From bugzilla-daemon at cs.uiuc.edu  Thu Mar  1 11:12:06 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 1 Mar 2007 11:12:06 -0600
Subject: [LLVMbugs] [Bug 1233] NEW: multi-dimensional VLA indicies not
	scaled by llvm-gcc
Message-ID: <200703011712.l21HC61W009170@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1233

           Summary: multi-dimensional VLA indicies not scaled by llvm-gcc
           Product: tools
           Version: 1.9
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: djg at cray.com


I typed this into the online demo:

int foo(int v, int w, int x, int y, int z,
        float A[][w][x][y][z],
        int g, int h, int i, int j, int k)
{
  return A[g][h][i][j][k];
}

It gave me this:

int %foo(int %v, int %w, int %x, int %y, int %z, float* %A, int %g, int %h, int
%i, int %j, int %k) {
entry:
	%tmp64.sum = add int %h, %g		; <int> [#uses=1]
	%tmp62.sum = add int %tmp64.sum, %i		; <int> [#uses=1]
	%tmp60.sum = add int %tmp62.sum, %j		; <int> [#uses=1]
	%tmp54.sum = add int %tmp60.sum, %k		; <int> [#uses=1]
	%tmp66 = getelementptr float* %A, int %tmp54.sum		; <float*> [#uses=1]
	%tmp = load float* %tmp66		; <float> [#uses=1]
	%tmp67 = cast float %tmp to int		; <int> [#uses=1]
	ret int %tmp67
}

The VLA indicies are not being scaled; it's just adding them all up verbatim and
doing a one-dimensional getelementptr.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar  1 12:52:07 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 1 Mar 2007 12:52:07 -0600
Subject: [LLVMbugs] [Bug 1234] NEW: Build of ADT/SmallPtrSet fails
Message-ID: <200703011852.l21Iq7C7011047@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1234

           Summary: Build of ADT/SmallPtrSet fails
           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: aeriksson at gmail.com


Building CVS HEAD on MacOSX fails with the following error:

$ make
make[1]: Nothing to be done for `all'.
llvm[1]: Compiling SmallPtrSet.cpp for Debug build 
/Users/andreas/build/llvmHEAD/llvm/include/llvm/ADT/SmallPtrSet.h:191: error: non-constant ???((N | 
(N - 1)) + 1)??? cannot be used as template argument
/Users/andreas/build/llvmHEAD/llvm/include/llvm/ADT/SmallPtrSet.h:191: error: ???Val??? is not a 
member of ???<declaration error>???
make[1]: *** [/Users/andreas/build/llvmHEAD/llvm/lib/Support/Debug/SmallPtrSet.o] Error 1
make: *** [all] Error 1



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  2 00:48:57 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Mar 2007 00:48:57 -0600
Subject: [LLVMbugs] [Bug 1235] NEW: ICE on invalid code
Message-ID: <200703020648.l226mvVn002626@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1235

           Summary: ICE on invalid code
           Product: tools
           Version: 1.0
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


llvm-gcc crashes on:

float foo( float A[][w], int g, int h)
{
  return A[g][h];
}



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  2 01:33:05 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Mar 2007 01:33:05 -0600
Subject: [LLVMbugs] [Bug 1235] ICE on invalid code
Message-ID: <200703020733.l227X5ir010271@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1235

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
           Keywords|                            |crash-on-invalid
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0



------- Additional Comments From sabre at nondot.org  2007-03-02 01:33 -------
Fixed, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070226/045384.html

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  2 01:37:55 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Mar 2007 01:37:55 -0600
Subject: [LLVMbugs] [Bug 1233] multi-dimensional VLA indicies not scaled by
	llvm-gcc
Message-ID: <200703020737.l227bt0R011324@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1233

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0



------- Additional Comments From sabre at nondot.org  2007-03-02 01:37 -------
Nice catch.  Testcase here:
CFrontend/2007-03-01-VarSizeArrayIdx.c

Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070226/045387.html

The patch generates pretty horrible LLVM code (in the VLA case only), but the code generator groks it 
well.  We compile your testcase to (with fomit-frame-pointer):

_foo:
        movl 8(%esp), %eax
        imull 12(%esp), %eax
        imull 16(%esp), %eax
        imull 20(%esp), %eax
        movl 36(%esp), %ecx
        addl 32(%esp), %ecx
        imull 28(%esp), %eax
        addl 40(%esp), %ecx
        leal (,%eax,4), %eax
        addl 44(%esp), %ecx
        addl 24(%esp), %eax
        cvttss2si (%eax,%ecx,4), %eax
        ret

GCC apparently misses reassociating some arithmetic and does significantly worse:

_foo:
        pushl   %esi
        movl    24(%esp), %eax
        movl    %eax, %ecx
        imull   20(%esp), %ecx
        movl    %ecx, %edx
        imull   16(%esp), %edx
        movl    %edx, %esi
        imull   12(%esp), %esi
        imull   32(%esp), %esi
        imull   44(%esp), %eax
        imull   40(%esp), %ecx
        imull   36(%esp), %edx
        addl    %edx, %eax
        addl    %ecx, %eax
        addl    48(%esp), %eax
        sall    $2, %eax
        leal    (%eax,%esi,4), %esi
        movl    28(%esp), %eax
        cvttss2si       (%esi,%eax), %eax
        popl    %esi
        ret

I'd appreciate it if you could verify that the code we now produce is correct.  I verified a simpler case.

Thanks again for noticing this,

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  2 01:56:46 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Mar 2007 01:56:46 -0600
Subject: [LLVMbugs] [Bug 1236] NEW: vector of ints not supported
Message-ID: <200703020756.l227ukQG015181@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1236

           Summary: vector of ints not supported
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: isanbard at gmail.com


$ llvm-gcc -S -mmmx -O2 t.c
Assertion failed: (0 && "This action is not supported yet!"), function LegalizeOp, file /Volumes/Gir/
devel/llvm/llvm.src/lib/CodeGen/SelectionDAG/LegalizeDAG.cpp, line 1510.
t.c:9: internal compiler error: Abort trap
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://llvm.org/bugs> for instructions.
$ cat t.c
#include <mmintrin.h>

typedef int V __attribute__ ((__vector_size__ (8)));

void foo(V *A, V *B)
{
  *A = *B;
  _mm_empty();
}

GCC generates this:

$ gcc -S -mmmx -O2 t.c
$ more t.s
        .text
        .align 4,0x90
.globl _foo
_foo:
        pushl   %ebp
        movl    %esp, %ebp
        subl    $8, %esp
        movl    12(%ebp), %eax
        movl    (%eax), %edx
        movl    4(%eax), %ecx
        movl    8(%ebp), %eax
        movl    %edx, (%eax)
        movl    %ecx, 4(%eax)
        emms
        leave
        ret
        .subsections_via_symbols



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  2 09:29:25 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Mar 2007 09:29:25 -0600
Subject: [LLVMbugs] [Bug 1233] multi-dimensional VLA indicies not scaled by
	llvm-gcc
Message-ID: <200703021529.l22FTPuR018726@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1233

djg at cray.com changed:

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



------- Additional Comments From djg at cray.com  2007-03-02 09:29 -------
That new LLVM code still isn't right. The g index is scaled now, but h, i, and j
are still being added without scaling.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  2 12:18:11 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Mar 2007 12:18:11 -0600
Subject: [LLVMbugs] [Bug 1234] Build of ADT/SmallPtrSet fails
Message-ID: <200703021818.l22IIBuP022049@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1234

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
          Component|new bugs                    |Support Libraries
           Keywords|                            |build-problem
            Product|new-bugs                    |libraries
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0
            Version|unspecified                 |1.0



------- Additional Comments From sabre at nondot.org  2007-03-02 12:18 -------
Okay, very strange compiler bug.  Easy enough to work around though:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070226/045400.html

Thanks!

BTW, I recommend downloading the latest Xcode from developer.apple.com, there are several issues with 
older versions.  See the Getting Started guide for details on specific broken ones.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  2 14:57:50 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Mar 2007 14:57:50 -0600
Subject: [LLVMbugs] [Bug 1237] NEW: Uninitialized variables in
	lib/CodeGen/MachOWriter.cpp
Message-ID: <200703022057.l22KvoTV024794@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1237

           Summary: Uninitialized variables in lib/CodeGen/MachOWriter.cpp
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: minor
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: me22.ca at gmail.com


make[2]: Entering directory `/home/me22/programming/llvm-cvs/lib/CodeGen'
llvm[2]: Compiling MachOWriter.cpp for Release build 
MachOWriter.cpp: In member function ???void
llvm::MachOWriter::CalculateRelocations(llvm::MachOWriter::MachOSection&)???:
MachOWriter.cpp:740: warning: ???TargetIndex??? may be used uninitialized in this
function
MachOWriter.cpp:739: warning: ???TargetAddr??? may be used uninitialized in this
function

The code does look like they're used w/o being initialized.  I'm not certain it
matters, but eliminating the warnings would be good anyways.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  2 17:40:51 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 Mar 2007 17:40:51 -0600
Subject: [LLVMbugs] [Bug 1233] multi-dimensional VLA indicies not scaled by
	llvm-gcc
Message-ID: <200703022340.l22NepA0027894@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1233

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|                            |FIXED
            Version|1.9                         |1.0



------- Additional Comments From sabre at nondot.org  2007-03-02 17:40 -------
Okay, here's the second half of the patch, which handles the ARRAY_REF array case (previous patch 
handled ARRAY_REF pointer case):
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070226/045418.html

With this, we now compile:

int foo(int v, int w, int x, int y, int z,
        float A[][w][x],
        int g, int h, int i, int j, int k){
  return A[g][10000][0];
}

into:

_foo:
        movl 28(%esp), %eax
        imull 8(%esp), %eax
        leal 40000(,%eax,4), %eax
        imull 12(%esp), %eax
        movl 24(%esp), %ecx
        cvttss2si (%ecx,%eax), %eax
        ret

... which looks right to me.

GCC produces this, which is also right, but just worse code:

_foo:
        movl    12(%esp), %eax
        movl    %eax, %edx
        imull   8(%esp), %edx
        imull   $40000, %eax, %eax
        imull   28(%esp), %edx
        leal    (%eax,%edx,4), %edx
        movl    24(%esp), %eax
        cvttss2si       (%edx,%eax), %eax
        ret??

We now compile your full case to:

_foo2:
        movl 8(%esp), %eax
        imull 28(%esp), %eax
        addl 32(%esp), %eax
        imull 12(%esp), %eax
        addl 36(%esp), %eax
        imull 16(%esp), %eax
        addl 40(%esp), %eax
        imull 20(%esp), %eax
        addl 44(%esp), %eax
        movl 24(%esp), %ecx
        cvttss2si (%ecx,%eax,4), %eax
        ret

GCC produces:

_foo2:
        pushl   %esi
        movl    24(%esp), %eax
        movl    %eax, %ecx
        imull   20(%esp), %ecx
        imull   44(%esp), %eax
        movl    %ecx, %edx
        imull   16(%esp), %edx
        imull   40(%esp), %ecx
        movl    %edx, %esi
        imull   36(%esp), %edx
        imull   12(%esp), %esi
        imull   32(%esp), %esi
        addl    %edx, %eax
        addl    %ecx, %eax
        addl    48(%esp), %eax
        sall    $2, %eax
        leal    (%eax,%esi,4), %esi
        movl    28(%esp), %eax
        cvttss2si       (%esi,%eax), %eax
        popl    %esi
        ret

Please verify that the new codegen is correct.  Thanks again,

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From isanbard at gmail.com  Fri Mar  2 19:20:39 2007
From: isanbard at gmail.com (Bill Wendling)
Date: Fri, 2 Mar 2007 17:20:39 -0800
Subject: [LLVMbugs] Segfault during bootstrap
Message-ID: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>

Hi all,

I'm getting a segfault during a bootstrap. Here's the compile line:

$ /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCReader.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipo.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCWriter.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMX86.o
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMPowerPC.o
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSelectionDAG.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCodeGen.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMScalarOpts.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTransformUtils.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipa.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMAnalysis.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTarget.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCore.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSupport.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMbzip2.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSystem.a
powerpc-apple-darwin9-c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings
-DHAVE_CONFIG_H -Wno-unused -DTARGET_NAME=\"powerpc-apple-darwin9\"
-DNDEBUG -DENABLE_LLVM -D__STDC_LIMIT_MACROS
-DLLVM_VERSION_INFO='"1033"' -I. -I.
-I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc
-I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/.
-I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/../include
-I./../intl -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/../libcpp/include
 -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/include
-I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/include
-o libllvmgcc.dylib \
           /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCReader.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipo.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCWriter.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMX86.o
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMPowerPC.o
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSelectionDAG.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCodeGen.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMScalarOpts.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTransformUtils.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipa.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMAnalysis.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTarget.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCore.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSupport.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMbzip2.a
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSystem.a
 -dynamiclib -single_module \
           -install_name @executable_path/libllvmgcc.dylib

And here's the error message:

/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp:
In function 'void dummy_function()':
/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp:41:
internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://llvm.org/bugs> for instructions.
make[2]: *** [libllvmgcc.dylib] Error 1
make[2]: *** Waiting for unfinished jobs....
rm cpp.pod gcov.pod gfdl.pod fsf-funding.pod gpl.pod gcc.pod
make[1]: *** [all-gcc] Error 2
+ exit 1
make: *** [install] Error 1

-bw

From rspencer at reidspencer.com  Fri Mar  2 19:44:22 2007
From: rspencer at reidspencer.com (Reid Spencer)
Date: Fri, 02 Mar 2007 17:44:22 -0800
Subject: [LLVMbugs] Segfault during bootstrap
In-Reply-To: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>
References: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>
Message-ID: <1172886262.17762.49.camel@bashful.x10sys.com>

Bill,

My bootstrap is still running. No issues yet. I noticed some regressions
in InstCombine, perhaps related to Chris' changes. Perhaps to mine. I'm
looking into it.

Reid.

On Fri, 2007-03-02 at 17:20 -0800, Bill Wendling wrote:
> Hi all,
> 
> I'm getting a segfault during a bootstrap. Here's the compile line:
> 
> $ /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCReader.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipo.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCWriter.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMX86.o
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMPowerPC.o
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSelectionDAG.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCodeGen.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMScalarOpts.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTransformUtils.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipa.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMAnalysis.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTarget.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCore.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSupport.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMbzip2.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSystem.a
> powerpc-apple-darwin9-c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings
> -DHAVE_CONFIG_H -Wno-unused -DTARGET_NAME=\"powerpc-apple-darwin9\"
> -DNDEBUG -DENABLE_LLVM -D__STDC_LIMIT_MACROS
> -DLLVM_VERSION_INFO='"1033"' -I. -I.
> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc
> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/.
> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/../include
> -I./../intl -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/../libcpp/include
>  -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/include
> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/include
> -o libllvmgcc.dylib \
>            /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCReader.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipo.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCWriter.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMX86.o
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMPowerPC.o
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSelectionDAG.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCodeGen.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMScalarOpts.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTransformUtils.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipa.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMAnalysis.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTarget.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCore.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSupport.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMbzip2.a
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSystem.a
>  -dynamiclib -single_module \
>            -install_name @executable_path/libllvmgcc.dylib
> 
> And here's the error message:
> 
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp:
> In function 'void dummy_function()':
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp:41:
> internal compiler error: Segmentation fault
> Please submit a full bug report,
> with preprocessed source if appropriate.
> See <URL:http://llvm.org/bugs> for instructions.
> make[2]: *** [libllvmgcc.dylib] Error 1
> make[2]: *** Waiting for unfinished jobs....
> rm cpp.pod gcov.pod gfdl.pod fsf-funding.pod gpl.pod gcc.pod
> make[1]: *** [all-gcc] Error 2
> + exit 1
> make: *** [install] Error 1
> 
> -bw
> _______________________________________________
> LLVMbugs mailing list
> LLVMbugs at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs


From sabre at nondot.org  Fri Mar  2 20:37:09 2007
From: sabre at nondot.org (Chris Lattner)
Date: Fri, 2 Mar 2007 18:37:09 -0800 (PST)
Subject: [LLVMbugs] Segfault during bootstrap
In-Reply-To: <1172886262.17762.49.camel@bashful.x10sys.com>
References: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>
	<1172886262.17762.49.camel@bashful.x10sys.com>
Message-ID: <Pine.LNX.4.62.0703021836520.9405@nondot.org>

On Fri, 2 Mar 2007, Reid Spencer wrote:
> My bootstrap is still running. No issues yet. I noticed some regressions
> in InstCombine, perhaps related to Chris' changes. Perhaps to mine. I'm
> looking into it.

I just did an darwin/x86 bootstrap, works for me.

