From bugzilla-daemon at cs.uiuc.edu  Thu May  1 01:46:33 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 1 May 2008 01:46:33 -0500
Subject: [LLVMbugs] [Bug 2272] New: Re-enable inlining at -O2 at above
Message-ID: <bug-2272-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2272

           Summary: Re-enable inlining at -O2 at above
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Interprocedural Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: evan.cheng at apple.com
                CC: llvmbugs at cs.uiuc.edu


We've disabled inlining for llvm-gcc optimization level < 3 for now. There are
several reasons for this:
1. This matches gcc behavior so it would not be a surprise to users.
2. Some projects with buggy code, e.g. misuses of inline asm, expect to build
correctly at -O2 when inlining is off. We'll encourage users fixed these
projects.
3. Some gcc testsuites depend on this -O2 behavior. Some tests would otherwise
expose llvm bugs are masked by inlining. We should fix the testsuites.
4. llvm inlining default is too aggressive. At -O2, it should not turn on
optimization that will increase code size.

We should re-enable inlining at -O2 but with inlining threshold turned down.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  1 13:56:36 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 1 May 2008 13:56:36 -0500
Subject: [LLVMbugs] [Bug 2269] Assertion `DestReg == X86::ST0 && " Can only
 copy to TOS right now"' failed.
In-Reply-To: <bug-2269-206@http.llvm.org/bugs/>
Message-ID: <200805011856.m41IuaGQ014663@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2269


T??r??k Edwin <edwintorok at gmail.com> changed:

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




--- Comment #1 from T??r??k Edwin <edwintorok at gmail.com>  2008-05-01 13:56:36 ---
This looks like the same bug, except the type is x86_fp80 instead of double.

*** This bug has been marked as a duplicate of bug 2255 ***


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  1 14:14:20 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 1 May 2008 14:14:20 -0500
Subject: [LLVMbugs] [Bug 2271] bug?
In-Reply-To: <bug-2271-206@http.llvm.org/bugs/>
Message-ID: <200805011914.m41JEKXC015350@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2271


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #6 from Dan Gohman <gohman at apple.com>  2008-05-01 14:14:19 ---
This is fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080428/061860.html


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  1 16:41:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 1 May 2008 16:41:12 -0500
Subject: [LLVMbugs] [Bug 2266] semantic analyzer assertion on malformed
 code: " The next DeclContext should be directly contained in the current
 one."
In-Reply-To: <bug-2266-206@http.llvm.org/bugs/>
Message-ID: <200805012141.m41LfCYc020385@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2266


Ted Kremenek <kremenek at cs.stanford.edu> changed:

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




--- Comment #1 from Ted Kremenek <kremenek at cs.stanford.edu>  2008-05-01 16:41:12 ---
Fixed by Argiris in r50538:

http://llvm.org/viewvc/llvm-project?rev=50538&view=rev


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  2 00:04:10 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 May 2008 00:04:10 -0500
Subject: [LLVMbugs] [Bug 2273] New: Segmentation Fault in PassManager.cpp
Message-ID: <bug-2273-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2273

           Summary: Segmentation Fault in PassManager.cpp
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Core LLVM classes
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: shackan at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1613)
 --> (http://llvm.org/bugs/attachment.cgi?id=1613)
test case

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb7d016c0 (LWP 5846)]
0x082483ed in llvm::PMStack::top (this=0x8371ce0) at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/include/llvm/PassManagers.h:130
130       inline PMDataManager *top() { return S.back(); }
(gdb) bt
#0  0x082483ed in llvm::PMStack::top (this=0x8371ce0) at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/include/llvm/PassManagers.h:130
#1  0x082473b3 in llvm::CallGraphSCCPass::assignPassManager (this=0x8371ff8,
PMS=@0x8371ce0, PreferredType=llvm::PMT_Unknown)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/Analysis/IPA/CallGraphSCCPass.cpp:161
#2  0x082c69f8 in llvm::FunctionPassManagerImpl::addTopLevelPass
(this=0x8371c70, P=0x8371ff8)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:166
#3  0x082bca43 in llvm::PMTopLevelManager::schedulePass (this=0x8371cdc,
P=0x8371ff8)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:456
#4  0x082bc9eb in llvm::PMTopLevelManager::schedulePass (this=0x8371cdc,
P=0x8371c30)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:449
#5  0x082c6a37 in llvm::FunctionPassManagerImpl::add (this=0x8371c70,
P=0x8371c30)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:134
#6  0x082bcbcd in llvm::MPPassManager::addLowerLevelRequiredPass
(this=0x8371b20, P=0x8371bb8, RequiredPass=0x8371c30)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:1283
#7  0x082be01c in llvm::PMDataManager::add (this=0x8371b38, P=0x8371bb8,
ProcessAnalysis=true)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:742
#8  0x082be4a2 in llvm::ModulePass::assignPassManager (this=0x8371bb8,
PMS=@0x8371888, PreferredType=llvm::PMT_Unknown)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:1453
#9  0x082c6902 in llvm::PassManagerImpl::addTopLevelPass (this=0x8371818,
P=0x8371bb8)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:298
#10 0x082bca43 in llvm::PMTopLevelManager::schedulePass (this=0x8371884,
P=0x8371bb8)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:456
#11 0x082c6a19 in llvm::PassManagerImpl::add (this=0x8371818, P=0x8371bb8)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:274
#12 0x082bca6f in llvm::PassManager::add (this=0xbfa26638, P=0x8371bb8)
    at
/mnt/shared/repos/svn/llvm/llvm-project/llvm/trunk/lib/VMCore/PassManager.cpp:1347
#13 0x081c4f40 in main () at dummy/main.cpp:86


