From bugzilla-daemon at cs.uiuc.edu  Fri Feb  1 05:08:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 1 Feb 2008 05:08:12 -0600
Subject: [LLVMbugs] [Bug 1971] New: EQUIVALENCE not supported in
	llvm-gfortran
Message-ID: <bug-1971-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1971

           Summary: EQUIVALENCE not supported in llvm-gfortran
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1365)
 --> (http://llvm.org/bugs/attachment.cgi?id=1365)
Testcase

Current gimple=>LLVM IR lowering does not support handling of fortran
EQUIVALENCE blocks. This looks like the last 'lack of feature' bug know so far.

Testcase from polyhedron 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  Fri Feb  1 15:29:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 1 Feb 2008 15:29:48 -0600
Subject: [LLVMbugs] [Bug 1970] CBackend doesn't handle properly unaligned
	load/stores
In-Reply-To: <bug-1970-206@http.llvm.org/bugs/>
Message-ID: <200802012129.m11LTmgT010944@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1970


Lauro Venancio <lauro.venancio at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |2.3




--- Comment #1 from Lauro Venancio <lauro.venancio at gmail.com>  2008-02-01 15:29:47 ---
Fixed.

http://llvm.org/viewvc/llvm-project?rev=46646&view=rev


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  1 17:56:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 1 Feb 2008 17:56:19 -0600
Subject: [LLVMbugs] [Bug 1972] New: Invalid alloca instcombine
Message-ID: <bug-1972-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1972

           Summary: Invalid alloca instcombine
           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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


@g = global i32* zeroinitializer

define i1 @b(i32 %t) {
entry:
  %r = mul i32 %t, 2863311532
  %alloc = alloca i8, i32 %r
  %casted = bitcast i8* %alloc to i32*
  store i32* %casted, i32** @g
  ret i1 1
}

instcombine combines the alloc to
%tmp = mul i32 %t, 715827883            ; <i32> [#uses=1]
%alloc = alloca i32, i32 %tmp           ; <i32*> [#uses=1]

which is wrong. In this case, this turns what might be an allocation of 4 bytes
(if t is 3) into an extremely large allocation (which would crash in an
implementation that checked stack allocations, or possibly some 64-bit
implementations); in other cases, which I haven't bothered to figure out the
exact numbers for, it could cause a buffer overflow vulnerability.

The only situation under which this optimization would be correct would be if
the multiply could be guaranteed not to overflow, I think (unless we guarantee
that alloca's do 32-bit multiplication and ignore overflow, which seems a
strange at best).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  1 17:58:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 1 Feb 2008 17:58:45 -0600
Subject: [LLVMbugs] [Bug 1973] New: Fix for PR1942 causes miscompilation of
	clang
Message-ID: <bug-1973-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1973

           Summary: Fix for PR1942 causes miscompilation of clang
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: miscompilation
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu


The (second) fix for PR1942 causes llvm-gcc emit emit memcpy's of things with
ctors, this is very bad.  The -O0 .ll code compiled for
_Z10DoStuffolaRSt3mapIjN5clang13RewriteBufferESt4lessIjESaISt4pairIKjS1_EEESt17_Rb_tree_iteratorIS6_EjRS1_
ends up constructing memtmp, memcpy'ing it to tmp1, then destroying tmp1.

-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  Fri Feb  1 22:34:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 1 Feb 2008 22:34:03 -0600
Subject: [LLVMbugs] [Bug 1972] Invalid alloca instcombine
In-Reply-To: <bug-1972-206@http.llvm.org/bugs/>
Message-ID: <200802020434.m124Y3mu023955@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1972


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-01 22:34:03 ---
Currently the size is always computed as a zero extended 32-bit value, so these
are equivalent.  This is obviously badness for 64-bit hosts: we should 
generalize alloca and malloc to take either a 32-bit or 64-bit integer operand
like GEP.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  2 08:22:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 2 Feb 2008 08:22:07 -0600
Subject: [LLVMbugs] [Bug 1974] New: Anders pass can produce miscompiled code
Message-ID: <bug-1974-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1974

           Summary: Anders pass can produce miscompiled code
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Keywords: miscompilation
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: robert.a.zeh at gmail.com
                CC: llvmbugs at cs.uiuc.edu


The attached bitcode is a little strange.  It's the result of combining a small
expression language with llvm.  The expression language uses dynamic_cast while
building up expressions, and I've inlined the gcc code for dynamic_cast.  I
expect the bitcode will only work when it is attached to the gnu tool chain
because it includes references to the GNU C++ runtime.

The expected (and correct) output of the attached bitcode is an empty file.
After running it through the anders pass, it incorrectly throws an exception. 
Here is what I'm talking about:

Macintosh:GQ robertzeh$ lli bad-2.bc
Macintosh:GQ robertzeh$ 

Fine, life is good.  Now I try to apply the anders pass along with a few other
pieces of code:
Macintosh:GQ robertzeh$ opt -f -anders-aa -load-vn -gcse -die bad-2.bc -o
bad-bad-2.bc
Macintosh:GQ robertzeh$ lli bad-bad-2.bc
terminate called after throwing an instance of 'Q::Exception'
0   lli                                 0x00471f71
_ZN40_GLOBAL__N_Signals.cpp_00000000_257E67F615PrintStackTraceEv + 45
1   lli                                 0x00472317
_ZN40_GLOBAL__N_Signals.cpp_00000000_257E67F613SignalHandlerEi + 323
2   libSystem.B.dylib                   0x93c4097b _sigtramp + 43
3   ???                                 0xffffffff 0x0 + 4294967295
4   libstdc++.6.dylib                   0x943f410c __gxx_personality_v0 + 1108
5   libstdc++.6.dylib                   0x943f414b _ZSt9terminatev + 29
6   libstdc++.6.dylib                   0x943f4261 __cxa_throw + 101
7   ???                                 0x01880358 0x0 + 25690968
8   lli                                 0x0013a2e3
_ZN4llvm15ExecutionEngine17runFunctionAsMainEPNS_8FunctionERKSt6vectorISsSaISsEEPKPKc
+ 1101
9   lli                                 0x00002744 main + 1186
10  lli                                 0x00002236 start + 54

The problem isn't with the exception handling --- the problem is that the code
shouldn't be trying to throw an exception.

After running things through bugpoint (where happyOutput is an empty file):
bugpoint -run-cbe -anders-aa -load-vn -gcse -die bad-2.bc --output happyOutput

....
*** The following function is being miscompiled: 
main__ZN1Q13SememeFactory9CreateAddEPNS_6SememeES2_i_2E_exit_2E_ce
<cbe><gcc>You can reproduce the problem with the command line: 
  lli -load ./bugpoint.safe.bc-CtCTaP.cbe.c.dylib bugpoint.test.bc-wYsqlo
The shared object was created with:
  llc -march=c bugpoint.safe.bc-CtCTaP -o temporary.c
  gcc -xc temporary.c -O2 -o ./bugpoint.safe.bc-CtCTaP.cbe.c.dylib -shared
-fno-strict-aliasing
Macintosh:GQ robertzeh$ 

The bugpoint reduced bitcode is only 2 bytes shorter, so I've attached the
original.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  2 15:04:18 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 2 Feb 2008 15:04:18 -0600
Subject: [LLVMbugs] [Bug 1942] problem with assigning returned object to
	*this
In-Reply-To: <bug-1942-206@http.llvm.org/bugs/>
Message-ID: <200802022104.m12L4IHP028978@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1942


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #9 from Duncan Sands <baldrick at free.fr>  2008-02-02 15:04:15 ---
This needs further investigation: the fix (which copies what gcc appears
to do) breaks C++ code for which the return value has constructors/
destructors, so I've reverted 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  Sat Feb  2 21:33:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 2 Feb 2008 21:33:19 -0600
Subject: [LLVMbugs] [Bug 1975] New: dag isel emitter isels wrong flag result
Message-ID: <bug-1975-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1975

           Summary: dag isel emitter isels wrong flag result
           Product: tools
           Version: 2.2
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: compile-fail
          Severity: normal
          Priority: P2
         Component: TableGen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: evan.cheng at apple.com, llvmbugs at cs.uiuc.edu


The dag isel emitter is generating code that uses RAUW on a the flag result of
the wrong node in some cases.  For example, in the X86 r/m/w pattern for
"adde", the adde node generates a flag.  However, the selector code (generated
by tblgen) looks like this:

  ReplaceUses(SDOperand(N10.Val, 1), SDOperand(ResNode, 0));
  ReplaceUses(SDOperand(N.Val, 1), InFlag);  <---
  ReplaceUses(SDOperand(N.Val, 0), SDOperand(Chain10.Val, Chain10.ResNo));
  return ResNode;

That is replacing the "flag of the store" with the input flag.  This is
incorrect, as the store doesn't have a flag!  I'm committing a patch soon with
a horrible hack that works around this, but this really needs to be fixed.

-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  Sat Feb  2 22:41:40 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 2 Feb 2008 22:41:40 -0600
Subject: [LLVMbugs] [Bug 1976] New: instcombine doesn't fold x*y+x*z to
	x*(y+z)
Message-ID: <bug-1976-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1976

           Summary: instcombine doesn't fold x*y+x*z to x*(y+z)
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: nicholas at mxc.ca
        ReportedBy: nicholas at mxc.ca
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1369)
 --> (http://llvm.org/bugs/attachment.cgi?id=1369)
proposed patch (a little ugly)

Instcombine should be able to fold:

define i8 @test1(i8 %x, i8 %y, i8 %z) {
  %A = mul i8 %x, %y
  %B = mul i8 %x, %z
  %C = add i8 %A, %B
  ret i8 %C
}

into this:

define i8 @test2(i8 %x, i8 %y, i8 %z) {
  %A = add i8 %y, %z
  %B = mul i8 %x, %A
  ret i8 %B
}

It currently implements this only when %y and %z are ConstantInt. That's wrong,
this transformation works in general.

Patch attached. (Note that we shouldn't remove the existing "X*C1 + X*C2 --> X
* (C1+C2)" trafo because it also does shift-left combining.)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  3 01:32:40 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 3 Feb 2008 01:32:40 -0600
Subject: [LLVMbugs] [Bug 1973] Fix for PR1942 causes miscompilation of clang
In-Reply-To: <bug-1973-206@http.llvm.org/bugs/>
Message-ID: <200802030732.m137WeRf014851@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1973


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #5 from Chris Lattner <sabre at nondot.org>  2008-02-03 01:32:40 ---
Duncan fixed this by reverting the patches:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080128/057865.html

I verified that the testcase is compiled to correct code now.  Thanks Duncan!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  3 01:43:17 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 3 Feb 2008 01:43:17 -0600
Subject: [LLVMbugs] [Bug 1976] instcombine doesn't fold x*y+x*z to x*(y+z)
In-Reply-To: <bug-1976-206@http.llvm.org/bugs/>
Message-ID: <200802030743.m137hHoS015183@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1976


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #3 from Nick Lewycky <nicholas at mxc.ca>  2008-02-03 01:43:17 ---
Fixed. Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080128/057877.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 Feb  3 03:00:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 3 Feb 2008 03:00:38 -0600
Subject: [LLVMbugs] [Bug 1966] -pedantic-errors unusable
In-Reply-To: <bug-1966-206@http.llvm.org/bugs/>
Message-ID: <200802030900.m1390c4b025059@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1966


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-02-03 03:00:37 ---
This is gross, but should fix it, please verify:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080128/004083.html

Thanks Neil!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  3 10:34:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 3 Feb 2008 10:34:27 -0600
Subject: [LLVMbugs] [Bug 1949] instcombine doesn't fold away add/icmp ult
In-Reply-To: <bug-1949-206@http.llvm.org/bugs/>
Message-ID: <200802031634.m13GYRa5031677@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1949


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #7 from Nick Lewycky <nicholas at mxc.ca>  2008-02-03 10:34:26 ---
Fixed. Patch here:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080128/057879.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 neil at daikokuya.co.uk  Sun Feb  3 06:12:34 2008
From: neil at daikokuya.co.uk (Neil Booth)
Date: Sun, 3 Feb 2008 21:12:34 +0900
Subject: [LLVMbugs] [Bug 1966] -pedantic-errors unusable
In-Reply-To: <200802030900.m1390c4b025059@zion.cs.uiuc.edu>
References: <bug-1966-206@http.llvm.org/bugs/>
	<200802030900.m1390c4b025059@zion.cs.uiuc.edu>
Message-ID: <20080203121234.GD23617@daikokuya.co.uk>

bugzilla-daemon at cs.uiuc.edu wrote:-

> http://llvm.org/bugs/show_bug.cgi?id=1966
> 
> 
> Chris Lattner <sabre at nondot.org> changed:
> 
>            What    |Removed                     |Added
> ----------------------------------------------------------------------------
>              Status|NEW                         |RESOLVED
>          Resolution|                            |FIXED
> 
> 
> 
> 
> --- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-02-03 03:00:37 ---
> This is gross, but should fix it, please verify:
> http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080128/004083.html
> 
> Thanks Neil!

Works well, thanks.

From bugzilla-daemon at cs.uiuc.edu  Sun Feb  3 12:49:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 3 Feb 2008 12:49:58 -0600
Subject: [LLVMbugs] [Bug 1977] New: Hang compiling llvm-gcc-4.2
Message-ID: <bug-1977-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1977

           Summary: Hang compiling llvm-gcc-4.2
           Product: tools
           Version: trunk
          Platform: DEC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: arthur.loiret at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1376)
 --> (http://llvm.org/bugs/attachment.cgi?id=1376)
llvm-dis crtbegin.bc

Hi,

llvm-gcc-4.2 (trunk) built with llvm-2.2 (trunk) hangs in an endless loop
compiling gcc/crtstuff.c.

The llvm-dis (crtbegin.ll) output is joined, following Duncan Sands's
instructions:

[arthur at sarah sid llvm/build-gcc/gcc]% /home/arthur/llvm/build-gcc/./gcc/xgcc
-emit-llvm -B/home/arthur/llvm/build-gcc/./gcc/
-B/usr/lib/llvm/llvm-gcc-4.2/alpha-linux-gnu/bin/
-B/usr/lib/llvm/llvm-gcc-4.2/alpha-linux-gnu/lib/ -isystem
/usr/lib/llvm/llvm-gcc-4.2/alpha-linux-gnu/include -isystem
/usr/lib/llvm/llvm-gcc-4.2/alpha-linux-gnu/sys-include -O2 -O2 -g -O2  -mieee
-DIN_GCC    -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes
-Wold-style-definition  -isystem ./include  -I. -I. -I../../llvm-gcc-4.2/gcc
-I../../llvm-gcc-4.2/gcc/. -I../../llvm-gcc-4.2/gcc/../include
-I../../llvm-gcc-4.2/gcc/../libcpp/include 
-I../../llvm-gcc-4.2/gcc/../libdecnumber -I../libdecnumber
-I/home/arthur/llvm/llvm-2.2/include -I/usr/lib/llvm/include  -g0
-finhibit-size-directive -fno-inline-functions -fno-exceptions
-fno-zero-initialized-in-bss -fno-toplevel-reorder   \   
          -c ../../llvm-gcc-4.2/gcc/crtstuff.c -DCRT_BEGIN \
          -o crtbegin.bc
[arthur at sarah sid llvm/build-gcc/gcc]% /usr/lib/llvm/bin/llvm-dis crtbegin.bc

?? llc crtbegin.bc ?? hangs in an endless loop in the same way than gcc
bootstrap build.


Arthur.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  3 14:35:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 3 Feb 2008 14:35:43 -0600
Subject: [LLVMbugs] [Bug 1978] New: tramp3d fails at -O0 but passes at -O3
Message-ID: <bug-1978-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1978

           Summary: tramp3d fails at -O0 but passes at -O3
           Product: tools
           Version: 2.2
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: miscompilation
          Severity: normal
          Priority: P2
         Component: llvm-g++
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu


I think this is likely to be a C++ FE bug, probably something similar to the
struct copy thing.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  3 19:57:25 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 3 Feb 2008 19:57:25 -0600
Subject: [LLVMbugs] [Bug 1979] New: Potential loop codegen improvement
Message-ID: <bug-1979-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1979

           Summary: Potential loop codegen improvement
           Product: new-bugs
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Comparing LLVM trunk (clang -emit-llvm-bc | opt -std-compile-opts | llc) and
gcc 4.1 (gcc -O3 -S) for the following code, gcc generates a shorter main loop.

Source:
#include <stdio.h>
#include <time.h>

static int parity(unsigned x) {
  x ^= x >> 1;
  x ^= x >> 2;
  return ((x & 0x11111111) * 0x88888888) >> 31;
}

int main(void) {
  unsigned i;
  unsigned x = 0;
  time_t t1=time(0), t2;
  for (i = 0; i < 0xFFFFFFFF; i+= 1)
    x += parity(i);
  t2=time(0);
  printf("TIME: %d, RESULT: %u\n", t2-t1, x);
}

LLVM's main loop:
.LBB2_3:        # forbody14
        movl    %eax, %edx
        shrl    %edx
        xorl    %eax, %edx
        movl    %edx, %esi
        shrl    $2, %esi
        xorl    %edx, %esi
        andl    $286331153, %esi
        imull   $2290649224, %esi, %esi
        shrl    $31, %esi
        addl    %ecx, %esi
        incl    %eax
        cmpl    $4294967295, %eax
        movl    %esi, %ecx
        jne     .LBB2_3 # forbody14

gcc's main loop:
.L2:
        movl    %ecx, %eax
        shrl    %eax
        xorl    %ecx, %eax
        addl    $1, %ecx
        movl    %eax, %edx
        shrl    $2, %edx
        xorl    %edx, %eax
        andl    $286331153, %eax
        imull   $-2004318072, %eax, %eax
        shrl    $31, %eax
        addl    %eax, %ebx
        cmpl    $-1, %ecx
        jne     .L2

gcc's loop is one instruction shorter; LLVM's loop has an unnecessary movl.

This is an extremely synthetic/worthless benchmark, of course, but this bug
looks like it could affect real code.

(In case anyone's wondering, the included "parity" function computes the parity
of the input integer, i.e. it returns one if the number of set bits is odd, and
zero otherwise.)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  4 04:32:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 04:32:16 -0600
Subject: [LLVMbugs] [Bug 1980] New: DSE handleEndBlock uses freed memory
Message-ID: <bug-1980-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1980

           Summary: DSE handleEndBlock uses freed memory
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


Here's what valgrind says:

Invalid read of size 4
   at 0x85D2CAB: llvm::ilist_iterator<llvm::Instruction>::operator++()
(Instruction.h:223)
   by 0x87ECBDB: llvm::ilist_iterator<llvm::Instruction>::operator++(int)
(ilist:152)
   by 0x8BAE2CA:
llvm::MemoryDependenceAnalysis::removeInstruction(llvm::Instruction*)
(MemoryDependenceAnalysis.cpp:469)
   by 0x8ABD37A: (anonymous namespace)::DSE::handleEndBlock(llvm::BasicBlock&,
llvm::SetVector<llvm::Instruction*, std::vector<llvm::Instruction*,
std::allocator<llvm::Instruction*> >, llvm::SmallSet<llvm::Instruction*, 16>
>&) (DeadStoreElimination.cpp:280)
   by 0x8ABE293: (anonymous namespace)::DSE::runOnBasicBlock(llvm::BasicBlock&)
(DeadStoreElimination.cpp:182)
   by 0x8ABE36E: (anonymous namespace)::DSE::runOnFunction(llvm::Function&)
(DeadStoreElimination.cpp:46)
   by 0x8C434A7: llvm::FPPassManager::runOnFunction(llvm::Function&)
(PassManager.cpp:1184)
   by 0x8C43649: llvm::FPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1204)
   by 0x8C4317D: llvm::MPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1254)
   by 0x8C43335: llvm::PassManagerImpl::run(llvm::Module&)
(PassManager.cpp:1328)
   by 0x8C43387: llvm::PassManager::run(llvm::Module&) (PassManager.cpp:1360)
   by 0x8597C5F: llvm_asm_file_end (llvm-backend.cpp:561)
 Address 0x7483194 is 36 bytes inside a block of size 72 free'd
   at 0x402231C: operator delete(void*) (vg_replace_malloc.c:342)
   by 0x8C2EAD8: llvm::StoreInst::~StoreInst() (Instructions.h:293)
   by 0x87CF1EC: llvm::iplist<llvm::Instruction,
llvm::ilist_traits<llvm::Instruction>
>::erase(llvm::ilist_iterator<llvm::Instruction>) (ilist:368)
   by 0x8C1E7A8: llvm::Instruction::eraseFromParent() (Instruction.cpp:68)
   by 0x8ABE13A: (anonymous namespace)::DSE::runOnBasicBlock(llvm::BasicBlock&)
(DeadStoreElimination.cpp:156)
   by 0x8ABE36E: (anonymous namespace)::DSE::runOnFunction(llvm::Function&)
(DeadStoreElimination.cpp:46)
   by 0x8C434A7: llvm::FPPassManager::runOnFunction(llvm::Function&)
(PassManager.cpp:1184)
   by 0x8C43649: llvm::FPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1204)
   by 0x8C4317D: llvm::MPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1254)
   by 0x8C43335: llvm::PassManagerImpl::run(llvm::Module&)
(PassManager.cpp:1328)
   by 0x8C43387: llvm::PassManager::run(llvm::Module&) (PassManager.cpp:1360)
   by 0x8597C5F: llvm_asm_file_end (llvm-backend.cpp:561)

I hope this is enough: it is not easy to reproduce using opt.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  4 11:23:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 11:23:39 -0600
Subject: [LLVMbugs] [Bug 1942] problem with assigning returned object to
	*this
In-Reply-To: <bug-1942-206@http.llvm.org/bugs/>
Message-ID: <200802041723.m14HNdtb026293@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1942


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #11 from Duncan Sands <baldrick at free.fr>  2008-02-04 11:23:32 ---
Fixed here, take 3:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080204/057898.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 Feb  4 14:20:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 14:20:28 -0600
Subject: [LLVMbugs] [Bug 1981] New: Objects of arbitrary type can be
	bitfields
Message-ID: <bug-1981-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1981

           Summary: Objects of arbitrary type can be bitfields
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


The Ada front-end expects the back-end to handle bitfields of
arbitrary type.  In the following example, a bitfield of struct type:

package Bit_Struct is
   type R is record
      A : Integer;
      B : String (1 .. 3);
   end record;

   type S is record
      An_R : R;
   end record;
   for S use record
      An_R at 0 range 4 .. 60;
   end record;

   An_S : constant S := (An_R => (A => 0, B => "Yo!"));
end;


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  4 14:24:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 14:24:16 -0600
Subject: [LLVMbugs] [Bug 1982] New: Stack backtraces in exceptions not
	supported
Message-ID: <bug-1982-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1982

           Summary: Stack backtraces in exceptions not supported
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


The Ada -E binder option stores a stack traceback in the exception.
This doesn't work with llvm-gcc.  Testcase:

procedure Trace is
begin
   raise Program_Error;
end;

$ gnatmake trace.adb -bargs -E
gcc -c trace.adb
gnatbind -aO./ -E -I- -x trace.ali
gnatlink trace.ali
$ ./trace
Segmentation fault (core dumped)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  4 20:40:18 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 20:40:18 -0600
Subject: [LLVMbugs] [Bug 1962] clang struct init codegen + padding broken
In-Reply-To: <bug-1962-206@http.llvm.org/bugs/>
Message-ID: <200802050240.m152eItX002083@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1962


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #1 from Devang Patel <dpatel at apple.com>  2008-02-04 20:40:17 ---
Fixed. Committed revision 46736.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  4 22:31:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 22:31:00 -0600
Subject: [LLVMbugs] [Bug 1983] New: Teach LLVM about parity
Message-ID: <bug-1983-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1983

           Summary: Teach LLVM about parity
           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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Spun off of Bug 1979, because I don't want to pollute that bug with a lot of
irrelevant discussion.

(In reply to comment #2)
> Eli: some random unrelated thoughts :).  If you're interested in parity, I
> wonder if llvm-gcc generates efficient code for __builtin_parity.  Could you
> check?

It appears llvm-gcc doesn't support __builtin_parity... at least the version at
http://llvm.org/demo/index.cgi (I don't have a trunk build).

LLVM currently generates bad code for llvm.ctpop.i32(x) & 1; it just generates
the ctpop and ands it with 1, which is really slow for cpu's without a popcount
instruction.

(In reply to comment #3)
> It would also be useful to teach instcombine about various idioms for parity,
> turning it into llvm.popct(x)&1

What exactly are common idioms for parity? The only idioms that would be easy
to detect are popcount & 1 and maybe the pure shift+xor method.


I've dome some rough testing of various ways of computing parity using a
benchmark that looks like the testcase for this bug.

For x86, something like the following is probably best; it appears to be as
fast as a lookup table.

static unsigned parity2(unsigned x) {
  int result;
  asm ("movl    %1, %%ecx\n"
       "shrl    $16, %%ecx\n"
       "xorl    %1, %%ecx\n"
       "xorb    %%ch, %%cl\n"
       "setnp   %%cl\n"
       "movzbl  %%cl, %0\n"
        :"=r"(result)    /* output */
        :"g" (x)         /* input */
        :"%ecx"         /* clobbered register */
       );
  return result;
}

I don't know how difficult it would be to make LLVM generate something like
that, though.

Besides that, the fastest method is a lookup table; however, a full lookup
table is at least 256 bytes.  Without a lookup table, the method from my
testcase is the fastest. The method from
http://graphics.stanford.edu/~seander/bithacks.html#ParityParallel is a little
slower. Purely using xor+shift is a little slower than that (it adds up to
being about twice as slow as the asm or table lookup).

That said, the way I'm benchmarking is not at all representative of real-world
code, so my results might be a bit off.

The version of gcc I have converts __builtin_parity into a call to __paritysi2,
a method in libgcc which uses a lookup table; however, I think that's changed 
(http://www.archivum.info/gcc-patches at gcc.gnu.org/2007-02/msg00999.html).

Is there actually any real-world use for 32-bit parity, though?  I just looked
at it because I was curious...


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  4 22:34:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 22:34:34 -0600
Subject: [LLVMbugs] [Bug 1980] DSE handleEndBlock uses freed memory
In-Reply-To: <bug-1980-206@http.llvm.org/bugs/>
Message-ID: <200802050434.m154YYAu005848@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1980


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #4 from Owen Anderson <resistor at mac.com>  2008-02-04 22:34:33 ---
Fixed in r46738.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  4 22:55:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 22:55:30 -0600
Subject: [LLVMbugs] [Bug 1978] tramp3d fails at -O0 but passes at -O3
In-Reply-To: <bug-1978-206@http.llvm.org/bugs/>
Message-ID: <200802050455.m154tURP006641@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1978


Chris Lattner <sabre at nondot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
          Component|llvm-g++                    |Scalar Optimizations
            Product|tools                       |libraries
         Resolution|                            |FIXED
   Target Milestone|---                         |2.3




--- Comment #6 from Chris Lattner <sabre at nondot.org>  2008-02-04 22:55:29 ---
Woot, deleting that code from instcombine fixed it.  Next question, why does
-O0 run instcombine? :)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  4 23:49:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 4 Feb 2008 23:49:43 -0600
Subject: [LLVMbugs] [Bug 1984] New: kaleidoscope tutorial, segfault on dump
Message-ID: <bug-1984-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1984

           Summary: kaleidoscope tutorial, segfault on dump
           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: tbptbp at gmail.com
                CC: llvmbugs at cs.uiuc.edu


The kaleidoscope tutorial starts to segfault at stage #4,
http://llvm.org/docs/tutorial/LangImpl4.html, with a fresh (as of now) trunk
checkout on my x86-64 linux box.

