int) + 242
7   llc       0xffffffff llvm::APInt::zext(unsigned int) + 4150368815


The complete backtrace is:

#0  0xffffe410 in __kernel_vsyscall ()
#1  0xb7d06d90 in raise () from /lib/i686/libc.so.6
#2  0xb7d087f8 in abort () from /lib/i686/libc.so.6
#3  0xb7cffcce in __assert_fail () from /lib/i686/libc.so.6
#4  0x089e66c2 in llvm::APInt::zext ()
#5  0x0867cb1b in llvm::TargetLowering::SimplifyDemandedBits ()
#6  0x0867f3b4 in llvm::TargetLowering::SimplifyDemandedBits ()
#7  0x0859924f in (anonymous namespace)::DAGCombiner::SimplifyDemandedBits ()
#8  0x085ad7b3 in (anonymous namespace)::DAGCombiner::visitXOR ()
#9  0x085c3dac in (anonymous namespace)::DAGCombiner::visit ()
#10 0x085c4844 in (anonymous namespace)::DAGCombiner::combine ()
#11 0x085c6096 in (anonymous namespace)::DAGCombiner::Run ()
#12 0x085c65d0 in llvm::SelectionDAG::Combine ()
#13 0x08663f72 in llvm::SelectionDAGISel::CodeGenAndEmitDAG ()
#14 0x08665c68 in llvm::SelectionDAGISel::SelectBasicBlock ()
#15 0x08667be3 in llvm::SelectionDAGISel::SelectAllBasicBlocks ()
#16 0x08668f80 in llvm::SelectionDAGISel::runOnFunction ()
#17 0x089a5f2e in llvm::FPPassManager::runOnFunction ()
#18 0x089a6b4c in llvm::FunctionPassManagerImpl::run ()
#19 0x089a6d7f in llvm::FunctionPassManager::run ()
#20 0x08118cb4 in main ()


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 12:51:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 12:51:20 -0600
Subject: [LLVMbugs] [Bug 3305] clang emits wrong type warning: incompatible
 pointer types passing 'struct logfile *', expected '  struct logfile *'
In-Reply-To: <bug-3305-206@http.llvm.org/bugs/>
Message-ID: <200901121851.n0CIpKPj006053@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3305


Douglas Gregor <dgregor at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Douglas Gregor <dgregor at apple.com>  2009-01-12 12:51:19 ---
Fixed in r62091


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 12:51:33 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 12:51:33 -0600
Subject: [LLVMbugs] =?utf-8?q?=5BBug_3312=5D__Recent_breakage_related_to_?=
 =?utf-8?b?4oCYZW51bXPigJk=?=
In-Reply-To: <bug-3312-206@http.llvm.org/bugs/>
Message-ID: <200901121851.n0CIpXEq006087@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3312


Douglas Gregor <dgregor at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #6 from Douglas Gregor <dgregor at apple.com>  2009-01-12 12:51:33 ---
Fixed in r62091


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 13:47:09 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 13:47:09 -0600
Subject: [LLVMbugs] [Bug 3318] New: Aggregate return error in
	CCState::AnalyzeCallResult
Message-ID: <bug-3318-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3318

           Summary: Aggregate return error in CCState::AnalyzeCallResult
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: markleone at gmail.com
                CC: llvmbugs at cs.uiuc.edu


I'm getting an error ("Call result #2 has unhandled type i32") for a function
that returns an aggregate in CCState::AnalyzeCallResult.  For example,

declare [3 x i32] @test1()

define void @test2() {
        %x = call [3 x i32] @test1()
        ret void
}

The error is also triggered for aggregates like {i32, i32, i32}, but it works
fine for [2 x i32] or {i32, i32}.  Here's the stack trace:

 llvm::CCState::AnalyzeCallResult(llvm::CallSDNode*, bool (*)(unsigned int,
llvm::MVT, llvm::MVT, llvm::CCValAssign::LocInfo, llvm::ISD::ArgFlagsTy,
llvm::CCState&)) + 309
 llvm::X86TargetLowering::LowerCallResult(llvm::SDValue, llvm::SDValue,
llvm::CallSDNode*, unsigned int, llvm::SelectionDAG&) + 138
 llvm::X86TargetLowering::LowerCALL(llvm::SDValue, llvm::SelectionDAG&) + 7702
 llvm::X86TargetLowering::LowerOperation(llvm::SDValue, llvm::SelectionDAG&) +
1761
 llvm::cast_retty<llvm::SubprogramDesc, llvm::DebugInfoDesc*>::ret_type
llvm::cast<llvm::SubprogramDesc, llvm::DebugInfoDesc*>(llvm::DebugInfoDesc*
const&) + 45271
 llvm::cast_retty<llvm::SubprogramDesc, llvm::DebugInfoDesc*>::ret_type
llvm::cast<llvm::SubprogramDesc, llvm::DebugInfoDesc*>(llvm::DebugInfoDesc*
const&) + 223622
 llvm::cast_retty<llvm::SubprogramDesc, llvm::DebugInfoDesc*>::ret_type
llvm::cast<llvm::SubprogramDesc, llvm::DebugInfoDesc*>(llvm::DebugInfoDesc*
const&) + 224700
 llvm::SelectionDAG::Legalize(bool) + 64
 llvm::SelectionDAGISel::CodeGenAndEmitDAG() + 2080
 llvm::SelectionDAGISel::SelectBasicBlock(llvm::BasicBlock*,
llvm::ilist_iterator<llvm::Instruction>,
llvm::ilist_iterator<llvm::Instruction>) + 575
 llvm::SelectionDAGISel::SelectAllBasicBlocks(llvm::Function&,
llvm::MachineFunction&, llvm::MachineModuleInfo*, llvm::TargetInstrInfo const&)
+ 2100


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 15:13:40 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 15:13:40 -0600
Subject: [LLVMbugs] [Bug 3319] New: remove useless warning in printf checker
Message-ID: <bug-3319-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3319

           Summary: remove useless warning in printf checker
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: FreeBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: rdivacky at freebsd.org
                CC: llvmbugs at cs.uiuc.edu


clang emits a warning for this:

string-literal.c:5:14: warning: format string is not a string literal
(potentially insecure)
      printf(i == 1 ? "yes" : "no");
             ^~~~~~~~~~~~~~~~~~~~~

the attached patch implements checking of ConditionalOperators, test case is
also attached


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 16:20:46 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 16:20:46 -0600
Subject: [LLVMbugs] [Bug 3318] Aggregate return error in
	CCState::AnalyzeCallResult
In-Reply-To: <bug-3318-206@http.llvm.org/bugs/>
Message-ID: <200901122220.n0CMKkmK014911@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3318


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |gohman at apple.com
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #1 from Dan Gohman <gohman at apple.com>  2009-01-12 16:20:46 ---
This problem is being tracked with PR2660.

*** This bug has been marked as a duplicate of bug 2660 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 17:10:31 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 17:10:31 -0600
Subject: [LLVMbugs] [Bug 3319] remove useless warning in printf checker
In-Reply-To: <bug-3319-206@http.llvm.org/bugs/>
Message-ID: <200901122310.n0CNAV6N017066@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3319


Ted Kremenek <kremenek at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #6 from Ted Kremenek <kremenek at apple.com>  2009-01-12 17:10:30 ---
(In reply to comment #5)
> Created an attachment (id=2384)
 --> (http://llvm.org/bugs/attachment.cgi?id=2384) [details]
> new version of the patch
> 

Applied:

http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090112/010779.html

Thanks so much!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 17:43:10 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 17:43:10 -0600
Subject: [LLVMbugs] [Bug 1750] clang rejects ia32_signal.i (anonymous unions)
In-Reply-To: <bug-1750-206@http.llvm.org/bugs/>
Message-ID: <200901122343.n0CNhAYU018361@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1750


Douglas Gregor <dgregor at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #15 from Douglas Gregor <dgregor at apple.com>  2009-01-12 17:43:09 ---
Closing this bug; we've dealt with all of the subproblems.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 21:59:55 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 21:59:55 -0600
Subject: [LLVMbugs] [Bug 3311] llvm-gcc bootstrap broken on x86-64 linux
In-Reply-To: <bug-3311-206@http.llvm.org/bugs/>
Message-ID: <200901130359.n0D3xtaI027014@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3311


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #14 from Evan Cheng <evan.cheng at apple.com>  2009-01-12 21:59:55 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072061.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 12 23:19:41 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Jan 2009 23:19:41 -0600
Subject: [LLVMbugs] [Bug 3320] New: volatile array bug in llvm-g++
Message-ID: <bug-3320-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3320

           Summary: volatile array bug in llvm-g++
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


We'd expect the compiler to generate code reading each element of a[] twice,
but that is not what happens on Ubuntu Hardy on x86 using r62045.  Oddly,
llvm-gcc gets this case right.

regehr at john-home:~$ cat vol.c
#define N 3

typedef volatile int vint;

int sum(vint a[N]) {
   int sum = 0;
   unsigned i;
   for (i = 0; i < N; i++)
     sum += a[i] + a[i];
   return sum;
}

regehr at john-home:~$ llvm-g++ -O3 vol.c -S -o - -fomit-frame-pointer

        .ident  "GCC: (GNU) 4.2.1 (Based on Apple Inc. build 5628) (LLVM
build)"
        .file   "vol.c"


        .text
        .align  16
        .globl  _Z3sumPVi
        .type   _Z3sumPVi, at function
_Z3sumPVi:
        movl    4(%esp), %eax
        movl    4(%eax), %ecx
        movl    (%eax), %edx
        addl    %edx, %edx
        leal    (%edx,%ecx,2), %ecx
        movl    8(%eax), %eax
        leal    (%ecx,%eax,2), %eax
        ret
        .size   _Z3sumPVi, .-_Z3sumPVi

        .section        .note.GNU-stack,"", at progbits
regehr at john-home:~$


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 13 06:55:23 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Jan 2009 06:55:23 -0600
Subject: [LLVMbugs] [Bug 3321] New: incorrect optimisation of linked list
	code
Message-ID: <bug-3321-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3321

           Summary: incorrect optimisation of linked list code
           Product: libraries
           Version: 2.4
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Interprocedural Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: jay.foad at antixlabs.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2385)
 --> (http://llvm.org/bugs/attachment.cgi?id=2385)
source file

The attached program should return 42. But if I build it with llvm-gcc on
Linux/x86, I get non-deterministic results:

$ ~/llvm/llvm-gcc4.2-2.4-x86-linux-RHEL4/bin/llvm-gcc -O2 -o f f.c
$ ./f ; echo $?
212
$ ./f ; echo $?
164
$ ./f ; echo $?
212
$ ./f ; echo $?
180

It looks like the optimizers have replaced all my code with "ret i32 undef".


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 13 10:12:13 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Jan 2009 10:12:13 -0600
Subject: [LLVMbugs] [Bug 3251] ice: Invalid node ID for RAUW analysis!
In-Reply-To: <bug-3251-206@http.llvm.org/bugs/>
Message-ID: <200901131612.n0DGCDGc031174@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3251


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Duncan Sands <baldrick at free.fr>  2009-01-13 10:12:12 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072095.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 13 14:01:18 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Jan 2009 14:01:18 -0600
Subject: [LLVMbugs] [Bug 3310] declaration of vars with undefined structured
 type is rejected
In-Reply-To: <bug-3310-206@http.llvm.org/bugs/>
Message-ID: <200901132001.n0DK1Ix4006650@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3310


Nuno Lopes <nunoplopes at sapo.pt> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |llvmbugs at cs.uiuc.edu,
                   |                            |nunoplopes at sapo.pt
         AssignedTo|kremenek at apple.com          |unassignedclangbugs at nondot.o
                   |                            |rg
          Component|Static Analyzer             |Semantic Analyzer
            Summary|Analyzer failures when      |declaration of vars with
                   |scanning coreboot firmware  |undefined structured type is
                   |                            |rejected




--- Comment #1 from Nuno Lopes <nunoplopes at sapo.pt>  2009-01-13 14:01:17 ---
the parser rejects are all reduced to this simple test case:

struct foo bar;

'gcc -fsyntax-only' accepts this with *no warning* (weird..), while clang
rejects it.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 13 14:04:58 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Jan 2009 14:04:58 -0600
Subject: [LLVMbugs] [Bug 1954] incorrect type+warning for "unnamed" union
In-Reply-To: <bug-1954-206@http.llvm.org/bugs/>
Message-ID: <200901132004.n0DK4wng006810@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1954


Nuno Lopes <nunoplopes at sapo.pt> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #8 from Nuno Lopes <nunoplopes at sapo.pt>  2009-01-13 14:04:58 ---
I believe this is now fixed.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 13 14:19:14 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Jan 2009 14:19:14 -0600
Subject: [LLVMbugs] [Bug 2248] Clang cannot compile its preprocessed output
 with same flags
In-Reply-To: <bug-2248-206@http.llvm.org/bugs/>
Message-ID: <200901132019.n0DKJELU007472@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2248


Nuno Lopes <nunoplopes at sapo.pt> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |nunoplopes at sapo.pt
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Nuno Lopes <nunoplopes at sapo.pt>  2009-01-13 14:19:13 ---
this bug is not reproduceable with bundled headers.
This was probably related with gcc's PR35460 (i.e. broken headers in some
systems).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 13 18:13:23 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Jan 2009 18:13:23 -0600
Subject: [LLVMbugs] [Bug 3321] incorrect optimisation of linked list code
In-Reply-To: <bug-3321-206@http.llvm.org/bugs/>
Message-ID: <200901140013.n0E0DNnn016914@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3321


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Chris Lattner <clattner at apple.com>  2009-01-13 18:13:23 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072133.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 14 00:37:44 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 Jan 2009 00:37:44 -0600
Subject: [LLVMbugs] [Bug 3322] New: __dyld_misaligned_stack_error on Mac OS
 X 10. 5 x86 when using system call
Message-ID: <bug-3322-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3322

           Summary: __dyld_misaligned_stack_error on Mac OS X 10.5 x86 when
                    using system call
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Interpreter
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: alexei.svitkine at gmail.com
                CC: llvmbugs at cs.uiuc.edu


When using the Interpeter to execute the following code compiled with clang:

#include <stdio.h
int main(void)
{
  perror(0);
  return 0;
}

I get a crash as a result of a __dyld_misaligned_stack_error(). However, the
same code runs fine using when ForceInterpreter is set to false when creating
the ExecutionEngine.

A Google search suggests that the stack must be 16-byte aligned for system
calls to work correctly on Mac OS X x86:

http://stackoverflow.com/questions/5649/x86-assembly-on-a-mac

The back trace is as follows:

#0  0x8fe18c02 in __dyld_misaligned_stack_error ()
#1  0x00204e4f in llvm::PATypeHolder::get (this=0xbffff130) at Type.h:406
#2  0x009f5de1 in llvm::APInt::initSlowCase (this=0xbffff130, that=@0xbffff0cc)
at APInt.cpp:54
#3  0x0028fc6c in llvm::APInt::APInt (this=0xbffff130, that=@0xbffff0cc) at
ADT/APInt.h:250
#4  0x002efe02 in llvm::GenericValue::GenericValue (this=0xbffff128) at
ExecutionEngine/GenericValue.h:26
#5  0x002ee8c7 in llvm::Interpreter::callExternalFunction (this=0x1c054f0,
F=0x1c0a4b0, ArgVals=@0xbffff214) at ExternalFunctions.cpp:119
#6  0x002e93c7 in llvm::Interpreter::callFunction (this=0x1c054f0, F=0x1c0a4b0,
ArgVals=@0xbffff214) at Execution.cpp:1309
#7  0x002e9ada in llvm::Interpreter::visitCallSite (this=0x1c054f0,
CS=@0xbffff2bc) at Execution.cpp:922
#8  0x002f60d6 in llvm::Interpreter::visitCallInst (this=0x1c054f0,
I=@0x1c0a518) at Interpreter.h:164
#9  0x002f60f0 in llvm::InstVisitor<llvm::Interpreter, void>::visitCall
(this=0x1c054f0, I=@0x1c0a518) at Instruction.def:159
#10 0x002f6b2a in llvm::InstVisitor<llvm::Interpreter, void>::visit
(this=0x1c054f0, I=@0x1c0a518) at Instruction.def:159
#11 0x002e9e17 in llvm::Interpreter::run (this=0x1c054f0) at Execution.cpp:1345
#12 0x002eeda4 in llvm::Interpreter::runFunction (this=0x1c054f0, F=0x1c09290,
ArgValues=@0xbffff3ec) at Interpreter.cpp:100
#13 0x0031639b in llvm::ExecutionEngine::runFunctionAsMain (this=0x1c054f0,
Fn=0x1c09290, argv=@0xbffff614, envp=0x0) at ExecutionEngine.cpp:357


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 14 04:54:49 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 Jan 2009 04:54:49 -0600
Subject: [LLVMbugs] [Bug 3323] New: AliasAnalysis pass removed before
 function pass can use it
Message-ID: <bug-3323-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3323

           Summary: AliasAnalysis pass removed before function pass can use
                    it
           Product: libraries
           Version: 2.3
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: hvdieren at elis.ugent.be
                CC: hvdieren at elis.ugent.be, llvmbugs at cs.uiuc.edu


The following command line suggests that the LICM pass can use the Andersens
Alias Analysis (for instance) to do its alias analysis:

$ opt -debug-pass=Structure -disable-opt -basicaa -anders-aa -licm in.bc -f -o
out.bc

But as the output shows, the AndersensAA, which is a ModulePass, is destroyed
before the FunctionPasses are run!

Pass Arguments:  -anders-aa -domtree -loops -loopsimplify -domfrontier
-scalar-evolution -licm -preverify -domtree -verify
Target Data Layout
Basic Alias Analysis (default AA impl)
  ModulePass Manager
    Andersen's Interprocedural Alias Analysis
--    Andersen's Interprocedural Alias Analysis
    FunctionPass Manager
      Dominator Tree Construction
      Natural Loop Construction
      Canonicalize natural loops
      Dominance Frontier Construction
      Scalar Evolution Analysis
      Loop Pass Manager
        Loop Invariant Code Motion
--        Loop Invariant Code Motion
--      Canonicalize natural loops
--      Basic Alias Analysis (default AA impl)
--      Natural Loop Construction
--      Dominator Tree Construction
--      Dominance Frontier Construction
--      Scalar Evolution Analysis
      Preliminary module verification
      Dominator Tree Construction
      Module Verifier
--      Preliminary module verification
--      Dominator Tree Construction
--      Module Verifier
    Bitcode Writer
--    Bitcode Writer

The problem is thus that FunctionPasses (and other finer-grain passes) can
never benefit from detailed AliasAnalysis passes, which are invariably
ModulePasses.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 14 05:02:11 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 Jan 2009 05:02:11 -0600
Subject: [LLVMbugs] [Bug 3324] New: [XCore] assertion rematerialising LDWSP
	instruction
Message-ID: <bug-3324-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3324

           Summary: [XCore] assertion rematerialising LDWSP instruction
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: richard at xmos.com
                CC: llvmbugs at cs.uiuc.edu


In the target description LDWSP implicitly uses the SP register. This causes
LiveIntervals to assert when rematerialising stack slot loads with the
following code:

define double @f1(double %a, double %b, double %c, double %d, double %e, double
%f, double %g) nounwind {
entry:
        br i1 false, label %bb113, label %bb129

bb113:          ; preds = %entry
        ret double 0.000000e+00

bb129:          ; preds = %entry
        %tmp134 = sub double %b, %a             ; <double> [#uses=1]
        %tmp136 = sub double %tmp134, %c                ; <double> [#uses=1]
        %tmp138 = add double %tmp136, %d                ; <double> [#uses=1]
        %tmp140 = sub double %tmp138, %e                ; <double> [#uses=1]
        %tmp142 = add double %tmp140, %f                ; <double> [#uses=1]
        %tmp.0 = mul double %tmp142, 0.000000e+00               ; <double>
[#uses=1]
        ret double %tmp.0
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 14 05:09:16 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 Jan 2009 05:09:16 -0600
Subject: [LLVMbugs] [Bug 3325] New: ipsccp vs invoke
Message-ID: <bug-3325-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3325

           Summary: ipsccp vs invoke
           Product: libraries
           Version: 2.4
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: jay.foad at antixlabs.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2387)
 --> (http://llvm.org/bugs/attachment.cgi?id=2387)
test case

The attached test program should return 42:

$ llvm-as -f -o invoke.bc invoke.ll
$ lli invoke.bc ; echo $?
42

But ipsccp breaks it:

$ opt -f -o opt.bc invoke.bc -ipsccp
$ lli opt.bc ; echo $?
16


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From xaero_1 at hotmail.com  Wed Jan 14 02:31:53 2009
From: xaero_1 at hotmail.com (Lee Wanghao)
Date: Wed, 14 Jan 2009 16:31:53 +0800
Subject: [LLVMbugs] Ubuntu LLVM compiled a page of Simit-ARM with errors
Message-ID: <BAY116-W429B225B8FDDAA1102DC7AC0D60@phx.gbl>


Hi Everyone,



I am new to LLVM and I am not sure if its capable of being a compilation

strategy to accelerate Dynamic Compiled Instruction set Simulation.



I have tried changing the source code of: http://simit-arm.sourceforge.net/ to

compile pages of C++ code (translated from Arm Assembly) on the fly using

llvm-g++ and it gave the following errors:



In file included from

/home/wanghao/Desktop/SimIt-ARM-3.0/build/include/bittypes.h:22,

                 from

/home/wanghao/Desktop/SimIt-ARM-3.0/build/include/misc.h:20,

                 from

/home/wanghao/Desktop/SimIt-ARM-3.0/build/include/armemul.h:20,

                 from

/home/wanghao/Desktop/SimIt-ARM-3.0/build/include/arch.hpp:120,

                 from

/home/wanghao/Desktop/SimIt-ARM-3.0/build/include/arch_jit.hpp:20,

                 from /home/wanghao/.ema/u512//Xcompiled_0.cpp:1:

/usr/include/inttypes.h:385: error: `__extern_inline' does not name a type

/usr/include/inttypes.h:401: error: `__extern_inline' does not name a type

/usr/include/inttypes.h:415: error: `__extern_inline' does not name a type

/usr/include/inttypes.h:432: error: `__extern_inline' does not name a type

In file included from

/home/wanghao/Desktop/SimIt-ARM-3.0/build/include/dyn_arm_emul.hpp:18,

                 from /home/wanghao/.ema/u512//Xcompiled_0.cpp:2:

/usr/include/pthread.h:1110: error: expected constructor, destructor, or type

conversion

/usr/include/pthread.h:1110: error: expected `,' or `;'

Total user time  : 3.188 sec.

Total system time: 0.472 sec.

Simulation user time  : 2.892 sec.

Simulation system time: 0.400 sec.

Caching user time  : 0.296 sec.

Caching system time: 0.072 sec.

Simulation speed : 2.285e+07 inst/sec.

Total instructions : 83634854 (83M)

Requested blocks  : 1.

Cache-hit count : 0.

Connected servers/threads : 2 of 2.

Compilation count :

  thread 0: 0

  thread 1: 0



Please also advise on whether LLVM-GCC is able to do any aid to Instruction Set

Simulation. Thanks in advance.



Regards

Wang Hao Lee
_________________________________________________________________
More than messages?check out the rest of the Windows Live?.
http://www.microsoft.com/windows/windowslive/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20090114/93a325b2/attachment.html 

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 14 12:34:01 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 Jan 2009 12:34:01 -0600
Subject: [LLVMbugs] [Bug 3324] [XCore] assertion rematerialising LDWSP
	instruction
In-Reply-To: <bug-3324-206@http.llvm.org/bugs/>
Message-ID: <200901141834.n0EIY1gh032252@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3324


Richard Osborne <richard at xmos.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Richard Osborne <richard at xmos.com>  2009-01-14 12:34:01 ---
Fixed in r62238

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072150.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 14 15:01:50 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 Jan 2009 15:01:50 -0600
Subject: [LLVMbugs] [Bug 3325] ipsccp vs invoke
In-Reply-To: <bug-3325-206@http.llvm.org/bugs/>
Message-ID: <200901142101.n0EL1o1a004876@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3325


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Chris Lattner <clattner at apple.com>  2009-01-14 15:01:49 ---
Nice catch, your patch looks great!  I committed it with a testcase here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072158.html

Thanks!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From echeng at apple.com  Wed Jan 14 12:20:45 2009
From: echeng at apple.com (Evan Cheng)
Date: Wed, 14 Jan 2009 10:20:45 -0800
Subject: [LLVMbugs] Ubuntu LLVM compiled a page of Simit-ARM with errors
In-Reply-To: <BAY116-W429B225B8FDDAA1102DC7AC0D60@phx.gbl>
References: <BAY116-W429B225B8FDDAA1102DC7AC0D60@phx.gbl>
Message-ID: <1F4C6884-0634-4117-871A-CC9DA856A8BE@apple.com>

Place post these types of questions to llvmdev.

That said, this looks like a frontend issue.

Evan

On Jan 14, 2009, at 12:31 AM, Lee Wanghao wrote:

> Hi Everyone,
>
> I am new to LLVM and I am not sure if its capable of being a  
> compilation
> strategy to accelerate Dynamic Compiled Instruction set Simulation.
>
> I have tried changing the source code of: http://simit-arm.sourceforge.net/ 
>  to
> compile pages of C++ code (translated from Arm Assembly) on the fly  
> using
> llvm-g++ and it gave the following errors:
>
> In file included from
> /home/wanghao/Desktop/SimIt-ARM-3.0/build/include/bittypes.h:22,
>                  from
> /home/wanghao/Desktop/SimIt-ARM-3.0/build/include/misc.h:20,
>                  from
> /home/wanghao/Desktop/SimIt-ARM-3.0/build/include/armemul.h:20,
>                  from
> /home/wanghao/Desktop/SimIt-ARM-3.0/build/include/arch.hpp:120,
>                  from
> /home/wanghao/Desktop/SimIt-ARM-3.0/build/include/arch_jit.hpp:20,
>                  from /home/wanghao/.ema/u512//Xcompiled_0.cpp:1:
> /usr/include/inttypes.h:385: error: `__extern_inline' does not name  
> a type
> /usr/include/inttypes.h:401: error: `__extern_inline' does not name  
> a type
> /usr/include/inttypes.h:415: error: `__extern_inline' does not name  
> a type
> /usr/include/inttypes.h:432: error: `__extern_inline' does not name  
> a type
> In file included from
> /home/wanghao/Desktop/SimIt-ARM-3.0/build/include/dyn_arm_emul.hpp:18,
>                  from /home/wanghao/.ema/u512//Xcompiled_0.cpp:2:
> /usr/include/pthread.h:1110: error: expected constructor,  
> destructor, or type
> conversion
> /usr/include/pthread.h:1110: error: expected `,' or `;'
> Total user time  : 3.188 sec.
> Total system time: 0.472 sec.
> Simulation user time  : 2.892 sec.
> Simulation system time: 0.400 sec.
> Caching user time  : 0.296 sec.
> Caching system time: 0.072 sec.
> Simulation speed : 2.285e+07 inst/sec.
> Total instructions : 83634854 (83M)
> Requested blocks  : 1.
> Cache-hit count : 0.
> Connected servers/threads : 2 of 2.
> Compilation count :
>   thread 0: 0
>   thread 1: 0
>
> Please also advise on whether LLVM-GCC is able to do any aid to  
> Instruction Set
> Simulation. Thanks in advance.
>
> Regards
> Wang Hao Lee
> check out the rest of the Windows Live?. More than mail?Windows  
> Live? goes way beyond your inbox.More than messages  
> _______________________________________________
> LLVMbugs mailing list
> LLVMbugs at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20090114/67c4bc8c/attachment.html 

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 14 23:15:24 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 Jan 2009 23:15:24 -0600
Subject: [LLVMbugs] [Bug 2746] handling casts to union transparently
In-Reply-To: <bug-2746-206@http.llvm.org/bugs/>
Message-ID: <200901150515.n0F5FOsJ021022@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2746


Seo Sanghyeon <sanxiyn at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sanxiyn at gmail.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Seo Sanghyeon <sanxiyn at gmail.com>  2009-01-14 23:15:22 ---
Fixed in r62255.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 03:25:30 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 03:25:30 -0600
Subject: [LLVMbugs] [Bug 3326] New: missed optimization with -m32: empty
 functions shouldn' t contain push/mov/pop
Message-ID: <bug-3326-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3326

           Summary: missed optimization with -m32: empty functions shouldn't
                    contain push/mov/pop
           Product: new-bugs
           Version: unspecified
          Platform: PC
               URL: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38854
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


LLVM exhibits same missed optimization as gcc with -m32.

Testcase:
void noop(char* foo) {}

$ llvm-gcc -O2 x.c -S -m32
noop:
        pushl   %ebp
        movl    %esp, %ebp
        popl    %ebp
        ret

It shouldn't contain the push/mov/pop, and indeed without -m32 it doesn't:
noop:
.Leh_func_begin1:
.Llabel1:
        ret

However I can't reproduce this with
llvm-gcc -S -emit-llvm -m32 -O2 | llvm-as | llc -march=x86

I tried -mcpu=i486 too.

Is gcc still using LLVM for code generation with -m32?

What is special in llvm-gcc's invocation of the code generator?

Here is the bitcode, and llc correctly generates the code w/o push/mov/pop

target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
target triple = "i386-unknown-linux-gnu"

define void @noop(i8* nocapture %foo) nounwind readnone {
entry:
        ret void
}

Both llvm and llvm-gcc were compiled from same revision (r62238).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 03:50:01 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 03:50:01 -0600
Subject: [LLVMbugs] [Bug 3327] New: ia64 target problems with ELF sections
Message-ID: <bug-3327-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3327

           Summary: ia64 target problems with ELF sections
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: IA64
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: romixlev at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Hi,

Compilation of the the following very simple file crashes with
--march=ia64 on my x86 machine.

; ModuleID = 'bugpoint-reduced-simplified.bc'
target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
target triple = "i386-pc-linux-gnu"
@x = weak global [8 x i32] zeroinitializer, align 32            ; <[8
x i32]*> [#uses=0]
@.str = internal constant [4 x i8] c"%c \00"            ; <[4 x i8]*> [#uses=0]
@rows = weak global [8 x i32] zeroinitializer, align 32         ; <[8
x i32]*> [#uses=0]
@up = weak global [15 x i32] zeroinitializer, align 32          ; <[15
x i32]*> [#uses=0]
@down = weak global [15 x i32] zeroinitializer, align 32
 ; <[15 x i32]*> [#uses=0]

declare i32 @print() nounwind

declare i32 @printf(i8*, ...) nounwind

declare i32 @putchar(i32)

declare i32 @queens(i32) nounwind

define i32 @main() nounwind {
entry:
       unreachable
}

I use this command to compile:
llc -f bugpoint-reduced-simplified.bc  --regalloc=linearscan --march=ia64

The assertion is:

llc: /opt/llvm/lib/Target/ELFTargetAsmInfo.cpp:133: const
llvm::Section* llvm::ELFTargetAsmInfo::MergeableStringSection(const
llvm::GlobalVariable*) const: Assertion `getCStringSection() &&
"Should have string section prefix"' failed.

Please see more details and test BC file at:
http://lists.cs.uiuc.edu/pipermail/llvmdev/2009-January/019382.html

-Roman


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 03:58:19 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 03:58:19 -0600
Subject: [LLVMbugs] [Bug 3326] missed optimization with -m32: empty
 functions shouldn' t contain push/mov/pop
In-Reply-To: <bug-3326-206@http.llvm.org/bugs/>
Message-ID: <200901150958.n0F9wJeO007941@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3326


Anton Korobeynikov <asl at math.spbu.ru> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |asl at math.spbu.ru
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




--- Comment #1 from Anton Korobeynikov <asl at math.spbu.ru>  2009-01-15 03:58:18 ---
This is proper behaviour. LLC defaults to FP elimination, gcc - not.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 04:47:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 04:47:20 -0600
Subject: [LLVMbugs] [Bug 3328] New: LLVM optmization loop optimization
Message-ID: <bug-3328-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3328

           Summary: LLVM optmization loop optimization
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: manoel at fonetica.com.br
                CC: llvmbugs at cs.uiuc.edu


#include <stdio.h>
#include <stdlib.h>
//
int  TESTE ( int  parami ,int  paraml   ) 
{
  int varx=0;
  int nI =0;

  if( parami  > 0 )
  {
    varx = parami;
  }
  else
  {
    varx = 1;
   }

  for( nI = 1 ;   nI <=  paraml; nI++) //here
  {
    varx =  varx +  parami+ 1;
  }

return varx ;
}

int main(int argc, char **argv)
{
   if( argc < 3 )
     return 0;
   return TESTE(atoi(argv[1]),atoi(argv[2]));
}

LLVM should eliminate the loop and do a direct calculation instead.
The MSVC is great in this example and gcc-4.3 also eliminates the loop entirely
and directly calculates the value of varx.
It looks like LLVM's scalar evolution code isn't handling your testcase well.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 07:00:25 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 07:00:25 -0600
Subject: [LLVMbugs] [Bug 3329] New: g++.dg/warn/pr23075.C fails
Message-ID: <bug-3329-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3329

           Summary: g++.dg/warn/pr23075.C fails
           Product: tools
           Version: 2.4
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: minor
          Priority: P2
         Component: llvm-g++
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: jay.foad at antixlabs.com
                CC: llvmbugs at cs.uiuc.edu


The GCC 4.2.1 testsuite test g++.dg/warn/pr23075.C (attached) passed in llvm
2.3 but fails in llvm 2.4, because it produces a bogus "no return statement"
warning:

$ llvm-g++ -c -O2 -Wreturn-type pr23075.C 
pr23075.C: In function ???int foo()???:
pr23075.C:8: error: return-statement with no value, in function returning
???int???
pr23075.C:9: warning: no return statement in function returning non-void


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 08:30:26 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 08:30:26 -0600
Subject: [LLVMbugs] [Bug 3330] New: ImplicitCastExpr with NULL pointer to
	subexpression.
Message-ID: <bug-3330-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3330

           Summary: ImplicitCastExpr with NULL pointer to subexpression.
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: AST
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bolzoni at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bagnara at cs.unipr.it


Test case:
--->
enum E {
 A,
 B = A + (((-2UL << 20) - (-10UL << 20)) >> 12) - 1,
 C
};
---<

to see the bug you need to modify the function
void DeclContextPrinter::PrintDeclContext(const DeclContext*, unsigned)
in the file Driver/ASTConsumer.cpp.

seek for `case Decl::EnumConstant:', add
--->
const Expr* e = ECD-> getInitExpr();
if (e) {
  e-> dumpAll();
}
---<
just before the break statement.
Recompile.