-Chris

> On Fri, 2007-03-02 at 17:20 -0800, Bill Wendling wrote:
>> Hi all,
>>
>> I'm getting a segfault during a bootstrap. Here's the compile line:
>>
>> $ /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCReader.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipo.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCWriter.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMX86.o
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMPowerPC.o
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSelectionDAG.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCodeGen.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMScalarOpts.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTransformUtils.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipa.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMAnalysis.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTarget.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCore.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSupport.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMbzip2.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSystem.a
>> powerpc-apple-darwin9-c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings
>> -DHAVE_CONFIG_H -Wno-unused -DTARGET_NAME=\"powerpc-apple-darwin9\"
>> -DNDEBUG -DENABLE_LLVM -D__STDC_LIMIT_MACROS
>> -DLLVM_VERSION_INFO='"1033"' -I. -I.
>> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc
>> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/.
>> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/../include
>> -I./../intl -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/../libcpp/include
>>  -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/include
>> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/include
>> -o libllvmgcc.dylib \
>>            /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCReader.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipo.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCWriter.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMX86.o
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/LLVMPowerPC.o
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSelectionDAG.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCodeGen.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMScalarOpts.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTransformUtils.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMipa.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMAnalysis.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMTarget.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMCore.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSupport.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMbzip2.a
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMSystem.a
>>  -dynamiclib -single_module \
>>            -install_name @executable_path/libllvmgcc.dylib
>>
>> And here's the error message:
>>
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp:
>> In function 'void dummy_function()':
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp:41:
>> internal compiler error: Segmentation fault
>> Please submit a full bug report,
>> with preprocessed source if appropriate.
>> See <URL:http://llvm.org/bugs> for instructions.
>> make[2]: *** [libllvmgcc.dylib] Error 1
>> make[2]: *** Waiting for unfinished jobs....
>> rm cpp.pod gcov.pod gfdl.pod fsf-funding.pod gpl.pod gcc.pod
>> make[1]: *** [all-gcc] Error 2
>> + exit 1
>> make: *** [install] Error 1
>>
>> -bw
>> _______________________________________________
>> LLVMbugs mailing list
>> LLVMbugs at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
>
> _______________________________________________
> LLVMbugs mailing list
> LLVMbugs at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
>

-Chris

-- 
http://nondot.org/sabre/
http://llvm.org/

From dpatel at apple.com  Fri Mar  2 19:52:59 2007
From: dpatel at apple.com (Devang Patel)
Date: Fri, 2 Mar 2007 17:52:59 -0800
Subject: [LLVMbugs] Segfault during bootstrap
In-Reply-To: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>
References: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>
Message-ID: <FCF2D631-7276-46BF-A62C-C014A8DC1029@apple.com>


On Mar 2, 2007, at 5:20 PM, Bill Wendling wrote:

> powerpc-apple-darwin9-c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings

<snip>

> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/ 
> include
> -o libllvmgcc.dylib \
>            /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/ 
> gcc/llvm-linker-hack.cpp
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/ 
> Release/lib/libLLVMBCReader.a

<snip>

> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm- 
> linker-hack.cpp:
> In function 'void dummy_function()':
> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm- 
> linker-hack.cpp:41:
> internal compiler error: Segmentation fault

powerpc-apple-darwin9-c++ is crashing. Is it part of llvm-gcc or plain  
gcc ?
-
Devang

> Please submit a full bug report,
> with preprocessed source if appropriate.
> See <URL:http://llvm.org/bugs> for instructions.
> make[2]: *** [libllvmgcc.dylib] Error 1
> make[2]: *** Waiting for unfinished jobs....

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20070302/f57aa903/attachment.html 

From rspencer at reidspencer.com  Fri Mar  2 20:37:40 2007
From: rspencer at reidspencer.com (Reid Spencer)
Date: Fri, 02 Mar 2007 18:37:40 -0800
Subject: [LLVMbugs] Segfault during bootstrap
In-Reply-To: <FCF2D631-7276-46BF-A62C-C014A8DC1029@apple.com>
References: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>
	<FCF2D631-7276-46BF-A62C-C014A8DC1029@apple.com>
Message-ID: <1172889460.17762.51.camel@bashful.x10sys.com>

My bootstrap on Linux completed successfully. 

Reid.

On Fri, 2007-03-02 at 17:52 -0800, Devang Patel wrote:
> 
> On Mar 2, 2007, at 5:20 PM, Bill Wendling wrote:
> 
> > powerpc-apple-darwin9-c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings
> 
> 
> <snip>
> 
> > -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/include
> > -o libllvmgcc.dylib \
> > 
> >   /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp
> > /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/Release/lib/libLLVMBCReader.a
> 
> 
> <snip>
> 
> 
> > /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp:
> > In function 'void dummy_function()':
> > /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm-linker-hack.cpp:41:
> > internal compiler error: Segmentation fault
> 
> 
> powerpc-apple-darwin9-c++ is crashing. Is it part of llvm-gcc or plain
> gcc ?
> -
> Devang
> 
> > Please submit a full bug report,
> > with preprocessed source if appropriate.
> > See <URL:http://llvm.org/bugs> for instructions.
> > make[2]: *** [libllvmgcc.dylib] Error 1
> > make[2]: *** Waiting for unfinished jobs....
> 
> _______________________________________________
> LLVMbugs mailing list
> LLVMbugs at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs


From isanbard at gmail.com  Fri Mar  2 23:53:46 2007
From: isanbard at gmail.com (Bill Wendling)
Date: Fri, 2 Mar 2007 21:53:46 -0800
Subject: [LLVMbugs] Segfault during bootstrap
In-Reply-To: <FCF2D631-7276-46BF-A62C-C014A8DC1029@apple.com>
References: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>
	<FCF2D631-7276-46BF-A62C-C014A8DC1029@apple.com>
Message-ID: <346BC2D5-D087-4AD5-964D-A61D4B2944E2@gmail.com>

It might be that it's the system compiler...I thought that it was the  
built one...Notice, this isn't just a normal bootstrap but one built  
4-way fat...

-bw

On Mar 2, 2007, at 5:52 PM, Devang Patel wrote:

>
> On Mar 2, 2007, at 5:20 PM, Bill Wendling wrote:
>
>> powerpc-apple-darwin9-c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings
>
> <snip>
>
>> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/ 
>> include
>> -o libllvmgcc.dylib \
>>            /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/ 
>> src/gcc/llvm-linker-hack.cpp
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/ 
>> Release/lib/libLLVMBCReader.a
>
> <snip>
>
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm- 
>> linker-hack.cpp:
>> In function 'void dummy_function()':
>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/llvm- 
>> linker-hack.cpp:41:
>> internal compiler error: Segmentation fault
>
> powerpc-apple-darwin9-c++ is crashing. Is it part of llvm-gcc or  
> plain gcc ?
> -
> Devang
>
>> Please submit a full bug report,
>> with preprocessed source if appropriate.
>> See <URL:http://llvm.org/bugs> for instructions.
>> make[2]: *** [libllvmgcc.dylib] Error 1
>> make[2]: *** Waiting for unfinished jobs....
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cs.uiuc.edu/pipermail/llvmbugs/attachments/20070302/40761671/attachment.html 

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 00:18:40 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 00:18:40 -0600
Subject: [LLVMbugs] [Bug 1237] Uninitialized variables in
	lib/CodeGen/MachOWriter.cpp
Message-ID: <200703030618.l236IexB002787@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1237

natebegeman at mac.com changed:

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





------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From isanbard at gmail.com  Sat Mar  3 01:20:05 2007
From: isanbard at gmail.com (Bill Wendling)
Date: Fri, 2 Mar 2007 23:20:05 -0800
Subject: [LLVMbugs] Another Breakage
Message-ID: <829E83C3-741C-48B2-A716-7CF4B6E3BB84@gmail.com>

Here ya go:

g++ -I/private/tmp/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/lib/ 
ExecutionEngine -I/tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/ 
lib/ExecutionEngine -I/private/tmp/llvm-1033-03.roots/ 
llvm-1033-03~obj/obj-llvm/include -I/tmp/llvm-1033-03.roots/ 
llvm-1033-03~obj/src/llvm/include -I/tmp/llvm-1033-03.roots/ 
llvm-1033-03~obj/obj-llvm/include -I/tmp/llvm-1033-03.roots/ 
llvm-1033-03~obj/src/llvm/include  -D_GNU_SOURCE - 
D__STDC_LIMIT_MACROS -O2  -fno-exceptions -D_DEBUG - 
DLLVM_VERSION_INFO='" Apple Build #1033-03"' -Woverloaded-virtual - 
pedantic -Wall -W -Wwrite-strings -Wno-long-long -Wunused -Wno-unused- 
parameter  -fstrict-aliasing -arch ppc -arch i386 -arch ppc64 -arch  
x86_64 -c /tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/lib/ 
ExecutionEngine/ExecutionEngine.cpp -o /private/tmp/ 
llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/lib/ExecutionEngine/ 
Release/ExecutionEngine.o
/tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/lib/ExecutionEngine/ 
ExecutionEngine.cpp: In member function 'llvm::GenericValue  
llvm::ExecutionEngine::getConstantValue(const llvm::Constant*)':
/tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/lib/ExecutionEngine/ 
ExecutionEngine.cpp:428: error: 'union llvm::GenericValue' has no  
member named 'APIntVal'
/tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/lib/ExecutionEngine/ 
ExecutionEngine.cpp: In member function 'void  
llvm::ExecutionEngine::StoreValueToMemory(llvm::GenericValue,  
llvm::GenericValue*, const llvm::Type*)':


From rspencer at reidspencer.com  Sat Mar  3 01:37:05 2007
From: rspencer at reidspencer.com (Reid Spencer)
Date: Fri, 02 Mar 2007 23:37:05 -0800
Subject: [LLVMbugs] Another Breakage
In-Reply-To: <829E83C3-741C-48B2-A716-7CF4B6E3BB84@gmail.com>
References: <829E83C3-741C-48B2-A716-7CF4B6E3BB84@gmail.com>
Message-ID: <1172907425.17762.61.camel@bashful.x10sys.com>

Gah, sorry. Forgot to check in the header. It's in now.

Reid.

On Fri, 2007-03-02 at 23:20 -0800, Bill Wendling wrote:
> Here ya go:
> 
> g++ -I/private/tmp/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/lib/ 
> ExecutionEngine -I/tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/ 
> lib/ExecutionEngine -I/private/tmp/llvm-1033-03.roots/ 
> llvm-1033-03~obj/obj-llvm/include -I/tmp/llvm-1033-03.roots/ 
> llvm-1033-03~obj/src/llvm/include -I/tmp/llvm-1033-03.roots/ 
> llvm-1033-03~obj/obj-llvm/include -I/tmp/llvm-1033-03.roots/ 
> llvm-1033-03~obj/src/llvm/include  -D_GNU_SOURCE - 
> D__STDC_LIMIT_MACROS -O2  -fno-exceptions -D_DEBUG - 
> DLLVM_VERSION_INFO='" Apple Build #1033-03"' -Woverloaded-virtual - 
> pedantic -Wall -W -Wwrite-strings -Wno-long-long -Wunused -Wno-unused- 
> parameter  -fstrict-aliasing -arch ppc -arch i386 -arch ppc64 -arch  
> x86_64 -c /tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/lib/ 
> ExecutionEngine/ExecutionEngine.cpp -o /private/tmp/ 
> llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/lib/ExecutionEngine/ 
> Release/ExecutionEngine.o
> /tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/lib/ExecutionEngine/ 
> ExecutionEngine.cpp: In member function 'llvm::GenericValue  
> llvm::ExecutionEngine::getConstantValue(const llvm::Constant*)':
> /tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/lib/ExecutionEngine/ 
> ExecutionEngine.cpp:428: error: 'union llvm::GenericValue' has no  
> member named 'APIntVal'
> /tmp/llvm-1033-03.roots/llvm-1033-03~obj/src/llvm/lib/ExecutionEngine/ 
> ExecutionEngine.cpp: In member function 'void  
> llvm::ExecutionEngine::StoreValueToMemory(llvm::GenericValue,  
> llvm::GenericValue*, const llvm::Type*)':
> 
> _______________________________________________
> LLVMbugs mailing list
> LLVMbugs at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs


From isanbard at gmail.com  Sat Mar  3 04:56:36 2007
From: isanbard at gmail.com (Bill Wendling)
Date: Sat, 3 Mar 2007 02:56:36 -0800
Subject: [LLVMbugs] Segfault during bootstrap
In-Reply-To: <1172889460.17762.51.camel@bashful.x10sys.com>
References: <16e5fdf90703021720x239b5ddeg468665f07fa75d89@mail.gmail.com>
	<FCF2D631-7276-46BF-A62C-C014A8DC1029@apple.com>
	<1172889460.17762.51.camel@bashful.x10sys.com>
Message-ID: <F50D5144-DF11-4EAD-83D9-23918B60AB81@gmail.com>

My buildit compleated finally without error. I guess it was  
transitory...

-bw

On Mar 2, 2007, at 6:37 PM, Reid Spencer wrote:

> My bootstrap on Linux completed successfully.
>
> Reid.
>
> On Fri, 2007-03-02 at 17:52 -0800, Devang Patel wrote:
>>
>> On Mar 2, 2007, at 5:20 PM, Bill Wendling wrote:
>>
>>> powerpc-apple-darwin9-c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings
>>
>>
>> <snip>
>>
>>> -I/Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/ 
>>> include
>>> -o libllvmgcc.dylib \
>>>
>>>   /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/ 
>>> llvm-linker-hack.cpp
>>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/obj-llvm/ 
>>> Release/lib/libLLVMBCReader.a
>>
>>
>> <snip>
>>
>>
>>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/ 
>>> llvm-linker-hack.cpp:
>>> In function 'void dummy_function()':
>>> /Users/admin/LLVM/llvm-1033-03.roots/llvm-1033-03~obj/src/gcc/ 
>>> llvm-linker-hack.cpp:41:
>>> internal compiler error: Segmentation fault
>>
>>
>> powerpc-apple-darwin9-c++ is crashing. Is it part of llvm-gcc or  
>> plain
>> gcc ?
>> -
>> Devang
>>
>>> Please submit a full bug report,
>>> with preprocessed source if appropriate.
>>> See <URL:http://llvm.org/bugs> for instructions.
>>> make[2]: *** [libllvmgcc.dylib] Error 1
>>> make[2]: *** Waiting for unfinished jobs....
>>
>> _______________________________________________
>> LLVMbugs mailing list
>> LLVMbugs at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
>


From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 12:02:31 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 12:02:31 -0600
Subject: [LLVMbugs] [Bug 1238] NEW: assertion compiling debug code for
	ppc/linux
Message-ID: <200703031802.l23I2VHm024138@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1238

           Summary: assertion compiling debug code for ppc/linux
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: PowerPC
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicholas at mxc.ca


I encountered an assertion in when compiling llvm-gcc:

$ llc -mtriple=powerpc-gnu-linux bugpoint-reduced-simplified.bc
llc: LegalizeDAG.cpp:833:
llvm::SDOperand<unnamed>::SelectionDAGLegalize::LegalizeOp(llvm::SDOperand):
Assertion `0 && "This action is not supported yet!"' failed.
llvm/Debug/bin/llc((anonymous namespace)::PrintStackTrace()+0x1a)[0x886b354]
llvm/Debug/bin/llc((anonymous namespace)::SignalHandler(int)+0x110)[0x886b618]
[0xffffe500]
Aborted



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 12:14:06 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 12:14:06 -0600
Subject: [LLVMbugs] [Bug 1064] ppc linux doesn't handle float parameter to
	varargs function
Message-ID: <200703031814.l23IE6IZ024353@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1064

nicholas at mxc.ca changed:

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



------- Additional Comments From nicholas at mxc.ca  2007-03-03 12:14 -------
This works now, presumably through the many Linux/PPC patches by Nicolas Geoffray.




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 13:15:30 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 13:15:30 -0600
Subject: [LLVMbugs] [Bug 1233] multi-dimensional VLA indicies not scaled by
	llvm-gcc
Message-ID: <200703031915.l23JFUg4025400@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1233

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-03 13:15 -------
This still doesn't work on i686-pc-linux-gnu



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 13:23:59 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 13:23:59 -0600
Subject: [LLVMbugs] [Bug 1238] assertion compiling debug code for ppc/linux
Message-ID: <200703031923.l23JNxPJ025633@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1238

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
           Keywords|                            |compile-fail
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0



------- Additional Comments From sabre at nondot.org  2007-03-03 13:23 -------
Fixed, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070226/045464.html

However, you may now get undefined label errors when assembling the resultant .s file.  If so, go into 
PPCISelLowering.cpp, and delete this:

  if (!TM.getSubtarget<PPCSubtarget>().isDarwin()) {
    setOperationAction(ISD::LABEL, MVT::Other, Expand);

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 17:31:59 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 17:31:59 -0600
Subject: [LLVMbugs] [Bug 1239] NEW: ICE compiling LLVM with LLVM-GCC
Message-ID: <200703032331.l23NVxPl029786@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1239

           Summary: ICE compiling LLVM with LLVM-GCC
           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: me22.ca at gmail.com


Compiling LLVM (cvs) using a bootstrapped LLVM-GCC (svn):

llvm-cvs/lib/Analysis $ llvm-g++ -I/home/me22/programming/llvm-cvs/lib/Analysis
-I/home/me22/programming/llvm-cvs/lib/Analysis
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include  -D_GNU_SOURCE -D__STDC_LIMIT_MACROS
-O3 -fomit-frame-pointer -fno-exceptions -D_DEBUG  -Woverloaded-virtual
-pedantic -Wall -W -Wwrite-strings -Wno-long-long -Wunused -Wno-unused-parameter
 -c -MD -MT
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.o -MP -MF
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.LACXXd
ConstantFolding.cpp -o
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.o
cc1plus: ScheduleDAG.cpp:378: void
llvm::ScheduleDAG::AddOperand(llvm::MachineInstr*, llvm::SDOperand, unsigned
int, const llvm::TargetInstrDescriptor*, llvm::DenseMap<llvm::SDNode*, unsigned
int, llvm::DenseMapKeyInfo<llvm::SDNode*> >&): Assertion
`RegMap->getRegClass(VReg) == RC && "Register class of operand and regclass of
use don't agree!"' failed.
ConstantFolding.cpp:487: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://llvm.org/bugs> for instructions.

aKor suggesting trying with -D__NO_MATH_INLINES:

llvm-cvs/lib/Analysis $ llvm-g++ -I/home/me22/programming/llvm-cvs/lib/Analysis
-I/home/me22/programming/llvm-cvs/lib/Analysis
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include  -D_GNU_SOURCE -D__STDC_LIMIT_MACROS
-O3 -fomit-frame-pointer -fno-exceptions -D_DEBUG  -Woverloaded-virtual
-pedantic -Wall -W -Wwrite-strings -Wno-long-long -Wunused -Wno-unused-parameter
 -c -MD -MT
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.o -MP -MF
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.LACXXd
ConstantFolding.cpp -o
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.o
-D__NO_MATH_INLINES
NODE: 0x8eb6600: f32 = fpowi 0x8eb1f20, 0x8eb55b0
cc1plus: LegalizeDAG.cpp:3159:
llvm::SDOperand<unnamed>::SelectionDAGLegalize::PromoteOp(llvm::SDOperand):
Assertion `0 && "Do not know how to promote this operator!"' failed.
ConstantFolding.cpp:487: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://llvm.org/bugs> for instructions.

bytecode file attached, generated with:
llvm-g++ -I/home/me22/programming/llvm-cvs/lib/Analysis
-I/home/me22/programming/llvm-cvs/lib/Analysis
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include  -D_GNU_SOURCE -D__STDC_LIMIT_MACROS
-O3 -fomit-frame-pointer -fno-exceptions -D_DEBUG  -Woverloaded-virtual
-pedantic -Wall -W -Wwrite-strings -Wno-long-long -Wunused -Wno-unused-parameter
 -c -MD -MT
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.o -MP -MF
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.LACXXd
ConstantFolding.cpp --emit-llvm

recompiling with GCC to get a working llc & bugpoint to narrow it further...