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  2 11:52:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 May 2008 11:52:47 -0500
Subject: [LLVMbugs] [Bug 2274] New: another possible code generation error
Message-ID: <bug-2274-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2274

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


Using recent svn, targeting x86-32, running on Ubuntu Feisty, this program
produces different output at -O0 and -O1:

#include <stdio.h>

static unsigned long
div_rhs (long rhs)
{
  if (rhs == 0)
    return 1;
  return rhs;
}

void func_38 (unsigned int p_39);

void func_38 (unsigned int p_39)
{
  unsigned long tmp___0 = div_rhs (-1583256716);
   if (p_39 / tmp___0)
    {
      printf ("foo\n");
    }
  else
    {
      printf ("bar\n");
    }
}

int main (void)
{
  func_38 (-1);
  return 0;
}


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  2 12:48:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 May 2008 12:48:23 -0500
Subject: [LLVMbugs] [Bug 2275] New: assertion failure with C++ namespaces
Message-ID: <bug-2275-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2275

           Summary: assertion failure with C++ namespaces
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: kremenek at cs.stanford.edu
                CC: llvmbugs at cs.uiuc.edu


The semantic analyzer crashes on the following input (also attached):

  long long f(long long);

  namespace __gnu_cxx {
    inline long long   f(long long __x) {}
  }

What we see:

$ clang t.cpp

Assertion failed: (( (isa<ObjCMethodDecl>(DC) &&
isa<TranslationUnitDecl>(CurContext)) || DC->getParent() == CurContext ) &&
"The next DeclContext should be directly contained in the current one."),
function PushDeclContext, file SemaDecl.cpp, line 48.
0   clang                               0x0029b785 _ZNK4llvm3sys4PathneERKS1_ +
79
1   clang                               0x0029b981
_ZN4llvm3sys28PrintStackTraceOnErrorSignalEv + 387
2   libSystem.B.dylib                   0x948dd5eb _sigtramp + 43
3   ???                                 0xffffffff 0x0 + 4294967295
4   libSystem.B.dylib                   0x949563f2 raise + 26
5   libSystem.B.dylib                   0x949659af abort + 73
6   libSystem.B.dylib                   0x94957593 __assert_rtn + 101
7   clang                               0x000e3f7f
_ZN5clang4Sema15PushDeclContextEPNS_11DeclContextE + 135
8   clang                               0x000e43de
_ZN5clang4Sema23ActOnStartOfFunctionDefEPNS_5ScopeERNS_10DeclaratorE + 1104
9   clang                               0x00174f45
_ZN5clang6Parser23ParseFunctionDefinitionERNS_10DeclaratorE + 505
10  clang                               0x001754ab
_ZN5clang6Parser36ParseDeclarationOrFunctionDefinitionEv + 1333
11  clang                               0x00175605
_ZN5clang6Parser24ParseExternalDeclarationEv + 85
12  clang                               0x0016441a
_ZN5clang6Parser14ParseNamespaceEj + 406
13  clang                               0x001629ba
_ZN5clang6Parser16ParseDeclarationEj + 66
14  clang                               0x001757d6
_ZN5clang6Parser24ParseExternalDeclarationEv + 550
15  clang                               0x00175823
_ZN5clang6Parser17ParseTopLevelDeclERPv + 61
16  clang                               0x000d9247
_ZN5clang8ParseASTERNS_12PreprocessorEPNS_11ASTConsumerEb + 339
17  clang                               0x000297d8
_Z16InitializeGCModeRN5clang11LangOptionsE + 5620
18  clang                               0x0002c30b main + 1345
19  clang                               0x00002356 start + 54