The output of clang -print-decl-contexts will be:
--->
(ImplicitCastExpr 0x1e84af0 'int'
  (BinaryOperator 0x1e849f0 'unsigned long' '-'
    (BinaryOperator 0x1e84930 'unsigned long' '+'
      (ImplicitCastExpr 0x1e848f0 'unsigned long'
        (DeclRefExpr 0x1e844f0 'int' EnumConstant='A' 0x1e84490))
      (ParenExpr 0x1e848b0 'unsigned long'
        (BinaryOperator 0x1e84870 'unsigned long' '>>'
          (ParenExpr 0x1e847f0 'unsigned long'
            (BinaryOperator 0x1e847b0 'unsigned long' '-'
              (ParenExpr 0x1e84630 'unsigned long'
                (BinaryOperator 0x1e845f0 'unsigned long' '<<'
                  (UnaryOperator 0x1e84570 'unsigned long' prefix '-'
                    (IntegerLiteral 0x1e84530 'unsigned long' 2))
                  (IntegerLiteral 0x1e845b0 'int' 20)))
              (ParenExpr 0x1e84770 'unsigned long'
                (BinaryOperator 0x1e84730 'unsigned long' '<<'
                  (UnaryOperator 0x1e846b0 'unsigned long' prefix '-'
                    (IntegerLiteral 0x1e84670 'unsigned long' 10))
                  (IntegerLiteral 0x1e846f0 'int' 20)))))
          (IntegerLiteral 0x1e84830 'int' 12))))
    (ImplicitCastExpr 0x1e849b0 'unsigned long'
      (IntegerLiteral 0x1e84970 'int' 1))))
(ImplicitCastExpr 0x1e84b30 'int'
  <<<NULL>>>)
[translation unit] 0x1e6dfa0
    [struct] __va_list_tag
        <field> [omitted]
    <typedef> __builtin_va_list
    [enum] E
        <enum constant> A
        <enum constant> B
        <enum constant> C
---<

The last ImplicitCastExpr has NULL as sub-expression.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 10:26:09 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 10:26:09 -0600
Subject: [LLVMbugs] [Bug 3332] New: Maintenance of CallGraph after
	self-inlining broken
Message-ID: <bug-3332-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3332

           Summary: Maintenance of CallGraph after self-inlining broken
           Product: libraries
           Version: trunk
          Platform: All
               URL: http://paste.lisp.org/display/73611
        OS/Version: All
            Status: NEW
          Severity: major
          Priority: P2
         Component: Transformation Utilities
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ggreif at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Obviously my private changes have unearthed a latent bug in InlineFunction.


Here is the relevant gdb session:

Breakpoint 5, UpdateCallGraphAfterInlining (CS=@0xbffff33c,
FirstNewBlock={<bidirectional_iterator<llvm::BasicBlock,ptrdiff_t>> =
{<iterator<std::bidirectional_iterator_tag,llvm::BasicBlock,ptrdiff_t,llvm::BasicBlock*,llvm::BasicBlock&>>
= {<No data fields>}, <No data fields>}, NodePtr = 0x6306850},
ValueMap=@0xbffff2d8, CG=@0x6303130) at InlineFunction.cpp:151
151       const Function *Caller =
CS.getInstruction()->getParent()->getParent();
(gdb) n
152       const Function *Callee = CS.getCalledFunction();
(gdb)
153       CallGraphNode *CalleeNode = CG[Callee];
(gdb)
154       CallGraphNode *CallerNode = CG[Caller];
(gdb)
158       for (CallGraphNode::iterator I = CalleeNode->begin(),
(gdb)
159            E = CalleeNode->end(); I != E; ++I) {
(gdb) call CalleeNode->dump()
Call graph node for function: 'fib'
  Calls function 'fib'
  Calls function 'fib'

(gdb) n
158       for (CallGraphNode::iterator I = CalleeNode->begin(),
(gdb) p CalleeNode->begin()
$99 = {
  _M_current = 0x6302d90
}
(gdb) p CalleeNode->end()
$100 = {
  _M_current = 0x6302da0
}
(gdb) n
160         const Instruction *OrigCall = I->first.getInstruction();
(gdb)
162         DenseMap<const Value*, Value*>::iterator VMI =
ValueMap.find(OrigCall);
(gdb)
164         if (VMI != ValueMap.end() && VMI->second) {
(gdb)
167           if (Instruction *NewCall = dyn_cast<Instruction>(VMI->second))
(gdb)
168             CallerNode->addCalledFunction(CallSite::get(NewCall),
I->second);
(gdb)
158       for (CallGraphNode::iterator I = CalleeNode->begin(),
(gdb) p CalleeNode->begin()
$101 = {
  _M_current = 0x6306f50
}
(gdb) p CalleeNode->end()
$102 = {
  _M_current = 0x6306f68
}
(gdb) p I
$103 = {
  _M_current = 0x6302d90
}
(gdb) p E
$104 = {
  _M_current = 0x6302da0
}
(gdb) 

Preliminary analysis is that the callsites in the CallGraphNode are iterated
over, but in the iteration the very vector of callsites is appended to, which
causes invalidation of the iterators.

This triggers for me on an existing testcase, and I'll paste it here later
along with a patch for a fix.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 12:42:04 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 12:42:04 -0600
Subject: [LLVMbugs] [Bug 3332] Maintenance of CallGraph after self-inlining
	broken
In-Reply-To: <bug-3332-206@http.llvm.org/bugs/>
Message-ID: <200901151842.n0FIg4GS027827@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3332


Gabor Greif <ggreif at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Gabor Greif <ggreif at gmail.com>  2009-01-15 12:42:03 ---
fixed by r62271.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 13:20:38 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 13:20:38 -0600
Subject: [LLVMbugs] [Bug 3330] ImplicitCastExpr with NULL pointer to
	subexpression.
In-Reply-To: <bug-3330-206@http.llvm.org/bugs/>
Message-ID: <200901151920.n0FJKcmv029608@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3330


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-15 13:20:37 ---
Fixed, thanks!
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090112/010859.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 13:43:34 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 13:43:34 -0600
Subject: [LLVMbugs] [Bug 3333] New: No warning when shifting invalid amount
Message-ID: <bug-3333-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3333

           Summary: No warning when shifting invalid amount
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: alexei.svitkine at gmail.com
                CC: llvmbugs at cs.uiuc.edu


When using operators ">>" or "<<", the valid range for the shift amount is [0,
# bits in type).

GCC gives appropriate warnings when the shift amount is an out-of-range
constant, such as:

"warning: right shift count is negative"

and

"warning: left shift count >= width of type"

However, clang does not present any warnings on the same code. An example
input:

#include <stdio.h>

int main(void)
{
        int i = 1;
        printf("%d\n", i>>-1);
        return 0;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 13:48:45 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 13:48:45 -0600
Subject: [LLVMbugs] [Bug 3334] New: Interpreter asserts on out of range
	shift amount
Message-ID: <bug-3334-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3334

           Summary: Interpreter asserts on out of range shift amount
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Interpreter
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: alexei.svitkine at gmail.com
                CC: llvmbugs at cs.uiuc.edu


When using the Interpreter variety of ExecutionEngine, an out of range shift
amount causes an assert to trigger. However, the same inputs to the regular
ExecutionEngine runs fine (producing an undefined result, but that's OK).

For example, the issue is visible when the following code is compiled with
clang and then executed through llvm:

int main(void) { int i=1; i<<100; return 0; }

When using the Interpreter, the following assert is triggered:

Assertion failed: (shiftAmt <= BitWidth && "Invalid shift amount"), function
shl, file /llvm/include/llvm/ADT/APInt.h, line 681.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 14:29:59 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 14:29:59 -0600
Subject: [LLVMbugs] [Bug 3335] New: instcombine ignores addrspace in
 InstCombineStoreToCast
Message-ID: <bug-3335-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3335

           Summary: instcombine ignores addrspace in InstCombineStoreToCast
           Product: new-bugs
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Keywords: miscompilation
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: luked at cs.rochester.edu
                CC: llvmbugs at cs.uiuc.edu, luked at cs.rochester.edu


Created an attachment (id=2391)
 --> (http://llvm.org/bugs/attachment.cgi?id=2391)
Simple IR file that demonstrates problem

instcombine drops addrspace in load/stores that have a bitcast address
parameter.

Run testcase with opt -instcombine -debug to see this behavior.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 15:23:06 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 15:23:06 -0600
Subject: [LLVMbugs] [Bug 3338] New: crash during compilation with clang
 0.139 (second crash)
Message-ID: <bug-3338-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3338

           Summary: crash during compilation with clang 0.139 (second crash)
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: jordan.breeding at mac.com
                CC: llvmbugs at cs.uiuc.edu


Here is the text:

ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ParserCreate
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ParserCreateNS
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ParserCreate_MM
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c parserCreate
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c parserInit
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c moveToFreeBindingList
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ParserReset
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetEncoding
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_ExternalEntityParserCreate
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c destroyBindings
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ParserFree
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_UseParserAsHandlerArg
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_UseForeignDTD
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetReturnNSTriplet
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetUserData
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetBase
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetBase
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_GetSpecifiedAttributeCount
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetIdAttributeIndex
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetElementHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetStartElementHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetEndElementHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetCharacterDataHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetProcessingInstructionHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetCommentHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetCdataSectionHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetStartCdataSectionHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetEndCdataSectionHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetDefaultHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetDefaultHandlerExpand
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetDoctypeDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetStartDoctypeDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetEndDoctypeDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetUnparsedEntityDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetNotationDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetNamespaceDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetStartNamespaceDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetEndNamespaceDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetNotStandaloneHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetExternalEntityRefHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetExternalEntityRefHandlerArg
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetSkippedEntityHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetUnknownEncodingHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetElementDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetAttlistDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetEntityDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_SetXmlDeclHandler
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_SetParamEntityParsing
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_Parse
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ParseBuffer
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetBuffer
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_StopParser
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ResumeParser
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetParsingStatus
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetErrorCode
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetCurrentByteIndex
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetCurrentByteCount
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetInputContext
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetCurrentLineNumber
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
XML_GetCurrentColumnNumber
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_FreeContentModel
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_MemMalloc
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_MemRealloc
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_MemFree
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_DefaultCurrent
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ErrorString
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ExpatVersion
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_ExpatVersionInfo
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c XML_GetFeatureList
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c storeRawNames
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c contentProcessor
ANALYZE: /Volumes/Secure Image/Source
Code/git/svn/Max/Frameworks/expat/expat/lib/xmlparse.c
externalEntityInitProcessor
Assertion failed: (isa<X>(Val) && "cast<Ty>() argument of incompatible type!"),
function cast, file
/private/tmp/checker-0.139-build/include/llvm/Support/Casting.h, line 199.
0   clang             0x00a90cd2 std::vector<llvm::sys::Path,
std::allocator<llvm::sys::Path>
>::_M_insert_aux(__gnu_cxx::__normal_iterator<llvm::sys::Path*,
std::vector<llvm::sys::Path, std::allocator<llvm::sys::Path> > >,
llvm::sys::Path const&) + 7746
1   libSystem.B.dylib 0x926be2bb _sigtramp + 43
2   libSystem.B.dylib 0xffffffff _sigtramp + 1838423407
3   libSystem.B.dylib 0x9273223a raise + 26
4   libSystem.B.dylib 0x9273e679 abort + 73
5   libSystem.B.dylib 0x927333db __assert_rtn + 101
6   clang             0x002d84fb
clang::DiagCollector::HandleDiagnostic(clang::Diagnostic::Level,
clang::DiagnosticInfo const&) + 28763
7   clang             0x002dffb1
clang::DiagCollector::HandleDiagnostic(clang::Diagnostic::Level,
clang::DiagnosticInfo const&) + 60177
8   clang             0x0030d69e clang::CFGBlock::operator[](unsigned long)
const + 81054
9   clang             0x0030d28c clang::CFGBlock::operator[](unsigned long)
const + 80012
10  clang             0x0030d28c clang::CFGBlock::operator[](unsigned long)
const + 80012
11  clang             0x0030d28c clang::CFGBlock::operator[](unsigned long)
const + 80012
12  clang             0x0030d28c clang::CFGBlock::operator[](unsigned long)
const + 80012
13  clang             0x0030e988 clang::CFGBlock::operator[](unsigned long)
const + 85896
14  clang             0x003041c3 clang::CFGBlock::operator[](unsigned long)
const + 42947
15  clang             0x00304995 clang::CFGBlock::operator[](unsigned long)
const + 44949
16  clang             0x003040bf clang::CFGBlock::operator[](unsigned long)
const + 42687
17  clang             0x0030f120 clang::CFGBlock::operator[](unsigned long)
const + 87840
18  clang             0x003159a1
clang::GRCoreEngine<clang::GRExprEngine>::ProcessStmt(clang::Stmt*,
clang::GRStmtNodeBuilderImpl&) + 97
19  clang             0x002f8e86
llvm::ImutAVLFactory<llvm::ImutKeyValueInfo<unsigned int, unsigned int>
>::Add_internal(std::pair<unsigned int, unsigned int> const&,
llvm::ImutAVLTree<llvm::ImutKeyValueInfo<unsigned int, unsigned int> >*) + 9414
20  clang             0x002f9098
llvm::ImutAVLFactory<llvm::ImutKeyValueInfo<unsigned int, unsigned int>
>::Add_internal(std::pair<unsigned int, unsigned int> const&,
llvm::ImutAVLTree<llvm::ImutKeyValueInfo<unsigned int, unsigned int> >*) + 9944
21  clang             0x0000a0b7 llvm::cast_retty<clang::CXXRecordType,
clang::Type*>::ret_type llvm::cast<clang::CXXRecordType,
clang::Type*>(clang::Type* const&) + 9911
22  clang             0x00009968 llvm::cast_retty<clang::CXXRecordType,
clang::Type*>::ret_type llvm::cast<clang::CXXRecordType,
clang::Type*>(clang::Type* const&) + 8040
23  clang             0x00009db5 llvm::cast_retty<clang::CXXRecordType,
clang::Type*>::ret_type llvm::cast<clang::CXXRecordType,
clang::Type*>(clang::Type* const&) + 9141
24  clang             0x0034f56b
std::vector<clang::IdentifierResolver::IdDeclInfo,
std::allocator<clang::IdentifierResolver::IdDeclInfo>
>::_M_fill_insert(__gnu_cxx::__normal_iterator<clang::IdentifierResolver::IdDeclInfo*,
std::vector<clang::IdentifierResolver::IdDeclInfo,
std::allocator<clang::IdentifierResolver::IdDeclInfo> > >, unsigned long,
clang::IdentifierResolver::IdDeclInfo const&) + 1483
25  clang             0x0004c6e6
clang::TokenRewriter::AddTokenAfter(std::_List_const_iterator<clang::Token>,
char const*) + 5622
26  clang             0x0004f79d
clang::TokenRewriter::AddTokenAfter(std::_List_const_iterator<clang::Token>,
char const*) + 18093
27  clang             0x00002536 _mh_execute_header + 5430
28  clang             0x0000001a _mh_execute_header + 18446744073709547546


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 15 16:45:19 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 Jan 2009 16:45:19 -0600
Subject: [LLVMbugs] [Bug 3339] New: Bug when compiling OS kernel code
Message-ID: <bug-3339-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3339

           Summary: Bug when compiling OS kernel code
           Product: libraries
           Version: 2.4
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Keywords: compile-fail
          Severity: major
          Priority: P2
         Component: Backend: C
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: yim6 at illinois.edu
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2392)
 --> (http://llvm.org/bugs/attachment.cgi?id=2392)
Testcase that causes a compile error

The attached is a testcase which causes a back-end error in LLVM 2.4.
 These types of assembly routines are mainly used in OS kernels. 
 I checked that GCC 4.3 works for this routine.

 The error messsage is ...
         Couldn't allocate output reg for constraint 'A'!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 03:48:39 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 03:48:39 -0600
Subject: [LLVMbugs] [Bug 3340] New: "warning: cannot pass object of non-POD
 type 'char []' through variadic function;
 call will abort at runtime" suddenly appears
Message-ID: <bug-3340-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3340

           Summary: "warning: cannot pass object of non-POD type 'char []'
                    through variadic function; call will abort at runtime"
                    suddenly appears
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: rdivacky at freebsd.org
                CC: llvmbugs at cs.uiuc.edu


when compiling screen with clang I am suddenly getting this: 

screen.c:3577:36: warning: cannot pass object of non-POD type 'char []' through
variadic function; call will abort at runtime
     Panic(0, "Screen version %s", version);
                                   ^



this wasnt the case a week or so ago... I attached "ccc -E" produced version of
screen.c as a testcase


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 09:34:28 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 09:34:28 -0600
Subject: [LLVMbugs] [Bug 3341] New: clang gets confused by redefinition of
 extern inline functions.
Message-ID: <bug-3341-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3341

           Summary: clang gets confused by redefinition of extern inline
                    functions.
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: AST
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bolzoni at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu


--->
/* from c torture. */
inline extern int foo () { return 0; }
inline extern int bar () { return 0; }
inline static int bar () { return foo(); }
---<
$ gcc -W -Wall -pedantic -c -o file.o file.c
no problem

$ clang file.c 
20021120-1.c:4:19: error: redefinition of 'bar'
inline static int bar () { return foo(); }
                  ^
20021120-1.c:3:19: note: previous definition is here
inline extern int bar () { return 0; }
                  ^
2 diagnostics generated.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 10:50:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 10:50:21 -0600
Subject: [LLVMbugs] [Bug 3340] "warning: cannot pass object of non-POD type
 'char []' through variadic function;
 call will abort at runtime" suddenly appears
In-Reply-To: <bug-3340-206@http.llvm.org/bugs/>
Message-ID: <200901161650.n0GGoLPb018791@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3340


Anders Carlsson <andersca at mac.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |andersca at mac.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Anders Carlsson <andersca at mac.com>  2009-01-16 10:50:20 ---
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090112/010881.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 11:09:01 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 11:09:01 -0600
Subject: [LLVMbugs] [Bug 3339] Bug when compiling OS kernel code
In-Reply-To: <bug-3339-206@http.llvm.org/bugs/>
Message-ID: <200901161709.n0GH91CF019518@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3339


Andrew Lenharth <alenhar2 at uiuc.edu> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |alenhar2 at uiuc.edu
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #1 from Andrew Lenharth <alenhar2 at uiuc.edu>  2009-01-16 11:09:00 ---
As mentioned on the email list, this has been reported multiple times and
fixed.

*** This bug has been marked as a duplicate of bug 1779 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 13:25:43 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 13:25:43 -0600
Subject: [LLVMbugs] [Bug 3342] New: Distinguish pointer incompatibilities
 where only sign differs from other incompatibilities
Message-ID: <bug-3342-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3342

           Summary: Distinguish pointer incompatibilities where only sign
                    differs from other incompatibilities
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: danchr at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2397)
 --> (http://llvm.org/bugs/attachment.cgi?id=2397)
First attempt???

Lots of legacy code happily compares loads of pointers of incompatible signs.
The attached patch is a rough stab at implementing the -Wno-pointer-sign flag
that GCC has for suppressing these.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 13:28:40 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 13:28:40 -0600
Subject: [LLVMbugs] [Bug 3338] crash in CFGBlock with clang 0.139 (second
	crash)
In-Reply-To: <bug-3338-206@http.llvm.org/bugs/>
Message-ID: <200901161928.n0GJSeFo024272@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3338


Ted Kremenek <kremenek at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #3 from Ted Kremenek <kremenek at apple.com>  2009-01-16 13:28:40 ---
This looks like a duplicate of PR 3337.  This should be fixed in checker-0.140.

*** This bug has been marked as a duplicate of bug 3337 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 14:13:41 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 14:13:41 -0600
Subject: [LLVMbugs] [Bug 3335] instcombine ignores addrspace in
	InstCombineStoreToCast
In-Reply-To: <bug-3335-206@http.llvm.org/bugs/>
Message-ID: <200901162013.n0GKDfci026098@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3335


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-16 14:13:40 ---
Fixed, thanks for the small testcase:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072267.html

Note that instcombine probably does a similar xform for loads.  If you notice
this, please file another bugzilla with a test case and I'll fix asap.  Thanks!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 14:17:49 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 14:17:49 -0600
Subject: [LLVMbugs] [Bug 3334] Interpreter asserts on out of range shift
	amount
In-Reply-To: <bug-3334-206@http.llvm.org/bugs/>
Message-ID: <200901162017.n0GKHnRC026464@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3334


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-16 14:17:48 ---
These have undefined behavior, but should not crash the interpreter obviously. 
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072268.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 14:43:10 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 14:43:10 -0600
Subject: [LLVMbugs] [Bug 3343] New: llc abort with -pre-alloc-split
Message-ID: <bug-3343-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3343

           Summary: llc abort with -pre-alloc-split
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: gohman at apple.com
                CC: resistor at mac.com, llvmbugs at cs.uiuc.edu


While testing MachineLICM I found a regression in
test/CodeGen/X86/pre-split2.ll.  By making a minor edit
to this testcase, I can reproduce the failure without running
MachineLICM.  Attached is the modified testcase.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 16 16:40:32 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Jan 2009 16:40:32 -0600
Subject: [LLVMbugs] [Bug 2477] clang misparses "//*" in C89 mode
In-Reply-To: <bug-2477-206@http.llvm.org/bugs/>
Message-ID: <200901162240.n0GMeWgZ031529@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2477


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Chris Lattner <clattner at apple.com>  2009-01-16 16:40:29 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090112/010918.html

Thanks Eli,


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 17 10:10:13 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 Jan 2009 10:10:13 -0600
Subject: [LLVMbugs] [Bug 3344] New: clang unconditionally rejects code with
 too many braces that is accepted by gcc
Message-ID: <bug-3344-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3344

           Summary: clang unconditionally rejects code with too many braces
                    that is accepted by gcc
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


$ cat bug1.c
struct S
{
  int i;
};

struct S s =
{
  {{ 1 }}
};
$ gcc -w -c bug1.c
$ clang -w bug1.c
bug1.c:8:3: error: too many braces around scalar initializer
  {{ 1 }}
  ^~~~~~~
1 diagnostic generated.
$


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 17 10:42:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 Jan 2009 10:42:35 -0600
Subject: [LLVMbugs] [Bug 3345] New: clang rejects asm code accepted by gcc
	and icc
Message-ID: <bug-3345-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3345

           Summary: clang rejects asm code accepted by gcc and icc
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


This code comes straight from the Linux kernel.

$ cat /tmp/bug2.c 
unsigned long __clear_user(void *addr, unsigned long size)
{
 long __d0;


 asm volatile(
  "     testq  %[size8],%[size8]\n"
  "     jz     4f\n"
  "0:   movq %[zero],(%[dst])\n"
  "     addq   %[eight],%[dst]\n"
  "     decl %%ecx ; jnz   0b\n"
  "4:   movq  %[size1],%%rcx\n"
  "     testl %%ecx,%%ecx\n"
  "     jz     2f\n"
  "1:   movb   %b[zero],(%[dst])\n"
  "     incq   %[dst]\n"
  "     decl %%ecx ; jnz  1b\n"
  "2:\n"
  ".section .fixup,\"ax\"\n"
  "3:   lea 0(%[size1],%[size8],8),%[size8]\n"
  "     jmp 2b\n"
  ".previous\n"
  ".section __ex_table,\"a\"\n"
  "       .align 8\n"
  "     .quad 0b,3b\n"
  "     .quad 1b,2b\n"
  ".previous"
  : [size8] "=c"(size), [dst] "=&D" (__d0)
  : [size1] "r"(size & 7), "[size8]" (size / 8), "[dst]"(addr),
    [zero] "r" (0UL), [eight] "r" (8UL));
 return size;
}
$ gcc -w -c /tmp/bug2.c
$ icc -w -c /tmp/bug2.c
$ clang /tmp/bug2.c
/tmp/bug2.c:29:28: error: invalid input constraint '[size8]' in asm
  : [size1] "r"(size & 7), "[size8]" (size / 8), "[dst]"(addr),
                           ^
1 diagnostic generated.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 17 20:07:01 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 Jan 2009 20:07:01 -0600
Subject: [LLVMbugs] [Bug 3345] clang rejects asm code accepted by gcc and icc
In-Reply-To: <bug-3345-206@http.llvm.org/bugs/>
Message-ID: <200901180207.n0I271xm027660@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3345


Anders Carlsson <andersca at mac.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |andersca at mac.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Anders Carlsson <andersca at mac.com>  2009-01-17 20:07:00 ---
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090112/010970.html
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090112/010973.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 17 21:47:37 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 Jan 2009 21:47:37 -0600
Subject: [LLVMbugs] [Bug 3346] New: Possible bug in linear scan allocator
Message-ID: <bug-3346-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3346

           Summary: Possible bug in linear scan allocator
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Register Allocator
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: resistor at mac.com
                CC: llvmbugs at cs.uiuc.edu


If you apply the attached patch to enable pre-alloc splitting with phi join
creation, telecom-FFT is miscompiled.  This occurs at iteration 5 of pre-alloc
splitting (-pre-split-limit=5).  However, looking at the diffs from
before/after that iteration, the transformation looks sounds.  Could this be a
bug later in codegen (regalloc?) that's being exposed?


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 08:41:11 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 08:41:11 -0600
Subject: [LLVMbugs] [Bug 3347] New: llvm-config 2.4 - libpsapi,
 libimagehlp in wrong order
Message-ID: <bug-3347-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3347

           Summary: llvm-config 2.4 - libpsapi, libimagehlp in wrong order
           Product: tools
           Version: 2.4
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-config
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: duncan at duncanpierce.org
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2408)
 --> (http://llvm.org/bugs/attachment.cgi?id=2408)
Corrects order for --libs, may be wrong for --libnames, --libfiles

llvm-config --cxxflags --ldflags --libs core gives:

-I/include  -D_DEBUG  -D_GNU_SOURCE -D__STDC_LIMIT_MACROS -O2
-fomit-frame-pointer  -Woverloaded-virtual
-L/lib  -lpsapi -limagehlp -lm
-lLLVMCore -lLLVMSupport -lLLVMSystem

The -lpsapi -limagehlp need to be at the end for ld to resolve all the symbols
correctly. (Probably the -lm too but I didn't experience problems with this).

A patch for llvm-config.in.in is attached which gets things working on MinGW
but it's not tested elsewhere.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 08:51:29 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 08:51:29 -0600
Subject: [LLVMbugs] [Bug 3348] New: llvm-config 2.4 in msys. bat - double
 slashes in generated paths
Message-ID: <bug-3348-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3348

           Summary: llvm-config 2.4 in msys.bat - double slashes in
                    generated paths
           Product: tools
           Version: 2.4
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-config
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: duncan at duncanpierce.org
                CC: llvmbugs at cs.uiuc.edu


When run from msys.bat (but not when run from DOS prompt), llvm-config outputs
double slashes for pathnames:

$ llvm-config --cxxflags --ldflags --libs
-I//include [...]
-L//lib [...]
-lLLVMLinker [...]
//lib/LLVMExecutionEngine.o [...]

g++ (3.4.5 at least) fails to understand these. llvm-config produces single
slashes from DOS prompt, and g++ works correctly.

My llvm-config was built using ./configure --prefix=
This may not be the correct way of configuring for "top level" installation.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 13:19:02 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 13:19:02 -0600
Subject: [LLVMbugs] [Bug 3322] [interpreter] __dyld_misaligned_stack_error
 on Mac OS X 10. 5 x86 when using system call
In-Reply-To: <bug-3322-206@http.llvm.org/bugs/>
Message-ID: <200901181919.n0IJJ2WA003579@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3322


Alexei Svitkine <alexei.svitkine at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #6 from Alexei Svitkine <alexei.svitkine at gmail.com>  2009-01-18 13:19:01 ---


*** This bug has been marked as a duplicate of bug 2128 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 13:42:22 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 13:42:22 -0600
Subject: [LLVMbugs] [Bug 2510] Non-power-of-2 _Bool vector asserts
In-Reply-To: <bug-2510-206@http.llvm.org/bugs/>
Message-ID: <200901181942.n0IJgMVM004629@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2510


Nuno Lopes <nunoplopes at sapo.pt> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |nunoplopes at sapo.pt
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Nuno Lopes <nunoplopes at sapo.pt>  2009-01-18 13:42:21 ---
Nate Begeman fixed this yesterday.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 14:00:36 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 14:00:36 -0600
Subject: [LLVMbugs] [Bug 3349] New: clang rejects apparently valid code
Message-ID: <bug-3349-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3349

           Summary: clang rejects apparently valid code
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


This but was discovered trying to parse diffutils-2.8.1:

$ cat bug.c
struct re_pattern_buffer ignore_regexp;

struct re_pattern_buffer {
  unsigned char *buffer;
};
$ gcc -Wall -W -c  bug.c
$ icc -w2 -c bug.c
$ clang -w bug.c
bug.c:1:26: error: variable has incomplete type 'struct re_pattern_buffer'
struct re_pattern_buffer ignore_regexp;
                         ^
1 diagnostic generated.
$

This code parses OK also with the Comeau C compiler.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 14:25:44 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 14:25:44 -0600
Subject: [LLVMbugs] [Bug 3350] New: clang rejects valid code involving va_end
Message-ID: <bug-3350-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3350

           Summary: clang rejects valid code involving va_end
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


This bug was discovered trying to parse coreutils-6.12:

$ cat bug.c
#include <stdarg.h>

void
foo(const char *command_name, va_list authors) {
  (void) command_name;
  va_end (authors);
}
$ gcc -c -W -Wall bug.c
$ icc -c bug.c
$ clang bug.c
bug.c:6:3: error: incompatible type passing 'struct __va_list_tag *', expected
'__builtin_va_list &'
  va_end (authors);
  ^~~~~~
1 diagnostic generated.
$ 

This code parses OK also with the Comeau C compiler.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 14:29:45 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 14:29:45 -0600
Subject: [LLVMbugs] [Bug 3351] New: constant folding should fold this
	constant expr
Message-ID: <bug-3351-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3351

           Summary: constant folding should fold this constant expr
           Product: libraries
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: code-quality
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


483.xalancbmk retains this awesome constant expr until codegen:

        %0 = load i16* getelementptr ([7 x i16]*
@_ZN11xercesc_2_5L17gIdeographicCharsE, i32 0, i32 add (i32 ashr (i32 sub (i32
ptrtoint (i16* getelementptr ([7 x i16]*
@_ZN11xercesc_2_5L17gIdeographicCharsE, i32 0, i32 4) to i32), i32 ptrtoint ([7
x i16]* @_ZN11xercesc_2_5L17gIdeographicCharsE to i32)), i32 1), i32 1)), align
4              ; <i16> [#uses=0]

This should get folded at the sub.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 14:35:15 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 14:35:15 -0600
Subject: [LLVMbugs] [Bug 3349] clang rejects apparently valid code
In-Reply-To: <bug-3349-206@http.llvm.org/bugs/>
Message-ID: <200901182035.n0IKZFSl006840@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3349


Daniel Dunbar <daniel at zuster.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |daniel at zuster.org
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #2 from Daniel Dunbar <daniel at zuster.org>  2009-01-18 14:35:15 ---


*** This bug has been marked as a duplicate of bug 3310 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 18:25:12 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 18:25:12 -0600
Subject: [LLVMbugs] [Bug 3352] New: Not implemented? "Cannot yet select"
Message-ID: <bug-3352-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3352

           Summary: Not implemented? "Cannot yet select"
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


When trying to generate code for the abs function for a 4 element vector of
float, I get the message below.