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 17:44:29 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 17:44:29 -0600
Subject: [LLVMbugs] [Bug 1239] Cannot promote llvm.powi.f32
Message-ID: <200703032344.l23NiTjb030136@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1239

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
          Component|new bugs                    |Common Code Generator Code
           Keywords|                            |compile-fail
         OS/Version|Linux                       |All
            Product|new-bugs                    |libraries
           Platform|PC                          |All
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0
            Version|unspecified                 |1.9



------- Additional Comments From sabre at nondot.org  2007-03-03 17:44 -------
Fixed, testcase here: CodeGen/Generic/fpowi-promote.ll

Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070226/045469.html

BTW, you'll get much better code out of LLVM if you configure llvm-gcc to produce code for a target with 
SSE2.

Thanks,

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 19:02:33 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 19:02:33 -0600
Subject: [LLVMbugs] [Bug 1179] -indvars pass fails to replace exit value of
	loop when profitable
Message-ID: <200703040102.l2412XmV031689@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1179

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0



------- Additional Comments From sabre at nondot.org  2007-03-03 19:02 -------
Implemented.  Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070226/045475.html

Testcases here: test/Transforms/IndVarsSimplify/loop_evaluate_[234].ll

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 23:11:23 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 23:11:23 -0600
Subject: [LLVMbugs] [Bug 1239] Cannot promote llvm.powi.f32
Message-ID: <200703040511.l245BNYl008427@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1239

me22.ca at gmail.com changed:

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



------- Additional Comments From me22.ca at gmail.com  2007-03-03 23:11 -------
I'm still getting this, after updating and rebuilding llvm-gcc and llvm:
if llvm-g++ -I/home/me22/programming/llvm-cvs/lib/Analysis
-I/home/me22/programming/llvm-cvs/lib/Analysis
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include
-I/home/me22/programming/llvm-cvs/include  -D_GNU_SOURCE -D__STDC_LIMIT_MACROS
-O3 -fomit-frame-pointer -fno-exceptions -D_DEBUG -D__NO_MATH_INLINES
-Woverloaded-virtual -pedantic -Wall -W -Wwrite-strings -Wno-long-long -Wunused
-Wno-unused-parameter  -c -MD -MT
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.o -MP -MF
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.LACXXd
ConstantFolding.cpp -o
/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.o ;\
          then /usr/bin/mv -f
"/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.LACXXd"
"/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.d"; \
          else /usr/bin/rm -f
"/home/me22/programming/llvm-cvs/lib/Analysis/Release/ConstantFolding.LACXXd";
exit 1; fi
NODE: 0x8e431c0: f32 = fpowi 0x8bf21b0, 0x8e4b628
cc1plus: LegalizeDAG.cpp:3162:
llvm::SDOperand<unnamed>::SelectionDAGLegalize::PromoteOp(llvm::SDOperand):
Assertion `0 && "Do not know how to promote this operator!"' failed.
ConstantFolding.cpp:487: internal compiler error: Aborted




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar  3 23:25:11 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 Mar 2007 23:25:11 -0600
Subject: [LLVMbugs] [Bug 1239] Cannot promote llvm.powi.f32
Message-ID: <200703040525.l245PBA3009213@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1239

sabre at nondot.org changed:

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



------- Additional Comments From sabre at nondot.org  2007-03-03 23:25 -------
This really is fixed, please verify you have the patch.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar  4 00:37:42 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 Mar 2007 00:37:42 -0600
Subject: [LLVMbugs] [Bug 1240] NEW: -debug-only=jit emits unpleasent output
Message-ID: <200703040637.l246bg6J010994@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1240

           Summary: -debug-only=jit emits unpleasent output
           Product: libraries
           Version: 1.0
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Target-Independent JIT
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


I get stuff like:

JIT: Stub emitted at [0x7087fd8] for function 'test'
JIT: Finished CodeGen of [0x7088028] Function: main: 88 bytes of text, 9 relocations
Disassembled code:
No disassembler available. See configure help for avaiable options.
JIT: Lazily resolving function 'test' In stub ptr = 0x7087ff0 actual ptr = 0x7087fd8
JIT: Finished CodeGen of [0x7088090] Function: test: 12 bytes of text, 0 relocations
Disassembled code:
No disassembler available. See configure help for avaiable options.

It would be much better if the "Disassembled code/No disassembler available" lines just were not 
printed on targets without a disassembler.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From asl at math.spbu.ru  Sun Mar  4 04:15:38 2007
From: asl at math.spbu.ru (Anton Korobeynikov)
Date: Sun, 04 Mar 2007 13:15:38 +0300
Subject: [LLVMbugs] llvm-gcc is broken
Message-ID: <1173003338.12999.61.camel@asl.dorms.spbu.ru>

Hello, All.

I'm getting this on TOT:

libbackend.a(llvm-backend.o): In function `CreateStructorsList':
../../gccsrc/gcc/llvm-backend.cpp:422: undefined reference to
`llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
libbackend.a(llvm-convert.o): In function `CopyAggregate':
../../gccsrc/gcc/llvm-convert.cpp:1030: undefined reference to
`llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
../../gccsrc/gcc/llvm-convert.cpp:1032: undefined reference to
`llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
../../gccsrc/gcc/llvm-convert.cpp:1039: undefined reference to
`llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
../../gccsrc/gcc/llvm-convert.cpp:1041: undefined reference to
`llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
libbackend.a(llvm-convert.o):../../gccsrc/gcc/llvm-convert.cpp:1102:
more undefined references to `llvm::ConstantInt::get(llvm::Type const*,
unsigned long long)' follow

Reid, is it yours? :)

-- 
With best regards, Anton Korobeynikov.

Faculty of Mathematics & Mechanics, Saint Petersburg State University.



From rspencer at reidspencer.com  Sun Mar  4 08:55:28 2007
From: rspencer at reidspencer.com (Reid Spencer)
Date: Sun, 04 Mar 2007 06:55:28 -0800
Subject: [LLVMbugs] llvm-gcc is broken
In-Reply-To: <1173003338.12999.61.camel@asl.dorms.spbu.ru>
References: <1173003338.12999.61.camel@asl.dorms.spbu.ru>
Message-ID: <1173020128.17762.85.camel@bashful.x10sys.com>

Hi Anton,

On Sun, 2007-03-04 at 13:15 +0300, Anton Korobeynikov wrote:
> Hello, All.
> 
> I'm getting this on TOT:
> 
> libbackend.a(llvm-backend.o): In function `CreateStructorsList':
> ../../gccsrc/gcc/llvm-backend.cpp:422: undefined reference to
> `llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'

We haven't removed this method. Its still in my tree. Must be something
local. Try rebuilding libLLVMCore.

> libbackend.a(llvm-convert.o): In function `CopyAggregate':
> ../../gccsrc/gcc/llvm-convert.cpp:1030: undefined reference to
> `llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
> ../../gccsrc/gcc/llvm-convert.cpp:1032: undefined reference to
> `llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
> ../../gccsrc/gcc/llvm-convert.cpp:1039: undefined reference to
> `llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
> ../../gccsrc/gcc/llvm-convert.cpp:1041: undefined reference to
> `llvm::ConstantInt::get(llvm::Type const*, unsigned long long)'
> libbackend.a(llvm-convert.o):../../gccsrc/gcc/llvm-convert.cpp:1102:
> more undefined references to `llvm::ConstantInt::get(llvm::Type const*,
> unsigned long long)' follow
> 
> Reid, is it yours? :)

Don't think so.

Reid.

> 


From bugzilla-daemon at cs.uiuc.edu  Sun Mar  4 17:16:43 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 Mar 2007 17:16:43 -0600
Subject: [LLVMbugs] [Bug 1241] NEW: LLVM-GCC on PowerPC doesn't support the
	-mmmx option
Message-ID: <200703042316.l24NGhgp030505@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1241

           Summary: LLVM-GCC on PowerPC doesn't support the -mmmx option
           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: isanbard at gmail.com


For this program:

#include <mmintrin.h>

void foo()
{
  _mm_empty();
}

GCC compiles it like so:

$  gcc -arch i386 -mmmx -S t.c
$

But LLVM-GCC gives an error:

$ llvm-gcc -arch i386 -mmmx -S t.c
cc1: error: invalid option 'mmx'
$

This is almost certainly a front-end issue.

-bw



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar  4 17:17:43 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 Mar 2007 17:17:43 -0600
Subject: [LLVMbugs] [Bug 1241] LLVM-GCC on PowerPC doesn't support the -mmmx
	option
Message-ID: <200703042317.l24NHhdU030552@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1241

sabre at nondot.org changed:

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



------- Additional Comments From sabre at nondot.org  2007-03-04 17:17 -------
You need to build llvm-gcc "apple-style" to get it to understand "-arch".

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar  5 16:24:13 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Mar 2007 16:24:13 -0600
Subject: [LLVMbugs] [Bug 1186] Assertion failure in PassManager.cpp:643
Message-ID: <200703052224.l25MODFH003007@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1186

dpatel at apple.com changed:

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



------- Additional Comments From dpatel at apple.com  2007-03-05 16:24 -------
tst1 is not preserved by tst2. So after ts2, tst1 is removed from the list of available analysis. This is why 
PassManager assserts while adding tst3.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar  5 19:02:20 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Mar 2007 19:02:20 -0600
Subject: [LLVMbugs] [Bug 1242] NEW: llvm-gcc emits llvm assembly with
	incomplte target data layout
Message-ID: <200703060102.l2612KpV005942@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1242

           Summary: llvm-gcc emits llvm assembly with incomplte target data
                    layout
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


When emitting llvm code (-emit-llvm), llvm-gcc generates a Target Data Layout of
"e-p:32:32" for x86-pc-linux-gnu. This is insufficient for some programs as the
defaults will use f64:64:64 whereas the correct alignment is f64:32:64. This
causes a problem in accessing fields of structures that use double/long long.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar  5 19:09:45 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Mar 2007 19:09:45 -0600
Subject: [LLVMbugs] [Bug 1242] llvm-gcc emits llvm assembly with incomplete
	target data layout
Message-ID: <200703060109.l2619j2l006171@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1242

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
           Keywords|                            |miscompilation
         OS/Version|Linux                       |All
           Platform|PC                          |All
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0



------- Additional Comments From sabre at nondot.org  2007-03-05 19:09 -------
Fixed, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045573.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar  5 23:34:06 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 Mar 2007 23:34:06 -0600
Subject: [LLVMbugs] [Bug 1240] -debug-only=jit emits unpleasent output
Message-ID: <200703060534.l265Y69J011265@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1240

asl at math.spbu.ru changed:

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



------- Additional Comments From asl at math.spbu.ru  2007-03-05 23:34 -------
Fixed with

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045594.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045595.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045596.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From isanbard at gmail.com  Tue Mar  6 01:17:10 2007
From: isanbard at gmail.com (Bill Wendling)
Date: Mon, 5 Mar 2007 23:17:10 -0800
Subject: [LLVMbugs] Build broke again
Message-ID: <36D68C24-7B3A-44F1-B815-2AD9C02955BE@gmail.com>

I'm getting this:

llvm[2]: Compiling fibonacci.cpp for Debug build
/Users/wendling/llvm/llvm.src/examples/Fibonacci/fibonacci.cpp: In  
function 'int main(int, char**)':
/Users/wendling/llvm/llvm.src/examples/Fibonacci/fibonacci.cpp:115:  
error: 'struct llvm::GenericValue' has no member named 'Int32Val'
/Users/wendling/llvm/llvm.src/examples/Fibonacci/fibonacci.cpp:119:  
error: 'struct llvm::GenericValue' has no member named 'Int32Val'
make[2]: *** [/Users/wendling/llvm/llvm.obj/examples/Fibonacci/Debug/ 
fibonacci.o] Error 1
make[1]: *** [Fibonacci/.makeall] Error 2
make: *** [all] Error 1


-bw

From sabre at nondot.org  Tue Mar  6 02:37:53 2007
From: sabre at nondot.org (Chris Lattner)
Date: Tue, 6 Mar 2007 00:37:53 -0800 (PST)
Subject: [LLVMbugs] Build broke again
In-Reply-To: <36D68C24-7B3A-44F1-B815-2AD9C02955BE@gmail.com>
References: <36D68C24-7B3A-44F1-B815-2AD9C02955BE@gmail.com>
Message-ID: <Pine.LNX.4.62.0703060037200.19751@nondot.org>

On Mon, 5 Mar 2007, Bill Wendling wrote:
> I'm getting this:
> llvm[2]: Compiling fibonacci.cpp for Debug build
> /Users/wendling/llvm/llvm.src/examples/Fibonacci/fibonacci.cpp:119:
> error: 'struct llvm::GenericValue' has no member named 'Int32Val'
> make[2]: *** [/Users/wendling/llvm/llvm.obj/examples/Fibonacci/Debug/

I temporarily disabled building of the examples until Reid has a chance to 
fix the problem.  Reid, are you building from the top of the tree?

-Chris

-- 
http://nondot.org/sabre/
http://llvm.org/

From rspencer at reidspencer.com  Tue Mar  6 11:27:04 2007
From: rspencer at reidspencer.com (Reid Spencer)
Date: Tue, 06 Mar 2007 09:27:04 -0800
Subject: [LLVMbugs] Build broke again
In-Reply-To: <Pine.LNX.4.62.0703060037200.19751@nondot.org>
References: <36D68C24-7B3A-44F1-B815-2AD9C02955BE@gmail.com>
	<Pine.LNX.4.62.0703060037200.19751@nondot.org>
Message-ID: <1173202024.24497.143.camel@bashful.x10sys.com>

On Tue, 2007-03-06 at 00:37 -0800, Chris Lattner wrote:
> On Mon, 5 Mar 2007, Bill Wendling wrote:
> > I'm getting this:
> > llvm[2]: Compiling fibonacci.cpp for Debug build
> > /Users/wendling/llvm/llvm.src/examples/Fibonacci/fibonacci.cpp:119:
> > error: 'struct llvm::GenericValue' has no member named 'Int32Val'
> > make[2]: *** [/Users/wendling/llvm/llvm.obj/examples/Fibonacci/Debug/
> 
> I temporarily disabled building of the examples until Reid has a chance to 
> fix the problem.  Reid, are you building from the top of the tree?

Sorry about that guys. My standard build alias is for tools-only. I'll
try to be more careful. Its all fixed now.

Reid.

> 
> -Chris
> 


From bugzilla-daemon at cs.uiuc.edu  Tue Mar  6 13:47:41 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Mar 2007 13:47:41 -0600
Subject: [LLVMbugs] [Bug 1243] NEW: llvm-upgrade issues bogus warnings about
	renaming internal symbols
Message-ID: <200703061947.l26JlfYj004861@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1243

           Summary: llvm-upgrade issues bogus warnings about renaming
                    internal symbols
           Product: tools
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-upgrade
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


While running this test: 
  test/ExecutionEngine/2003-05-06-LivenessClobber.llx
llvm-upgrade issues this warning:
<stdin>:9: warning: Renaming function 'test' as 'test.upgrd.1' may cause linkage
errors while reading token '{'.

It really shouldn't as the function is internal and renaming it cannot cause any
harm. Before issuing the warning, a check should be made to see if its internal
and the warning squelched if it is.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar  6 20:49:42 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 Mar 2007 20:49:42 -0600
Subject: [LLVMbugs] [Bug 978] X86/ELF Dwarf Support Is Lacking
Message-ID: <200703070249.l272ng5E012868@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=978

asl at math.spbu.ru changed:

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



------- Additional Comments From asl at math.spbu.ru  2007-03-06 20:49 -------
Fixed with

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045671.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045673.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045672.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045674.html

Probably, the same should be done for ppc/Linux. I'm waiting for Nicolas's response.

Reopen in case of any bad things.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar  8 21:05:58 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 Mar 2007 21:05:58 -0600
Subject: [LLVMbugs] [Bug 1244] NEW: Problem in instructioncombining
Message-ID: <200703090305.l2935wex005957@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1244

           Summary: Problem in instructioncombining
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: critical
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: wjiang at cs.ucla.edu


hi, I found a bug in TOT (instuction combiner). It tried to optimize (a<b)||
(a=b) to (a<=b), but the new predicate is wrong. 
The original code:
   %tmp266.i = icmp slt i32 %c.3.i, %d.292.2.i     
   %tmp276.i = icmp eq i32 %c.3.i, %d.292.2.i 
   %sel_tmp80 = or i1 %sel_tmp78, %sel_tmp79 
New code: 
   %sel_tmp187.demorgan.i = icmp ule i32 %c.3.i, %d.292.2.i 
                           --(should be sle)--

I looked the source codes, I guess the bug is in these lines:
struct FoldICmpLogical {
    Instruction *apply(Instruction &Log) const {
    Value *RV = getICmpValue(ICmpInst::isSignedPredicate(pred), 
           Code, LHS, RHS);
          ....
    }
}
It only looks the second predicate: pred = eq for this example, so it return 
non-signed, but actually it's signed because the other preciate is slt. I guess 
both predicates should be looked at to get the correct results. I changed the 
code locally, it works
   Value *RV = getICmpValue(
      ICmpInst::isSignedPredicate(ICI->getPredicate())
   || ICmpInst::isSignedPredicate(cast<ICmpInst>(Log.getOperand(1))-
>getPredicate()),
      Code, LHS, RHS);

Thanks.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  9 05:31:43 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Mar 2007 05:31:43 -0600
Subject: [LLVMbugs] [Bug 1245] NEW: Invalid bytecode assembled
Message-ID: <200703091131.l29BVh8C015649@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1245

           Summary: Invalid bytecode assembled
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-as
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


Consider the attached .ll file.