My version of clang:

(kremenek at grue:clang)$ svn info
Path: .
URL: https://kremenek at llvm.org/svn/llvm-project/cfe/trunk
Repository Root: https://kremenek at llvm.org/svn/llvm-project
Repository UUID: 91177308-0d34-0410-b5e6-96231b3b80d8
Revision: 50576
Node Kind: directory
Schedule: normal
Last Changed Author: kremenek
Last Changed Rev: 50557
Last Changed Date: 2008-05-01 16:38:35 -0700 (Thu, 01 May 2008)


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  2 16:32:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 May 2008 16:32:30 -0500
Subject: [LLVMbugs] [Bug 2274] another possible code generation error
In-Reply-To: <bug-2274-206@http.llvm.org/bugs/>
Message-ID: <200805022132.m42LWUdD009926@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2274


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #1 from Dan Gohman <gohman at apple.com>  2008-05-02 16:32:30 ---
This is fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080428/061896.html


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  2 17:14:54 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 May 2008 17:14:54 -0500
Subject: [LLVMbugs] [Bug 2273] Segmentation Fault in PassManager.cpp
In-Reply-To: <bug-2273-206@http.llvm.org/bugs/>
Message-ID: <200805022214.m42MEshX011312@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2273


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #1 from Devang Patel <dpatel at apple.com>  2008-05-02 17:14:53 ---
I added assert message.

A function pass, that is required by a module pass, can not require SCC pass.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  2 23:20:33 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 2 May 2008 23:20:33 -0500
Subject: [LLVMbugs] [Bug 2276] New: probable integer codegen bug
Message-ID: <bug-2276-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2276

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


Using recent svn targeting x86-32 running on Ubuntu Linux this program behaves
differently at -O0 and -O1:

int g_91 = 1;

void func_64 (void);
void func_64 (void)
{
  long tmp = 1;
  if ((1 | g_91) % tmp)
    {
      printf ("foo\n");
    }
  else
    {
      printf ("bar\n");
    }
}
int
main (void)
{
  func_64 ();
  return (0);
}


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

From bugzilla-daemon at cs.uiuc.edu  Sat May  3 14:56:31 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 May 2008 14:56:31 -0500
Subject: [LLVMbugs] [Bug 2277] New: llc -enable-correct-eh-support: " Not
 all catch info was assigned to a landing pad!"' failed
Message-ID: <bug-2277-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2277

           Summary: llc -enable-correct-eh-support: "Not all catch info was
                    assigned to a landing pad!"' failed
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1614)
 --> (http://llvm.org/bugs/attachment.cgi?id=1614)
bugpoint reduced testcase

On SVN r50604 I get assertion failure with llc -enable-correct-eh-support in
the testsuite.

$ ~/llvm-svn/llvm/Release/bin/llc -enable-correct-eh-support
bugpoint-reduced-simplified.bc -f
llc: SelectionDAGISel.cpp:4556: virtual bool
llvm::SelectionDAGISel::runOnFunction(llvm::Function&): Assertion
`FuncInfo.CatchInfoFound.size() == FuncInfo.CatchInfoLost.size() && "Not all
catch info was assigned to a landing pad!"' failed.
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x9b4351]
/lib/libc.so.6[0x3ccc631240]
/lib/libc.so.6(gsignal+0x35)[0x3ccc6311d5]
/lib/libc.so.6(abort+0x110)[0x3ccc632680]
/lib/libc.so.6(__assert_fail+0xef)[0x3ccc62a75f]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x73b7ae]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x4e3bf3]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm13FPPassManager13runOnFunctionERNS_8FunctionE+0x265)[0x95d975]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm23FunctionPassManagerImpl3runERNS_8FunctionE+0x75)[0x95de05]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm19FunctionPassManager3runERNS_8FunctionE+0x31)[0x95df71]
/home/edwin/llvm-svn/llvm/Release/bin/llc(main+0x989)[0x4d4539]
/lib/libc.so.6(__libc_start_main+0xf4)[0x3ccc61e1c4]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZNSt8ios_base4InitD1Ev+0x39)[0x4d1f19]
Aborted

This occurs 24 times on the testsuite:
$ grep "Not all catch info" report.nightly.txt |wc -l
24

$ uname -a
Linux lightspeed2 2.6.25-rc8 #5 Tue Apr 8 21:16:47 EEST 2008 x86_64 GNU/Linux