LLVM code:

define <4 x float> @vabs(<4 x float>) {
L1:
        %1 = fcmp olt <4 x float> %0, zeroinitializer
        %2 = sub <4 x float> zeroinitializer, %0
        %3 = select <4 x i1> %1, <4 x float> %2, <4 x float> %0
        ret <4 x float> %3
}


Error:
Cannot yet select: 0x100b97c: i32 = X86ISD::CMP 0x100b86c, 0x100b42c
0   llc                                 0x008f1c5c
_ZNSt8_Rb_treeIN4llvm3sys4PathES2_St9_IdentityIS2_ESt4lessIS2_ESaIS2_EE13insert_uniqueERKS2_
+ 7948
1   libSystem.B.dylib                   0x9526d2bb _sigtramp + 43
2   ???                                 0xffffffff 0x0 + 4294967295
3   libSystem.B.dylib                   0x952e123a raise + 26
4   libSystem.B.dylib                   0x952ed679 abort + 73
5   llc                                 0x0030c6d0
_ZNK4llvm14ilist_iteratorINS_6SDNodeEEcvPS1_Ev + 176
6   llc                                 0x00365b3f
_ZN48_GLOBAL__N_X86ISelDAGToDAG.cpp_00000000_274AD75D18X86ISelAddressMode4dumpEv
+ 362831
7   llc                                 0x0040f1d6
_ZN48_GLOBAL__N_X86ISelDAGToDAG.cpp_00000000_274AD75D18X86ISelAddressMode4dumpEv
+ 1056742
8   llc                                 0x002a5c82
_ZNK4llvm13TargetMachine12getSubtargetINS_12X86SubtargetEEERKT_v + 133762
9   llc                                 0x00413cb3
_ZN48_GLOBAL__N_X86ISelDAGToDAG.cpp_00000000_274AD75D18X86ISelAddressMode4dumpEv
+ 1075907
10  llc                                 0x002a8d03
_ZNK4llvm13TargetMachine12getSubtargetINS_12X86SubtargetEEERKT_v + 146179
11  llc                                 0x005fd97b
_ZNK4llvm14TargetLowering15getNumRegistersENS_3MVTE + 22779
12  llc                                 0x005ff977
_ZNK4llvm14TargetLowering15getNumRegistersENS_3MVTE + 30967
13  llc                                 0x00600127
_ZNK4llvm14TargetLowering15getNumRegistersENS_3MVTE + 32935
14  llc                                 0x006017fc
_ZNK4llvm14TargetLowering15getNumRegistersENS_3MVTE + 38780
15  llc                                 0x0086e19f _ZN4llvm12FunctionPassD1Ev +
33839
16  llc                                 0x0086e738 _ZN4llvm12FunctionPassD1Ev +
35272
17  llc                                 0x0086e8e6 _ZN4llvm12FunctionPassD1Ev +
35702
18  llc                                 0x00002b67 _mh_execute_header + 7015
19  llc                                 0x00001836 _mh_execute_header + 2102


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 18:37:11 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 18:37:11 -0600
Subject: [LLVMbugs] [Bug 3292] Assertion `i*2 < getNumOperands() && "Invalid
 value number!"' failed.
In-Reply-To: <bug-3292-206@http.llvm.org/bugs/>
Message-ID: <200901190037.n0J0bBg8014554@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3292


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Chris Lattner <clattner at apple.com>  2009-01-18 18:37:11 ---
Fixed here, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072327.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 18:42:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 18:42:27 -0600
Subject: [LLVMbugs] [Bug 3302] yet another math bug??
In-Reply-To: <bug-3302-206@http.llvm.org/bugs/>
Message-ID: <200901190042.n0J0gRWt014774@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3302


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-18 18:42:26 ---
I think this got fixed by the recent SCEV reversions.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 18:45:02 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 18:45:02 -0600
Subject: [LLVMbugs] [Bug 3296] another math bug
In-Reply-To: <bug-3296-206@http.llvm.org/bugs/>
Message-ID: <200901190045.n0J0j2pF014876@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3296


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Chris Lattner <clattner at apple.com>  2009-01-18 18:45:01 ---
I think this got fixed with the recent SCEV reversions.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 18:47:11 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 18:47:11 -0600
Subject: [LLVMbugs] [Bug 3294] another math bug
In-Reply-To: <bug-3294-206@http.llvm.org/bugs/>
Message-ID: <200901190047.n0J0lBis014992@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3294


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #7 from Chris Lattner <clattner at apple.com>  2009-01-18 18:47:10 ---
This now works, I think this was fixed by the SCEV reversions.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 18:52:07 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 18:52:07 -0600
Subject: [LLVMbugs] [Bug 3207] Crash on invalid in DeclGroupOwningRef
In-Reply-To: <bug-3207-206@http.llvm.org/bugs/>
Message-ID: <200901190052.n0J0q73E015298@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3207


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-18 18:52:06 ---
this got fixed along the way


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 18:55:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 18:55:21 -0600
Subject: [LLVMbugs] [Bug 3119] maybe a loop miscompilation
In-Reply-To: <bug-3119-206@http.llvm.org/bugs/>
Message-ID: <200901190055.n0J0tL2f015478@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3119


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #6 from Chris Lattner <clattner at apple.com>  2009-01-18 18:55:20 ---
this works for me with mainline.  I'm assuming it was fixed by the reversion of
the buggy SCEV stuff.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 20:47:14 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 20:47:14 -0600
Subject: [LLVMbugs] [Bug 3016] simplifycfg creates dead code and then
 creates invalid code because of it
In-Reply-To: <bug-3016-206@http.llvm.org/bugs/>
Message-ID: <200901190247.n0J2lE88019978@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3016


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
            Summary|null ptr dereference        |simplifycfg creates dead
                   |                            |code and then creates
                   |                            |invalid code because of it




--- Comment #6 from Chris Lattner <clattner at apple.com>  2009-01-18 20:47:12 ---
Fixed here, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072338.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 20:51:17 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 20:51:17 -0600
Subject: [LLVMbugs] [Bug 3006] llc: assertion failure with inline asm
In-Reply-To: <bug-3006-206@http.llvm.org/bugs/>
Message-ID: <200901190251.n0J2pHXS020177@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3006


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-18 20:51:16 ---
This should be rejected by the front-end now.  llc produces:

Unsupported asm: input constraint with a matching output constraint of
incompatible type!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 20:52:29 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 20:52:29 -0600
Subject: [LLVMbugs] [Bug 2999] miscompilation due to gcc inliner?
In-Reply-To: <bug-2999-206@http.llvm.org/bugs/>
Message-ID: <200901190252.n0J2qTCl020258@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2999


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #8 from Chris Lattner <clattner at apple.com>  2009-01-18 20:52:29 ---
This appears to work on mainline.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 18 22:01:31 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Jan 2009 22:01:31 -0600
Subject: [LLVMbugs] [Bug 3295] loop-reduce / scalar-evolution crash
 expanding {0, +, 2, +, 2} addrec
In-Reply-To: <bug-3295-206@http.llvm.org/bugs/>
Message-ID: <200901190401.n0J41VP7022791@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3295


Nick Lewycky <nicholas at mxc.ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #6 from Nick Lewycky <nicholas at mxc.ca>  2009-01-18 22:01:31 ---
Yes, this was fixed by the recent SCEV reversion.

After the branch, I plan to fix this properly by removing the "round up to next
power of 2" logic in SCEV, but I consider that too dangerous this close to the
release.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 11:17:08 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 11:17:08 -0600
Subject: [LLVMbugs] [Bug 3353] New: Infinite loop while building libxml
Message-ID: <bug-3353-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3353

           Summary: Infinite loop while building libxml
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: opt
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rafael.espindola at gmail.com
                CC: llvmbugs at cs.uiuc.edu


To reproduce:

opt test.bc -inline -jump-threading -licm -loop-unswitch -gvn -simplifycfg
-verify -o t.bc


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 14:19:28 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 14:19:28 -0600
Subject: [LLVMbugs] [Bug 3354] New: apparent bad interaction between math
 and short-circuit
Message-ID: <bug-3354-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3354

           Summary: apparent bad interaction between math and short-circuit
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


The attached program displays this behavior on x86 Linux:

[regehr at babel tmp30]$ llvm-gcc -O small.c -o small
small.c: In function ???main???:
small.c:19: warning: division by zero
small.c:19: warning: division by zero
[regehr at babel tmp30]$ ./small
x=1
Floating point exception

The thing is, the divide-by-zero behaviors are all protected by
short-circuiting operations, and should never fire.  I claim that LLVM's -O0
behavior is what must occur at all optimization levels:

[regehr at babel tmp30]$ llvm-gcc -O0 small.c -o small
small.c: In function ???main???:
small.c:19: warning: division by zero
small.c:19: warning: division by zero
[regehr at babel tmp30]$ ./small
x=1
x=0

The macro in the attached code is simply a macroized version of the 2nd safe
multiplication fragment here:

https://www.securecoding.cert.org/confluence/display/cplusplus/INT32-CPP.+Ensure+that+operations+on+signed+integers+do+not+result+in+overflow


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 15:07:02 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 15:07:02 -0600
Subject: [LLVMbugs] [Bug 2980] llc crashes with Cannot yet select:
 0xa60d794: ch = declare 0xa60d568, 0xa60d70c, 0xa60d8a4
In-Reply-To: <bug-2980-206@http.llvm.org/bugs/>
Message-ID: <200901192107.n0JL724x007531@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2980


Devang Patel <dpatel at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Devang Patel <dpatel at apple.com>  2009-01-19 15:07:01 ---
Added verification check to flag invalid debug info.
rev. 62526.

If llvm-gcc FE is generating this invalid debug info, then pl. file separate PR
to track llvm-gcc bug. Thanks.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 15:20:59 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 15:20:59 -0600
Subject: [LLVMbugs] [Bug 3316] Constant folding for 'frem' is flaky
In-Reply-To: <bug-3316-206@http.llvm.org/bugs/>
Message-ID: <200901192120.n0JLKxYt008024@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3316


Dale Johannesen <dalej at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #6 from Dale Johannesen <dalej at apple.com>  2009-01-19 15:20:58 ---
I see, currently the FP-to-integer conversions are always called with
TowardsZero, except this one place.  So the bug calling it with
NearestTiesToEven would never show up anywhere else.  This fixes both bugs.
Neil, please review.
http://llvm.org/viewvc/llvm-project?view=rev&revision=62528


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 15:21:02 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 15:21:02 -0600
Subject: [LLVMbugs] [Bug 3353] Infinite loop while building libxml
In-Reply-To: <bug-3353-206@http.llvm.org/bugs/>
Message-ID: <200901192121.n0JLL27P008043@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3353


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Chris Lattner <clattner at apple.com>  2009-01-19 15:21:02 ---
Fixed, sorry for the breakage!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072364.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 15:56:28 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 15:56:28 -0600
Subject: [LLVMbugs] [Bug 3354] apparent bad interaction between math and
	short-circuit
In-Reply-To: <bug-3354-206@http.llvm.org/bugs/>
Message-ID: <200901192156.n0JLuSJV009357@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3354


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Chris Lattner <clattner at apple.com>  2009-01-19 15:56:27 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072366.html

Reassociate was moving a constant expression to a different block under the
assumption that constants can't have side effects.  I think that this is a
reasonable expectation, so I changed x/0 to constant fold into undef as is
allowed by langref.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 16:04:56 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 16:04:56 -0600
Subject: [LLVMbugs] [Bug 3018] bug in SelectionDAG::ReplaceAllUsesWith
In-Reply-To: <bug-3018-206@http.llvm.org/bugs/>
Message-ID: <200901192204.n0JM4uv9009675@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3018


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 16:05:51 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 16:05:51 -0600
Subject: [LLVMbugs] [Bug 3354] apparent bad interaction between math and
	short-circuit
In-Reply-To: <bug-3354-206@http.llvm.org/bugs/>
Message-ID: <200901192205.n0JM5pCL009739@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3354


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |




--- Comment #4 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-19 16:05:51 ---
(In reply to comment #3)
> Reassociate was moving a constant expression to a different block under the
> assumption that constants can't have side effects.  I think that this is a
> reasonable expectation, so I changed x/0 to constant fold into undef as is
> allowed by langref.

That's only a workaround... what if the denominator is an unfoldable constant?


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 16:56:17 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 16:56:17 -0600
Subject: [LLVMbugs] [Bug 3355] New: llvm handles inline functions in C99
	poorly
Message-ID: <bug-3355-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3355

           Summary: llvm handles inline functions in C99 poorly
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Support Libraries
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: dalej at apple.com
                CC: llvmbugs at cs.uiuc.edu


A refresher:  the semantics of function definitions declared "inline" in C99
are that the body may be substituted into callers within that compilation unit,
but this should not create an external definition for use in other compilation
units.  There must be a (non-inline) definition in some other compilation unit,
and there is no requirement that it be functionally equivalent to the one in
this file; it's "unspecified behavior" if that happens (not "undefined").

Note that gcc-89 provides equivalent functionality using "extern inline" in a
way that's incompatible with C99, and that comments within the gcc sources use
"extern inline" for these semantics.  Also note C++ semantics for inline are
completely different, but you knew that.

Moving the inlining phase from gcc to llvm sort of broke this.  Currently we
(i.e. cgraph in llvm-gcc) treats these functions by not passing them into
llvm's inliner, which is correct but not very useful.  If we do pass them in we
should have a way to tell llvm not to generate a definition for them, and I
don't think there is one.  (Although pretending they are static may work OK in
practice; if something doesn't get inlined that will wind up invoking a
different leg of unspecified behavior, which is conformant but may be
surprising.  It's a hack, though.)

For always-inline C99 inline functions, llvm needs to see the bodies for
correctness.  I've fixed that case using the "static" hack (a static definition
may be emitted, but that's harmless since nothing can reference it).  But even
here I'd prefer to have a way to tell llvm not to emit the body, for QOI if
nothing else.

So I'm suggesting an IR extension here.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From neil at daikokuya.co.uk  Mon Jan 19 17:09:45 2009
From: neil at daikokuya.co.uk (Neil Booth)
Date: Tue, 20 Jan 2009 08:09:45 +0900
Subject: [LLVMbugs] [Bug 3316] Constant folding for 'frem' is flaky
In-Reply-To: <200901192120.n0JLKxYt008024@zion.cs.uiuc.edu>
References: <bug-3316-206@http.llvm.org/bugs/>
	<200901192120.n0JLKxYt008024@zion.cs.uiuc.edu>
Message-ID: <20090119230945.GB14812@daikokuya.co.uk>

bugzilla-daemon at cs.uiuc.edu wrote:-

> http://llvm.org/bugs/show_bug.cgi?id=3316
> 
> 
> Dale Johannesen <dalej at apple.com> changed:
> 
>            What    |Removed                     |Added
> ----------------------------------------------------------------------------
>              Status|NEW                         |RESOLVED
>          Resolution|                            |FIXED
> 
> 
> 
> 
> --- Comment #6 from Dale Johannesen <dalej at apple.com>  2009-01-19 15:20:58 ---
> I see, currently the FP-to-integer conversions are always called with
> TowardsZero, except this one place.  So the bug calling it with
> NearestTiesToEven would never show up anywhere else.  This fixes both bugs.
> Neil, please review.
> http://llvm.org/viewvc/llvm-project?view=rev&revision=62528

I'll take a look later; no time now.  But I left a comment in / above
the frem implementation pointing out that it isn't close to being
correct.  It will need a complete rewrite and will have to look like
the other operations before it can be called correct.

Neil.

From dalej at apple.com  Mon Jan 19 17:19:26 2009
From: dalej at apple.com (Dale Johannesen)
Date: Mon, 19 Jan 2009 15:19:26 -0800
Subject: [LLVMbugs] [Bug 3316] Constant folding for 'frem' is flaky
In-Reply-To: <20090119230945.GB14812@daikokuya.co.uk>
References: <bug-3316-206@http.llvm.org/bugs/>
	<200901192120.n0JLKxYt008024@zion.cs.uiuc.edu>
	<20090119230945.GB14812@daikokuya.co.uk>
Message-ID: <61EFD60E-4E39-4479-BDB5-265EAD23DB74@apple.com>


On Jan 19, 2009, at 3:09 PMPST, Neil Booth wrote:

> bugzilla-daemon at cs.uiuc.edu wrote:-
>
>> http://llvm.org/bugs/show_bug.cgi?id=3316
>>
>> --- Comment #6 from Dale Johannesen <dalej at apple.com>  2009-01-19  
>> 15:20:58 ---
>> I see, currently the FP-to-integer conversions are always called with
>> TowardsZero, except this one place.  So the bug calling it with
>> NearestTiesToEven would never show up anywhere else.  This fixes  
>> both bugs.
>> Neil, please review.
>> http://llvm.org/viewvc/llvm-project?view=rev&revision=62528
>
> I'll take a look later; no time now.  But I left a comment in / above
> the frem implementation pointing out that it isn't close to being
> correct.  It will need a complete rewrite and will have to look like
> the other operations before it can be called correct.

Ah, that's what TRT means.
Could you be more specific about what doesn't work (when you have time)?


From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 18:54:45 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 18:54:45 -0600
Subject: [LLVMbugs] [Bug 2128] interpreter should bind functions by type
In-Reply-To: <bug-2128-206@http.llvm.org/bugs/>
Message-ID: <200901200054.n0K0sjRM015246@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2128


Nick Lewycky <nicholas at mxc.ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #9 from Nick Lewycky <nicholas at mxc.ca>  2009-01-19 18:54:45 ---
Fixed in r62553!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 19:12:01 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 19:12:01 -0600
Subject: [LLVMbugs] [Bug 3216] Incorrect codegen with shifting+sign
 extension ( possibly DAGCombiner?)
In-Reply-To: <bug-3216-206@http.llvm.org/bugs/>
Message-ID: <200901200112.n0K1C1hd015843@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3216


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Dan Gohman <gohman at apple.com>  2009-01-19 19:12:01 ---
It turns out there was already code that attempted to avoid
creating non-power-of-2-sized loads, but it was checking
the wrong type. This is fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072385.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 19:16:41 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 19:16:41 -0600
Subject: [LLVMbugs] [Bug 3354] apparent bad interaction between math and
	short-circuit
In-Reply-To: <bug-3354-206@http.llvm.org/bugs/>
Message-ID: <200901200116.n0K1GfiP016030@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3354


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #9 from Chris Lattner <clattner at apple.com>  2009-01-19 19:16:40 ---
This patch fixes the rest of it:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072389.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 20:18:04 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 20:18:04 -0600
Subject: [LLVMbugs] [Bug 3016] simplifycfg creates dead code and then
 creates invalid code because of it
In-Reply-To: <bug-3016-206@http.llvm.org/bugs/>
Message-ID: <200901200218.n0K2I4dl018264@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3016


Bill Wendling <wendling at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |wendling at apple.com
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |




--- Comment #7 from Bill Wendling <wendling at apple.com>  2009-01-19 20:18:04 ---
I had to revert this patch. It was breaking the release bootstrap.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 19 21:48:07 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Jan 2009 21:48:07 -0600
Subject: [LLVMbugs] [Bug 3356] New: using llvm-config --libs results in
	missing symbols
Message-ID: <bug-3356-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3356

           Summary: using llvm-config --libs results in missing symbols
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: alexei.svitkine at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Consider the following example:

echo "int main(void) { return 0; }" > a.cpp
g++ a.cpp `llvm-config --ldflags` `llvm-config --libs`

The expected result is that a.cpp will compile and link properly. However,
instead there are link errors.

It turns out that LLVMCBase.o and LLVMCClang.o which are both output by
`llvm-config --libs` depend on symbols that are not found in any library given
by `llvm-config --libs`.

Some of the undefined symbols are:

  "llvmc::CompilationGraph::insertEdge(std::basic_string<char,
std::char_traits<char>, std::allocator<char> > const&, llvmc::Edge*)",
  "llvmc::CompilationGraph::insertNode(llvmc::Tool*)",
  "_OutputFilename",
  "llvmc::Tool::OutFilename(llvm::sys::Path const&, llvm::sys::Path const&,
bool, char const*) const",

Either the output of `llvm-config --libs` should include the missing libraries,
or LLVMCBase.o and LLVMCClang.o should be removed from that output. This would
allow projects using LLVM to link properly based on the output of `llvm-config
--libs`.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 01:47:05 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 01:47:05 -0600
Subject: [LLVMbugs] [Bug 3350] clang rejects valid code involving va_end
In-Reply-To: <bug-3350-206@http.llvm.org/bugs/>
Message-ID: <200901200747.n0K7l5sY028159@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3350


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-20 01:47:05 ---
Fixed in r62574.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 03:22:15 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 03:22:15 -0600
Subject: [LLVMbugs] [Bug 3357] New: Kaleidoscope Tutorial Ch3 doesn't compile
Message-ID: <bug-3357-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3357

           Summary: Kaleidoscope Tutorial Ch3 doesn't compile
           Product: Documentation
           Version: 2.4
          Platform: PC
               URL: http://llvm.org/docs/tutorial/LangImpl3.html
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: General docs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rmann at latencyzero.com
                CC: llvmbugs at cs.uiuc.edu


On line 350 of the code on the webpage, the declaration of Builder doesn't
account for the fact that IRBuilder is now templetized, and therefore needs at
least<>, but probably should get <true,NoFolder>. Doing the latter requires
adding

#include "llvm/Support/NoFolder.h"

Near the top of the file.

I jokingly proposed this idea in the chat room: that the Kaleidoscope tutorial
code be part of a regular test suite, and even that it be automatically
injected into the tutorial's markup. baldrick thought it was a good idea, so
I'm recording it in this bug.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 03:54:26 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 03:54:26 -0600
Subject: [LLVMbugs] [Bug 3358] New: invalid use of iterators in gvn
Message-ID: <bug-3358-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3358

           Summary: invalid use of iterators in gvn
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2417)
 --> (http://llvm.org/bugs/attachment.cgi?id=2417)
testcase .ll

When built with ENABLE_EXPENSIVE_CHECKS=1, the attached testcase (reduced
from the llvm-gcc build) gives:

$ opt -gvn regexp.bc -disable-output
/usr/include/c++/4.3/bits/stl_algo.h:2180:error: elements in iterator range     
    [__first, __last) are not partitioned by the value __val.                   

Objects involved in the operation:
iterator "__first" @ 0x0x7fffe6d01ca0 {
type =
N11__gnu_debug14_Safe_iteratorIN9__gnu_cxx17__normal_iteratorIPSt4pairIPN4llvm10BasicBlockENS4_12MemDepResultEENSt6__norm6vectorIS8_SaIS8_EEEEENSt7__debug6vectorIS8_SC_EEEE
(mutable iterator);                                                             
  state = dereferenceable (start-of-sequence);                                  
  references sequence with type
`NSt7__debug6vectorISt4pairIPN4llvm10BasicBlockENS2_12MemDepResultEESaIS6_EEE'
@ 0x0x7fffe6d01ca0             
}                                                                               
iterator "__last" @ 0x0x7fffe6d01c40 {                                          
type =
N11__gnu_debug14_Safe_iteratorIN9__gnu_cxx17__normal_iteratorIPSt4pairIPN4llvm10BasicBlockENS4_12MemDepResultEENSt6__norm6vectorIS8_SaIS8_EEEEENSt7__debug6vectorIS8_SC_EEEE
(mutable iterator);                                                             
  state = dereferenceable;                                                      
  references sequence with type
`NSt7__debug6vectorISt4pairIPN4llvm10BasicBlockENS2_12MemDepResultEESaIS6_EEE'
@ 0x0x7fffe6d01c40             
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 05:56:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 05:56:35 -0600
Subject: [LLVMbugs] [Bug 3359] New: fmod in APFloat is badly broken
Message-ID: <bug-3359-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3359

           Summary: fmod in APFloat is badly broken
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Support Libraries
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


Just some of the things wrong with it:

a) It takes a rounding mode.  The result is always exact, so a rounding mode
should not be an argument.

b) It doesn't handle specials (NaN, zero, infinity).

c) It gets standard cases wrong, e.g.


#include <math.h>
#include <stdio.h>

int main (void)
{
  float x = 0x1.345672p30;
  float y = 0x1.56789ap-20;
  float z1 = fmod (x, y), z2 = remainder (x,y );

  printf ("%a %a: fmod = %a remainder = %a\n", x, y, z1, z2);

  return 0;  
};

gives on NetBSD:

0x1.345672p+30 0x1.56789ap-20: fmod = 0x1.0f7dbap-20 remainder = -0x1.1beb8p-22

and for APFloat:

int main (void)
{
  char dst[30];

  APFloat x (APFloat::IEEEsingle, "0x1.345672p+30");
  APFloat y (APFloat::IEEEsingle, "0x1.56789ap-20");

  x.convertToHexString (dst, 0, false, APFloat::rmNearestTiesToEven);
  printf ("x: %s\n", dst);
  y.convertToHexString (dst, 0, false, APFloat::rmNearestTiesToEven);
  printf ("y: %s\n", dst);
  x.fmod(y, APFloat::rmNearestTiesToEven);
  x.convertToHexString (dst, 0, false, APFloat::rmNearestTiesToEven);
  printf ("mod: %s\n", dst);

  return 0;
}

gives:

x: 0x1.345672p30
y: 0x1.56789ap-20
mod: 0x0p0


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 07:07:26 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 07:07:26 -0600
Subject: [LLVMbugs] [Bug 3360] New: g++.dg/abi/local1.C fails
Message-ID: <bug-3360-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3360

           Summary: g++.dg/abi/local1.C fails
           Product: tools
           Version: 2.4
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-g++
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: jay.foad at antixlabs.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2418)
 --> (http://llvm.org/bugs/attachment.cgi?id=2418)
sources

The GCC 4.2.1 testsuite test g++.dg/abi/local1.C fails:

$ ~/llvm/llvm-gcc4.2-2.4-x86-linux-RHEL4/bin/llvm-g++ -o local1 local1.C
local1-a.cc 
$ ./local1 ; echo $?
1

It seems to be checking that identically named local classes in different
translation units get distinct typeids.

I've attached the sources for this test.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 11:04:50 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 11:04:50 -0600
Subject: [LLVMbugs] [Bug 3361] New: Should be possible to share stack slots
 when variables have non-overlapping liveness
Message-ID: <bug-3361-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3361

           Summary: Should be possible to share stack slots when variables
                    have non-overlapping liveness
           Product: new-bugs
           Version: unspecified
          Platform: PC
               URL: http://lkml.org/lkml/2009/1/12/344
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


See the URL for a detailed description of the problem, criticising gcc.

In short: stack usage should be "b + c", but it is "a + b (+ c)"
a is no longer live when b and c are used, thus they could share stack slots.
(they would have gotten different addresses, if they weren't inlined).

I tested with LLVM-gcc, and it seems that 'gcc -fno-strict-aliasing' performs
better in reducing stack size (however it is slightly better than gcc with
default strict-aliasing)

$ llvm-gcc -O2 -S -m32  linus.c; grep subl linus.s
        subl    $1632, %esp

$ gcc-4.3 -O2 linus.c; grep subl linus.s
        subl    $1640, %esp

$ gcc-4.3 -O2 -S -fno-strict-aliasing -m32  linus.c
        subl    $1240, %esp

Are there any flags to get the same (or better) stack usage as gcc's with
-fno-strict-aliasing? 

Here is the testcase as bitcode:
; ModuleID = 'linus.c'
target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
target triple = "i386-unknown-linux-gnu"
        %struct.a = type { i32, [200 x i32] }
        %struct.b = type { i32, [100 x i32] }

define i32 @fn(i32 %flag) nounwind {
entry:
        %c.i = alloca %struct.b         ; <%struct.b*> [#uses=1]
        %b.i = alloca %struct.b         ; <%struct.b*> [#uses=1]
        %a.i = alloca %struct.a         ; <%struct.a*> [#uses=1]
        %a1.i = bitcast %struct.a* %a.i to i8*          ; <i8*> [#uses=1]
        %0 = call i32 @fn3(i32 %flag, i8* %a1.i) nounwind               ; <i32>
[#uses=0]
        %1 = and i32 %flag, 1           ; <i32> [#uses=1]
        %toBool = icmp eq i32 %1, 0             ; <i1> [#uses=1]
        br i1 %toBool, label %bb1, label %bb2

bb1:            ; preds = %entry
        %b1.i = bitcast %struct.b* %b.i to i8*          ; <i8*> [#uses=1]
        %2 = call i32 @fn4(i32 %flag, i8* %b1.i) nounwind               ; <i32>
[#uses=1]
        %c2.i = bitcast %struct.b* %c.i to i8*          ; <i8*> [#uses=1]
        %3 = call i32 @fn4(i32 %flag, i8* %c2.i) nounwind               ; <i32>
[#uses=1]
        %4 = add i32 %3, %2             ; <i32> [#uses=1]
        ret i32 %4

bb2:            ; preds = %entry
        ret i32 0
}

declare i32 @fn3(i32, i8*)

declare i32 @fn4(i32, i8*)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 13:06:03 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 13:06:03 -0600
Subject: [LLVMbugs] [Bug 3154] Assertion `I->contains(Start) &&
 I->contains(End-1) && " Range is not entirely in interval!"' failed in llc
In-Reply-To: <bug-3154-206@http.llvm.org/bugs/>
Message-ID: <200901201906.n0KJ63tV028878@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3154


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Evan Cheng <evan.cheng at apple.com>  2009-01-20 13:06:02 ---
Unfortunately some of us have added responsibilities at work. I would welcome
someone else to be working in this area.

This is passing for me now though. Please check.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 14:06:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 14:06:21 -0600
Subject: [LLVMbugs] [Bug 3362] New: Front page link to "front-end
	development" broken
Message-ID: <bug-3362-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3362

           Summary: Front page link to "front-end development" broken
           Product: Documentation
           Version: 2.4
          Platform: PC
               URL: http://llvm.org/
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: General docs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rmann at latencyzero.com
                CC: llvmbugs at cs.uiuc.edu


It references "http://llvm.org/docs/Stacker.html", which, according to bug
#2259, is no longer a part of the main tree.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 14:14:18 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 14:14:18 -0600
Subject: [LLVMbugs] [Bug 3363] New: 2008-07-29-EHLabel.ll is non-portable
Message-ID: <bug-3363-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3363

           Summary: 2008-07-29-EHLabel.ll is non-portable
           Product: Test Suite
           Version: trunk
          Platform: PC
        OS/Version: other
            Status: NEW
          Severity: normal
          Priority: P2
         Component: DejaGNU
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: mrs at apple.com
                CC: llvmbugs at cs.uiuc.edu


Running as isn't strictly portable to systems that default to x86-64 codegen.

  test/CodeGen/Generic/2008-07-29-EHLabel.ll

is the only testcase that appears to do this.  I get:

FAIL:
/Volumes/mrs5/net/clang/clang/clang/test/CodeGen/Generic/2008-07-29-EHLabel.ll
for PR2609

Doing diffs in test:
--- test/CodeGen/Generic/2008-07-29-EHLabel.ll.~1~      2009-01-02
11:47:41.000000000 -0800
+++ test/CodeGen/Generic/2008-07-29-EHLabel.ll  2009-01-20 11:59:27.000000000
-0800
@@ -1,4 +1,4 @@
-; RUN: llvm-as < %s | llc -o - | as -o /dev/null
+; RUN: llvm-as < %s | llc -o - | as -arch i386 -o /dev/null
 ; PR2609
        %struct..0._11 = type { i32 }
        %struct..1__pthread_mutex_s = type { i32, i32, i32, i32, i32,
%struct..0._11 }
--------------

fixes it for me, but -arch i386 might not be as portable as I'd hope.  Can
someone try that on linux say, and if it works there, we can drive it that way,
otherwise, we need to call a driver with more beef.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 15:16:54 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 15:16:54 -0600
Subject: [LLVMbugs] [Bug 3364] New: getPointerToGlobal missing in C API
Message-ID: <bug-3364-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3364

           Summary: getPointerToGlobal missing in C API
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


The function getPointerToGlobal is missing in C API.