Do:
llvm-as foo.ll -o - | ./llvm-dif -o foo-cmp.ll

I have (x86-linux):

....
-       %ov35 = add i32 %ov16, -29              ; <i32> [#uses=1]
+       %ov35 = add i32 %ov16, -93              ; <i32> [#uses=1]
        store i32 %ov35, i32* %state
        %pcc8 = add i32 %pcv7, 1                ; <i32> [#uses=1]
        store i32 %tmp20, i32* %tmp20_c
@@ -321,7 +321,7 @@
        %ov7 = xor i32 %ov18_cc, 251            ; <i32> [#uses=3]
        %ov41 = xor i32 %ov7, 116               ; <i32> [#uses=1]
        %ov63 = sub i32 139, %ov7               ; <i32> [#uses=2]
-       %hvar68 = add i32 %ov63, -1             ; <i32> [#uses=1]
+       %hvar68 = add i32 %ov63, -9             ; <i32> [#uses=1]

and so on. Valgrind is quite. It's definitely an assembler bug, since output
binary code is broken.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From asl at math.spbu.ru  Fri Mar  9 06:53:52 2007
From: asl at math.spbu.ru (Anton Korobeynikov)
Date: Fri, 09 Mar 2007 15:53:52 +0300
Subject: [LLVMbugs] Warnings
Message-ID: <1173444832.12999.168.camel@asl.dorms.spbu.ru>

Hello, Everyone.

I'm getting these warnings during fresh build:

1. 
/home/asl/proj/llvm/src/lib/ExecutionEngine/Interpreter/Execution.cpp:
In member function `llvm::GenericValue
llvm::Interpreter::executeIntToPtrInst(llvm::Value*, const llvm::Type*,
llvm::ExecutionContext&)':
/home/asl/proj/llvm/src/lib/ExecutionEngine/Interpreter/Execution.cpp:1065: warning: cast to pointer from integer of different size
2.
/home/asl/proj/llvm/src/lib/Transforms/Scalar/InstructionCombining.cpp:981: warning: 'bool MaskedValueIsZero(llvm::Value*, const llvm::APInt&, unsigned int)' defined but not used

-- 
With best regards, Anton Korobeynikov.

Faculty of Mathematics & Mechanics, Saint Petersburg State University.



From bugzilla-daemon at cs.uiuc.edu  Fri Mar  9 07:07:12 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Mar 2007 07:07:12 -0600
Subject: [LLVMbugs] [Bug 1246] NEW: Bug in LoopStrengthReduce.
Message-ID: <200703091307.l29D7CBn017418@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1246

           Summary: Bug in LoopStrengthReduce.
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


Consider the attached bytecode. Be careful with re-assembling - you can hit PR1245.

Everything is ok with "llc -fast" or with disabled LoopStrengthReduce pass (or
cbe), but segfault with enabled.

Segfault occurs in bb43 BB. The variable %i1.0_cc becames uninitialized.
Initialization occurs in the meshBB BB, but it seems, that LSR incorrectly moves
it somewhere. 

Target triple is, unfortunately, set to mingw32, but bytecode should run ok any
ia32 platform, it doesn't contain any target-specific code. Unfortunately, until
PR1245 is fixed, I cannot change it :(

Please also note, that "opt -std-compile-opts" is buggy on this bytecode, but
it's subject for future investigation (maybe due to PR1245).



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  9 11:28:54 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Mar 2007 11:28:54 -0600
Subject: [LLVMbugs] [Bug 1247] NEW: build failure on darwin 8.8.1
Message-ID: <200703091728.l29HSsmc022060@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1247

           Summary: build failure on darwin 8.8.1
           Product: tools
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: ebner at complang.tuwien.ac.at


after upgrading of llvm-gcc to svn-head (revision 290), i see the following
build failure:

build/genattrtab /Users/ebner/dev/opt/llvm-cvs/gcc/config/i386/i386.md >
tmp-attrtab.c
genattrtab: Automaton `ppro_load': Insn `ppro_fmov_XF_load' will never be issued
genattrtab: Automaton `ppro_load': Insn `ppro_sse_icvt_SF' will never be issued
genattrtab: Automaton `ppro_load': Insn `ppro_sse_add_V4SF_load' will never be
issued
genattrtab: Automaton `ppro_load': Insn `ppro_sse_cmp_V4SF_load' will never be
issued
genattrtab: Automaton `ppro_load': Insn `ppro_sse_mul_V4SF_load' will never be
issued
genattrtab: Automaton `ppro_load': Insn `ppro_sse_div_V4SF_load' will never be
issued
genattrtab: Automaton `ppro_load': Insn `ppro_sse_mov_V4SF_load' will never be
issued
genattrtab: Automaton `ppro_store': Insn `ppro_fmov_XF_store' will never be issued
genattrtab: Automaton `k6_integer_units': Insn `k6_alux_only_load' will never be
issued
genattrtab: Automaton `k6_integer_units': Insn `k6_alux_only_store' will never
be issued
make[1]: *** [s-attrtab] Error 1

configure line: /Users/ebner/dev/opt/llvm-cvs/configure  
--prefix=/Users/ebner/dev/install/ --program-prefix=llvm-
--enable-llvm=/Users/ebner/dev/build/llvm/ --enable-checking --disable-multilib
--with-gxx-include-dir=/usr/include/c++/4.0.0 --build=i686-apple-darwin8
--host=i686-apple-darwin8 --target=i686-apple-darwin8
--with-gcc-version-trigger=/Users/ebner/dev/llvm-gcc4/gcc/version.c
--enable-languages=c,c++

$ uname -a
Darwin cdbook2 8.8.1 Darwin Kernel Version 8.8.1: Mon Sep 25 19:42:00 PDT 2006;
root:xnu-792.13.8.obj~1/RELEASE_I386 i386 i386 MacBookPro1,1 Darwin

$ gcc ---version
i686-apple-darwin8-gcc-4.0.1 (GCC) 4.0.1 LLVM (Apple Computer, Inc. build 5421)
Copyright (C) 2005 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar  9 13:30:43 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 Mar 2007 13:30:43 -0600
Subject: [LLVMbugs] [Bug 1245] llvm-as mis-reads some negative integers
Message-ID: <200703091930.l29JUhcu024181@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1245

rspencer at x10sys.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED
            Summary|Invalid bytecode assembled  |llvm-as mis-reads some
                   |                            |negative integers



------- Additional Comments From rspencer at x10sys.com  2007-03-09 13:30 -------
This problem only affected negative integers whose "next to last" significant
bit was set. In such cases the number of bits computed as needed by the value
was off-by-one as it did not account for the sign bit. Consequently, the number
was parsed correctly by the "fromString" constructor but truncated to the wrong
length which changed the value.

This simple patch fixes it:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070305/045759.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 10 12:25:05 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 Mar 2007 12:25:05 -0600
Subject: [LLVMbugs] [Bug 1248] NEW: Another InstCombine issue
Message-ID: <200703101825.l2AIP5Ad014420@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1248

           Summary: Another InstCombine issue
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


Consider the attached bytecode. This is just -O0 compiled bytecode from gcc
testsuite.

Running "opt -instcombine" will lead to "eq" comparison turn to "slt", which is
definitely wrong.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 10 12:53:05 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 Mar 2007 12:53:05 -0600
Subject: [LLVMbugs] [Bug 1249] NEW: Invalid transformation for vaargs
	functions.
Message-ID: <200703101853.l2AIr5ce014898@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1249

           Summary: Invalid transformation for vaargs functions.
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


Consider the attached C source (from gcc testsuite).

The call to function f is converted to LLVM IR incorrectly: i8/16 args will be
extended to 32 bits, but args handling logic in f() itself assumes, that they
are all in place.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 10 13:02:17 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 Mar 2007 13:02:17 -0600
Subject: [LLVMbugs] [Bug 1250] NEW: Invalid conversion of mode attribute.
Message-ID: <200703101902.l2AJ2Htg015124@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1250

           Summary: Invalid conversion of mode attribute.
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


Consider the attached C source (from gcc testsuite).

llvm-gcc turns type of struct S to:

%struct.S = type { i8, i8, i8, [4 x i8], [18446744073709551613 x i8] }

I don't think, it's correct. Valgrind is quite here.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 10 13:02:18 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 Mar 2007 13:02:18 -0600
Subject: [LLVMbugs] [Bug 1251] NEW: Invalid conversion of mode attribute.
Message-ID: <200703101902.l2AJ2ITB015140@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1251

           Summary: Invalid conversion of mode attribute.
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


Consider the attached C source (from gcc testsuite).

llvm-gcc turns type of struct S to:

%struct.S = type { i8, i8, i8, [4 x i8], [18446744073709551613 x i8] }

I don't think, it's correct. Valgrind is quite here.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 10 13:08:17 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 Mar 2007 13:08:17 -0600
Subject: [LLVMbugs] [Bug 1251] Invalid conversion of mode attribute.
Message-ID: <200703101908.l2AJ8HBM015358@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1251

asl at math.spbu.ru changed:

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



------- Additional Comments From asl at math.spbu.ru  2007-03-10 13:08 -------
Sorry, dup



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 10 15:23:21 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 Mar 2007 15:23:21 -0600
Subject: [LLVMbugs] [Bug 1252] NEW: Read-after-free bug in llvm-gcc
Message-ID: <200703102123.l2ALNLXD018077@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1252

           Summary: Read-after-free bug in llvm-gcc
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


Consider attached .c file. It causes assertion at -O3 optimization level.
Assertion was due to uninitialized read in EmitLV_DECL() routine (line numbers
may be incorrect):

==6803==    at 0x834B8B8: TreeToLLVM::EmitLV_DECL(tree_node*) (Value.h:190)
==6803==    by 0x836B37F: TreeToLLVM::EmitLV(tree_node*) (llvm-convert.cpp:846)
==6803==    by 0x8370EE2: TreeToLLVM::EmitADDR_EXPR(tree_node*)
(llvm-convert.cpp:2280)
==6803==    by 0x835A207: TreeToLLVM::Emit(tree_node*, llvm::Value*)
(llvm-convert.cpp:702)
==6803==    by 0x8359A54: TreeToLLVM::EmitCALL_EXPR(tree_node*, llvm::Value*)
(llvm-convert.cpp:2305)
==6803==    by 0x835A40A: TreeToLLVM::Emit(tree_node*, llvm::Value*)
(llvm-convert.cpp:703)
==6803==    by 0x836EE9F: TreeToLLVM::EmitMODIFY_EXPR(tree_node*, llvm::Value*)
(llvm-convert.cpp:2673)
==6803==    by 0x835A686: TreeToLLVM::Emit(tree_node*, llvm::Value*)
(llvm-convert.cpp:704)
==6803==    by 0x83684D5: TreeToLLVM::EmitSTATEMENT_LIST(tree_node*,
llvm::Value*) (llvm-convert.cpp:1464)
==6803==    by 0x835A4FE: TreeToLLVM::Emit(tree_node*, llvm::Value*)
(llvm-convert.cpp:673)
==6803==    by 0x8368F03: TreeToLLVM::EmitBIND_EXPR(tree_node*, llvm::Value*)
(llvm-convert.cpp:1439)
==6803==    by 0x835A3EC: TreeToLLVM::Emit(tree_node*, llvm::Value*)
(llvm-convert.cpp:672)
==6803==  Address 0x43386A4 is 4 bytes inside a block of size 100 free'd
==6803==    at 0x402167A: operator delete(void*) (in
/usr/lib/valgrind/x86-linux/vgpreload_memcheck.so)
==6803==    by 0x88629B0: llvm::Function::~Function() (Function.cpp:111)
==6803==    by 0x84F605A: llvm::iplist<llvm::Function,
llvm::ilist_traits<llvm::Function>
>::erase(llvm::ilist_iterator<llvm::Function>) (ilist:323)
==6803==    by 0x8862AE3: llvm::Function::eraseFromParent() (Function.cpp:138)
==6803==    by 0x83694B8: TreeToLLVM::StartFunctionBody() (llvm-convert.cpp:470)
==6803==    by 0x83425BC: llvm_emit_code_for_current_function (llvm-backend.cpp:501)
==6803==    by 0x80DDFE8: tree_rest_of_compilation (in
/home/asl/proj/llvm/install_debug/libexec/gcc/i686-pc-linux-gnu/4.0.1/cc1)

The statement in question is "  if (GlobalValue *GV =
dyn_cast<GlobalValue>(Decl)) {" line. It seems, that Decl, returned by DECL_LLVM
is dead somehow...



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 12 16:29:40 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Mar 2007 16:29:40 -0500
Subject: [LLVMbugs] [Bug 1253] NEW: "optimized" range comparison not
	simplified
Message-ID: <200703122129.l2CLTe9D029216@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1253

           Summary: "optimized" range comparison not simplified
           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


The condition a<=x<=b can be transformed into (unsigned)n-a<=b-a.
However if a=INT_MIN then this "optimization" leads to suboptimal
code:

#include <limits.h>
int bad (int n) {
  return ((unsigned)n-INT_MIN<=-1-INT_MIN);
}

->

define i32 @bad(i32 %n) {
entry:
        %tmp2 = xor i32 %n, -2147483648         ; <i32> [#uses=1]
        icmp sgt i32 %tmp2, -1          ; <i1>:0 [#uses=1]
        zext i1 %0 to i32               ; <i32>:0 [#uses=1]
        ret i32 %0
}

This would be better as: "return n<0;"



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 12 18:24:51 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 Mar 2007 18:24:51 -0500
Subject: [LLVMbugs] [Bug 1254] NEW: CBE handling of inline asm immediate
	operand broken
Message-ID: <200703122324.l2CNOpsF031023@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1254

           Summary: CBE handling of inline asm immediate operand broken
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: major
          Priority: P2
         Component: Backend: C
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: evan.cheng at apple.com


define void @test4(i16 %S) {
       call i16 asm "rolw   $$8, ${0:w}", "=r,0,~{dirflag},~{fpsr},~{flags}"(
i16 %S )
       ret void
}

CBE produces:

void test4(unsigned short llvm_cbe_S) {
  unsigned short ltmp_0_1;

  __asm__ volatile ("rolw   %%8, %w0"
        :"=r"(ltmp_0_1)
        :"0"(llvm_cbe_S));
  return;
}

gcc compiles inline asm statement to:
rolw %8, %ax

It should have been rolw $8, %ax.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 13 09:28:34 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Mar 2007 09:28:34 -0500
Subject: [LLVMbugs] [Bug 1244] InstructionCombining miscompiles (x < y) ||
	(x == y)
Message-ID: <200703131428.l2DESYOK015931@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1244

sabre at nondot.org changed:

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



------- Additional Comments From sabre at nondot.org  2007-03-13 09:28 -------
This is an obvious bug from the signless types change.

Testcase here:
Transforms/InstCombine/2007-03-13-CompareMerge.ll

Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070312/045812.html

Thank you for expertly reducing this down!

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From sabre at nondot.org  Tue Mar 13 11:15:48 2007
From: sabre at nondot.org (Chris Lattner)
Date: Tue, 13 Mar 2007 08:15:48 -0800 (PST)
Subject: [LLVMbugs] Warnings
In-Reply-To: <1173444832.12999.168.camel@asl.dorms.spbu.ru>
References: <1173444832.12999.168.camel@asl.dorms.spbu.ru>
Message-ID: <Pine.LNX.4.62.0703130814270.17545@nondot.org>

On Fri, 9 Mar 2007, Anton Korobeynikov wrote:

> Hello, Everyone.
>
> I'm getting these warnings during fresh build:
>
> 1.
> /home/asl/proj/llvm/src/lib/ExecutionEngine/Interpreter/Execution.cpp:
> In member function `llvm::GenericValue
> llvm::Interpreter::executeIntToPtrInst(llvm::Value*, const llvm::Type*,
> llvm::ExecutionContext&)':
> /home/asl/proj/llvm/src/lib/ExecutionEngine/Interpreter/Execution.cpp:1065: warning: cast to pointer from integer of different size

I don't see this one anymore.  If it still exists, plz ping reid.

> 2.
> /home/asl/proj/llvm/src/lib/Transforms/Scalar/InstructionCombining.cpp:981: warning: 'bool MaskedValueIsZero(llvm::Value*, const llvm::APInt&, unsigned int)' defined but not used

I just fixed this one.  Reid/Sheng, plz be careful that you don't 
introduce warnings into the build.

-Chris

-- 
http://nondot.org/sabre/
http://llvm.org/

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 13 10:19:46 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Mar 2007 10:19:46 -0500
Subject: [LLVMbugs] [Bug 1233] multi-dimensional VLA indicies not scaled by
	llvm-gcc
Message-ID: <200703131519.l2DFJkts018225@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1233

sabre at nondot.org changed:

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



------- Additional Comments From sabre at nondot.org  2007-03-13 10:19 -------
Sounds good, I will assume it was just something transient in Reid's tree. 



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 13 10:31:48 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 Mar 2007 10:31:48 -0500
Subject: [LLVMbugs] [Bug 1255] NEW: Should enhance LLVM switch instruction
	to take case ranges
Message-ID: <200703131531.l2DFVm9l018451@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1255

           Summary: Should enhance LLVM switch instruction to take case
                    ranges
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


The LLVM switch instruction should handle ranges of case values that go to the same destination without 
enumerating all of them.

This would be good to do before 2.0, and would be a great project for someone who wants to learn/touch 
more of LLVM.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 14 08:18:01 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 Mar 2007 08:18:01 -0500
Subject: [LLVMbugs] [Bug 1256] NEW: llvm-upgrade cannot distinguish between
	int tmp & uint tmp etc.
Message-ID: <200703141318.l2EDI1Ut012040@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1256

           Summary: llvm-upgrade cannot distinguish between int tmp & uint
                    tmp etc.
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: major
          Priority: P2
         Component: llvm-upgrade
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: patil.nikhil at gmail.com


(Mis-upgradation. No warning issued.)

The following testcase should be self-explanatory... llvm-upgrade can't pick 
the correct %tmp when both int %tmp and uint %tmp are defined. Similar problem 
occurs with sbyte and ubyte.