I've exhausted my options, fails in debug and release, with
Target: x86_64-linux-gnu
Configured with: ../src/configure -v
--enable-languages=c,c++,fortran,objc,obj-c++,treelang --prefix=/usr
--enable-shared --with-system-zlib --libexecdir=/usr/lib
--without-included-gettext --enable-threads=posix --enable-nls
--with-gxx-include-dir=/usr/include/c++/4.2 --program-suffix=-4.2
--enable-clocale=gnu --enable-libstdcxx-debug --enable-objc-gc --enable-mpfr
--disable-libmudflap --enable-checking=release --build=x86_64-linux-gnu
--host=x86_64-linux-gnu --target=x86_64-linux-gnu
Thread model: posix
gcc version 4.2.3 (Debian 4.2.3-1)
and

Target: x86_64-linux-gnu
Configured with: ../src/configure -v
--enable-languages=c,c++,fortran,objc,obj-c++,treelang --prefix=/usr
--enable-shared --with-system-zlib --libexecdir=/usr/lib
--without-included-gettext --enable-threads=posix --enable-nls
--with-gxx-include-dir=/usr/include/c++/4.1.3 --program-suffix=-4.1
--enable-__cxa_atexit --enable-clocale=gnu --enable-libstdcxx-debug
--enable-mpfr --enable-checking=release x86_64-linux-gnu
Thread model: posix
gcc version 4.1.3 20080114 (prerelease) (Debian 4.1.2-19)

configured with ../configure --prefix=/usr/local/llvm --enable-jit

Simply ^D right away and

(gdb) bt full
#0  0x00000000007f62cc in llvm::isa<llvm::Type, llvm::Type const*> (Val=@0x28)
    at /usr/src/tbp/llvm/llvm/include/llvm/Support/Casting.h:116
No locals.
#1  0x00000000007f6cbb in llvm::cast<llvm::Type, llvm::Type const*> (Val=@0x28)
    at /usr/src/tbp/llvm/llvm/include/llvm/Support/Casting.h:199
        __PRETTY_FUNCTION__ = "typename llvm::cast_retty<To, From>::ret_type
llvm::cast(const Y&) [with X = llvm::Type, Y = const llvm::Type*]"
#2  0x00000000007f42f5 in fillTypeNameTable (M=0xd49670,
TypeNames=@0x7fffada4de28)
    at /usr/src/tbp/llvm/llvm/lib/VMCore/AsmWriter.cpp:252
        Ty = (const class llvm::Type *) 0x7f5d73
        ST = (const class llvm::TypeSymbolTable &) @0xd49710:
{<llvm::AbstractTypeUser> = {
    _vptr.AbstractTypeUser = 0xd5e2a0}, tmap = {_M_t = {
      _M_impl = {<std::allocator<std::_Rb_tree_node<std::pair<const
std::basic_string<char, std::char_traits<char>, std::allocator<char> >, const
llvm::Type*> > >> =
{<__gnu_cxx::new_allocator<std::_Rb_tree_node<std::pair<const
std::basic_string<char, std::char_traits<char>, std::allocator<char> >, const
llvm::Type*> > >> = {<No data fields>}, <No data fields>}, 
        _M_key_compare = {<std::binary_function<const std::basic_string<char,
std::char_traits<char>, std::allocator<char> >,const std::basic_string<char,
std::char_traits<char>, std::allocator<char> >,bool>> = {<No data fields>}, <No
data fields>}, _M_header = {_M_color = 13923072, _M_parent = 0x0, _M_left =
0x0, _M_right = 0xd5de20}, _M_node_count = 13932304}}}, 
  LastUnique = 0}
        TI = {_M_node = 0x0}
#3  0x00000000007f8b9d in AssemblyWriter (this=0x7fffada4de10, o=@0xd42600,
Mac=@0x7fffada4dd80, M=0xd49670, AAW=0x0)
    at /usr/src/tbp/llvm/llvm/lib/VMCore/AsmWriter.cpp:733
No locals.
#4  0x00000000007f47e3 in llvm::Module::print (this=0xd49670, o=@0xd42600,
AAW=0x0)
    at /usr/src/tbp/llvm/llvm/lib/VMCore/AsmWriter.cpp:1443
        SlotTable = {TheModule = 0xd49670, TheFunction = 0x0, FunctionProcessed
= false, mMap = {_M_t = {
      _M_impl = {<std::allocator<std::_Rb_tree_node<std::pair<const
llvm::Value* const, unsigned int> > >> =
{<__gnu_cxx::new_allocator<std::_Rb_tree_node<std::pair<const llvm::Value*
const, unsigned int> > >> = {<No data fields>}, <No data fields>}, 
        _M_key_compare = {<std::binary_function<const llvm::Value*,const
llvm::Value*,bool>> = {<No data fields>}, <No data fields>}, _M_header =
{_M_color = std::_S_red, _M_parent = 0x0, _M_left = 0x7fffada4dda0, 
          _M_right = 0x7fffada4dda0}, _M_node_count = 0}}}, mNext = 0, fMap =
{_M_t = {
      _M_impl = {<std::allocator<std::_Rb_tree_node<std::pair<const
llvm::Value* const, unsigned int> > >> =
{<__gnu_cxx::new_allocator<std::_Rb_tree_node<std::pair<const llvm::Value*
const, unsigned int> > >> = {<No data fields>}, <No data fields>}, 
        _M_key_compare = {<std::binary_function<const llvm::Value*,const
llvm::Value*,bool>> = {<No data fields>}, <No data fields>}, _M_header =
{_M_color = std::_S_red, _M_parent = 0x0, _M_left = 0x7fffada4ddd8, 
          _M_right = 0x7fffada4ddd8}, _M_node_count = 0}}}, fNext = 0}
        W = {Out = @0xd42600, Machine = @0x7fffada4dd80, TheModule = 0xd49670,
TypeNames = {_M_t = {
      _M_impl = {<std::allocator<std::_Rb_tree_node<std::pair<const llvm::Type*
const, std::basic_string<char, std::char_traits<char>, std::allocator<char> > >
> >> = {<__gnu_cxx::new_allocator<std::_Rb_tree_node<std::pair<const
llvm::Type* const, std::basic_string<char, std::char_traits<char>,
std::allocator<char> > > > >> = {<No data fields>}, <No data fields>}, 
        _M_key_compare = {<std::binary_function<const llvm::Type*,const
llvm::Type*,bool>> = {<No data fields>}, <No data fields>}, _M_header =
{_M_color = std::_S_red, _M_parent = 0x0, _M_left = 0x7fffada4de30, _M_right =
0x7fffada4de30}, 
        _M_node_count = 0}}}, AnnotationWriter = 0x0}
#5  0x00000000007f8bc2 in llvm::Module::print (this=0xd49670, OS=@0xd42600)
    at /usr/src/tbp/llvm/llvm/include/llvm/Module.h:353
No locals.
#6  0x0000000000848b24 in llvm::Module::dump (this=0xd49670) at
/usr/src/tbp/llvm/llvm/lib/VMCore/Module.cpp:94
No locals.
#7  0x00000000004091d3 in main () at toy4.cc:597
        OurModuleProvider = {<llvm::ModuleProvider> = {_vptr.ModuleProvider =
0x9bcb10, 
    TheModule = 0xd49670}, <No data fields>}
        OurFPM = {FPM = 0xd5e100, MP = 0x7fffada4def0}
(gdb)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  5 00:29:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 5 Feb 2008 00:29:19 -0600
Subject: [LLVMbugs] [Bug 1984] kaleidoscope tutorial, segfault on dump
In-Reply-To: <bug-1984-206@http.llvm.org/bugs/>
Message-ID: <200802050629.m156TJVG009411@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1984


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #2 from Chris Lattner <sabre at nondot.org>  2008-02-05 00:29:19 ---
Fixed, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080204/057925.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 Feb  5 00:38:17 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 5 Feb 2008 00:38:17 -0600
Subject: [LLVMbugs] [Bug 1968] clang codegen for static forward declaration
	broken
In-Reply-To: <bug-1968-206@http.llvm.org/bugs/>
Message-ID: <200802050638.m156cHMv009842@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1968


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-05 00:38:16 ---
Fixed in r46742


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  5 16:50:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 5 Feb 2008 16:50:55 -0600
Subject: [LLVMbugs] [Bug 1975] dag isel emitter isels wrong flag result
In-Reply-To: <bug-1975-206@http.llvm.org/bugs/>
Message-ID: <200802052250.m15MotFo020263@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1975


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>  2008-02-05 16:50:52 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080204/057958.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 Feb  5 21:31:06 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 5 Feb 2008 21:31:06 -0600
Subject: [LLVMbugs] [Bug 1985] New: llvm-ar is unkillable and doesn't halt
Message-ID: <bug-1985-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1985

           Summary: llvm-ar is unkillable and doesn't halt
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: major
          Priority: P2
         Component: llvm-ar
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: alenhar2 at uiuc.edu
                CC: llvmbugs at cs.uiuc.edu


any invocation of llvm-ar rcs seems to use 100% of the cpu and never halt.  It
is also unkillable.  This is on linux x86.  even just creating a new archive
with one file in it triggers this behavior.  When run in gdb, Ctl-C doesn't
cause a break.  kill -9 doesn't work either.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  6 09:48:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 09:48:16 -0600
Subject: [LLVMbugs] [Bug 1986] New: Incorrect type generated for anonymous
	struct
Message-ID: <bug-1986-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1986

           Summary: Incorrect type generated for anonymous struct
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: andersca at mac.com
                CC: llvmbugs at cs.uiuc.edu


The following code

struct s
{
 struct {
  int a;
  int b;
 };
 void *d;
};

is translated to 

        %struct.s = type { i8* }

by clang. Also, the warning

Macintosh:clang andersca$ clang -emit-llvm test.c -o -test.c:4:2: warning:
declaration does not declare anything
 struct {
 ^

is emitted. GCC correctly emits the following type

        %struct.anon = type { i32, i32 }
        %struct.s = type { %struct.anon, 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  Wed Feb  6 12:30:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 12:30:08 -0600
Subject: [LLVMbugs] [Bug 1987] New: sema crashes on if() without then
Message-ID: <bug-1987-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1987

           Summary: sema crashes on if() without then
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Keywords: crash-on-invalid
          Severity: normal
          Priority: P2
         Component: AST
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


clang crashes with the following invalid code:


   struct _zend_module_entry {
  }
     typedef struct _zend_function_entry {
  }
    static void zm_globals_dtor_pcre(zend_pcre_globals *pcre_globals ) {


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  6 12:41:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 12:41:37 -0600
Subject: [LLVMbugs] [Bug 1988] New: clang crashes on invalid struct
	definition
Message-ID: <bug-1988-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1988

           Summary: clang crashes on invalid struct definition
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Keywords: crash-on-invalid
          Severity: enhancement
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1379)
 --> (http://llvm.org/bugs/attachment.cgi?id=1379)
test case

clang crashes with the attached invalid code (reduced with delta).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  6 13:13:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 13:13:43 -0600
Subject: [LLVMbugs] [Bug 1985] llvm-ar is unkillable and doesn't halt
In-Reply-To: <bug-1985-206@http.llvm.org/bugs/>
Message-ID: <200802061913.m16JDhj3029857@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1985


Andrew Lenharth <alenhar2 at uiuc.edu> changed:

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




--- Comment #4 from Andrew Lenharth <alenhar2 at uiuc.edu>  2008-02-06 13:13:43 ---
The hangs are caused by a bug introduced in linux 2.6.24.  The patch in the
previous comment fixes 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  Wed Feb  6 13:36:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 13:36:53 -0600
Subject: [LLVMbugs] [Bug 1925] Assertion failed in RegisterScavenging with
 regalloc=local
In-Reply-To: <bug-1925-206@http.llvm.org/bugs/>
Message-ID: <200802061936.m16JarxE030680@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1925


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

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




--- Comment #1 from Evan Cheng <evan.cheng at apple.com>  2008-02-06 13:36:52 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080204/057979.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 Feb  6 13:47:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 13:47:09 -0600
Subject: [LLVMbugs] [Bug 1964] constant expression error
In-Reply-To: <bug-1964-206@http.llvm.org/bugs/>
Message-ID: <200802061947.m16Jl9qS031111@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1964


Nuno Lopes <nunoplopes at sapo.pt> changed:

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




--- Comment #3 from Nuno Lopes <nunoplopes at sapo.pt>  2008-02-06 13:47:09 ---
fixed in
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080204/004113.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 Feb  6 14:05:17 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 14:05:17 -0600
Subject: [LLVMbugs] [Bug 1989] New: " All operands to PHI node must be the
 same type as the PHI node" assertion triggered
Message-ID: <bug-1989-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1989

           Summary: "All operands to PHI node must be the same type as the
                    PHI node" assertion triggered
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Keywords: compile-fail
          Severity: enhancement
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1380)
 --> (http://llvm.org/bugs/attachment.cgi?id=1380)
test case

nuno at linux /tmp/llvm $ clang -emit-llvm a.c
clang: /cvs/llvm/include/llvm/Instructions.h:1312: void
llvm::PHINode::addIncoming(llvm::Value*, llvm::BasicBlock*): Assertion
`getType() == V->getType() && 
"All operands to PHI node must be the same type as the PHI node!"' failed.
clang((anonymous namespace)::PrintStackTrace()+0x22)[0x84eb458]
clang((anonymous namespace)::SignalHandler(int)+0x110)[0x84eb71c]
[0xb7fb9420]
/lib/libc.so.6(abort+0x108)[0xb7d53978]
/lib/libc.so.6(__assert_fail+0xf5)[0xb7d4b7b5]
clang(llvm::PHINode::addIncoming(llvm::Value*,
llvm::BasicBlock*)+0x62)[0x8302a70]
clang[0x8308851]
clang(clang::StmtVisitor<(anonymous namespace)::ScalarExprEmitter,
llvm::Value*>::Visit(clang::Stmt*)+0x9ea)[0x830ba3a]
clang[0x830bc6c]
clang(clang::StmtVisitor<(anonymous namespace)::ScalarExprEmitter,
llvm::Value*>::Visit(clang::Stmt*)+0x900)[0x830b950]
clang[0x8307e8f]
clang[0x83082b9]
clang(clang::StmtVisitor<(anonymous namespace)::ScalarExprEmitter,
llvm::Value*>::Visit(clang::Stmt*)+0xa04)[0x830ba54]
clang[0x83088e7]
clang(clang::StmtVisitor<(anonymous namespace)::ScalarExprEmitter,
llvm::Value*>::Visit(clang::Stmt*)+0x248)[0x830b298]
clang(clang::CodeGen::CodeGenFunction::EmitScalarExpr(clang::Expr
const*)+0x55)[0x830831d]
clang(clang::CodeGen::CodeGenFunction::EmitStmt(clang::Stmt
const*)+0xa0)[0x830ddca]
clang(clang::CodeGen::CodeGenFunction::EmitCompoundStmt(clang::CompoundStmt
const&, bool, llvm::Value*, bool)+0x97)[0x830f45d]
clang(clang::CodeGen::CodeGenFunction::EmitStmt(clang::Stmt
const*)+0x177)[0x830dea1]
clang(clang::CodeGen::CodeGenFunction::GenerateCode(clang::FunctionDecl
const*)+0x3be)[0x82f646c]
clang(clang::CodeGen::CodeGenModule::EmitFunction(clang::FunctionDecl
const*)+0x47)[0x82e3a7f]
clang((anonymous
namespace)::CodeGenerator::HandleTopLevelDecl(clang::Decl*)+0x45)[0x82e281f]


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  6 14:20:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 14:20:11 -0600
Subject: [LLVMbugs] [Bug 1990] New: codegen: structures with bitfields
 trigger an assertion
Message-ID: <bug-1990-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1990

           Summary: codegen: structures with bitfields trigger an assertion
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: compile-fail
          Severity: enhancement
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1381)
 --> (http://llvm.org/bugs/attachment.cgi?id=1381)
test case (delta reduced)

nuno at linux /tmp/llvm $ clang -emit-llvm a.c
clang: /cvs/llvm/include/llvm/ADT/APInt.h:377: static llvm::APInt
llvm::APInt::getBitsSet(uint32_t, uint32_t, uint32_t): Assertion `hiBit <
numBits && "hiBit 
out of range"' failed.
clang((anonymous namespace)::PrintStackTrace()+0x22)[0x84eb458]
clang((anonymous namespace)::SignalHandler(int)+0x110)[0x84eb71c]
[0xb7eea420]
/lib/libc.so.6(abort+0x108)[0xb7c84978]
/lib/libc.so.6(__assert_fail+0xf5)[0xb7c7c7b5]
clang(llvm::APInt::getBitsSet(unsigned int, unsigned int, unsigned
int)+0x33)[0x82fda77]
clang(clang::CodeGen::CodeGenFunction::EmitStoreThroughBitfieldLValue(clang::CodeGen::RValue,
clang::CodeGen::LValue, clang::QualType)+0x110)[0x82f93d2]
clang(clang::CodeGen::CodeGenFunction::EmitStoreThroughLValue(clang::CodeGen::RValue,
clang::CodeGen::LValue, clang::QualType)+0x171)[0x82f9609]
clang[0x830895b]
clang(clang::StmtVisitor<(anonymous namespace)::ScalarExprEmitter,
llvm::Value*>::Visit(clang::Stmt*)+0x248)[0x830b298]
clang(clang::CodeGen::CodeGenFunction::EmitScalarExpr(clang::Expr
const*)+0x55)[0x830831d]
clang(clang::CodeGen::CodeGenFunction::EmitStmt(clang::Stmt
const*)+0xa0)[0x830ddca]
clang(clang::CodeGen::CodeGenFunction::EmitCompoundStmt(clang::CompoundStmt
const&, bool, llvm::Value*, bool)+0x97)[0x830f45d]
clang(clang::CodeGen::CodeGenFunction::EmitStmt(clang::Stmt
const*)+0x177)[0x830dea1]
clang(clang::CodeGen::CodeGenFunction::GenerateCode(clang::FunctionDecl
const*)+0x3be)[0x82f646c]
clang(clang::CodeGen::CodeGenModule::EmitFunction(clang::FunctionDecl
const*)+0x47)[0x82e3a7f]
clang((anonymous
namespace)::CodeGenerator::HandleTopLevelDecl(clang::Decl*)+0x45)[0x82e281f]


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  6 14:32:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 14:32:45 -0600
Subject: [LLVMbugs] [Bug 1991] New: Assertion `E &&
 !hasAggregateLLVMType(E->getType()) && " Invalid scalar expression to
 emit"' failed
Message-ID: <bug-1991-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1991

           Summary: Assertion `E && !hasAggregateLLVMType(E->getType()) &&
                    "Invalid scalar expression to emit"' failed
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: compile-fail
          Severity: enhancement
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1382)
 --> (http://llvm.org/bugs/attachment.cgi?id=1382)
test case (delta reduced)

nuno at linux /tmp/llvm $ clang -emit-llvm a.c
clang: CGExprScalar.cpp:1060: llvm::Value*
clang::CodeGen::CodeGenFunction::EmitScalarExpr(const clang::Expr*): Assertion
`E && 
!hasAggregateLLVMType(E->getType()) && "Invalid scalar expression to emit"'
failed.
clang((anonymous namespace)::PrintStackTrace()+0x22)[0x84eb458]
clang((anonymous namespace)::SignalHandler(int)+0x110)[0x84eb71c]
[0xb7f66420]
/lib/libc.so.6(abort+0x108)[0xb7d00978]
/lib/libc.so.6(__assert_fail+0xf5)[0xb7cf87b5]
clang(clang::CodeGen::CodeGenFunction::EmitScalarExpr(clang::Expr
const*)+0x7e)[0x8308346]
clang[0x82fe1e7]
clang(clang::StmtVisitor<(anonymous namespace)::AggExprEmitter,
void>::Visit(clang::Stmt*)+0x941)[0x830034d]
clang(clang::CodeGen::CodeGenFunction::EmitAggExpr(clang::Expr const*,
llvm::Value*, bool)+0x6c)[0x82fe8c2]
clang(clang::CodeGen::CodeGenFunction::EmitLocalBlockVarDecl(clang::BlockVarDecl
const&)+0x25d)[0x82f8199]
clang(clang::CodeGen::CodeGenFunction::EmitBlockVarDecl(clang::BlockVarDecl
const&)+0x9e)[0x82f84cc]
clang(clang::CodeGen::CodeGenFunction::EmitDecl(clang::Decl
const&)+0xaf)[0x82f85a3]
clang(clang::CodeGen::CodeGenFunction::EmitDeclStmt(clang::DeclStmt
const&)+0x28)[0x830d72e]
clang(clang::CodeGen::CodeGenFunction::EmitStmt(clang::Stmt
const*)+0x272)[0x830df9c]
clang(clang::CodeGen::CodeGenFunction::EmitCompoundStmt(clang::CompoundStmt
const&, bool, llvm::Value*, bool)+0x97)[0x830f45d]
clang(clang::CodeGen::CodeGenFunction::EmitStmt(clang::Stmt
const*)+0x177)[0x830dea1]
clang(clang::CodeGen::CodeGenFunction::GenerateCode(clang::FunctionDecl
const*)+0x3be)[0x82f646c]
clang(clang::CodeGen::CodeGenModule::EmitFunction(clang::FunctionDecl
const*)+0x47)[0x82e3a7f]
clang((anonymous
namespace)::CodeGenerator::HandleTopLevelDecl(clang::Decl*)+0x45)[0x82e281f]


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  6 14:42:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 14:42:55 -0600
Subject: [LLVMbugs] [Bug 1992] New: switch() { case sizeof("str") }
 generates spurious warnings
Message-ID: <bug-1992-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1992

           Summary: switch() { case sizeof("str") } generates spurious
                    warnings
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


The following code generates spurious warnings (duplicate case value):

void func()
{
 switch (2) {
 case sizeof("abc"): break;
 case sizeof("loooong"): break;
 }
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  6 19:20:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 6 Feb 2008 19:20:55 -0600
Subject: [LLVMbugs] [Bug 1993] New: Variants are inappropriately pruned
Message-ID: <bug-1993-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1993

           Summary: Variants are inappropriately pruned
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: major
          Priority: P2
         Component: TableGen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: scottm at aero.org
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1383)
 --> (http://llvm.org/bugs/attachment.cgi?id=1383)
Testcase to demonstrate variant pattern pruning bug

Consider the following instruction patterns:

-------
class SELBVecInst<ValueType vectype>:
  SELBInst<(outs VECREG:$rT), (ins VECREG:$rA, VECREG:$rB, VECREG:$rC),
           [(set (vectype VECREG:$rT),
                 (or (and (vectype VECREG:$rA), (vectype VECREG:$rC)),
                     (and (vectype VECREG:$rB), (vnot (vectype
VECREG:$rC)))))]>;

def SELBv16i8: SELBVecInst<v16i8>;


class SPUVecInst2<ValueType vectype>:
  SELBInst<(outs VECREG:$rT), (ins VECREG:$rA, VECREG:$rB, VECREG:$rC),
           [(set (vectype VECREG:$rT),
                 (or (and (vectype VECREG:$rA), (vnot (vectype VECREG:$rC))),
                     (and (vectype VECREG:$rB), (vectype VECREG:$rC))))]>;

def SELBv16i8_2 : SPUVecInst2<v16i8>;
------

tblgen creates 8 variants for SELBv16i8 and 8 variants for SELBv16i8_v2. The
problem revolves around matching this input:

(or                               N
  (and                            N0
     VECREG:$rA,                  N00
     (xor                         N01
       (build_vector <-1, ...>),  N011, N0110
       VECREG:$rC))               N010
    (and                          N1
       VECREG:$rC,                N10
       VECREG:$rB))               N11

Using tblgen's labeling in the instruction selector, the operands need to match
as follows:

  N.getOpcode() == ISD::OR
  N0.getOpcode() == ISD::AND
  N1.getOpcode() == ISD::AND
  N01.getOpcode() == ISD::XOR
  N011.getOpcode() == BUILD_VECTOR
  Predicate_isImmAllOnesV(N0110)
  N010 == N10

However, this particular instruction matching sequence is never generated
because the pattern to match is considered isomorphic to another variant
pattern (see attached output from tblgen).

Not sure how to fix CodeGenDAGPatterns.cpp or isIsomorphicTo() so that all
variant patterns are actually matched...


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  7 06:17:21 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 7 Feb 2008 06:17:21 -0600
Subject: [LLVMbugs] [Bug 1994] New: llvm-gcc-4.2 build broken on x86-64-linux
Message-ID: <bug-1994-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1994

           Summary: llvm-gcc-4.2 build broken on x86-64-linux
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: evan.cheng at apple.com, llvmbugs at cs.uiuc.edu


c++ -c -g -O2 -DIN_GCC -W -Wall -Wwrite-strings -Wmissing-format-attribute
-fno-common -DHAVE_CONFIG_H -Wno-unused
-DTARGET_NAME=\"x86_64-unknown-linux-gnu\" -I. -I. -I../../llvm-gcc-4.2/gcc
-I../../llvm-gcc-4.2/gcc/. -I../../llvm-gcc-4.2/gcc/../include
-I../../llvm-gcc-4.2/gcc/../libcpp/include 
-I../../llvm-gcc-4.2/gcc/../libdecnumber -I../libdecnumber
-I/home/baldrick/src/llvm/include -I/home/baldrick/src/llvm-objects/include
-DENABLE_LLVM -I/home/baldrick/src/llvm-objects/../llvm/include  -D_DEBUG 
-D_GNU_SOURCE -D__STDC_LIMIT_MACROS   -I. -I. -I../../llvm-gcc-4.2/gcc
-I../../llvm-gcc-4.2/gcc/. -I../../llvm-gcc-4.2/gcc/../include
-I../../llvm-gcc-4.2/gcc/../libcpp/include 
-I../../llvm-gcc-4.2/gcc/../libdecnumber -I../libdecnumber
-I/home/baldrick/src/llvm/include -I/home/baldrick/src/llvm-objects/include \
                ../../llvm-gcc-4.2/gcc/config/i386/llvm-i386.cpp -o llvm-i386.o
../../llvm-gcc-4.2/gcc/config/i386/llvm-i386.cpp: In function ???bool
llvm_x86_should_pass_aggregate_in_memory(tree_node*, const llvm::Type*)???:
../../llvm-gcc-4.2/gcc/config/i386/llvm-i386.cpp:711: error:
???llvm_x86_32_should_pass_aggregate_in_mixed_regs??? was not declared in this
scope


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  7 12:50:36 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 7 Feb 2008 12:50:36 -0600
Subject: [LLVMbugs] [Bug 1990] codegen: structures with bitfields trigger an
	assertion
In-Reply-To: <bug-1990-206@http.llvm.org/bugs/>
Message-ID: <200802071850.m17Ioaaa014883@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1990


Lauro Venancio <lauro.venancio at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|enhancement                 |normal
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Lauro Venancio <lauro.venancio at gmail.com>  2008-02-07 12:50:36 ---
Fixed.

http://llvm.org/viewvc/llvm-project?rev=46856&view=rev


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  8 02:24:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 8 Feb 2008 02:24:27 -0600
Subject: [LLVMbugs] [Bug 1994] llvm-gcc-4.2 build broken on x86-64-linux
In-Reply-To: <bug-1994-206@http.llvm.org/bugs/>
Message-ID: <200802080824.m188ORKf006781@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1994


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #2 from Duncan Sands <baldrick at free.fr>  2008-02-08 02:24:26 ---
This is now fixed, probably by
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080204/058018.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 Feb  8 13:46:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 8 Feb 2008 13:46:27 -0600
Subject: [LLVMbugs] [Bug 1995] New: LoopIndexSplit crashes
Message-ID: <bug-1995-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1995

           Summary: LoopIndexSplit crashes
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: alenhar2 at uiuc.edu
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1390)
 --> (http://llvm.org/bugs/attachment.cgi?id=1390)
reduction

LoopIndexSplit 

/home/andrewl/Research/cfe-4.2/install/bin/llvm-gcc -O1 -emit-llvm -c test.c -o
foo.o

causes

cc1: LoopIndexSplit.cpp:1648:
void<unnamed>::LoopIndexSplit::updatePHINodes(llvm::BasicBlock*,
llvm::BasicBlock*, llvm::BasicBlock*, llvm::PHINode*, llvm::Instruction*):
Assertion `0 && "Unexpected third use of this PHINode"' 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  Fri Feb  8 16:49:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 8 Feb 2008 16:49:53 -0600
Subject: [LLVMbugs] [Bug 1995] LoopIndexSplit crashes
In-Reply-To: <bug-1995-206@http.llvm.org/bugs/>
Message-ID: <200802082249.m18MnrFn007773@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1995


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #1 from Devang Patel <dpatel at apple.com>  2008-02-08 16:49:52 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080204/058032.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 Feb  8 20:39:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 8 Feb 2008 20:39:38 -0600
Subject: [LLVMbugs] [Bug 1996] New: llvm doesn't optmize out loads of undef
Message-ID: <bug-1996-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1996

           Summary: llvm doesn't optmize out loads of undef
           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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


define i32 @b() {
entry:
        %b = alloca i32         ; <i32*> [#uses=3]
        %tmp2 = load i32* %b, align 8           ; <i32> [#uses=1]
        %tmp3 = or i32 %tmp2, 11                ; <i32> [#uses=1]
        %tmp4 = and i32 %tmp3, -5               ; <i32> [#uses=1]
        store i32 %tmp4, i32* %b, align 8
        %call = call i32 (...)* @a( i32* %b )           ; <i32> [#uses=0]
        ret i32 undef
}

declare i32 @a(...)

run through opt -std-compile-opts, doesn't change.  The load should get
optimized to undef because it is easy to prove that nothing has been stored
into that location.

This bug affects bitfields.  Original C code:

struct s {unsigned a : 4, b : 4;};
int a();
int b() {
  struct s b;
  b.a = 11;
  a(&b);
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  8 21:14:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 8 Feb 2008 21:14:22 -0600
Subject: [LLVMbugs] [Bug 1997] New: Cannot codegen simple initializer
Message-ID: <bug-1997-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1997

           Summary: Cannot codegen simple initializer
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


typedef struct
{
  unsigned int line_number;
} c_coords;
void foo (void)
{
  c_coords coords = { 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 Feb  8 21:17:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 8 Feb 2008 21:17:23 -0600
Subject: [LLVMbugs] [Bug 1998] New: We codegen unused stuff
Message-ID: <bug-1998-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1998

           Summary: We codegen unused stuff
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


e.g. inline functions in standard headers, and (according to sabre) unused
static functions.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  8 21:19:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 8 Feb 2008 21:19:53 -0600
Subject: [LLVMbugs] [Bug 1999] New: Rejects valid C90.
Message-ID: <bug-1999-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=1999

           Summary: Rejects valid C90.
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


void g1 (register);


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  8 21:21:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 8 Feb 2008 21:21:30 -0600
Subject: [LLVMbugs] [Bug 2000] New: Rejects valid C90 and C99
Message-ID: <bug-2000-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2000

           Summary: Rejects valid C90 and C99
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: minor
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


extern const void cv1;
const void *f (void)
{
  return &cv1;
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  9 04:04:24 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 9 Feb 2008 04:04:24 -0600
Subject: [LLVMbugs] [Bug 2001] New: clang bitfield assignment produces wrong
	value
Message-ID: <bug-2001-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2001

           Summary: clang bitfield assignment produces wrong value
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


int a(int x) {
struct {signed int x:5;} s;
return s.x = x;
}

clang -emit-llvm-bc | opt -std-compile-opts reduces this to "ret i32 %x", which
is wrong; the result should be something more like what llvm-gcc produces:

define i32 @a(i32 %x) {
entry:
        %tmp14 = shl i32 %x, 27
        %tmp15 = ashr i32 %tmp14, 27
        %tmp1516 = trunc i32 %tmp15 to i8
        %tmp151617 = sext i8 %tmp1516 to i32
        ret i32 %tmp151617
}

(Actually, ideally it would be smart enough to figure out that the trunc+sext
don't do anything, but that's a separate 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  Sat Feb  9 13:57:57 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 9 Feb 2008 13:57:57 -0600
Subject: [LLVMbugs] [Bug 1924] Assertion failure at ParseStmt.cpp:983
In-Reply-To: <bug-1924-206@http.llvm.org/bugs/>
Message-ID: <200802091957.m19JvvOA020570@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1924


Anders Carlsson <andersca at mac.com> changed:

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




--- Comment #2 from Anders Carlsson <andersca at mac.com>  2008-02-09 13:57:57 ---
Fixed in 46910. Thanks for the report!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  9 15:13:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 9 Feb 2008 15:13:27 -0600
Subject: [LLVMbugs] [Bug 2002] New: Incorrect iteration count for loop with
 signed ICMP condition
Message-ID: <bug-2002-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2002

           Summary: Incorrect iteration count for loop with signed ICMP
                    condition
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Keywords: miscompilation
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: wmatyjewicz at fastmail.fm
                CC: nicholas at mxc.ca, llvmbugs at cs.uiuc.edu


For the following LLVM code:

define void @foo(i8 signext %n) {
entry:
        br label %header
header:
        %i = phi i8 [ -100, %entry ], [ %i.inc, %next ]
        %cond = icmp slt i8 %i, %n
        br i1 %cond, label %next, label %return
next:
        %i.inc = add i8 %i, 1
        br label %header
return:
        ret void
}

which more or less represents this high-level loop:

char n = ...;
for (char i = -100; i < n; ++i)
  ;

scalar evolution determines loop iteration count as: ( 0 smax ( 100 +  %n)).

In my opinion this is incorrect. Suppose, %n=100. Then 100 + %n becomes -56
(unsigned 200) and loop iteration count is evaluated to 0.

If someone confirm the bug, I'll try to fix 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  Sat Feb  9 17:05:42 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 9 Feb 2008 17:05:42 -0600
Subject: [LLVMbugs] [Bug 2003] New: Incorrect iteration count for loop with
 unsigned ICMP condition
Message-ID: <bug-2003-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2003

           Summary: Incorrect iteration count for loop with unsigned ICMP
                    condition
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Keywords: miscompilation
          Severity: normal
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: wmatyjewicz at fastmail.fm
                CC: nicholas at mxc.ca, llvmbugs at cs.uiuc.edu


For the following LLVM code:

define void @foo(i32 %n) {
entry:
        br label %header
header:
        %i = phi i32 [ 100, %entry ], [ %i.inc, %next ]
        %cond = icmp ult i32 %i, %n
        br i1 %cond, label %next, label %return
next:
        %i.inc = add i32 %i, 1
        br label %header
return:
        ret void
}

which contains loop of this form:

unsigned n = ...;
for (unsigned i = 100; i < n; ++i)
  ;

scalar evolution determines loop iteration count as: (-100 + %n).
This isn't correct, because for %n < 100 we'll get negative number of
iterations.

One way to fix it is to add a 'umax' SCEV similar to the 'smax' one. Using it,
the answer for the example would be: (100 umax %n) - 100.

Any other ideas?


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  9 18:32:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 9 Feb 2008 18:32:39 -0600
Subject: [LLVMbugs] [Bug 2000] Rejects valid C90 and C99
In-Reply-To: <bug-2000-206@http.llvm.org/bugs/>
Message-ID: <200802100032.m1A0WdBY028170@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2000


snaroff at apple.com <snaroff at apple.com> changed:

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




--- Comment #1 from snaroff at apple.com  2008-02-09 18:32:39 ---
http://llvm.org/viewvc/llvm-project?view=rev&revision=46917


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb  9 19:11:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 9 Feb 2008 19:11:32 -0600
Subject: [LLVMbugs] [Bug 2004] New: missed instcombine with trunc+sext
Message-ID: <bug-2004-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2004

           Summary: missed instcombine with trunc+sext
           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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


define i32 @a(i32 %x) {
entry:
        %tmp14 = shl i32 %x, 27
        %tmp15 = ashr i32 %tmp14, 27
        %tmp1516 = trunc i32 %tmp15 to i8
        %tmp151617 = sext i8 %tmp1516 to i32
        ret i32 %tmp151617
}

run with "opt -std-compile-opts" is left unchanged; the trunc+sext should be
eliminated.  (Note that CodeGen is actually smart enough to do this, but it
would be nice if it worked on the IR level as 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  Sun Feb 10 03:04:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 10 Feb 2008 03:04:37 -0600
Subject: [LLVMbugs] [Bug 2005] New: Bad codegen on x86 with _Bool
Message-ID: <bug-2005-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2005

           Summary: Bad codegen on x86 with _Bool
           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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Take the following C method:

_Bool a(_Bool a, _Bool b, _Bool c) {return a ? b : c;}

clang compiles it to the following (except that clang screws up the attributes
at the moment):

define i1 @a(i1 zeroext %a, i1 zeroext %b, i1 zeroext %c) nounwind zeroext {
entry:
        %retval = select i1 %a, i1 %b, i1 %c
        ret i1 %retval
}

This gets compiled to the following x86 code:
        movb    12(%esp), %al
        movzbw  %al, %ax
        movb    8(%esp), %cl
        movzbw  %cl, %cx
        movb    4(%esp), %dl
        testb   %dl, %dl
        cmovne  %cx, %ax
        andl    $1, %eax
        ret

This is correct, but it's about twice as long as it needs to be.