For example it fails on SingleSource/Benchmarks/ShootOut-C++/wordfreq

Bugpoint reduced testcase attached


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

From bugzilla-daemon at cs.uiuc.edu  Sat May  3 14:59:57 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 May 2008 14:59:57 -0500
Subject: [LLVMbugs] [Bug 2278] New: llc: Cannot yet select: 0x1d1ca60: i64 =
 GlobalTLSAddress < i32* @i.2609> 0
Message-ID: <bug-2278-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2278

           Summary: llc: Cannot yet select: 0x1d1ca60: i64 =
                    GlobalTLSAddress <i32* @i.2609> 0
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1615)
 --> (http://llvm.org/bugs/attachment.cgi?id=1615)
bugpoint reduced testcase

On x86-64 Linux with SVN r50604 I get a failure in the testsuite:

make[4]: Entering directory
`/home/edwin/llvm-svn/llvm/projects/llvm-test/SingleSource/UnitTests/Threads'
/home/edwin/llvm-svn/llvm/Release/bin/llc  -f Output/tls.llvm.bc -o
Output/tls.llc.s
Cannot yet select: 0x2c9e870: i64 = GlobalTLSAddress <i32* @i.2609> 0
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x9b4351]
/lib/libc.so.6[0x3ccc631240]
/lib/libc.so.6(gsignal+0x35)[0x3ccc6311d5]
/lib/libc.so.6(abort+0x110)[0x3ccc632680]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x591427]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x5bd591]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x5c11e6]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x5c16ab]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm16SelectionDAGISel17CodeGenAndEmitDAGERNS_12SelectionDAGE+0x100)[0x70fab0]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm16SelectionDAGISel16SelectBasicBlockEPNS_10BasicBlockERNS_15MachineFunctionERNS_20FunctionLoweringInfoE+0x310)[0x738940]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm16SelectionDAGISel13runOnFunctionERNS_8FunctionE+0x215)[0x73b465]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x4e3bf3]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm13FPPassManager13runOnFunctionERNS_8FunctionE+0x265)[0x95d975]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm23FunctionPassManagerImpl3runERNS_8FunctionE+0x75)[0x95de05]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm19FunctionPassManager3runERNS_8FunctionE+0x31)[0x95df71]
/home/edwin/llvm-svn/llvm/Release/bin/llc(main+0x989)[0x4d4539]
/lib/libc.so.6(__libc_start_main+0xf4)[0x3ccc61e1c4]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZNSt8ios_base4InitD1Ev+0x39)[0x4d1f19]
make[4]: *** [Output/tls.llc.s] Aborted

To reproduce:
$ ~/llvm-svn/llvm/Release/bin/llc bugpoint-reduced-simplified.bc
Cannot yet select: 0xf420f0: i64 = GlobalTLSAddress <i32* @i.2609> 0
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x9b4351]
/lib/libc.so.6[0x3ccc631240]
/lib/libc.so.6(gsignal+0x35)[0x3ccc6311d5]
/lib/libc.so.6(abort+0x110)[0x3ccc632680]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x591427]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x5bd591]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x5c11e6]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x5c16ab]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm16SelectionDAGISel17CodeGenAndEmitDAGERNS_12SelectionDAGE+0x100)[0x70fab0]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm16SelectionDAGISel16SelectBasicBlockEPNS_10BasicBlockERNS_15MachineFunctionERNS_20FunctionLoweringInfoE+0x310)[0x738940]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm16SelectionDAGISel13runOnFunctionERNS_8FunctionE+0x215)[0x73b465]
/home/edwin/llvm-svn/llvm/Release/bin/llc[0x4e3bf3]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm13FPPassManager13runOnFunctionERNS_8FunctionE+0x265)[0x95d975]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm23FunctionPassManagerImpl3runERNS_8FunctionE+0x75)[0x95de05]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZN4llvm19FunctionPassManager3runERNS_8FunctionE+0x31)[0x95df71]
/home/edwin/llvm-svn/llvm/Release/bin/llc(main+0x989)[0x4d4539]
/lib/libc.so.6(__libc_start_main+0xf4)[0x3ccc61e1c4]
/home/edwin/llvm-svn/llvm/Release/bin/llc(_ZNSt8ios_base4InitD1Ev+0x39)[0x4d1f19]
Aborted


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

From bugzilla-daemon at cs.uiuc.edu  Sat May  3 20:14:04 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 May 2008 20:14:04 -0500
Subject: [LLVMbugs] [Bug 2050] incorrect "error: initializer element is not
	constant"
In-Reply-To: <bug-2050-206@http.llvm.org/bugs/>
Message-ID: <200805040114.m441E4ok005611@zion.cs.uiuc.edu>

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


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-03 20:14:04 ---
Fixed, testcase added here:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080428/005537.html