Testcase: zxc.ll:
target datalayout = "e-p:32:32"
target endian = little
target pointersize = 32
target triple = "i686-pc-linux-gnu"

implementation   ; Functions:

void %_Z4func() {
entry:
        %tmp = add int 0, 0
        %tmp = add uint 1, 1
        %tmp = add ubyte 1, 2
        %tmp = add sbyte 2, 3
        call void %f (int %tmp)
        call void %g (ubyte %tmp)
        ret void
}

declare void %f(int)
declare void %g(ubyte)


Output of $llvm-upgrade zxc.ll
; ModuleID = 'zxc.ll'
target datalayout = "e-p:32:32"
target triple = "i686-pc-linux-gnu"

implementation   ; Functions:

define void @_Z4func() {
entry:
        %tmp = add i32 0, 0             ; <i32> [#uses=0]
        %tmp.upgrd.1 = add i32 1, 1             ; <i32> [#uses=1]
        %tmp.upgrd.2 = add i8 1, 2              ; <i8> [#uses=0]
        %tmp.upgrd.3 = add i8 2, 3              ; <i8> [#uses=1]
        call void @f( i32 %tmp.upgrd.1 )
        call void @g( i8 %tmp.upgrd.3 )
        ret void
}

declare void @f(i32)

declare void @g(i8)


Expected output:

; ModuleID = 'zxc.ll'
target datalayout = "e-p:32:32"
target triple = "i686-pc-linux-gnu"

implementation   ; Functions:

define void @_Z4func() {
entry:
        %tmp = add i32 0, 0             ; <i32> [#uses=0]
        %tmp.upgrd.1 = add i32 1, 1             ; <i32> [#uses=1]
        %tmp.upgrd.2 = add i8 1, 2              ; <i8> [#uses=0]
        %tmp.upgrd.3 = add i8 2, 3              ; <i8> [#uses=1]
        call void @f( i32 %tmp )         ; *XXX*
        call void @g( i8 %tmp.upgrd.2 )  ; *XXX*
        ret void
}

declare void @f(i32)

declare void @g(i8)



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From asl at math.spbu.ru  Wed Mar 14 16:22:03 2007
From: asl at math.spbu.ru (Anton Korobeynikov)
Date: Thu, 15 Mar 2007 00:22:03 +0300
Subject: [LLVMbugs] Warnings
In-Reply-To: <Pine.LNX.4.62.0703130814270.17545.SS888SS@nondot.org>
References: <1173444832.12999.168.camel@asl.dorms.spbu.ru>
	<Pine.LNX.4.62.0703130814270.17545.SS888SS@nondot.org>
Message-ID: <1173907323.23851.73.camel@asl.dorms.spbu.ru>

Hello, Chris.

> > 1.
> > /home/asl/proj/llvm/src/lib/ExecutionEngine/Interpreter/Execution.cpp:
> > In member function `llvm::GenericValue
> > llvm::Interpreter::executeIntToPtrInst(llvm::Value*, const llvm::Type*,
> > llvm::ExecutionContext&)':
> > /home/asl/proj/llvm/src/lib/ExecutionEngine/Interpreter/Execution.cpp:1065: warning: cast to pointer from integer of different size
> 
> I don't see this one anymore.  If it still exists, plz ping reid.
It's still here.

-- 
With best regards, Anton Korobeynikov.

Faculty of Mathematics & Mechanics, Saint Petersburg State University.



From rspencer at reidspencer.com  Wed Mar 14 16:44:40 2007
From: rspencer at reidspencer.com (Reid Spencer)
Date: Wed, 14 Mar 2007 14:44:40 -0700
Subject: [LLVMbugs] Warnings
In-Reply-To: <1173907323.23851.73.camel@asl.dorms.spbu.ru>
References: <1173444832.12999.168.camel@asl.dorms.spbu.ru>
	<Pine.LNX.4.62.0703130814270.17545.SS888SS@nondot.org>
	<1173907323.23851.73.camel@asl.dorms.spbu.ru>
Message-ID: <1173908680.13912.149.camel@bashful.x10sys.com>

On Thu, 2007-03-15 at 00:22 +0300, Anton Korobeynikov wrote:
> Hello, Chris.
> 
> > > 1.
> > > /home/asl/proj/llvm/src/lib/ExecutionEngine/Interpreter/Execution.cpp:
> > > In member function `llvm::GenericValue
> > > llvm::Interpreter::executeIntToPtrInst(llvm::Value*, const llvm::Type*,
> > > llvm::ExecutionContext&)':
> > > /home/asl/proj/llvm/src/lib/ExecutionEngine/Interpreter/Execution.cpp:1065: warning: cast to pointer from integer of different size
> > 
> > I don't see this one anymore.  If it still exists, plz ping reid.
> It's still here.

I don't see it either. lib/ExecutionEngine build is clean. :)

Reid

> 


From bugzilla-daemon at cs.uiuc.edu  Fri Mar 16 15:57:13 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 Mar 2007 15:57:13 -0500
Subject: [LLVMbugs] [Bug 1257] NEW: llc crashes when it compiles the
	attached file
Message-ID: <200703162057.l2GKvDYs031848@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1257

           Summary: llc crashes when it compiles the attached file
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lauro.venancio at gmail.com


laurov at laurov-desktop:/tmp$ llc < bugpoint-reduced-simplified.bc 
llc((anonymous namespace)::PrintStackTrace()+0x1a)[0x88b3a1a]
llc((anonymous namespace)::SignalHandler(int)+0x112)[0x88b3ce0]
[0xffffe420]
llc(llvm::LiveIntervals::JoinCopy(llvm::MachineInstr*, unsigned int, unsigned
int)+0x4d7)[0x8739267]
llc(llvm::LiveIntervals::joinIntervals()+0x3c0)[0x8739af6]
llc(llvm::LiveIntervals::runOnMachineFunction(llvm::MachineFunction&)+0x416)[0x873c63c]
llc(llvm::MachineFunctionPass::runOnFunction(llvm::Function&)+0x28)[0x83f51e6]
llc(llvm::FPPassManager::runOnFunction(llvm::Function&)+0x13a)[0x884042e]
llc(llvm::FunctionPassManagerImpl::run(llvm::Function&)+0x6e)[0x88406b2]
llc(llvm::FunctionPassManager::run(llvm::Function&)+0x88)[0x8840772]
llc(main+0xbaa)[0x83c6b9e]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xdc)[0xb7d2e8cc]
llc(__gxx_personality_v0+0x171)[0x83c4f11]
Segmentation fault (core dumped)



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar 18 09:52:28 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Mar 2007 09:52:28 -0500
Subject: [LLVMbugs] [Bug 1258] NEW: type planes not collapsed for unnamed
	variables
Message-ID: <200703181452.l2IEqSZB001247@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1258

           Summary: type planes not collapsed for unnamed variables
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nicholas at mxc.ca


The following program assembles with llvm-as:

  define i32 @test1(i32 %a, i32 %b) {
  entry:
          icmp eq i32 %b, %a              ; <i1>:0 [#uses=1]
          zext i1 %0 to i32               ; <i32>:0 [#uses=1]
          ret i32 %0
  }

The type of %0 is simultaneously i1 and i32. By contrast, naming the variable %A
makes it invalid. This does not assemble

  define i32 @test2(i32 %a, i32 %b) {
  entry:
          %A = icmp eq i32 %b, %a
          %A = zext i1 %A to i32
          ret i32 %A
  }

with the error "Redefinition of value 'A' of type 'i32'" which is desired. I'd
like this C code:

  int test1(int a, int b) { return a == b; }

to produce this llvm assembly:

  define i32 @test1(i32 %a, i32 %b) {
  entry:
          icmp eq i32 %b, %a              ; <i1>:0 [#uses=1]
          zext i1 %1 to i32               ; <i32>:1 [#uses=1]
          ret i32 %1
  }



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar 18 12:27:57 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Mar 2007 12:27:57 -0500
Subject: [LLVMbugs] [Bug 1259] NEW: LiveVars assertion handling inline
	assembler.
Message-ID: <200703181727.l2IHRvhk004327@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1259

           Summary: LiveVars assertion handling inline assembler.
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru
                CC: evan.cheng at apple.com


Consider attached bytecode. LLC results with:

./llc bugpoint-reduced-simplified.bc
llc: /home/asl/proj/llvm/src/lib/CodeGen/LiveInterval.cpp:189: llvm::LiveRange*
llvm::LiveInterval::addRangeFrom(llvm::LiveRange, llvm::LiveRange*): Assertion
`B->end <= Start && "Cannot overlap two LiveRanges with differing ValID's" "
(did you def the same reg twice in a MachineInstr?)"' failed.
./llc((anonymous namespace)::PrintStackTrace()+0x1f)[0x863177f]
/lib/libc.so.6(abort+0xe9)[0xb7d25229]
/lib/libc.so.6(__assert_fail+0xeb)[0xb7d1d4e3]
./llc(llvm::LiveInterval::addRangeFrom(llvm::LiveRange,
llvm::LiveRange*)+0x1ee)[0x84e4a0e]

I don't see anything wrong with the inline assembler itself.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar 18 16:59:29 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Mar 2007 16:59:29 -0500
Subject: [LLVMbugs] [Bug 1260] NEW: Need MMX support for Qt operations
Message-ID: <200703182159.l2ILxThU025699@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1260

           Summary: Need MMX support for Qt operations
           Product: libraries
           Version: 1.0
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


We should at least support enough MMX to build Qt with MMX intrinsics enabled.  Here is a testcase to get 
started:
http://lists.cs.uiuc.edu/pipermail/llvmdev/2007-March/008196.html

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar 18 19:19:57 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 Mar 2007 19:19:57 -0500
Subject: [LLVMbugs] [Bug 1226] scalarrepl should be able to scalarrepl
	aggregates with memcpy uses
Message-ID: <200703190019.l2J0JvLu009213@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1226

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0



------- Additional Comments From sabre at nondot.org  2007-03-18 19:19 -------
Here is the final piece:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070312/045911.html

Testcases here:
Transforms/ScalarRepl/memset-aggregate.ll
Transforms/ScalarRepl/memset-aggregate-byte-leader.ll

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 19 13:42:20 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Mar 2007 13:42:20 -0500
Subject: [LLVMbugs] [Bug 1258] type planes not collapsed for unnamed
	variables
Message-ID: <200703191842.l2JIgKRj017350@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1258

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-19 13:42 -------
Fixed with these patches:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/045938.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/045939.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/045940.html

Test case here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/045942.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 19 13:46:44 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Mar 2007 13:46:44 -0500
Subject: [LLVMbugs] [Bug 1261] NEW: instcombine miscompilation
Message-ID: <200703191846.l2JIkiGQ017448@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1261

           Summary: instcombine miscompilation
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: major
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: wenwenti at hotmail.com


I found an error in instcombine pass. Please look at the b.ll file attached. At
line 956, the extension operation is sext. After instcombine, it has been
changed to zext which causes semantic errors.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 19 17:22:00 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Mar 2007 17:22:00 -0500
Subject: [LLVMbugs] [Bug 1262] NEW: LTO example in the LTO docs fails to
	build
Message-ID: <200703192222.l2JMM0WP022319@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1262

           Summary: LTO example in the LTO docs fails to build
           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


Near the start of http://llvm.org/docs/LinkTimeOptimization.html there is
"Example of link time optimization".  I gave it a whirl and got:
$ gcc --emit-llvm -c a.c -o a.o
$ gcc -c main.c -o main.o
$ gcc a.o main.o -o main
a.o: file not recognized: File format not recognized
collect2: ld returned 1 exit status



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 19 18:19:21 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 Mar 2007 18:19:21 -0500
Subject: [LLVMbugs] [Bug 1261] instcombine miscompilation
Message-ID: <200703192319.l2JNJLNb025890@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1261

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-19 18:19 -------


*** This bug has been marked as a duplicate of 1205 ***



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 11:36:01 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 11:36:01 -0500
Subject: [LLVMbugs] [Bug 1263] NEW: missed bitcast optimization
Message-ID: <200703211636.l2LGa1da030026@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1263

           Summary: missed bitcast 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


In the attached .ll you can find:
        %tmp64 = bitcast i32* %tmp1 to { i32 }*         ; <{ i32 }*> [#uses=1]
        %tmp65 = getelementptr { i32 }* %tmp64, i32 0, i32 0            ; 
<i32*> [#uses=1]
Thus %tmp65 is equal to %tmp1, but the optimizers miss this.  This strange
construction was produced by llvm-convert, or so it seems.  I will try to
find out why it is being produced in the first place.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 11:56:54 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 11:56:54 -0500
Subject: [LLVMbugs] [Bug 1264] NEW: converting 64-bit constant pool entry to
	32-bit not necessarily beneficial
Message-ID: <200703211656.l2LGusZK032415@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1264

           Summary: converting 64-bit constant pool entry to 32-bit not
                    necessarily beneficial
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: djg at cray.com


For this test case:

define double @foo(double %x) {
        %y = mul double %x, 5.000000e-01
        ret double %y
}

llc -march=x86-64 currently produces a 32-bit constant pool entry and this code:

        cvtss2sd .LCPI1_0(%rip), %xmm1
        mulsd %xmm1, %xmm0

instead of just using a 64-bit constant pool entry with this:

        mulsd .LCPI1_0(%rip), %xmm0

This is due to the code in ExpandConstantFP in LegalizeDAG.cpp. It notices that
x86-64 indeed has an instruction to load a 32-bit float from memory and convert
it into a 64-bit float in a register, however it doesn't notice that this isn't
beneficial because it prevents the load from being folded into the multiply.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 12:16:34 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 12:16:34 -0500
Subject: [LLVMbugs] [Bug 1256] llvm-upgrade cannot distinguish between int
	tmp & uint tmp etc.
Message-ID: <200703211716.l2LHGYJi000414@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1256

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-21 12:16 -------
Test Case Here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070312/045862.html

Patches Here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046051.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 12:27:36 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 12:27:36 -0500
Subject: [LLVMbugs] [Bug 1243] llvm-upgrade issues bogus warnings about
	renaming internal symbols
Message-ID: <200703211727.l2LHRaOh000627@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1243

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-21 12:27 -------
Patches Here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046051.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046053.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 15:17:48 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 15:17:48 -0500
Subject: [LLVMbugs] [Bug 1265] NEW: pointless masking being generated for
	zext values
Message-ID: <200703212017.l2LKHm4K004708@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1265

           Summary: pointless masking being generated for zext values
           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


Running opt -std-compile-opts on the attached bytecode results in:

        tail call void @report__test( i64 or (i64 and (i64 zext (i32 ptrtoint 
([7 x i8]* @str1 to i32) to i64), i64 4294967295), i64 shl (i64 zext (i32 
ptrtoint (%struct.string___XUB* @C.1.384 to i32) to i64), i64 32)) )

Note how an i32 is being zero extended to an i64, then the upper 32 bits of
the result, which are necessarily zero, are being pointlessly masked off.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 18:21:54 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 18:21:54 -0500
Subject: [LLVMbugs] [Bug 1248] Another InstCombine issue
Message-ID: <200703212321.l2LNLsH0008141@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1248

rspencer at x10sys.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         OS/Version|Linux                       |All
           Platform|PC                          |All
         Resolution|                            |FIXED



------- Additional Comments From rspencer at x10sys.com  2007-03-21 18:21 -------
Turns out there were two bugs. The AddWithOverflow needed to be APIntified and
to deal with the sign properly. Also, in the case of:
   %Y = sdiv iN %X, C1
   icmp pred iN %Y, -C2
there was an off-by-one error in the high bound for the range test that is
produced in the replacement. 

This patch was applied:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046072.html




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 19:25:39 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 19:25:39 -0500
Subject: [LLVMbugs] [Bug 1223] verifier produces bytecode that causes a
	bytecode reader assertion failure
Message-ID: <200703220025.l2M0PdsL009451@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1223

rspencer at x10sys.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
           Keywords|                            |invalid-bug
         Resolution|                            |INVALID
   Target Milestone|---                         |2.0



------- Additional Comments From rspencer at x10sys.com  2007-03-21 19:25 -------
Ryan,

I cannot reproduce this bug. Please see the attached test program and its input
data. I tried to replicate exactly what you suggested was going on but the
verifier doesn't complain and the assembler doesn't crash. I'm marking this bug
as invalid. If you think I got something wrong, please reopen it and explain.

Some possibilities:
1. The bug has been fixed in the last month
2. You weren't using LLVM head version
3. Test program doesn't replicate your exact situation.
4. Something else.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 19:42:25 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 19:42:25 -0500
Subject: [LLVMbugs] [Bug 1060] ParseAssemblyString does not allow already
	existing names
Message-ID: <200703220042.l2M0gPcv009839@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1060

rspencer at x10sys.com changed:

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





------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 20:53:32 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 20:53:32 -0500
Subject: [LLVMbugs] [Bug 1257] llc crashes when it compiles the attached file
Message-ID: <200703220153.l2M1rWTm031658@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1257

evan.cheng at apple.com changed:

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



------- Additional Comments From evan.cheng at apple.com  2007-03-21 20:53 -------
Fixed.