I'd expect something like the following (untested):
        movzbl  12(%esp), %eax
        movzbl  8(%esp), %ecx
        cmpl    $0, 4(%esp)
        cmovne  %ecx, %eax
        ret

llvm-gcc compiles the C code to something a bit different:
define i8 @a(i8 zeroext  %a, i8 zeroext  %b, i8 zeroext  %c) zeroext  {
entry:
        %tmp3 = icmp eq i8 %a, 0                ; <i1> [#uses=1]
        %iftmp.16.0.in.in = select i1 %tmp3, i8 %c, i8 %b
        %iftmp.16.0.in = icmp ne i8 %iftmp.16.0.in.in, 0
        %retval1819 = zext i1 %iftmp.16.0.in to i8
        ret i8 %retval1819
}

Which gets compiled to the following x86:
        movzbw  12(%esp), %ax
        movzbw  8(%esp), %cx
        cmpb    $0, 4(%esp)
        cmove   %ax, %cx
        testb   %cl, %cl
        setne   %al
        movzbl  %al, %eax
        ret


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 10 14:11:20 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 10 Feb 2008 14:11:20 -0600
Subject: [LLVMbugs] [Bug 2011] New: Loop Index Splitting Pass asserts on
	SPASS
Message-ID: <bug-2011-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2011

           Summary: Loop Index Splitting Pass asserts on SPASS
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: evan.cheng at apple.com
                CC: llvmbugs at cs.uiuc.edu


Compile rules-inf.c at -O3:

~/LLVM/install/bin/llvm-gcc -DCLOCK_NO_TIMING -fno-strict-aliasing -w
-I/Volumes/Muggles/LLVM/llvm/projects/llvm-test/MultiSource/Applications/SPASS
-I/U\
sers/echeng/LLVM/llvm/projects/llvm-test/MultiSource/Applications/SPASS
-I/Users/echeng/LLVM/llvm/include
-I/Users/echeng/LLVM/llvm/projects/llvm-test/include -I../../../inc\
lude -I/Users/echeng/LLVM/llvm/include -D_GNU_SOURCE -D__STDC_LIMIT_MACROS
-DNDEBUG  -O3 -mdynamic-no-pic -fomit-frame-pointer -c rules-inf.c -o
Output/rules-inf.bc -emit-ll\
vm -mllvm -debug-pass=Arguments
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -preverify -domtree -verify -simplifycfg -domtree -domfrontier
-scalarrepl -instcombine
Pass Arguments:  -raiseallocs -simplifycfg -domtree -domfrontier -mem2reg
-globalopt -globaldce -ipconstprop -deadargelim -instcombine -simplifycfg
-basiccg -prune-eh -inlin\
e -simplify-libcalls -basiccg -argpromotion -tailduplicate -instcombine
-simplifycfg -domtree -domfrontier -scalarrepl -instcombine -break-crit-edges
-condprop -tailcallelim\
 -simplifycfg -reassociate -domtree -loops -loopsimplify -domfrontier
-scalar-evolution -lcssa -loop-rotate -licm -lcssa -loop-unswitch
-scalar-evolution -lcssa -loop-index-\
split -instcombine -scalar-evolution -lcssa -indvars -scalar-evolution -lcssa
-loop-unroll -instcombine -domtree -memdep -gvn -sccp -instcombine
-break-crit-edges -condprop \
-memdep -dse -mergereturn -postdomtree -postdomfrontier -adce -simplifycfg
-constmerge -strip-dead-prototypes
Assertion failed: (NewV && "Unable to find new incoming value for exit block
PHI"), function updatePHINodes, file LoopIndexSplit.cpp, line 1651.
rules-inf.c:4281: internal compiler error: Abort trap
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 Feb 10 17:08:50 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 10 Feb 2008 17:08:50 -0600
Subject: [LLVMbugs] [Bug 1999] Rejects valid C90.
In-Reply-To: <bug-1999-206@http.llvm.org/bugs/>
Message-ID: <200802102308.m1AN8oI1001040@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1999


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #2 from Chris Lattner <sabre at nondot.org>  2008-02-10 17:08:49 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080204/004222.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 Feb 10 18:02:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 10 Feb 2008 18:02:39 -0600
Subject: [LLVMbugs] [Bug 1992] switch() { case sizeof("str") } generates
 spurious warnings
In-Reply-To: <bug-1992-206@http.llvm.org/bugs/>
Message-ID: <200802110002.m1B02dsm002710@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1992


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #2 from Chris Lattner <sabre at nondot.org>  2008-02-10 18:02:39 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080204/004225.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 Feb 10 18:07:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 10 Feb 2008 18:07:28 -0600
Subject: [LLVMbugs] [Bug 1988] clang crashes on invalid struct definition
In-Reply-To: <bug-1988-206@http.llvm.org/bugs/>
Message-ID: <200802110007.m1B07Sj6002869@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1988


snaroff at apple.com <snaroff at apple.com> changed:

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




--- Comment #1 from snaroff at apple.com  2008-02-10 18:07:28 ---
Fixed in http://llvm.org/viewvc/llvm-project?view=rev&revision=46942


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 10 19:10:17 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 10 Feb 2008 19:10:17 -0600
Subject: [LLVMbugs] [Bug 2012] New: implicit int rules still not right
Message-ID: <bug-2012-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2012

           Summary: implicit int rules still not right
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


The following is invalid in C90 yet clang accepts it silently:

void f (int, x);


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 04:00:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 04:00:07 -0600
Subject: [LLVMbugs] [Bug 2013] New: Missing source line in diagnostic
Message-ID: <bug-2013-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2013

           Summary: Missing source line in diagnostic
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: minor
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


Playing with clang after the conversation of cfe-dev I tried:

struct S {
  int a;
  int foo();
};

struct S s; 

void bar (void) { s.foo = &s; }

$ ~/src/nobackup/llvm/Debug/bin/clang -pedantic-errors /tmp/bug.c 
/tmp/bug.c:3:7: error: field 'foo' declared as a function
  int foo();
      ^
/tmp/bug.c:8:25: error: non-object type 'int ()' is not assignable
2 diagnostics generated.

The original source line is not displayed for the 2nd error.  It would be nice
if the error were suppressed entirely, but that's a minor and debatable point.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 11:36:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 11:36:08 -0600
Subject: [LLVMbugs] [Bug 2014] New: triggered assertion in getAlignment()
Message-ID: <bug-2014-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2014

           Summary: triggered assertion in getAlignment()
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: compile-fail
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1394)
 --> (http://llvm.org/bugs/attachment.cgi?id=1394)
test case (delta reduced)

here is on more test case that triggers an assertion in codegen. the problem is
related with function receiving struct pointers as argument.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 12:02:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 12:02:09 -0600
Subject: [LLVMbugs] [Bug 1925] Assertion failed in RegisterScavenging with
 regalloc=local
In-Reply-To: <bug-1925-206@http.llvm.org/bugs/>
Message-ID: <200802111802.m1BI2963009494@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1925


Lauro Venancio <lauro.venancio at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         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 Feb 11 12:30:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 12:30:12 -0600
Subject: [LLVMbugs] [Bug 2015] New: __attribute__ ((__transparent_union__))
 not implemented
Message-ID: <bug-2015-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2015

           Summary: __attribute__ ((__transparent_union__)) not implemented
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1396)
 --> (http://llvm.org/bugs/attachment.cgi?id=1396)
test case

__attribute__ ((__transparent_union__)) is not implemented (found in some glibc
headers).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 16:19:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 16:19:28 -0600
Subject: [LLVMbugs] [Bug 2013] Missing source line in diagnostic
In-Reply-To: <bug-2013-206@http.llvm.org/bugs/>
Message-ID: <200802112219.m1BMJShI017340@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2013


snaroff at apple.com <snaroff at apple.com> changed:

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




--- Comment #1 from snaroff at apple.com  2008-02-11 16:19:26 ---
http://llvm.org/viewvc/llvm-project?view=rev&revision=46972


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 18:50:49 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 18:50:49 -0600
Subject: [LLVMbugs] [Bug 2016] New: clang doesn't support gcc's computed
	goto tables
Message-ID: <bug-2016-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2016

           Summary: clang doesn't support gcc's computed goto tables
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


The following code is rejected by clang:

#include <stdio.h>

int main()
{
        static void *table[] = {&&label1, &&label2};

label1:
        goto *(table[1]);
        printf("skip this\n");
label2:
        printf("hello\n");

        return 0;
}


gcc compiles this just fine, and the program outputs "hello"


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 22:11:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 22:11:39 -0600
Subject: [LLVMbugs] [Bug 1987] sema crashes on invalid function definition
In-Reply-To: <bug-1987-206@http.llvm.org/bugs/>
Message-ID: <200802120411.m1C4BdCP027422@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1987


snaroff at apple.com <snaroff at apple.com> changed:

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




--- Comment #1 from snaroff at apple.com  2008-02-11 22:11:38 ---
http://llvm.org/viewvc/llvm-project?view=rev&revision=46984


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 22:18:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 22:18:48 -0600
Subject: [LLVMbugs] [Bug 2017] New: clang crashes on array with void size
Message-ID: <bug-2017-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2017

           Summary: clang crashes on array with void size
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:

void x(); int a() {int r[x()];}

crashes clang -fsyntax-only.

We probably shouldn't be calling isIntegerConstantExpr on expressions that
don't have integer 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  Mon Feb 11 22:49:42 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 22:49:42 -0600
Subject: [LLVMbugs] [Bug 2018] New: Crash with void compound literal
Message-ID: <bug-2018-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2018

           Summary: Crash with void compound literal
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:

int a() {
(void) { 1 };
}

Should give an error, not crash.

Also, the behavior for the following testcase is odd:

int a() {
(struct a) {1};
}

Probably just need a check for the restriction in 6.5.2.5p1.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 23:09:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 23:09:23 -0600
Subject: [LLVMbugs] [Bug 2019] New: Crash in codegen with inline asm using
 "+S" and "+D" operands
Message-ID: <bug-2019-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2019

           Summary: Crash in codegen with inline asm using "+S" and "+D"
                    operands
           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


C code at http://llvm.org/bugs/attachment.cgi?id=1397. llvm-gcc and clang
compile this code to different asm statements, leading to different asserts.

With llvm-gcc, this triggers a "matching constraints for memory operands unimp"
assertion in SelectionDAGISel.  Despite this, I'm pretty sure this is unrelated
to Bug 1133.

clang, on the other hand, triggers a "Unknown constraint type" in
SelectionDAGISel.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 11 23:53:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 11 Feb 2008 23:53:09 -0600
Subject: [LLVMbugs] [Bug 2003] Incorrect iteration count for loop with
 unsigned ICMP condition
In-Reply-To: <bug-2003-206@http.llvm.org/bugs/>
Message-ID: <200802120553.m1C5r9bj007664@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2003


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #1 from Nick Lewycky <nicholas at mxc.ca>  2008-02-11 23:53:09 ---
If %n = 0, then we get %n - 100 = -100 = +156 (unsigned). And this is the
correct loop trip count; if you start at 100, it'll take you 156 (256-100)
iterations to get back around to zero.

Reopen if you find a counter-example.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 12 00:20:15 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 12 Feb 2008 00:20:15 -0600
Subject: [LLVMbugs] [Bug 2020] New: clang crash redeclaring a union as an
	enum
Message-ID: <bug-2020-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2020

           Summary: clang crash redeclaring a union as an enum
           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


Testcase:

union u0;
enum u0 { U0A };


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 12 00:35:40 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 12 Feb 2008 00:35:40 -0600
Subject: [LLVMbugs] [Bug 2021] New: clang prints bogus type compatibility
 warning between constant and variable array types
Message-ID: <bug-2021-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2021

           Summary: clang prints bogus type compatibility warning between
                    constant and variable array types
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: AST
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
int a(int i) {
  int (*pf)[2];
  int (*pv)[i];
  (i ? pf : pv);
}

This shouldn't warn because the pointed-to types are compatible.  Easy fix to
ASTContext, which I'll sort out in a bit.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 12 02:23:59 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 12 Feb 2008 02:23:59 -0600
Subject: [LLVMbugs] [Bug 2021] clang prints bogus type compatibility warning
 between constant and variable array types
In-Reply-To: <bug-2021-206@http.llvm.org/bugs/>
Message-ID: <200802120823.m1C8Nxk5003636@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2021


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2008-02-12 02:23:59 ---
Okay, checked in.  (I guess I didn't really need to open a bug, but I didn't
want to lose track of 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 Feb 12 06:10:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 12 Feb 2008 06:10:00 -0600
Subject: [LLVMbugs] [Bug 1996] llvm doesn't optmize out loads of undef
In-Reply-To: <bug-1996-206@http.llvm.org/bugs/>
Message-ID: <200802121210.m1CCA09B018246@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1996


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #9 from Eli Friedman <sharparrow1 at yahoo.com>  2008-02-12 06:09:59 ---
Commited.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 12 07:41:02 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 12 Feb 2008 07:41:02 -0600
Subject: [LLVMbugs] [Bug 2003] Incorrect iteration count for loop with
 unsigned ICMP condition
In-Reply-To: <bug-2003-206@http.llvm.org/bugs/>
Message-ID: <200802121341.m1CDf2RS021667@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2003


Wojciech Matyjewicz <wmatyjewicz at fastmail.fm> changed:

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




--- Comment #2 from Wojciech Matyjewicz <wmatyjewicz at fastmail.fm>  2008-02-12 07:40:54 ---
> If %n = 0, then we get %n - 100 = -100 = +156 (unsigned). And this is the
> correct loop trip count; if you start at 100, it'll take you 156 (256-100)
> iterations to get back around to zero.

Please, note that the comparison operator is ULT and not NE. What you are
saying is true in the second case. Suppose %n = 0. Then the first execution of:
        %cond = icmp ult i32 %i, %n
defines %cond as false and causes the loop to end. So, the trip count should be
0 in this case.


-- 
Configure bugmail: 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  Tue Feb 12 07:44:37 2008
From: neil at daikokuya.co.uk (Neil Booth)
Date: Tue, 12 Feb 2008 22:44:37 +0900
Subject: [LLVMbugs] [Bug 2021] New: clang prints bogus type
	compatibility warning between constant and variable array types
In-Reply-To: <bug-2021-206@http.llvm.org/bugs/>
References: <bug-2021-206@http.llvm.org/bugs/>
Message-ID: <20080212134437.GD7679@daikokuya.co.uk>

bugzilla-daemon at cs.uiuc.edu wrote:-

> http://llvm.org/bugs/show_bug.cgi?id=2021
> 
>            Summary: clang prints bogus type compatibility warning between
>                     constant and variable array types
>            Product: clang
>            Version: unspecified
>           Platform: PC
>         OS/Version: Linux
>             Status: NEW
>           Severity: normal
>           Priority: P2
>          Component: AST
>         AssignedTo: unassignedbugs at nondot.org
>         ReportedBy: sharparrow1 at yahoo.com
>                 CC: llvmbugs at cs.uiuc.edu
> 
> 
> Testcase:
> int a(int i) {
>   int (*pf)[2];
>   int (*pv)[i];
>   (i ? pf : pv);
> }
> 
> This shouldn't warn because the pointed-to types are compatible.  Easy fix to
> ASTContext, which I'll sort out in a bit.

It's best to tighten up the testcases and ensure they can only
complain about the thing being tested.  For example this code
has good reason to complain twice:

c$ ~/src/cfe/cfe /tmp/bug.c
"/tmp/bug.c", line 4: warning: expression has no effect
  (i ? pf : pv);
  ^
"/tmp/bug.c", line 5: warning: function "a" should return a value
}
^

Neil.

From bugzilla-daemon at cs.uiuc.edu  Tue Feb 12 09:14:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 12 Feb 2008 09:14:43 -0600
Subject: [LLVMbugs] [Bug 2002] Incorrect iteration count for loop with
 signed ICMP condition
In-Reply-To: <bug-2002-206@http.llvm.org/bugs/>
Message-ID: <200802121514.m1CFEh8M025574@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2002


Wojciech Matyjewicz <wmatyjewicz at fastmail.fm> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |2.3




--- Comment #6 from Wojciech Matyjewicz <wmatyjewicz at fastmail.fm>  2008-02-12 09:14:34 ---
Thanks, Nick!

Applied here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058176.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 Feb 12 16:53:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 12 Feb 2008 16:53:11 -0600
Subject: [LLVMbugs] [Bug 2022] New: 2.2 does not build cleanly with GCC 4.3
Message-ID: <bug-2022-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2022

           Summary: 2.2 does not build cleanly with GCC 4.3
           Product: libraries
           Version: 2.2
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: major
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: bos at serpentine.com
                CC: llvmbugs at cs.uiuc.edu


It needs some love from someone who can follow
http://gcc.gnu.org/gcc-4.3/porting_to.html

This is currently blocking the build of 2.2 for Fedora 9.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 13 06:14:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 06:14:37 -0600
Subject: [LLVMbugs] [Bug 2023] New: Crash in llc with unusual initializer
Message-ID: <bug-2023-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2023

           Summary: Crash in llc with unusual initializer
           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


Testcase:

@b =  global i16 ptrtoint (i16* @b to i16)

run through llvm-as | llc, should generate code that initializes the global
appropriately.

Current output:
llc: /home/eli/llvm/lib/CodeGen/AsmPrinter.cpp:789: void
llvm::AsmPrinter::EmitConstantValueOnly(const llvm::Constant*): Assertion `0 &&
"FIXME: Don't yet support this kind of constant cast expr"' failed.

Fixing this isn't really a high priority for me, but I would like to know
whether the fact that this doesn't work is considered a bug; see
http://lists.cs.uiuc.edu/pipermail/cfe-dev/2008-February/001054.html.  I'm in
the process of considering how to handle constant expressions in clang, and
this affects what clang should accept as a constant 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  Wed Feb 13 14:32:24 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 14:32:24 -0600
Subject: [LLVMbugs] [Bug 2024] New: Problem with ELF section attribute
Message-ID: <bug-2024-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2024

           Summary: Problem with ELF section attribute
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: hehaifeng2nd at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Hi,

I am having problem of using LLVM on a program which contains
customized code section.
In addition to ".text" section for keeping code, the program has
another code section, called ".init.text".
The functions in this section are defined as
  "int foo (int, int) __attribute__ ((section (".init.text")));"

If I compile the program with gcc, this section(".init.text") will
have attributes of "AX" (means allocate space and
executable). However, if I first compile the program into llvm IR,
then convert IR to native assembly code using
llc, this section will not have "AX" attribute. This causes the native
executable fails when it tries to call a function
in .init.text section. Any suggestion? (btw, my platform is Linux
running on x86 machine.)

Haifeng


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 13 15:19:52 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 15:19:52 -0600
Subject: [LLVMbugs] [Bug 2025] New: llvm-gcc4.2 won' t build if using LLVM
 release build and boostrap enabled
Message-ID: <bug-2025-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2025

           Summary: llvm-gcc4.2 won't build if using LLVM release build and
                    boostrap enabled
           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: tonic at nondot.org
                CC: llvmbugs at cs.uiuc.edu


I have a release build of LLVM and I'm trying to build a release build of
llvm-gcc4.2 (svn TOT). It gives this error:

[cut]
dsymutil is not available
configure: error: You must specify valid path to your LLVM tree with
--enable-llvm=DIR
yes
checking for strings.h... (cached) yes
checking for string.h... (cached) yes
checking sys/file.h usability... yes
[cut]
Found Release LLVM Tree in /Users/lattner/work/release_build/llvm
updating cache ./config.cache
configure: creating ./config.status
config.status: creating Makefile
config.status: creating config.h
config.status: executing depdir commands
make[1]: *** [stage1-bubble] Error 2
make: *** [all] Error 2


If I add --disable-bootstrap to my configure line, it builds just fine.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 13 16:28:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 16:28:22 -0600
Subject: [LLVMbugs] [Bug 2011] Loop Index Splitting Pass asserts on SPASS
In-Reply-To: <bug-2011-206@http.llvm.org/bugs/>
Message-ID: <200802132228.m1DMSM82023730@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2011


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #1 from Devang Patel <dpatel at apple.com>  2008-02-13 16:28:21 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058262.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058262.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058274.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 Feb 13 20:06:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 20:06:55 -0600
Subject: [LLVMbugs] [Bug 2026] New: passmanager crash on domfrontier
	requiring domtree?
Message-ID: <bug-2026-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2026

           Summary: passmanager crash on domfrontier requiring domtree?
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nlewycky at google.com
                CC: dpatel at apple.com, llvmbugs at cs.uiuc.edu


$ opt b.bc -loopsimplify -domtree -lcssa -domfrontier -ipconstprop

opt: /usr/local/src/llvm/include/llvm/PassAnalysisSupport.h:193: AnalysisType&
llvm::Pass::getAnalysisID(const llvm::PassInfo*) const [with AnalysisType =
llvm::DominatorTree]: Assertion `ResultPass && "getAnalysis*() called on an
analysis that was not " "'required' by pass!"' failed.
opt((anonymous namespace)::PrintStackTrace()+0x22)[0x8682eee]
opt((anonymous namespace)::SignalHandler(int)+0x110)[0x8683216]
[0xffffe500]
[0x6]
/lib/tls/i686/cmov/libc.so.6(abort+0xe9)[0x41e982b9]
/lib/tls/i686/cmov/libc.so.6(__assert_fail+0x101)[0x41e8ff51]
opt(llvm::DominatorTree&
llvm::Pass::getAnalysisID<llvm::DominatorTree>(llvm::PassInfo const*)
const+0xa0)[0x83ae944]
opt(llvm::DominatorTree& llvm::Pass::getAnalysis<llvm::DominatorTree>()
const+0x48)[0x83ae9fc]
opt(llvm::DominanceFrontier::runOnFunction(llvm::Function&)+0x1f)[0x85d4451]
opt(llvm::FPPassManager::runOnFunction(llvm::Function&)+0x122)[0x8610ad2]
opt(llvm::FPPassManager::runOnModule(llvm::Module&)+0x6e)[0x8610c82]
opt(llvm::MPPassManager::runOnModule(llvm::Module&)+0x108)[0x861079c]
opt(llvm::PassManagerImpl::run(llvm::Module&)+0x6e)[0x8610956]
opt(llvm::PassManager::run(llvm::Module&)+0x1a)[0x86109aa]
opt(main+0xa8f)[0x8364959]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xd2)[0x41e82ea2]
opt(__gxx_personality_v0+0x141)[0x8356d6d]
Aborted

The LLVM asm to trigger this is:

define i32 @test() {
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  Wed Feb 13 21:05:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 21:05:23 -0600
Subject: [LLVMbugs] [Bug 2027] New: -find-bugs forgets pass order when it
 comes time to reduce
Message-ID: <bug-2027-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2027

           Summary: -find-bugs forgets pass order when it comes time to
                    reduce
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: bugpoint
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nlewycky at google.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1400)
 --> (http://llvm.org/bugs/attachment.cgi?id=1400)
proposed fix

bugpoint -find-bugs works by reordering the sequence of passes randomly and
optimizing and executing the result. Once it finds a failure (optz'n crash or
miscompile) it then starts to reduce it.

Unfortunately, what it reduces is the original pass list that the user passed
in, not the shuffled version that actually failed.

This patch modifies bugpoint to remember its ordering. I'm not sure it's the
best fix. Could we remove TempPass entirely and just modify PassesToRun in
place? Why doesn't the code already do 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  Wed Feb 13 21:21:26 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 21:21:26 -0600
Subject: [LLVMbugs] [Bug 2028] New: passmanager asserts that pass requires
 lower level pass
Message-ID: <bug-2028-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2028

           Summary: passmanager asserts that pass requires lower level pass
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nlewycky at google.com
                CC: dpatel at apple.com, llvmbugs at cs.uiuc.edu


Similar to PR2026, this sequence fails on any bytecode with at least a function
in it:

$ opt b2.bc -loop-unroll -loop-rotate -simplifycfg
opt: PassManager.cpp:934: virtual void
llvm::PMDataManager::addLowerLevelRequiredPass(llvm::Pass*, llvm::Pass*):
Assertion `0 && "Unable to handle Pass that requires lower level Analysis
pass"' failed.
opt((anonymous namespace)::PrintStackTrace()+0x22)[0x8682eee]
opt((anonymous namespace)::SignalHandler(int)+0x110)[0x8683216]
[0xffffe500]
[0x6]
/lib/tls/i686/cmov/libc.so.6(abort+0xe9)[0x41e982b9]
/lib/tls/i686/cmov/libc.so.6(__assert_fail+0x101)[0x41e8ff51]
opt(llvm::PMStack::dump()+0x0)[0x860e91a]
opt(llvm::PMDataManager::add(llvm::Pass*, bool)+0x2f9)[0x861166d]
opt(llvm::LoopPass::assignPassManager(llvm::PMStack&,
llvm::PassManagerType)+0x1af)[0x8542809]
opt(llvm::PassManagerImpl::addTopLevelPass(llvm::Pass*)+0xe8)[0x861af98]
opt(llvm::PMTopLevelManager::schedulePass(llvm::Pass*)+0x166)[0x860f26c]
opt(llvm::PassManagerImpl::add(llvm::Pass*)+0x1b)[0x861527b]
opt(llvm::PassManager::add(llvm::Pass*)+0x1a)[0x860f296]
opt((anonymous namespace)::addPass(llvm::PassManager&,
llvm::Pass*)+0x18)[0x83660ce]
opt(main+0x6ce)[0x8364598]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xd2)[0x41e82ea2]
opt(__gxx_personality_v0+0x141)[0x8356d6d]
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  Wed Feb 13 21:29:52 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 21:29:52 -0600
Subject: [LLVMbugs] [Bug 2029] New: CBE fails with -enable-correct-eh-support
Message-ID: <bug-2029-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2029

           Summary: CBE fails with -enable-correct-eh-support
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: C
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nlewycky at google.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1401)
 --> (http://llvm.org/bugs/attachment.cgi?id=1401)
testcase

llc -march=c crashes compiling the attached testcase with the
-enable-correct-eh-support option.

$ llc b.bc -march=c -f -o b.c -enable-correct-eh-support
llc: Value.cpp:57: virtual llvm::Value::~Value(): Assertion `use_empty() &&
"Uses remain when a value is destroyed!"' failed.
llc((anonymous namespace)::PrintStackTrace()+0x22)[0x8a873e2]
llc((anonymous namespace)::SignalHandler(int)+0x110)[0x8a8770a]
[0xffffe500]
[0x6]
/lib/tls/i686/cmov/libc.so.6(abort+0xe9)[0x41e982b9]
/lib/tls/i686/cmov/libc.so.6(__assert_fail+0x101)[0x41e8ff51]
llc(llvm::Value::~Value()+0x164)[0x8a35758]
llc(llvm::BasicBlock::~BasicBlock()+0x51)[0x89b3c29]
llc(llvm::iplist<llvm::BasicBlock, llvm::ilist_traits<llvm::BasicBlock>
>::erase(llvm::ilist_iterator<llvm::BasicBlock>)+0x33)[0x88de98d]
llc[0x892c447]
llc[0x892c4d0]
llc(llvm::FPPassManager::runOnFunction(llvm::Function&)+0x122)[0x8a1515e]
llc(llvm::FPPassManager::runOnModule(llvm::Module&)+0x6e)[0x8a1530e]
llc(llvm::MPPassManager::runOnModule(llvm::Module&)+0x108)[0x8a14e28]
llc(llvm::PassManagerImpl::run(llvm::Module&)+0x6e)[0x8a14fe2]
llc(llvm::PassManager::run(llvm::Module&)+0x1a)[0x8a15036]
llc(main+0x5b0)[0x8424a3c]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xd2)[0x41e82ea2]
llc[0x8423d51]
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  Wed Feb 13 21:35:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 21:35:11 -0600
Subject: [LLVMbugs] [Bug 2030] New: loop index split tries to mangle PHIs
Message-ID: <bug-2030-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2030

           Summary: loop index split tries to mangle PHIs
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nlewycky at google.com
                CC: dpatel at apple.com, llvmbugs at cs.uiuc.edu


Created an attachment (id=1402)
 --> (http://llvm.org/bugs/attachment.cgi?id=1402)
testcase

Loop index splitting dies on this testcase:

$ opt b.bc -loop-index-split

opt: /usr/local/src/llvm/include/llvm/Instructions.h:1285: llvm::Value*
llvm::PHINode::getIncomingValue(unsigned int) const: Assertion `i*2 <
getNumOperands() && "Invalid value number!"' failed.

Program received signal SIGABRT, Aborted.
0xffffe410 in __kernel_vsyscall ()
(gdb) bt
#0  0xffffe410 in __kernel_vsyscall ()
#1  0x41e969a1 in raise () from /lib/tls/i686/cmov/libc.so.6
#2  0x41e982b9 in abort () from /lib/tls/i686/cmov/libc.so.6
#3  0x41e8ff51 in __assert_fail () from /lib/tls/i686/cmov/libc.so.6
#4  0x08397b95 in llvm::PHINode::getIncomingValue (this=0x8774a28, 
    i=4294967295) at /usr/local/src/llvm/include/llvm/Instructions.h:1285
#5  0x08397d25 in llvm::PHINode::getIncomingValueForBlock (this=0x8774a28, 
    BB=0x876f618) at /usr/local/src/llvm/include/llvm/Instructions.h:1349
#6  0x08457063 in (anonymous namespace)::LoopIndexSplit::splitLoop (
    this=0x876c050, SD=@0x876c090) at LoopIndexSplit.cpp:1468
#7  0x084581e3 in (anonymous namespace)::LoopIndexSplit::runOnLoop (
    this=0x876c050, IncomingLoop=0x877e028, LPM_Ref=@0x876cf60)
    at LoopIndexSplit.cpp:272
#8  0x08541d60 in llvm::LPPassManager::runOnFunction (this=0x876cf60, 
    F=@0x876d5f0) at LoopPass.cpp:225
#9  0x08610ad2 in llvm::FPPassManager::runOnFunction (this=0x876ee78, 
    F=@0x876d5f0) at PassManager.cpp:1184
#10 0x08610c82 in llvm::FPPassManager::runOnModule (this=0x876ee78, 
    M=@0x876bf98) at PassManager.cpp:1204
#11 0x0861079c in llvm::MPPassManager::runOnModule (this=0x876ca70, 
    M=@0x876bf98) at PassManager.cpp:1254
#12 0x08610956 in llvm::PassManagerImpl::run (this=0x876c298, M=@0x876bf98)
    at PassManager.cpp:1328
#13 0x086109aa in llvm::PassManager::run (this=0xffffd61c, M=@0x876bf98)
    at PassManager.cpp:1360
#14 0x08364959 in main (argc=3, argv=0xffffd804) at opt.cpp:426


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 13 21:49:51 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 21:49:51 -0600
Subject: [LLVMbugs] [Bug 2031] New: dominator frontier not updated correctly
Message-ID: <bug-2031-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2031

           Summary: dominator frontier not updated correctly
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nlewycky at google.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1403)
 --> (http://llvm.org/bugs/attachment.cgi?id=1403)
testcase

One of these passes isn't updating the domfrontier properly and the problem is
manifesting itself as a crash in loop-unswitch.

$ opt -scalarrepl -loop-unswitch -break-crit-edges b.bc

opt: /usr/local/src/llvm/include/llvm/Analysis/Dominators.h:881: void
llvm::DominanceFrontierBase::removeFromFrontier(std::_Rb_tree_iterator<std::pair<llvm::BasicBlock*
const, std::set<llvm::BasicBlock*, std::less<llvm::BasicBlock*>,
std::allocator<llvm::BasicBlock*> > > >, llvm::BasicBlock*): Assertion
`I->second.count(Node) && "Node is not in DominanceFrontier of BB"' failed.