Patch included to add it.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 15:25:33 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 15:25:33 -0600
Subject: [LLVMbugs] [Bug 3243] null ptr dereference
In-Reply-To: <bug-3243-206@http.llvm.org/bugs/>
Message-ID: <200901202125.n0KLPXG9001214@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3243


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Evan Cheng <evan.cheng at apple.com>  2009-01-20 15:25:32 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072428.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 15:37:40 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 15:37:40 -0600
Subject: [LLVMbugs] [Bug 3362] Front page link to "front-end development"
	broken
In-Reply-To: <bug-3362-206@http.llvm.org/bugs/>
Message-ID: <200901202137.n0KLbeKX001607@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3362


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-20 15:37:40 ---
fixed, thanks


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 15:44:25 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 15:44:25 -0600
Subject: [LLVMbugs] [Bug 3363] 2008-07-29-EHLabel.ll is non-portable
In-Reply-To: <bug-3363-206@http.llvm.org/bugs/>
Message-ID: <200901202144.n0KLiP7c001891@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3363


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 15:51:24 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 15:51:24 -0600
Subject: [LLVMbugs] [Bug 2191] missed optimization: b &= ~a
In-Reply-To: <bug-2191-206@http.llvm.org/bugs/>
Message-ID: <200901202151.n0KLpOAE002136@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2191


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #16 from Evan Cheng <evan.cheng at apple.com>  2009-01-20 15:51:24 ---
This has been fixed a while back.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 16:06:06 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 16:06:06 -0600
Subject: [LLVMbugs] [Bug 2074] (X86?) Backend generates duplicate labels
In-Reply-To: <bug-2074-206@http.llvm.org/bugs/>
Message-ID: <200901202206.n0KM66Dp002783@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2074


Evan Cheng <evan.cheng at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |WONTFIX




--- Comment #7 from Evan Cheng <evan.cheng at apple.com>  2009-01-20 16:06:06 ---
Mark this WONTFIX now that debug info has changed significantly.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 16:09:26 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 16:09:26 -0600
Subject: [LLVMbugs] [Bug 2172] Dwarf emitter assumes all functions have
	debug info
In-Reply-To: <bug-2172-206@http.llvm.org/bugs/>
Message-ID: <200901202209.n0KM9Qje002930@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2172


Devang Patel <dpatel at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Devang Patel <dpatel at apple.com>  2009-01-20 16:09:25 ---
This seems to work now.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 20 18:29:36 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 Jan 2009 18:29:36 -0600
Subject: [LLVMbugs] [Bug 3365] New: getSourceRange() doesn't contain full
	DeclStmt
Message-ID: <bug-3365-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3365

           Summary: getSourceRange() doesn't contain full DeclStmt
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: AST
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: alexei.svitkine at gmail.com
                CC: llvmbugs at cs.uiuc.edu


When the following line is parsed (in some context):

int i = 0, *p = (int *) 0x100, c;

The result is:

(DeclStmt 0x1d06f80
  0x1d06e70 "int i =
    (IntegerLiteral 0x1d06ea0 'int' 0)"  0x1d06ee0 "int *p =
    (CStyleCastExpr 0x1d06f30 'int *'
      (IntegerLiteral 0x1d06f10 'int' 256))"  0x1d06f50 "int c"

That is, the DeclStmt contains the three nested Decls. However,
getSourceRange() spans only the first 'int' portion of the line, which seems
incorrect to me.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 05:23:53 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 05:23:53 -0600
Subject: [LLVMbugs] [Bug 3366] New: llc segfault in CheckForPhysRegDependency
Message-ID: <bug-3366-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3366

           Summary: llc segfault in CheckForPhysRegDependency
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2421)
 --> (http://llvm.org/bugs/attachment.cgi?id=2421)
testcase .ll

Several Ada testcases are failing due to a segfault in llc.  The
reduced testcase (attached) is pretty simple.  Backtrace:

_ZL25CheckForPhysRegDependencyPN4llvm6SDNodeES1_jPKNS_18TargetRegisterInfoEPKNS_15TargetInstrInfoERjRi
(ScheduleDAGSDNodes.cpp:57)                                                     
llvm::ScheduleDAGSDNodes::AddSchedEdges() (ScheduleDAGSDNodes.cpp:188)          
llvm::ScheduleDAGSDNodes::BuildSchedGraph() (ScheduleDAGSDNodes.cpp:213)        
(anonymous namespace)::ScheduleDAGRRList::Schedule()
(ScheduleDAGRRList.cpp:172)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 05:36:17 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 05:36:17 -0600
Subject: [LLVMbugs] [Bug 3367] New: assertion failure: Cannot find callsite
	to remove!
Message-ID: <bug-3367-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3367

           Summary: assertion failure: Cannot find callsite to remove!
           Product: libraries
           Version: 2.4
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Interprocedural Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: jay.foad at antixlabs.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2422)
 --> (http://llvm.org/bugs/attachment.cgi?id=2422)
bitcode file

With the attached bitcode file I get:

$ opt -f -o /dev/null stripped.bc -inline -prune-eh
opt: /home/foad/svn/llvm-project/llvm/trunk/lib/Analysis/IPA/CallGraph.cpp:264:
void llvm::CallGraphNode::removeCallEdgeFor(llvm::CallSite): Assertion `i &&
"Cannot find callsite to remove!"' failed.
/home/foad/llvm/objdir-svn/Debug/bin/opt[0x86d2b66]
/home/foad/llvm/objdir-svn/Debug/bin/opt[0x86d2c96]
[0xb7fa5400]
/lib/i686/cmov/libc.so.6(abort+0x188)[0xb7d26018]
/lib/i686/cmov/libc.so.6(__assert_fail+0xee)[0xb7d1d5be]
/home/foad/llvm/objdir-svn/Debug/bin/opt(_ZN4llvm13CallGraphNode17removeCallEdgeForENS_8CallSiteE+0x41)[0x8561d41]
/home/foad/llvm/objdir-svn/Debug/bin/opt[0x83f8e31]
/home/foad/llvm/objdir-svn/Debug/bin/opt[0x83f9239]
/home/foad/llvm/objdir-svn/Debug/bin/opt[0x83f9506]
/home/foad/llvm/objdir-svn/Debug/bin/opt(_ZN105_GLOBAL__N__home_foad_svn_llvm_project_llvm_trunk_lib_Analysis_IPA_CallGraphSCCPass.cpp_00000000_EE43971E13CGPassManager11runOnModuleERN4llvm6ModuleE+0x161)[0x8564bb7]
/home/foad/llvm/objdir-svn/Debug/bin/opt(_ZN4llvm13MPPassManager11runOnModuleERNS_6ModuleE+0xcd)[0x865f461]
/home/foad/llvm/objdir-svn/Debug/bin/opt(_ZN4llvm15PassManagerImpl3runERNS_6ModuleE+0x74)[0x8660df6]
/home/foad/llvm/objdir-svn/Debug/bin/opt(_ZN4llvm11PassManager3runERNS_6ModuleE+0x1b)[0x8660e4d]
/home/foad/llvm/objdir-svn/Debug/bin/opt(main+0x11e3)[0x839ce81]
/lib/i686/cmov/libc.so.6(__libc_start_main+0xe5)[0xb7d0f455]
/home/foad/llvm/objdir-svn/Debug/bin/opt(__gxx_personality_v0+0x355)[0x838d361]
Aborted

This happens with both the 2.4 release and a fairly recent head of svn build.
I'm using Linux/x86.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 06:50:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 06:50:21 -0600
Subject: [LLVMbugs] [Bug 3368] New: can't build with just the C backend
Message-ID: <bug-3368-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3368

           Summary: can't build with just the C backend
           Product: tools
           Version: 2.4
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: jay.foad at antixlabs.com
                CC: llvmbugs at cs.uiuc.edu


On Linux/x86 I did:

$ ../llvm-2.4/configure --enable-targets=cbe ; make

It fails with:

llvm[2]: Linking Release executable llc (without symbols)
/home/foad/llvm/objdir-poo/tools/llc/Release/llc.o: In function `global
constructors keyed to FileType':
llc.cpp:(.text+0x93d): undefined reference to
`llvm::createBURRListDAGScheduler(llvm::SelectionDAGISel*, llvm::SelectionDAG*,
llvm::MachineBasicBlock*, bool)'
llc.cpp:(.text+0x961): undefined reference to
`llvm::createTDRRListDAGScheduler(llvm::SelectionDAGISel*, llvm::SelectionDAG*,
llvm::MachineBasicBlock*, bool)'
llc.cpp:(.text+0x985): undefined reference to
`llvm::createTDListDAGScheduler(llvm::SelectionDAGISel*, llvm::SelectionDAG*,
llvm::MachineBasicBlock*, bool)'
llc.cpp:(.text+0x9a9): undefined reference to
`llvm::createFastDAGScheduler(llvm::SelectionDAGISel*, llvm::SelectionDAG*,
llvm::MachineBasicBlock*, bool)'
llc.cpp:(.text+0x9cd): undefined reference to
`llvm::createDefaultScheduler(llvm::SelectionDAGISel*, llvm::SelectionDAG*,
llvm::MachineBasicBlock*, bool)'
llc.cpp:(.text+0x9d7): undefined reference to `llvm::linkOcamlGCPrinter()'
collect2: ld returned 1 exit status
make[2]: *** [/home/foad/llvm/objdir-poo/Release//bin/llc] Error 1

I would find it useful to be able to build llc without any of the native
codegen backends.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 07:08:08 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 07:08:08 -0600
Subject: [LLVMbugs] [Bug 3356] [2.5] using llvm-config --libs results in
	missing symbols
In-Reply-To: <bug-3356-206@http.llvm.org/bugs/>
Message-ID: <200901211308.n0LD88Eb010958@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3356


Mikhail Glushenkov <foldr at codedgers.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Mikhail Glushenkov <foldr at codedgers.com>  2009-01-21 07:08:07 ---
Fixed in:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072472.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 09:18:47 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 09:18:47 -0600
Subject: [LLVMbugs] [Bug 3366] llc segfault in CheckForPhysRegDependency
In-Reply-To: <bug-3366-206@http.llvm.org/bugs/>
Message-ID: <200901211518.n0LFIldD015005@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3366


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |gohman at apple.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Dan Gohman <gohman at apple.com>  2009-01-21 09:18:46 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072474.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 09:19:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 09:19:27 -0600
Subject: [LLVMbugs] [Bug 3317] APint error during x86 lowering
In-Reply-To: <bug-3317-206@http.llvm.org/bugs/>
Message-ID: <200901211519.n0LFJRsV015034@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3317


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Dan Gohman <gohman at apple.com>  2009-01-21 09:19:27 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072475.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 10:08:31 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 10:08:31 -0600
Subject: [LLVMbugs] [Bug 3369] New: math error at all optimization levels
Message-ID: <bug-3369-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3369

           Summary: math error at all optimization levels
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


Using r62644 on Ubuntu Hardy on x86 I get func_1() returning 1 at all
optimization levels.  I believe it should return -1.  The safe mod macro is
just a macroized version of the safe signed mod function from here:

https://www.securecoding.cert.org/confluence/display/seccode/INT32-C.+Ensure+that+operations+on+signed+integers+do+not+result+in+overflow

#include <stdint.h>
#include <stdio.h>

#define safe_mod_macro_int_s_s(si1,si2) \
  (((((int)(si2)) == ((int)0)) || ((((int)(si1)) == (INT32_MIN)) &&
(((int)(si2)) == ((int)-1)))) \
  ? ((int)(si1)) \
  : (((int)(si1)) % ((int)(si2))))

int func_1 (void);
int func_1 (void)
{
  unsigned int l_174 = -7;
  return safe_mod_macro_int_s_s (l_174, 2);
}

int
main (void)
{  
  printf ("checksum = %d\n", func_1 ());
  return 0;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 12:11:41 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 12:11:41 -0600
Subject: [LLVMbugs] [Bug 3364] getPointerToGlobal missing in C API
In-Reply-To: <bug-3364-206@http.llvm.org/bugs/>
Message-ID: <200901211811.n0LIBfau021591@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3364


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-21 12:11:37 ---
Applied, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072481.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 12:20:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 12:20:27 -0600
Subject: [LLVMbugs] [Bug 2648] Scalarrepl and inliner produce messy IR when
 returning small structs
In-Reply-To: <bug-2648-206@http.llvm.org/bugs/>
Message-ID: <200901211820.n0LIKRPx022023@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2648


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Chris Lattner <clattner at apple.com>  2009-01-21 12:20:26 ---
I think this is fixed, if not, please file new bugs for any individual issues
that remain.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 21 18:53:04 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Jan 2009 18:53:04 -0600
Subject: [LLVMbugs] [Bug 3190] VECTOR_SAVE_INLINE not defined for
 other-than-darwin targets
In-Reply-To: <bug-3190-206@http.llvm.org/bugs/>
Message-ID: <200901220053.n0M0r4Qd005030@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3190


Dale Johannesen <dalej at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Dale Johannesen <dalej at apple.com>  2009-01-21 18:52:54 ---
I've defined it to 1 , haven't checked anything except that it still builds on
Darwin.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 00:06:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 00:06:21 -0600
Subject: [LLVMbugs] [Bug 3370] New: Eliminate the vicmp and vfcmp
	instructions
Message-ID: <bug-3370-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3370

           Summary: Eliminate the vicmp and vfcmp instructions
           Product: libraries
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


There is general consensus that vicmp/vfcmp should be eliminated and code
should use icmp/fcmp instead.  We need to get rid of v[if]cmp and move all
vector codegen to work with [if]cmp.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 01:04:43 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 01:04:43 -0600
Subject: [LLVMbugs] [Bug 3358] invalid use of iterators in gvn
In-Reply-To: <bug-3358-206@http.llvm.org/bugs/>
Message-ID: <200901220704.n0M74hcY018415@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3358


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #9 from Chris Lattner <clattner at apple.com>  2009-01-22 01:04:43 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072525.html

I will request this to be pulled into 2.5


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 08:05:11 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 08:05:11 -0600
Subject: [LLVMbugs] [Bug 3371] New: clang rejects pointer arithmetic code (
 from the Linux kernel)
Message-ID: <bug-3371-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3371

           Summary: clang rejects pointer arithmetic code (from the Linux
                    kernel)
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


$ cat /tmp/bug4.c
void
sys_init_module() {
  extern int (*init)(void);

  if (init+0 <= init) {
    ;
  }
}
$ gcc -W -Wall -c /tmp/bug4.c   
$ clang /tmp/bug4.c 
/tmp/bug4.c:6:11: error: arithmetic on pointer to function type 'int (*)(void)'
  if (init+0 <= init) {
      ~~~~^
1 diagnostic generated.
$


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 09:02:50 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 09:02:50 -0600
Subject: [LLVMbugs] [Bug 3372] New: .ll parser error when referencing '@0'
Message-ID: <bug-3372-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3372

           Summary: .ll parser error when referencing '@0'
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM assembly language parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicolas.geoffray at lip6.fr
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2425)
 --> (http://llvm.org/bugs/attachment.cgi?id=2425)
test case

There seems to be a bug in the .ll parser when referencing '@0'. The attached
test case sends an error when running with llvm-as:

test.ll:2:17: forward reference and definition of global have different types

However, if you change the '@0' reference to a '@1' reference (thus referencing
the next global variable) in the 'Try' variable, it works.

(And to make sure '@0' is the first unnamed global variable, changing '@0' to
'@2' returns an error: "use of undefined value '@2'")


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 10:52:02 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 10:52:02 -0600
Subject: [LLVMbugs] [Bug 3373] New: frontend doesn't cast inline asm
 arguments, thus failing with an unsupported asm message
Message-ID: <bug-3373-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3373

           Summary: frontend doesn't cast inline asm arguments, thus failing
                    with an unsupported asm message
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Using yesterday's llvm-gcc (SVN r62706), I fail to build the following:

int foo(unsigned long *addr)
{
  int __ret_pu;
  unsigned long __pu_val;
  return   ({asm volatile("call __put_user_" "8" : "=a" (__ret_pu) :"0"
                          (__pu_val), "c"(addr) : "ebx"); __ret_pu;});
}

$ llvm-gcc x.i -O0 -emit-llvm -o x.bc
x.i: In function ???foo???:
x.i:5: error: unsupported inline asm: input constraint with a matching output
constraint of incompatible type!

I thought the frontend was supposed to cast the argument to the proper type to
avoid this error message.

LLVM 2.4 accepted this code, and appears to be the kernel people's favourite
way to use inline asm (i.e. return type is int, but input is something else).

Can this be fixed for 2.5 please?


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 12:26:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 12:26:35 -0600
Subject: [LLVMbugs] [Bug 3374] New: -ffast-math x86 isel crash
Message-ID: <bug-3374-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3374

           Summary: -ffast-math x86 isel crash
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


llc -enable-unsafe-fp-math crashes on the attached .bc file:

Cannot yet select: 0x14c39d8: f64 = fneg 0x14c4638


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 15:44:38 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 15:44:38 -0600
Subject: [LLVMbugs] [Bug 3375] New: opt crash with use-after-free in MemDep
Message-ID: <bug-3375-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3375

           Summary: opt crash with use-after-free in MemDep
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2428)
 --> (http://llvm.org/bugs/attachment.cgi?id=2428)
bugpoint reduced testcase

With TOT opt crashes when optimizing clamscan:
bugpoint --enable-valgrind in MultiSource/Applications/Clamav:

If you just use 'make bugpoint-opt' it'll show a crash in simplifycfg, or
lcssa, etc. valgrind shows the problem in memdep.

$ /home/edwin/llvm-svn/llvm/Release/bin/bugpoint x.bc -gvn -memdep -memcpyopt
-sccp -append-exit-code -Xlinker=-lz  -input=/dev/null
-output=Output/clamscan.out-nat -timeout=500 -mlimit=0 --enable-valgrind

Checking for crash with only these blocks: bb68 bb58 bb3.outer bb11 bb6 bb2
bb84 bb26 bb114 bb242... <191 total>: ^C==11910== Invalid read of size 8
==11910==    at 0x6B69E9:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6EFA:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDependency(llvm::Value*,
bool, llvm::BasicBlock*, llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*,
llvm::MemDepResult> >&) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x54A85B: (anonymous
namespace)::GVN::processNonLocalLoad(llvm::LoadInst*,
llvm::SmallVectorImpl<llvm::Instruction*>&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x54C020: (anonymous
namespace)::GVN::iterateOnFunction(llvm::Function&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x54E352: (anonymous
namespace)::GVN::runOnFunction(llvm::Function&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x76E26A: llvm::FPPassManager::runOnFunction(llvm::Function&)
(in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x76E4A5: llvm::FPPassManager::runOnModule(llvm::Module&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x76DE46: llvm::MPPassManager::runOnModule(llvm::Module&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x76E045: llvm::PassManagerImpl::run(llvm::Module&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x4AF0CA:
llvm::BugDriver::runPassesAsChild(std::vector<llvm::PassInfo const*,
std::allocator<llvm::PassInfo const*> > const&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x494F98: llvm::BugDriver::run() (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x4B8865: main (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==  Address 0x650f788 is 1,464 bytes inside a block of size 2,560 free'd
==11910==    at 0x4C2111D: operator delete(void*) (vg_replace_malloc.c:342)
==11910==    by 0x6BA847: llvm::DenseMap<llvm::PointerIntPair<llvm::Value*, 1u,
bool>, std::pair<llvm::PointerIntPair<llvm::BasicBlock*, 1u, bool>,
std::vector<std::pair<llvm::BasicBlock*, llvm::MemDepResult>,
std::allocator<std::pair<llvm::BasicBlock*, llvm::MemDepResult> > > >,
llvm::DenseMapInfo<llvm::PointerIntPair<llvm::Value*, 1u, bool> >,
llvm::DenseMapInfo<std::pair<llvm::PointerIntPair<llvm::BasicBlock*, 1u, bool>,
std::vector<std::pair<llvm::BasicBlock*, llvm::MemDepResult>,
std::allocator<std::pair<llvm::BasicBlock*, llvm::MemDepResult> > > > >
>::grow(unsigned int) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6BA96A: llvm::DenseMap<llvm::PointerIntPair<llvm::Value*, 1u,
bool>, std::pair<llvm::PointerIntPair<llvm::BasicBlock*, 1u, bool>,
std::vector<std::pair<llvm::BasicBlock*, llvm::MemDepResult>,
std::allocator<std::pair<llvm::BasicBlock*, llvm::MemDepResult> > > >,
llvm::DenseMapInfo<llvm::PointerIntPair<llvm::Value*, 1u, bool> >,
llvm::DenseMapInfo<std::pair<llvm::PointerIntPair<llvm::BasicBlock*, 1u, bool>,
std::vector<std::pair<llvm::BasicBlock*, llvm::MemDepResult>,
std::allocator<std::pair<llvm::BasicBlock*, llvm::MemDepResult> > > > >
>::operator[](llvm::PointerIntPair<llvm::Value*, 1u, bool> const&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6300:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==
==11910== Invalid read of size 8
==11910==    at 0x6B69ED:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6EFA:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDependency(llvm::Value*,
bool, llvm::BasicBlock*, llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*,
llvm::MemDepResult> >&) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x54A85B: (anonymous
namespace)::GVN::processNonLocalLoad(llvm::LoadInst*,
llvm::SmallVectorImpl<llvm::Instruction*>&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x54C020: (anonymous
namespace)::GVN::iterateOnFunction(llvm::Function&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x54E352: (anonymous
namespace)::GVN::runOnFunction(llvm::Function&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x76E26A: llvm::FPPassManager::runOnFunction(llvm::Function&)
(in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x76E4A5: llvm::FPPassManager::runOnModule(llvm::Module&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x76DE46: llvm::MPPassManager::runOnModule(llvm::Module&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x76E045: llvm::PassManagerImpl::run(llvm::Module&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x4AF0CA:
llvm::BugDriver::runPassesAsChild(std::vector<llvm::PassInfo const*,
std::allocator<llvm::PassInfo const*> > const&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x494F98: llvm::BugDriver::run() (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x4B8865: main (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==  Address 0x650f780 is 1,456 bytes inside a block of size 2,560 free'd
==11910==    at 0x4C2111D: operator delete(void*) (vg_replace_malloc.c:342)
==11910==    by 0x6BA847: llvm::DenseMap<llvm::PointerIntPair<llvm::Value*, 1u,
bool>, std::pair<llvm::PointerIntPair<llvm::BasicBlock*, 1u, bool>,
std::vector<std::pair<llvm::BasicBlock*, llvm::MemDepResult>,
std::allocator<std::pair<llvm::BasicBlock*, llvm::MemDepResult> > > >,
llvm::DenseMapInfo<llvm::PointerIntPair<llvm::Value*, 1u, bool> >,
llvm::DenseMapInfo<std::pair<llvm::PointerIntPair<llvm::BasicBlock*, 1u, bool>,
std::vector<std::pair<llvm::BasicBlock*, llvm::MemDepResult>,
std::allocator<std::pair<llvm::BasicBlock*, llvm::MemDepResult> > > > >
>::grow(unsigned int) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6BA96A: llvm::DenseMap<llvm::PointerIntPair<llvm::Value*, 1u,
bool>, std::pair<llvm::PointerIntPair<llvm::BasicBlock*, 1u, bool>,
std::vector<std::pair<llvm::BasicBlock*, llvm::MemDepResult>,
std::allocator<std::pair<llvm::BasicBlock*, llvm::MemDepResult> > > >,
llvm::DenseMapInfo<llvm::PointerIntPair<llvm::Value*, 1u, bool> >,
llvm::DenseMapInfo<std::pair<llvm::PointerIntPair<llvm::BasicBlock*, 1u, bool>,
std::vector<std::pair<llvm::BasicBlock*, llvm::MemDepResult>,
std::allocator<std::pair<llvm::BasicBlock*, llvm::MemDepResult> > > > >
>::operator[](llvm::PointerIntPair<llvm::Value*, 1u, bool> const&) (in
/home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6300:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
==11910==    by 0x6B6A49:
llvm::MemoryDependenceAnalysis::getNonLocalPointerDepFromBB(llvm::Value*,
unsigned long, bool, llvm::BasicBlock*,
llvm::SmallVectorImpl<std::pair<llvm::BasicBlock*, llvm::MemDepResult> >&,
llvm::DenseMap<llvm::BasicBlock*, llvm::Value*,
llvm::DenseMapInfo<llvm::BasicBlock*>, llvm::DenseMapInfo<llvm::Value*> >&,
bool) (in /home/edwin/llvm-svn/llvm/Release/bin/bugpoint)
Exited with error code '1'


*** Reduction Interrupted, cleaning up...

Emitted bitcode to 'bugpoint-reduced-simplified.bc'

*** You can reproduce the problem with: opt bugpoint-reduced-simplified.bc -gvn


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 16:01:31 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 16:01:31 -0600
Subject: [LLVMbugs] [Bug 3374] [2.5] -ffast-math x86 isel crash
In-Reply-To: <bug-3374-206@http.llvm.org/bugs/>
Message-ID: <200901222201.n0MM1Vfh028366@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3374


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Dan Gohman <gohman at apple.com>  2009-01-22 16:01:30 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072553.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 17:45:15 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 17:45:15 -0600
Subject: [LLVMbugs] [Bug 2749] sema support for designated initializers
In-Reply-To: <bug-2749-206@http.llvm.org/bugs/>
Message-ID: <200901222345.n0MNjFf7031942@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2749


Douglas Gregor <dgregor at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #14 from Douglas Gregor <dgregor at apple.com>  2009-01-22 17:45:14 ---
As of r62812, we have complete semantic analysis for C99 designated
initializers.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 18:27:24 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 18:27:24 -0600
Subject: [LLVMbugs] [Bug 3375] opt crash with use-after-free in MemDep
In-Reply-To: <bug-3375-206@http.llvm.org/bugs/>
Message-ID: <200901230027.n0N0ROqp001022@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3375


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Chris Lattner <clattner at apple.com>  2009-01-22 18:27:23 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072569.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 18:38:28 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 18:38:28 -0600
Subject: [LLVMbugs] [Bug 3371] clang rejects pointer arithmetic code (from
 the Linux kernel)
In-Reply-To: <bug-3371-206@http.llvm.org/bugs/>
Message-ID: <200901230038.n0N0cSgW001370@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3371


Douglas Gregor <dgregor at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dgregor at apple.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Douglas Gregor <dgregor at apple.com>  2009-01-22 18:38:28 ---
As of r62823, Clang accepts pointer arithmetic on function pointers in C mode
as a GNU extension.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 18:44:03 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 18:44:03 -0600
Subject: [LLVMbugs] [Bug 3376] New: Cannot yet select: 0x411b508c: f64 =
	fneg 0x411b5950
Message-ID: <bug-3376-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3376

           Summary: Cannot yet select: 0x411b508c: f64 = fneg 0x411b5950
           Product: new-bugs
           Version: unspecified
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: howarth at bromo.msbb.uc.edu
                CC: llvmbugs at cs.uiuc.edu


The gfortran compiler in current llvm/llvm-gcc-4.2 ICEs when compiling the
xtarget.f sourcefile of xplor-nih with -O3 -ffast-math. The exact -ffast-math
option causing the ICE appears to be -ffinite-math-only as -ffast-math
-fno-finite-math-only compiles without ICEs. The subroutine causing the ICE
with...

lvm-gfortran -c xtarget_reduced.f -O3 -ffast-math -ffixed-line-length-120
-fno-second-underscore
Cannot yet select: 0x411b508c: f64 = fneg 0x411b5950
xtarget_reduced.f:156: internal compiler error: Abort trap
Please submit a full bug report,
with preprocessed source if appropriate.

is attached a xtarget-reduced.f.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 22 22:22:57 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Jan 2009 22:22:57 -0600
Subject: [LLVMbugs] [Bug 3214] Exception handling doesn't preserve
 callee-saved registers
In-Reply-To: <bug-3214-206@http.llvm.org/bugs/>
Message-ID: <200901230422.n0N4MvAc008321@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3214


Patrick Walton <pcwalton at cs.ucla.edu> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pcwalton at cs.ucla.edu
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




--- Comment #3 from Patrick Walton <pcwalton at cs.ucla.edu>  2009-01-22 22:22:55 ---
Not an LLVM bug. Turns out this was a bug in the D runtime.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 03:50:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 03:50:20 -0600
Subject: [LLVMbugs] [Bug 3377] New: Assertion `BitWidth == RHS.BitWidth && "
 Comparison requires equal bit widths"' failed
Message-ID: <bug-3377-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3377

           Summary: Assertion `BitWidth == RHS.BitWidth && "Comparison
                    requires equal bit widths"' failed
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


This bug affects r62839.  The reduced testcase comes from linux-2.4.37.


$ cat /tmp/bug10.c 
typedef unsigned long long u64;

struct tg3_ethtool_stats {
 u64 rx_octets;
};


struct {
 char string[32];
} ethtool_stats_keys[(sizeof(struct tg3_ethtool_stats)/sizeof(u64))] = {
 { "rx_octets" },
};
$ gcc -c -W -Wall /tmp/bug10.c 
$ ~/eclair/eclair/src/parser/dest/bin/clang -w /tmp/bug10.c  
clang: /home/roberto/parser/src/include/llvm/ADT/APInt.h:793: bool
llvm::APInt::operator==(const llvm::APInt&) const: Assertion `BitWidth ==
RHS.BitWidth && "Comparison requires equal bit widths"' failed.
0   clang     0x0000000000d87641
1   libc.so.6 0x0000003e33c32f90
2   libc.so.6 0x0000003e33c32f05 gsignal + 53
3   libc.so.6 0x0000003e33c34a73 abort + 387
4   libc.so.6 0x0000003e33c2bef9 __assert_fail + 233
5   clang     0x000000000076cda2
6   clang     0x00000000008312f5
clang::InitListChecker::CheckArrayType(clang::InitListExpr*, clang::QualType&,
llvm::APSInt, bool, unsigned int&) + 1141
7   clang     0x0000000000830354
clang::InitListChecker::CheckListElementTypes(clang::InitListExpr*,
clang::QualType&, bool, unsigned int&) + 852
8   clang     0x0000000000830a1d
clang::InitListChecker::CheckExplicitInitList(clang::InitListExpr*,
clang::QualType&, unsigned int&) + 93
9   clang     0x0000000000831689
clang::InitListChecker::InitListChecker(clang::Sema*, clang::InitListExpr*,
clang::QualType&) + 41
10  clang     0x00000000007f45f5
clang::Sema::CheckInitializerTypes(clang::Expr*&, clang::QualType&,
clang::SourceLocation, clang::DeclarationName, bool) + 725
11  clang     0x00000000007f4b7d clang::Sema::AddInitializerToDecl(void*,
clang::ASTOwningPtr<&(clang::ActionBase::DeleteExpr(void*))>, bool) + 733
12  clang     0x00000000007f4c06 clang::Sema::AddInitializerToDecl(void*,
clang::ASTOwningPtr<&(clang::ActionBase::DeleteExpr(void*))>) + 38
13  clang     0x00000000008c83fd
clang::Parser::ParseInitDeclaratorListAfterFirstDeclarator(clang::Declarator&)
+ 1341
14  clang     0x00000000008c3b8b
clang::Parser::ParseDeclarationOrFunctionDefinition(llvm::SmallVector<void*,
4u>*) + 1339
15  clang     0x00000000008c3f48 clang::Parser::ParseExternalDeclaration() +
104
16  clang     0x00000000008c4386 clang::Parser::ParseTopLevelDecl(void*&) + 22
17  clang     0x00000000007e52db clang::ParseAST(clang::Preprocessor&,
clang::ASTConsumer*, bool, bool) + 219
18  clang     0x0000000000589ad4
19  clang     0x000000000058ea29 main + 1417
20  libc.so.6 0x0000003e33c1e576 __libc_start_main + 230
21  clang     0x000000000054d3e9
Aborted
$


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 04:43:22 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 04:43:22 -0600
Subject: [LLVMbugs] [Bug 3378] New: Assertion `IsUnsigned == RHS.IsUnsigned
 && " Signedness mismatch!"' failed
Message-ID: <bug-3378-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3378

           Summary: Assertion `IsUnsigned == RHS.IsUnsigned && "Signedness
                    mismatch!"' failed
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


This bug affects r62839.  The reduced testcase comes from linux-2.4.37.

$ cat /tmp/bug11.c
typedef unsigned int brlock_read_lock_t;

enum brlock_indices {
 BR_GLOBALIRQ_LOCK,
 BR_NETPROTO_LOCK,

 __BR_END
};

brlock_read_lock_t
__brlock_array[(4)][(((sizeof(brlock_read_lock_t)*__BR_END
                       + (1 << ((7)))-1) & ~((1 << ((7)))-1))
                     / sizeof(brlock_read_lock_t))] = {
  [0 ... (4)-1] = {
    [0 ... (((sizeof(brlock_read_lock_t)*__BR_END +
              (1 << ((7)))-1) & ~((1 << ((7)))-1))
            / sizeof(brlock_read_lock_t))-1] = 0
  }
};
$ gcc -c -W -Wall /tmp/bug11.c
$ clang -w /tmp/bug11.c  
clang: /home/roberto/parser/src/include/llvm/ADT/APSInt.h:121: bool
llvm::APSInt::operator<(const llvm::APSInt&) const: Assertion `IsUnsigned ==
RHS.IsUnsigned && "Signedness mismatch!"' failed.
0   clang     0x0000000000d87641
1   libc.so.6 0x0000003e33c32f90
2   libc.so.6 0x0000003e33c32f05 gsignal + 53
3   libc.so.6 0x0000003e33c34a73 abort + 387
4   libc.so.6 0x0000003e33c2bef9 __assert_fail + 233
5   clang     0x000000000082e193
6   clang     0x000000000082e5ff
clang::Sema::ActOnDesignatedInitializer(clang::Designation&,
clang::SourceLocation, bool,
clang::ASTOwningResult<&(clang::ActionBase::DeleteExpr(void*))>) + 1071
7   clang     0x00000000008ddea7
clang::Parser::ParseInitializerWithPotentialDesignator(clang::InitListDesignations&,
unsigned int) + 4551
8   clang     0x00000000008de67e clang::Parser::ParseBraceInitializer() + 702
9   clang     0x00000000008df5c8 clang::Parser::ParseInitializer() + 72
10  clang     0x00000000008dde59
clang::Parser::ParseInitializerWithPotentialDesignator(clang::InitListDesignations&,
unsigned int) + 4473
11  clang     0x00000000008de67e clang::Parser::ParseBraceInitializer() + 702
12  clang     0x00000000008c8418
clang::Parser::ParseInitDeclaratorListAfterFirstDeclarator(clang::Declarator&)
+ 1368
13  clang     0x00000000008c3b8b
clang::Parser::ParseDeclarationOrFunctionDefinition(llvm::SmallVector<void*,
4u>*) + 1339
14  clang     0x00000000008c3f48 clang::Parser::ParseExternalDeclaration() +
104
15  clang     0x00000000008c4386 clang::Parser::ParseTopLevelDecl(void*&) + 22
16  clang     0x00000000007e52db clang::ParseAST(clang::Preprocessor&,
clang::ASTConsumer*, bool, bool) + 219
17  clang     0x0000000000589ad4
18  clang     0x000000000058ea29 main + 1417
19  libc.so.6 0x0000003e33c1e576 __libc_start_main + 230
20  clang     0x000000000054d3e9
Aborted
$


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 10:54:23 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 10:54:23 -0600
Subject: [LLVMbugs] [Bug 3377] Assertion `BitWidth == RHS.BitWidth && "
 Comparison requires equal bit widths"' failed
In-Reply-To: <bug-3377-206@http.llvm.org/bugs/>
Message-ID: <200901231654.n0NGsNOd010836@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3377


Douglas Gregor <dgregor at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from Douglas Gregor <dgregor at apple.com>  2009-01-23 10:54:23 ---
Fixed in r62851.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 10:56:11 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 10:56:11 -0600
Subject: [LLVMbugs] [Bug 3267] extra symbol in llvm-gcc output
In-Reply-To: <bug-3267-206@http.llvm.org/bugs/>
Message-ID: <200901231656.n0NGuBcW010915@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3267


Rafael ??vila de Esp??ndola <rafael.espindola at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #20 from Rafael ??vila de Esp??ndola <rafael.espindola at gmail.com>  2009-01-23 10:56:10 ---
Fixed in revision 62850


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 15:04:33 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 15:04:33 -0600
Subject: [LLVMbugs] [Bug 3378] Assertion `IsUnsigned == RHS.IsUnsigned &&
 "Signedness mismatch! "' failed
In-Reply-To: <bug-3378-206@http.llvm.org/bugs/>
Message-ID: <200901232104.n0NL4X7I022949@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3378


Douglas Gregor <dgregor at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dgregor at apple.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Douglas Gregor <dgregor at apple.com>  2009-01-23 15:04:32 ---
Fixed in r62876.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 16:01:24 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 16:01:24 -0600
Subject: [LLVMbugs] [Bug 3379] New: Invalid operand found in inline asm:
 'movl %gs:${1:P}, $0'
Message-ID: <bug-3379-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3379

           Summary: Invalid operand found in inline asm: 'movl
                    %gs:${1:P},$0'
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Another one from the kernel

typedef unsigned long long u64;
    extern __typeof__(int) per_cpu__cpu_number;
    static u64 __attribute__((__section__(".data.read_mostly")))
boot_pat_state;
    void pat_init(void) {
    u64 pat;
    printk("<6>" "x86 PAT enabled: cpu %d, old 0x%Lx, new 0x%Lx\n",         (({
  typeof(per_cpu__cpu_number) ret__;
  switch (sizeof(per_cpu__cpu_number)) {
 case 4: asm("mov" "l ""%%""gs"":%P" "1"",%0" : "=r" (ret__) : "m"
(per_cpu__cpu_number));
 }
  ret__;
  }
  )), boot_pat_state, pat);
   }

$ llvm-gcc testcase-min.i
Invalid operand found in inline asm: 'movl %gs:${1:P},$0'
  INLINEASM <es:movl %gs:${1:P},$0>, 10, %ECX<def>, 36, %reg0, 1, %reg0,
<ga:per_cpu__cpu_number>

With %1 gcc generates this:
  movl %gs:per_cpu__cpu_number,%eax

Without %P gcc generates this:
  movl %gs:per_cpu__cpu_number(%rip),%eax


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 16:31:59 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 16:31:59 -0600
Subject: [LLVMbugs] [Bug 2995] gcc/config/i386//darwin.h should not set
 -Wformat, -Wformat-security, -Wno-format-extra-args,
 -Wno-format-zero-length and -Wno-nonnull
In-Reply-To: <bug-2995-206@http.llvm.org/bugs/>
Message-ID: <200901232231.n0NMVxIj026344@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2995


Jack Howarth <howarth at bromo.msbb.uc.edu> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 16:38:46 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 16:38:46 -0600
Subject: [LLVMbugs] [Bug 3380] New: LangRef doesn't mention backreference
	syntax
Message-ID: <bug-3380-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3380

           Summary: LangRef doesn't mention backreference syntax
           Product: Documentation
           Version: trunk
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: minor
          Priority: P2
         Component: General docs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: skaflotten at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2436)
 --> (http://llvm.org/bugs/attachment.cgi?id=2436)
New section describing recursive structures

The LangRef documentation does not mention the \N* syntax for declaring
recursive structure types.