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

From bugzilla-daemon at cs.uiuc.edu  Sat May  3 20:29:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 3 May 2008 20:29:34 -0500
Subject: [LLVMbugs] [Bug 2279] New: Conformance problems with pp expressions
Message-ID: <bug-2279-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2279

           Summary: Conformance problems with pp expressions
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: preprocessor
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


These are conformance problems I'm aware of regarding PP expressions.  If you
need a testcase I can provide, but they should be obvious from the
descriptions:

a) Comma subexpressions are permitted in C99 as long as they are unevaluated. 
Clang rejects them, which is correct behaviour for C90 but not C99 (see 6.6).
b) Clang doesn't follow the grammar for the 2nd operand of the conditional
operator.
c) Clang doesn't follow the rules for the signedness of the result of the shift
operators (no UAC).
d) Clang seems to have issues with unevaluated subexpressions.  Here are two
examples:

#if 0 ? 1/0: 2
#else
#error
#endif

#if 1 || 1/0
#else
#error
#endif


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 04:53:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 04:53:22 -0500
Subject: [LLVMbugs] [Bug 2280] New: clang: no output should be produced when
 errors occurs
Message-ID: <bug-2280-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2280

           Summary: clang: no output should be produced when errors occurs
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


When an error occurs, clang still creates the output file, which is a problem
because make will believe that compiling succeeded the 2nd time I run it.

$ make others.o
source='../../../trunk/clamscan/others.c' object='others.o' libtool=no \
        DEPDIR=.deps depmode=none /bin/sh ../../../trunk/config/depcomp \
        /home/edwin/llvm-svn/llvm/tools/clang/utils/ccc -DHAVE_CONFIG_H
-DCL_NOTHREADS -I. -I../../../trunk/clamscan -I..  -I../../../trunk
-I../../../trunk/shared -I../../../trunk/libclamav   -g -O2 -c
../../../trunk/clamscan/others.c
../../../trunk/clamscan/others.c
clang -emit-llvm-bc -x c -o others.o ../../../trunk/clamscan/others.c
-DHAVE_CONFIG_H -DCL_NOTHREADS -I. -I../../../trunk/clamscan -I..
-I../../../trunk -I../../../trunk/shared -I../../../trunk/libclamav
int (identifier)(char const *, short)
../../../trunk/clamscan/others.c:115:8: error: incompatible type passing 'int
*', expected '__WAIT_STATUS'
                wait(&status);
                     ^~~~~~~
1 diagnostic generated.
make: *** [others.o] Error 1

$ make others.o
make: `others.o' is up to date.

The generated file contains some functions, except the one where the failure
occured. In the end I get linker errors because of the missing function, and
make no longer tells me which file failed to compile.
A workaround is to 'make clean' each time, but I think it would be best to
delete the output file in case of an error.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 05:00:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 05:00:39 -0500
Subject: [LLVMbugs] [Bug 2281] New: Charconsts in pp expressions
Message-ID: <bug-2281-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2281

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


The following doesn't error with pedantic-errors, it just warns.  It fails
constraint 6.4.4.4p9.

#if '\x100'
#endif


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 12:17:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 12:17:12 -0500
Subject: [LLVMbugs] [Bug 2282] New: llvm-gcc build failure on x86 linux
 (@llvm.x86.sse.comile. ss)
Message-ID: <bug-2282-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2282

           Summary: llvm-gcc build failure on x86 linux
                    (@llvm.x86.sse.comile.ss)
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


The LLVM build fails on x86-32 and x86-64 linux:

cc1: llvm/lib/VMCore/Instructions.cpp:301: void
llvm::CallInst::init(llvm::Value                             *, llvm::Value*):
Assertion `(FTy->getNumParams() == 1 || (FTy->isVarArg() && FTy->getNumParams()
== 0)) &&                              "Calling a function with bad signature"'
failed.
gcc/config/i386/crtfastmath.c: In function ???set_fast_math???:
gcc/config/i386/crtfastmath.c:48: internal compiler error: Aborted

The problem is that this code (64 bit case):

  unsigned int mxcsr = __builtin_ia32_stmxcsr ();
  mxcsr |= MXCSR_DAZ | MXCSR_FTZ;
  __builtin_ia32_ldmxcsr (mxcsr);

is generating a call to

  declare i32 @llvm.x86.sse.comile.ss(<4 x float>, <4 x float>) nounwind
readnone

but is only passing one parameter to it:
  %tmp6 = bitcast i32* %memtmp5 to i8*            ; <i8*> [#uses=1]


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 12:24:04 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 12:24:04 -0500
Subject: [LLVMbugs] [Bug 2283] New: Problem compiling some GCC regtests at
	-O0
Message-ID: <bug-2283-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2283