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046078.html




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 21 21:21:22 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 Mar 2007 21:21:22 -0500
Subject: [LLVMbugs] [Bug 1145] Support nothrow attribute on function calls
Message-ID: <200703220221.l2M2LMTj002189@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1145

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-21 21:21 -------
These patches add assembly support for nounwind and noreturn attributes:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046082.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046084.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046086.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046088.html

Test case is here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046087.html

Both nothrow and noreturn are now fully supported in VMCore, Bytecode, and
Assembly. I'm closing this bug as done since all that it asks has been
completed. Changes to the prune-eh pass should be done under a separate PR.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar 22 08:55:24 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Mar 2007 08:55:24 -0500
Subject: [LLVMbugs] [Bug 1266] NEW: assertion failed in RegisterScavenging
Message-ID: <200703221355.l2MDtO9M029727@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1266

           Summary: assertion failed in RegisterScavenging
           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: lauro.venancio at gmail.com


laurov at edgy-laptop:/tmp$ llc bugpoint-reduced-simplified.bc 
llc: /home/laurov/llvm/llvm/lib/CodeGen/RegisterScavenging.cpp:138: void
llvm::RegScavenger::forward(): Assertion `isUsed(Reg)' failed.
llc((anonymous namespace)::PrintStackTrace()+0x1a)[0x88b929a]
llc((anonymous namespace)::SignalHandler(int)+0x112)[0x88b9560]
[0xffffe420]
/lib/tls/i686/cmov/libc.so.6(abort+0x103)[0xb7da1ef3]
/lib/tls/i686/cmov/libc.so.6(__assert_fail+0xfb)[0xb7d99dbb]
llc(llvm::RegScavenger::forward()+0x389)[0x8723507]
llc(llvm::RegScavenger::forward(llvm::ilist_iterator<llvm::MachineInstr>)+0x13)[0x84475c3]
llc[0x870dd3e]
llc[0x870fcf2]
llc(llvm::MachineFunctionPass::runOnFunction(llvm::Function&)+0x28)[0x83f802e]
llc(llvm::FPPassManager::runOnFunction(llvm::Function&)+0x13a)[0x8845ac6]
llc(llvm::FunctionPassManagerImpl::run(llvm::Function&)+0x6e)[0x8845d4a]
llc(llvm::FunctionPassManager::run(llvm::Function&)+0x88)[0x8845e0a]
llc(main+0xbaa)[0x83c920e]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xdc)[0xb7d8c8cc]
llc(__gxx_personality_v0+0x17d)[0x83c7581]
Aborted (core dumped)



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar 22 17:11:33 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 Mar 2007 17:11:33 -0500
Subject: [LLVMbugs] [Bug 1267] NEW: Fix Bug 930
Message-ID: <200703222211.l2MMBX1K012745@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1267

           Summary: Fix Bug 930
           Product: Bugzilla Admin
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: UI
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


gccbot can't read bug 930 because it has an invalid character in the second
comment (attachment comment). The character is 0xa0 located after the ). We just
need to go into the database and hack the comment text.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 23 16:28:35 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Mar 2007 16:28:35 -0500
Subject: [LLVMbugs] [Bug 1268] NEW: slow compilation coming from dse
Message-ID: <200703232128.l2NLSZwg011741@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1268

           Summary: slow compilation coming from dse
           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


opt -dse takes about 15 minutes on the original byte code (see
orig.ll attachment).  I reduced it somewhat using bugpoint,
and it now takes about 2 minutes (reduced.ll).  I reduced it
completely using bugpoint with a timeout of 5 seconds, but
it seems like it was reduced too much, but I attach it anyway
as too_reduced.ll.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 23 18:20:02 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 Mar 2007 18:20:02 -0500
Subject: [LLVMbugs] [Bug 1268] slow compilation coming from dse
Message-ID: <200703232320.l2NNK2EK025948@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1268

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-23 18:20 -------
After analyzing this a bit more carefully, this just seems to be due to general
algorithmic deficiencies in AliasSetTracker and/or DSE.  For 1700 store
instructions that it added to the AliasSetTracker, it is causing millions of
calls to functions that don't individually take much time. Seems roughly an
O(N^2) issue.

This appears to be a duplicate of PR930. Marking it as such.

*** This bug has been marked as a duplicate of 930 ***



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 24 16:12:20 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Mar 2007 16:12:20 -0500
Subject: [LLVMbugs] [Bug 1269] NEW: Remove "invoke",
	add "unwind to" for labels
Message-ID: <200703242112.l2OLCKm4009353@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1269

           Summary: Remove "invoke", add "unwind to" for labels
           Product: libraries
           Version: trunk
          Platform: All
               URL: http://nondot.org/~sabre/LLVMNotes/ExceptionHandlingChan
                    ges.txt
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: LLVM assembly language parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


I would like to see Chris' notes on Exception Handling implemented (see the link
above), preferably for 2.0. My reasons for this are two fold:

1. It fits in the same category of disruptive assembly/bytecode changes as the 
   other major changes in LLVM 2.0; so, including it in 2.0 means less future
   impact for 2.0 users. 
2. As I start to develop the HLVM exception model, I'd prefer to do it on an EH
   model in LLVM that is simpler for the kinds of languages HLVM targets.

Yes, I'm signing up to do this; hopefully, before 2.0 is released.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 24 16:28:25 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 Mar 2007 16:28:25 -0500
Subject: [LLVMbugs] [Bug 1270] NEW: Allow AP integers <= 64 bits for any
	target
Message-ID: <200703242128.l2OLSPfw009766@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1270

           Summary: Allow AP integers <= 64 bits for any target
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


Currently, the various targets only handle arbitrary precision integers of
discrete byte sizes (1, 2, 4, 8 bytes). However, the LLVM IR can now handle
integer bit widths from 1 bit to 8 million bits. In cases where the integer bit
size doesn't match the discrete byte size of a "known" integer, the common code
generator will generate an assertion because the integer size doesn't map to
something the machine recognizes.

This problem needs to be fixed because:

a) it prevents codegen for programs with "funny" bit sizes.
b) it prevents bit-size optimizations from working. For example, a bit size
   minimization pass might be able to determine that an i32 value really only
   needs 14 bits. It would thus fit into an i16 instead of an i32. However the
   pass would generate i14 which would cause the common code gen to assert.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar 25 00:49:08 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Mar 2007 00:49:08 -0500
Subject: [LLVMbugs] [Bug 1265] pointless masking being generated for zext
	values
Message-ID: <200703250549.l2P5n8JQ019450@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1265

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement
             Status|NEW                         |RESOLVED
          Component|new bugs                    |Core LLVM classes
           Keywords|                            |quality-of-implementation
         OS/Version|Linux                       |All
            Product|new-bugs                    |libraries
           Platform|Other                       |All
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0
            Version|unspecified                 |1.0



------- Additional Comments From sabre at nondot.org  2007-03-25 00:49 -------
Implemented, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046236.html

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar 25 00:50:39 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Mar 2007 00:50:39 -0500
Subject: [LLVMbugs] [Bug 1254] CBE handling of inline asm immediate operand
	broken
Message-ID: <200703250550.l2P5odbc019499@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1254

sabre at nondot.org changed:

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



------- Additional Comments From sabre at nondot.org  2007-03-25 00:50 -------


*** This bug has been marked as a duplicate of 802 ***



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar 25 14:00:00 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Mar 2007 14:00:00 -0500
Subject: [LLVMbugs] [Bug 1271] NEW: Instcombine broken after APIntification
Message-ID: <200703251900.l2PJ007a026300@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1271

           Summary: Instcombine broken after APIntification
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


instcombine is broken now.

This leads to big miscompilations here and there.
The smallest testcase is:
Regression/C/2003-05-21-UnionBitfields.c



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sun Mar 25 15:44:26 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 Mar 2007 15:44:26 -0500
Subject: [LLVMbugs] [Bug 1263] missed bitcast optimization
Message-ID: <200703252044.l2PKiQJl028652@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1263

sabre at nondot.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement
             Status|NEW                         |RESOLVED
          Component|new bugs                    |Scalar Optimizations
           Keywords|                            |code-quality
         OS/Version|Linux                       |All
            Product|new-bugs                    |libraries
           Platform|Other                       |All
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0
            Version|unspecified                 |1.0



------- Additional Comments From sabre at nondot.org  2007-03-25 15:44 -------
Nice catch.  Testcase here: Transforms/InstCombine/cast2.ll:test3

Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046249.html

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 00:34:48 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 00:34:48 -0500
Subject: [LLVMbugs] [Bug 1271] Instcombine broken after APIntification
Message-ID: <200703260534.l2Q5YmGi005390@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1271

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-26 00:34 -------
Test cases here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046251.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046257.html

Fixes here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046250.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070319/046252.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046256.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 02:41:22 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 02:41:22 -0500
Subject: [LLVMbugs] [Bug 1193] missing optimization for bit-accurate types
Message-ID: <200703260741.l2Q7fMlP007704@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1193

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-26 02:41 -------
This appears to have been fixed. Running "opt -instcombine" on the test program
yields this code:

define void @test_add_sub(i11 sext  %a, i11 sext  %b, i11* %r1, i11* %r2) {
entry:
        %tmp45 = add i11 %b, %a         ; <i11> [#uses=1]
        store i11 %tmp45, i11* %r1
        %tmp113 = sub i11 %a, %b                ; <i11> [#uses=1]
        store i11 %tmp113, i11* %r2
        ret void
}

I don't think it gets much better than that.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 06:22:44 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 06:22:44 -0500
Subject: [LLVMbugs] [Bug 1271] Instcombine broken after APIntification
Message-ID: <200703261122.l2QBMiZX012350@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1271

asl at math.spbu.ru changed:

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



------- Additional Comments From asl at math.spbu.ru  2007-03-26 06:22 -------
Unfortunately, things are still broken: Qt crashes, llvm-gcc bootstrap too.
Haven't tried mozilla, but I think it's broken too. I'm working on testcase.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 09:08:44 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 09:08:44 -0500
Subject: [LLVMbugs] [Bug 1272] NEW: Assertion failed generating code
	(ScheduleDAG.cpp:377)
Message-ID: <200703261408.l2QE8iRx015898@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1272

           Summary: Assertion failed generating code (ScheduleDAG.cpp:377)
           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: nipo at ssji.net


I'm rather new to LLVM and trying to compile some C++ software I already got.
llvm-g++ produced a bytecode file happily, but I get an assertion failure when
translating the bytecode to x86 machine code, either through lli or llc.

Assertion failure is:
llc: /dsk/l1/compilation/cvs/llvm/lib/CodeGen/SelectionDAG/ScheduleDAG.cpp:378:
void llvm::ScheduleDAG::AddOperand(llvm::MachineInstr*, llvm::SDOperand,
unsigned int, const llvm::TargetInstrDescriptor*, llvm::DenseMap<llvm::SDNode*,
unsigned int, llvm::DenseMapKeyInfo<llvm::SDNode*> >&): Assertion
`RegMap->getRegClass(VReg) == RC && "Register class of operand and regclass of
use don't agree!"' failed.
llc((anonymous namespace)::PrintStackTrace()+0x1a)[0x88dfe50]
llc((anonymous namespace)::SignalHandler(int)+0x110)[0x88e0114]
[0xb7f4e420]
/lib/libc.so.6(abort+0x109)[0xb7d17c69]
/lib/libc.so.6(__assert_fail+0xfc)[0xb7d0feac]
llc(llvm::ScheduleDAG::AddOperand(llvm::MachineInstr*, llvm::SDOperand, unsigned
int, llvm::TargetInstrDescriptor const*, llvm::DenseMap<llvm::SDNode*, unsigned
int, llvm::DenseMapKeyInfo<llvm::SDNode*> >&)+0x7c7)[0x86d2dff]
llc(llvm::ScheduleDAG::EmitNode(llvm::SDNode*, llvm::DenseMap<llvm::SDNode*,
unsigned int, llvm::DenseMapKeyInfo<llvm::SDNode*> >&)+0x28d)[0x86d3095]
llc(llvm::ScheduleDAG::EmitSchedule()+0x2a8)[0x86d3c72]
llc[0x8640e51]
llc(llvm::ScheduleDAG::Run()+0x94)[0x86d1a4c]
llc(llvm::SelectionDAGISel::ScheduleAndEmitDAG(llvm::SelectionDAG&)+0x73)[0x8662def]
llc[0x850f31e]
llc(llvm::SelectionDAGISel::CodeGenAndEmitDAG(llvm::SelectionDAG&)+0x137)[0x8662f49]
llc(llvm::SelectionDAGISel::SelectBasicBlock(llvm::BasicBlock*,
llvm::MachineFunction&, llvm::FunctionLoweringInfo&)+0xad)[0x8677dad]
llc(llvm::SelectionDAGISel::runOnFunction(llvm::Function&)+0x5cf)[0x867940f]
llc[0x853fcf5]
llc(llvm::FPPassManager::runOnFunction(llvm::Function&)+0x13a)[0x886d506]
llc(llvm::FunctionPassManagerImpl::run(llvm::Function&)+0x6e)[0x886d788]
llc(llvm::FunctionPassManager::run(llvm::Function&)+0x88)[0x886d848]
llc(main+0xbc3)[0x83d8445]
/lib/libc.so.6(__libc_start_main+0xd8)[0xb7d03878]
llc(__gxx_personality_v0+0x179)[0x83d67e1]
Aborted

Putting a bp in gdb just before the assertion, I get:

Breakpoint 3, llvm::ScheduleDAG::AddOperand (this=0x92affc8, MI=0x8d42f00,
Op={Val = 0x92af680, ResNo = 0}, IIOpNum=4, II=0x894cdc0, VRBaseMap=@0xbfe67e50)
    at /dsk/l1/compilation/cvs/llvm/lib/CodeGen/SelectionDAG/ScheduleDAG.cpp:377
377           assert(RegMap->getRegClass(VReg) == RC &&
3: RC = (const llvm::TargetRegisterClass *) 0x8a03e80
1: this->RegMap->getRegClass (VReg) = (const llvm::TargetRegisterClass *) 0x8a04080

Dumping some variables around:

(gdb) print *RC
$1 = {_vptr.TargetRegisterClass = 0x895d988, ID = 2, isSubClass = false, VTs =
0x8959200, SubClasses = 0x8959208, SuperClasses = 0x895920c, RegSize = 8,
Alignment = 8, RegsBegin = 0x89591c0, 
  RegsEnd = 0x8959200}
(gdb) print *this->RegMap->getRegClass (VReg) 
$2 = {_vptr.TargetRegisterClass = 0x895d910, ID = 10, isSubClass = false, VTs =
0x8958da0, SubClasses = 0x8958da8, SuperClasses = 0x8958dac, RegSize = 8,
Alignment = 4, RegsBegin = 0x8958d80, 
  RegsEnd = 0x8958da0}
(gdb) print VReg
$3 = 1069
(gdb) print Op
$4 = {Val = 0x92af680, ResNo = 0}
(gdb) print *MI
$5 = {TID = 0x894cdc0, NumImplicitOps = 0, Operands =
{<std::_Vector_base<llvm::MachineOperand,std::allocator<llvm::MachineOperand> >> = {
      _M_impl = {<std::allocator<llvm::MachineOperand>> =
{<__gnu_cxx::new_allocator<llvm::MachineOperand>> = {<No data fields>}, <No data
fields>}, _M_start = 0x92b04e8, _M_finish = 0x92b0538, 
        _M_end_of_storage = 0x92b0538}}, <No data fields>}, prev = 0x0, next =
0x928ec01, parent = 0x0}
(gdb) print IIOpNum
$6 = 4
(gdb) print (void)Op.Val->dump(&DAG)  
0x92af680: f64,ch,flag = CopyFromReg 0x92af7e0, 0x92af538, 0x92af7e0:1
$7 = void



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 12:26:23 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 12:26:23 -0500
Subject: [LLVMbugs] [Bug 1272] Assertion failed generating code
	(ScheduleDAG.cpp:377)
Message-ID: <200703261726.l2QHQNQZ019753@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1272

asl at math.spbu.ru changed:

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



------- Additional Comments From asl at math.spbu.ru  2007-03-26 12:26 -------
Please compile your sources with -D__NO_MATH_INLINES until PR879 will be fixed

*** This bug has been marked as a duplicate of 879 ***



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 13:10:43 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 13:10:43 -0500
Subject: [LLVMbugs] [Bug 1273] NEW: CBE fails on
	SingleSource/Regression/C/2003-05-22-VarSizeArray
Message-ID: <200703261810.l2QIAhIG020708@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1273

           Summary: CBE fails on SingleSource/Regression/C/2003-05-22-
                    VarSizeArray
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: C
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


This test case has been failing for CBE (only) for some time now. The problem
appears to be getting GCC 4.0.3 to compile the generated code:

Output/2003-05-22-VarSizeArray.cbe.c:148: internal compiler error: in
lhd_set_decl_assembler_name, at langhooks.c:191

Details to follow.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 13:14:03 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 13:14:03 -0500
Subject: [LLVMbugs] [Bug 1273] CBE fails on
	SingleSource/Regression/C/2003-05-22-VarSizeArray
Message-ID: <200703261814.l2QIE3ZC020816@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1273

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-26 13:14 -------
I should have looked at the code first. The problem is:

  __builtin_stack_restore(llvm_cbe_savedstack);



*** This bug has been marked as a duplicate of 1028 ***



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 15:21:14 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 15:21:14 -0500
Subject: [LLVMbugs] [Bug 915] llvm-gcc doesn't support nested functions
Message-ID: <200703262021.l2QKLEQc023438@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=915

sabre at nondot.org changed:

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



------- Additional Comments From sabre at nondot.org  2007-03-26 15:21 -------
Verified, this example doesn't use trampolines, but causes infinite recursion in the FE.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 15:25:08 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 15:25:08 -0500
Subject: [LLVMbugs] [Bug 1274] NEW: crash in lower_nested_functions
Message-ID: <200703262025.l2QKP8Pw023525@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1274

           Summary: crash in lower_nested_functions
           Product: tools
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


This example causes infinite recursion in walk_tree, called from lower_nested_functions:
static void foo(void *a){
  inline void foo_bar() {
    a += 1;
  }
}

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 15:25:30 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 15:25:30 -0500
Subject: [LLVMbugs] [Bug 915] llvm-gcc doesn't support nested functions
Message-ID: <200703262025.l2QKPUoO023551@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=915

sabre at nondot.org changed:

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



------- Additional Comments From sabre at nondot.org  2007-03-26 15:25 -------
Actually, this looks like an unrelated bug.  I've filed it as Bug 1274.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 17:03:05 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 17:03:05 -0500
Subject: [LLVMbugs] [Bug 1275] NEW: Combine zero/sign/any extended rotl/rotr
	patterns
Message-ID: <200703262203.l2QM35bU025628@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1275

           Summary: Combine zero/sign/any extended rotl/rotr patterns
           Product: libraries
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: scottm at aero.org


The attached patch adds the ability to fold zero/sign/any extended rotl/rotr 
patterns that occur in the test/CodeGen/X86/rotate.ll test file on other 
platforms (PPC, CellSPU). Also added the ability to custom lower rotl and rotr, 
since the Cell only supports 16 and 32-bit rotate instructions (needed to 
custom lower for 8-bit, for example.)



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 17:10:08 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 17:10:08 -0500
Subject: [LLVMbugs] [Bug 1276] NEW: docs/Makefile minor mod
Message-ID: <200703262210.l2QMA8DC025772@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1276

           Summary: docs/Makefile minor mod
           Product: Documentation
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Doxygen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: scottm at aero.org


Tweaked the makefile so that it's easier to regenerate the doxygen-generated 
documentation instead of having to go through an install.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 17:14:41 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 17:14:41 -0500
Subject: [LLVMbugs] [Bug 1277] NEW: Print error message if bytecode not read
	correctly
Message-ID: <200703262214.l2QMEfVQ025874@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1277

           Summary: Print error message if bytecode not read correctly
           Product: tools
           Version: trunk
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: scottm at aero.org


Attached patch prints error string generated by bytecode reader when bytecode 
fails to read correctly.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 17:39:24 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 17:39:24 -0500
Subject: [LLVMbugs] [Bug 1277] Print error message if bytecode not read
	correctly
Message-ID: <200703262239.l2QMdOvG026558@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1277

rspencer at x10sys.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
           Keywords|                            |code-cleanup
         Resolution|                            |FIXED



------- Additional Comments From rspencer at x10sys.com  2007-03-26 17:39 -------
Patch applied. Thanks, Scott.