The attached patch adds a short explanation with suggestions from nlewycky.

Thanks,
Stein Roger


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 23 19:03:57 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Jan 2009 19:03:57 -0600
Subject: [LLVMbugs] [Bug 3237] llvm-debug breaks on enums for which
	TREE_CHAIN is null
In-Reply-To: <bug-3237-206@http.llvm.org/bugs/>
Message-ID: <200901240103.n0O13vXD031405@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3237


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #7 from Duncan Sands <baldrick at free.fr>  2009-01-23 19:03:57 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072597.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072604.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 04:13:12 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 04:13:12 -0600
Subject: [LLVMbugs] [Bug 3381] New: instcombine: DerivedTypes.h:243: const
 llvm::Type* llvm:: StructType::getElementType(unsigned int) const:
 Assertion `N < NumContainedTys && "Element number out of range!"' failed
Message-ID: <bug-3381-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3381

           Summary: instcombine: DerivedTypes.h:243: const llvm::Type*
                    llvm::StructType::getElementType(unsigned int) const:
                    Assertion `N < NumContainedTys && "Element number out of
                    range!"' failed
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2437)
 --> (http://llvm.org/bugs/attachment.cgi?id=2437)
bugpoint-reduced-simplified.bc

Another testcase from the kernel:
$ ~/llvm-bootstrap/install/bin/opt -instcombine bugpoint-reduced-simplified.bc
opt: /home/edwin/llvm-bootstrap/llvm/include/llvm/DerivedTypes.h:243: const
llvm::Type* llvm::StructType::getElementType(unsigned int) const: Assertion `N
< NumContainedTys && "Element number out of range!"' failed.
0   opt       0x0000000000783b8f
1   libc.so.6 0x00007f7124d6cf60
2   libc.so.6 0x00007f7124d6ced5 gsignal + 53
3   libc.so.6 0x00007f7124d6e3f3 abort + 387
4   libc.so.6 0x00007f7124d65dc9 __assert_fail + 233
5   opt       0x000000000052d068
6   opt       0x000000000053bd96
7   opt       0x00000000005517a5
8   opt       0x0000000000552dab
9   opt       0x00000000005534be
10  opt       0x00000000007216bb
llvm::FPPassManager::runOnFunction(llvm::Function&) + 347
11  opt       0x00000000007218f6
llvm::FPPassManager::runOnModule(llvm::Module&) + 166
12  opt       0x0000000000721297
llvm::MPPassManager::runOnModule(llvm::Module&) + 199
13  opt       0x0000000000721496 llvm::PassManagerImpl::run(llvm::Module&) +
134
14  opt       0x0000000000497176 main + 3046
15  libc.so.6 0x00007f7124d591a6 __libc_start_main + 230
16  opt       0x000000000048be49 sinh + 65
Aborted


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 04:19:44 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 04:19:44 -0600
Subject: [LLVMbugs] [Bug 3382] New: clang rejects parentheses around
	initializer
Message-ID: <bug-3382-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3382

           Summary: clang rejects parentheses around initializer
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


This comes from linux-2.6.28.1:

$ cat /tmp/bug23.c 
#define N_(text) (text)

char nohelp_text[] = N_("Sorry, no help available for this option yet.\n");
$ gcc -c -W -Wall /tmp/bug23.c
$ icc -c -w2 /tmp/bug23.c 
$ clang -w /tmp/bug23.c
/tmp/bug23.c:3:22: error: initialization with '{...}' expected for array
char nohelp_text[] = N_("Sorry, no help available for this option yet.\n");
                     ^~
1 diagnostic generated.
$ ~/eclair/eclair/src/parser/dest/bin/clang -E /tmp/bug23.c
# 1 "/tmp/bug23.c"
# 1 "/tmp/bug23.c" 1
# 0 "/tmp/bug23.c"
# 1 "<predefines>" 1
# 1 "/tmp/bug23.c" 2


char nohelp_text[] = ("Sorry, no help available for this option yet.\n");
$


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 04:32:07 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 04:32:07 -0600
Subject: [LLVMbugs] [Bug 3383] New: Assertion failed: Only trivial
 vector-to-scalar conversions should get here!
Message-ID: <bug-3383-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3383

           Summary: Assertion failed: Only trivial vector-to-scalar
                    conversions should get here!
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


With LLVM 2.4, generating assembly (x86) for this generates a failed assertion:

define <4 x i1> @fun(<4 x float>) {
_L1:
        %1 = extractelement <4 x float> %0, i32 0
        %2 = fcmp olt float %1, 0.000000e+00
        %3 = insertelement <4 x i1> undef, i1 %2, i32 0
        %4 = extractelement <4 x float> %0, i32 1
        %5 = fcmp olt float %4, 0.000000e+00
        %6 = insertelement <4 x i1> %3, i1 %5, i32 1
        %7 = extractelement <4 x float> %0, i32 2
        %8 = fcmp olt float %7, 0.000000e+00
        %9 = insertelement <4 x i1> %6, i1 %8, i32 2
        %10 = extractelement <4 x float> %0, i32 3
        %11 = fcmp olt float %10, 0.000000e+00
        %12 = insertelement <4 x i1> %9, i1 %11, i32 3
        ret <4 x i1> %12
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 04:37:55 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 04:37:55 -0600
Subject: [LLVMbugs] [Bug 3384] New: Assertion failed: It's impossible to
	expand bools
Message-ID: <bug-3384-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3384

           Summary: Assertion failed: It's impossible to expand bools
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


With LLVM 2.4, generating code (x86) for the following generates a failed
assertion:

define <4 x float> @fun(<4 x float>) {
_L1:
        %1 = extractelement <4 x float> %0, i32 0
        %2 = fcmp olt float %1, 0.000000e+00
        %3 = insertelement <4 x i1> undef, i1 %2, i32 0
        %4 = extractelement <4 x float> %0, i32 1
        %5 = fcmp olt float %4, 0.000000e+00
        %6 = insertelement <4 x i1> %3, i1 %5, i32 1
        %7 = extractelement <4 x float> %0, i32 2
        %8 = fcmp olt float %7, 0.000000e+00
        %9 = insertelement <4 x i1> %6, i1 %8, i32 2
        %10 = extractelement <4 x float> %0, i32 3
        %11 = fcmp olt float %10, 0.000000e+00
        %12 = insertelement <4 x i1> %9, i1 %11, i32 3
        %13 = sub <4 x float> < float -0.000000e+00, float -0.000000e+00, float
-0.000000e+00, float -0.000000e+00 >, %0
        %14 = select <4 x i1> %12, <4 x float> %13, <4 x float> %0
        ret <4 x float> %14
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 04:53:32 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 04:53:32 -0600
Subject: [LLVMbugs] [Bug 3385] New: clang rejects input constraint 'er' in
	asm
Message-ID: <bug-3385-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3385

           Summary: clang rejects input constraint 'er' in asm
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


This comes from linux-2.6.28.1:

$ cat /tmp/bug27.c 
typedef struct {
        long counter;
} atomic64_t;

inline void atomic64_sub(long i, atomic64_t *v)
{
        asm volatile("subq %1,%0"
                     : "=m" (v->counter)
                     : "er" (i), "m" (v->counter));
}
$ gcc -c -W -Wall  /tmp/bug27.
$ icc -c -Wall  /tmp/bug27.c
$ clang /tmp/bug27.c
/tmp/bug27.c:9:24: error: invalid input constraint 'er' in asm
                     : "er" (i), "m" (v->counter));
                       ^
1 diagnostic generated.
$


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 05:22:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 05:22:27 -0600
Subject: [LLVMbugs] [Bug 3386] New: clang rejects a sensible application of
	__alignof__
Message-ID: <bug-3386-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3386

           Summary: clang rejects a sensible application of __alignof__
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: bagnara at cs.unipr.it
                CC: llvmbugs at cs.uiuc.edu, bolzoni at cs.unipr.it


This comes from linux-2.6.28.1:

$ cat /tmp/bug28.c 
struct crypto_tfm {
        int crt_flags;
        void *__crt_ctx[];
};

inline unsigned int crypto_tfm_ctx_alignment(void)
{
        struct crypto_tfm *tfm;
        return __alignof__(tfm->__crt_ctx);
}
$ gcc -c -W -Wall /tmp/bug28.c 
$ icc -c -Wall /tmp/bug28.c 
/tmp/bug28.c(9): warning #1549: alignment-of operator applied to incomplete
type
          return __alignof__(tfm->__crt_ctx);
                 ^

$ clang /tmp/bug28.c
/tmp/bug28.c:9:16: error: invalid application of '__alignof' to an incomplete
type 'void *[]'
        return __alignof__(tfm->__crt_ctx);
               ^          ~~~~~~~~~~~~~~~~
1 diagnostic generated.
$ 

(I miss the point of the warning given by the Intel C compiler, by the way.)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 05:44:41 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 05:44:41 -0600
Subject: [LLVMbugs] [Bug 3387] New: ICE when ENABLE_EXPENSIVE_CHECKS is
	enabled
Message-ID: <bug-3387-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3387

           Summary: ICE when ENABLE_EXPENSIVE_CHECKS is enabled
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: wendling at apple.com
                CC: llvmbugs at cs.uiuc.edu


$ ./llvm.obj/Debug+Checks/bin/llc -o - bugpoint-reduced-simplified.bc 
/usr/include/c++/4.0.0/bits/stl_heap.h:326:error: elements in iterator 
    range [__first, __last) do not form a heap with respect to the
    predicate __comp.

Objects involved in the operation:
iterator "__first" @ 0x0xbffff0ec {
type =
N11__gnu_debug14_Safe_iteratorIN9__gnu_cxx17__normal_iteratorIPPN4llvm5SUnitEN10__gnu_norm6vectorIS5_SaIS5_EEEEE\
N15__gnu_debug_def6vectorIS5_S9_EEEE (mutable iterator);
  state = dereferenceable (start-of-sequence);
  references sequence with type
`N15__gnu_debug_def6vectorIPN4llvm5SUnitESaIS3_EEE' @ 0x0xbffff0ec
}
iterator "__last" @ 0x0xbffff0d8 {
type =
N11__gnu_debug14_Safe_iteratorIN9__gnu_cxx17__normal_iteratorIPPN4llvm5SUnitEN10__gnu_norm6vectorIS5_SaIS5_EEEEE\
N15__gnu_debug_def6vectorIS5_S9_EEEE (mutable iterator);
  state = past-the-end;
  references sequence with type
`N15__gnu_debug_def6vectorIPN4llvm5SUnitESaIS3_EEE' @ 0x0xbffff0d8
}
0   llc               0x00a4c1dc char const* std::find<char const*, char>(char
const*, char const*, char const&) + 296
1   llc               0x00a4c98d llvm::sys::RemoveFileOnSignal(llvm::sys::Path
const&, std::string*) + 517
2   libSystem.B.dylib 0x924522bb _sigtramp + 43
3   libSystem.B.dylib 0xffffffff _sigtramp + 1840962927
4   libSystem.B.dylib 0x924c623a raise + 26
5   libSystem.B.dylib 0x924d2679 abort + 73
6   libstdc++.6.dylib 0x9417e6f9 __gnu_debug::_Error_formatter::_M_error()
const + 465
7   llc               0x005b73bd void
std::pop_heap<__gnu_debug::_Safe_iterator<__gnu_cxx::__normal_iterator<llvm::SUni\
t**, __gnu_norm::vector<llvm::SUnit*, std::allocator<llvm::SUnit*> > >,
__gnu_debug_def::vector<llvm::SUnit*, std::allo\
cator<llvm::SUnit*> > >, (anonymous
namespace)::bu_ls_rr_sort>(__gnu_debug::_Safe_iterator<__gnu_cxx::__normal_iterator\
<llvm::SUnit**, __gnu_norm::vector<llvm::SUnit*, std::allocator<llvm::SUnit*> >
>, __gnu_debug_def::vector<llvm::SUnit*\
, std::allocator<llvm::SUnit*> > >,
__gnu_debug::_Safe_iterator<__gnu_cxx::__normal_iterator<llvm::SUnit**,
__gnu_norm:\
:vector<llvm::SUnit*, std::allocator<llvm::SUnit*> > >,
__gnu_debug_def::vector<llvm::SUnit*, std::allocator<llvm::SUni\
t*> > >, (anonymous namespace)::bu_ls_rr_sort) + 701
8   llc               0x005b75a2 std::priority_queue<llvm::SUnit*,
__gnu_debug_def::vector<llvm::SUnit*, std::allocator\
<llvm::SUnit*> >, (anonymous namespace)::bu_ls_rr_sort>::pop() + 244
9   llc               0x005b77ca llvm::PriorityQueue<llvm::SUnit*,
__gnu_debug_def::vector<llvm::SUnit*, std::allocator\
<llvm::SUnit*> >, (anonymous namespace)::bu_ls_rr_sort>::erase_one(llvm::SUnit*
const&) + 514
10  llc               0x005b0d78 (anonymous
namespace)::bu_ls_rr_sort::operator()(llvm::SUnit const*, llvm::SUnit const\
*) const + 9562
11  llc               0x005b14ff (anonymous
namespace)::bu_ls_rr_sort::operator()(llvm::SUnit const*, llvm::SUnit const\
*) const + 11489
12  llc               0x007a1cf4 llvm::ScheduleDAG::Run(llvm::SelectionDAG*,
llvm::MachineBasicBlock*, llvm::ilist_iter\
ator<llvm::MachineInstr>, llvm::ilist_iterator<llvm::MachineInstr>) + 186
13  llc               0x0062e665 llvm::SelectionDAGISel::Schedule() + 155
14  llc               0x00630e04 llvm::SelectionDAGISel::CodeGenAndEmitDAG() +
3276
15  llc               0x006329f9
llvm::SelectionDAGISel::SelectBasicBlock(llvm::BasicBlock*,
llvm::ilist_iterator<llvm:\
:Instruction>, llvm::ilist_iterator<llvm::Instruction>) + 575
16  llc               0x00633233
llvm::SelectionDAGISel::SelectAllBasicBlocks(llvm::Function&,
llvm::MachineFunction&, \
llvm::MachineModuleInfo*, llvm::DwarfWriter*, llvm::TargetInstrInfo const&) +
2085
17  llc               0x00633603
llvm::SelectionDAGISel::runOnFunction(llvm::Function&) + 893
18  llc               0x0099c5b8
llvm::FPPassManager::runOnFunction(llvm::Function&) + 270
19  llc               0x0099c80e
llvm::FunctionPassManagerImpl::run(llvm::Function&) + 116
20  llc               0x0099c994
llvm::FunctionPassManager::run(llvm::Function&) + 156
21  llc               0x000037dd main + 2639
22  llc               0x00002536 start + 54
23  llc               0x00000004 start + 18446744073709542148


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 06:46:17 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 06:46:17 -0600
Subject: [LLVMbugs] [Bug 3388] New: Unsupported bith width
Message-ID: <bug-3388-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3388

           Summary: Unsupported bith width
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


So I guess some bit widths don't work.  I quote from the LLVM language
reference "Any bit width from 1 bit to 2^23-1 (about 8 million) can be
specified.".
Is there's a difference between something that can be specified and something
that actually works?

Simple example failing on LLVM 2.4, x86:

define i320 @fun(i320, i320) {
_L1:
        %2 = add i320 %1, %0
        ret i320 %2
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 06:53:44 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 06:53:44 -0600
Subject: [LLVMbugs] [Bug 3389] New: Assertion failed: Invalid truncate node,
	src < dst!
Message-ID: <bug-3389-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3389

           Summary: Assertion failed: Invalid truncate node, src < dst!
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


LLVM 2.4, x86 code generation:

define i33 @fun(i33, i33) {
_L1:
        %2 = udiv i33 %1, %0
        ret i33 %2
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 07:21:23 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 07:21:23 -0600
Subject: [LLVMbugs] [Bug 3388] Unsupported bith width
In-Reply-To: <bug-3388-206@http.llvm.org/bugs/>
Message-ID: <200901241321.n0ODLNEq031692@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3388


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |baldrick at free.fr
             Status|NEW                         |RESOLVED
         Resolution|                            |WORKSFORME




--- Comment #1 from Duncan Sands <baldrick at free.fr>  2009-01-24 07:21:20 ---
The documentation refers to current svn, not the
2.4 release.  This works fine with llvm from svn.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 07:22:04 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 07:22:04 -0600
Subject: [LLVMbugs] [Bug 3389] Assertion failed: Invalid truncate node,
	src < dst!
In-Reply-To: <bug-3389-206@http.llvm.org/bugs/>
Message-ID: <200901241322.n0ODM4ZF031731@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3389


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |baldrick at free.fr
             Status|NEW                         |RESOLVED
         Resolution|                            |WORKSFORME




--- Comment #1 from Duncan Sands <baldrick at free.fr>  2009-01-24 07:22:03 ---
The documentation refers to current svn, not the
2.4 release.  This works fine with llvm from svn.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 07:26:00 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 07:26:00 -0600
Subject: [LLVMbugs] [Bug 3388] Unsupported bith width
In-Reply-To: <bug-3388-206@http.llvm.org/bugs/>
Message-ID: <200901241326.n0ODQ0pN031873@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3388


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|WORKSFORME                  |DUPLICATE




--- Comment #2 from Duncan Sands <baldrick at free.fr>  2009-01-24 07:25:59 ---
Sorry, that comment was for PR3389!  This bug is a duplicate of PR2660.

*** This bug has been marked as a duplicate of bug 2660 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 08:47:23 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 08:47:23 -0600
Subject: [LLVMbugs] [Bug 3390] New: -m32 inline asm: Do not know how to
 expand the result of this operator
Message-ID: <bug-3390-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3390

           Summary: -m32 inline asm: Do not know how to expand the result of
                    this operator
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


After my attempt to fix the kernel to use same types for input/output regs, I
get a failure on 32-bit:

Am I doing something wrong?

$ llvm-gcc testcase-min.i -m32
ExpandIntegerResult #0: 0x28ce838: i64 = Register  RCX
cc1:
/home/edwin/llvm-bootstrap/llvm/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp:930:
void llvm::DAGTypeLegalizer::ExpandIntegerResult(llvm::SDNode*, unsigned int):
Assertion `0 && "Do not know how to expand the result of this operator!"'
failed.
testcase-min.i:17: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.

C reduced testcase below, bytecode attached:
/* testcase */
typedef unsigned int __kernel_size_t;
typedef long long __kernel_loff_t;
typedef __kernel_loff_t loff_t;
typedef __kernel_size_t size_t;
__attribute__ ((regparm (0)))
     long sys_sendfile64 (int out_fd, int in_fd, loff_t * offset,
                          size_t count)
{
  if (offset)
    {
      if (__builtin_expect (!!((
                                 {
                                 __typeof__ (*(offset)) __ret_pu;
                                 __typeof__ (*(offset)) __pu_val;
                                 switch (sizeof (*(offset)))
                                 {
case 8:
      asm volatile ("call __put_user_8": "=a" (__ret_pu): "A" ((typeof
(*(offset))) (__pu_val)), "c" (offset):"ebx");}
                                 (int) __ret_pu;}
                               )), 0))
        return -14;
    }
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 09:46:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 09:46:21 -0600
Subject: [LLVMbugs] [Bug 3391] New: llvm-gcc -m32 exponential memory use /
	llc segfault
Message-ID: <bug-3391-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3391

           Summary: llvm-gcc -m32 exponential memory use / llc segfault
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2440)
 --> (http://llvm.org/bugs/attachment.cgi?id=2440)
testcase-min.i

$ (ulimit -v 1024768; llvm-gcc -m32 testcase-min.i -O1)
testcase-min.i:50: warning: excess elements in struct initializer
testcase-min.i:50: warning: (near initialization for ???pci_indirect???)
testcase-min.i:50: warning: excess elements in struct initializer
testcase-min.i:50: warning: (near initialization for ???pci_indirect???)
llvm-gcc: Internal error: Segmentation fault (program cc1)
Please submit a full bug report.
See <URL:http://developer.apple.com/bugreporter> for instructions

Without ulimit it fills entire mem.

Bugpoint reduced testcase below, preprocessed C attached:
; ModuleID = 'bugpoint-reduced-simplified.bc'
target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
target triple = "i386-unknown-linux-gnu"
@pci_indirect = external global { }             ; <{ }*> [#uses=1]
@pcibios_last_bus = external global i32         ; <i32*> [#uses=2]

define void @pci_pcbios_init() nounwind section ".init.text" {
entry:
        br label %bb1.i

bb1.i:          ; preds = %bb6.i.i, %bb1.i, %entry
        %0 = load i32* null, align 8            ; <i32> [#uses=1]
        %1 = icmp ugt i32 %0, 1048575           ; <i1> [#uses=1]
        br i1 %1, label %bb2.i, label %bb1.i

bb2.i:          ; preds = %bb1.i
        %asmtmp.i.i = tail call { i32, i32, i32, i32 } asm "lcall *(%edi);
cld\0A\09jc 1f\0A\09xor %ah, %ah\0A1:",
"={dx},={ax},={bx},={cx},1,{di},~{dirflag},~{fpsr},~{flags},~{memory}"(i32
45313, { }* @pci_indirect) nounwind             ; <{ i32, i32, i32, i32 }>
[#uses=2]
        %asmresult2.i.i = extractvalue { i32, i32, i32, i32 } %asmtmp.i.i, 1   
        ; <i32> [#uses=1]
        %2 = lshr i32 %asmresult2.i.i, 8                ; <i32> [#uses=1]
        %3 = trunc i32 %2 to i8         ; <i8> [#uses=1]
        %4 = load i32* @pcibios_last_bus, align 4               ; <i32>
[#uses=1]
        %5 = icmp slt i32 %4, 0         ; <i1> [#uses=1]
        br i1 %5, label %bb5.i.i, label %bb6.i.i

bb5.i.i:                ; preds = %bb2.i
        %asmresult4.i.i = extractvalue { i32, i32, i32, i32 } %asmtmp.i.i, 3   
        ; <i32> [#uses=1]
        %6 = and i32 %asmresult4.i.i, 255               ; <i32> [#uses=1]
        store i32 %6, i32* @pcibios_last_bus, align 4
        br label %bb6.i.i

bb6.i.i:                ; preds = %bb5.i.i, %bb2.i
        %7 = icmp eq i8 %3, 0           ; <i1> [#uses=1]
        %or.cond.i.i = and i1 %7, false         ; <i1> [#uses=1]
        br i1 %or.cond.i.i, label %bb1.i, label %bb8.i.i

bb8.i.i:                ; preds = %bb6.i.i
        unreachable
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 11:48:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 11:48:21 -0600
Subject: [LLVMbugs] [Bug 3382] clang rejects parentheses around initializer
In-Reply-To: <bug-3382-206@http.llvm.org/bugs/>
Message-ID: <200901241748.n0OHmLZN007557@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3382


Anders Carlsson <andersca at mac.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |andersca at mac.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Anders Carlsson <andersca at mac.com>  2009-01-24 11:48:20 ---
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090119/011271.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 12:04:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 12:04:35 -0600
Subject: [LLVMbugs] [Bug 3385] clang rejects input constraint 'er' in asm
In-Reply-To: <bug-3385-206@http.llvm.org/bugs/>
Message-ID: <200901241804.n0OI4ZSi008184@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3385


Anders Carlsson <andersca at mac.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |andersca at mac.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Anders Carlsson <andersca at mac.com>  2009-01-24 12:04:35 ---
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090119/011272.html

Z was already added.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 13:27:53 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 13:27:53 -0600
Subject: [LLVMbugs] [Bug 3381] instcombine: DerivedTypes.h:243: const
 llvm::Type* llvm:: StructType::getElementType(unsigned int) const:
 Assertion `N < NumContainedTys && "Element number out of range!"' failed
In-Reply-To: <bug-3381-206@http.llvm.org/bugs/>
Message-ID: <200901241927.n0OJRrcs011044@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3381


T??r??k Edwin <edwintorok at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #4 from T??r??k Edwin <edwintorok at gmail.com>  2009-01-24 13:27:53 ---
This was fixed, and testcase is added also.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 14:52:34 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 14:52:34 -0600
Subject: [LLVMbugs] [Bug 3393] New: llc abort with
	-enable-legalize-types-checking
Message-ID: <bug-3393-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3393

           Summary: llc abort with -enable-legalize-types-checking
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: gohman at apple.com
                CC: llvmbugs at cs.uiuc.edu


The following testcase gets an llc abort on x86-32 with
-enable-legalize-types-checking enabled.

define void @foo(i32 inreg %x) {
        %t709 = select i1 false, i32 0, i32 %x          ; <i32> [#uses=1]
        %t711 = add i32 %t709, 1                ; <i32> [#uses=4]
        %t801 = icmp slt i32 %t711, 0           ; <i1> [#uses=1]
        %t712 = zext i32 %t711 to i64           ; <i64> [#uses=1]
        %t804 = select i1 %t801, i64 0, i64 %t712       ; <i64> [#uses=1]
        store i64 %t804, i64* null
        %t815 = icmp slt i32 %t711, 0           ; <i1> [#uses=1]
        %t814 = sext i32 %t711 to i64           ; <i64> [#uses=1]
        %t816 = select i1 %t815, i64 0, i64 %t814               ; <i64>
[#uses=1]
        store i64 %t816, i64* null
        unreachable
}

llc: lib/CodeGen/SelectionDAG/LegalizeTypes.cpp:86: void
llvm::DAGTypeLegalizer::PerformExpensiveChecks(): Assertion `UI->getNodeId() ==
NewNode && "Remapped value has non-trivial use!"' failed.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 15:09:04 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 15:09:04 -0600
Subject: [LLVMbugs] [Bug 3394] New: introduce -fconserve-stack flag
Message-ID: <bug-3394-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3394

           Summary: introduce -fconserve-stack flag
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


gcc-4.4 has a -fconserve-stack flag, and even though gcc-4.2 doesn't support
it,
we could introduce it, and convey the flag's status to the LLVM inliner.
It could then switch to use another policy, that also takes stack usage into
consideration.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 15:51:36 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 15:51:36 -0600
Subject: [LLVMbugs] [Bug 3395] New: ccc: -include not working
Message-ID: <bug-3395-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3395

           Summary: ccc: -include not working
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Neither of these work:

$ /home/edwin/llvm-svn/llvm/tools/clang/utils/ccc -include ./x.h x.c
<predefines>:95:10: error: './x.h' file not found
#include "./x.h"
         ^
1 diagnostic generated

$ /home/edwin/llvm-svn/llvm/tools/clang/utils/ccc -include x.h x.c
<predefines>:95:10: error: 'x.h' file not found
#include "x.h"
         ^
1 diagnostic generated

Adding -I. works, but -include shouldn't need -I

FWIW, this causes ccc-analyzer to say this on linux-2.6:
<predefines>:102:10: error: './include/linux/autoconf.h' file not found


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 15:55:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 15:55:27 -0600
Subject: [LLVMbugs] [Bug 3386] clang rejects a sensible application of
	__alignof__
In-Reply-To: <bug-3386-206@http.llvm.org/bugs/>
Message-ID: <200901242155.n0OLtRl4015796@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3386


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-24 15:55:26 ---
This works now.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 16:36:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 16:36:20 -0600
Subject: [LLVMbugs] [Bug 3396] New: Crash in Sema with non-constant offsetof
Message-ID: <bug-3396-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3396

           Summary: Crash in Sema with non-constant offsetof
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:

struct sockaddr_un {
 unsigned char sun_len;
 char sun_path[104];
};
int a() {
int a[__builtin_offsetof(struct sockaddr_un, sun_path[len+1])];
}

Crashes with the following message:
clang:
/home/eli/llvm/tools/clang/lib/CodeGen/../../include/clang/AST/Expr.h:169:
llvm::APSInt clang::Expr::getIntegerConstantExprValue(clang::ASTContext&)
const: Assertion `success && "Illegal argument to getIntegerConstantExpr"'
failed.

Note that non-constant offsetof is a gcc extension.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 16:40:10 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 16:40:10 -0600
Subject: [LLVMbugs] [Bug 2910] clang codegen doesn't handle non-constant
	offsetof
In-Reply-To: <bug-2910-206@http.llvm.org/bugs/>
Message-ID: <200901242240.n0OMeAFu017559@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2910


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-24 16:40:09 ---
Fixed in r62942, but see also bug 3396.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 16:45:13 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 16:45:13 -0600
Subject: [LLVMbugs] [Bug 3301] x(x){} confuses clang.
In-Reply-To: <bug-3301-206@http.llvm.org/bugs/>
Message-ID: <200901242245.n0OMjD9C017723@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3301


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-24 16:45:13 ---
Fixed in r62849.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 17:09:26 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 17:09:26 -0600
Subject: [LLVMbugs] [Bug 3062] ({}) should be rejected at file scope.
In-Reply-To: <bug-3062-206@http.llvm.org/bugs/>
Message-ID: <200901242309.n0ON9Qkj018490@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3062


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-24 17:09:25 ---
Fixed in r62943.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 17:51:04 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 17:51:04 -0600
Subject: [LLVMbugs] [Bug 2100] Sema doesn't merge array definitions
In-Reply-To: <bug-2100-206@http.llvm.org/bugs/>
Message-ID: <200901242351.n0ONp4Cr019789@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2100


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-24 17:51:03 ---
Fixed in r62947.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 18:28:40 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 18:28:40 -0600
Subject: [LLVMbugs] [Bug 2657] Error when accessing vector component from
 member variable.
In-Reply-To: <bug-2657-206@http.llvm.org/bugs/>
Message-ID: <200901250028.n0P0SeeA021273@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2657


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |WORKSFORME




--- Comment #4 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-24 18:28:40 ---
It seems like this was fixed; we aren't printing an error anymore.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 19:07:48 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 19:07:48 -0600
Subject: [LLVMbugs] [Bug 3397] New: Wide multiplications fail
Message-ID: <bug-3397-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3397

           Summary: Wide multiplications fail
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


The LLVM 2.6svn rev 62947, generating x86 code for

define i1 @fun(i128, i128, i128) {
_L1:
        %3 = mul i128 %1, %0
        %4 = icmp eq i128 %2, %3
        ret i1 %4
}

gives this:
Call result #2 has unhandled type i32


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 20:39:01 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 20:39:01 -0600
Subject: [LLVMbugs] [Bug 3398] New: IRGen crash with pointer cast to int in
	static var
Message-ID: <bug-3398-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3398

           Summary: IRGen crash with pointer cast to int in static var
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
int a() {static unsigned b = (unsigned)a;}

Result with "clang -emit-llvm":
clang: /home/eli/llvm/tools/clang/lib/CodeGen/CGDecl.cpp:93: llvm::GlobalValue*
clang::CodeGen::CodeGenFunction::GenerateStaticBlockVarDecl(const
clang::VarDecl&, bool, const char*): Assertion
`getContext().getLangOptions().CPlusPlus && "only C++ supports non-constant
static initializers!"' failed.

This bug is of a similar nature to a couple I've been committing fixes for:
roughly, if Sema::CheckForConstantInitializer thinks an initializer for a
static local variable is constant, but Expr::isConstantInitializer disagrees,
we hit this assertion in CodeGen.

I think this is basically the last significant case where this disagreement
exists; the reason I'm filing it rather than fixing it myself is that I'm not
sure what the best way to fix this particular bug is.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 24 23:24:58 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Jan 2009 23:24:58 -0600
Subject: [LLVMbugs] [Bug 3399] New: Assertion `S1->getType() ==
 S2->getType() && " Cannot create binary operator with two operands of
 differing type!"' failed .
Message-ID: <bug-3399-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3399

           Summary: Assertion `S1->getType() == S2->getType() && "Cannot
                    create binary operator with two operands of differing
                    type!"' failed.
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


Seen using r62920 on Ubuntu Hardy on x86.

regehr at john-home:~/volatile/tmp131$ llvm-gcc -O2 small.c
small.c:13: warning: ???bar??? used but never defined
cc1: Instructions.cpp:1575: static llvm::BinaryOperator*
llvm::BinaryOperator::Create(llvm::Instruction::BinaryOps, llvm::Value*,
llvm::Value*, const std::string&, llvm::Instruction*): Assertion `S1->getType()
== S2->getType() && "Cannot create binary operator with two operands of
differing type!"' failed.
small.c:25: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.
regehr at john-home:~/volatile/tmp131$ cat small.c
static short baz (short x, int y) 
{
  return 1 >= 1*8  ? : x;
}

static int biz (int x, int y) 
{
  return (1 & -y) < 0 ? : x - 1;
}

volatile int g_53;

static int bar (void);

int foo (void) 
{
  int l_2;
  for (l_2 = 0; l_2 != 1; l_2 = biz (l_2, 1))                    
    if (bar()) {
      if (baz (0 ? : g_53 * (short)l_2, 1)) 
        for (0; 1; 1);
      return 1;
    }
  return 0;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 01:11:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 01:11:20 -0600
Subject: [LLVMbugs] [Bug 3400] New: code generator bug using __thread
	variables
Message-ID: <bug-3400-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3400

           Summary: code generator bug using __thread variables
           Product: tools
           Version: 2.2
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: bhurt at spnz.org
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2442)
 --> (http://llvm.org/bugs/attachment.cgi?id=2442)
Tar of attachments

Attempting to access global variables declared to be thread-local using
__thread in llvm-gcc causes the code generator to abort with the error message:

Cannot yet select: 0x285c510: i64 = GlobalTLSAddress <i64** @young_limit> 0
temp3.c:19: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.

The original bug is reproduced with the following attached file temp3.c, which
is part of an Ocaml-like parallel garbage collector I'm writing.  Note that if
you remove the __thread keywords from the source code, it compiles without
error, so it almost certainly a problem with the thread local storage code. 
Also attached should be a reduced byte code file, produced by bugpoint.  As
bugzilla doesn't want me to add multiple attachments, I tar'ed and gzip'd the
two files together.

This is LLVM 2.2 (Ubuntu 8.10 default, x86 64-bit).  I haven't tested the
problem with more recent versions of the compiler.  I did both google the bug
and search the bugzilla database.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 03:43:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 03:43:20 -0600
Subject: [LLVMbugs] [Bug 3397] Wide multiplications fail
In-Reply-To: <bug-3397-206@http.llvm.org/bugs/>
Message-ID: <200901250943.n0P9hKCu015989@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3397


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |baldrick at free.fr
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #1 from Duncan Sands <baldrick at free.fr>  2009-01-25 03:43:19 ---
The i128 multiplication gets turned into a libcall
(a call into the gcc library).  This results in a
function call that returns i128.  That this does
not work on x86-32 is PR2660.  It most likely still
will not work once PR2660 is fixed, because the gcc
library for x86-32 does not currently support 128 bit
multiplications.  This is one of the limitations of
codegen of arbitrary precision integers: it relies
on external libraries which do not support arbitrarily
large sizes.