Program received signal SIGABRT, Aborted.
0xffffe410 in __kernel_vsyscall ()
(gdb) bt
#0  0xffffe410 in __kernel_vsyscall ()
#1  0x41e969a1 in raise () from /lib/tls/i686/cmov/libc.so.6
#2  0x41e982b9 in abort () from /lib/tls/i686/cmov/libc.so.6
#3  0x41e8ff51 in __assert_fail () from /lib/tls/i686/cmov/libc.so.6
#4  0x08459777 in llvm::DominanceFrontierBase::removeFromFrontier (
    this=0x876c168, I={_M_node = 0x8772cc8}, Node=0x8770d08)
    at /usr/local/src/llvm/include/llvm/Analysis/Dominators.h:881
#5  0x084762c5 in (anonymous
namespace)::LoopUnswitch::UnswitchNontrivialCondition (this=0x876d190,
LIC=0x876e370, Val=0x87719d0, L=0x8771238)
    at LoopUnswitch.cpp:897
#6  0x08476990 in (anonymous namespace)::LoopUnswitch::UnswitchIfProfitable (
    this=0x876d190, LoopCond=0x876e370, Val=0x87719d0, L=0x8771238)
    at LoopUnswitch.cpp:433
#7  0x08476aa0 in (anonymous namespace)::LoopUnswitch::processLoop (
    this=0x876d190, L=0x8771238) at LoopUnswitch.cpp:215
#8  0x08476d6c in (anonymous namespace)::LoopUnswitch::runOnLoop (
    this=0x876d190, L=0x8771238, LPM_Ref=@0x876f818) at LoopUnswitch.cpp:191
#9  0x08541d60 in llvm::LPPassManager::runOnFunction (this=0x876f818, 
    F=@0x876d8b0) at LoopPass.cpp:225
#10 0x08610ad2 in llvm::FPPassManager::runOnFunction (this=0x876e8e8, 
    F=@0x876d8b0) at PassManager.cpp:1184
#11 0x08610c82 in llvm::FPPassManager::runOnModule (this=0x876e8e8, 
    M=@0x876bf98) at PassManager.cpp:1204
---Type <return> to continue, or q <return> to quit---
#12 0x0861079c in llvm::MPPassManager::runOnModule (this=0x876e3f0, 
    M=@0x876bf98) at PassManager.cpp:1254
#13 0x08610956 in llvm::PassManagerImpl::run (this=0x876c090, M=@0x876bf98)
    at PassManager.cpp:1328
#14 0x086109aa in llvm::PassManager::run (this=0xffffd60c, M=@0x876bf98)
    at PassManager.cpp:1360
#15 0x08364959 in main (argc=5, argv=0xffffd7f4) at opt.cpp:426

Testcase 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 Feb 13 23:03:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 23:03:05 -0600
Subject: [LLVMbugs] [Bug 2027] -find-bugs forgets pass order when it comes
 time to reduce
In-Reply-To: <bug-2027-206@http.llvm.org/bugs/>
Message-ID: <200802140503.m1E535KN006539@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2027


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #3 from Nick Lewycky <nicholas at mxc.ca>  2008-02-13 23:03:04 ---
Thanks Chris! Patch checked in:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058286.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 Feb 13 23:48:42 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 13 Feb 2008 23:48:42 -0600
Subject: [LLVMbugs] [Bug 2032] New: Andersens assertion failure
Message-ID: <bug-2032-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2032

           Summary: Andersens assertion failure
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Interprocedural Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicholas at mxc.ca
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1405)
 --> (http://llvm.org/bugs/attachment.cgi?id=1405)
testcase

This small testcase fails with -anders-aa:

$ opt -anders-aa -gvn b.bc

        %tmp27.rle = phi float* [ null, %bb ], [ %sx, %entry ]          ;
<float*> [#uses=1]

opt: Andersens.cpp:497: unsigned
int<unnamed>::Andersens::getNode(llvm::Value*): Assertion `0 && "Value does not
have a node in the points-to graph!"' failed.
[New Thread 0xf7c596c0 (LWP 21274)]

Program received signal SIGABRT, Aborted.
[Switching to Thread 0xf7c596c0 (LWP 21274)]
0xffffe405 in __kernel_vsyscall ()
(gdb) bt
#0  0xffffe405 in __kernel_vsyscall ()
#1  0xf7c85f15 in raise () from /lib/i686/cmov/libc.so.6
#2  0xf7c87891 in abort () from /lib/i686/cmov/libc.so.6
#3  0xf7c7f0ee in __assert_fail () from /lib/i686/cmov/libc.so.6
#4  0x0848bc4e in getNode (this=0x86eac38, V=0x86f47a0) at Andersens.cpp:497
#5  0x084909cc in alias (this=0x86eac38, V1=0x86f47a0, V1Size=4, V2=0x86eaf90, 
    V2Size=4) at Andersens.cpp:609
#6  0x084cdccc in llvm::MemoryDependenceAnalysis::getDependency (
    this=0x86e93f8, query=0x86ea960, start=0x0, block=0x86eaec8)
    at MemoryDependenceAnalysis.cpp:431
#7  0x084cdf5d in llvm::MemoryDependenceAnalysis::nonLocalHelper (
    this=0x86e93f8, query=0x86ea960, block=0x86eaf18, resp=@0xffc376d4)
    at MemoryDependenceAnalysis.cpp:201
#8  0x084ce473 in llvm::MemoryDependenceAnalysis::getNonLocalDependency (
    this=0x86e93f8, query=0x86ea960, resp=@0xffc376d4)
    at MemoryDependenceAnalysis.cpp:288
#9  0x0837adfe in processNonLocalLoad (this=0x86e9378, L=0x86ea960, 
    toErase=@0xffc378a0) at GVN.cpp:894
#10 0x0837b2ec in processLoad (this=0x86e9378, L=0x86ea960, 
    lastLoad=@0xffc378d4, toErase=@0xffc378a0) at GVN.cpp:966
#11 0x0837b611 in processInstruction (this=0x86e9378, I=0x86ea960, 
    currAvail=@0x86eeb34, lastSeenLoad=@0xffc378d4, toErase=@0xffc378a0)
    at GVN.cpp:1129
#12 0x0837baf2 in iterateOnFunction (this=0x86e9378, F=@0x86eb300)
    at GVN.cpp:1233
#13 0x0837bc2e in runOnFunction (this=0x86e9378, F=@0x86eb300) at GVN.cpp:1195
#14 0x0858b582 in llvm::FPPassManager::runOnFunction (this=0x86e9140, 
    F=@0x86eb300) at PassManager.cpp:1184
#15 0x0858b724 in llvm::FPPassManager::runOnModule (this=0x86e9140, 
    M=@0x86e8ff8) at PassManager.cpp:1204
#16 0x0858b258 in llvm::MPPassManager::runOnModule (this=0x86e9ae0, 
    M=@0x86e8ff8) at PassManager.cpp:1254
#17 0x0858b410 in llvm::PassManagerImpl::run (this=0x86eab48, M=@0x86e8ff8)
    at PassManager.cpp:1328
#18 0x0858b462 in llvm::PassManager::run (this=0xffc37af8, M=@0x86e8ff8)
    at PassManager.cpp:1360
#19 0x082edf87 in main (argc=4, argv=0xffc37cf4) at opt.cpp:430

Testcase 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  Thu Feb 14 00:01:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 00:01:00 -0600
Subject: [LLVMbugs] [Bug 2033] New: corrupted CallGraph
Message-ID: <bug-2033-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2033

           Summary: corrupted CallGraph
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicholas at mxc.ca
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1406)
 --> (http://llvm.org/bugs/attachment.cgi?id=1406)
testcase

Testcase attached.

$ opt -basiccg -instcombine -inline b.bc

opt: CallGraph.cpp:288: void
llvm::CallGraphNode::removeCallEdgeTo(llvm::CallGraphNode*): Assertion `i &&
"Cannot find callee to remove!"' failed.
[New Thread 4157462208 (LWP 22976)]

Program received signal SIGABRT, Aborted.
[Switching to Thread 4157462208 (LWP 22976)]
0xffffe405 in __kernel_vsyscall ()
(gdb) bt
#0  0xffffe405 in __kernel_vsyscall ()
#1  0xf7d09f15 in raise () from /lib/i686/cmov/libc.so.6
#2  0xf7d0b891 in abort () from /lib/i686/cmov/libc.so.6
#3  0xf7d030ee in __assert_fail () from /lib/i686/cmov/libc.so.6
#4  0x08499d8d in llvm::CallGraphNode::removeCallEdgeTo (this=0x86ebd60, 
    Callee=0x86e9228) at CallGraph.cpp:288
#5  0x0845320f in UpdateCallGraphAfterInlining (Caller=0x86e9e70, 
    Callee=0x86ebe78, FirstNewBlock=
      {<bidirectional_iterator<llvm::BasicBlock,int>> =
{<std::iterator<std::bidirectional_iterator_tag,llvm::BasicBlock,int,llvm::BasicBlock*,llvm::BasicBlock&>>
= {<No data fields>}, <No data fields>}, NodePtr = 0x86e9348}, 
    ValueMap=@0xfff18080, CG=@0x86e93d0) at InlineFunction.cpp:154
#6  0x08454237 in llvm::InlineFunction (CS=@0xfff18280, CG=0x86e93d0, 
    TD=0x86e9b00) at InlineFunction.cpp:306
#7  0x0832c069 in InlineCallIfPossible (CS=@0xfff18324, CG=@0x86e93d0, 
    SCCFunctions=@0xfff182e0, TD=@0x86e9b00) at Inliner.cpp:59
#8  0x0832c90c in llvm::Inliner::runOnSCC (this=0x86e9098, SCC=@0xfff1841c)
    at Inliner.cpp:151
#9  0x0849cb48 in CGPassManager::runOnModule (this=0x86e9128, M=@0x86e9aa8)
    at CallGraphSCCPass.cpp:100
#10 0x0858b258 in llvm::MPPassManager::runOnModule (this=0x86ebfc0, 
    M=@0x86e9aa8) at PassManager.cpp:1254
#11 0x0858b410 in llvm::PassManagerImpl::run (this=0x86ebee8, M=@0x86e9aa8)
    at PassManager.cpp:1328
#12 0x0858b462 in llvm::PassManager::run (this=0xfff185b8, M=@0x86e9aa8)
    at PassManager.cpp:1360
#13 0x082edf87 in main (argc=5, argv=0xfff187b4) at opt.cpp:430


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 14 01:10:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 01:10:27 -0600
Subject: [LLVMbugs] [Bug 2034] New: module-level analysis invalidated but
	not re-run
Message-ID: <bug-2034-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2034

           Summary: module-level analysis invalidated but not re-run
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicholas at mxc.ca
                CC: dpatel at apple.com, llvmbugs at cs.uiuc.edu


Here's an interesting one:

$ llvm-commit/Debug/bin/opt -anders-aa -instcombine -gvn b.bc
-debug-pass=Structure

Now before I get to the output, what's happening is that
  a) anders-aa registers all the Values up front
  b) instcombine folds together two GEPs (%tmp10 and %tmp11) creating a new one
(%tmp11) which anders-aa hasn't seen
  c) gvn queries anders-aa about the new %tmp11

instcombine is marked as preserving the CFG, but not any AA stuff, so really
anders-aa should be re-run before gvn gets to it. Here's the output:

Pass Arguments:  -anders-aa -instcombine -domtree -memdep -gvn -preverify
-domtree -verify
Target Data Layout
Basic Alias Analysis (default AA impl)
  ModulePass Manager
    Andersen's Interprocedural Alias Analysis
    FunctionPass Manager
      Combine redundant instructions
--      Combine redundant instructions
      Dominator Tree Construction
      Memory Dependence Analysis
      Global Value Numbering
--      Global Value Numbering
--      Dominator Tree Construction
--      Memory Dependence Analysis
      Preliminary module verification
      Dominator Tree Construction
      Module Verifier
--      Dominator Tree Construction
--      Module Verifier
--      Preliminary module verification
--    Andersen's Interprocedural Alias Analysis
--    Target Data Layout
        %tmp11 = getelementptr %struct.FULL* %tmp9, i32 0, i32 2, i32 0       
; <float**> [#uses=1]

opt: Andersens.cpp:497: unsigned
int<unnamed>::Andersens::getNode(llvm::Value*): Assertion `0 && "Value does not
have a node in the points-to graph!"' failed.
llvm-commit/Debug/bin/opt[0x85f97ee]
llvm-commit/Debug/bin/opt[0x85f9ab4]
[0xffffe500]
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  Thu Feb 14 01:15:59 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 01:15:59 -0600
Subject: [LLVMbugs] [Bug 2032] GVN bug
In-Reply-To: <bug-2032-206@http.llvm.org/bugs/>
Message-ID: <200802140715.m1E7FxlG011355@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2032


Nick Lewycky <nicholas at mxc.ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
          Component|Scalar Optimizations        |Interprocedural Analyses
         Resolution|                            |FIXED




--- Comment #4 from Nick Lewycky <nicholas at mxc.ca>  2008-02-14 01:15:59 ---
Fixed. Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058293.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 Feb 14 08:39:59 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 08:39:59 -0600
Subject: [LLVMbugs] [Bug 2035] New: Suboptimal code storing to a bitfield
Message-ID: <bug-2035-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2035

           Summary: Suboptimal code storing to a bitfield
           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


Take the following code:

void x(struct {int a : 5, b : 5;} *c, int x) {c->a = x;}

llvm-gcc generates the following code sequence:

x:
        movl    4(%esp), %eax
        movw    $65504, %cx
        andw    (%eax), %cx
        movw    $31, %dx
        andw    8(%esp), %dx
        orw     %dx, %cx
        movw    %cx, (%eax)
        ret

gcc generates a similar sequence.  However, there is a straightforward sequence
that is two instructions shorter:

x:
movl    4(%esp), %eax
movw    8(%esp), %dx
xorw    (%eax), %dx
andw    $31, %dx
xorw    %dx, (%eax)
ret

This doesn't actually save any logical operations, but it doesn't move stuff
around as much.

I'm not sure if this bug affects real code, since I just came up with it by
experimenting, but it doesn't seem like an unusual sort of construct.  I'm also
not sure if it's actually worth using; it's quite possible I'm missing
something.  That said, it seems like something interesting to investigate.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 14 13:10:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 13:10:01 -0600
Subject: [LLVMbugs] [Bug 2035] Suboptimal code storing to a bitfield
In-Reply-To: <bug-2035-206@http.llvm.org/bugs/>
Message-ID: <200802141910.m1EJA1AT012734@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2035


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #2 from Chris Lattner <sabre at nondot.org>  2008-02-14 13:10:01 ---
Actually, the code sequence suggested is not quite correct either.  It will
toggle any bits set in x, but it won't clear any bits from c->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  Thu Feb 14 13:18:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 13:18:37 -0600
Subject: [LLVMbugs] [Bug 2029] CBE fails with -enable-correct-eh-support
In-Reply-To: <bug-2029-206@http.llvm.org/bugs/>
Message-ID: <200802141918.m1EJIbRc013029@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2029


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-02-14 13:18:37 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058308.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 Feb 14 16:38:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 16:38:08 -0600
Subject: [LLVMbugs] [Bug 2036] New: Incompatible function decls not diagnosed
Message-ID: <bug-2036-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2036

           Summary: Incompatible function decls not diagnosed
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


Invalid in C90 and C99 requiring a diagnostic.

void f();
void f (const char *fmt, ...) {}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 14 17:19:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 17:19:48 -0600
Subject: [LLVMbugs] [Bug 2030] loop index split tries to mangle PHIs
In-Reply-To: <bug-2030-206@http.llvm.org/bugs/>
Message-ID: <200802142319.m1ENJmdQ022798@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2030


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #1 from Devang Patel <dpatel at apple.com>  2008-02-14 17:19:48 ---
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058313.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 Feb 14 17:39:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 17:39:29 -0600
Subject: [LLVMbugs] [Bug 2026] passmanager crash on domfrontier requiring
	domtree?
In-Reply-To: <bug-2026-206@http.llvm.org/bugs/>
Message-ID: <200802142339.m1ENdTpL023495@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2026


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #2 from Devang Patel <dpatel at apple.com>  2008-02-14 17:39:28 ---
You ran into this earlier also :)

As a work around, drop -domtree from
  $ opt b.bc -loopsimplify -domtree -lcssa -domfrontier -ipconstprop
and drop -loops from
  $ opt -loopsimplify -loops -lcssa -instcombine -lcssa -inline



*** This bug has been marked as a duplicate of bug 1503 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 14 17:50:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 17:50:08 -0600
Subject: [LLVMbugs] [Bug 1437] RunSafely.sh doesn't run safely
In-Reply-To: <bug-1437-206@http.llvm.org/bugs/>
Message-ID: <200802142350.m1ENo8MI023901@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1437


Lauro Venancio <lauro.venancio at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |lauro.venancio at gmail.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |2.3




--- Comment #2 from Lauro Venancio <lauro.venancio at gmail.com>  2008-02-14 17:50:07 ---
Fixed.
http://llvm.org/viewvc/llvm-project?rev=47145&view=rev


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 14 19:25:13 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 19:25:13 -0600
Subject: [LLVMbugs] [Bug 2028] passmanager asserts that pass requires lower
	level pass
In-Reply-To: <bug-2028-206@http.llvm.org/bugs/>
Message-ID: <200802150125.m1F1PDXE026761@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2028


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #1 from Devang Patel <dpatel at apple.com>  2008-02-14 19:25:12 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058323.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 Feb 14 20:34:42 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 20:34:42 -0600
Subject: [LLVMbugs] [Bug 2037] New: LLVM-GCC4. 2 instructions forget to
 mention --program-prefix option.
Message-ID: <bug-2037-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2037

           Summary: LLVM-GCC4.2 instructions forget to mention --program-
                    prefix option.
           Product: Build scripts
           Version: 2.2
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Makefiles
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ttmrichter at gmail.com
                CC: llvmbugs at cs.uiuc.edu


While upgrading to LLVM-GCC-4.2 and LLVM-2.2 I stupidly followed the
instructions without giving a lot of thought and failed to notice that the
README.LLVM doesn't mention --program-prefix as an option to ./configure
anymore.  The result was me getting horribly confused as all my llvm-gcc calls
still reported a version 4.0.1 GCC.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 14 20:49:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 14 Feb 2008 20:49:38 -0600
Subject: [LLVMbugs] [Bug 2038] New: Ruby-1. 9 builds and passes light checks
 but explodes in the regression suite.
Message-ID: <bug-2038-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2038

           Summary: Ruby-1.9 builds and passes light checks but explodes in
                    the regression suite.
           Product: libraries
           Version: 2.2
          Platform: PC
               URL: http://snakeratpig.blogspot.com/2008/02/alternative-
                    compiler-suites.html
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ttmrichter at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Executive summary:
Ruby 1.9 compiles under llvm-gcc-4.2/llvm-2.2 just fine and passes sanity
checks.  Running the full regression suite fails with an "Illegal instruction
(core dumped)" message.  Running the same full regression suite with plain old
GCC causes a segmentation fault far later in the process.

Details:
Currently the test case, such as it is, is as follows:
<ol>
<li>svn co http://svn.ruby-lang.org/repos/ruby/trunk ruby-1.9</li>
<li>cd ruby-1.9</li>
<li>autoconf</li>
<li>CC=llvm-gcc ./configure</li>
<li>make</li>
<li>make test</li>
<li>make test-all</li>
</ol>

In the "make test" there are 7 errors out of 800-something tests.  This is
identical to what happens when GCC is used.  I consider this a validation of
llvm-gcc or a condemnation of badly-written test suites (your choice).  It is
the "make test-all" that shows behaviour sharply divergent from GCC.  About 2
minutes into the test (on my machine) the test suite fails with a sudden and
very complete core dump.  The nature of the test is such that I haven't yet
figured out which file is being run and what it's doing, but that is on my "to
do" list.  Any further information I get will be appended to this report.

Incidentally, GCC also fails on that test suite, but it fails after going
through a lot more tests and with a completely different error that reports
where it was and what it was doing, so whatever is happening with LLVM-GCC is
pretty drastic given its inability to report where it was when it died.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 15 00:56:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 00:56:23 -0600
Subject: [LLVMbugs] [Bug 2036] Incompatible function decls not diagnosed
In-Reply-To: <bug-2036-206@http.llvm.org/bugs/>
Message-ID: <200802150656.m1F6uNkV005431@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2036


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2008-02-15 00:56:23 ---
Fix committed.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 15 03:33:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 03:33:43 -0600
Subject: [LLVMbugs] [Bug 2039] New: clang doesn't propagate qualifiers for
 arrays correctly
Message-ID: <bug-2039-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2039

           Summary: clang doesn't propagate qualifiers for arrays correctly
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:

void a() {typedef int a[5]; typedef const a b; b r; r[0]=10;}

r's type is equivalent to "const int[5]", so the assignment should be rejected.

Not sure where exactly the bug is... maybe b is ending up with the wrong type? 
I need to think about it a bit more. This construct is kind of messy.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 15 05:01:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 05:01:38 -0600
Subject: [LLVMbugs] [Bug 2040] New: Docs for building llvm-gcc-4. 2
 incorrectly explain bootstrap option.
Message-ID: <bug-2040-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2040

           Summary: Docs for building llvm-gcc-4.2 incorrectly explain
                    bootstrap option.
           Product: Documentation
           Version: 2.2
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: General docs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ttmrichter at gmail.com
                CC: llvmbugs at cs.uiuc.edu


For GCC 4.0.1 the instructions in README.LLVM are correct (barring the
oversight reported earlier).  Things have changed for 4.2, however, as regards
bootstrapping.  Where in the old system you just had to make
LLVM_VERSION_INFO=n.n bootstrap/bootstrap2 after doing your configure, the new
system has an extra configure option required: --enable-bootstrap.  Without
that option set up, the build fails with bizarre error messages like "unknown
target '!'" and the like.  Documentation should be updated to reflect this
change.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 15 08:49:57 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 08:49:57 -0600
Subject: [LLVMbugs] [Bug 2041] New: restrict is not a C90 keyword
Message-ID: <bug-2041-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2041

           Summary: restrict is not a C90 keyword
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


Clang rejects:

int *restrict;


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 15 10:35:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 10:35:22 -0600
Subject: [LLVMbugs] [Bug 2042] New: clang incorrectly considers f(){} to
 have a prototype
Message-ID: <bug-2042-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2042

           Summary: clang incorrectly considers f(){} to have a prototype
           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


Per http://www.open-std.org/jtc1/sc22/wg14/www/docs/dr_317.htm, the following:

void f(){}
void g(){if(0)f(1);}

is valid, so clang should not print out an error.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 15 12:04:02 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 12:04:02 -0600
Subject: [LLVMbugs] [Bug 2041] restrict is not a C90 keyword
In-Reply-To: <bug-2041-206@http.llvm.org/bugs/>
Message-ID: <200802151804.m1FI42F1003335@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2041


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-15 12:04:01 ---
Fixed, thanks Neil!
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080211/004295.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 Feb 15 13:48:14 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 13:48:14 -0600
Subject: [LLVMbugs] [Bug 2043] New: llvm-gcc4.2 boostrap fails: Assertion
 'castIsValid( getOpcode(), S, Ty) && "Illegal BitCast"'
Message-ID: <bug-2043-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2043

           Summary: llvm-gcc4.2 boostrap fails: Assertion
                    'castIsValid(getOpcode(), S, Ty) && "Illegal BitCast"'
           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=1410)
 --> (http://llvm.org/bugs/attachment.cgi?id=1410)
preprocessed crtbegin.c causing compiler error

Bootstrap of llvm-gcc4.2 r47164 with llvm r47170 fails.

Compiling the preprocessed source with an earlier llvm-gcc4.2 succeeds.

Compiling with -O0 shows same error, and --disable-llvm-optzns is not accepted.
How can I get to see the bitcode causing this failure?

BTW shouldn't the bugreport url say llvm.org/bugs?

Earlier llvm-gcc that successfully compiles the preprocessed source:
$ ../../install/bin/gcc -v 
Using built-in specs.
Target: i686-pc-linux-gnu
Configured with: ../llvm-gcc4.2/configure
--prefix=/home/edwin/llvm/obj42/../install
--enable-llvm=/home/edwin/llvm/llvm-svn/trunk/ --enable-languages=c,c++ :
(reconfigured) ../llvm-gcc4.2/configure
--prefix=/home/edwin/llvm/obj42/../install
--enable-llvm=/home/edwin/llvm/llvm-svn/trunk/ --enable-languages=c :
(reconfigured) ../llvm-gcc4.2/configure
--prefix=/home/edwin/llvm/obj42/../install
--enable-llvm=/home/edwin/llvm/llvm-svn/trunk/ --enable-languages=c --no-create
--no-recursion
Thread model: posix
gcc version 4.2.1 (Based on Apple Inc. build 5546) (LLVM build)

$ uname -a
Linux thunder 2.6.25-rc1-00266-g4ee29f6-dirty #2 SMP PREEMPT Fri Feb 15
19:49:39 EET 2008 i686 GNU/Linux

$ make
 ./xgcc -B./ -B/home/edwin/llvm/obj42/../install/i686-pc-linux-gnu/bin/
-isystem /home/edwin/llvm/obj42/../install/i686-pc-linux-gnu/include -isystem
/home/edwin/llvm/obj42/../install/i686-pc-linux-gnu/sys-include
-L/home/edwin/llvm/obj42/gcc/../ld -O2  -DIN_GCC    -W -Wall -Wwrite-strings
-Wstrict-prototypes -Wmissing-prototypes -Wold-style-definition  -isystem
./include  -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/llvm-svn/trunk/include
-I/home/edwin/llvm/llvm-svn/trunk//include  -g0 -finhibit-size-directive
-fno-inline-functions -fno-exceptions -fno-zero-initialized-in-bss
-fno-toplevel-reorder  -fno-omit-frame-pointer \
          -c ../../llvm-gcc4.2/gcc/crtstuff.c -DCRT_BEGIN \
          -o crtbegin.o
cc1: Instructions.cpp:2260: llvm::BitCastInst::BitCastInst(llvm::Value*, const
llvm::Type*, const std::string&, llvm::Instruction*): Assertion
`castIsValid(getOpcode(), S, Ty) && "Illegal BitCast"' failed.
../../llvm-gcc4.2/gcc/crtstuff.c: In function ???__do_global_dtors_aux???:
../../llvm-gcc4.2/gcc/crtstuff.c:295: 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: *** [crtbegin.o] Error 1

Invoking compiler on preprocessed output:

$ /home/edwin/llvm/obj42/./gcc/xgcc -B/home/edwin/llvm/obj42/./gcc/
-B/home/edwin/llvm/obj42/../install/i686-pc-linux-gnu/bin/
-B/home/edwin/llvm/obj42/../install/i686-pc-linux-gnu/lib/ -O0 -c x.i
cc1: Instructions.cpp:2260: llvm::BitCastInst::BitCastInst(llvm::Value*, const
llvm::Type*, const std::string&, llvm::Instruction*): Assertion
`castIsValid(getOpcode(), S, Ty) && "Illegal BitCast"' failed.
../../llvm-gcc4.2/gcc/crtstuff.c: In function ???__do_global_dtors_aux???:
../../llvm-gcc4.2/gcc/crtstuff.c:295: 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  Fri Feb 15 17:09:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 17:09:34 -0600
Subject: [LLVMbugs] [Bug 2044] New: clang incorrectly accepts function
 returning pointer to vla
Message-ID: <bug-2044-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2044

           Summary: clang incorrectly accepts function returning pointer to
                    vla
           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


Testcase:

int a(int b) {int (*c(void))[b];**c() = 2;}

This violates the constraint in C99 6.7.5.2p2 because the function c has
linkage.

gcc also messes this up.

Not that we do much checking at all for this stuff at the moment, but I figured
I should file it so I don't lose 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 Feb 15 19:10:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 19:10:45 -0600
Subject: [LLVMbugs] [Bug 2045] New: #undef syntax violation
Message-ID: <bug-2045-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2045

           Summary: #undef syntax violation
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: preprocessor
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


Clang permits the following which doesn't match the #undef syntax:
#define b
#undef a b


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 15 19:21:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 15 Feb 2008 19:21:37 -0600
Subject: [LLVMbugs] [Bug 2045] #undef syntax violation
In-Reply-To: <bug-2045-206@http.llvm.org/bugs/>
Message-ID: <200802160121.m1G1Lb9q018229@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2045


Chris Lattner <sabre at nondot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
           Keywords|                            |accepts-invalid
         Resolution|                            |FIXED




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-15 19:21:36 ---
fixed, thanks!
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080211/004321.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 Feb 16 02:20:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 16 Feb 2008 02:20:30 -0600
Subject: [LLVMbugs] [Bug 2046] New: clang incorrectly rejects references to
 earlier paramters in a prototype
Message-ID: <bug-2046-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2046

           Summary: clang incorrectly rejects references to earlier
                    paramters in a prototype
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
int aa(int b, int x[sizeof b]) {}

This is perfectly legal C90 and C99 code, but clang claims the b is undeclared.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 16 04:11:35 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 16 Feb 2008 04:11:35 -0600
Subject: [LLVMbugs] [Bug 2048] New: Invalid division incombine with division
 by constants
Message-ID: <bug-2048-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2048

           Summary: Invalid division incombine with division by constants
           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


Take the following function:

int i(int a) {return a/-1431655765/3;}

LLVM concludes that this is equivalent to the following:

define i32 @i(i32 %a) nounwind  {
entry:
        ret i32 %a
}

This is clearly 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 Feb 16 04:02:15 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 16 Feb 2008 04:02:15 -0600
Subject: [LLVMbugs] [Bug 2047] New: Instcombine doesn't know that
	"a-b-c-a==-b-c"
Message-ID: <bug-2047-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2047

           Summary: Instcombine doesn't know that "a-b-c-a==-b-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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Per summary: llvm doesn't simplify a function like the following:

int l(int a,int b,int c) {return a-b-c-a;}

to not reference a.

Probably not that important... but possibly interesting.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 16 06:15:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 16 Feb 2008 06:15:45 -0600
Subject: [LLVMbugs] [Bug 2049] New: clang: CGExprScalar.cpp:1060:
 llvm::Value* clang::CodeGen:: CodeGenFunction::EmitScalarExpr(const
 clang::Expr*): Assertion `E && !hasAggregateLLVMType(E->getType()) && "
 Invalid scalar expression to emit"' failed.
Message-ID: <bug-2049-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2049

           Summary: clang: CGExprScalar.cpp:1060: llvm::Value*
                    clang::CodeGen::CodeGenFunction::EmitScalarExpr(const
                    clang::Expr*): Assertion `E && !hasAggregateLLVMType(E-
                    >getType()) && "Invalid scalar expression to emit"'
                    failed.
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1413)
 --> (http://llvm.org/bugs/attachment.cgi?id=1413)
reduced testcase

Using clang SVN r47211, llvm SVN r47211 I get a codegen assert failure:

$ uname -a
Linux lightspeed2 2.6.24-1-amd64 #1 SMP Mon Feb 11 13:47:43 UTC 2008 x86_64
GNU/Linux

$ clang -emit-llvm-bc testcase-min.i
clang: CGExprScalar.cpp:1060: llvm::Value*
clang::CodeGen::CodeGenFunction::EmitScalarExpr(const clang::Expr*): Assertion
`E && !hasAggregateLLVMType(E->getType()) && "Invalid scalar expression to
emit"' failed.
clang[0x68b306]
/lib/libc.so.6[0x3491231240]
/lib/libc.so.6(gsignal+0x35)[0x34912311d5]
/lib/libc.so.6(abort+0x110)[0x3491232680]
/lib/libc.so.6(__assert_fail+0xef)[0x349122a75f]
clang(_ZN5clang7CodeGen15CodeGenFunction14EmitScalarExprEPKNS_4ExprE+0x3f)[0x4d585f]
clang[0x4c918c]
clang(_ZN5clang7CodeGen15CodeGenFunction11EmitAggExprEPKNS_4ExprEPN4llvm5ValueEb+0x72)[0x4c9762]
clang(_ZN5clang7CodeGen15CodeGenFunction8EmitStmtEPKNS_4StmtE+0x13c)[0x4db38c]
clang(_ZN5clang7CodeGen15CodeGenFunction16EmitCompoundStmtERKNS_12CompoundStmtEbPN4llvm5ValueEb+0x71)[0x4dccf1]
clang(_ZN5clang7CodeGen15CodeGenFunction8EmitStmtEPKNS_4StmtE+0x2bb)[0x4db50b]
clang(_ZN5clang7CodeGen15CodeGenFunction12GenerateCodeEPKNS_12FunctionDeclE+0x2ed)[0x4c0d5d]
clang(_ZN5clang7CodeGen13CodeGenModule12EmitFunctionEPKNS_12FunctionDeclE+0x33)[0x4b3903]
clang[0x4b33cb]
clang(_ZN5clang8ParseASTERNS_12PreprocessorEPNS_11ASTConsumerEb+0x446)[0x513ab6]
clang[0x4aa0df]
clang(main+0xa3a)[0x4acaea]
/lib/libc.so.6(__libc_start_main+0xf4)[0x349121e1c4]
clang[0x48e789]
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 Feb 16 06:27:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 16 Feb 2008 06:27:03 -0600
Subject: [LLVMbugs] [Bug 2050] New: incorrect "error: initializer element is
 not constant"
Message-ID: <bug-2050-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2050

           Summary: incorrect  "error: initializer element is not constant"
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Using clang r47211, and llvm r47211 I cannot compile the code below.
gcc and icc compile the code without errors.