           Summary: Problem compiling some GCC regtests at -O0
           Product: tools
           Version: 2.2
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu


>From PR915:

> Incidentally, many tests from the gcc testsuite fail at -O0, e.g.
> gcc.c-torture/20010605-1.c

This fires an assertion I have here locally that checks that
CreateTemporary is not used when the gcc type has variable
size.  So at first glance it looks to me like this isn't really
a problem with nested functions as such.  Can you please open
a new PR for this.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 13:37:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 13:37:19 -0500
Subject: [LLVMbugs] [Bug 2279] Conformance problems with pp expressions
In-Reply-To: <bug-2279-206@http.llvm.org/bugs/>
Message-ID: <200805041837.m44IbJxu009196@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2279


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-05-04 13:37:18 ---
I think these fix the rest of the issues, thanks Neil!
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080428/005542.html
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080428/005543.html

-Chris


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 16:28:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 16:28:03 -0500
Subject: [LLVMbugs] [Bug 2282] llvm-gcc build failure on x86 linux
 (@llvm.x86.sse.comile.ss)
In-Reply-To: <bug-2282-206@http.llvm.org/bugs/>
Message-ID: <200805042128.m44LS38N014882@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2282


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

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




--- Comment #5 from Anton Korobeynikov <asl at math.spbu.ru>  2008-05-04 16:28:02 ---
Erm, really fixed :)


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 17:55:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 17:55:34 -0500
Subject: [LLVMbugs] [Bug 1098] Incorrect post-dominators with multiple
	returns
In-Reply-To: <bug-1098-206@http.llvm.org/bugs/>
Message-ID: <200805042255.m44MtYVv017393@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1098


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #11 from Owen Anderson <resistor at mac.com>  2008-05-04 17:55:34 ---
Fixed in r50628.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 18:38:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 18:38:43 -0500
Subject: [LLVMbugs] [Bug 2279] Conformance problems with pp expressions
In-Reply-To: <bug-2279-206@http.llvm.org/bugs/>
Message-ID: <200805042338.m44Nchj2018586@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2279


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

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




--- Comment #4 from Neil Booth <neil at daikokuya.co.uk>  2008-05-04 18:38:43 ---
Still have issues:

a) It seems now you accept , operator you're doing the UAC there too.
b) Grammar still not right for ? :. 

#if 1 ? 2 ? 3 : 4 : 5
#endif

is rejected for example.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 18:52:21 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 18:52:21 -0500
Subject: [LLVMbugs] [Bug 2281] Charconsts in pp expressions
In-Reply-To: <bug-2281-206@http.llvm.org/bugs/>
Message-ID: <200805042352.m44NqLIB019127@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2281


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-04 18:52:21 ---
Fixed, thanks!
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080428/005548.html


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

From bugzilla-daemon at cs.uiuc.edu  Sun May  4 23:11:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 4 May 2008 23:11:28 -0500
Subject: [LLVMbugs] [Bug 2279] Conformance problems with pp expressions
In-Reply-To: <bug-2279-206@http.llvm.org/bugs/>
Message-ID: <200805050411.m454BSCx029210@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2279


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #5 from Chris Lattner <sabre at nondot.org>  2008-05-04 23:11:28 ---
Fixed:
Comma UACs:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080428/005546.html

?: parsing:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080428/005550.html


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 02:12:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 02:12:12 -0500
Subject: [LLVMbugs] [Bug 2284] New: Conditional operator issues
Message-ID: <bug-2284-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2284

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


The following is not diagnosed (excess tokens):

#if 1 ? 2 : 1, 3
#endif

"/tmp/bug.c", line 1: warning: excess tokens in preprocessing directive
#if 1 ? 2 : 1, 3
             ^

The following is diagnosed:

#if 1 ? 0: 1 ? 1/0: 1/0
#endif


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 07:47:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 07:47:47 -0500
Subject: [LLVMbugs] [Bug 2285] New: Jump threading crash with multiple
	return values
Message-ID: <bug-2285-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2285

           Summary: Jump threading crash with multiple return 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
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1618)
 --> (http://llvm.org/bugs/attachment.cgi?id=1618)
testcase .ll

Running opt -jump-threading on the attached testcase results in:

opt: lib/VMCore/Value.cpp:44: llvm::Value::Value(const llvm::Type*, unsigned
int): Assertion `(Ty->isFirstClassType() || Ty == Type::VoidTy ||
isa<OpaqueType>(ty)) && "Cannot create non-first-class values except for
constants!"' failed.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 08:11:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 08:11:38 -0500
Subject: [LLVMbugs] [Bug 2286] New: FrontendObjC/nonfragile-ivar.m fails on
	x86-32 linux
Message-ID: <bug-2286-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2286