*** This bug has been marked as a duplicate of bug 2660 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 04:51:37 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 04:51:37 -0600
Subject: [LLVMbugs] [Bug 3401] New: Wrong code for i288 addition
Message-ID: <bug-3401-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3401

           Summary: Wrong code for i288 addition
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


LLVM 2.6svn, x86 code generation.

Here's a small test program:
-------
define void @add(i288, i288, i288*) {
_L1:
        %3 = add i288 %0, %1
        store i288 %3, i288* %2
        ret void
}
-------

And some C code to test it:
-------
#include <stdio.h>

#define N 9

struct big { unsigned int x[N]; };

void add(struct big, struct big, struct big *);

main()
{
  struct big a, b, c;
  int i;
  for (i = 0; i < N; i++) {
    a.x[i] = 0x11111111;
    b.x[i] = 0x22222222;
    c.x[i] = 0x5a5a5a5a;
  }
  add(a,b,&c);
  for (i = N-1; i >= 0; i--)
    printf("%08x ", c.x[i]);
  printf("\n");
}
-------

This prints
fffffffe 00000000 00000001 ffffffff ffffffff ffffffff ffffffff fffffffe
33333333 
which is clearly wrong.  Changing i288 to i256 (and N to 8) makes it work.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 06:23:01 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 06:23:01 -0600
Subject: [LLVMbugs] [Bug 3400] code generator bug using __thread variables
In-Reply-To: <bug-3400-206@http.llvm.org/bugs/>
Message-ID: <200901251223.n0PCN1G3013336@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3400


Anton Korobeynikov <asl at math.spbu.ru> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |WORKSFORME




--- Comment #1 from Anton Korobeynikov <asl at math.spbu.ru>  2009-01-25 06:23:00 ---
Works for me on top of the tree. Please consider trying current version before
filling PR.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 07:35:53 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 07:35:53 -0600
Subject: [LLVMbugs] [Bug 3402] New: sse cannot be turned off for x86-64 ->
 linux kernel panics
Message-ID: <bug-3402-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3402

           Summary: sse cannot be turned off for x86-64 -> linux kernel
                    panics
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


[    0.004000] Call Trace:
[    0.004000]  [<ffffffff80212f37>] init_fpu+0x107/0x140
[    0.004000]  [<ffffffff8020d449>] math_state_restore+0x89/0xc0
[    0.004000]  [<ffffffff805950c9>] do_device_not_available+0x9/0x10
[    0.004000]  [<ffffffff8020cfe5>] device_not_available+0x15/0x20
[    0.004000]  [<ffffffff80472c69>] ? uart_set_options+0x9/0xe0
[    0.004000]  [<ffffffff8047557f>] ? uart_parse_options+0x2f/0xa0
[    0.004000]  [<ffffffff80732ee8>] serial8250_console_setup+0xa8/0xc0
[    0.004000]  [<ffffffff80239676>] register_console+0x216/0x300
[    0.004000]  [<ffffffff8073306b>] serial8250_console_init+0x2b/0x30
[    0.004000]  [<ffffffff8073208f>] console_init+0x22/0x36
[    0.004000]  [<ffffffff80713a76>] start_kernel+0x22b/0x3e4
[    0.004000]  [<ffffffff80713284>] x86_64_start_reservations+0x94/0xb9
[    0.004000]  [<ffffffff8071337c>] x86_64_start_kernel+0xd3/0xda
[    0.004000] Code: 00 00 00 00 00 55 48 89 e5 48 83 ec 20 4c 89 6c 24 10 48
89 1c 24 4c 89 64 24 08 4c 89 74 24 18 41 89 f5 48 8b 4d 08 9c 41 5c fa <48> 8b
9f e0 00 00 00 44 8b b7 f8 00 00 00 48 85 db 74 47 8b 87
[    0.004000] RIP  [<ffffffff802b57d6>] kmem_cache_alloc+0x26/0xa0

init_fpu gave me a hint, and looking at the disassembly confirmed it:
ffffffff80472c60 <uart_set_options>:
ffffffff80472c60:       55                      push   %rbp
ffffffff80472c61:       48 89 e5                mov    %rsp,%rbp
ffffffff80472c64:       53                      push   %rbx
ffffffff80472c65:       48 83 ec 38             sub    $0x38,%rsp
ffffffff80472c69:       0f 57 c0                xorps  %xmm0,%xmm0
ffffffff80472c6c:       0f 29 45 d0             movaps %xmm0,-0x30(%rbp)
ffffffff80472c70:       0f 29 45 c0             movaps %xmm0,-0x40(%rbp)

The kernel is built (on purpose) with -mno-sse -mno-mmx -mno-sse2 -mno-3dnow,
however llvm-gcc didn't take that into consideration.

Testcase:
/* llvm-gcc -m64 -O1 -mno-sse -mno-mmx -mno-sse2 -mno-3dnow -S < %s | not grep
xmm0 */
typedef unsigned long __kernel_size_t;
typedef __kernel_size_t size_t;
void *memset(void *s, int c, size_t n);

typedef unsigned char cc_t;
typedef unsigned int speed_t;
typedef unsigned int tcflag_t;

struct ktermios {
 tcflag_t c_iflag;
 tcflag_t c_oflag;
 tcflag_t c_cflag;
 tcflag_t c_lflag;
 cc_t c_line;
 cc_t c_cc[19];
 speed_t c_ispeed;
 speed_t c_ospeed;
};
void bar(struct ktermios*);
void foo()
{
    struct ktermios termios;
    memset(&termios, 0, sizeof(termios));
    bar(&termios);
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 12:48:00 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 12:48:00 -0600
Subject: [LLVMbugs] [Bug 3403] New: llc: "Do not know how to split the
 result of this operator! "
Message-ID: <bug-3403-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3403

           Summary: llc: "Do not know how to split the result of this
                    operator!"
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Due to my recent change for #3402, it is now possible to  use -mattr=-sse on
x86-64.
However in some cases this causes assertion failures in codegen (all ok with
defaults, and -m32):

Testcase: 
$ llc -march=x86-64 x.bc -f -mattr=-sse
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-unknown-linux-gnu"

define void @acpi_os_printf(i8* %fmt, ...) nounwind optsize {
entry:
        ret void
}
llc:
/home/edwin/llvm-bootstrap/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp:358:
void llvm::DAGTypeLegalizer::SplitVectorResult(llvm::SDNode*, unsigned int):
Assertion `0 && "Do not know how to split the result of this operator!"'
failed.
0   llc       0x0000000000c6e99f
1   libc.so.6 0x00007fac0e79df60
2   libc.so.6 0x00007fac0e79ded5 gsignal + 53
3   libc.so.6 0x00007fac0e79f3f3 abort + 387
4   libc.so.6 0x00007fac0e796dc9 __assert_fail + 233
5   llc       0x00000000009e50aa
6   llc       0x00000000009d6c2d
7   llc       0x00000000009d76cf llvm::SelectionDAG::LegalizeTypes() + 847
8   llc       0x0000000000973528 llvm::SelectionDAGISel::CodeGenAndEmitDAG() +
568
9   llc       0x0000000000975612
llvm::SelectionDAGISel::SelectBasicBlock(llvm::BasicBlock*,
llvm::ilist_iterator<llvm::Instruction>,
llvm::ilist_iterator<llvm::Instruction>) + 562
10  llc       0x0000000000975a0c
llvm::SelectionDAGISel::SelectAllBasicBlocks(llvm::Function&,
llvm::MachineFunction&, llvm::MachineModuleInfo*, llvm::DwarfWriter*,
llvm::TargetInstrInfo const&) + 748
11  llc       0x0000000000976f9e
llvm::SelectionDAGISel::runOnFunction(llvm::Function&) + 1214
12  llc       0x0000000000c0a2fb
llvm::FPPassManager::runOnFunction(llvm::Function&) + 347
13  llc       0x0000000000c0a6f6
llvm::FunctionPassManagerImpl::run(llvm::Function&) + 134
14  llc       0x0000000000c0a87e
llvm::FunctionPassManager::run(llvm::Function&) + 62
15  llc       0x00000000004fbc31 main + 4193
16  libc.so.6 0x00007fac0e78a1a6 __libc_start_main + 230
17  llc       0x00000000004fa129 std::ios_base::Init::~Init() + 57
Aborted

Without -mattr=-sse, llc generates this:
acpi_os_printf:
        subq    $184, %rsp
        movaps  %xmm7, 160(%rsp)
        movaps  %xmm6, 144(%rsp)
        movaps  %xmm5, 128(%rsp)
        movaps  %xmm4, 112(%rsp)
        movaps  %xmm3, 96(%rsp)
        movaps  %xmm2, 80(%rsp)
        movaps  %xmm1, 64(%rsp)
        movaps  %xmm0, 48(%rsp)
        movq    %r9, 40(%rsp)
        movq    %r8, 32(%rsp)
        movq    %rcx, 24(%rsp)
        movq    %rdx, 16(%rsp)
        movq    %rsi, 8(%rsp)
        addq    $184, %rsp
        ret

gcc generates this:
$ gcc  x.c -S -o - -m64 -mno-sse
acpi_os_printf:
.LFB2:
        pushq   %rbp
.LCFI0:
        movq    %rsp, %rbp
.LCFI1:
        subq    $72, %rsp
.LCFI2:
        movq    %rsi, -168(%rbp)
        movq    %rdx, -160(%rbp)
        movq    %rcx, -152(%rbp)
        movq    %r8, -144(%rbp)
        movq    %r9, -136(%rbp)
        movq    %rdi, -184(%rbp)
        leave
        ret

Or with -O1:
acpi_os_printf:
.LFB2:
        rep
        ret

To reproduce with llvm-gcc:
/* testcase */
void acpi_os_printf(unsigned char* fmt, ...) {
}

$ llvm-gcc  x.c -S -o - -m64 -mno-sse
        .ident  "GCC: (GNU) 4.2.1 (Based on Apple Inc. build 5636) (LLVM
build)"
SplitVectorResult #0: 0x2772730: v4f32 = Register  #1038
cc1:
/home/edwin/llvm-bootstrap/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp:358:
void llvm::DAGTypeLegalizer::SplitVectorResult(llvm::SDNode*, unsigned int):
Assertion `0 && "Do not know how to split the result of this operator!"'
failed.
x.c:2: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 18:38:16 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 18:38:16 -0600
Subject: [LLVMbugs] [Bug 3405] New: Assertion `FromBr->isUnconditional() &&
 " FromBB should end with uncond br!"' failed.
Message-ID: <bug-3405-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3405

           Summary: Assertion `FromBr->isUnconditional() && "FromBB should
                    end with uncond br!"' failed.
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sliedes at cc.hut.fi
                CC: llvmbugs at cs.uiuc.edu


Running opt -condprop on the attached bitcode file causes an assertion failure.

Revision is today's trunk and architecture is amd64.

---------------------------------------------------------
$ opt -condprop bugpoint-reduced-simplified.bc
opt: CondPropagate.cpp:197:
void<unnamed>::CondProp::RevectorBlockTo(llvm::BasicBlock*, llvm::BasicBlock*):
Assertion `FromBr->isUnconditional() && "FromBB should end with uncond br!"'
failed.
0   opt       0x000000000078440f
1   libc.so.6 0x00007f98a5b7ef60
2   libc.so.6 0x00007f98a5b7eed5 gsignal + 53
3   libc.so.6 0x00007f98a5b803f3 abort + 387
4   libc.so.6 0x00007f98a5b77dc9 __assert_fail + 233
5   opt       0x00000000004f3713
6   opt       0x00000000004f3bb0
7   opt       0x000000000072199b
llvm::FPPassManager::runOnFunction(llvm::Function&) + 347
8   opt       0x0000000000721bd6
llvm::FPPassManager::runOnModule(llvm::Module&) + 166
9   opt       0x0000000000721577
llvm::MPPassManager::runOnModule(llvm::Module&) + 199
10  opt       0x0000000000721776 llvm::PassManagerImpl::run(llvm::Module&) +
134
11  opt       0x0000000000497166 main + 3046
12  libc.so.6 0x00007f98a5b6b1a6 __libc_start_main + 230
13  opt       0x000000000048be39 sinh + 65
Aborted
---------------------------------------------------------

Here's the bitcode disassembled:

---------------------------------------------------------
; ModuleID = 'bugpoint-reduced-simplified.bc'
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-unknown-linux-gnu"

define i32 @main(i32 %argc, i8** %argv) nounwind {
entry:
        br label %bb2

bb2:            ; preds = %bb.bb2_crit_edge, %entry
        br i1 false, label %bb5.thread2, label %bb

bb:             ; preds = %bb2
        br i1 false, label %bb3, label %bb.bb2_crit_edge

bb.bb2_crit_edge:               ; preds = %bb
        br label %bb2

bb3:            ; preds = %bb
        %.lcssa4 = phi i1 [ false, %bb ]                ; <i1> [#uses=1]
        br i1 %.lcssa4, label %bb5.thread, label %bb6

bb5.thread:             ; preds = %bb3
        br label %bb7

bb7:            ; preds = %bb5.thread2, %bb5.thread
        br label %UnifiedReturnBlock

bb6:            ; preds = %bb3
        br label %UnifiedReturnBlock

bb5.thread2:            ; preds = %bb2
        br label %bb7

UnifiedReturnBlock:             ; preds = %bb6, %bb7
        ret i32 0
}
---------------------------------------------------------


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 18:51:08 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 18:51:08 -0600
Subject: [LLVMbugs] [Bug 3406] New: Assertion `0 && "Requiring domfrontiers
 but not idom/ domtree/domset." " not implemented yet!"' failed.
Message-ID: <bug-3406-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3406

           Summary: Assertion `0 && "Requiring domfrontiers but not
                    idom/domtree/domset." " not implemented yet!"' failed.
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sliedes at cc.hut.fi
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2444)
 --> (http://llvm.org/bugs/attachment.cgi?id=2444)
test case

Revision is trunk from today and arch is x86-64.

On running

  opt -loop-index-split -break-crit-edges -scalarrepl -postdomtree

on the attached bitcode, the following assertion failure happens:

------------------------------------------------------------
$ opt bugpoint-reduced-simplified.bc -loop-index-split -break-crit-edges
-scalarrepl -postdomtree
opt: BreakCriticalEdges.cpp:226: bool
llvm::SplitCriticalEdge(llvm::TerminatorInst*, unsigned int, llvm::Pass*,
bool): Assertion `0 && "Requiring domfrontiers but not idom/domtree/domset." "
not implemented yet!"' failed.
0   opt       0x000000000078440f
1   libc.so.6 0x00007f0764a24f60
2   libc.so.6 0x00007f0764a24ed5 gsignal + 53
3   libc.so.6 0x00007f0764a263f3 abort + 387
4   libc.so.6 0x00007f0764a1ddc9 __assert_fail + 233
5   opt       0x00000000005cf646 llvm::SplitCriticalEdge(llvm::TerminatorInst*,
unsigned int, llvm::Pass*, bool) + 5398
6   opt       0x00000000005cf777
7   opt       0x000000000072199b
llvm::FPPassManager::runOnFunction(llvm::Function&) + 347
8   opt       0x0000000000721bd6
llvm::FPPassManager::runOnModule(llvm::Module&) + 166
9   opt       0x0000000000721577
llvm::MPPassManager::runOnModule(llvm::Module&) + 199
10  opt       0x0000000000721776 llvm::PassManagerImpl::run(llvm::Module&) +
134
11  opt       0x0000000000497166 main + 3046
12  libc.so.6 0x00007f0764a111a6 __libc_start_main + 230
13  opt       0x000000000048be39 sinh + 65
Aborted
------------------------------------------------------------


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 19:02:39 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 19:02:39 -0600
Subject: [LLVMbugs] [Bug 3407] New: Assertion `0 && "ERROR: Bad TargetData
 ctor used. " " Tool did not specify a TargetData to use?"' failed.
Message-ID: <bug-3407-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3407

           Summary: Assertion `0 && "ERROR: Bad TargetData ctor used.  "
                    "Tool did not specify a TargetData to use?"' failed.
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Target Description Classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sliedes at cc.hut.fi
                CC: llvmbugs at cs.uiuc.edu


Running

  opt -targetdata

on the attached bitcode causes this assertion failure:

------------------------------------------------------------
$ opt bugpoint-reduced-simplified.bc -targetdata
opt: /home/sliedes/llvm-svn/src/llvm/include/llvm/Target/TargetData.h:114:
llvm::TargetData::TargetData(): Assertion `0 && "ERROR: Bad TargetData ctor
used.  " "Tool did not specify a TargetData to use?"' failed.
0   opt       0x000000000078440f
1   libc.so.6 0x00007f35bd7d8f60
2   libc.so.6 0x00007f35bd7d8ed5 gsignal + 53
3   libc.so.6 0x00007f35bd7da3f3 abort + 387
4   libc.so.6 0x00007f35bd7d1dc9 __assert_fail + 233
5   opt       0x000000000069a85d
6   opt       0x000000000049699a main + 1050
7   libc.so.6 0x00007f35bd7c51a6 __libc_start_main + 230
8   opt       0x000000000048be39 sinh + 65
Aborted
------------------------------------------------------------

Disassembled:

------------------------------------------------------------
; ModuleID = 'bugpoint-reduced-simplified.bc'
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-unknown-linux-gnu"
        %struct.Cls = type { i32, i8, [2 x %struct.Cls*], [2 x %struct.Lit*] }
        %struct.Lit = type { i8 }

define fastcc i32 @cmp_activity(%struct.Cls* %c, %struct.Cls* %d) nounwind {
entry:
        unreachable
}
------------------------------------------------------------


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 19:27:26 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 19:27:26 -0600
Subject: [LLVMbugs] [Bug 3408] New: looprotate: Assertion `index != -1 && "
 Expected incoming value from Original PreHeader"' failed.
Message-ID: <bug-3408-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3408

           Summary: looprotate: Assertion `index != -1 && "Expected incoming
                    value from Original PreHeader"' failed.
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sliedes at cc.hut.fi
                CC: llvmbugs at cs.uiuc.edu


Running

  opt -loop-rotate

on the attached bitcode file causes this assertion failure:

------------------------------------------------------------
$ opt bugpoint-reduced-simplified.bc -loop-rotate
opt: LoopRotation.cpp:473:
void<unnamed>::LoopRotate::preserveCanonicalLoopForm(llvm::LPPassManager&):
Assertion `index != -1 && "Expected incoming value from Original PreHeader"'
failed.
0   opt       0x000000000078440f
1   libc.so.6 0x00007f7ab36e9f60
2   libc.so.6 0x00007f7ab36e9ed5 gsignal + 53
3   libc.so.6 0x00007f7ab36eb3f3 abort + 387
4   libc.so.6 0x00007f7ab36e2dc9 __assert_fail + 233
5   opt       0x0000000000571165
6   opt       0x00000000005725a6
7   opt       0x0000000000572b33
8   opt       0x000000000065dcbb
llvm::LPPassManager::runOnFunction(llvm::Function&) + 955
9   opt       0x000000000072199b
llvm::FPPassManager::runOnFunction(llvm::Function&) + 347
10  opt       0x0000000000721bd6
llvm::FPPassManager::runOnModule(llvm::Module&) + 166
11  opt       0x0000000000721577
llvm::MPPassManager::runOnModule(llvm::Module&) + 199
12  opt       0x0000000000721776 llvm::PassManagerImpl::run(llvm::Module&) +
134
13  opt       0x0000000000497166 main + 3046
14  libc.so.6 0x00007f7ab36d61a6 __libc_start_main + 230
15  opt       0x000000000048be39 sinh + 65
Aborted
------------------------------------------------------------

Here's the bitcode disassembled:

------------------------------------------------------------
; ModuleID = 'bugpoint-reduced-simplified.bc'
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-unknown-linux-gnu"
        %struct.Cls = type { i32, i8, [2 x %struct.Cls*], [2 x %struct.Lit*] }
        %struct.Lit = type { i8 }

define void @picosat_main_bb13.i.i71.outer_bb132.i.i.i.outer(%struct.Cls**,
%struct.Cls**, i32 %collect.i.i.i.1.lcssa, i32 %lcollect.i.i.i.2.lcssa,
%struct.Cls*** %rhead.tmp.0236.out, i32* %collect.i.i.i.2.out, i32*
%lcollect.i.i.i.3.ph.ph.ph.out) nounwind {
newFuncRoot:
        br label %codeRepl

bb133.i.i.i.exitStub:           ; preds = %codeRepl
        ret void

bb130.i.i.i:            ; preds = %codeRepl
        %rhead.tmp.0236.lcssa82 = phi %struct.Cls** [ null, %codeRepl ]        
; <%struct.Cls**> [#uses=0]
        br label %codeRepl

codeRepl:               ; preds = %bb130.i.i.i, %newFuncRoot
        br i1 false, label %bb130.i.i.i, label %bb133.i.i.i.exitStub
}
------------------------------------------------------------


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 19:34:17 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 19:34:17 -0600
Subject: [LLVMbugs] [Bug 3269] clang::CompoundLiteralExpr::Init can be NULL,
 consequently getInitializer() must use cast_or_null.
In-Reply-To: <bug-3269-206@http.llvm.org/bugs/>
Message-ID: <200901260134.n0Q1YHAv008391@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3269


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-25 19:34:16 ---
Fixed in r62981, by eliminating the case where the initializer for a compound
literal could be null.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 20:12:22 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 20:12:22 -0600
Subject: [LLVMbugs] [Bug 3408] looprotate: Assertion `index != -1 && "
 Expected incoming value from Original PreHeader"' failed.
In-Reply-To: <bug-3408-206@http.llvm.org/bugs/>
Message-ID: <200901260212.n0Q2CMR9009721@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3408


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Chris Lattner <clattner at apple.com>  2009-01-25 20:12:22 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072659.html

Thanks!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 20:18:37 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 20:18:37 -0600
Subject: [LLVMbugs] [Bug 3405] Assertion `FromBr->isUnconditional() && "
 FromBB should end with uncond br!"' failed.
In-Reply-To: <bug-3405-206@http.llvm.org/bugs/>
Message-ID: <200901260218.n0Q2Ib54009925@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3405


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-25 20:18:37 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090119/072660.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 20:43:52 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 20:43:52 -0600
Subject: [LLVMbugs] [Bug 3409] New: gvnpre: segmentation fault
Message-ID: <bug-3409-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3409

           Summary: gvnpre: segmentation fault
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sliedes at cc.hut.fi
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2446)
 --> (http://llvm.org/bugs/attachment.cgi?id=2446)
test case

Running "opt bugpoint-reduced-simplified.bc -gvnpre" on the attached bitcode
file causes a segmentation fault.

Here's the bitcode disassembled:

------------------------------------------------------------
; ModuleID = 'bugpoint-reduced-simplified.bc'
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-unknown-linux-gnu"

define i32 @base2flt1(i32 %e) nounwind readnone {
entry:
        br i1 false, label %entry.bb12_crit_edge, label %bb1

entry.bb12_crit_edge:           ; preds = %entry
        unreachable

bb1:            ; preds = %entry
        br i1 false, label %bb1.bb10_crit_edge, label %bb1.bb2_crit_edge

bb1.bb2_crit_edge:              ; preds = %bb1
        br label %bb2

bb1.bb10_crit_edge:             ; preds = %bb1
        unreachable

bb2:            ; preds = %bb4.bb2_crit_edge, %bb1.bb2_crit_edge
        br i1 false, label %bb4, label %bb2.bb12_crit_edge

bb2.bb12_crit_edge:             ; preds = %bb2
        br label %bb12

bb4:            ; preds = %bb2
        br i1 false, label %bb4.bb11_crit_edge, label %bb4.bb2_crit_edge

bb4.bb2_crit_edge:              ; preds = %bb4
        br label %bb2

bb4.bb11_crit_edge:             ; preds = %bb4
        br label %bb11

bb7:            ; preds = %bb10
        br i1 false, label %bb7.bb12_crit_edge, label %bb9

bb7.bb12_crit_edge:             ; preds = %bb7
        br label %bb12

bb9:            ; preds = %bb7
        %0 = lshr i32 %0, 1             ; <i32> [#uses=2]
        br label %bb10

bb10:           ; preds = %bb9
        br i1 false, label %bb7, label %bb10.bb11_crit_edge

bb10.bb11_crit_edge:            ; preds = %bb10
        br label %bb11

bb11:           ; preds = %bb10.bb11_crit_edge, %bb4.bb11_crit_edge
        %m_addr.1 = phi i32 [ 0, %bb4.bb11_crit_edge ], [ %0,
%bb10.bb11_crit_edge ]            ; <i32> [#uses=1]
        %1 = and i32 %m_addr.1, -16777217               ; <i32> [#uses=0]
        br label %UnifiedReturnBlock

bb12:           ; preds = %bb7.bb12_crit_edge, %bb2.bb12_crit_edge
        br label %UnifiedReturnBlock

UnifiedReturnBlock:             ; preds = %bb12, %bb11
        ret i32 0
}
------------------------------------------------------------


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sun Jan 25 22:11:55 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Jan 2009 22:11:55 -0600
Subject: [LLVMbugs] [Bug 3101] cell spu mul lowering issues
In-Reply-To: <bug-3101-206@http.llvm.org/bugs/>
Message-ID: <200901260411.n0Q4BtM4013647@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3101


Scott Michel <scooter.phd at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #15 from Scott Michel <scooter.phd at gmail.com>  2009-01-25 22:11:54 ---
(In reply to comment #14)
> Thanks Scott, with your recent commits, the fix is much more obvious.  If you
> look at the -view-legalize-dags output from LLC, you'll see that it contains a
> sign_extend_inreg(i64 mul, i32).  I think it makes a lot of sense for you to
> add a dag combine xform that sees this and transforms it into an i32 mul when
> the i64 mul isn't legal.

Capitulate to the wackiness and just implement a full-on i64 mul, which is
indeed a sight to behold in the generated tblgen output.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 00:52:59 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 00:52:59 -0600
Subject: [LLVMbugs] [Bug 3410] New: Better 'expected ';
	' at end of statement diagnostic
Message-ID: <bug-3410-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3410

           Summary: Better 'expected ';' at end of statement diagnostic
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: quality-of-implementation
          Severity: enhancement
          Priority: P2
         Component: parser
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


Consider this erroneous program:

void foo() {
  int X;
  X = 4
}

We currently emit:

t.c:4:1: error: expected ';' after expression
}
^

This is ok as far as basic stuff goes, but the caret diagnostics would be much
nicer if the caret was pointing to the end of the previous line (right after
the 4) instead of at the }.  This happens very frequently if you forget a ;.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 09:31:48 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 09:31:48 -0600
Subject: [LLVMbugs] [Bug 3411] New: ice: Invalid node ID for RAUW analysis!
Message-ID: <bug-3411-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3411

           Summary: ice: Invalid node ID for RAUW analysis!
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


Seen using r62977 on Ubuntu Hardy on x86.

regehr at john-home:~/volatile/tmp132$ llvm-gcc -O small.c
cc1: LegalizeTypes.cpp:701: void
llvm::DAGTypeLegalizer::ReplaceValueWith(llvm::SDValue, llvm::SDValue):
Assertion `N->getNodeId() != DAGTypeLegalizer::ReadyToProcess && N->getNodeId()
!= DAGTypeLegalizer::Processed && "Invalid node ID for RAUW analysis!"' failed.
small.c:19: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.
regehr at john-home:~/volatile/tmp132$ cat small.c
int g_3;
volatile uint32;

int foo (int p_81);

void bar (long long p_66)
{
  char l_90 = 0;
  if (foo (l_90))
    {
    }
  else
    {
      foo(0);
    }

  foo (1 ^ p_66 & (p_66 | g_3) << 8 -
       p_66 ? p_66 | g_3 : (p_66 | g_3) - p_66);
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 13:00:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 13:00:35 -0600
Subject: [LLVMbugs] [Bug 3412] New: Assertion
 `isa<PointerType>(V->getType()) && " Unexpected operand type!"' failed.
Message-ID: <bug-3412-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3412

           Summary: Assertion `isa<PointerType>(V->getType()) && "Unexpected
                    operand type!"' failed.
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sliedes at cc.hut.fi
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2448)
 --> (http://llvm.org/bugs/attachment.cgi?id=2448)
test case

On running

  opt -licm -loop-reduce

on the attached bitcode, the following assertion failure happens:

------------------------------------------------------------
opt: Value.cpp:370: llvm::Value* llvm::Value::getUnderlyingObject(): Assertion
`isa<PointerType>(V->getType()) && "Unexpected operand type!"' failed.
0   opt       0x000000000078440f
1   libc.so.6 0x00007f08bf276f60
2   libc.so.6 0x00007f08bf276ed5 gsignal + 53
3   libc.so.6 0x00007f08bf2783f3 abort + 387
4   libc.so.6 0x00007f08bf26fdc9 __assert_fail + 233
5   opt       0x0000000000735e94 llvm::Value::getUnderlyingObject() + 484
6   opt       0x000000000063da18
7   opt       0x00000000006389de
llvm::AliasSet::aliasesCallSite(llvm::CallSite, llvm::AliasAnalysis&) const +
414
8   opt       0x0000000000639806
llvm::AliasSetTracker::findAliasSetForCallSite(llvm::CallSite) + 150
9   opt       0x0000000000639ba6 llvm::AliasSetTracker::add(llvm::CallSite) +
118
10  opt       0x000000000063a763 llvm::AliasSetTracker::add(llvm::Instruction*)
+ 211
11  opt       0x000000000063a84a llvm::AliasSetTracker::add(llvm::BasicBlock&)
+ 58
12  opt       0x0000000000565b25
13  opt       0x000000000065dcbb
llvm::LPPassManager::runOnFunction(llvm::Function&) + 955
14  opt       0x000000000072199b
llvm::FPPassManager::runOnFunction(llvm::Function&) + 347
15  opt       0x0000000000721bd6
llvm::FPPassManager::runOnModule(llvm::Module&) + 166
16  opt       0x0000000000721577
llvm::MPPassManager::runOnModule(llvm::Module&) + 199
17  opt       0x0000000000721776 llvm::PassManagerImpl::run(llvm::Module&) +
134
18  opt       0x0000000000497166 main + 3046
19  libc.so.6 0x00007f08bf2631a6 __libc_start_main + 230
20  opt       0x000000000048be39 sinh + 65
Aborted
------------------------------------------------------------

Here's the bitcode disassembled:

------------------------------------------------------------
; ModuleID = 'bugpoint-reduced-simplified.bc'
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-unknown-linux-gnu"

declare i64 @strlen(i8* nocapture) nounwind readonly

define void @relem515(double %val) nounwind {
entry:
        br i1 false, label %bb1.i3.i, label %bb8.i11

bb1.i3.i:               ; preds = %bb4.i6.i.i, %entry
        %tmp4.i = tail call i64 @strlen(i8* null) nounwind readonly            
; <i64> [#uses=0]
        br label %bb4.i6.i.i

bb3.i5.i.i:             ; preds = %bb4.i6.i.i
        store i8 0, i8* %tmp5.i.i9, align 1
        %indvar.next76 = add i64 %indvar75, 1           ; <i64> [#uses=1]
        br label %bb4.i6.i.i

bb4.i6.i.i:             ; preds = %bb3.i5.i.i, %bb1.i3.i
        %indvar75 = phi i64 [ 0, %bb1.i3.i ], [ %indvar.next76, %bb3.i5.i.i ]  
        ; <i64> [#uses=2]
        %p.i.0.i.i.rec = mul i64 %indvar75, -1          ; <i64> [#uses=1]
        %tmp5.i.i9.rec = add i64 %p.i.0.i.i.rec, -1             ; <i64>
[#uses=1]
        %tmp5.i.i9 = getelementptr i8* null, i64 %tmp5.i.i9.rec         ; <i8*>
[#uses=1]
        br i1 false, label %bb3.i5.i.i, label %bb1.i3.i

bb8.i11:                ; preds = %entry
        ret void
}
------------------------------------------------------------


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 13:00:44 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 13:00:44 -0600
Subject: [LLVMbugs] [Bug 3413] New: llvm-gcc: x86_64 passing of struct {
 char a, b, c; } generates invalid read
Message-ID: <bug-3413-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3413

           Summary: llvm-gcc: x86_64 passing of struct { char a, b, c; }
                    generates invalid read
           Product: new-bugs
           Version: unspecified
          Platform: Macintosh
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


llvm-gcc generates an out of bounds read when passing struct { char a, b, c; }.

--
ddunbar at ozzy:tmp$ echo 'struct { char a,b,c; } f0() {}' | llvm-gcc -m64 -S
-emit-llvm -o - -x c -

target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-apple-darwin10.0"
        %struct.anon = type { i8, i8, i8 }

define i32 @f0() nounwind {
entry:
        %retval = alloca %struct.anon           ; <%struct.anon*> [#uses=1]
        %"alloca point" = bitcast i32 0 to i32          ; <i32> [#uses=0]
        br label %return

return:         ; preds = %entry
        %retval1 = bitcast %struct.anon* %retval to i32*                ;
<i32*> [#uses=1]
        %retval2 = load i32* %retval1           ; <i32> [#uses=1]
        ret i32 %retval2
}
--
The read is accessing an extra byte after the alloca.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 15:55:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 15:55:21 -0600
Subject: [LLVMbugs] [Bug 3393] llc abort with -enable-legalize-types-checking
In-Reply-To: <bug-3393-206@http.llvm.org/bugs/>
Message-ID: <200901262155.n0QLtLhQ026982@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3393


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Duncan Sands <baldrick at free.fr>  2009-01-26 15:55:21 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090126/072690.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 18:56:59 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 18:56:59 -0600
Subject: [LLVMbugs] [Bug 3414] New: missing x86-64 ABI alignment for arrays
Message-ID: <bug-3414-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3414

           Summary: missing x86-64 ABI alignment for arrays
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: gohman at apple.com
                CC: llvmbugs at cs.uiuc.edu


The x86-64 ABI states

"An array uses the same alignment as its elements, except that a local or
global 
array variable of length at least 16 bytes or a C99 variable-length array
variable 
always has alignment of at least 16 bytes."