$ clang -fsyntax-only testcase-min.i
testcase-min.i:10:9: error: initializer element is not constant
        {"OPEN", 1, &cdiff_cmd_open }
        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1 diagnostic generated.

$ gcc -Wall -c testcase-min.i
testcase-min.i:9: warning: ???commands??? defined but not used
$ icc -w2 -c testcase-min.i
testcase-min.i(9): remark #177: variable "commands" was declared but never
referenced
  static struct cdiff_cmd commands[] = {
                          ^

$ cat testcase-min.i
struct cdiff_cmd {
          const char *name;
          unsigned short argc;
          int (*handler)();
};
static int cdiff_cmd_open() {
  return 0;
}
static struct cdiff_cmd commands[] = {
        {"OPEN", 1, &cdiff_cmd_open }
};


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 16 14:35:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 16 Feb 2008 14:35:08 -0600
Subject: [LLVMbugs] [Bug 2051] New: llvm-gcc configure should default to
 native processor
Message-ID: <bug-2051-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2051

           Summary: llvm-gcc configure should default to native processor
           Product: tools
           Version: 2.2
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu


We should change the llvm-gcc configure process to build code for the current
processor unless otherwise specified.  This would allow users who have machines
with SSE2+ to automatically get good FP code without having to explicitly
specify 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  Sun Feb 17 07:44:21 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 17 Feb 2008 07:44:21 -0600
Subject: [LLVMbugs] [Bug 2052] New: Invalid instcombine for float division
	with undef
Message-ID: <bug-2052-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2052

           Summary: Invalid instcombine for float division with undef
           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


define float @a(float %x) {
%res = fdiv float undef, %x
ret float %res
}

instcombines to:

define float @a(float %x) nounwind  {
        ret float 0.000000e+00
}

If I understand the rules for undef correctly, this is an invalid
transformation because since x/NaN is NaN for all x, the division returns an
impossible result in the case where %x is an NaN.

I think a correct transformation is undef/X->NaN, because NaN/x is NaN for all
x. In the case of signaling floating point, we can't actually do the
transformation at all, because the operation might be required to signal.  But
we normally assume no floating point exceptions, right?


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 17 09:20:04 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 17 Feb 2008 09:20:04 -0600
Subject: [LLVMbugs] [Bug 2053] New: Inefficient code generated checking
 32-bit multiply for signed overflow on x86
Message-ID: <bug-2053-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2053

           Summary: Inefficient code generated checking 32-bit multiply for
                    signed overflow on x86
           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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


The generated code on x86 for checking for signed overflow on a multiply the
obvious way is much longer than it needs to be.

Original C code:

int x(int a, int b) {long long prod = (long long)a*b; return  prod > 0x7FFFFFFF
|| prod < (-0x7FFFFFFF-1);}

Generated LLVM IL (generated using clang -emit-llvm-bc | opt
-std-compile-opts):
define i32 @x(i32 %a, i32 %b) nounwind  {
entry:
        %conv = sext i32 %a to i64              ; <i64> [#uses=1]
        %conv2 = sext i32 %b to i64             ; <i64> [#uses=1]
        %mul = mul i64 %conv2, %conv            ; <i64> [#uses=2]
        %cmp = icmp sgt i64 %mul, 2147483647            ; <i1> [#uses=1]
        br i1 %cmp, label %UnifiedReturnBlock, label %lor_rhs

lor_rhs:                ; preds = %entry
        %cmp6 = icmp slt i64 %mul, -2147483648          ; <i1> [#uses=1]
        %phitmp = zext i1 %cmp6 to i32          ; <i32> [#uses=1]
        ret i32 %phitmp

UnifiedReturnBlock:             ; preds = %entry
        ret i32 1
}

Code generated by llc:
x:
        pushl   %ebx
        pushl   %esi
        movl    16(%esp), %eax
        imull   12(%esp)
        testl   %eax, %eax
        sets    %cl
        movzbw  %cl, %cx
        testl   %edx, %edx
        setg    %bl
        movzbw  %bl, %si
        cmove   %cx, %si
        movw    %si, %cx
        testb   $1, %cl
        je      .LBB1_3 # lor_rhs
.LBB1_1:        # UnifiedReturnBlock
        movl    $1, %eax
.LBB1_2:        # UnifiedReturnBlock
        popl    %esi
        popl    %ebx
        ret
.LBB1_3:        # lor_rhs
        cmpl    $2147483648, %eax
        setb    %al
        movzbw  %al, %ax
        cmpl    $4294967295, %edx
        setl    %cl
        movzbw  %cl, %cx
        cmove   %ax, %cx
        movw    %cx, %ax
        andl    $1, %eax
        jmp     .LBB1_2 # UnifiedReturnBlock

Code generated by gcc:
x:
        movl    8(%esp), %eax
        imull   4(%esp)
        addl    $-2147483648, %eax
        adcl    $0, %edx
        xorl    %eax, %eax
        cmpl    $0, %edx
        seta    %al
        ret

LLVM seems to be able to deal with the following variant much more easily:

int x(int a, int b) {long long prod = (long long)a*b; return (prod >
0x7FFFFFFF) | (prod < (-0x7FFFFFFF-1));}

Note that this code uses "|" instead of "||".

Generated LLVM IL (generated using clang -emit-llvm-bc | opt
-std-compile-opts):
define i32 @x(i32 %a, i32 %b) nounwind  {
entry:
        %conv = sext i32 %a to i64              ; <i64> [#uses=1]
        %conv2 = sext i32 %b to i64             ; <i64> [#uses=1]
        %mul = mul i64 %conv2, %conv            ; <i64> [#uses=1]
        %mul.off = add i64 %mul, 2147483648             ; <i64> [#uses=1]
        %or1 = icmp ugt i64 %mul.off, 4294967295                ; <i1>
[#uses=1]
        %or = zext i1 %or1 to i32               ; <i32> [#uses=1]
        ret i32 %or
}

Code generated by llc:
 x:
        movl    8(%esp), %eax
        imull   4(%esp)
        addl    $2147483648, %eax
        adcl    $0, %edx
        testl   %edx, %edx
        setne   %al
        movzbl  %al, %eax
        ret

Here's the ideal code as far as I know (handwritten by myself; not really
tested, but all it does is use the overflow flag set by imull):
x:
        movl    8(%esp), %eax
        imull   4(%esp)
        setc    %al
        movzbl  %al, %eax
        ret


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 17 13:32:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 17 Feb 2008 13:32:01 -0600
Subject: [LLVMbugs] [Bug 2042] clang incorrectly considers f(){} to have a
	prototype
In-Reply-To: <bug-2042-206@http.llvm.org/bugs/>
Message-ID: <200802171932.m1HJW1UR004931@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2042


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-17 13:32:01 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080211/004330.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 Feb 17 13:46:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 17 Feb 2008 13:46:28 -0600
Subject: [LLVMbugs] [Bug 2053] Inefficient code generated checking 32-bit
 multiply for signed overflow on x86
In-Reply-To: <bug-2053-206@http.llvm.org/bugs/>
Message-ID: <200802171946.m1HJkSH9005396@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2053


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-17 13:46:28 ---
I moved this to the X86 readme file:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058422.html

which is where we generally track non-trivial target-specific
micro-optimizations.  We also have lib/Target/README.txt for target-indep
microoptimizations which can be useful as well.

It's never clear what optimizations should go in a readme vs what should go in
bugzilla, but a general rule of thumb is that things should go in bugzilla if
they block other bigger optimization issues (where bugzillas dependency
tracking stuff can be useful) or when they themselves are a big deal for one
reason or another.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 17 15:04:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 17 Feb 2008 15:04:16 -0600
Subject: [LLVMbugs] [Bug 2047] reassociate doesn't know that "a-b-c-a==-b-c"
In-Reply-To: <bug-2047-206@http.llvm.org/bugs/>
Message-ID: <200802172104.m1HL4G19007799@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2047


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #4 from Chris Lattner <sabre at nondot.org>  2008-02-17 15:04:16 ---
second half fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080211/058429.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 Feb 17 15:51:59 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 17 Feb 2008 15:51:59 -0600
Subject: [LLVMbugs] [Bug 2054] New: llvm-ld crashes with Assertion `0 &&
 "Unmapped global?"' failed. when linking code with weakrefs on linux amd64
Message-ID: <bug-2054-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2054

           Summary: llvm-ld crashes with Assertion `0 && "Unmapped global?"'
                    failed. when linking code with weakrefs on linux amd64
           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: belyshev at depni.sinp.msu.ru
                CC: llvmbugs at cs.uiuc.edu


Testcase:

int foo;
static int bar __attribute__ ((weakref ("foo")));

int baz (void) 
{
  return bar;
}

$ gcc -emit-llvm -c bug.c           
$ llvm-ld bug.o
llvm-ld: LinkModules.cpp:309: llvm::Value* RemapOperand(const llvm::Value*,
std::map<const llvm::Value*, llvm::Value*, std::less<const llvm::Value*>,
std::allocator<std::pair<const llvm::Value* const, llvm::Value*> > >&):
Assertion `0 && "Unmapped global?"' failed.
llvm-ld[0x66e0a1]
/lib/libc.so.6[0x7f5ba9692240]
/lib/libc.so.6(gsignal+0x35)[0x7f5ba96921d5]
/lib/libc.so.6(abort+0x110)[0x7f5ba9693680]
/lib/libc.so.6(__assert_fail+0xef)[0x7f5ba968b75f]
llvm-ld[0x4a3671]
llvm-ld[0x4a397e]
llvm-ld(_ZN4llvm6Linker11LinkModulesEPNS_6ModuleES2_PSs+0x9f4)[0x4a7e44]
llvm-ld(_ZN4llvm6Linker10LinkInFileERKNS_3sys4PathERb+0x3ce)[0x4a02ce]
llvm-ld(_ZN4llvm6Linker11LinkInItemsERKSt6vectorISt4pairISsbESaIS3_EERS5_+0xd9)[0x4a1049]
llvm-ld(main+0x9b3)[0x4851d3]
/lib/libc.so.6(__libc_start_main+0xf4)[0x7f5ba967f1c4]
llvm-ld(sinh+0x41)[0x47da29]
Abort 
$

gcc version is: gcc version 4.2.1 (Based on Apple Inc. build 5546) (LLVM build)
llvm version is: trunk r47234
platform is: linux amd64 (x86_64-unknown-linux-gnu)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 17 20:10:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 17 Feb 2008 20:10:29 -0600
Subject: [LLVMbugs] [Bug 2055] New: LLVM should have a nocapture parameter
 attribute for functions that don' t let a pointer value escape
Message-ID: <bug-2055-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2055

           Summary: LLVM should have a nocapture parameter attribute for
                    functions that don't let a pointer value escape
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: resistor at mac.com
                CC: llvmbugs at cs.uiuc.edu


Chris pointed out that it would be useful to have a "nocapture" parameter
attribute.  Its semantics would be that a pointer value passed in with the
"nocapture" attribute can be assumed never to escape from the callee (never
stored to a global or passed to another function where it could escape).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:26:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:26:16 -0600
Subject: [LLVMbugs] [Bug 2056] New: Significant missed block merging
	opportunity
Message-ID: <bug-2056-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2056

           Summary: Significant missed block merging opportunity
           Product: libraries
           Version: 2.2
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: code-quality
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu


Applying this patch to simplifycfg disables cloning of return instructions into
predecessor blocks.  This should be a simple code size win, and it allows
simplifycfg to be better at turning simple cfg's into select instructions:

--- SimplifyCFG.cpp     (revision 47234)
+++ SimplifyCFG.cpp     (working copy)
@@ -1245,7 +1245,7 @@
       }

       // If we found some, do the transformation!
-      if (!UncondBranchPreds.empty()) {
+      if (0 && !UncondBranchPreds.empty()) {
         while (!UncondBranchPreds.empty()) {
           BasicBlock *Pred = UncondBranchPreds.back();
           DOUT << "FOLDING: " << *BB

However, this causes significant code size pessimizations.  In the case where
the return  block is cloned, the .llvm.bc file ends up with lots of blocks like
this (examples from 'agrep' in multisource/benchmarks):

bb393.i.i:              ; preds = %bb388.i.i
        %tmp394.i.i = load i32* @num_of_matched, align 4                ; <i32>
[#uses=1]
        %tmp395.i.i = add i32 %tmp394.i.i, 1            ; <i32> [#uses=1]
        store i32 %tmp395.i.i, i32* @num_of_matched, align 4
        %tmp396.i.i = call i32 @puts( i8* getelementptr ([256 x i8]*
@CurrentFileName, i32 0, i32 0) ) nounwind                 ; <i32> [#uses=0]
        ret void

When the return is not cloned, we end up with lots of code like this:

bb393.i.i:              ; preds = %bb388.i.i
        %tmp394.i.i = load i32* @num_of_matched, align 4                ; <i32>
[#uses=1]
        %tmp395.i.i = add i32 %tmp394.i.i, 1            ; <i32> [#uses=1]
        store i32 %tmp395.i.i, i32* @num_of_matched, align 4
        %tmp396.i.i = call i32 @puts( i8* getelementptr ([256 x i8]*
@CurrentFileName, i32 0, i32 0) ) nounwind                 ; <i32> [#uses=0]
        br label %UnifiedReturnBlock

where UnifiedReturnBlock does a simple return.

The major difference between these is that in the former case, the code
generator's block merging pass is able to merge away all of these common
blocks, but when the return is shared, this optimization doesn't happen.  Since
this sort of thing really happens due to inlining (e.g where F calls G multiple
times), we should be able to eliminate this redundancy at the LLVM level, for
example in simplifycfg.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:38:44 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:38:44 -0600
Subject: [LLVMbugs] [Bug 410] Improvements in the MachineInstr/Operand
	classes
In-Reply-To: <bug-410-206@http.llvm.org/bugs/>
Message-ID: <200802180638.m1I6ci3D024502@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=410


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #5 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:38:44 ---
#5 is done with MemOperand.  Closing.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:42:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:42:11 -0600
Subject: [LLVMbugs] [Bug 698] Clean up LLVM error and warning output
In-Reply-To: <bug-698-206@http.llvm.org/bugs/>
Message-ID: <200802180642.m1I6gBoM024608@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=698


Chris Lattner <sabre at nondot.org> changed:

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




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:43:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:43:34 -0600
Subject: [LLVMbugs] [Bug 731] Enhanced Makefile System
In-Reply-To: <bug-731-206@http.llvm.org/bugs/>
Message-ID: <200802180643.m1I6hYCA024667@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=731


Chris Lattner <sabre at nondot.org> changed:

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




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:44:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:44:53 -0600
Subject: [LLVMbugs] [Bug 758] [spiller] Spiller could optimize spill code
 generation for 2-addr folded instructions better
In-Reply-To: <bug-758-206@http.llvm.org/bugs/>
Message-ID: <200802180644.m1I6ir6H024734@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=758


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #2 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:44:53 ---
I think that has been implemented, and the .bc file is a 1.x file that can't be
read to verify 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  Mon Feb 18 00:45:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:45:38 -0600
Subject: [LLVMbugs] [Bug 762] TargetData needs to be able to specify
 alignment of vector data types
In-Reply-To: <bug-762-206@http.llvm.org/bugs/>
Message-ID: <200802180645.m1I6jcXX024764@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=762


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:45:38 ---
TargetData does this.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:46:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:46:43 -0600
Subject: [LLVMbugs] [Bug 768] Register scavenging
In-Reply-To: <bug-768-206@http.llvm.org/bugs/>
Message-ID: <200802180646.m1I6khx9024809@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=768


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #4 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:46:42 ---
We have a register scavenger now, targets can switch over to using it as they
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  Mon Feb 18 00:48:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:48:12 -0600
Subject: [LLVMbugs] [Bug 776] LLVM register allocator issues
In-Reply-To: <bug-776-206@http.llvm.org/bugs/>
Message-ID: <200802180648.m1I6mCLc024861@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=776


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #6 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:48:12 ---
Evan, please file relevant subtasks of this bug as their own separate bugs. 
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 Feb 18 00:49:24 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:49:24 -0600
Subject: [LLVMbugs] [Bug 793] Factor LeakDetector.cpp out of lib/VMCore and
 into lib/Support
In-Reply-To: <bug-793-206@http.llvm.org/bugs/>
Message-ID: <200802180649.m1I6nOdW024909@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=793


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #12 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:49:24 ---
There isn't a great need for this.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:50:41 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:50:41 -0600
Subject: [LLVMbugs] [Bug 903] Better way to handle asm writing from the
	tools.
In-Reply-To: <bug-903-206@http.llvm.org/bugs/>
Message-ID: <200802180650.m1I6ofdB024965@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=903


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:50:41 ---
There isn't interest in this.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:54:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:54:34 -0600
Subject: [LLVMbugs] [Bug 1125] Update autoconf and libtool
In-Reply-To: <bug-1125-206@http.llvm.org/bugs/>
Message-ID: <200802180654.m1I6sYif025110@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1125


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #4 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:54:33 ---
Upgrading for the sake of upgrading isn't interesting.  


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 00:59:40 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 00:59:40 -0600
Subject: [LLVMbugs] [Bug 1353] Code generator needs DominatorTree
In-Reply-To: <bug-1353-206@http.llvm.org/bugs/>
Message-ID: <200802180659.m1I6xeo6025316@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1353


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #2 from Chris Lattner <sabre at nondot.org>  2008-02-18 00:59:40 ---
Owen implemented this.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 01:01:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 01:01:22 -0600
Subject: [LLVMbugs] [Bug 1372] Adding PR to Dejagnu tests has impacted
	nightly tester
In-Reply-To: <bug-1372-206@http.llvm.org/bugs/>
Message-ID: <200802180701.m1I71Mtp025394@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1372


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-18 01:01:21 ---
The nightly tester seems fine.  It's strange to consider PR#s as part of the
test name, but doesn't seem to be a problem in practice.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 01:02:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 01:02:55 -0600
Subject: [LLVMbugs] [Bug 1433] X86 isel problem with flags
In-Reply-To: <bug-1433-206@http.llvm.org/bugs/>
Message-ID: <200802180702.m1I72tVE025456@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1433


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #4 from Chris Lattner <sabre at nondot.org>  2008-02-18 01:02:55 ---
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  Mon Feb 18 01:06:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 01:06:19 -0600
Subject: [LLVMbugs] [Bug 1547] Documentation Requirements
In-Reply-To: <bug-1547-206@http.llvm.org/bugs/>
Message-ID: <200802180706.m1I76JCE025639@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1547


Chris Lattner <sabre at nondot.org> changed:

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




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 01:07:18 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 01:07:18 -0600
Subject: [LLVMbugs] [Bug 1552] [meta] Fix llvm-gcc 4.2 to be usable
In-Reply-To: <bug-1552-206@http.llvm.org/bugs/>
Message-ID: <200802180707.m1I77IIh025725@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1552


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #19 from Chris Lattner <sabre at nondot.org>  2008-02-18 01:07:18 ---
4.2 is "usable" 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  Mon Feb 18 01:16:18 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 01:16:18 -0600
Subject: [LLVMbugs] [Bug 1747] verification does not detect calling
 functions with incorrect calling convention
In-Reply-To: <bug-1747-206@http.llvm.org/bugs/>
Message-ID: <200802180716.m1I7GIAV026112@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1747


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #7 from Chris Lattner <sabre at nondot.org>  2008-02-18 01:16:18 ---
Verifier shouldn't reject this, it is valid .ll 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  Mon Feb 18 01:20:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 01:20:37 -0600
Subject: [LLVMbugs] [Bug 1844] Build problem on Fedora core 7 x86_64
In-Reply-To: <bug-1844-206@http.llvm.org/bugs/>
Message-ID: <200802180720.m1I7KbWH026388@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1844


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #4 from Chris Lattner <sabre at nondot.org>  2008-02-18 01:20:37 ---
Devang removed ltdl.c, so this should be 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 Feb 18 01:21:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 01:21:11 -0600
Subject: [LLVMbugs] [Bug 1855] __sync_synchronize()
In-Reply-To: <bug-1855-206@http.llvm.org/bugs/>
Message-ID: <200802180721.m1I7LBxw026494@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1855


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-18 01:21:11 ---
I think that Andrew implemented this.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 01:28:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 01:28:23 -0600
Subject: [LLVMbugs] [Bug 1521] Revamp LLVM by-value structure passing
In-Reply-To: <bug-1521-206@http.llvm.org/bugs/>
Message-ID: <200802180728.m1I7SNDM026907@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1521


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

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




--- Comment #22 from Evan Cheng <evan.cheng at apple.com>  2008-02-18 01:28:23 ---
Nothing.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 07:59:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 07:59:34 -0600
Subject: [LLVMbugs] [Bug 2000] Rejects valid C90 and C99
In-Reply-To: <bug-2000-206@http.llvm.org/bugs/>
Message-ID: <200802181359.m1IDxYrq017855@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2000


Neil Booth <neil at daikokuya.co.uk> changed:

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




--- Comment #8 from Neil Booth <neil at daikokuya.co.uk>  2008-02-18 07:59:33 ---
Steve - I don't think your fix is transparent to typedefs?


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 08:40:06 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 08:40:06 -0600
Subject: [LLVMbugs] [Bug 2057] New: miscompilation by licm
Message-ID: <bug-2057-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2057

           Summary: miscompilation by licm
           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=1420)
 --> (http://llvm.org/bugs/attachment.cgi?id=1420)
testcase .ll

Some recent change introduced a problem with licm.

$ llvm-as testcase.ll
$ llc testcase.bc
$ gcc -o testcase testcase.s
$ ./testcase
$ opt -licm testcase.bc > badcase.bc
$ llc badcase.bc
$ gcc -o badcase badcase.s
$ ./badcase
Aborted (core dumped)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 09:16:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 09:16:34 -0600
Subject: [LLVMbugs] [Bug 2000] Rejects valid C90 and C99
In-Reply-To: <bug-2000-206@http.llvm.org/bugs/>
Message-ID: <200802181516.m1IFGY8b020210@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2000


snaroff at apple.com <snaroff at apple.com> changed:

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




--- Comment #9 from snaroff at apple.com  2008-02-18 09:16:33 ---
Fixed in http://llvm.org/viewvc/llvm-project?view=rev&revision=47269.

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 Feb 18 10:00:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 10:00:30 -0600
Subject: [LLVMbugs] [Bug 2041] restrict is not a C90 keyword
In-Reply-To: <bug-2041-206@http.llvm.org/bugs/>
Message-ID: <200802181600.m1IG0UN5022036@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2041


Neil Booth <neil at daikokuya.co.uk> changed:

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




--- Comment #2 from Neil Booth <neil at daikokuya.co.uk>  2008-02-18 10:00:29 ---
__restrict is not being accepted in C90 mode.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 11:12:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 11:12:23 -0600
Subject: [LLVMbugs] [Bug 2058] New: Missed memcpy optimization
Message-ID: <bug-2058-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2058

           Summary: Missed memcpy optimization
           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=1421)
 --> (http://llvm.org/bugs/attachment.cgi?id=1421)
testcase .ll

In the attached testcase there are two memcpy's.  In theory one
could be eliminated, for example by using %agg.result instead of
%temp everywhere.  I admit that this is tricky - but not impossible!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 11:25:06 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 11:25:06 -0600
Subject: [LLVMbugs] [Bug 2059] New: Missed sret memcpy optimization
Message-ID: <bug-2059-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2059

           Summary: Missed sret memcpy optimization
           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=1423)
 --> (http://llvm.org/bugs/attachment.cgi?id=1423)
testcase .ll

The recent sret+memcpy simplification assumes that the memcpy is to
a noalias sret parameter.  But there seems to be no real reason for
that, it just needs to be to something that can be proved not to be
accessed by the called function.  For example another temporary.
Here is a testcase: both of the memcpy's could be removed by passing
the destination directly to the called function.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 11:38:15 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 11:38:15 -0600
Subject: [LLVMbugs] [Bug 1579] Use DESTDIR,
	not --prefix on Darwin llvm-gcc builds
In-Reply-To: <bug-1579-206@http.llvm.org/bugs/>
Message-ID: <200802181738.m1IHcFS2025247@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1579


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #2 from Devang Patel <dpatel at apple.com>  2008-02-18 11:38:14 ---
I do not see any issue here. 

$ llvm-gcc-4.2 -v -isysroot /blah foo.c
Using built-in specs.
Target: i686-apple-darwin9
Configured with: /Volumes/Nanpura/w/llvmgcc42.roots/llvmgcc42~obj/src/configure
--disable-checking --enable-werror --prefix=/Developer/usr/llvm-gcc-4.2
--mandir=/Developer/usr/llvm-gcc-4.2/share/man
--enable-languages=c,objc,c++,obj-c++ --program-prefix=llvm-
--program-transform-name=/^[cg][^.-]*$/s/$/-4.2/
--with-gxx-include-dir=/usr/include/c++/4.0.0 --with-slibdir=/usr/lib
--build=i686-apple-darwin9 --enable-llvm=/usr/local --host=i686-apple-darwin9
--target=i686-apple-darwin9
Thread model: posix
gcc version 4.2.1 (Based on Apple Inc. build 5546) (LLVM build 00)
 /Developer/usr/llvm-gcc-4.2/libexec/gcc/i686-apple-darwin9/4.2.1/cc1 -quiet -v
-isysroot /blah -D__DYNAMIC__ -isysroot /blah foo.c -fPIC -quiet -dumpbase
foo.c -mmacosx-version-min=10.5.1 -mtune=core2 -auxbase foo -version -o
/var/tmp//cc7txUYD.s
ignoring nonexistent directory "/blah/usr/local/include"
ignoring nonexistent directory "/blah/Developer/usr/llvm-gcc-4.2/include"
ignoring nonexistent directory
"/blah/Developer/usr/llvm-gcc-4.2/lib/gcc/i686-apple-darwin9/4.2.1/include"
ignoring nonexistent directory
"/blah/Developer/usr/llvm-gcc-4.2/lib/gcc/i686-apple-darwin9/4.2.1/../../../../i686-apple-darwin9/include"
ignoring nonexistent directory "/blah/usr/include"
ignoring nonexistent directory "/blah/System/Library/Frameworks"
ignoring nonexistent directory "/blah/Library/Frameworks"
#include "..." search starts here:
#include <...> search starts here:
End of search list.
GNU C version 4.2.1 (Based on Apple Inc. build 5546) (LLVM build 00)
(i686-apple-darwin9)
        compiled by GNU C version 4.2.1 (Based on Apple Inc. build 5546) (LLVM
build 00).
GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 12:03:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 12:03:43 -0600
Subject: [LLVMbugs] [Bug 1720] lli gives double free error on AMD64
In-Reply-To: <bug-1720-206@http.llvm.org/bugs/>
Message-ID: <200802181803.m1II3haA026097@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1720


Sterling Stuart Stein <stein at ir.iit.edu> changed:

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




--- Comment #8 from Sterling Stuart Stein <stein at ir.iit.edu>  2008-02-18 12:03:42 ---
The bug appears to be fixed now.  Last weekend, the test suite
ran without me having to manually intervene.  I ran make on
MultiSource/d and it went smoothly.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 18 16:48:52 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 16:48:52 -0600
Subject: [LLVMbugs] [Bug 2048] Invalid division incombine with division by
	constants
In-Reply-To: <bug-2048-206@http.llvm.org/bugs/>
Message-ID: <200802182248.m1IMmqZ6006588@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2048


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #13 from Nick Lewycky <nicholas at mxc.ca>  2008-02-18 16:48:52 ---
Thanks for the review, Eli! Patch checked in:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058483.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 Feb 18 18:06:49 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 18:06:49 -0600
Subject: [LLVMbugs] [Bug 2060] New: llvm-as crash
Message-ID: <bug-2060-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2060

           Summary: llvm-as crash
           Product: tools
           Version: 2.2
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: minor
          Priority: P2
         Component: llvm-as
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1425)
 --> (http://llvm.org/bugs/attachment.cgi?id=1425)
Crashing .ll file

llvm-as crashes on the following program instead of giving a type error:
--
define i8* @foo() {
        ret i8* 0
}
--
assert log:
--
ddunbar at ozzy:lli$ llvm-as bug.ll
llvm-as: /home/ddunbar/public/llvm/llvm/include/llvm/Support/Casting.h:199:
typename llvm::cast_retty<To, From>::ret_type llvm::cast(const Y&) [with X =
llvm::IntegerType, Y = const llvm::Type*]: Assertion `isa<X>(Val) &&
"cast<Ty>() argument of incompatible type!"' failed.
llvm-as((anonymous namespace)::PrintStackTrace()+0x1a)[0x817ac7a]
/lib/tls/i686/cmov/libc.so.6(abort+0x109)[0xb7d00fb9]
/lib/tls/i686/cmov/libc.so.6(__assert_fail+0x10f)[0xb7cf8fbf]
llvm-as(llvm::ConstantInt::isValueValidForType(llvm::Type const*, long
long)+0x0)[0x80cca30]
[0x0]
Aborted
--
tested on 2.2 and 2.1


-- 
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  Mon Feb 18 22:36:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 18 Feb 2008 22:36:53 -0600
Subject: [LLVMbugs] [Bug 2060] llvm-as crash
In-Reply-To: <bug-2060-206@http.llvm.org/bugs/>
Message-ID: <200802190436.m1J4ar9D017599@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2060


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-18 22:36:53 ---
Fixed, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058502.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 Feb 19 00:23:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 00:23:27 -0600
Subject: [LLVMbugs] [Bug 2052] suboptimal instcombine for float division
 with undef, bad fold of uitofp(undef)
In-Reply-To: <bug-2052-206@http.llvm.org/bugs/>
Message-ID: <200802190623.m1J6NRfQ020569@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2052


Chris Lattner <sabre at nondot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
            Summary|Invalid instcombine for     |suboptimal instcombine for
                   |float division with undef   |float division with undef,
                   |                            |bad fold of uitofp(undef)




--- Comment #7 from Chris Lattner <sabre at nondot.org>  2008-02-19 00:23:26 ---
Fixed, patches here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058506.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058507.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 Feb 19 00:46:41 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 00:46:41 -0600
Subject: [LLVMbugs] [Bug 2041] restrict is not a C90 keyword
In-Reply-To: <bug-2041-206@http.llvm.org/bugs/>
Message-ID: <200802190646.m1J6kffe021434@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2041


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-02-19 00:46:41 ---
Fixed, patch here:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080218/004353.html

Thanks Neil


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 00:47:35 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 00:47:35 -0600
Subject: [LLVMbugs] [Bug 2057] miscompilation by basicaa
In-Reply-To: <bug-2057-206@http.llvm.org/bugs/>
Message-ID: <200802190647.m1J6lZYG021478@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2057


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #3 from Owen Anderson <resistor at mac.com>  2008-02-19 00:47:35 ---
Fixed with r47317.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 01:26:36 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 01:26:36 -0600
Subject: [LLVMbugs] [Bug 2061] New: Remove the tail duplication pass for
 llvm-gcc & -std-compile-opts
Message-ID: <bug-2061-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2061

           Summary: Remove the tail duplication pass for llvm-gcc & -std-
                    compile-opts
           Product: libraries
           Version: 2.2
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: code-quality
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu
            Blocks: 1622,2056


We should remove taildup from being run by default.  It causes lots of problems
(e.g. Bug 1622) and causes code growth.  We also want to do code factoring (Bug
2056), but it doesn't make a lot of sense to do code factoring while we're
still duplicating code around willy nilly.