If you're interested, it would be useful to review the other tools to see if
they have the same problem.

Reid.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 17:43:04 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 17:43:04 -0500
Subject: [LLVMbugs] [Bug 1266] assertion failed in RegisterScavenging
Message-ID: <200703262243.l2QMh4VA026712@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1266

evan.cheng at apple.com changed:

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



------- Additional Comments From evan.cheng at apple.com  2007-03-26 17:43 -------
Fixed.

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046281.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046282.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046284.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046285.html




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 18:38:30 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 18:38:30 -0500
Subject: [LLVMbugs] [Bug 1246] Bug in LoopSimplfy pass
Message-ID: <200703262338.l2QNcUI2027782@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1246

dalej at apple.com changed:

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



------- Additional Comments From dalej at apple.com  2007-03-26 18:38 -------
Yes, that fixes it.  thanks!




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Mon Mar 26 20:17:55 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 Mar 2007 20:17:55 -0500
Subject: [LLVMbugs] [Bug 1278] NEW: struct layout suboptimality
Message-ID: <200703270117.l2R1HtOe030290@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1278

           Summary: struct layout suboptimality
           Product: tools
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


On darwin ppc32, this code:

struct s {
  double d1;
  int s1;
};

struct s foo(void) {
  struct s S = {1.1, 2};
  return S;
}

Compiles "s" to type { double, i32, [4 x i8] }.  There are two problems with this:

1. [4 x i8] can be int, which would be more efficient since it is just tail padding.  This tail padding exists 
because the struct has 8-byte alignment, but llvm thinks both double/int have 4 byte alignment.  If the 
padding bytes weren't made explicit, the struct would be too small.

2. CopyAggregate copies by the llvm fields even if there are no source-level fields that correspond to 
them.

The end result of this is that the example compiles into:

define void @foo(%struct.s* %agg.result) {
entry:
        %tmp20 = getelementptr %struct.s* %agg.result, i32 0, i32 0             ; <double*> [#uses=1]
        store double 1.100000e+00, double* %tmp20
        %tmp23 = getelementptr %struct.s* %agg.result, i32 0, i32 1             ; <i32*> [#uses=1]
        store i32 2, i32* %tmp23
        %tmp28 = getelementptr %struct.s* %agg.result, i32 0, i32 2, i32 0              ; <i8*> [#uses=1]
        store i8 0, i8* %tmp28
        %tmp31 = getelementptr %struct.s* %agg.result, i32 0, i32 2, i32 1              ; <i8*> [#uses=1]
        store i8 0, i8* %tmp31
        %tmp34 = getelementptr %struct.s* %agg.result, i32 0, i32 2, i32 2              ; <i8*> [#uses=1]
        store i8 0, i8* %tmp34
        %tmp37 = getelementptr %struct.s* %agg.result, i32 0, i32 2, i32 3              ; <i8*> [#uses=1]
        store i8 0, i8* %tmp37
        ret void
}

Note that the 4 byte stores are completely dead.  If #1 were solved at least we'd have a single dead 
word store :)

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 01:06:52 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 01:06:52 -0500
Subject: [LLVMbugs] [Bug 1276] docs/Makefile minor mod
Message-ID: <200703270606.l2R66qS7005827@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1276

rspencer at x10sys.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         OS/Version|Linux                       |All
           Platform|Other                       |All
         Resolution|                            |FIXED



------- Additional Comments From rspencer at x10sys.com  2007-03-27 01:06 -------
Patch applied. Thanks, Scott.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 01:12:05 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 01:12:05 -0500
Subject: [LLVMbugs] [Bug 1271] Instcombine broken after APIntification
Message-ID: <200703270612.l2R6C5oU005946@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1271

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-27 01:12 -------
Anton,

This has passed llvm/test, llvm-test, and a bootstrap build of llvm-gcc.

I'm calling it done.  If you find more bugs in instcombine (e.g. Qt, SeaMonkey),
they are probably separate issues, please file a new bug.

Thanks,

Reid.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 09:04:07 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 09:04:07 -0500
Subject: [LLVMbugs] [Bug 1247] build failure on darwin 8.8.1
Message-ID: <200703271404.l2RE47UA014799@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1247

ebner at complang.tuwien.ac.at changed:

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



------- Additional Comments From ebner at complang.tuwien.ac.at  2007-03-27 09:04 -------
the failure described above occurs if there's the install dir of a previous
llvm(-gcc) installation in the PATH. building with a clean environment solved
the problem for me. there are similir issuses when building llvm, e.g., tblgen
appears to be taken from the former installation which causes build failures.




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 14:46:59 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 14:46:59 -0500
Subject: [LLVMbugs] [Bug 1279] NEW: another assertion failed in
	RegisterScavenging
Message-ID: <200703271946.l2RJkxh9031167@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1279

           Summary: another assertion failed in RegisterScavenging
           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: lauro.venancio at gmail.com


laurov at laurov-desktop:~/llvm/llvm/build$ llc bugpoint-reduced-simplified.bc
llc: /home/laurov/llvm/llvm/lib/CodeGen/RegisterScavenging.cpp:141: void
llvm::RegScavenger::forward(): Assertion `isUnused(Reg) || isReserved(Reg)' failed.
llc((anonymous namespace)::PrintStackTrace()+0x1a)[0x88d7886]
llc((anonymous namespace)::SignalHandler(int)+0x112)[0x88d7b4c]
[0xffffe420]
/lib/tls/i686/cmov/libc.so.6(abort+0x103)[0xb7d8cef3]
/lib/tls/i686/cmov/libc.so.6(__assert_fail+0xfb)[0xb7d84dbb]
llc(llvm::RegScavenger::forward()+0x3df)[0x8741249]
llc(llvm::RegScavenger::forward(llvm::ilist_iterator<llvm::MachineInstr>)+0x13)[0x8455267]
llc[0x872b9ea]
llc[0x872d99e]
llc(llvm::MachineFunctionPass::runOnFunction(llvm::Function&)+0x28)[0x8405a8e]
llc(llvm::FPPassManager::runOnFunction(llvm::Function&)+0x13a)[0x88639a6]
llc(llvm::FunctionPassManagerImpl::run(llvm::Function&)+0x6e)[0x8863c2a]
llc(llvm::FunctionPassManager::run(llvm::Function&)+0x88)[0x8863cea]
llc(main+0xbe9)[0x83d6bdd]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xdc)[0xb7d778cc]
llc(__gxx_personality_v0+0x175)[0x83d4f11]
Aborted (core dumped)



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 15:27:50 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 15:27:50 -0500
Subject: [LLVMbugs] [Bug 1280] NEW: llc can't compile "sext i4 to i64"
	emitted by llvm-gcc
Message-ID: <200703272027.l2RKRoGx013552@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1280

           Summary: llc can't compile "sext i4 to i64" emitted by llvm-gcc
           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: lauro.venancio at gmail.com


laurov at laurov-desktop:~/llvm/llvm/build$ llc -march=x86
bugpoint-reduced-simplified.bc
llc: /home/laurov/llvm/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp:2257:
llvm::MVT::ValueType llvm::TargetLowering::getValueType(const llvm::Type*)
const: Assertion `0 && "Invalid width for value type"' failed.
llc((anonymous namespace)::PrintStackTrace()+0x1a)[0x88d7886]
llc((anonymous namespace)::SignalHandler(int)+0x112)[0x88d7b4c]
[0xffffe420]
/lib/tls/i686/cmov/libc.so.6(abort+0x103)[0xb7cdfef3]
/lib/tls/i686/cmov/libc.so.6(__assert_fail+0xfb)[0xb7cd7dbb]
llc(llvm::TargetLowering::getValueType(llvm::Type const*) const+0xbf)[0x868dff9]
llc(llvm::SelectionDAGLowering::getValue(llvm::Value const*)+0xaa)[0x8660a3c]
llc(llvm::SelectionDAGLowering::visitSExt(llvm::User&)+0x30)[0x866b834]
llc(llvm::SelectionDAGLowering::visit(unsigned int, llvm::User&)+0x2eb)[0x8688df9]
llc(llvm::SelectionDAGLowering::visit(llvm::Instruction&)+0x28)[0x8688fde]
llc(llvm::SelectionDAGISel::BuildSelectionDAG(llvm::SelectionDAG&,
llvm::BasicBlock*, std::vector<std::pair<llvm::MachineInstr*, unsigned int>,
std::allocator<std::pair<llvm::MachineInstr*, unsigned int> > >&,
llvm::FunctionLoweringInfo&)+0x14a)[0x866f26e]
llc(llvm::SelectionDAGISel::SelectBasicBlock(llvm::BasicBlock*,
llvm::MachineFunction&, llvm::FunctionLoweringInfo&)+0x98)[0x866fcd4]
llc(llvm::SelectionDAGISel::runOnFunction(llvm::Function&)+0x5d2)[0x86714e6]
llc[0x853a0ed]
llc(llvm::FPPassManager::runOnFunction(llvm::Function&)+0x13a)[0x88639a6]
llc(llvm::FunctionPassManagerImpl::run(llvm::Function&)+0x6e)[0x8863c2a]
llc(llvm::FunctionPassManager::run(llvm::Function&)+0x88)[0x8863cea]
llc(main+0xbe9)[0x83d6bdd]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xdc)[0xb7cca8cc]
llc(__gxx_personality_v0+0x175)[0x83d4f11]
Aborted (core dumped)



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 17:48:42 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 17:48:42 -0500
Subject: [LLVMbugs] [Bug 1281] NEW: Doxygen page gives 403 error
Message-ID: <200703272248.l2RMmgEN000992@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1281

           Summary: Doxygen page gives 403 error
           Product: Documentation
           Version: trunk
          Platform: All
               URL: http://llvm.org/doxygen/
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Doxygen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: resistor at mac.com


The LLVM Doxygen page is currently giving a 403 error.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 18:21:31 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 18:21:31 -0500
Subject: [LLVMbugs] [Bug 1281] Doxygen page gives 403 error
Message-ID: <200703272321.l2RNLVwL013921@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1281

tonic at nondot.org changed:

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



------- Additional Comments From tonic at nondot.org  2007-03-27 18:21 -------
Fix here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046311.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 20:33:17 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 20:33:17 -0500
Subject: [LLVMbugs] [Bug 1279] another assertion failed in RegisterScavenging
Message-ID: <200703280133.l2S1XHoT018025@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1279

evan.cheng at apple.com changed:

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



------- Additional Comments From evan.cheng at apple.com  2007-03-27 20:33 -------
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046321.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Tue Mar 27 20:45:03 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 Mar 2007 20:45:03 -0500
Subject: [LLVMbugs] [Bug 1280] llc can't compile "sext i4 to i64" emitted by
	llvm-gcc
Message-ID: <200703280145.l2S1j3W2018595@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1280

rspencer at x10sys.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|rspencer at x10sys.com         |
             Status|NEW                         |RESOLVED
          Component|new bugs                    |Scalar Optimizations
            Product|new-bugs                    |libraries
         Resolution|                            |FIXED
   Target Milestone|---                         |2.0
            Version|unspecified                 |trunk



------- Additional Comments From rspencer at x10sys.com  2007-03-27 20:45 -------
Fixed with this patch:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046325.html

Test case here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046326.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 28 03:30:03 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Mar 2007 03:30:03 -0500
Subject: [LLVMbugs] [Bug 1282] NEW: Extremly slow codegen
Message-ID: <200703280830.l2S8U3Gm027873@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1282

           Summary: Extremly slow codegen
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru


Consider attached file. I have:

 $ ./llc -time-passes sha512-opt.bc
===-------------------------------------------------------------------------===
                      ... Pass execution timing report ...
===-------------------------------------------------------------------------===
  Total Execution Time: 553.6418 seconds (554.1256 wall clock)

   ---User Time---   --System Time--   --User+System--   ---Wall Time---  ---
Name ---
  553.4408 ( 99.9%)   0.0899 ( 91.8%)  553.5308 ( 99.9%)  554.0134 ( 99.9%)  X86
DAG->DAG Instruction Selection
   0.0439 (  0.0%)   0.0000 (  0.0%)   0.0439 (  0.0%)   0.0466 (  0.0%)  Linear
Scan Register Allocator
   0.0169 (  0.0%)   0.0000 (  0.0%)   0.0169 (  0.0%)   0.0213 (  0.0%)  Live
Interval Analysis
   0.0099 (  0.0%)   0.0019 (  2.0%)   0.0119 (  0.0%)   0.0126 (  0.0%)  X86
AT&T-Style Assembly Printer
   0.0119 (  0.0%)   0.0000 (  0.0%)   0.0119 (  0.0%)   0.0079 (  0.0%)  Live
Variable Analysis
   0.0019 (  0.0%)   0.0000 (  0.0%)   0.0019 (  0.0%)   0.0047 (  0.0%)  Module
Verifier
   0.0029 (  0.0%)   0.0000 (  0.0%)   0.0029 (  0.0%)   0.0035 (  0.0%)  Loop
Strength Reduction
   0.0009 (  0.0%)   0.0010 (  1.0%)   0.0019 (  0.0%)   0.0032 (  0.0%)  Live
Variable Analysis
   0.0040 (  0.0%)   0.0000 (  0.0%)   0.0040 (  0.0%)   0.0027 (  0.0%) 
Two-Address instruction pass
   0.0009 (  0.0%)   0.0000 (  0.0%)   0.0009 (  0.0%)   0.0023 (  0.0%) 
Machine Code Deleter
   0.0030 (  0.0%)   0.0019 (  2.0%)   0.0050 (  0.0%)   0.0013 (  0.0%) 
Prolog/Epilog Insertion & Frame Finalization
   0.0009 (  0.0%)   0.0010 (  1.0%)   0.0019 (  0.0%)   0.0005 (  0.0%)  Scalar
Evolution Analysis
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0004 (  0.0%)  ET
Forest Construction
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0004 (  0.0%)  ET
Forest Construction
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0004 (  0.0%) 
Immediate Dominators Construction
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0003 (  0.0%) 
Eliminate PHI nodes for register allocation
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0003 (  0.0%) 
Control Flow Optimizer
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0003 (  0.0%) 
Immediate Dominators Construction
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0003 (  0.0%)  Label
Folder
   0.0009 (  0.0%)   0.0000 (  0.0%)   0.0009 (  0.0%)   0.0003 (  0.0%)  X86 FP
Stackifier
   0.0009 (  0.0%)   0.0000 (  0.0%)   0.0009 (  0.0%)   0.0003 (  0.0%) 
Dominator Set Construction
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0003 (  0.0%) 
Natural Loop Construction
   0.0010 (  0.0%)   0.0000 (  0.0%)   0.0010 (  0.0%)   0.0002 (  0.0%) 
Natural Loop Construction
   0.0009 (  0.0%)   0.0000 (  0.0%)   0.0009 (  0.0%)   0.0002 (  0.0%) 
Dominator Tree Construction
   0.0009 (  0.0%)   0.0000 (  0.0%)   0.0009 (  0.0%)   0.0002 (  0.0%)  Remove
unreachable blocks from the CFG
   0.0000 (  0.0%)   0.0010 (  1.0%)   0.0010 (  0.0%)   0.0002 (  0.0%) 
Canonicalize natural loops
   0.0000 (  0.0%)   0.0010 (  1.0%)   0.0010 (  0.0%)   0.0002 (  0.0%)  Lower
invoke and unwind, for unwindless code generators
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0001 (  0.0%)  Lower
GC intrinsics, for GCless code generators
   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0001 (  0.0%)  Target
Data Layout
  553.5438 (100.0%)   0.0979 (100.0%)  553.6418 (100.0%)  554.1256 (100.0%)  TOTAL

This is something like 30% from all compilation time of Mozilla here.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 28 08:23:03 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Mar 2007 08:23:03 -0500
Subject: [LLVMbugs] [Bug 1279] another assertion failed in RegisterScavenging
Message-ID: <200703281323.l2SDN30S004861@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1279

lauro.venancio at gmail.com changed:

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



------- Additional Comments From lauro.venancio at gmail.com  2007-03-28 08:23 -------
This assertion remains failing in other testcase.

laurov at edgy-laptop:~$ llc bugpoint-reduced-simplified2.bc 
llc: /home/laurov/llvm/llvm/lib/CodeGen/RegisterScavenging.cpp:141: void
llvm::RegScavenger::forward(): Assertion `isUnused(Reg) || isReserved(Reg)' failed.
...



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 28 09:31:04 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Mar 2007 09:31:04 -0500
Subject: [LLVMbugs] [Bug 1283] NEW: Something broke nightly tester (no rule
	to build target)
Message-ID: <200703281431.l2SEV4x8006206@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1283

           Summary: Something broke nightly tester (no rule to build target)
           Product: Test Suite
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Nightly Tester
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


Last night's run of my nightly test yielded errors like:

*** No rule to make target `Output/sumarray-dbl.linked.rbc', needed by
`Output/sumarray-dbl.linked.bc'

for every single test program. The consequence was no results. The only change
to the makefile system was this:

--- Makefile.programs   21 Mar 2007 00:19:32 -0000      1.263
+++ Makefile.programs   28 Mar 2007 08:31:17 -0000      1.264
@@ -221,12 +221,11 @@ LLCBETAOPTION := -sched=list-td
 endif
 ifeq ($(ARCH),IA64)
 LLCBETAOPTION := -sched=simple
 endif
 ifeq ($(ARCH),x86)
-LLCBETAOPTION := -enable-rematerialization
-# -enable-tail-merge
+LLCBETAOPTION := -enable-tail-merge
 #-regalloc=local -fast
 endif
 ifeq ($(ARCH),Sparc)
 LLCBETAOPTION := -enable-sparc-v9-insts
 endif

But, I don't see how that could cause this problem.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 28 11:45:30 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Mar 2007 11:45:30 -0500
Subject: [LLVMbugs] [Bug 1284] NEW: Support bitwidth attribute in llvm-gcc
Message-ID: <200703281645.l2SGjUht008663@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1284

           Summary: Support bitwidth attribute in llvm-gcc
           Product: tools
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


The llvm-gcc tool should support a bitwidth attribute that permits arbitrary
precision bit width integers to be used. This PR is to track the development of
that feature. 