For the following testcase:

void frob(void *);

void foo(void) {
  char x[17];
  frob(x);
}

llvm-gcc -m64 emits

[...]
   subq $32, $rsp
   leaq -17(%rsp), %rdi
   call frob

which shows that the array is not being aligned on a 16 byte boundary.
The LLVM IR has this:

  %x = alloca [17 x 18]


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 19:13:04 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 19:13:04 -0600
Subject: [LLVMbugs] [Bug 3415] New: Infinite loop in escape analysis
Message-ID: <bug-3415-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3415

           Summary: Infinite loop in escape analysis
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sliedes at cc.hut.fi
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2451)
 --> (http://llvm.org/bugs/attachment.cgi?id=2451)
test case

Running

  opt bugpoint-reduced-simplified.bc -escape-analysis

on the attached bitcode file causes opt to enter an infinite (or very long)
loop.

Architecture is x86-64 and version is recent svn trunk.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 21:57:12 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 21:57:12 -0600
Subject: [LLVMbugs] [Bug 3024] possible volatile bug
In-Reply-To: <bug-3024-206@http.llvm.org/bugs/>
Message-ID: <200901270357.n0R3vCVl006420@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3024


John Regehr <regehr at cs.utah.edu> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #8 from John Regehr <regehr at cs.utah.edu>  2009-01-26 21:57:11 ---
Yes fixed.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 21:59:25 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 21:59:25 -0600
Subject: [LLVMbugs] [Bug 3289] Assertion `LRegs.size() == 1 && "Can't handle
 this yet!"' failed .
In-Reply-To: <bug-3289-206@http.llvm.org/bugs/>
Message-ID: <200901270359.n0R3xPFW006525@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3289


John Regehr <regehr at cs.utah.edu> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from John Regehr <regehr at cs.utah.edu>  2009-01-26 21:59:24 ---
Yes fixed.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 22:17:57 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 22:17:57 -0600
Subject: [LLVMbugs] [Bug 3416] New: ice: Assertion `vi.AliveBlocks.none() &&
 "Shouldn' t be alive across any blocks!"' failed.
Message-ID: <bug-3416-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3416

           Summary: ice: Assertion `vi.AliveBlocks.none() && "Shouldn't be
                    alive across any blocks!"' failed.
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


Seen using r63072 on Ubuntu Hardy on x86.

regehr at john-home:~/volatile/tmp133$ llvm-gcc -O3 small.c
cc1: LiveIntervalAnalysis.cpp:425: void
llvm::LiveIntervals::handleVirtualRegisterDef(llvm::MachineBasicBlock*,
llvm::ilist_iterator<llvm::MachineInstr>, unsigned int, llvm::MachineOperand&,
unsigned int, llvm::LiveInterval&): Assertion `vi.AliveBlocks.none() &&
"Shouldn't be alive across any blocks!"' failed.
small.c:28: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.

regehr at john-home:~/volatile/tmp133$ cat small.c

long long foo (long long _si1, int _si2)
{
  return _si1 > 9223372036854775807LL - 1 ? _si1 : _si1 + _si2;
}

int bar (unsigned long long _ui1, int _ui2)
{
  return _ui1 * _ui2;
}

int g_78; 

int func_31 (int32p_32)
{
  unsigned long long l_33;
  for (l_33 = 1; l_33 < -13; l_33 = foo (l_33, 1))
    {
    }
  return l_33;
}

baz (uint32p_4)
{
  if (func (bar (g_78, func_31 (1))))
    {
    }
  return 0;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Mon Jan 26 23:36:42 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Jan 2009 23:36:42 -0600
Subject: [LLVMbugs] [Bug 3275] math bug
In-Reply-To: <bug-3275-206@http.llvm.org/bugs/>
Message-ID: <200901270536.n0R5agT1009632@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3275


Nick Lewycky <nicholas at mxc.ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 01:56:32 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 01:56:32 -0600
Subject: [LLVMbugs] [Bug 3418] New: Strange parse error with __extension__
	operator
Message-ID: <bug-3418-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3418

           Summary: Strange parse error with __extension__ operator
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
int a() {
  __extension__ (*(char*)0) = 1;
}

clang output:
<stdin>:2:29: error: expected ';' after expression
  __extension__ (*(char*)0) = 1;
                            ^
<stdin>:2:29: error: expected expression
<stdin>:2:3: warning: expression result unused
  __extension__ (*(char*)0) = 1;
  ^             ~~~~~~~~~~~
3 diagnostics generated.

gcc accepts this without any errors/warnings.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 02:44:21 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 02:44:21 -0600
Subject: [LLVMbugs] [Bug 3418] Strange parse error with __extension__
	operator
In-Reply-To: <bug-3418-206@http.llvm.org/bugs/>
Message-ID: <200901270844.n0R8iLJQ018249@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3418


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-27 02:44:21 ---
Fixed in r63100.  (In retrospect, I didn't really need to file this, but I was
expecting it to be a bit more difficult.)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 03:44:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 03:44:20 -0600
Subject: [LLVMbugs] [Bug 3419] New: llc: assertion failure with i128 "Too
 many bits for int64_t "
Message-ID: <bug-3419-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3419

           Summary: llc: assertion failure with i128 "Too many bits for
                    int64_t"
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


When compiling part of LibTomMath using llvm-gcc -O1  I get a failure in
Codegen:

cc1: /home/edwin/llvm-bootstrap/llvm/include/llvm/ADT/APInt.h:1051: int64_t
llvm::APInt::getSExtValue() const: Assertion `getMinSignedBits() <= 64 && "Too
many bits for int64_t"' failed.
../../../clamav-devel/libclamav/bignum.c:9506: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.

/* testcase */
typedef unsigned long mp_digit;
typedef unsigned long mp_word __attribute__ ((mode (TI)));
typedef struct
{
  int used, alloc, sign;
  mp_digit *dp;
}
mp_int;
int
mp_div_3 (mp_int * a, mp_int * c, mp_digit * d)
{
  mp_word w, t;
  int res, ix;
  for (ix = a->used - 1; ix >= 0; ix--)
    {
      w = (w << ((mp_word) 60)) | ((mp_word) a->dp[ix]);
      if (w >= 3)
        {
          while (w >= 3)
            {
              w -= 3;
            }
        }
    }
}

Bugpoint reduced testcase:
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-unknown-linux-gnu"
        %struct.mp_int = type { i32, i32, i32, i64* }

define i32 @mp_div_3(%struct.mp_int* nocapture %a, %struct.mp_int* nocapture
%c, i64* nocapture %d) nounwind readonly {
entry:
        br label %bb

bb:             ; preds = %bb1, %bb, %entry
        br i1 false, label %bb1, label %bb

bb1:            ; preds = %bb1, %bb
        %indvar = phi i128 [ 0, %bb ], [ %indvar.next, %bb1 ]           ;
<i128> [#uses=2]
        %tmp = mul i128 %indvar, -3             ; <i128> [#uses=1]
        %w.17 = add i128 %tmp, 0                ; <i128> [#uses=1]
        %0 = add i128 %w.17, -3         ; <i128> [#uses=1]
        %1 = icmp ugt i128 %0, 2                ; <i1> [#uses=1]
        %indvar.next = add i128 %indvar, 1              ; <i128> [#uses=1]
        br i1 %1, label %bb1, label %bb
}


llc: /home/edwin/llvm-svn/llvm/include/llvm/ADT/APInt.h:1051: int64_t
llvm::APInt::getSExtValue() const: Assertion `getMinSig
nedBits() <= 64 && "Too many bits for int64_t"' failed.
#0  0x00007f018e79fed5 in raise () from /lib/libc.so.6
#1  0x00007f018e7a13f3 in abort () from /lib/libc.so.6
#2  0x00007f018e798dc9 in __assert_fail () from /lib/libc.so.6
#3  0x000000000083b638 in llvm::APInt::getSExtValue (this=0x353da78) at
/home/edwin/llvm-svn/llvm/include/llvm/ADT/APInt.h:10
51
#4  0x000000000083b657 in llvm::ConstantInt::getSExtValue (this=0x353da40) at
/home/edwin/llvm-svn/llvm/include/llvm/Constant
s.h:83
#5  0x0000000000f499de in CheckForIVReuse (this=0x3551a30, HasBaseReg=false,
AllUsesAreAddresses=false, AllUsesAreOutsideLoop
=false, Stride=@0x356fef0, IV=@0x7fff97ab5570, Ty=0x353ccb0,
UsersToProcess=@0x7fff97ab5610) at LoopStrengthReduce.cpp:1259
#6  0x0000000000f4c967 in StrengthReduceStridedIVUsers (this=0x3551a30,
Stride=@0x356fef0, Uses=@0x356fef8, L=0x356ebb0, isOn
lyStride=true) at LoopStrengthReduce.cpp:1487
#7  0x0000000000f4eb00 in runOnLoop (this=0x3551a30, L=0x356ebb0,
LPM=@0x3553110) at LoopStrengthReduce.cpp:2308
#8  0x0000000000fa61c8 in llvm::LPPassManager::runOnFunction (this=0x3553110,
F=@0x353d460) at LoopPass.cpp:226
#9  0x000000000104c9a1 in llvm::FPPassManager::runOnFunction (this=0x35490b0,
F=@0x353d460) at PassManager.cpp:1323
#10 0x000000000104d4e4 in llvm::FunctionPassManagerImpl::run (this=0x353e160,
F=@0x353d460) at PassManager.cpp:1281
#11 0x000000000104d642 in llvm::FunctionPassManager::run (this=0x7fff97ab5cd0,
F=@0x353d460) at PassManager.cpp:1226
#12 0x00000000008106a6 in main (argc=2, argv=0x7fff97ab5ef8) at llc.cpp:317


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 04:09:52 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 04:09:52 -0600
Subject: [LLVMbugs] [Bug 3420] New: clang: unsupported machine mode 'TI'
Message-ID: <bug-3420-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3420

           Summary: clang: unsupported machine mode 'TI'
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


/* testcase */
typedef unsigned long      mp_word __attribute__ ((mode(TI)))

$ /home/edwin/llvm-svn/llvm/tools/clang/utils/ccc -DHAVE_CONFIG_H -I.
-I../../../clamav-devel/libclamav -I..  -I../../../clamav-devel
-I../../../clamav-devel/libclamav/nsis -I../../../clamav-devel/libclamav/lzma 
-DWARN_DLOPEN_FAIL   -O2  -c ../../../clamav-devel/libclamav/bignum.c
In file included from ../../../clamav-devel/libclamav/bignum.c:17:
../../../clamav-devel/libclamav/bignum.h:94:55: error: unsupported machine mode
'TI'
   typedef unsigned long      mp_word __attribute__ ((mode(TI)))


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 04:31:33 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 04:31:33 -0600
Subject: [LLVMbugs] [Bug 3421] New: miscompilation on x86 (involves
	addressing?)
Message-ID: <bug-3421-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3421

           Summary: miscompilation on x86 (involves addressing?)
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2452)
 --> (http://llvm.org/bugs/attachment.cgi?id=2452)
Testcase

Testcase attached. Compiled+run through plain llc, results in a program that
prints out "0".  Compiled+run any other way (for example, with "llc -fast"),
results in a program that prints out "1".


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 04:43:16 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 04:43:16 -0600
Subject: [LLVMbugs] [Bug 3422] New: clang: static analyzer crash when
	analyzing aspack.c
Message-ID: <bug-3422-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3422

           Summary: clang: static analyzer crash when analyzing aspack.c
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2453)
 --> (http://llvm.org/bugs/attachment.cgi?id=2453)
preprocessed file

On SVN r63101, x86-64 Linux:

ANALYZE: ../../../clamav-devel/libclamav/aspack.c readstream
ANALYZE: ../../../clamav-devel/libclamav/aspack.c getdec
ANALYZE: ../../../clamav-devel/libclamav/aspack.c build_decrypt_array
../../../clamav-devel/libclamav/aspack.c:147:3: warning: dead assignment
  i = 0;
  ^   ~
ANALYZE: ../../../clamav-devel/libclamav/aspack.c getbits
ANALYZE: ../../../clamav-devel/libclamav/aspack.c build_decrypt_dictionaries
ANALYZE: ../../../clamav-devel/libclamav/aspack.c decrypt
ANALYZE: ../../../clamav-devel/libclamav/aspack.c decomp_block
ANALYZE: ../../../clamav-devel/libclamav/aspack.c unaspack212
clang: /home/edwin/llvm-svn/llvm/include/llvm/Support/Casting.h:199: typename
llvm::cast_retty<To, From>::ret_type llvm::cast(const Y&) [with X =
clang::PointerType, Y = clang::Type*]: Assertion `isa<X>(Val) && "cast<Ty>()
argument of incompatible type!"' failed.
0   clang     0x0000000001036e4f
1   libc.so.6 0x00002b8d4aafbf60
2   libc.so.6 0x00002b8d4aafbed5 gsignal + 53
3   libc.so.6 0x00002b8d4aafd3f3 abort + 387
4   libc.so.6 0x00002b8d4aaf4dc9 __assert_fail + 233
5   clang     0x0000000000a6bb86
6   clang     0x0000000000a6a4d1
clang::ElementRegion::getRValueType(clang::ASTContext&) const + 97
7   clang     0x0000000000a6bb9a
clang::TypedRegion::getLValueType(clang::ASTContext&) const + 10
8   clang     0x0000000000a14e8f
9   clang     0x0000000000a52533 clang::GRExprEngine::VisitCast(clang::Expr*,
clang::Expr*, clang::ExplodedNode<clang::GRState>*,
clang::ExplodedNodeSet<clang::GRState>&) + 2675
10  clang     0x0000000000a4d317 clang::GRExprEngine::Visit(clang::Stmt*,
clang::ExplodedNode<clang::GRState>*, clang::ExplodedNodeSet<clang::GRState>&)
+ 951
11  clang     0x0000000000a502a1
clang::GRExprEngine::VisitBinaryOperator(clang::BinaryOperator*,
clang::ExplodedNode<clang::GRState>*, clang::ExplodedNodeSet<clang::GRState>&)
+ 545
12  clang     0x0000000000a4d2e8 clang::GRExprEngine::Visit(clang::Stmt*,
clang::ExplodedNode<clang::GRState>*, clang::ExplodedNodeSet<clang::GRState>&)
+ 904
13  clang     0x0000000000a539cf clang::GRExprEngine::ProcessStmt(clang::Stmt*,
clang::GRStmtNodeBuilder<clang::GRState>&) + 591
14  clang     0x0000000000a58c65
clang::GRCoreEngine<clang::GRExprEngine>::ProcessStmt(clang::Stmt*,
clang::GRStmtNodeBuilderImpl&) + 85
15  clang     0x0000000000a4469d
clang::GRCoreEngineImpl::HandlePostStmt(clang::PostStmt const&,
clang::CFGBlock*, unsigned int, clang::ExplodedNodeImpl*) + 205
16  clang     0x0000000000a44b84
clang::GRCoreEngineImpl::ExecuteWorkList(unsigned int) + 484
17  clang     0x000000000059ad0c
18  clang     0x000000000059a715
19  clang     0x000000000059aac6
20  clang     0x0000000000a862e5 clang::ParseAST(clang::Preprocessor&,
clang::ASTConsumer*, bool, bool) + 245
21  clang     0x00000000005cfe14
22  clang     0x00000000005d4d67 main + 1383
23  libc.so.6 0x00002b8d4aae81a6 __libc_start_main + 230
24  clang     0x0000000000592f89 std::ios_base::Init::~Init() + 65


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 06:23:37 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 06:23:37 -0600
Subject: [LLVMbugs] [Bug 3423] New: link fails on ARM EABI machine
Message-ID: <bug-3423-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3423

           Summary: link fails on ARM EABI machine
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Target-Independent JIT
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: thebohemian at gmx.net
                CC: llvmbugs at cs.uiuc.edu


Hi,
currently linking llvm on ARM EABI machines fails because of the missing symbol

__register_frame

which is not available on that platform. The suggested workaround is to disable
its use depending on the __ARM_EABI__ macro. See patch.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 06:28:02 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 06:28:02 -0600
Subject: [LLVMbugs] [Bug 3424] New: C++ static initialization order problem
Message-ID: <bug-3424-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3424

           Summary: C++ static initialization order problem
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Linker
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: thebohemian at gmx.net
                CC: llvmbugs at cs.uiuc.edu


Due to the order of static initialization being undefined in C++ I get a
problem where g_symbol is not initialized when it is used for the first time.

According to cppreference.com one can work around this issue by moving the uses
of g_symbol into a accessor method. This is what I did in the attached patch
and that solved my issue.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 07:48:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 07:48:27 -0600
Subject: [LLVMbugs] [Bug 3425] New: Inaccurate storage class for
 redeclaration of function
Message-ID: <bug-3425-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3425

           Summary: Inaccurate storage class for redeclaration of function
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2456)
 --> (http://llvm.org/bugs/attachment.cgi?id=2456)
WIP

Testcase:
static void a(void);
void a(void) {};

Run through clang -emit-llvm, this should result in empty output; however, this
isn't the case with trunk clang.

The issue is basically that Sema::MergeFunctionDecl isn't taking into account
the requirements in C99 6.2.2p4 and p5, which taken together mean that "a"
should have internal linkage.

I've attached a patch, but I'm not really sure if it's correct.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 09:34:10 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 09:34:10 -0600
Subject: [LLVMbugs] [Bug 3426] New: lli fails cast assertion.
Message-ID: <bug-3426-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3426

           Summary: lli fails cast assertion.
           Product: new-bugs
           Version: unspecified
          Platform: PC
               URL: http://shootout.alioth.debian.org/u32q/benchmark.php?tes
                    t=binarytrees&lang=gcc&id=5
        OS/Version: Linux
            Status: NEW
          Keywords: regression
          Severity: critical
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: micketeer at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Compiling and trying to run the bitcode of the binarytrees benchmark crashes
lli.

The following bitcode was compiled both with version 2.2 and 2.4 from Debian
unstable and experimental respectively.

$ lli -version
Low Level Virtual Machine (http://llvm.org/):
  llvm version 2.4 (Debian 2.4-1~exp1)
  Optimized build with assertions.

Compiling to a native binary with llvm-gcc works fine.
Using version 2.2 there was no errors (in fact lli was faster than gcc)
Step 1, compile
$ llvm-gcc binarytrees.c -emit-llvm -c -o binarytrees.bc

Step 2, run with a large input (for me, the limit is at 18, 17 and below works
and 18 and above crashes)
$ lli binarytrees.bc 18
stretch tree of depth 19         check: -1
lli: /build/buildd/llvm-2.4/include/llvm/Support/Casting.h:199: typename
llvm::cast_retty<To, From>::ret_type llvm::cast(const Y&) [with X =
llvm::Instruction, Y = llvm::User*]: Assertion `isa<X>(Val) && "cast<Ty>()
argument of incompatible type!"' failed.
lli[0x85972aa]
/lib/libc.so.6(abort+0x188)[0xb7d6dd78]
/lib/libc.so.6(__assert_fail+0xf0)[0xb7d65590]
lli[0x84b0a2b]
[0x9e69bc0]
[0xa]
Aborted

Running the file with the latest lli (from svn)
$ /usr/local/bin/lli binarytrees.bc 18
stretch tree of depth 19         check: -1
lli: /build/llvm/include/llvm/Support/Casting.h:199: typename
llvm::cast_retty<To, From>::ret_type llvm::cast(const Y&) [with X =
llvm::Instruction, Y = llvm::User*]: Assertion `isa<X>(Val) && "cast<Ty>()
argument of incompatible type!"' failed.
0   lli       0x08973066
1   lli       0x089733a8
2             0xb7f68400 __kernel_sigreturn + 0
3   libc.so.6 0xb7d05d78 abort + 392
4   libc.so.6 0xb7cfd590 __assert_fail + 240
5   lli       0x086510a7
6   lli       0x088163f5
7   lli       0x08816cf1
8   lli       0x08817de1
9   lli       0x089101ad llvm::FPPassManager::runOnFunction(llvm::Function&) +
249
10  lli       0x08910c22 llvm::FunctionPassManagerImpl::run(llvm::Function&) +
116
11  lli       0x08910d83 llvm::FunctionPassManager::run(llvm::Function&) + 137
12  lli       0x085b3f23 llvm::JIT::runJITOnFunction(llvm::Function*) + 119
13  lli       0x085b4315 llvm::JIT::getPointerToFunction(llvm::Function*) + 649
14  lli       0x085bd431
15  lli       0x0855b8f1 X86CompilationCallback2 + 142
16  lli       0x0855b529
17  lli       0xb6d57ffd
18  libc.so.6 0xb7da4bee clone + 94
Aborted


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 11:56:00 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 11:56:00 -0600
Subject: [LLVMbugs] [Bug 3421] miscompilation on x86 (involves addressing?)
In-Reply-To: <bug-3421-206@http.llvm.org/bugs/>
Message-ID: <200901271756.n0RHu0Kc013777@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3421


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




--- Comment #3 from Eli Friedman <sharparrow1 at yahoo.com>  2009-01-27 11:55:59 ---
Yeah, this is my own fault :(


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 12:30:14 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 12:30:14 -0600
Subject: [LLVMbugs] [Bug 3422] clang: static analyzer crash when analyzing
	aspack.c
In-Reply-To: <bug-3422-206@http.llvm.org/bugs/>
Message-ID: <200901271830.n0RIUENY015218@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3422


Ted Kremenek <kremenek at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kremenek at apple.com
             Status|NEW                         |RESOLVED
          Component|new bugs                    |Static Analyzer
            Product|new-bugs                    |clang
         Resolution|                            |FIXED




--- Comment #3 from Ted Kremenek <kremenek at apple.com>  2009-01-27 12:30:14 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090126/011396.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 14:10:17 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 14:10:17 -0600
Subject: [LLVMbugs] [Bug 3427] New: Assertion failed: (0 && "Incomplete
 types have no size!"), function getTypeInfo, file ASTContext.cpp, line 296.
Message-ID: <bug-3427-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3427

           Summary: Assertion failed: (0 && "Incomplete types have no
                    size!"), function getTypeInfo, file ASTContext.cpp, line
                    296.
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: FreeBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: rdivacky at freebsd.org
                CC: llvmbugs at cs.uiuc.edu


I am getting this when trying to compile freebsd kernel with CC =
/usr/local/bin/ccc (the old one not the new one). Unfortunatelly the single
file is not compilable and "ccc -E" processed file does not trigger the assert.
Anyway, I am attaching both..

ccc -fsyntax-only does NOT trigger the assert so I guess its the CodeGen


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 18:38:52 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 18:38:52 -0600
Subject: [LLVMbugs] [Bug 3427] Assertion failed: (0 && "Incomplete types
 have no size!"), function getTypeInfo, file ASTContext.cpp, line 296.
In-Reply-To: <bug-3427-206@http.llvm.org/bugs/>
Message-ID: <200901280038.n0S0cqPP030160@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3427


Nuno Lopes <nunoplopes at sapo.pt> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 18:43:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 18:43:35 -0600
Subject: [LLVMbugs] [Bug 3428] New: inline asm crash
Message-ID: <bug-3428-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3428

           Summary: inline asm crash
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


llvm-gcc crashes on this asm in selectiondagbuild:

        #include <stdint.h>

        typedef struct { uint64_t a0, a1; } u128;

        u128 mul(uint64_t a, uint64_t b)
        {
                u128 result;
                __asm__("mulq %[b]"
                        : "=a" (result.a0), "=d" (result.a1)
                        : "a" (a), [b] "rm" (b));
                return result;
        }

SelectionDAGBuild.cpp:532: failed assertion `TLI.isTypeLegal(PartVT) &&
"Copying to an illegal type!"'


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 18:45:09 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 18:45:09 -0600
Subject: [LLVMbugs] [Bug 2502] Function pointer types are wrongly compared
In-Reply-To: <bug-2502-206@http.llvm.org/bugs/>
Message-ID: <200901280045.n0S0j9Lq030468@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2502


Nuno Lopes <nunoplopes at sapo.pt> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |nunoplopes at sapo.pt
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #10 from Nuno Lopes <nunoplopes at sapo.pt>  2009-01-27 18:45:09 ---
this is already fixed. I've added the given test case to the test suite.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Tue Jan 27 19:34:05 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Jan 2009 19:34:05 -0600
Subject: [LLVMbugs] [Bug 3429] New: __label__ not supported
Message-ID: <bug-3429-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3429

           Summary: __label__ not supported
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


the following doesnt work (-fsyntax-only):

void foo() {
        __label__ l;
        int v;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 02:20:48 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 02:20:48 -0600
Subject: [LLVMbugs] [Bug 3419] llc: assertion failure with i128 "Too many
 bits for int64_t"
In-Reply-To: <bug-3419-206@http.llvm.org/bugs/>
Message-ID: <200901280820.n0S8Kmk5016703@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3419


T??r??k Edwin <edwintorok at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #5 from T??r??k Edwin <edwintorok at gmail.com>  2009-01-28 02:20:47 ---
(In reply to comment #4)
> Great!  Please close if it looks ok, thanks Edwin!
> 

I think it is ok.
Apparently the unit tests are not run by 'make check', or the nightly tester.
It would be nice to have 'make unittests' run by the nightly tester.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 05:38:52 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 05:38:52 -0600
Subject: [LLVMbugs] [Bug 3415] Infinite loop in escape analysis
In-Reply-To: <bug-3415-206@http.llvm.org/bugs/>
Message-ID: <200901281138.n0SBcqHY002095@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3415


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Duncan Sands <baldrick at free.fr>  2009-01-28 05:38:26 ---
Fixed here by deleting escape analysis:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090126/072806.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 07:59:26 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 07:59:26 -0600
Subject: [LLVMbugs] [Bug 3430] New: top-level 'struct x;
 ' decl shadows other non-top level '  struct s {...}' decls
Message-ID: <bug-3430-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3430

           Summary: top-level 'struct x;' decl shadows other non-top level
                    'struct s {...}' decls
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


'clang -fsyntax-only' does eat the following:

struct s {
        struct st {
                int v;
        } *ts;
};

// it works if next line is removed
struct st;

int foo() {
        struct st *f;
        return f->v + f[0].v;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 11:57:29 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 11:57:29 -0600
Subject: [LLVMbugs] [Bug 3431] New: Kaleidascope tutorial doesn't mention
	-rdynamic
Message-ID: <bug-3431-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3431

           Summary: Kaleidascope tutorial doesn't mention -rdynamic
           Product: Documentation
           Version: trunk
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: General docs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: skaflotten at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2464)
 --> (http://llvm.org/bugs/attachment.cgi?id=2464)
Mentions rdynamic

Please see

http://www.nabble.com/Kaleidoscope-doesn%27t-work-properly-td19669800.html#a19671105

The cause of this, on Linux, is that the -rdynamic is not specified as a g++
option (required, since the tutorial doesn't provide global mappings.)

A docu-fix is attached.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 11:58:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 11:58:35 -0600
Subject: [LLVMbugs] [Bug 3376] Cannot yet select: 0x411b508c: f64 = fneg
	0x411b5950
In-Reply-To: <bug-3376-206@http.llvm.org/bugs/>
Message-ID: <200901281758.n0SHwZ7k015562@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3376


Dan Gohman <gohman at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #15 from Dan Gohman <gohman at apple.com>  2009-01-28 11:58:35 ---
The second half of the fix is now done, here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090126/072817.html

Theoretically, the DAGCombiner and Legalizer could run
iteratively, allowing the DAGCombiner to create new
Custom nodes which are subsequently Legalized, however
situations where this would help are uncommon. In this
PR, it was only visible before the earlier fixes were
applied.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 13:41:05 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 13:41:05 -0600
Subject: [LLVMbugs] [Bug 3432] New: llvm-gcc is overly pessimistic of union
	alignment
Message-ID: <bug-3432-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3432

           Summary: llvm-gcc is overly pessimistic of union alignment
           Product: new-bugs
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: major
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: evan.cheng at apple.com
                CC: llvmbugs at cs.uiuc.edu


#include <xmmintrin.h>

typedef union
{
  int i[4];
  float f[4];
  __m128 v;
} vector4_t;

__m128
swizzle (const void *a, vector4_t * b)
{
  return _mm_loadl_pi (b->v, (__m64 *) a);
}

On x86 / Mac OS X, this compiles to:

target triple = "i386-apple-darwin9.6"
        %struct.vector4_t = type { <4 x float> }

define <4 x float> @swizzle(i8* nocapture %a, %struct.vector4_t* nocapture %b)
nounwind readonly {
entry:
        %0 = getelementptr %struct.vector4_t* %b, i32 0, i32 0          ; <<4 x
float>*> [#uses=1]
        %1 = load <4 x float>* %0, align 4              ; <<4 x float>>
[#uses=1]
        %tmp.i = bitcast i8* %a to double*              ; <double*> [#uses=1]
        %tmp1.i = load double* %tmp.i           ; <double> [#uses=1]
        %2 = insertelement <2 x double> undef, double %tmp1.i, i32 0           
; <<2 x double>> [#uses=1]
        %tmp2.i = bitcast <2 x double> %2 to <4 x float>                ; <<4 x
float>> [#uses=1]
        %3 = shufflevector <4 x float> %1, <4 x float> %tmp2.i, <4 x i32> < i32
4, i32 5, i32 2, i32 3 >                ; <<4 x float>> [#uses=1]
        ret <4 x float> %3
}

Notice the alignment of the load instruction is 4. This is overly strict. We
ended up issuing a movups instead of a movaps.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 13:45:55 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 13:45:55 -0600
Subject: [LLVMbugs] [Bug 3433] New: __alignof__ for basic types doesn't
	match gcc
Message-ID: <bug-3433-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3433

           Summary: __alignof__ for basic types doesn't match gcc
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


For example, on x86-32 gcc reports __alignof__(double) as 8, whereas the actual
alignment is 4.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 15:35:28 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 15:35:28 -0600
Subject: [LLVMbugs] [Bug 3434] New: Wrong x86 code generated
Message-ID: <bug-3434-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3434

           Summary: Wrong x86 code generated
           Product: new-bugs
           Version: unspecified
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lennart at augustsson.net
                CC: llvmbugs at cs.uiuc.edu


This is LLVM 2.6svn, rev 63214, generating code for x86.

The code below together with the test program should print 8.
It does print if 8 if optimized (with opt -std-compile-opts),
but with the code straight from the source it prints 1.
Looking at the generated assembly code it looks like the second loop, the one
with the compare instructions, has an empty body.

Maybe there's something wrong with my code, but since it works with
optimization I don't think so.

C code:
-------
#include <stdio.h>

int _fun1(int);

main()
{
  printf("%d\n", _fun1(3));
}


LLVM code:
----------
define i32 @_fun1(i32) {
_L1:
        br label %_L2

_L2:
        %1 = phi i32 [ 0, %_L1 ], [ %6, %_L3 ]
        %2 = phi <4 x i32> [ undef, %_L1 ], [ %5, %_L3 ]
        %3 = icmp ne i32 %1, 4
        br i1 %3, label %_L3, label %_L4

_L3:
        %4 = extractelement <4 x i32> < i32 0, i32 1, i32 2, i32 3 >, i32 %1
        %5 = insertelement <4 x i32> %2, i32 %0, i32 %1
        %6 = add i32 %1, 1
        br label %_L2

_L4:
        br label %_L5

_L5:
        %7 = phi i32 [ 0, %_L4 ], [ %14, %_L6 ]
        %8 = phi <4 x i1> [ undef, %_L4 ], [ %13, %_L6 ]
        %9 = icmp ne i32 %7, 4
        br i1 %9, label %_L6, label %_L7

_L6:
        %10 = extractelement <4 x i32> < i32 0, i32 1, i32 2, i32 3 >, i32 %7
        %11 = extractelement <4 x i32> %2, i32 %7
        %12 = icmp eq i32 %10, %11
        %13 = insertelement <4 x i1> %8, i1 %12, i32 %7
        %14 = add i32 %7, 1
        br label %_L5

_L7:
        %15 = bitcast <4 x i1> %8 to i4
        %16 = zext i4 %15 to i32
        ret i32 %16
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Wed Jan 28 22:44:14 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Jan 2009 22:44:14 -0600
Subject: [LLVMbugs] [Bug 3424] C++ static initialization order problem
In-Reply-To: <bug-3424-206@http.llvm.org/bugs/>
Message-ID: <200901290444.n0T4iEMU005327@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3424


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #9 from Chris Lattner <clattner at apple.com>  2009-01-28 22:44:13 ---
I slightly modified the patch and applied here.  Please verify that it fixes
the problem you see, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090126/072851.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 29 01:57:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Jan 2009 01:57:20 -0600
Subject: [LLVMbugs] [Bug 3432] llvm-gcc is overly pessimistic of union
	alignment
In-Reply-To: <bug-3432-206@http.llvm.org/bugs/>
Message-ID: <200901290757.n0T7vKeG012680@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3432


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-29 01:57:20 ---
This was an llvm-gcc bug that may be a regression caused by the lvalue
alignment stuff.  Evan, I checked in some patches for this, please commit the
testcase, thanks!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 29 07:59:18 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Jan 2009 07:59:18 -0600
Subject: [LLVMbugs] [Bug 3436] New: clang: cannot codegen address of global
	variable
Message-ID: <bug-3436-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3436

           Summary: clang: cannot codegen address of global variable
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu, rdivacky at freebsd.org


Created an attachment (id=2469)
 --> (http://llvm.org/bugs/attachment.cgi?id=2469)
fixed preprocessed file

Using SVN r63306 on x86-64:

/usr/local/bin/ccc -c -O2 -pipe -fno-strict-aliasing -march=native -std=c99 -g
-Wall -Wredundant-decls -Wnested-externs -Wstrict-prototypes 
-Wmissing-prototypes -Wpointer-arith -Winline -Wcast-qual  -Wundef
-Wno-pointer-sign -fformat-extensions -nostdinc  -I. -I../../..
-I../../../contrib/altq -D_KERNEL -DHAVE_KERNEL_OPTION_HEADERS -include
opt_global.h -fno-common -finline-limit=8000 --param inline-unit-growth=100
--param large-func

./../../dev/aac/aac.c:2831:17: warning: incompatible pointer types assigning
'int8_t [18]', expected 'char *'
                         adapter_type = ((struct aac_supplement_adapter_info *)
                                ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
error: cannot codegen this constant expression yet
error: cannot codegen this constant expression yet
error: cannot codegen this constant expression yet
ERROR: Error opening output file 'aac.o.bc'
*** Error code 1


$ clang -emit-llvm testcase-min.i
...
error: cannot codegen this constant expression yet

Reduced testcase from the aac.i file rdivacky sent me:

typedef unsigned int u_int;
    typedef unsigned long u_long;
    typedef char *caddr_t;
    typedef int d_open_t (struct cdev *dev, int oflags, int devtype,           
      struct thread *td);
    typedef int d_fdopen_t (struct cdev *dev, int oflags, struct thread *td,   
                struct file *fp);
    typedef int d_close_t (struct cdev *dev, int fflag, int devtype,           
       struct thread *td);
    typedef int d_ioctl_t (struct cdev *dev, u_long cmd, caddr_t data, int
fflag,                      struct thread *td);
    typedef int d_poll_t (struct cdev *dev, int events, struct thread *td);
    struct cdevsw {
    int d_version;
    u_int d_flags;
    const char *d_name;
    d_open_t *d_open;
    d_fdopen_t *d_fdopen;
    d_close_t *d_close;
    d_ioctl_t *d_ioctl;
    d_poll_t *d_poll;
  }
    __attribute__ ((__packed__));
    static d_open_t aac_open;
    static d_close_t aac_close;
    static d_ioctl_t aac_ioctl;
    static d_poll_t aac_poll;
    static struct cdevsw aac_cdevsw = {
    .d_version = 0x17032005,.d_flags = 0x00400000,.d_open = aac_open,.d_close =
    aac_close,.d_ioctl = aac_ioctl,.d_poll = aac_poll,.d_name = "aac", };
    int aac_attach (struct aac_softc *sc) {
    &aac_cdevsw;
  }


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 29 08:01:19 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Jan 2009 08:01:19 -0600
Subject: [LLVMbugs] [Bug 3437] New: clang creates preprocessed file that is
 refused by itself: # 0
Message-ID: <bug-3437-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3437

           Summary: clang creates preprocessed file that is refused by
                    itself: # 0
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu, rdivacky at freebsd.org


Created an attachment (id=2470)
 --> (http://llvm.org/bugs/attachment.cgi?id=2470)
original preprocessed file

While investigating the message from PR3436, clang refused the original aac.i
file from rdivacky:

aac.i.1:3:3: error: line marker directive requires a positive integer argument
# 0 "../../../dev/aac/aac.c"

After manually removing the # 0 line, clang accepted the file.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 29 09:14:51 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Jan 2009 09:14:51 -0600
Subject: [LLVMbugs] [Bug 3438] New: SROA could peel one level off struct,
	but doesn't
Message-ID: <bug-3438-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3438

           Summary: SROA could peel one level off struct, but doesn't
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


In the following testcase, the alloca could be
turned into an alloca of type %bounds, and one
GEP index dropped.