The hard part of doing this isn't disabling the pass itself, it is doing the
performance analysis to make sure that we're not losing anything important.  If
we do lose some performance in some case that matters, we should figure out if
a limited form of tail dup is reasonable or if the optimization can be done
another 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  Tue Feb 19 01:40:20 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 01:40:20 -0600
Subject: [LLVMbugs] [Bug 1705] lower-packed pass fails assertion on bitcast
In-Reply-To: <bug-1705-206@http.llvm.org/bugs/>
Message-ID: <200802190740.m1J7eKsK023184@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1705


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #8 from Chris Lattner <sabre at nondot.org>  2008-02-19 01:40:19 ---
Based on Comment #5, I see you're not using it.  Fixed here: :)
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058515.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 Feb 19 03:36:46 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 03:36:46 -0600
Subject: [LLVMbugs] [Bug 928] Investigate cases where running llvm optimizer
 twice slows down code
In-Reply-To: <bug-928-206@http.llvm.org/bugs/>
Message-ID: <200802190936.m1J9akkD002586@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=928


Bill Wendling <isanbard at gmail.com> changed:

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




--- Comment #4 from Bill Wendling <isanbard at gmail.com>  2008-02-19 03:36:44 ---
This doesn't appear to be happening anymore. The fibo execution time with -O0
and with -O2 is pretty much the same:

[Gaz:Shootout-C++] time ./Output/fibo.llc  # With -O2
701408733

real    0m10.129s
user    0m9.935s
sys     0m0.034s
[Gaz:Shootout-C++] time ./fibo.O0.llc  # With -O0
701408733

real    0m10.125s
user    0m9.935s
sys     0m0.034s
[Gaz:Shootout-C++] time ./Output/fibo.native 
701408733

real    0m10.752s
user    0m10.339s
sys     0m0.041s


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 07:26:49 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 07:26:49 -0600
Subject: [LLVMbugs] [Bug 2062] New: build: cannot build llvm from source on
 read-only file system
Message-ID: <bug-2062-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2062

           Summary: build: cannot build llvm from source on read-only file
                    system
           Product: Build scripts
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Makefiles
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: markus at oberhumer.com
                CC: llvmbugs at cs.uiuc.edu


mfx at minerva:~/tmp > cd llvm-svn-ro
mfx at minerva:~/tmp/llvm-svn-ro > svn info | grep '^Rev'
Revision: 47328
mfx at minerva:~/tmp/llvm-svn-ro > cd ..
mfx at minerva:~/tmp > cd llvm-build
mfx at minerva:~/tmp/llvm-build > ../llvm-svn-ro/configure 
  [...]
mfx at minerva:~/tmp/llvm-build > make 
  [...]
llvm[2]: ======= Finished Linking Debug Executable llvm-PerfectShuffle
make[2]: Leaving directory `/home/mfx/tmp/llvm-build/utils/PerfectShuffle'
make[1]: Leaving directory `/home/mfx/tmp/llvm-build/utils'
make[1]: Entering directory `/home/mfx/tmp/llvm-build/lib/VMCore'
llvm[1]: Building Intrinsics.gen.tmp from Intrinsics.td
/bin/cp: cannot create regular file
`/mnt/net/smb-work-ro/cvs/llvm/llvm/llvm/include/llvm/Intrinsics.gen':
Read-only file system
make[1]: ***
[/mnt/net/smb-work-ro/cvs/llvm/llvm/llvm/include/llvm/Intrinsics.gen] Error 1
make[1]: Leaving directory `/home/mfx/tmp/llvm-build/lib/VMCore'
make: *** [all] Error 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 Feb 19 09:26:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 09:26:16 -0600
Subject: [LLVMbugs] [Bug 2063] New: Assertion
 `isa<llvm::ArrayType>(ElementType) && " Invalid initilizer"' failed
Message-ID: <bug-2063-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2063

           Summary: Assertion `isa<llvm::ArrayType>(ElementType) && "Invalid
                    initilizer"' failed
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Keywords: compile-fail
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1428)
 --> (http://llvm.org/bugs/attachment.cgi?id=1428)
test case (delta reduced)

clang: CGExprAgg.cpp:248:
llvm::Constant*<unnamed>::AggExprEmitter::GetConstantInit(clang::InitListExpr*,
const llvm::ArrayType*): Assertion 
`isa<llvm::ArrayType>(ElementType) && "Invalid initilizer"' 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  Tue Feb 19 09:30:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 09:30:38 -0600
Subject: [LLVMbugs] [Bug 2064] New: redifinition of typedef'ed types
Message-ID: <bug-2064-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2064

           Summary: redifinition of typedef'ed types
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


when compiling PHP's tidy extension I get the following:

In file included from /usr/local/include/tidy/tidy.h:70:
/usr/local/include/tidy/platform.h:509:22: error: redefinition of 'uint'
typedef unsigned int uint;
                     ^
In file included from /cvs/php5/ext/tidy/tidy.c:25:
In file included from /cvs/php5/main/php.h:34:
In file included from /cvs/php5/Zend/zend.h:53:
In file included from /cvs/php5/Zend/zend_config.h:1:
In file included from /cvs/php5/include/../main/php_config.h:2335:
In file included from /usr/include/stdlib.h:438:
/usr/include/sys/types.h:153:22: error: previous definition is here
typedef unsigned int uint;


GCC doesn't output any warning or error. Basically clang isn't considering
/usr/local/include as system headers.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 10:55:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 10:55:53 -0600
Subject: [LLVMbugs] [Bug 2065] New: cbe missing a address of operation
Message-ID: <bug-2065-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2065

           Summary: cbe missing a address of operation
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: C
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: alenhar2 at uiuc.edu
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1429)
 --> (http://llvm.org/bugs/attachment.cgi?id=1429)
extracted function

the attached testcase fails when the cbe tries:
struct* x = struct y;

where y is a function argument (the struct contains a single int).

when dealing with a phi node.

the failure is seen when compiling the resultant .c 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  Tue Feb 19 12:19:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 12:19:27 -0600
Subject: [LLVMbugs] [Bug 2066] New: Suboptimal code for factorial function
Message-ID: <bug-2066-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2066

           Summary: Suboptimal code for factorial function
           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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase (from http://llvm.org/demo/index.cgi):

int power(int X) {
  if (X == 0) return 1;
  return X*power(X-1);
}

currently compiles to (with clang -emit-llvm-bc | opt -std-compile-opts):
define i32 @power(i32 %X) nounwind  {
entry:
        %cmp3 = icmp eq i32 %X, 0
        br i1 %cmp3, label %ifthen, label %ifend

ifthen:         ; preds = %ifend, %entry
        %accumulator.tr.lcssa = phi i32 [ 1, %entry ], [ %mul, %ifend ]
        ret i32 %accumulator.tr.lcssa

ifend:          ; preds = %ifend, %entry
        %indvar = phi i32 [ %indvar.next, %ifend ], [ 0, %entry ]
        %accumulator.tr1 = phi i32 [ %mul, %ifend ], [ 1, %entry ]
        %X.tr2 = sub i32 %X, %indvar
        %mul = mul i32 %X.tr2, %accumulator.tr1
        %indvar.next = add i32 %indvar, 1
        %exitcond = icmp eq i32 %indvar.next, %X
        br i1 %exitcond, label %ifthen, label %ifend
}

Having two induction variables makes this code longer than it needs to be. 
Only one induction variable should be needed.

x86 asm for main loop:
.LBB1_4:        # ifend
        imull   %esi, %eax
        decl    %esi
        incl    %edx
        cmpl    %ecx, %edx
        jne     .LBB1_4 # ifend

compared to gcc:
.L5:
        imull   %edx, %eax
        subl    $1, %edx
        jne     .L5


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 12:55:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 12:55:28 -0600
Subject: [LLVMbugs] [Bug 2067] New: Loop unroller heuristics are
 over-aggressive with calls
Message-ID: <bug-2067-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2067

           Summary: Loop unroller heuristics are over-aggressive with calls
           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


Testcase:
void a(int a,int b,int c,int d,int e,int f,int g);
void b(void) {for (int i = 0; i < 20; i++)  a(1,2,3,4,5,6,7);}

This snippet of C code produces about 160 lines of x86 code using clang
-emit-llvm-bc | opt -std-compile-opts | llc.  I'm pretty sure this is due to
the loop unroller inaccurately predicting the cost of a call.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 13:01:50 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 13:01:50 -0600
Subject: [LLVMbugs] [Bug 2068] New: Passing thousands of individual params
 instead of by value
Message-ID: <bug-2068-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2068

           Summary: Passing thousands of individual params instead of by
                    value
           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: isanbard at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1430)
 --> (http://llvm.org/bugs/attachment.cgi?id=1430)
What my machine actually generates

2008-01-25-ByValReadNone.c is failing on my machine (PPC Mac G4 running Tiger):

FAIL: /Users/wendling/llvm/llvm.src/test/CFrontend/2008-01-25-
ByValReadNone.c
Failed with exit(1) at line 1
while running: /Users/wendling/llvm/llvm-gcc-4.2.install/bin/llvm-gcc
-emit-llvm -O3 -S -o - -emit-llvm /Users/wendling/llvm/llvm.src/test/
CFrontend/2008-01-25-ByValReadNone.c | not grep readonly
        %tmp1505 = call i32 @g( i64 %x.0.0, i64 %x.0.1, i64 %x.0.2,
i64 %x.0.3, i64 %x.0.4, i64 %x.0.5, i64 %x.0.6, i64 %x.0.7, i64 %x.
0.8, i64 %x.0.9, i64 %x.0.10, i64 %x.0.11, i64 %x.0.12, i64 %x.0.13,
i64 %x.0.14, i64 %x.0.15, i64 %x.0.16, i64 %x.0.17, i64 %x.0.18, i64 %
x.0.19, i64 %x.0.20, i64 %x.0.21, i64 %x.0.22, i64 %x.0.23, i64 %x.
0.24, i64 %x.0.25, i64 %x.0.26, i64 %x.0.27, i64 %x.0.28, i64 %x.
0.29, i64 %x.0.30, i64 %x.0.31, i64 %x.0.32, i64 %x.0.33, i64 %x.
0.34, i64 %x.0.35, i64 %x.0.36, i64 %x.0.37, i64 %x.0.38, i64 %x.
...
0.487, i64 %x.0.488, i64 %x.0.489, i64 %x.0.490, i64 %x.0.491, i64 %x.
0.492, i64 %x.0.493, i64 %x.0.494, i64 %x.0.495, i64 %x.0.496, i64 %x.
0.497, i64 %x.0.498, i64 %x.0.499 ) nounwind readonly        ; <i32>
[#uses=1]
...

It appears to be passing the humongous array into the functions by individual
values instead of by value.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 13:36:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 13:36:00 -0600
Subject: [LLVMbugs] [Bug 1909] Serious tail merging pessimization
In-Reply-To: <bug-1909-206@http.llvm.org/bugs/>
Message-ID: <200802191936.m1JJa07M007290@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1909


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

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




--- Comment #7 from Evan Cheng <evan.cheng at apple.com>  2008-02-19 13:36:00 ---
Ok. I'll close this bug then.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 13:46:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 13:46:55 -0600
Subject: [LLVMbugs] [Bug 2069] New: opt --std-compile-opts misses
 optimizations on first pass
Message-ID: <bug-2069-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2069

           Summary: opt --std-compile-opts misses optimizations on first
                    pass
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: opt
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1431)
 --> (http://llvm.org/bugs/attachment.cgi?id=1431)
Poorly optimized test case (hacked up pcregrep).

(I'm not sure whether I should consider this a bug, feedback appreciated)

On the given test file opt is significantly more effective (in terms of
emmitted bytecode size) when run twice.

--
ddunbar at ozzy:lli$ llvm-as -o=- pcregrep.ll | opt --std-compile-opts | llvm-dis
| wc -l
1011
ddunbar at ozzy:lli$ llvm-as -o=- pcregrep.ll | opt --std-compile-opts | opt
--std-compile-opts | llvm-dis | wc -l
77
--

Sorry for the complexity of the test case, it is hand trimmed from an llvm
built pcregrep and there is a lot of dead code in the initial test. I don't
have an easy way to trim further while still reproducing the 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 Feb 19 13:48:18 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 13:48:18 -0600
Subject: [LLVMbugs] [Bug 2065] cbe missing a address of operation
In-Reply-To: <bug-2065-206@http.llvm.org/bugs/>
Message-ID: <200802191948.m1JJmI7q007701@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2065


Andrew Lenharth <alenhar2 at uiuc.edu> changed:

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




--- Comment #6 from Andrew Lenharth <alenhar2 at uiuc.edu>  2008-02-19 13:48:17 ---
fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058528.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 Feb 19 14:18:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 14:18:19 -0600
Subject: [LLVMbugs] [Bug 2070] New: Indvars makes loop more complicated
Message-ID: <bug-2070-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2070

           Summary: Indvars makes loop more complicated
           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


Testcase:

int a(int x) {int z = 1; for (int i = 1; i < x; i++) z *= i; return z;}

leads to a really crappy loop.  The bug appears to be in indvars.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 20:50:18 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 20:50:18 -0600
Subject: [LLVMbugs] [Bug 2063] Assertion `isa<llvm::ArrayType>(ElementType)
 && " Invalid initilizer"' failed
In-Reply-To: <bug-2063-206@http.llvm.org/bugs/>
Message-ID: <200802200250.m1K2oIL2004193@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2063


Lauro Venancio <lauro.venancio at gmail.com> changed:

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




--- Comment #2 from Lauro Venancio <lauro.venancio at gmail.com>  2008-02-19 20:50:17 ---
Fixed.
http://llvm.org/viewvc/llvm-project?rev=47336&view=rev


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 19 20:54:10 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 20:54:10 -0600
Subject: [LLVMbugs] [Bug 1991] Assertion `E &&
 !hasAggregateLLVMType(E->getType()) && " Invalid scalar expression to
 emit"' failed
In-Reply-To: <bug-1991-206@http.llvm.org/bugs/>
Message-ID: <200802200254.m1K2sAlF004388@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1991


Lauro Venancio <lauro.venancio at gmail.com> changed:

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




--- Comment #2 from Lauro Venancio <lauro.venancio at gmail.com>  2008-02-19 20:54:09 ---
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 Feb 19 20:55:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 19 Feb 2008 20:55:38 -0600
Subject: [LLVMbugs] [Bug 1997] Cannot codegen simple initializer
In-Reply-To: <bug-1997-206@http.llvm.org/bugs/>
Message-ID: <200802200255.m1K2tcEF004473@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1997


Lauro Venancio <lauro.venancio at gmail.com> changed:

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




--- Comment #1 from Lauro Venancio <lauro.venancio at gmail.com>  2008-02-19 20:55:37 ---
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  Wed Feb 20 00:51:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 00:51:03 -0600
Subject: [LLVMbugs] [Bug 2003] Incorrect iteration count for loop with
 unsigned ICMP condition
In-Reply-To: <bug-2003-206@http.llvm.org/bugs/>
Message-ID: <200802200651.m1K6p3JQ011312@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2003


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #12 from Nick Lewycky <nicholas at mxc.ca>  2008-02-20 00:51:03 ---
Patch checked in:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058549.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 Feb 20 01:52:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 01:52:43 -0600
Subject: [LLVMbugs] [Bug 2071] New: llvm-gcc uses the -strip-dead-prototypes
 pass but opt doesn 't know it
Message-ID: <bug-2071-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2071

           Summary: llvm-gcc uses the -strip-dead-prototypes pass but opt
                    doesn't know it
           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


Running llvm-gcc with -fdebug-pass-arguments gives the list of passes
run.  Pasting this as a list of passes for opt I noticed that opt doesn't
know about the -strip-dead-prototypes pass.  It would be good 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  Wed Feb 20 02:02:54 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 02:02:54 -0600
Subject: [LLVMbugs] [Bug 2072] New: use-after-free in GVN
Message-ID: <bug-2072-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2072

           Summary: use-after-free 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=1433)
 --> (http://llvm.org/bugs/attachment.cgi?id=1433)
testcase .ll

$ valgrind opt use_after_free.bc -gvn
...
 Invalid read of size 2
    at 0x82D2A58: llvm::Value::getValueID() const (Value.h:208)
    by 0x82D2ACC: bool llvm::isa_impl<llvm::Instruction,
llvm::Value>(llvm::Value const&) (Value.h:257)
    by 0x82D3230: llvm::isa_impl_wrap<llvm::Instruction, llvm::Value const,
llvm::Value const>::doit(llvm::Value const&) (Casting.h:71)
    by 0x82D3246: bool
llvm::isa_impl_cl<llvm::Value>::isa<llvm::Instruction>(llvm::Value const&)
(Casting.h:83)
    by 0x82EE95E: bool llvm::isa_impl_cl<llvm::Value
const>::isa<llvm::Instruction>(llvm::Value const&) (Casting.h:92)
    by 0x82EE974: bool llvm::isa_impl_cl<llvm::Value
const*>::isa<llvm::Instruction>(llvm::Value const*) (Casting.h:101)
    by 0x82EE98C: bool llvm::isa<llvm::Instruction, llvm::Value
const*>(llvm::Value const* const&) (Casting.h:116)
    by 0x82FE180: llvm::CallInst::classof(llvm::Value const*)
(Instructions.h:980)
    by 0x82FE1C2: bool llvm::isa_impl<llvm::CallInst, llvm::Value>(llvm::Value
const&) (Casting.h:54)
    by 0x82FE1D8: llvm::isa_impl_wrap<llvm::CallInst, llvm::Value const,
llvm::Value const>::doit(llvm::Value const&) (Casting.h:71)
    by 0x82FE1EE: bool
llvm::isa_impl_cl<llvm::Value>::isa<llvm::CallInst>(llvm::Value const&)
(Casting.h:83)
    by 0x82FE232: bool llvm::isa_impl_cl<llvm::Value
const>::isa<llvm::CallInst>(llvm::Value const&) (Casting.h:92)
  Address 0x42d3f44 is 4 bytes inside a block of size 44 free'd
    at 0x402231C: operator delete(void*) (vg_replace_malloc.c:342)
    by 0x8556D6A: llvm::CallInst::~CallInst() (Instructions.cpp:239)
    by 0x83012D0: llvm::iplist<llvm::Instruction,
llvm::ilist_traits<llvm::Instruction>
>::erase(llvm::ilist_iterator<llvm::Instruction>) (ilist:368)
    by 0x8547494: llvm::Instruction::eraseFromParent() (Instruction.cpp:68)
    by 0x8364649: (anonymous
namespace)::GVN::iterateOnFunction(llvm::Function&) (GVN.cpp:1342)
    by 0x8364733: (anonymous namespace)::GVN::runOnFunction(llvm::Function&)
(GVN.cpp:1295)
    by 0x856C52B: llvm::FPPassManager::runOnFunction(llvm::Function&)
(PassManager.cpp:1184)
    by 0x856C6CD: llvm::FPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1204)
    by 0x856C201: llvm::MPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1254)
    by 0x856C3B9: llvm::PassManagerImpl::run(llvm::Module&)
(PassManager.cpp:1328)
    by 0x856C40B: llvm::PassManager::run(llvm::Module&) (PassManager.cpp:1360)
    by 0x82E017E: main (opt.cpp:426)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 02:11:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 02:11:12 -0600
Subject: [LLVMbugs] [Bug 2073] New: CBE barfs on constant struct containing
	long double
Message-ID: <bug-2073-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2073

           Summary: CBE barfs on constant struct containing long double
           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=1434)
 --> (http://llvm.org/bugs/attachment.cgi?id=1434)
testcase .ll

$ llc -march=c ld.bc
llc: /home/duncan/LLVM/llvm.top/llvm/lib/Target/CBackend/CBackend.cpp:923:
void<unnamed>::CWriter::printConstant(llvm::Constant*): Assertion
`FPC->getType() == Type::FloatTy || FPC->getType() == Type::DoubleTy' 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  Wed Feb 20 02:23:26 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 02:23:26 -0600
Subject: [LLVMbugs] [Bug 2072] use-after-free in GVN
In-Reply-To: <bug-2072-206@http.llvm.org/bugs/>
Message-ID: <200802200823.m1K8NQFK014237@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2072


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #1 from Owen Anderson <resistor at mac.com>  2008-02-20 02:23:26 ---
Fixed in r47364.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 03:59:52 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 03:59:52 -0600
Subject: [LLVMbugs] [Bug 2074] New: (X86?) Backend generates duplicate labels
Message-ID: <bug-2074-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2074

           Summary: (X86?) Backend generates duplicate labels
           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: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1435)
 --> (http://llvm.org/bugs/attachment.cgi?id=1435)
failing testcase

It appears as if the label counter is not being incremented somewhere.

--
ddunbar at ozzy:src$ llvm-as -f invalid-duplicate-labels.ll
ddunbar at ozzy:src$ llvm-ld --disable-opt -native -o rt
invalid-duplicate-labels.bc
rt.s: Assembler messages:
rt.s:46: Error: symbol `.Llabel1' is already defined
llvm-ld: 
--


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 05:13:56 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 05:13:56 -0600
Subject: [LLVMbugs] [Bug 2075] New: (X86?) Backend generates reference to
	missing label
Message-ID: <bug-2075-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2075

           Summary: (X86?) Backend generates reference to missing label
           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: daniel at zuster.org
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1436)
 --> (http://llvm.org/bugs/attachment.cgi?id=1436)
failing testcase

The backend generates an invalid assembly fail (in some way related to the
debug symbols).

--
ddunbar at ozzy:src$ llvm-as -f undefined-label.ll 
ddunbar at ozzy:src$ llvm-ld -native -o rt undefined-label.bc
/tmp/ccJhziqe.o:/home/ddunbar/private/klee-tests/MultiSource/vim/vim71/src//window.c:3090:
undefined reference to `.Llabel2'
collect2: ld returned 1 exit status
llvm-ld: 
--

This test case was aggressively trimmed from a large one (vim 7.1 compiled to
bitcode) and I'm guessing the problem has something to do with the inlining. If
I've destroyed the interesting structure of the test case by reducing it I can
provide the complete original input as 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  Wed Feb 20 08:15:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 08:15:00 -0600
Subject: [LLVMbugs] [Bug 2076] New: Crash with inline asm: assertion "
 attempt to spill already spilled interval!"
Message-ID: <bug-2076-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2076

           Summary: Crash with inline asm: assertion "attempt to spill
                    already spilled interval!"
           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=1438)
 --> (http://llvm.org/bugs/attachment.cgi?id=1438)
Testcase

llc: /home/eli/llvm/lib/CodeGen/LiveIntervalAnalysis.cpp:1188:
std::vector<llvm::LiveInterval*, std::allocator<llvm::LiveInterval*> >
llvm::LiveIntervals::addIntervalsForSpills(const llvm::LiveInterval&, const
llvm::MachineLoopInfo*, llvm::VirtRegMap&): Assertion `li.weight !=
(__builtin_huge_valf()) && "attempt to spill already spilled interval!"'
failed.

Testcase attached (reduced with bugpoint).

No clue why this is happening... there's nothing obviously suspicious about the
code (besides the fact that it uses an inline asm statement).

Test reduced from ffmpeg.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 08:49:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 08:49:22 -0600
Subject: [LLVMbugs] [Bug 2077] New: incorrect gvn memcpy simplification
Message-ID: <bug-2077-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2077

           Summary: incorrect gvn memcpy simplification
           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=1439)
 --> (http://llvm.org/bugs/attachment.cgi?id=1439)
testcase .ll

The function @badly_optimized does:

  initialize %memtmp (by calling @initialize)
  memcpy %memtmp to %tmp
  memcpy %tmp to %z
  pass %z to @passed_uninitialized

After GVN (opt -gvn) it does:

  initialize %tmp (by calling @initialize)
  memcpy %memtmp to %z   <= This is not "memcpy %tmp to %z"!
  pass %z to @passed_uninitialized


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 09:25:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 09:25:27 -0600
Subject: [LLVMbugs] [Bug 2078] New: LLVM allocates eax as an input register
 for inline asm when it is in the clobber list
Message-ID: <bug-2078-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2078

           Summary: LLVM allocates eax as an input register for inline asm
                    when it is in the clobber list
           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=1440)
 --> (http://llvm.org/bugs/attachment.cgi?id=1440)
Source for testcase

Steps to reproduce: run source through llvm-gcc -S, and note this code will
always instantly crash.

The asm expects to be able to use eax because it is in the clobber list, but
one of the inputs ends up getting allocated into eax.  I don't know the exact
rules here, but we should probably be compatible with gcc.

Testcase reduced from ffmpeg.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 14:54:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 14:54:47 -0600
Subject: [LLVMbugs] [Bug 2079] New: clang asset: initializer with more or
 less braces than default
Message-ID: <bug-2079-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2079

           Summary: clang asset: initializer with more or less braces than
                    default
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lauro.venancio at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1441)
 --> (http://llvm.org/bugs/attachment.cgi?id=1441)
test case

This bug makes 6 tests fail:
Benchmarks/MallocBench/espresso/espresso
Benchmarks/MallocBench/gs/gs
Benchmarks/Prolangs-C/cdecl/cdecl
Benchmarks/Prolangs-C/unix-smail/unix-smail
Benchmarks/VersaBench/bmm/bmm
Benchmarks/MiBench/consumer-lame/consumer-lame

I think, the Sema should emit a warning (as gcc does) and fix the AST.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 15:19:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 15:19:28 -0600
Subject: [LLVMbugs] [Bug 2080] New: clang assert: sizeof(void)
Message-ID: <bug-2080-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2080

           Summary: clang assert: sizeof(void)
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Basic
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lauro.venancio at gmail.com
                CC: llvmbugs at cs.uiuc.edu


This bug makes Benchmarks/MiBench/network-patricia test fail.

Obviously, there is a bug in network-patricia, but gcc accepts it without
warning.
Gcc returns 1 to sizeof(void). Clang should give a warning or an error.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 15:22:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 15:22:34 -0600
Subject: [LLVMbugs] [Bug 2081] New: clang assert: circular type definition
Message-ID: <bug-2081-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2081

           Summary: clang assert: circular type definition
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lauro.venancio at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1442)
 --> (http://llvm.org/bugs/attachment.cgi?id=1442)
test case

This bug makes Benchmarks/Olden/health/health test fail.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 20 22:57:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 22:57:01 -0600
Subject: [LLVMbugs] [Bug 2078] llvm allocates registers for inline asm even
 when alias is in clobber set
In-Reply-To: <bug-2078-206@http.llvm.org/bugs/>
Message-ID: <200802210457.m1L4v1vY028683@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2078


Chris Lattner <sabre at nondot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
          Component|new bugs                    |Common Code Generator Code
           Keywords|                            |miscompilation
            Product|new-bugs                    |libraries
         Resolution|                            |FIXED
            Summary|LLVM allocates eax as an    |llvm allocates registers for
                   |input register for inline   |inline asm even when alias
                   |asm when it is in the       |is in clobber set
                   |clobber list                |
   Target Milestone|---                         |2.3
            Version|unspecified                 |1.0




--- Comment #6 from Chris Lattner <sabre at nondot.org>  2008-02-20 22:57:00 ---
Fixed, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058622.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 Feb 20 23:28:06 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 23:28:06 -0600
Subject: [LLVMbugs] [Bug 1133] matching inline asm constraints with memory
 not implemented
In-Reply-To: <bug-1133-206@http.llvm.org/bugs/>
Message-ID: <200802210528.m1L5S6gM029565@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1133


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #11 from Chris Lattner <sabre at nondot.org>  2008-02-20 23:28:05 ---
Your patch looks great to me, applied:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058625.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  Wed Feb 20 23:45:54 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 20 Feb 2008 23:45:54 -0600
Subject: [LLVMbugs] [Bug 2080] clang assert: sizeof(void)
In-Reply-To: <bug-2080-206@http.llvm.org/bugs/>
Message-ID: <200802210545.m1L5js44030195@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2080


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-20 23:45:54 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080218/004379.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 Feb 21 03:57:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 21 Feb 2008 03:57:45 -0600
Subject: [LLVMbugs] [Bug 2078] llvm allocates registers for inline asm even
 when alias is in clobber set
In-Reply-To: <bug-2078-206@http.llvm.org/bugs/>
Message-ID: <200802210957.m1L9vjnU012646@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2078


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #7 from Eli Friedman <sharparrow1 at yahoo.com>  2008-02-21 03:57:44 ---
Fixing that issue just shuffles the problem... try the original testcase again.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 21 09:51:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 21 Feb 2008 09:51:47 -0600
Subject: [LLVMbugs] [Bug 2065] cbe missing a address of operation
In-Reply-To: <bug-2065-206@http.llvm.org/bugs/>
Message-ID: <200802211551.m1LFpl8F023928@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2065


Andrew Lenharth <alenhar2 at uiuc.edu> changed:

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




--- Comment #7 from Andrew Lenharth <alenhar2 at uiuc.edu>  2008-02-21 09:51:47 ---
Fix was backed out,so this still fails.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 21 13:45:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 21 Feb 2008 13:45:03 -0600
Subject: [LLVMbugs] [Bug 2078] llvm allocates registers for inline asm even
 when alias is in clobber set
In-Reply-To: <bug-2078-206@http.llvm.org/bugs/>
Message-ID: <200802211945.m1LJj3qw031325@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2078


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #8 from Chris Lattner <sabre at nondot.org>  2008-02-21 13:45:03 ---
Fixed again:  :)
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058640.html

Please verify, 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 akyrtzi at gmail.com  Fri Feb 22 00:21:20 2008
From: akyrtzi at gmail.com (Argiris Kirtzidis)
Date: Thu, 21 Feb 2008 22:21:20 -0800
Subject: [LLVMbugs] Patch for Win32/MSVC compilation
Message-ID: <47BE69E0.1060103@gmail.com>

Hi,

The attached patch fixes some issues for Visual C++ 2008 compilation.
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: win32.patch
Url: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20080221/bcbb543e/attachment.pl 

From asl at math.spbu.ru  Thu Feb 21 16:31:50 2008
From: asl at math.spbu.ru (Anton Korobeynikov)
Date: Fri, 22 Feb 2008 01:31:50 +0300
Subject: [LLVMbugs] Patch for Win32/MSVC compilation
In-Reply-To: <47BE69E0.1060103.SS2827SS@gmail.com>
References: <47BE69E0.1060103.SS2827SS@gmail.com>
Message-ID: <1203633110.24729.81.camel@asl.dorms.spbu.ru>

Hello, Argiris

> +#ifdef _MSC_VER
> +// Get std::equal_to
> +#include <functional>
> +#endif
Please, no compiler-specific defines outside libSystem / libSupport.
Most probably, we should pull <functional> unconditionally here, since
std::equal_to should be defined there according to standard.

> +  static BOOL CALLBACK ELM_Callback(
> +#ifdef __MINGW32__
> +									PSTR  ModuleName,
> +#else
> +									PCSTR  ModuleName,
> +#endif
>                                      ModuleBaseType ModuleBase,
>                                      ULONG ModuleSize,
>                                      PVOID UserContext)
Use spaces, not tabs. How is PCSTR defined? What's about VC2005 & 7.x
builds?

-- 
With best regards, Anton Korobeynikov.

Faculty of Mathematics & Mechanics, Saint Petersburg State University.



From bugzilla-daemon at cs.uiuc.edu  Thu Feb 21 16:49:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 21 Feb 2008 16:49:27 -0600
Subject: [LLVMbugs] [Bug 1874] Multiplication codegen issues
In-Reply-To: <bug-1874-206@http.llvm.org/bugs/>
Message-ID: <200802212249.m1LMnRZh004311@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1874


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #10 from Dan Gohman <gohman at apple.com>  2008-02-21 16:49:26 ---
This is now fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058473.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 akyrtzi at gmail.com  Fri Feb 22 03:33:11 2008
From: akyrtzi at gmail.com (Argiris Kirtzidis)
Date: Fri, 22 Feb 2008 01:33:11 -0800
Subject: [LLVMbugs] Patch for Win32/MSVC compilation
In-Reply-To: <1203633110.24729.81.camel@asl.dorms.spbu.ru>
References: <47BE69E0.1060103.SS2827SS@gmail.com>
	<1203633110.24729.81.camel@asl.dorms.spbu.ru>
Message-ID: <47BE96D7.1090703@gmail.com>

Hi,

'PCSTR' is defined as "CONST CHAR *" and is needed only when compiling 
with VS 2008.
I've corrected the patch (new one attached).