           Summary: FrontendObjC/nonfragile-ivar.m fails on x86-32 linux
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


It looks like this test assumes a 64 bit compiler:

while running: llvm-gcc  -emit-llvm -x objective-c -c
test/FrontendObjC/nonfragile-ivar.m -o /dev/null -m64
test/FrontendObjC/nonfragile-ivar.m:1: sorry, unimplemented: 64-bit mode not
compiled in
test/FrontendObjC/nonfragile-ivar.m:16: error: synthesized property 'whatever'
must either be named the same as a compatible ivar or must explicitly name an
ivar


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 08:47:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 08:47:55 -0500
Subject: [LLVMbugs] [Bug 2287] New: CodeGen/X86/2007-07-03-GR64ToVR64.ll
	fails
Message-ID: <bug-2287-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2287

           Summary: CodeGen/X86/2007-07-03-GR64ToVR64.ll fails
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


The test CodeGen/X86/2007-07-03-GR64ToVR64.ll is failing (x86-64 linux).
The assembler produced is

        .text
        .align  16
        .globl  foo
        .type   foo, at function
foo:
.Leh_func_begin1:
.Llabel1:
        movq    8(%rsp), %mm0
        paddusw 16(%rsp), %mm0
        movq    %mm0, R
        emms
        ret
        .size   foo, .-foo
.Leh_func_end1:
        .section        .eh_frame,"aw", at progbits
EH_frame0:
.Lsection_eh_frame:
.Leh_frame_common:
        .long   .Leh_frame_common_end-.Leh_frame_common_begin
.Leh_frame_common_begin:
        .long   0x0
        .byte   0x1
        .asciz  "zR"
        .uleb128        1
        .sleb128        -8
        .byte   0x10
        .uleb128        1
        .byte   0x1b
        .byte   0xc
        .uleb128        7
        .uleb128        8
        .byte   0x90
        .uleb128        1
        .align  8
.Leh_frame_common_end:

foo.eh = 0

        .section        .note.GNU-stack,"", at progbits

This does not pass any of the tests
grep {movd        %rsi, %mm0}
grep {movd        %rdi, %mm1}
grep {paddusw     %mm0, %mm1}


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 09:11:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 09:11:53 -0500
Subject: [LLVMbugs] [Bug 2288] New: LLVM type size doesn't match GCC type
	size!
Message-ID: <bug-2288-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2288

           Summary: LLVM type size doesn't match GCC type size!
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


I see this assertion (x86-32 linux) for the following testcases:

C++Frontend/2007-04-05-PackedBitFieldsSmall.cpp
C++Frontend/2007-10-01-StructResize.cpp
CFrontend/2006-09-11-BitfieldRefCrash.c
CFrontend/2008-01-07-UnusualIntSize.c


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 11:34:04 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 11:34:04 -0500
Subject: [LLVMbugs] [Bug 1132] SelectionDAGISel doesn' t handle illegal
 vectors of illegal types right
In-Reply-To: <bug-1132-206@http.llvm.org/bugs/>
Message-ID: <200805051634.m45GY41q028292@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1132


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #4 from Dan Gohman <gohman at apple.com>  2008-05-05 11:34:03 ---
The support for copying to/from virtual registers is now more
flexible, and this testcase now works even on i386 with <2 x i64>.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 12:06:33 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 12:06:33 -0500
Subject: [LLVMbugs] [Bug 2289] New: llc assertion failure: Multiple
 implicit_def defining same register?
Message-ID: <bug-2289-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2289

           Summary: llc assertion failure: Multiple implicit_def defining
                    same register?
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1619)
 --> (http://llvm.org/bugs/attachment.cgi?id=1619)
testcase .ll

Running llc on the attached testcase gives:

llc: lib/CodeGen/SimpleRegisterCoalescing.cpp:766: void
llvm::SimpleRegisterCoalescing::RemoveCopiesFromValNo(llvm::LiveInterval&,
llvm::VNInfo*): Assertion `!ImpDef && "Multiple implicit_def defining same
register?"' failed.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 15:07:59 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 15:07:59 -0500
Subject: [LLVMbugs] [Bug 2284] Conditional operator issues
In-Reply-To: <bug-2284-206@http.llvm.org/bugs/>
Message-ID: <200805052007.m45K7xmG003192@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2284


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-05 15:07:58 ---
Fixed, thanks!
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080505/005566.html