        %wrapper = type { %bounds }
        %bounds = type { i32, i32 }

define void @f() {
entry:
        %v = alloca %wrapper, align 8           ; <%wrapper*> [#uses=5]
        %p = getelementptr %wrapper* %v, i32 0, i32 0           ; <%bounds*>
[#uses=1]
        call void @use(%bounds* noalias sret %p)
        ret void
}

declare void @use(%bounds* noalias sret)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 29 11:21:23 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Jan 2009 11:21:23 -0600
Subject: [LLVMbugs] [Bug 3253] switch lowering places compare in different
 block than shift
In-Reply-To: <bug-3253-206@http.llvm.org/bugs/>
Message-ID: <200901291721.n0THLN5T009811@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3253


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #20 from Chris Lattner <clattner at apple.com>  2009-01-29 11:21:22 ---
I think that Dan's recent patch finished this off!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 29 11:42:35 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Jan 2009 11:42:35 -0600
Subject: [LLVMbugs] [Bug 3437] clang creates preprocessed file that is
 refused by itself: # 0
In-Reply-To: <bug-3437-206@http.llvm.org/bugs/>
Message-ID: <200901291742.n0THgZT3010715@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3437


Ted Kremenek <kremenek at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|llvmbugs at cs.uiuc.edu        |kremenek at apple.com,
                   |                            |clattner at apple.com
          Component|new bugs                    |preprocessor
            Product|new-bugs                    |clang




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 29 12:16:18 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Jan 2009 12:16:18 -0600
Subject: [LLVMbugs] [Bug 3439] New: math wrong-code bug
Message-ID: <bug-3439-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3439

           Summary: math wrong-code bug
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


On Ubuntu Hardy on x86, llvm-gcc -O3 from 1/29/09 miscompiles the code below
such that it prints 1 instead of 0.

#include <stdint.h>

volatile uint32_t g_22;
uint32_t g_46 = 1;

#define safe_mod_macro_int16_t_s_s(si1,si2) \
  (((((int16_t)(si2)) == ((int16_t)0)) || ((((int16_t)(si1)) == (INT16_MIN)) &&
(((int16_t)(si2)) == ((int16_t)-1)))) \
  ? ((int16_t)(si1)) \
  : (((int16_t)(si1)) % ((int16_t)(si2))))

static int16_t
safe_mod_func_int16_t_s_s (int16_t _si1, int16_t _si2)
{
  return safe_mod_macro_int16_t_s_s(_si1,_si2);
}

void func_35 (uint64_t p_36);
void func_35 (uint64_t p_36)
{
  uint32_t l_38;
  for (l_38 = 0; l_38 <= 0; l_38++)
    g_22 = 1 & safe_mod_func_int16_t_s_s (g_46, p_36);
}

int
main (void)
{
  func_35 (1);
  printf ("%d\n", g_22);
  return 0;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Thu Jan 29 12:47:32 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Jan 2009 12:47:32 -0600
Subject: [LLVMbugs] [Bug 3440] New: llvm miscompiles
	gcc.target/i386/sse4_1-mpsadbw.c
Message-ID: <bug-3440-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3440

           Summary: llvm miscompiles gcc.target/i386/sse4_1-mpsadbw.c
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: critical
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: evan.cheng at apple.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2472)
 --> (http://llvm.org/bugs/attachment.cgi?id=2472)
bugpoint-tonotoptimize.bc

Dale noticed lots of gcc sse4.1 tests are failing (when they are run on sse4.1
capable machines!). I looked into one of them: sse4_1-mpsadbw.c. It turns out
to be an optimizer bug.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 06:51:53 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 06:51:53 -0600
Subject: [LLVMbugs] [Bug 3436] clang: cannot codegen constant aggregates
 with designated initializers
In-Reply-To: <bug-3436-206@http.llvm.org/bugs/>
Message-ID: <200901301251.n0UCprrF027874@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3436


Nuno Lopes <nunoplopes at sapo.pt> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #6 from Nuno Lopes <nunoplopes at sapo.pt>  2009-01-30 06:51:53 ---
yes, this is fixed. I confirm it is generating good code (at least for the
tests I run).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 07:54:47 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 07:54:47 -0600
Subject: [LLVMbugs] [Bug 3441] New: mixing ANSI C declaration and K&R
 function is sometimes allowed and sometimes not
Message-ID: <bug-3441-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3441

           Summary: mixing ANSI C declaration and K&R function is sometimes
                    allowed and sometimes not
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: rdivacky at freebsd.org
                CC: llvmbugs at cs.uiuc.edu


sometimes it works, sometimes it doesnt:

lev /tmp$ cat y.c
static int i(int, int);

static int i(j, k)
   int j;
   int k;
{
   return 0;
}
lev /tmp$ ccc y.c
/usr/lib/crt1.o(.text+0x8a): In function `_start':
: undefined reference to `main'
lev /tmp$ cat yy.c
static int i(int, int, char);

static int i(j, k, l)
   int j;
   int k;
   char l;
{
   return 0;
}
lev /tmp$ ccc yy.c
yy.c:3:12: error: conflicting types for 'i'
static int i(j, k, l)
           ^
yy.c:1:12: note: previous declaration is here
static int i(int, int, char);
           ^
2 diagnostics generated.
lev /tmp$ 

there are similar bugs #2412 and #2821, dont know if it's related but it looks
like it is


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 08:56:36 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 08:56:36 -0600
Subject: [LLVMbugs] [Bug 3442] New: Assertion failed: ((i >=
 FTy->getNumParams() || FTy-> getParamType(i) == Params[i]->getType()) && "
 Calling a function with a bad signature!"), function init,
 file Instructions.cpp, line 294.
Message-ID: <bug-3442-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3442

           Summary: Assertion failed: ((i >= FTy->getNumParams() || FTy-
                    >getParamType(i) == Params[i]->getType()) && "Calling a
                    function with a bad signature!"), function init, file
                    Instructions.cpp, line 294.
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: rdivacky at freebsd.org
                CC: llvmbugs at cs.uiuc.edu


this assertion triggers when compiling the attached is compiled via clang


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 09:27:12 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 09:27:12 -0600
Subject: [LLVMbugs] [Bug 3443] New: recent regression: undeclared function
 isn' t infered the right type
Message-ID: <bug-3443-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3443

           Summary: recent regression: undeclared function isn't infered the
                    right type
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


I've here a recent regression in clang:

/cvs/php5/Zend/zend_ini_parser.c:1192:34: error: invalid operands to binary
expression ('int' and 'char *')
  return yystpcpy (yyres, yystr) - yyres;
         ~~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~~~


code:

unsigned yytnamerr(char *yyres, char *yystr) {
        return stpcpy(yyres, yystr) - yyres;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 09:57:45 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 09:57:45 -0600
Subject: [LLVMbugs] [Bug 3444] New: unterminated swtich statement crashes
	clang
Message-ID: <bug-3444-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3444

           Summary: unterminated swtich statement crashes clang
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


clang -fsyntax-only crashes with:

void foo() {
  switch (1) {
    case 2:


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 10:42:48 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 10:42:48 -0600
Subject: [LLVMbugs] [Bug 3442] Assertion failed: ((i >= FTy->getNumParams()
 || FTy-> getParamType(i) == Params[i]->getType()) && " Calling a function
 with a bad signature!"), function init, file Instructions.cpp, line 294.
In-Reply-To: <bug-3442-206@http.llvm.org/bugs/>
Message-ID: <200901301642.n0UGgmOB004258@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3442


Anders Carlsson <andersca at mac.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |andersca at mac.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Anders Carlsson <andersca at mac.com>  2009-01-30 10:42:47 ---
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090126/011565.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 11:51:12 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 11:51:12 -0600
Subject: [LLVMbugs] [Bug 3445] New: Making TargetData optional for passes
Message-ID: <bug-3445-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3445

           Summary: Making TargetData optional for passes
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicolas.geoffray at lip6.fr
                CC: llvmbugs at cs.uiuc.edu


Currently, optimization passes call getAnalysis<TargetData> when they want
architecture specific informations. For llvm-gcc that's OK, because the
generated bitcode is architecture-specific. However, a TargetData should be
optional for languages like eg Java if we want the bitcode result of
optimization passes to continue to be portable.

Doing this involves two things in LLVM passes:

1) Calling getAnalysisIfAvailable<TargetData> instead of
getAnalysis<TargetData>. If no TargetData are specified, getAnalysisIfAvailable
just returns null.

2) Rewriting passes that assume a non-null TargetData (like instcombine) to
only perform TargetData dependant optimizations if the TargetData is available.

Here's the discussion on the list:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090112/072146.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 13:43:03 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 13:43:03 -0600
Subject: [LLVMbugs] [Bug 3446] New: aliases with different types
Message-ID: <bug-3446-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3446

           Summary: aliases with different types
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: minor
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


$ llvm-gcc testcase-min.i  -c -O1
testcase-min.i:16: error:  Alias ???__thread_start__key??? used with invalid
type!

Testcase is from dietlibc, gcc accepts this:

typedef signed int int32_t;
     typedef int32_t pid_t;
     typedef struct _pthread_descr_struct*_pthread_descr;
     typedef int pthread_once_t;
     struct _pthread_descr_struct {
     pid_t pid;
   }
     __attribute__((aligned(32)));
     typedef struct __thread_manager_func {
   }
     __thread_manager_func;
     static _pthread_descr manager_thread;
     static pthread_once_t __thread_started=0;
     static int __thread_nop() {
   }
     void __thread_start__key(_pthread_descr td)
__attribute__((weak,alias("__thread_nop")));
     static void* __managed_start(void*arg) {
     _pthread_descr td;
     __thread_start__key(td);
   }
     static void __manager_thread_init() {
     char*stack;
     if
((manager_thread->pid=__clone(__managed_start,stack,(0x00000100|0x00000200|0x00000400|0x00000800)|((64
-1)-1),manager_thread))==-1) __libc_exit(43);
   }
     static void __thread_init() {
     __manager_thread_init();
   }
     static int __MGR_send(void(*f)(void*),void*arg) {
     __pthread_once(&__thread_started,__thread_init);
     return ({
  long ret;
  ret;
  }
  );
   }
     int __thread_send_manager(void(*f)(void*),void*arg)
__attribute__((alias("__MGR_send")))


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 15:24:01 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 15:24:01 -0600
Subject: [LLVMbugs] [Bug 3447] New: miscompilation of union { long double }
Message-ID: <bug-3447-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3447

           Summary: miscompilation of union { long double }
           Product: clang
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


clang miscompiles the following:
--
ddunbar at lordcrumb:return-types$ cat fail.c
#include <stdio.h>
typedef union T0 { long double field0; } T0;
static T0 T0_values[] = {
        { (long double) 32 },
        { (long double) 32 },
};
int main(int argc, char **argv) {
  printf("  fn424_RV: RV.field0 = %d\n", T0_values[1].field0 == (long double)
32);
  return 0;
}
ddunbar at lordcrumb:return-types$ for x in gcc xcc; do echo "-- $x --"; $x
-std=c99 -m64 fail.c && ./a.out; done
-- gcc --
  fn424_RV: RV.field0 = 1
-- xcc --
  fn424_RV: RV.field0 = 0
--


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 15:45:16 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 15:45:16 -0600
Subject: [LLVMbugs] [Bug 3032] Switch instruction gives an assertion error
 in SelectionDAGLowering:: handleBTSplitSwitchCase
In-Reply-To: <bug-3032-206@http.llvm.org/bugs/>
Message-ID: <200901302145.n0ULjGeb016240@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3032


Anton Korobeynikov <asl at math.spbu.ru> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #8 from Anton Korobeynikov <asl at math.spbu.ru>  2009-01-30 15:45:15 ---
This was fixed just before New Year 2009


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 15:51:09 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 15:51:09 -0600
Subject: [LLVMbugs] [Bug 3446] aliases with different types
In-Reply-To: <bug-3446-206@http.llvm.org/bugs/>
Message-ID: <200901302151.n0ULp9FW016480@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3446


Anton Korobeynikov <asl at math.spbu.ru> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Anton Korobeynikov <asl at math.spbu.ru>  2009-01-30 15:51:09 ---
Fixed in
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090126/072970.html

I decided not to put a testcase, since it seems to be platform & compilation
mode-dependent.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 15:54:14 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 15:54:14 -0600
Subject: [LLVMbugs] [Bug 3438] SROA could peel one level off struct,
	but doesn't
In-Reply-To: <bug-3438-206@http.llvm.org/bugs/>
Message-ID: <200901302154.n0ULsE9r016624@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3438


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |WONTFIX




--- Comment #11 from Duncan Sands <baldrick at free.fr>  2009-01-30 15:54:13 ---
This is more bother than it's worth.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 17:19:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 17:19:27 -0600
Subject: [LLVMbugs] [Bug 2862] [sema] vector types differing only in
 signedness should be compatible
In-Reply-To: <bug-2862-206@http.llvm.org/bugs/>
Message-ID: <200901302319.n0UNJRKv019946@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2862


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #6 from Chris Lattner <clattner at apple.com>  2009-01-30 17:19:27 ---
Fixed by anders here:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090126/011594.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 17:24:10 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 17:24:10 -0600
Subject: [LLVMbugs] [Bug 2930] clang does not allow VLA in unions/structs as
 gcc does ( violates C99)
In-Reply-To: <bug-2930-206@http.llvm.org/bugs/>
Message-ID: <200901302324.n0UNOAo1020150@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2930


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |WONTFIX




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-30 17:24:10 ---
I'm sorry, but this extension is just too insane to support.  I would much
rather that we never have this in Clang.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Fri Jan 30 18:25:52 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Jan 2009 18:25:52 -0600
Subject: [LLVMbugs] [Bug 3448] New: clang does not match gcc/llvm-gcc in
 returning struct { < 2 x float> }
Message-ID: <bug-3448-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3448

           Summary: clang does not match gcc/llvm-gcc in returning struct {
                    <2 x float> }
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedclangbugs at nondot.org
        ReportedBy: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


gcc/llvm-gcc return T0 in the following code in memory 
--
ddunbar at lordcrumb:return-types$ cat test.183.a.c
cat test.183.a.c
// ../ABITestGen.py --max-args 0 -o test.183.a.c -T test.183.b.c -D
test.183.driver.c --min=183 --count=1
// Generated: 2009-01-30 16:12
// Cardinality of function generator: inf
// Cardinality of type generator: inf

#include <stdio.h>

typedef float T1 __attribute__ ((vector_size (8)));
typedef struct T0 { T1 field0; } T0;
T0 T0_retval;
T0 fn183(void) {
  return T0_retval;
}

ddunbar at lordcrumb:return-types$ llvm-gcc -m64 -S -emit-llvm -o - test.183.a.c
llvm-gcc -m64 -S -emit-llvm -o - test.183.a.c
; ModuleID = 'test.183.a.c'
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
target triple = "x86_64-apple-darwin10.0"
        %struct.T0 = type { <2 x float> }
@T0_retval = common global %struct.T0 zeroinitializer           ; <%struct.T0*>
[#uses=1]

define void @fn183(%struct.T0* noalias sret %agg.result) nounwind {
entry:
        %0 = getelementptr %struct.T0* %agg.result, i32 0, i32 0               
; <<2 x float>*> [#uses=1]
        %1 = load <2 x float>* getelementptr (%struct.T0* @T0_retval, i32 0,
i32 0), align 8            ; <<2 x float>> [#uses=1]
        store <2 x float> %1, <2 x float>* %0, align 8
        br label %return

return:         ; preds = %entry
        ret void
}
--
while clang returns it in registers.

In my opinion this is clearly a bug in gcc/llvm-gcc and is both totally against
the spirit of the ABI (if not technically incorrect) and something which could
matter in practice.

icc doesn't support <2 x float>, so no help there.

My opinion is we should understand why gcc is classifying this this way to
determine if there is a good reason, and see if this has been recognized as a
gcc bug upstream.

Even if we decide to be bug compatible, we need to understand how gcc is
computing this because I cannot figure out what gcc is actually doing to come
up with this result. It correctly passes <2 x float> as a return type, so it is
not just a matter of the classification being wrong.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 00:38:54 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 00:38:54 -0600
Subject: [LLVMbugs] [Bug 3449] New: miscompilation of i512
Message-ID: <bug-3449-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3449

           Summary: miscompilation of i512
           Product: libraries
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: gohman at apple.com, baldrick at free.fr, llvmbugs at cs.uiuc.edu


sumarrray-dbl now fails because SROA is promoting a struct to an i512 and
codegen is shifting by the wrong amount.  Here's a testcase:

define void @test(<8 x double> *%P, i64* %Q) nounwind {
        %A = load <8 x double>* %P
        %B = bitcast <8 x double> %A to i512            ; <i512> [#uses=2]
        %C = lshr i512 %B, 448          ; <i512> [#uses=1]
        %D = trunc i512 %C to i64               ; <i64> [#uses=1]
        volatile store i64 %D, i64* %Q
        ret void
}

I get:
$ llvm-as < ~/t.ll | llc -march=x86-64

_test:
        movq    24(%rdi), %rax
        movq    %rax, (%rsi)
        ret

bit 448 = byte 56 not byte 24.  It looks like it is off by 32 or maybe
modulusing by 32 or something weird like that?


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 00:43:07 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 00:43:07 -0600
Subject: [LLVMbugs] [Bug 3450] New: codegen assertion on i960 value
Message-ID: <bug-3450-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3450

           Summary: codegen assertion on i960 value
           Product: libraries
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2480)
 --> (http://llvm.org/bugs/attachment.cgi?id=2480)
testcase 

Benchmarks/McCat/08-main is now failing because SROA makes an i960 and codegen
can't handle it:

SelectionDAG.cpp:863: failed assertion `(EltVT.getSizeInBits() >= 64 ||
(uint64_t)((int64_t)Val >> EltVT.getSizeInBits()) + 1 < 2) && "getConstant with
a uint64_t value that doesn't fit in the type!"'

This is dying in SelectionDAG::getConstant


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 02:34:34 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 02:34:34 -0600
Subject: [LLVMbugs] [Bug 3449] miscompilation of i512
In-Reply-To: <bug-3449-206@http.llvm.org/bugs/>
Message-ID: <200901310834.n0V8YYux006218@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3449


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-31 02:34:34 ---


*** This bug has been marked as a duplicate of bug 3401 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 02:36:07 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 02:36:07 -0600
Subject: [LLVMbugs] [Bug 3450] codegen assertion on i960 value
In-Reply-To: <bug-3450-206@http.llvm.org/bugs/>
Message-ID: <200901310836.n0V8a7Ot006633@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3450


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #3 from Duncan Sands <baldrick at free.fr>  2009-01-31 02:36:06 ---


*** This bug has been marked as a duplicate of bug 3401 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 02:50:20 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 02:50:20 -0600
Subject: [LLVMbugs] [Bug 3451] New: Instcombine should break up insane
	integer phi nodes
Message-ID: <bug-3451-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3451

           Summary: Instcombine should break up insane integer phi nodes
           Product: libraries
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: code-quality
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2481)
 --> (http://llvm.org/bugs/attachment.cgi?id=2481)
example

188.ammp has a few aggregates promoted to large integers, but they are live
across blocks.  If they were local to a block instcombine would be able to
eliminate the bitcast+zext+shift+shift+trunc+bitcast dance that is happening,
but it can't.  We should clean this up for a major perf boost (assuming the
loop is hot).  Search the attached .bc file for i576.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 03:33:30 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 03:33:30 -0600
Subject: [LLVMbugs] [Bug 3452] New:
	InstCombiner::SimplifyDemandedInstructionBit hangs
Message-ID: <bug-3452-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3452

           Summary: InstCombiner::SimplifyDemandedInstructionBit hangs
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=2482)
 --> (http://llvm.org/bugs/attachment.cgi?id=2482)
testcase: llvm-as <x.ll | opt -O1

llvm-gcc bootstrap hangs here:

/home/edwin/llvm-bootstrap/llvm-gcc-obj/./gcc/cc1 -quiet -I. -I.
-I../../llvm-gcc4.2/gcc -I../../llvm-gcc4.2/gcc/.
-I../../llvm-gcc4.2/gcc/../include -I../../llvm-gcc4.2/gcc/../libcpp/include
-I../../llvm-gcc4.2/gcc/../libdecnumber -I../libdecnumber
-I/home/edwin/llvm-bootstrap//llvm-obj//include
-I/home/edwin/llvm-bootstrap/llvm/include -iprefix
/home/edwin/llvm-bootstrap/llvm-gcc-obj/gcc/../lib/gcc/x86_64-unknown-linux-gnu/4.2.1/
-isystem /home/edwin/llvm-bootstrap/llvm-gcc-obj/./gcc/include -DIN_GCC
-DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -DL_moddi3
-DHIDE_EXPORTS -isystem
/home/edwin/llvm-bootstrap/llvm-gcc-obj/../install/x86_64-unknown-linux-gnu/include
-isystem
/home/edwin/llvm-bootstrap/llvm-gcc-obj/../install/x86_64-unknown-linux-gnu/sys-include
-isystem ./include ../../llvm-gcc4.2/gcc/libgcc2.c -quiet -dumpbase libgcc2.c
-mtune=generic -auxbase-strip libgcc/./_moddi3.o -g -g -O2 -O2 -O2 -W -Wall
-Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -Wold-style-definition
-fPIC -fexceptions -fnon-call-exceptions -fvisibility=hidden

It also reproduces with -O1, and doesn't with -O0.

If I do -O0 -emit-llvm-bc, and then 'opt -O1 | llc -relocation-model=pic' it
doesn't reproduce.

However I was able to dump the module from gdb.

A backtrace:

#0  0x0000000000a630c1 in llvm::APInt::clearUnusedBits (this=0x7fffc0f31c70) at
/home/edwin/llvm-bootstrap/llvm/include/llvm/ADT/APInt.h:127
#1  0x0000000000a744b6 in APInt (this=0x7fffc0f31c80, numBits=1, val=0,
isSigned=false) at /home/edwin/llvm-bootstrap/llvm/include/llvm/ADT/AP
Int.h:219
#2  0x0000000000fa36d6 in (anonymous
namespace)::InstCombiner::SimplifyDemandedUseBits ()
#3  0x0000000000faab51 in (anonymous
namespace)::InstCombiner::SimplifyDemandedBits ()
#4  0x0000000000fa4105 in (anonymous
namespace)::InstCombiner::SimplifyDemandedUseBits ()
#5  0x0000000000fa69cf in (anonymous
namespace)::InstCombiner::SimplifyDemandedInstructionBits ()
#6  0x0000000000faeb35 in (anonymous namespace)::InstCombiner::visitAnd ()
#7  0x0000000000fb9055 in llvm::InstVisitor<(anonymous
namespace)::InstCombiner, llvm::Instruction*>::visit ()
#8  0x0000000000fba7c0 in (anonymous namespace)::InstCombiner::DoOneIteration
()
#9  0x0000000000fbaeae in (anonymous namespace)::InstCombiner::runOnFunction ()
#10 0x000000000110ff7b in llvm::FPPassManager::runOnFunction ()
#11 0x0000000001059450 in (anonymous namespace)::CGPassManager::runOnModule ()
#12 0x000000000110fb57 in llvm::MPPassManager::runOnModule ()
#13 0x000000000110fd56 in llvm::PassManagerImpl::run ()
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, DW_AT_name missing from DW_TAG_base_type
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, DW_AT_type missing from DW_TAG_subrange_type.
During symbol reading, DW_AT_type missing from DW_TAG_subrange_type.
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, DW_AT_type missing from DW_TAG_subrange_type.
#14 0x0000000000a1d43e in llvm_asm_file_end () at
../../llvm-gcc4.2/gcc/llvm-backend.cpp:716
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
During symbol reading, unsupported tag: 'DW_TAG_const_type'.
During symbol reading, unsupported tag: 'DW_TAG_const_type'.
During symbol reading, DW_AT_type missing from DW_TAG_subrange_type.
During symbol reading, DW_AT_type missing from DW_TAG_subrange_type.
During symbol reading, unsupported tag: 'DW_TAG_const_type'.
During symbol reading, unsupported tag: 'DW_TAG_const_type'.
During symbol reading, DW_AT_type missing from DW_TAG_subrange_type.
#15 0x0000000000987042 in compile_file () at
../../llvm-gcc4.2/gcc/toplev.c:1191
#16 0x0000000000988d3a in do_compile () at ../../llvm-gcc4.2/gcc/toplev.c:2263
#17 0x0000000000988da1 in toplev_main (argc=49, argv=0x7fffc0f34e08) at
../../llvm-gcc4.2/gcc/toplev.c:2295
#18 0x00000000004da9c9 in main (argc=49, argv=0x7fffc0f34e08) at
../../llvm-gcc4.2/gcc/llvm-main.cpp:39


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 09:51:18 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 09:51:18 -0600
Subject: [LLVMbugs] [Bug 3401] Wrong code for i288 addition
In-Reply-To: <bug-3401-206@http.llvm.org/bugs/>
Message-ID: <200901311551.n0VFpIpQ029584@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3401


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #13 from Duncan Sands <baldrick at free.fr>  2009-01-31 09:51:17 ---
Fixed here using a different strategy:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090126/073021.html


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 13:06:19 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 13:06:19 -0600
Subject: [LLVMbugs] [Bug 3451] Instcombine should break up insane integer
	phi nodes
In-Reply-To: <bug-3451-206@http.llvm.org/bugs/>
Message-ID: <200901311906.n0VJ6JQ9004828@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3451


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <clattner at apple.com>  2009-01-31 13:06:18 ---
Fixed in 
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090126/073023.html

Thanks!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 13:07:33 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 13:07:33 -0600
Subject: [LLVMbugs] [Bug 3452] InstCombiner::SimplifyDemandedInstructionBit
	hangs
In-Reply-To: <bug-3452-206@http.llvm.org/bugs/>
Message-ID: <200901311907.n0VJ7XOm004905@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3452


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #3 from Chris Lattner <clattner at apple.com>  2009-01-31 13:07:33 ---
Fixed in 
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090126/073023.html

Thanks!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 13:07:28 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 13:07:28 -0600
Subject: [LLVMbugs] [Bug 3451] Instcombine should break up insane integer
	phi nodes
In-Reply-To: <bug-3451-206@http.llvm.org/bugs/>
Message-ID: <200901311907.n0VJ7S83004883@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=3451


Chris Lattner <clattner at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |




--- Comment #2 from Chris Lattner <clattner at apple.com>  2009-01-31 13:07:28 ---
Oops, wrong bug.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

From bugzilla-daemon at cs.uiuc.edu  Sat Jan 31 15:11:27 2009
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Jan 2009 15:11:27 -0600
Subject: [LLVMbugs] [Bug 3453] New: assert on and i288
Message-ID: <bug-3453-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=3453

           Summary: assert on and i288
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicholas at mxc.ca
                CC: llvmbugs at cs.uiuc.edu


The attached testcase is reduced from clamav on i686 Linux.

llc: /home/nicholas/llvm-commit/include/llvm/CodeGen/ValueTypes.h:450: uint64_t
llvm::MVT::getIntegerVTBitMask() const: Assertion `EltVT.getSizeInBits() <= 64
&& "getIntegerVTBitMask doesn't use APInt!"' failed.
[New Thread 0xf7cce6c0 (LWP 22303)]

Program received signal SIGABRT, Aborted.
[Switching to Thread 0xf7cce6c0 (LWP 22303)]
0xffffe425 in __kernel_vsyscall ()
(gdb) bt
#0  0xffffe425 in __kernel_vsyscall ()
#1  0xf7cfb640 in raise () from /lib/i686/cmov/libc.so.6
#2  0xf7cfd018 in abort () from /lib/i686/cmov/libc.so.6
#3  0xf7cf45be in __assert_fail () from /lib/i686/cmov/libc.so.6
#4  0x08940fcd in llvm::MVT::getIntegerVTBitMask (this=0xff9ce160)
    at /home/nicholas/llvm-commit/include/llvm/CodeGen/ValueTypes.h:450
#5  0x08974e5e in llvm::SelectionDAG::getNode (this=0x8f63608, Opcode=85, DL=
      {Idx = 0}, VT={{V = 6, SimpleTy = llvm::MVT::i128, LLVMTy = 0x6}}, N1=
      {Node = 0x8f7b05c, ResNo = 0}, N2={Node = 0x8f7b7cc, ResNo = 0})
    at SelectionDAG.cpp:2775
#6  0x08a452d1 in llvm::DAGTypeLegalizer::ExpandIntRes_Logical (
    this=0xff9ce350, N=0x8f7e1c4, Lo=@0xff9ce208, Hi=@0xff9ce200)
    at LegalizeIntegerTypes.cpp:1558
#7  0x08a528aa in llvm::DAGTypeLegalizer::ExpandIntegerResult (
    this=0xff9ce350, N=0x8f7e1c4, ResNo=0) at LegalizeIntegerTypes.cpp:996
#8  0x08a2cbfb in llvm::DAGTypeLegalizer::run (this=0xff9ce350)
    at LegalizeTypes.cpp:232
#9  0x08a2d719 in llvm::SelectionDAG::LegalizeTypes (this=0x8f63608)
    at LegalizeTypes.cpp:1055
#10 0x089c3077 in llvm::SelectionDAGISel::CodeGenAndEmitDAG (this=0x8f62ed0)
    at SelectionDAGISel.cpp:595
#11 0x089c5940 in llvm::SelectionDAGISel::SelectBasicBlock (this=0x8f62ed0, 
    LLVMBB=0x8f53ed8, Begin=
      {<bidirectional_iterator<llvm::Instruction, int>> =
{<std::iterator<std::bidirectional_iterator_tag, llvm::Instruction, int,
llvm::Instruction*, llvm::Instruction&>> = {<No data fields>}, <No data
fields>}, NodePtr = 0x8f55cd8}, End=
      {<bidirectional_iterator<llvm::Instruction, int>> =
{<std::iterator<std::bidirectional_iterator_tag, llvm::Instruction, int,
llvm::Instruction*, llvm::Instruction&>> = {<No data fields>}, <No data
fields>}, NodePtr = 0x8f55d68})
    at SelectionDAGISel.cpp:498
#12 0x089c6320 in llvm::SelectionDAGISel::SelectAllBasicBlocks (
    this=0x8f62ed0, Fn=@0x8f55108, MF=@0x8f73460, MMI=0x8f6e010, DW=0x8f67238, 
    TII=@0x8f56bc8) at SelectionDAGISel.cpp:854
#13 0x089c70ac in llvm::SelectionDAGISel::runOnFunction (this=0x8f62ed0, 
    Fn=@0x8f55108) at SelectionDAGISel.cpp:327
#14 0x08cbb659 in llvm::FPPassManager::runOnFunction (this=0x8f53ae0, 
    F=@0x8f55108) at PassManager.cpp:1323
#15 0x08cbc15c in llvm::FunctionPassManagerImpl::run (this=0x8f534c0, 
    F=@0x8f55108) at PassManager.cpp:1281
#16 0x08cbc2eb in llvm::FunctionPassManager::run (this=0xff9ce958, 
    F=@0x8f55108) at PassManager.cpp:1226
#17 0x084109a3 in main (argc=3, argv=0xff9cea74) at llc.cpp:317
(gdb) up 4
#4  0x08940fcd in llvm::MVT::getIntegerVTBitMask (this=0xff9ce160)
    at /home/nicholas/llvm-commit/include/llvm/CodeGen/ValueTypes.h:450
450                  "getIntegerVTBitMask doesn't use APInt!");
(gdb) p EltVT.getSizeInBits()
$1 = 128


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.