Anton Korobeynikov wrote:
> Hello, Argiris
>
>   
>> +#ifdef _MSC_VER
>> +// Get std::equal_to
>> +#include <functional>
>> +#endif
>>     
> Please, no compiler-specific defines outside libSystem / libSupport.
> Most probably, we should pull <functional> unconditionally here, since
> std::equal_to should be defined there according to standard.
>
>   
>> +  static BOOL CALLBACK ELM_Callback(
>> +#ifdef __MINGW32__
>> +									PSTR  ModuleName,
>> +#else
>> +									PCSTR  ModuleName,
>> +#endif
>>                                      ModuleBaseType ModuleBase,
>>                                      ULONG ModuleSize,
>>                                      PVOID UserContext)
>>     
> Use spaces, not tabs. How is PCSTR defined? What's about VC2005 & 7.x
> builds?
>
>   
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: win32-2.patch
Url: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20080222/b4517338/attachment.pl 

From bugzilla-daemon at cs.uiuc.edu  Thu Feb 21 17:59:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 21 Feb 2008 17:59:45 -0600
Subject: [LLVMbugs] [Bug 2082] New: Register allocation allocating stuff
 into the wrong register with inline asm
Message-ID: <bug-2082-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2082

           Summary: Register allocation allocating stuff into the wrong
                    register with inline asm
           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=1445)
 --> (http://llvm.org/bugs/attachment.cgi?id=1445)
Testcase

Testcase attached; generates wrong code with "llvm-gcc -S -o -".

On my computer, this ends up attempting to load from the address src_stride*3,
which is clearly not valid.

I'm pretty sure this is something funny going on with register allocation.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 22 08:51:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 22 Feb 2008 08:51:34 -0600
Subject: [LLVMbugs] [Bug 2083] New: Optimisation breaks debug records:
 region.start/region.end
Message-ID: <bug-2083-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2083

           Summary: Optimisation breaks debug records:
                    region.start/region.end
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Archive library
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: richard.smith at antixlabs.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1447)
 --> (http://llvm.org/bugs/attachment.cgi?id=1447)
Demo sourtce file

The test case is a C program which makes a series of calls to function check().
When compiled with no optimisation and disassembled thusly:

% llvm-gcc 20000223-1.c -g -c --emit-llvm
% llvm-dis -f 20000223-1.o

you will see that in the code for function check() there are debug intrinsic
functions for a region start and end, and the region.end is correctly matched
to the block (first and last lines of the code fragment):

20000223-1.o.ll (fragment):

        call void @llvm.dbg.region.start( {  }* bitcast (%llvm.dbg.block.type*
@llvm.dbg.block to {  }*) )
        call void @llvm.dbg.stoppoint( i32 12, i32 0, {  }* bitcast
(%llvm.dbg.compile_unit.type* @llvm.dbg.compile_unit to {  }*) )
        call void @abort( )
        unreachable
                ; No predecessors!
        call void @llvm.dbg.stoppoint( i32 12, i32 0, {  }* bitcast
(%llvm.dbg.compile_unit.type* @llvm.dbg.compile_unit to {  }*) )
        call void @llvm.dbg.region.end( {  }* bitcast (%llvm.dbg.block.type*
@llvm.dbg.block to {  }*) )

If you now optimise by running opt and disassemble:

opt -f -std-compile-opts 20000223-1.o -o 20000223-1.o
llvm-dis -f 20000223-1.o

you will see that the region.end has been amended and is now incorrectly paired
with a subprogram. Furthermore, this function check() is inlined into the main
function and the problem is replicated in each case.

20000223-1.o.ll (fragment):

cond_true:      ; preds = %entry
    call void @llvm.dbg.region.start( {  }* bitcast (%llvm.dbg.block.type*
@llvm.dbg.block to {  }*) )
    call void @llvm.dbg.stoppoint( i32 12, i32 0, {  }* bitcast
(%llvm.dbg.compile_unit.type* @llvm.dbg.compile_unit to {  }*) )
    call void @abort( )
    unreachable

return:     ; preds = %entry
    call void @llvm.dbg.stoppoint( i32 14, i32 0, {  }* bitcast
(%llvm.dbg.compile_unit.type* @llvm.dbg.compile_unit to {  }*) )
    call void @llvm.dbg.region.end( {  }* bitcast (%llvm.dbg.subprogram.type*
@llvm.dbg.subprogram to {  }*) )


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 22 12:40:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 22 Feb 2008 12:40:27 -0600
Subject: [LLVMbugs] [Bug 2071] llvm-gcc uses the -strip-dead-prototypes pass
 but opt doesn' t know it
In-Reply-To: <bug-2071-206@http.llvm.org/bugs/>
Message-ID: <200802221840.m1MIeRf9015201@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2071


Devang Patel <dpatel 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  Fri Feb 22 14:41:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 22 Feb 2008 14:41:08 -0600
Subject: [LLVMbugs] [Bug 2082] local RA allocating two values to
 sameregister with inline asm
In-Reply-To: <bug-2082-206@http.llvm.org/bugs/>
Message-ID: <200802222041.m1MKf8Tf019018@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2082


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

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




--- Comment #8 from Evan Cheng <evan.cheng at apple.com>  2008-02-22 14:41:08 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080218/058689.html

I am glad you found this bug. Local register allocator has not been using ESI,
EDI, and EBP because it thinks SIL, DIL, and BPL are always occupied! This
causes it to completely ran out of registers.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 22 15:02:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 22 Feb 2008 15:02:19 -0600
Subject: [LLVMbugs] [Bug 2084] New: Regalloc should deal with running out of
 registers more gracefully
Message-ID: <bug-2084-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2084

           Summary: Regalloc should deal with running out of registers more
                    gracefully
           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


Testcase (note that it is impossible to allocate 8 general regs to an asm on
x86):

int a() {int a,b,c,d,di,si,bp,sp;; asm volatile ("%0 %1 %2 %3 %4 %5 %6 %7" :
"=r"(a), "=r"(b), "=r"(c), "=r"(d), "=r"(di), "=r"(si), "=r"(bp), "=r"(sp) : );
return a;}

gcc gives the following error:
<stdin>: In function ???a???:
<stdin>:1: error: can't find a register in class ???GENERAL_REGS??? while
reloading ???asm???

llc -regalloc=linearscan hangs.

llc -regalloc=local silently puts multiple outputs into the same register.

llc -regalloc=simple gives the following assertion:
llc: /home/eli/llvm/lib/CodeGen/RegAllocSimple.cpp:128: unsigned
int<unnamed>::RegAllocSimple::getFreeReg(unsigned int): Assertion `RI+regIdx !=
RE && "Not enough registers!"' failed.

llc -regalloc=bigblock silently puts multiple outputs into the same register.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 22 15:36:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 22 Feb 2008 15:36:01 -0600
Subject: [LLVMbugs] [Bug 2085] New: standardize locales for build
Message-ID: <bug-2085-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2085

           Summary: standardize locales for build
           Product: Build scripts
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Makefiles
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nlewycky at google.com
                CC: llvmbugs at cs.uiuc.edu


llvm-gcc won't build given the locale settings:

  LC_ALL=et_EE.UTF-8
  LANG=et_EE.UTF-8
  unset LOCALE

because the sed pattern [a-z] doesn't match 'llvm'. In Estonian, 'z' is placed
right after 's'.

Putting 'export LANG=C' in the configure script fixes the 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  Fri Feb 22 15:49:51 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 22 Feb 2008 15:49:51 -0600
Subject: [LLVMbugs] [Bug 2086] New: undef doesn't work when a file is
	included twice
Message-ID: <bug-2086-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2086

           Summary: undef doesn't work when a file is included twice
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: preprocessor
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lauro.venancio at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1449)
 --> (http://llvm.org/bugs/attachment.cgi?id=1449)
teste6.c

This bug affects, for example, the following code on linux

#define __need_timeval
#include <bits/time.h>
#define __need_timeval
#include <bits/time.h>


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 22 18:42:02 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 22 Feb 2008 18:42:02 -0600
Subject: [LLVMbugs] [Bug 2087] New: gvn aborts in testcase with sret
Message-ID: <bug-2087-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2087

           Summary: gvn aborts in testcase with sret
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Keywords: compile-fail
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: gohman at apple.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1451)
 --> (http://llvm.org/bugs/attachment.cgi?id=1451)
bugpoint-reduced testcase from 252.eon

Running opt -gvn on the attached testcase aborts:

Instruction does not dominate all uses!
        %tmp1.i = getelementptr %struct.ggFrame3* %agg.result, i32 0, i32 1    
        ; <%struct.ggONB3*> [#uses=2]
        call void @_ZN12ggQuaternion7getONB3Ev( %struct.ggONB3* noalias sret 
%tmp1.i, %struct.ggQuaternion* %qresult ) nounwind 
Broken module found, compilation aborted!

This is a bugpoint-reduced testcase from 252.eon.


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

From akyrtzi at gmail.com  Sat Feb 23 04:44:49 2008
From: akyrtzi at gmail.com (Argiris Kirtzidis)
Date: Sat, 23 Feb 2008 02:44:49 -0800
Subject: [LLVMbugs] Fix for heap corruption bug
Message-ID: <47BFF921.30900@gmail.com>

Hi,

The attached patch fixes a very small bug:

---------------------------------------
  if (!cstr)
    cstr = new char[len + (isNullTerm ? 1 : 0)];
 
  assert (cstr != NULL);
 
  for (unsigned i = 0; i < len; ++i)
    cstr[i] = (char) ReadInt();
 
  if (isNullTerm)
    cstr[len+1] = '\0';
---------------------------------------

"cstr[len+1] = '\0';" writes one byte after the reserved block of 
memory; msvcrt was kind enough to report it with a "heap corruption" 
error ;)


-------------- next part --------------
A non-text attachment was scrubbed...
Name: heap-fix.zip
Type: application/octet-stream
Size: 318 bytes
Desc: not available
Url : http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20080223/f8109674/attachment.obj 
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: heap-fix.patch
Url: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20080223/f8109674/attachment.pl 

From kremenek at apple.com  Fri Feb 22 19:10:59 2008
From: kremenek at apple.com (Ted Kremenek)
Date: Fri, 22 Feb 2008 17:10:59 -0800
Subject: [LLVMbugs] Fix for heap corruption bug
In-Reply-To: <47BFF921.30900@gmail.com>
References: <47BFF921.30900@gmail.com>
Message-ID: <5C56DD2E-6274-42D9-BA51-B6445C9A63E0@apple.com>

Fixed.  Thanks!

On Feb 23, 2008, at 2:44 AM, Argiris Kirtzidis wrote:

> Hi,
>
> The attached patch fixes a very small bug:
>
> ---------------------------------------
> if (!cstr)
>   cstr = new char[len + (isNullTerm ? 1 : 0)];
> assert (cstr != NULL);
> for (unsigned i = 0; i < len; ++i)
>   cstr[i] = (char) ReadInt();
> if (isNullTerm)
>   cstr[len+1] = '\0';
> ---------------------------------------
>
> "cstr[len+1] = '\0';" writes one byte after the reserved block of  
> memory; msvcrt was kind enough to report it with a "heap corruption"  
> error ;)
>
>
> <heap-fix.zip>Index: lib/Bitcode/Reader/Deserialize.cpp
> ===================================================================
> --- lib/Bitcode/Reader/Deserialize.cpp	(revision 47480)
> +++ lib/Bitcode/Reader/Deserialize.cpp	(working copy)
> @@ -324,7 +324,7 @@
>     cstr[i] = (char) ReadInt();
>
>   if (isNullTerm)
> -    cstr[len+1] = '\0';
> +    cstr[len] = '\0';
>
>   return cstr;
> }
> _______________________________________________
> LLVMbugs mailing list
> LLVMbugs at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs


From bugzilla-daemon at cs.uiuc.edu  Sun Feb 24 08:53:15 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 24 Feb 2008 08:53:15 -0600
Subject: [LLVMbugs] [Bug 2088] New: Miscompilation with indvars
Message-ID: <bug-2088-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2088

           Summary: Miscompilation with indvars
           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=1455)
 --> (http://llvm.org/bugs/attachment.cgi?id=1455)
Testcase

To reproduce: compare the output of "lli abortbug.bc" and "opt -indvars
abortbug.bc -o - | lli" with the attached testcase.

I think it's somehow illegally rewriting the induction variable z.

Original source (from X86/README.txt):
main ()
{
  int i = 0;
  unsigned long int z = 0;

  do {
    z -= 0x00004000;
    i++;
    if (i > 0x00040000)
      abort ();
  } while (z > 0);
  exit (0);
}
compiled with 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  Sun Feb 24 10:08:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 24 Feb 2008 10:08:28 -0600
Subject: [LLVMbugs] [Bug 2089] New: sizeof('a') in C++ is == 1 and in C == 4
Message-ID: <bug-2089-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2089

           Summary: sizeof('a') in C++ is == 1 and in C == 4
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: miscompilation
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


printf("sizeof('a') == %d\n", sizeof('a'));

should ouput 1 in C++, but 4 in C. Currently it outputs 4 for both languages.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 24 10:14:13 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 24 Feb 2008 10:14:13 -0600
Subject: [LLVMbugs] [Bug 2090] New: Digraph punctuation tokens not recognized
Message-ID: <bug-2090-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2090

           Summary: Digraph punctuation tokens not recognized
           Product: clang
           Version: unspecified
          Platform: PC
               URL: http://david.tribble.com/text/cdiffs.htm
        OS/Version: All
            Status: NEW
          Keywords: missing-feature
          Severity: enhancement
          Priority: P2
         Component: parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


C++ and C99 accept the following weird syntax (which clang doesn't support):

%:include <stdio.h>

    %:ifndef BUFSIZE
     %:define BUFSIZE  512
    %:endif

    void copy(char d<::>, const char s<::>, int len)
    <%
        while (len-- >= 0)
        <%
            d<:len:> = s<:len:>;
        %>
    %>

C99: ??6.4.6
C++98: ??2.5, 2.12


-- 
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  Sun Feb 24 10:39:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 24 Feb 2008 10:39:09 -0600
Subject: [LLVMbugs] [Bug 2091] New: referenced structure should? be
	implicitly defined
Message-ID: <bug-2091-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2091

           Summary: referenced structure should? be implicitly defined
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Keywords: compile-fail
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1456)
 --> (http://llvm.org/bugs/attachment.cgi?id=1456)
test case (delta reduced)

The attached test case fails with clang, but both gcc and g++ accepts it.
It causes failure when parsing the stdio.h header in C++ mode.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 24 13:06:51 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 24 Feb 2008 13:06:51 -0600
Subject: [LLVMbugs] [Bug 2090] <% Digraph not recognized
In-Reply-To: <bug-2090-206@http.llvm.org/bugs/>
Message-ID: <200802241906.m1OJ6pOD006538@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2090


Chris Lattner <sabre at nondot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|enhancement                 |normal
             Status|NEW                         |RESOLVED
          Component|parser                      |preprocessor
         Resolution|                            |FIXED
            Summary|Digraph punctuation tokens  |<% Digraph not recognized
                   |not recognized              |




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-24 13:06:51 ---
Oops, a typo prevented recognition of <%.  Nice catch, patch here:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080218/004417.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 Feb 24 18:41:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 24 Feb 2008 18:41:01 -0600
Subject: [LLVMbugs] [Bug 2087] gvn aborts in testcase with sret
In-Reply-To: <bug-2087-206@http.llvm.org/bugs/>
Message-ID: <200802250041.m1P0f1QI015967@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2087


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #2 from Owen Anderson <resistor at mac.com>  2008-02-24 18:41:00 ---
Fixed in r47541.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 24 22:08:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 24 Feb 2008 22:08:29 -0600
Subject: [LLVMbugs] [Bug 2077] incorrect gvn memcpy simplification
In-Reply-To: <bug-2077-206@http.llvm.org/bugs/>
Message-ID: <200802250408.m1P48TT2021300@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2077


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #1 from Owen Anderson <resistor at mac.com>  2008-02-24 22:08:28 ---
Fixed in r47544.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 25 01:47:46 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 25 Feb 2008 01:47:46 -0600
Subject: [LLVMbugs] [Bug 1757] Missed optimization: any number greater than
 two is not equal to two
In-Reply-To: <bug-1757-206@http.llvm.org/bugs/>
Message-ID: <200802250747.m1P7lklH027104@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1757


Nick Lewycky <nicholas at mxc.ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|LATER                       |FIXED




--- Comment #3 from Nick Lewycky <nicholas at mxc.ca>  2008-02-25 01:47:45 ---
opt -gvn -simplifycfg can handle this case.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 25 11:46:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 25 Feb 2008 11:46:05 -0600
Subject: [LLVMbugs] [Bug 1757] Missed optimization: any number greater than
 two is not equal to two
In-Reply-To: <bug-1757-206@http.llvm.org/bugs/>
Message-ID: <200802251746.m1PHk5pY021071@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1757


Nick Lewycky <nicholas at mxc.ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|FIXED                       |LATER




--- Comment #5 from Nick Lewycky <nicholas at mxc.ca>  2008-02-25 11:46:05 ---
Oh I'm sorry. I accidentally put the wrong code into opt.

You're right, LLVM still can't handle this case out of the box.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 25 12:01:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 25 Feb 2008 12:01:38 -0600
Subject: [LLVMbugs] [Bug 1757] Missed optimization: any number greater than
 two is not equal to two
In-Reply-To: <bug-1757-206@http.llvm.org/bugs/>
Message-ID: <200802251801.m1PI1cij022188@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1757


Chris Lattner <sabre at nondot.org> changed:

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




-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 25 13:16:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 25 Feb 2008 13:16:47 -0600
Subject: [LLVMbugs] [Bug 2086] undef doesn't work when a file is included
	twice
In-Reply-To: <bug-2086-206@http.llvm.org/bugs/>
Message-ID: <200802251916.m1PJGlgp026122@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2086


Lauro Venancio <lauro.venancio at gmail.com> changed:

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




--- Comment #3 from Lauro Venancio <lauro.venancio at gmail.com>  2008-02-25 13:16:47 ---
Fixed.
http://llvm.org/viewvc/llvm-project?rev=47551&view=rev


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 25 15:47:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 25 Feb 2008 15:47:01 -0600
Subject: [LLVMbugs] [Bug 2092] New: DeclRefExpr and EnumConstantDecl have
 different signedness
Message-ID: <bug-2092-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2092

           Summary: DeclRefExpr and EnumConstantDecl have different
                    signedness
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: kremenek at cs.stanford.edu
                CC: llvmbugs at cs.uiuc.edu, snaroff at apple.com


Consider the following code:

  enum prefixes { PPS_LREP, PPS_SEG, PPS_OSIZE, PPS_ASIZE, MAXPREFIX };

  int assemble () {
    int j = 0;
    return j < MAXPREFIX;
  }


In the return statement, the AST node for MAXPREFIX is a DeclRefExpr that
references an EnumConstantDecl for MAXPREFIX.  This DeclRefExpr has type 'int',
which represents a signed integer.  The EnumConstantDecl for MAXPREFIX
constants an APSInt object that represents an integer constant; this APSInt
object, however, is marked "unsigned"  (i.e., the method APSINT::isUnsigned()
returns true).  Here is the AST dump:

  int assemble()
  (CompoundStmt 0x9065e0 <assemble.i:7:17, line:10:1>
    (DeclStmt 0x906350 <:0:0>
      0x906520 "int j =
        (IntegerLiteral 0x906550 <assemble.i:8:11> 'int' 0)"
    (ReturnStmt 0x9065d0 <line:9:3, col:14>
      (BinaryOperator 0x9065b0 <col:10, col:14> 'int' '<'
        (DeclRefExpr 0x906570 <col:10> 'int' BlockVar='j' 0x906520)
        (DeclRefExpr 0x906590 <col:14> 'int' EnumConstant='MAXPREFIX'
0x9063f0))))

The bug is that either the DeclRefExpr that wraps the EnumConstant should have
an unsigned type (and thus we should have an implicit cast somewhere) or that
the internal APSInt value of the EnumConstant should be marked signed.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 25 20:43:50 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 25 Feb 2008 20:43:50 -0600
Subject: [LLVMbugs] [Bug 2076] Crash with inline asm: assertion " attempt to
 spill already spilled interval!"
In-Reply-To: <bug-2076-206@http.llvm.org/bugs/>
Message-ID: <200802260243.m1Q2hoaM007307@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2076


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

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




--- Comment #9 from Evan Cheng <evan.cheng at apple.com>  2008-02-25 20:43:50 ---
Missed optimization. CodeGenPrepare should sink address computation into the
block whether the inline asm instruction resides. Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058784.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058785.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 Feb 25 21:49:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 25 Feb 2008 21:49:45 -0600
Subject: [LLVMbugs] [Bug 2093] New: Assertion with inline asm:
 `castIsValid(getOpcode(), S, Ty)   && "Illegal IntToPtr"' failed.
Message-ID: <bug-2093-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2093

           Summary: Assertion with inline asm: `castIsValid(getOpcode(), S,
                    Ty) && "Illegal IntToPtr"' failed.
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: evan.cheng at apple.com
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1461)
 --> (http://llvm.org/bugs/attachment.cgi?id=1461)
Testcase

Testcase attached, to reproduce, run llc over the attached, 99% sure this is a
regression from
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058785.html.
 I'm not quite sure what's going on here, but I think it's related to direct
and indirect memory operands.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 25 23:17:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 25 Feb 2008 23:17:09 -0600
Subject: [LLVMbugs] [Bug 2094] New: Inefficient code generated for inline
 asm with multiple in-out register operands
Message-ID: <bug-2094-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2094

           Summary: Inefficient code generated for inline asm with multiple
                    in-out register operands
           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


Testcase:
#include <stdint.h>

int sad16_sse2(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)
{
    int ret;
    asm volatile(
        "pxor %%xmm6, %%xmm6            \n\t"
        //ASMALIGN(4)
        "1:                             \n\t"
        "movdqu (%1), %%xmm0            \n\t"
        "movdqu (%1, %3), %%xmm1        \n\t"
        "psadbw (%2), %%xmm0            \n\t"
        "psadbw (%2, %3), %%xmm1        \n\t"
        "paddw %%xmm0, %%xmm6           \n\t"
        "paddw %%xmm1, %%xmm6           \n\t"
        "lea (%1,%3,2), %1              \n\t"
        "lea (%2,%3,2), %2              \n\t"
        "sub $2, %0                     \n\t"
        " jg 1b                         \n\t"
        : "+r" (h), "+r" (blk1), "+r" (blk2)
        : "r" ((long)stride)
    );
    asm volatile(
        "movhlps %%xmm6, %%xmm0         \n\t"
        "paddw   %%xmm0, %%xmm6         \n\t"
        "movd    %%xmm6, %0             \n\t"
        : "=r"(ret)
    );
    return ret;
}

Generated code:

        pushl   %esi
        subl    $8, %esp
        movl    20(%esp), %edx
        movl    %edx, 4(%esp)
        movl    24(%esp), %ecx
        movl    %ecx, (%esp)
        movl    32(%esp), %eax
        movl    28(%esp), %esi
        #APP
        pxor %xmm6, %xmm6            
        1:                             
        movdqu (%ecx), %xmm0            
        movdqu (%ecx, %esi), %xmm1        
        psadbw (%edx), %xmm0            
        psadbw (%edx, %esi), %xmm1        
        paddw %xmm0, %xmm6           
        paddw %xmm1, %xmm6           
        lea (%ecx,%esi,2), %ecx              
        lea (%edx,%esi,2), %edx              
        sub $2, %eax                     
         jg 1b                         

        #NO_APP
        movl    %edx, 4(%esp)
        movl    %ecx, (%esp)
        #APP
        movhlps %xmm6, %xmm0         
        paddw   %xmm0, %xmm6         
        movd    %xmm6, %eax             

        #NO_APP
        addl    $8, %esp
        popl    %esi
        ret


(We'll put aside for the moment the fact that this code is extremely dangerous
because a compiler using certain kinds of optimizations might actually end up
using the xmm regs between the two asm statements.)

The generated code ends up being rather inefficient in that it emits four
unnecessary stores to the stack, plus allocation for the necessary space. I
think it's because blk1 and blk2 have to be put into alloca's at the il level,
and codegen isn't smart enough to eliminate them.  Not sure what the right fix
is; maybe inline asm should take advantage of the multiple return value work?

(I don't know how much fixing this will help, but this function shows up at the
top of a profile in ffmpeg re-encoding from h.264 to mpeg4, so every bit likely
helps.)


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 26 00:06:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 00:06:07 -0600
Subject: [LLVMbugs] [Bug 2095] New: [meta] Multiple return value support
Message-ID: <bug-2095-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2095

           Summary: [meta] Multiple return value support
           Product: libraries
           Version: 2.2
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: new-feature
          Severity: enhancement
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: dpatel at apple.com, llvmbugs at cs.uiuc.edu


We need:

1) support in llvm IR to represent multiple return values
2) optimizations updated to not die in the face of MRVs
3) something that generates MRVs, I'd suggest enhancing arg promotion to
promote results to use them.
4) generic codegen support for MRVs including using the CallingConv.td support
5) target-specific support for them, as appropriate, such as support for the
x86-64 calling conv stuff
6) enhance inline asm to have result register values returned as an extra
result instead of using "=*r" constraints (Bug 2094).


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 26 00:09:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 00:09:00 -0600
Subject: [LLVMbugs] [Bug 2092] DeclRefExpr and EnumConstantDecl have
 different signedness
In-Reply-To: <bug-2092-206@http.llvm.org/bugs/>
Message-ID: <200802260609.m1Q690ZH012997@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2092


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-02-26 00:08:59 ---
no really, 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 Feb 26 02:28:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 02:28:30 -0600
Subject: [LLVMbugs] [Bug 2096] New: crash in llc with floating point
	arithmetic
Message-ID: <bug-2096-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2096

           Summary: crash in llc with floating point arithmetic
           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=1464)
 --> (http://llvm.org/bugs/attachment.cgi?id=1464)
Testcase

Testcase attached; crashes when processed with llc (testcase is
bugpoint-reduced).

No idea what the issue is; the only thing that strikes me in particular is the
negative zero.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 26 09:36:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 09:36:38 -0600
Subject: [LLVMbugs] [Bug 2097] New: crash in
 Transforms/Scalar/CodeGenPrepare.cpp:993 on sparc
Message-ID: <bug-2097-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2097

           Summary: crash in Transforms/Scalar/CodeGenPrepare.cpp:993 on
                    sparc
           Product: libraries
           Version: trunk
          Platform: Sun
               URL: http://llvm.org/viewvc/llvm-
                    project/llvm/trunk/lib/Transforms/Scalar/CodeGenPrepare.
                    cpp?view=markup&pathrev=47604
        OS/Version: Solaris
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: SparcV8
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ggreif at gmail.com
                CC: llvmbugs at cs.uiuc.edu


llc crashes on this test:

CodeGen/PowerPC/2006-12-07-LargeAlloca.ll

Reproducible as:

cd Debug/bin
llvm-upgrade < .../CodeGen/PowerPC/2006-12-07-LargeAlloca.ll | llvm-as > ggg.bc
llc -march=sparc -f ggg.bc


Here is the gdb session:

GNU gdb 6.6
Copyright (C) 2006 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "sparc-sun-solaris"...
(gdb) b .../llvm/lib/Transforms/Scalar/CodeGenPrepare.cpp:993
Breakpoint 1 at 0xa4285c: file CodeGenPrepare.cpp, line 993.
(gdb) run -march=sparc -f ../../Release/bin/ggg.bc
Starting program: .../llvm/Debug/bin/llc -march=sparc -f
../../Release/bin/ggg.bc
warning: Temporarily disabling breakpoints for unloaded shared library
"/usr/lib/ld.so.1"
warning: Lowest section in /lib/libpthread.so.1 is .dynamic at 00000074

Breakpoint 1, (anonymous namespace)::CodeGenPrepare::OptimizeExtUses
(this=0xd323d0, 
    I=0xd34f98) at CodeGenPrepare.cpp:993
993       if (!TLI->isTruncateFree(I->getType(), Src->getType()))
(gdb) p TLI
$1 = (const class llvm::TargetLowering *) 0x0
(gdb) p *this
$2 = {<llvm::FunctionPass> = {<llvm::Pass> = {_vptr.Pass = 0xcebd78, 
      Resolver = 0xd31a90, PassID = 13815488, 
      AnalysisImpls = {<std::_Vector_base<std::pair<const llvm::PassInfo*,
llvm::Pass*>,std::allocator<std::pair<const llvm::PassInfo*, llvm::Pass*> > >>
= {
          _M_impl = {<std::allocator<std::pair<const llvm::PassInfo*,
llvm::Pass*> >> = {<__gnu_cxx::new_allocator<std::pair<const llvm::PassInfo*,
llvm::Pass*> >> = {<No data fields>}, <No data fields>}, _M_start = 0x0,
_M_finish = 0x0, 
            _M_end_of_storage = 0x0}}, <No data fields>}}, <No data fields>}, 
  TLI = 0x0, static ID = 0 '\0'}
(gdb) p I
$3 = (class llvm::Instruction *) 0xd34f98
(gdb) p Src
$4 = (class llvm::Value *) 0xd2e738
(gdb) p *Src
$5 = {_vptr.Value = 0xd073c8, SubclassID = 58, SubclassData = 0, Ty = {Ty =
0xd30720}, 
  UseList = 0xd34fc0, Name = 0xd31dc0}
(gdb) call I->dump()
        %gep.upgrd.1 = zext i32 %iv. to i64             ; <i64> [#uses=1]

(gdb) call Src->dump()
        %iv. = phi i32 [ %iv..inc, %bb12.i ], [ 0, %bb19.bb12.i_crit_edge ]    
       ; <i32> [#uses=2]

(gdb) 
        %iv. = phi i32 [ %iv..inc, %bb12.i ], [ 0, %bb19.bb12.i_crit_edge ]    
       ; <i32> [#uses=2]



Obviously the TLI object pointer is not correctly set, or in case NULL is
admissible, it should be checked for.


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

From arvind.ayyangar at gmail.com  Tue Feb 26 05:29:39 2008
From: arvind.ayyangar at gmail.com (Arvind Ayyangar)
Date: Tue, 26 Feb 2008 16:59:39 +0530
Subject: [LLVMbugs] Rd and Rm should be different in mul (for ARM target)
Message-ID: <c34bb56f0802260329s33e0b7fey4862e9192999a23c@mail.gmail.com>

Hi all,
      I compiled libz using the llvm toolchain. Below are the steps :

llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o example.o example.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o adler32.o adler32.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o compress.o compress.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o crc32.o crc32.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o gzio.o gzio.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o uncompr.o uncompr.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o deflate.o deflate.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o trees.o trees.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o zutil.o zutil.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o inflate.o inflate.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o infback.o infback.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o inftrees.o inftrees.c
llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o inffast.o inffast.c
llvm-ld -o libz.so.1.2.3 adler32.o compress.o crc32.o gzio.o uncompr.o
deflate.o trees.o zutil.o inflate.o infback.o inftrees.o inffast.o
llc libz.so.1.2.3.bc -o llvmlibz.s -filtype=asm -march=arm


Compiling the assemble code generated using the arm-softfloat
toolchain gives the following assembler messages

gcc llvmlibz.s --shared -o libzllvm.so
llvmlibz.s: Assembler messages:
llvmlibz.s:46: Rd and Rm should be different in mul
llvmlibz.s:113: Rd and Rm should be different in mul
llvmlibz.s:117: Rd and Rm should be different in mul
llvmlibz.s:190: Rd and Rm should be different in mla
llvmlibz.s:202: Rd and Rm should be different in mul
llvmlibz.s:206: Rd and Rm should be different in mul
llvmlibz.s:261: Rd and Rm should be different in mul
llvmlibz.s:266: Rd and Rm should be different in mul
llvmlibz.s:4770: rdhi, rdlo and rm must all be different
llvmlibz.s:4834: Rd and Rm should be different in mla
llvmlibz.s:5130: Rd and Rm should be different in mul
llvmlibz.s:6103: Rd and Rm should be different in mla
llvmlibz.s:6817: Rd and Rm should be different in mla


I tried using the march=thumb, -soft-float and -mcpu options (thanks
to the LLVM IRC Channel) but they dint really help..