As with other type attributes, this feature requests that:

__attribute__((bitwidth(N)))

be recognized by llvm-gcc so that things like:

typedef int __attribute__((bitwidth(23))

could be used to make a signed 23 bit integer.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 28 12:54:32 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Mar 2007 12:54:32 -0500
Subject: [LLVMbugs] [Bug 1285] NEW: Fix consistency of FP operators
Message-ID: <200703281754.l2SHsWVN010024@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1285

           Summary: Fix consistency of FP operators
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


As part of the signless types work, some FP operators were split out from the integer counterparts.  
Because of this, we have fdiv/frem/fcmp, but some FP ops are still overloaded with integer ones (add/sub/
mul).

We should introduce explicit fadd/fsub/fmul operations, as well as a new unary fneg operator.  This will 
be important going forward, because we want to add flags to the various fp operations to support 
"fastmath" and more FP pedantic languages.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 28 13:39:00 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Mar 2007 13:39:00 -0500
Subject: [LLVMbugs] [Bug 1282] Extremly slow codegen
Message-ID: <200703281839.l2SId0aY011047@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1282

asl at math.spbu.ru changed:

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



------- Additional Comments From asl at math.spbu.ru  2007-03-28 13:38 -------
Fixed with.

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046388.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 28 14:07:06 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Mar 2007 14:07:06 -0500
Subject: [LLVMbugs] [Bug 1286] NEW: SSE Code is Horrid!
Message-ID: <200703281907.l2SJ76rh011609@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1286

           Summary: SSE Code is Horrid!
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: isanbard at gmail.com


This:

#include <xmmintrin.h>

void foo(__m128i *A, __m128i *B) {
  *A = _mm_sll_epi16 (*A, *B);
}

generates this code:

_foo:
        movl 8(%esp), %eax
        movdqa (%eax), %xmm0
        #IMPLICIT_DEF %eax
	pinsrw $2, %eax, %xmm0
        xorl %ecx, %ecx
	pinsrw $3, %ecx, %xmm0
        pinsrw $4, %eax, %xmm0
        pinsrw $5, %ecx, %xmm0
	pinsrw $6, %eax, %xmm0
        pinsrw $7, %ecx, %xmm0
	movl 4(%esp), %eax
        movdqa (%eax), %xmm1
        psllw %xmm0, %xmm1
	movdqa %xmm1, (%eax)
        ret

This is obviously bad! We should be as good GCC at least.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Wed Mar 28 14:16:07 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 Mar 2007 14:16:07 -0500
Subject: [LLVMbugs] [Bug 1287] NEW: llc can't compile llvm.powi.f64 when f64
	is a illegal type
Message-ID: <200703281916.l2SJG7nK011770@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1287

           Summary: llc can't compile llvm.powi.f64 when f64 is a illegal
                    type
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: lauro.venancio at gmail.com


laurov at edgy-laptop:/tmp$ llc fpowi.bc 
NODE: 0x8a09a70: f64 = fpowi 0x8a09a08, 0x8a09948:2
llc: /home/laurov/llvm/llvm/lib/CodeGen/SelectionDAG/LegalizeDAG.cpp:4690:
void<unnamed>::SelectionDAGLegalize::ExpandOp(llvm::SDOperand, llvm::SDOperand&,
llvm::SDOperand&): Assertion `0 && "Do not know how to expand this operator!"'
failed.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar 29 04:36:11 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Mar 2007 04:36:11 -0500
Subject: [LLVMbugs] [Bug 1288] NEW: Not so hot code produced for Ada array
	traversal
Message-ID: <200703290936.l2T9aBV9004285@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1288

           Summary: Not so hot code produced for Ada array traversal
           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


The attached .ll file contains unoptimized code produced by the Ada front-end
for the operation "loop over all elements of an array of ints, setting them to
zero", for several different array types.  The arrays are all of fixed size,
and differ only in the type of the indexing variable:
@sbytezero: zero an array indexed by a signed i8 (index range -128..127)
@ubytezero: zero an array indexed by an unsigned i8 (index range 0..255)
@sintzero: zero an array indexed by a signed i32
@uintzero: zero an array indexed by an unsigned i32
@srangezero: zero an array with index range -10..10
@urangezero: zero an array with index range 10..30

Thus all of these routines could in theory be optimized to a single memset.
That doesn't happen, but let's not ask for the moon!  They do get fairly
well optimized but some optimizations are missed.  I think this matters
because these kind of array traversals (only more complicated) come up all
the time in Ada code.  The main problems seem to come from code like this
(extracted from @sbytezero):
        %tmp = load i8* %i              ; <i8> [#uses=1]
        %tmp1 = sext i8 %tmp to i32             ; <i32> [#uses=5]
        %tmp3 = sub i32 %tmp1, -128             ; <i32> [#uses=1]
        %tmp4 = getelementptr [256 x i32]* %tmp2, i32 0, 
i32 %tmp3              ; <i32*> [#uses=1]
Here %i is the array index variable.  It runs from -128 to 127.  The sext to
i32 is generated by the Ada f-e, presumably to avoid overflow when subtracting
off the array lower bound of -128, which occurs in the next line.  This gets
optimized to:
        %tmp8 = add i8 %indvar, -127            ; <i8> [#uses=1]
        %tmp115 = sext i8 %tmp8 to i32          ; <i32> [#uses=1]
        %tmp316 = add i32 %tmp115, 128          ; <i32> [#uses=1]
        %tmp417 = getelementptr [256 x i32]* %a, i32 0, 
i32 %tmp316             ; <i32*> [#uses=1]
Here %indvar runs from 0 to 254 (nice optimization!), which really corresponds
to indices 1..255 in the [256 x i32] array (the first value was peeled off).
Thus %tmp8 is between -127 and 127.  Thus %tmp115 is also between -127 and 127.
Thus %tmp316 is between 1 and 255.  It is just %tmp8+1 zextended to i32.  Thus
the adding of -127 and 128 could have be turned into a single addition of 1.
In fact it would have been better not to peel off the first loop value and
have %indvar go from 0..255, then just zextend it for use in the getelementptr.

PS: there are a bunch of pointless compares in the unoptimized code, like
        icmp sle i32 -128, %tmp1                ; <i1>:1 [#uses=0]
        icmp sle i32 %tmp1, 127         ; <i1>:2 [#uses=0]
This is me playing with assertion generation in llvm-convert, i.e. one day
these may become
        iassert sle i32 -128, %tmp1
        iassert sle i32 %tmp1, 127
if we introduce an assert instruction.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar 29 10:40:51 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Mar 2007 10:40:51 -0500
Subject: [LLVMbugs] [Bug 1283] Something broke nightly tester (no rule to
	build target)
Message-ID: <200703291540.l2TFepE0011232@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1283

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-29 10:40 -------
This problem is now fixed. The configuration test for llvm-gcc includes a sanity
check that was looking for the "implementation" keyword. That keyword just got
removed from llvm. The fix is to make the sanity check look for "target
datalayout" which is probably more distinct anyway.

Patch here:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046405.html



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar 29 14:01:44 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Mar 2007 14:01:44 -0500
Subject: [LLVMbugs] [Bug 1289] NEW: Drop "implementation" keyword in
	AsmParser
Message-ID: <200703291901.l2TJ1iDi015610@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1289

           Summary: Drop "implementation" keyword in AsmParser
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: minor
          Priority: P2
         Component: LLVM assembly language parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


Before 2.0 ships, we need to drop support for parsing the implementation keyword
in AsmParser. The AsmWriter no longer generates it. It isn't needed. We also
don't want "checkpoint". Instead we do want the AsmParser to check for
incomplete types and values upon EOF. So, just move the checks done by
"implementation" into the code at the end of the parse. When this is done, the
test case:
  test/Assembler/2003-12-30-TypeMapInvalidMemory.llx
will need to be updated to drop use of "implementation".



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar 29 14:14:24 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Mar 2007 14:14:24 -0500
Subject: [LLVMbugs] [Bug 789] sys::Path enhancements
Message-ID: <200703291914.l2TJEOcs016021@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=789

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-29 14:14 -------
With these patches:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046438.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046446.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046448.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046449.html

all requested functionality for this PR has been implemented.




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Thu Mar 29 14:45:36 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 Mar 2007 14:45:36 -0500
Subject: [LLVMbugs] [Bug 1290] NEW: linklint reports many problems with
	llvm.org
Message-ID: <200703291945.l2TJja8K016760@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1290

           Summary: linklint reports many problems with llvm.org
           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


See attached error reports.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 01:43:10 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 01:43:10 -0500
Subject: [LLVMbugs] [Bug 1291] NEW: sys::Path enhancement: add PathWithStatus
Message-ID: <200703300643.l2U6hAGM028513@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1291

           Summary: sys::Path enhancement: add PathWithStatus
           Product: libraries
           Version: 1.0
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: trivial
          Priority: P2
         Component: System Library
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


Getting the status of a sys::path curently requires new'ing the status object.  It would be nice if we 
could make a subclass PathWithStatus, and move the 'status' related stuff there.  This would give us 
something like this:

class PathWithStatus : public Path {
   Status S;
   bool IsStatusValid;
}

If you sink the 'status gathering' methods out of Path into  
PathWithStatus, then you let the client decide whether they want to  
store a status or not, and if they want it, they don't have to pay for a separate malloc.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 10:30:27 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 10:30:27 -0500
Subject: [LLVMbugs] [Bug 1292] NEW: lli barfs on trivial main program with
	argc, argv, envp args
Message-ID: <200703301530.l2UFURKM017029@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1292

           Summary: lli barfs on trivial main program with argc,argv,envp
                    args
           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


define i32 @main(i32 %argc, i32 %argv, i32 %envp) {
        ret i32 0
}

$ lli m.bc
lli: lib/VMCore/Instructions.cpp:205: void llvm::CallInst::init(llvm::Value*, 
llvm::Value* const*, unsigned int): Assertion `(i >= FTy->getNumParams() || 
FTy->getParamType(i) == Params[i]->getType()) && "Calling a function with a bad 
signature!"' failed



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 10:51:06 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 10:51:06 -0500
Subject: [LLVMbugs] [Bug 1293] NEW: lli -force-interpreter=true crash at
	APInt::operator==
Message-ID: <200703301551.l2UFp64G017431@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1293

           Summary: lli -force-interpreter=true crash at APInt::operator==
           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


$ lli -force-interpreter=true ./apint.bc
lli: lib/Support/APInt.cpp:491: bool llvm::APInt::operator==(const 
llvm::APInt&) const: Assertion `BitWidth == RHS.BitWidth && "Comparison 
requires equal bit widths"' failed.

I couldn't get bugpoint to reduce the testcase, so I'm attaching
the compressed bytecode in all its massiveness.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 11:48:34 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 11:48:34 -0500
Subject: [LLVMbugs] [Bug 1293] lli -force-interpreter=true crash at
	APInt::operator==
Message-ID: <200703301648.l2UGmY8i019548@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1293

rspencer at x10sys.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
           Keywords|                            |code-cleanup
         Resolution|                            |FIXED



------- Additional Comments From rspencer at x10sys.com  2007-03-30 11:48 -------
Fixed with this patch:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046510.html




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 11:51:28 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 11:51:28 -0500
Subject: [LLVMbugs] [Bug 1292] lli barfs on program with malformed argv,
	envp args
Message-ID: <200703301651.l2UGpSiD019903@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1292

rspencer at x10sys.com changed:

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



------- Additional Comments From rspencer at x10sys.com  2007-03-30 11:51 -------
The FE should generate the right prototypes.

Closing this as WONTFIX.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 13:25:03 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 13:25:03 -0500
Subject: [LLVMbugs] [Bug 1294] NEW: ELF Writer Support
Message-ID: <200703301825.l2UIP3ba022265@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1294

           Summary: ELF Writer Support
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: christopher.lamb at gmail.com
                CC: christopher.lamb at gmail.com


This is a bug to track work on the target independent ELF file writer.
 
The initial work is slated to allow generation of ELF REL files based on a target specific configuration.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 15:20:19 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 15:20:19 -0500
Subject: [LLVMbugs] [Bug 1279] another assertion failed in RegisterScavenging
Message-ID: <200703302020.l2UKKJTV024762@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1279

evan.cheng at apple.com changed:

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



------- Additional Comments From evan.cheng at apple.com  2007-03-30 15:20 -------
Fixed.

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070326/046517.html




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 15:47:33 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 15:47:33 -0500
Subject: [LLVMbugs] [Bug 1295] NEW: Benchmarks/MiBench/consumer-typeset is
	set up wrong
Message-ID: <200703302047.l2UKlXOL025378@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1295

           Summary: Benchmarks/MiBench/consumer-typeset is set up wrong
           Product: Test Suite
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Programs Tests
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


I'm not sure what is going on here, but it looks like something is wrong with consumer-typeset.  Before 
Jeff C's change last night, the jit passed on darwin/x86 (but only because it ran first usually), now nothing 
does.

I don't think there is anything wrong with what he did, but I think an input file or output file or something 
is getting clobbered when one instance runs, messing up later runs of the test.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 20:18:19 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 20:18:19 -0500
Subject: [LLVMbugs] [Bug 1296] NEW: poor codegen for code size
Message-ID: <200703310118.l2V1IJVB030497@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1296

           Summary: poor codegen for code size
           Product: libraries
           Version: 1.0
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org


This contrived testcase:

int foo(int A, int B, int C) {
  int i;
  int result = 0;

  if (A == 1)  {
    for (i = 0; i < 1000; ++i) {
      if (i & B) {result = 1; break; }
    }
  } else if (A == 0) {
    for (i = 0; i < 1000; ++i) {
      if (i & C) {result = 1; break; }
   }
  } else if (A == 2) {
    for (i = 0; i < 1000; ++i) {
      if (i & C) {result = 1; break; }
    }
  }

out:
  return result;
}

Produces silly code like this:

LBB1_19:        #bb13.out.loopexit1_crit_edge
        movl $1, %eax
        jmp LBB1_15     #out
LBB1_20:        #bb.out.loopexit_crit_edge
        movl $1, %eax
        jmp LBB1_15     #out
LBB1_21:        #bb27.out.loopexit3_crit_edge
        movl $1, %eax
        jmp LBB1_15     #out



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Fri Mar 30 21:40:00 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 Mar 2007 21:40:00 -0500
Subject: [LLVMbugs] [Bug 1297] NEW: Support Overloaded Intrinsic Functions
Message-ID: <200703310240.l2V2e0rk032078@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1297

           Summary: Support Overloaded Intrinsic Functions
           Product: tools
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: TableGen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rspencer at x10sys.com


With there now being up to 8 million integer types, we need the ability to
overload intrinsic functions by the type of their arguments. To this end, this
PR tracks work that I'm doing to:

1. Create a new MVT::ValueType, iAny which is a place holder for an integer type
   of any bit width
2. Use iAny in the definition of "llvm_int_ty" in Intrinsics.td
3. Use llvm_int_ty in the definition of overloaded intrinsics that work on a 
   variety of integer bit widths. For example llvm.ctpop will now work on any
   integer bit width.
4. Convert existing intrinsics that should be overloaded to use the overloading
   mechanisms (bswap, ctpop, cttz, ctlz).
5. Enhance TableGen to deal with all of the above, especially the generation of
   the recognizers where only the prefix should be matched for overloaded
   intrinsics.
6. Implement new intrinsics to support bit manipulation of integer types:
     bit_select (select one bit from many) 
     bit_set (set a bit in an integer of arbitrary width)
     bit_part_set (set a range of bits in an arbitrary integer)
     bit_part_select (select a set of bits from an integer type)
     bit_concat (concat two integer values of arbitrary width)
     bit_and_reduce (reduce an iAny to i1 by applying AND bitwise)
     bit_or_reduce (reduce an iAny to i1 by applying OR bitwise)
     bit_xor_reduce (reduce an iAny to i1 by applying XOR bitwise)
     bit_nand_reduce (reduce an iAny to i1 by applying NAND bitwise)
     bit_nor_reduce (reduce an iAny to i1 by applying NOR bitwise)
     bit_nxor_reduce (reduce an iAny to i1 by applying NXOR bitwise)
7. Make the Verifier check the names of overloaded intrinsics to ensure that
   the suffix is of the correct form (. followed by the actual type for each
   iAny type).
8. Implement lowering of the overloaded intrinsics so that they can be executed
   regardless of whether a target supports lowering them itself (to an
   instruction).  Targets (and CBE and LLI) can decide which of these intrinsics
   they wish to support directly. The others will be lowered to internal 
   functions of a similar name (. -> _) with bodies defined in LLVM IR. These
   function bodies would then need to be code generated. Only those overloadings
   actually used (called) will be generated.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 31 20:35:30 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Mar 2007 20:35:30 -0500
Subject: [LLVMbugs] [Bug 935] Loop optimization deficiencies
Message-ID: <200704010135.l311ZUhg001604@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=935

sabre at nondot.org changed:

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



------- Additional Comments From sabre at nondot.org  2007-03-31 20:35 -------
Actually, the ppc code is fine.  It would be nice if the loop was rotated, but it doesn't actually impact 
codegen in this case.

-Chris



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

From bugzilla-daemon at cs.uiuc.edu  Sat Mar 31 22:54:35 2007
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 Mar 2007 22:54:35 -0500
Subject: [LLVMbugs] [Bug 1298] NEW: ARM bad code: constants materialization
	inside loops
Message-ID: <200704010354.l313sZ3p004068@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1298

           Summary: ARM bad code: constants materialization inside loops
           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: lauro.venancio at gmail.com


code generated for sieve test:

        mov r2, #0
        ldr r1, .LCPI1_0
.LBB1_3:        @bb7
        add r0, r1, r2
        mov r12, #1
        strb r12, [r0, #+2]
        add r2, r2, #1
        mov r0, #255
        orr r0, r0, #31, 24 @ 7936
        cmp r2, r0
        bne .LBB1_3     @bb7

can be:

        mov r2, #0
        ldr r1, .LCPI1_0
        mov r12, #1
        mov r0, #255
        orr r0, r0, #31, 24 @ 7936
.LBB1_3:        @bb7
        add r3, r1, r2
        strb r12, [r3, #+2]
        add r2, r2, #1
        cmp r2, r0
        bne .LBB1_3     @bb7



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.