-- 
Arvind

From bugzilla-daemon at cs.uiuc.edu  Tue Feb 26 11:11:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 11:11:03 -0600
Subject: [LLVMbugs] [Bug 2096] crash in llc with floating point arithmetic
In-Reply-To: <bug-2096-206@http.llvm.org/bugs/>
Message-ID: <200802261711.m1QHB3uW009544@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2096


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-26 11:11:02 ---
Fixed, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058804.html

Sorry for the breakage!


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 26 12:14:49 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 12:14:49 -0600
Subject: [LLVMbugs] [Bug 2098] New: Regression:  abort on incorrect code
Message-ID: <bug-2098-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2098

           Summary: Regression:  abort on incorrect code
           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: dalej at apple.com
                CC: llvmbugs at cs.uiuc.edu


This is incorrect code but should not abort.  Broken sometime between Feb 12
and Feb 21.  
x86-32 Darwin.

void
foo (long x)
{
  if (((x >> -2) & 1) != 0)
    bar ();
}

now what? /usr/local/bin/gcc uu.c
uu.c: In function ???foo???:
uu.c:4: warning: right shift count is negative
Assertion failed: (shiftAmt <= BitWidth && "Invalid shift amount"), function
shl, file APInt.cpp, line 1234.
uu.c: At top level:
uu.c:6: internal compiler error: Abort trap


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

From lattner at apple.com  Tue Feb 26 12:15:29 2008
From: lattner at apple.com (Tanya Lattner)
Date: Tue, 26 Feb 2008 10:15:29 -0800
Subject: [LLVMbugs] Rd and Rm should be different in mul (for ARM target)
In-Reply-To: <c34bb56f0802260329s33e0b7fey4862e9192999a23c@mail.gmail.com>
References: <c34bb56f0802260329s33e0b7fey4862e9192999a23c@mail.gmail.com>
Message-ID: <E1C4772D-1C1F-464C-B783-71850FECB0AA@apple.com>

You might want to email llvm-dev as it has a wider audience.

-Tanya

On Feb 26, 2008, at 3:29 AM, Arvind Ayyangar wrote:

> Hi all,
>       I compiled libz using the llvm toolchain. Below are the steps :
>
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o example.o  
> example.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o adler32.o  
> adler32.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o compress.o  
> compress.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o crc32.o crc32.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o gzio.o gzio.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o uncompr.o  
> uncompr.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o deflate.o  
> deflate.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o trees.o trees.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o zutil.o zutil.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o inflate.o  
> inflate.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o infback.o  
> infback.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o inftrees.o  
> inftrees.c
> llvm-gcc -emit-llvm -DNO_vsnprintf -DUSE_MMAP   -c -o inffast.o  
> inffast.c
> llvm-ld -o libz.so.1.2.3 adler32.o compress.o crc32.o gzio.o uncompr.o
> deflate.o trees.o zutil.o inflate.o infback.o inftrees.o inffast.o
> llc libz.so.1.2.3.bc -o llvmlibz.s -filtype=asm -march=arm
>
>
> Compiling the assemble code generated using the arm-softfloat
> toolchain gives the following assembler messages
>
> gcc llvmlibz.s --shared -o libzllvm.so
> llvmlibz.s: Assembler messages:
> llvmlibz.s:46: Rd and Rm should be different in mul
> llvmlibz.s:113: Rd and Rm should be different in mul
> llvmlibz.s:117: Rd and Rm should be different in mul
> llvmlibz.s:190: Rd and Rm should be different in mla
> llvmlibz.s:202: Rd and Rm should be different in mul
> llvmlibz.s:206: Rd and Rm should be different in mul
> llvmlibz.s:261: Rd and Rm should be different in mul
> llvmlibz.s:266: Rd and Rm should be different in mul
> llvmlibz.s:4770: rdhi, rdlo and rm must all be different
> llvmlibz.s:4834: Rd and Rm should be different in mla
> llvmlibz.s:5130: Rd and Rm should be different in mul
> llvmlibz.s:6103: Rd and Rm should be different in mla
> llvmlibz.s:6817: Rd and Rm should be different in mla
>
>
> I tried using the march=thumb, -soft-float and -mcpu options (thanks
> to the LLVM IRC Channel) but they dint really help..
>
>
>
> -- 
> Arvind
> _______________________________________________
> LLVMbugs mailing list
> LLVMbugs at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs


From bugzilla-daemon at cs.uiuc.edu  Tue Feb 26 12:31:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 12:31:39 -0600
Subject: [LLVMbugs] [Bug 2099] New: Incorrect codegen: inliner vs byval
	parameter
Message-ID: <bug-2099-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2099

           Summary: Incorrect codegen: inliner vs byval parameter
           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: dalej at apple.com
                CC: llvmbugs at cs.uiuc.edu


The following currently produces incorrect code at -Os or -O3 (x86-32 Darwin). 
The inliner is screwing up the byval parameter somehow.  Broken between Feb 12
and Feb 21.  This may be connected with my changes to parameter passing, I'll
look.

struct s {
  char text[11];
  int flag;
} cell;

int
check (struct s p)
{
  if (p.flag != 99)
    return 1;
  return strcmp (p.text, "0123456789");
}

main ()
{
  cell.flag = 99;
  strcpy (cell.text, "0123456789");

  if (check (cell))
    abort();
  exit (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 dalej at apple.com  Tue Feb 26 12:35:13 2008
From: dalej at apple.com (Dale Johannesen)
Date: Tue, 26 Feb 2008 10:35:13 -0800
Subject: [LLVMbugs] Rd and Rm should be different in mul (for ARM target)
In-Reply-To: <E1C4772D-1C1F-464C-B783-71850FECB0AA@apple.com>
References: <c34bb56f0802260329s33e0b7fey4862e9192999a23c@mail.gmail.com>
	<E1C4772D-1C1F-464C-B783-71850FECB0AA@apple.com>
Message-ID: <D568FC8A-576E-42AB-8A6E-45C60A03E4A8@apple.com>

>> llvmlibz.s:117: Rd and Rm should be different in mul
>> llvmlibz.s:190: Rd and Rm should be different in mla

This was a hardware restriction on early ARM chips that has been  
removed in later ones.  You need to tell the assembler that your  
target chip is one of the current ones somehow (or you may need a  
later assembler).  In some assembler versions at least, this message  
is only a warning and the assembled code is correct.


From bugzilla-daemon at cs.uiuc.edu  Tue Feb 26 12:51:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 12:51:27 -0600
Subject: [LLVMbugs] [Bug 2098] Regression:  abort on incorrect code
In-Reply-To: <bug-2098-206@http.llvm.org/bugs/>
Message-ID: <200802261851.m1QIpRbv013171@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2098


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #2 from Dan Gohman <gohman at apple.com>  2008-02-26 12:51:27 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058821.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 Feb 26 13:15:06 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 13:15:06 -0600
Subject: [LLVMbugs] [Bug 2097] crash in
 Transforms/Scalar/CodeGenPrepare.cpp:993 on sparc
In-Reply-To: <bug-2097-206@http.llvm.org/bugs/>
Message-ID: <200802261915.m1QJF6G3014179@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2097


Gabor Greif <ggreif at gmail.com> changed:

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




--- Comment #3 from Gabor Greif <ggreif at gmail.com>  2008-02-26 13:15:06 ---
fixen in r47615.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 26 13:17:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 13:17:05 -0600
Subject: [LLVMbugs] [Bug 2100] New: Sema doesn't merge array definitions
Message-ID: <bug-2100-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2100

           Summary: Sema doesn't merge array definitions
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu, sharparrow1 at yahoo.com,
                    snaroff at apple.com, andersca at mac.com


Codegen crashes on this:
static int a[10];
static int a[] = { 1, 2, 3 };

because sema is saying the type of the second 'a' is int[3] instead of int[10].
 Apparently sema should merge the previous int[10] into the array type before
analyzing the initializer.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 26 13:56:50 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 13:56:50 -0600
Subject: [LLVMbugs] [Bug 1915] Speedup linscan by using register iterators
In-Reply-To: <bug-1915-206@http.llvm.org/bugs/>
Message-ID: <200802261956.m1QJuoI8015559@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1915


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

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




--- Comment #2 from Evan Cheng <evan.cheng at apple.com>  2008-02-26 13:56:49 ---
This is done: r47416. I don't see any compile time impact 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  Tue Feb 26 16:32:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 16:32:12 -0600
Subject: [LLVMbugs] [Bug 2101] New: Assertion failure in codegen
Message-ID: <bug-2101-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2101

           Summary: Assertion failure in codegen
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


The following aborts with -emit-llvm:

typedef struct { unsigned int i: 1; } c;
const c d = { 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 Feb 26 17:01:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 17:01:58 -0600
Subject: [LLVMbugs] [Bug 2102] New: Inline asm constants incorrect in CBE
	output
Message-ID: <bug-2102-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2102

           Summary: Inline asm constants incorrect in CBE output
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: C
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: criswell at uiuc.edu
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1466)
 --> (http://llvm.org/bugs/attachment.cgi?id=1466)
LLVM Assembly Code that exhibits the problem

The C Backend does not correctly generate inline asm that uses constants.

For example, it takes this:

%tmp1 = call i32 asm sideeffect "add $$8, $0\0A",
"={ax},~{dirflag},~{fpsr},~{flags}"( ) nounwind     ; <i32> [#uses=1]

And translates it into this:

__asm__ volatile ("add %%8, %0\n"
        :"=a"(llvm_cbe_tmp1)
        :);

The 8 should be "$8" and not "%%8" in the output.

An LLVM assembly case is attached.  To reproduce:

llvm-as -o test.bc test.ll
llc -march=c -f -o test.cbe.c test.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  Tue Feb 26 17:06:44 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 17:06:44 -0600
Subject: [LLVMbugs] [Bug 2099] Codegen regression: incorrect const prop of
	string(?)
In-Reply-To: <bug-2099-206@http.llvm.org/bugs/>
Message-ID: <200802262306.m1QN6iEQ022144@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2099


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #5 from Owen Anderson <resistor at mac.com>  2008-02-26 17:06:44 ---
Fixed in r47639.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 26 17:15:18 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 26 Feb 2008 17:15:18 -0600
Subject: [LLVMbugs] [Bug 2102] Inline asm constants incorrect in CBE output
In-Reply-To: <bug-2102-206@http.llvm.org/bugs/>
Message-ID: <200802262315.m1QNFILd022450@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2102


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-26 17:15:17 ---
C backend support for inline asm is not really usable at this point.

*** This bug has been marked as a duplicate of bug 802 ***


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 27 00:06:57 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 00:06:57 -0600
Subject: [LLVMbugs] [Bug 2103] New: clang doesn't reject address of bitfield
 or vector element
Message-ID: <bug-2103-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2103

           Summary: clang doesn't reject address of bitfield or vector
                    element
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu, sharparrow1 at yahoo.com,
                    snaroff at apple.com, andersca at mac.com


clang doesn't reject:

struct x { 
  int y : 4;
};

void foo(struct x *P) {
  &P->y;
}

or:

void a() {
  typedef float v4sf __attribute__ ((vector_size (16)));
  static v4sf q;
  float* r = &q[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 Feb 27 00:21:02 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 00:21:02 -0600
Subject: [LLVMbugs] [Bug 2083] Optimisation breaks debug records:
 region.start/region.end
In-Reply-To: <bug-2083-206@http.llvm.org/bugs/>
Message-ID: <200802270621.m1R6L24j002312@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2083


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #12 from Chris Lattner <sabre at nondot.org>  2008-02-27 00:21:01 ---
> There is the comment on this bug "Optimization of the code with debug
> information currently is not supported at all". llvm-gcc will disable debugging
> if  -g and -O are specified simultaneously but we have been creating
> semi-optimised and debuggable code by: (a) passing -O0 to llvm-gcc; (b) passing
> the resultant bitcode file through opt; (c) passing optimisation flags to
> llvm-ld when linking. Is this recommended practice, or would it be more
> appropriate at this stage in the llvm development to ensure that optimisation
> and debugging are never used simultaneously?

The basic issue is that turning on debug info will cause severe code
pessimization, and some people build with "-O2 -g" regularly. For llvm-gcc we
can't take the risk that someone will do this, measure performance, then go
away saying "llvm generates crappy code" :).

Clients of the debug info should be able to grok arbitrarily modified debug
intrinsics though, so the optimizers should all be safe.  Unnesting is
something that your debug consumer will have to deal with, sorry! :(


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 27 00:50:46 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 00:50:46 -0600
Subject: [LLVMbugs] [Bug 1989] "All operands to PHI node must be the same
 type as the PHI node" assertion triggered
In-Reply-To: <bug-1989-206@http.llvm.org/bugs/>
Message-ID: <200802270650.m1R6ok5g003250@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1989


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #5 from Chris Lattner <sabre at nondot.org>  2008-02-27 00:50:45 ---
This apparently 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  Wed Feb 27 04:02:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 04:02:58 -0600
Subject: [LLVMbugs] [Bug 2104] New: Solaris test failure: Analysis/BasicAA/
 2007-08-01-NoAliasAndGEP.ll
Message-ID: <bug-2104-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2104

           Summary: Solaris test failure: Analysis/BasicAA/2007-08-01-
                    NoAliasAndGEP.ll
           Product: libraries
           Version: trunk
          Platform: Sun
        OS/Version: Solaris
            Status: NEW
          Severity: minor
          Priority: P2
         Component: Global Analyses
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ggreif at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Here is the failure:

Running .../llvm/test/Analysis/BasicAA/dg.exp ...
FAIL: .../llvm/test/Analysis/BasicAA/2007-08-01-NoAliasAndGEP.ll
Failed with exit(1) at line 3
while running: llvm-as
.../llvm/test/Analysis/BasicAA/2007-08-01-NoAliasAndGEP.ll -o - | opt -basicaa
-aa-eval -print-all-alias-modref-info -disable-output |& grep {MayAlias:.*i32\*
%Jpointer, i32\* %Ipointer}
child process exited abnormally


Looks like the analysis is correct, only the displaying is mixed up:

ggreif at my [!345] cd Debug/bin/
ggreif at my [!346] llvm-as
.../llvm/test/Analysis/BasicAA/2007-08-01-NoAliasAndGEP.ll -o - | opt -basicaa
-aa-eval -print-all-alias-modref-info -disable-output
Function: foo: 6 pointers, 0 call sites
  NoAlias:      i32* %Jpointer, i32* %qi
  MayAlias:     i32* %qj, i32* %qi
  NoAlias:      i32* %qj, i32* %Jpointer
  NoAlias:      i32* %Ipointer, i32* %qi
  MayAlias:     i32* %Ipointer, i32* %Jpointer    <<<<< HERE WE GO
  NoAlias:      i32* %Ipointer, i32* %qj
  MayAlias:     i32* %q, i32* %qi
  NoAlias:      i32* %q, i32* %Jpointer
  MayAlias:     i32* %q, i32* %qj
  NoAlias:      i32* %q, i32* %Ipointer
  NoAlias:      i32* %p, i32* %qi
  MayAlias:     i32* %p, i32* %Jpointer
  NoAlias:      i32* %p, i32* %qj
  MayAlias:     i32* %p, i32* %Ipointer
  NoAlias:      i32* %p, i32* %q
===== Alias Analysis Evaluator Report =====
  15 Total Alias Queries Performed
  9 no alias responses (60.0%)
  6 may alias responses (40.0%)
  0 must alias responses (0.0%)
  Alias Analysis Evaluator Pointer Alias Summary: 60%/40%/0%
  Alias Analysis Mod/Ref Evaluator Summary: no mod/ref!


My theory is that on solaris the map/set is built up in a different way,
indeterministically.
For output we might need to sort the results, or fix "operator <" on the key
type to start with.

I am looking into this.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 27 05:10:24 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 05:10:24 -0600
Subject: [LLVMbugs] [Bug 2105] New: Instruction selection failures on
	SPARC/Solaris
Message-ID: <bug-2105-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2105

           Summary: Instruction selection failures on SPARC/Solaris
           Product: new-bugs
           Version: unspecified
          Platform: Sun
        OS/Version: Solaris
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ggreif at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Some tests fail on SPARCv8:

1) ----------------------------------------------------
FAIL: .../llvm/test/CodeGen/Generic/2008-01-25-dag-combine-mul.ll
Failed with signal(SIGABRT) at line 1
while running: llvm-as <
.../llvm/test/CodeGen/Generic/2008-01-25-dag-combine-mul.ll | llc
Cannot yet select: 0xd42d28: i32,i32 = umul_lohi 0xd42c38, 0xd42a58

2) ----------------------------------------------------
FAIL: .../llvm/test/CodeGen/Generic/storetrunc-fp.ll
Failed with signal(SIGABRT) at line 1
while running: llvm-as < .../llvm/test/CodeGen/Generic/storetrunc-fp.ll | llc
Cannot yet select: 0xd3fe68: ch = store 0xd3fcf8, 0xd403d8, 0xd39320, 0xd404f8
<0xd36278:0> <trunc f32> alignment=4

3) ----------------------------------------------------
FAIL: .../llvm/test/DebugInfo/funccall.ll
Failed with signal(SIGABRT) at line 1
while running: llvm-upgrade < .../llvm/test/DebugInfo/funccall.ll | llvm-as |
llc
Cannot yet select llvm.dbg.declare: 0xd3adc8: ch = declare 0xd496b8, 0xd3d6c8,
0xd4b228


Anybody who is at home in this area, can you fix this?

Otherwise some guidance welcome and I try to do it myself.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 27 11:58:41 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 11:58:41 -0600
Subject: [LLVMbugs] [Bug 2106] New: ObjC invalid code segfaults clang
Message-ID: <bug-2106-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2106

           Summary: ObjC invalid code segfaults clang
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Keywords: crash-on-invalid
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


clang crashes with the following ObjC code:

- (void)compilerTestAgainst {


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 27 13:21:56 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 13:21:56 -0600
Subject: [LLVMbugs] [Bug 2107] New: Bad middle-end optimization for
	insertelement
Message-ID: <bug-2107-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2107

           Summary: Bad middle-end optimization for insertelement
           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


Testcase:
#include <xmmintrin.h>
__m128i doload64(unsigned long long x) { return _mm_loadl_epi64(&x);}

Generated il:
define <2 x i64> @doload64(i64 %x) nounwind  {
entry:
        %tmp717 = bitcast i64 %x to double
        %tmp8 = insertelement <2 x double> undef, double %tmp717, i32 0
        %tmp9 = insertelement <2 x double> %tmp8, double 0.000000e+00, i32 1
        %tmp11 = bitcast <2 x double> %tmp9 to <2 x i64>
        ret <2 x i64> %tmp11
}

This seems somewhat less than ideal.  Ideally, it would look something like the
following:

define <2 x i64> @doload64(i64 %x) nounwind  {
entry:
        %tmp8 = insertelement <2 x i64> zeroinitializer, i64 %x, i32 0
        ret <2 x i64> %tmp8
}

Note that this affects the codegen for this construct on x86-64: the orginial
codegens to the following:
doload64:
        movd    %rdi, %xmm1
        xorps   %xmm0, %xmm0
        movsd   %xmm1, %xmm0
        ret

and the simplified version codegens to the following:
doload64:
        movd    %rdi, %xmm0
        ret

Although maybe the backend should be smarter about this sort of thing as 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  Wed Feb 27 13:23:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 13:23:34 -0600
Subject: [LLVMbugs] [Bug 2108] New: Horrible x86 codegen for _mm_loadl_epi64
Message-ID: <bug-2108-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2108

           Summary: Horrible x86 codegen for _mm_loadl_epi64
           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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
#include <xmmintrin.h>
__m128i doload64(unsigned long long x) { return _mm_loadl_epi64(&x);}

Generated il:
define <2 x i64> @doload64(i64 %x) nounwind  {
entry:
        %tmp717 = bitcast i64 %x to double
        %tmp8 = insertelement <2 x double> undef, double %tmp717, i32 0
        %tmp9 = insertelement <2 x double> %tmp8, double 0.000000e+00, i32 1
        %tmp11 = bitcast <2 x double> %tmp9 to <2 x i64>
        ret <2 x i64> %tmp11
}

On x86, the this codegens to the following:

doload64:
        subl    $12, %esp
        movl    20(%esp), %eax
        movl    %eax, 4(%esp)
        movl    16(%esp), %eax
        movl    %eax, (%esp)
        movsd   (%esp), %xmm0
        addl    $12, %esp
        ret

which is 6 instructions longer than it needs to be.  The code generator should
really be smart enough to load from the original stack slot.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 27 21:52:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 21:52:03 -0600
Subject: [LLVMbugs] [Bug 2109] New: Missed optimization for extract/
 insertelements equivalent to movhps
Message-ID: <bug-2109-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2109

           Summary: Missed optimization for extract/insertelements
                    equivalent to movhps
           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


Take the following bitcode

define <4 x float> @_mm_loadh_pi22(<4 x float> %__A, <2 x i32>* %__P) nounwind 
{
entry:
        load <2 x i32>* %__P
        bitcast <2 x i32> %0 to <2 x float>
        extractelement <2 x float> %1, i32 0
        extractelement <2 x float> %1, i32 1
        insertelement <4 x float> %__A, float %2, i32 2
        insertelement <4 x float> %4, float %3, i32 3
        ret <4 x float> %5
}

This currently codegens to the following:
_mm_loadh_pi22:
        subl    $8, %esp
        movl    12(%esp), %eax
        movl    4(%eax), %ecx
        movl    %ecx, 4(%esp)
        movl    (%eax), %eax
        movl    %eax, (%esp)
        movss   (%esp), %xmm1
        movaps  %xmm0, %xmm2
        shufps  $3, %xmm1, %xmm2
        shufps  $36, %xmm2, %xmm0
        movss   4(%esp), %xmm1
        movaps  %xmm0, %xmm2
        shufps  $2, %xmm1, %xmm2
        shufps  $132, %xmm2, %xmm0
        addl    $8, %esp
        ret

Unless I've made a mistake, this should codegen to the following:
_mm_loadh_pi22:
        movl    4(%esp), %eax
        movhps  (%eax), %xmm0
        ret

Note that LLVM does know how to generate movhps... it does in fact generate it
for the following, which is functionally equivalent:
define <4 x float> @_mm_loadh_pi22(<4 x float> %__A, <1 x i64>* %__P) nounwind 
{
entry:
        %tmp4 = bitcast <1 x i64>* %__P to double*
        %tmp5 = load double* %tmp4
        %tmp6 = insertelement <2 x double> undef, double %tmp5, i32 0
        %tmp8 = bitcast <2 x double> %tmp6 to <4 x float>
        %tmp9 = shufflevector <4 x float> %__A, <4 x float> %tmp8, <4 x i32> <
i32 0, i32 1, i32 4, i32 5 >
        ret <4 x float> %tmp9
}


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 27 22:20:52 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 22:20:52 -0600
Subject: [LLVMbugs] [Bug 2110] New: Assertion `0 && "Cannot expand insert
 element operand"' failed.
Message-ID: <bug-2110-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2110

           Summary: Assertion `0 && "Cannot expand insert element operand"'
                    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: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1477)
 --> (http://llvm.org/bugs/attachment.cgi?id=1477)
Testcase

Per description: assertion/crash in LegalizeDAG (attached testcase
bug-point-reduced).  My guess is it's having trouble with the extractelement <2
x i64>, although I'm not sure.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 27 23:56:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 27 Feb 2008 23:56:16 -0600
Subject: [LLVMbugs] [Bug 2105] Instruction selection failures on
	SPARC/Solaris
In-Reply-To: <bug-2105-206@http.llvm.org/bugs/>
Message-ID: <200802280556.m1S5uGO0024258@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2105


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-02-27 23:56:15 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058920.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058921.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058922.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058923.html

Note that the sparc backend isn't generating the 32 x 32 -> 64 multiplies.  In
the meantime we generate correct but non crashing code.

Feel free to un-xfail the powf test if you're interested.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 28 02:41:41 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 28 Feb 2008 02:41:41 -0600
Subject: [LLVMbugs] [Bug 2104] Solaris test failure:
 Analysis/BasicAA/2007-08-01-NoAliasAndGEP. ll
In-Reply-To: <bug-2104-206@http.llvm.org/bugs/>
Message-ID: <200802280841.m1S8ffCw002834@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2104


Gabor Greif <ggreif at gmail.com> changed:

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




--- Comment #3 from Gabor Greif <ggreif at gmail.com>  2008-02-28 02:41:40 ---
fixed in 47712:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058928.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 Feb 28 10:49:02 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 28 Feb 2008 10:49:02 -0600
Subject: [LLVMbugs] [Bug 2111] New: ICE llvm-gcc (4.2)
Message-ID: <bug-2111-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2111

           Summary: ICE llvm-gcc (4.2)
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: alenhar2 at uiuc.edu
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1480)
 --> (http://llvm.org/bugs/attachment.cgi?id=1480)
one line reduction

The attached test case failes in llvm-gcc (4.2) if you give the -fno-common
flag.
-emit-llvm doesn't matter.  GNU gcc-4.2 doesn't mind the file.  reproduce with:
llvm-gcc -o /dev/null min.i -c -fno-common


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 28 11:58:10 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 28 Feb 2008 11:58:10 -0600
Subject: [LLVMbugs] [Bug 2111] ICE llvm-gcc (4.2)
In-Reply-To: <bug-2111-206@http.llvm.org/bugs/>
Message-ID: <200802281758.m1SHwAvB028135@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2111


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-02-28 11:58:09 ---
Fixed, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058938.html

Thanks for the nice reduction!

-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  Fri Feb 29 05:59:14 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 29 Feb 2008 05:59:14 -0600
Subject: [LLVMbugs] [Bug 2112] New: LoopAligner pass fallout on Sparc/Solaris
Message-ID: <bug-2112-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2112

           Summary: LoopAligner pass fallout on Sparc/Solaris
           Product: new-bugs
           Version: unspecified
          Platform: Sun
        OS/Version: Solaris
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ggreif at gmail.com
                CC: llvmbugs at cs.uiuc.edu


I get this crash on SPARCv8:


(gdb) run ggg.bc -f
Starting program: .../llvm/Debug/bin/llc ggg.bc -f
warning: Temporarily disabling breakpoints for unloaded shared library
"/usr/lib/ld.so.1"
warning: Lowest section in /lib/libpthread.so.1 is .dynamic at 00000074

Program received signal SIGSEGV, Segmentation fault.
0x0095fd68 in llvm::TargetLowering::getPrefLoopAlignment (this=0x0)
    at .../llvm/include/llvm/Target/TargetLowering.h:566
566       return PrefLoopAlignment;
(gdb) bt
#0  0x0095fd68 in llvm::TargetLowering::getPrefLoopAlignment (this=0x0)
    at .../llvm/include/llvm/Target/TargetLowering.h:566
#1  0x0095f67c in (anonymous namespace)::LoopAligner::runOnMachineFunction
(this=0xd387f0, MF=@0xd42e90)
    at LoopAligner.cpp:54
#2  0x00481e28 in llvm::MachineFunctionPass::runOnFunction (this=0xd387f0,
F=@0xd3b150)
    at .../llvm/include/llvm/CodeGen/MachineFunctionPass.h:41
#3  0x00ba4a70 in llvm::FPPassManager::runOnFunction (this=0xd3a8c8,
F=@0xd3b150) at PassManager.cpp:1170
#4  0x00ba4704 in llvm::FunctionPassManagerImpl::run (this=0xd3a5e0,
F=@0xd3b150) at PassManager.cpp:1128
#5  0x00ba4590 in llvm::FunctionPassManager::run (this=0xffbff400, F=@0xd3b150)
at PassManager.cpp:1073
#6  0x00413e9c in main (argc=3, argv=0xffbff4bc) at llc.cpp:296
(gdb) up
#1  0x0095f67c in (anonymous namespace)::LoopAligner::runOnMachineFunction
(this=0xd387f0, MF=@0xd42e90)
    at LoopAligner.cpp:54
54      unsigned Align =
MF.getTarget().getTargetLowering()->getPrefLoopAlignment();




It appears that on Sparc no TargetLowering object is allocated and LoopAligner
does not guard against this condition.

Three solution attempts come to my mind:

1) Install a TargetLowering for Sparc too.
2) Guard against TargetLowering == NULL in LoopAligner
3) Do not register with LoopAligner when there is no TargetLowering installed.

1) looks best to me, but other people may prefer 3). 2) probably makes little
sense.

Ah, the test in question is CodeGen/Generic/GC/alloc_loop.ll
but many others fail too.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 29 11:53:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 29 Feb 2008 11:53:37 -0600
Subject: [LLVMbugs] [Bug 2112] LoopAligner pass fallout on Sparc/Solaris
In-Reply-To: <bug-2112-206@http.llvm.org/bugs/>
Message-ID: <200802291753.m1THrbxi024890@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2112


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

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




--- Comment #2 from Evan Cheng <evan.cheng at apple.com>  2008-02-29 11:53:36 ---
Fixed. Thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058986.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 Feb 29 12:52:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 29 Feb 2008 12:52:29 -0600
Subject: [LLVMbugs] [Bug 1925] Assertion failed in RegisterScavenging with
 regalloc=local
In-Reply-To: <bug-1925-206@http.llvm.org/bugs/>
Message-ID: <200802291852.m1TIqT3T026392@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1925


Bill Wendling <isanbard at gmail.com> changed:

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




--- Comment #7 from Bill Wendling <isanbard at gmail.com>  2008-02-29 12:52:28 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080225/058990.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 Feb 29 15:50:56 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 29 Feb 2008 15:50:56 -0600
Subject: [LLVMbugs] [Bug 2106] ObjC invalid code segfaults clang
In-Reply-To: <bug-2106-206@http.llvm.org/bugs/>
Message-ID: <200802292150.m1TLouxn031605@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2106


snaroff at apple.com <snaroff at apple.com> changed:

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




--- Comment #2 from snaroff at apple.com  2008-02-29 15:50:56 ---
Fixed in http://llvm.org/viewvc/llvm-project?view=rev&revision=47768.

The compiler now says...

[snaroff:llvm/tools/clang] snarofflocal% ../../Debug/bin/clang
test/Sema/missing-method-context.m 
test/Sema/missing-method-context.m:2:1: error: missing context for method
declaration
- (void)compilerTestAgainst;  
^
test/Sema/missing-method-context.m:4:1: error: expected method body
void xx(); 
^
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 Feb 29 17:31:25 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 29 Feb 2008 17:31:25 -0600
Subject: [LLVMbugs] [Bug 2103] clang doesn't reject address of bitfield or
 vector element
In-Reply-To: <bug-2103-206@http.llvm.org/bugs/>
Message-ID: <200802292331.m1TNVP4L002287@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2103


snaroff at apple.com <snaroff at apple.com> changed:

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




--- Comment #1 from snaroff at apple.com  2008-02-29 17:31:24 ---
Fixed in http://llvm.org/viewvc/llvm-project?view=rev&revision=47775.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?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 Feb 29 18:40:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 29 Feb 2008 18:40:47 -0600
Subject: [LLVMbugs] [Bug 2113] New: verifyModule accepts "alloca void ()"
Message-ID: <bug-2113-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2113

           Summary: verifyModule accepts "alloca void ()"
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ofv at wanadoo.es
                CC: llvmbugs at cs.uiuc.edu, ofv at wanadoo.es


verifyModule does not detect this mistake:

        alloca void ()          ; <void ()*>:1 [#uses=1]

(Note that the type alloca'ted is a function type, not a function pointer type)

It bombs later with an assert:

Assertion failed: Ty->isSized() && "Cannot getTypeInfo() on a type that is
unsiz
ed!", file d:/dev/lib/llvm-2.1/lib/Target/TargetData.cpp, line 401

This was tested on 2.1 (from where the assert message was taken) and HEAD.


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