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 15:22:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 15:22:11 -0500
Subject: [LLVMbugs] [Bug 2285] Jump threading crash with multiple return
	values
In-Reply-To: <bug-2285-206@http.llvm.org/bugs/>
Message-ID: <200805052022.m45KMBs9003750@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2285


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-05-05 15:22:09 ---
This fixes the call case:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/061979.html

I don't think invoke will be threaded across.  If you get a case where invoke
shows up, plz file a bug and I will fix it.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 20:03:51 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 20:03:51 -0500
Subject: [LLVMbugs] [Bug 2276] probable integer codegen bug
In-Reply-To: <bug-2276-206@http.llvm.org/bugs/>
Message-ID: <200805060103.m4613ppO014692@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2276


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #2 from Dan Gohman <gohman at apple.com>  2008-05-05 20:03:50 ---
This is now fixed, here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062004.html


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 20:36:18 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 20:36:18 -0500
Subject: [LLVMbugs] [Bug 1871] Need to remove IncludeFile
In-Reply-To: <bug-1871-206@http.llvm.org/bugs/>
Message-ID: <200805060136.m461aIxg015920@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1871


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #5 from Dan Gohman <gohman at apple.com>  2008-05-05 20:36:17 ---
This patch:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062006.html
removes all the IncludeFile.h uses for passes that don't use AnalysisGroups.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 20:55:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 20:55:32 -0500
Subject: [LLVMbugs] [Bug 2290] New: llvm-gcc doesn't bootstrap on arm-qemu
	(linux)
Message-ID: <bug-2290-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2290

           Summary: llvm-gcc doesn't bootstrap on arm-qemu (linux)
           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: nlewycky at google.com
                CC: llvmbugs at cs.uiuc.edu


llvm-gcc build dails with the error "Did not get a target machine!" which is
emitted by this block of code in gcc/llvm-backend.cpp:

  // Create the TargetMachine we will be generating code with.
  // FIXME: Figure out how to select the target and pass down subtarget info.
  std::string Err;
  const TargetMachineRegistry::entry *TME = 
    TargetMachineRegistry::getClosestStaticTargetForModule(*TheModule, Err);
  if (!TME) {
    cerr << "Did not get a target machine!\n";
    exit(1);
  }

I wasn't quite able to figure out how getClosestStaticTargetForModule matches
the target type. The GCC target triple is "armv5tejl-unknown-linux-gnu". uname
-a says:

Linux debian-arm 2.6.18-6-versatile #1 Thu Feb 14 01:25:13 CET 2008 armv5tejl
GNU/Linux

I'd appreciate a hint on how to diagnose the problem.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May  5 22:07:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 5 May 2008 22:07:58 -0500
Subject: [LLVMbugs] [Bug 2265] Can't solve trip count for simple loop
In-Reply-To: <bug-2265-206@http.llvm.org/bugs/>
Message-ID: <200805060307.m4637wj1019073@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2265


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #4 from Owen Anderson <resistor at mac.com>  2008-05-05 22:07:56 ---
Turns out that SCEV is capable of solving this one.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 02:23:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 02:23:58 -0500
Subject: [LLVMbugs] [Bug 2287] CodeGen/X86/2007-07-03-GR64ToVR64.ll fails
In-Reply-To: <bug-2287-206@http.llvm.org/bugs/>
Message-ID: <200805060723.m467Nw0h028373@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2287


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

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




--- Comment #2 from Evan Cheng <evan.cheng at apple.com>  2008-05-06 02:23:58 ---
According to Agner's calling convention document, Linux x86-64 passes mmx
values in stack. I've fixed the test.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 04:33:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 04:33:48 -0500
Subject: [LLVMbugs] [Bug 2291] New: Undiagnosed syntax error
Message-ID: <bug-2291-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2291

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


#elif is not in a skipped group so the following must be diagnosed:

#if 1
#elif
#endif


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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 04:04:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 04:04:58 -0500
Subject: [LLVMbugs] [Bug 1643] Flags should be different from an expand
 parameter and a divided parameter
In-Reply-To: <bug-1643-206@http.llvm.org/bugs/>
Message-ID: <200805060904.m4694wml008588@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1643


varth <nicolas.geoffray at lip6.fr> changed:

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




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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 05:48:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 05:48:34 -0500
Subject: [LLVMbugs] [Bug 2292] New: tree check failure inlining pure function
Message-ID: <bug-2292-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2292

           Summary: tree check failure inlining pure function
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: baldrick at free.fr
                CC: llvmbugs at cs.uiuc.edu


This testcase

__inline__ __attribute__ ((__pure__)) int g (void) {}
void f (int k) { k = g (); }

crashes llvm-gcc at -O2 on x86-32 linux:

$ gcc -c -O2 dwv.c
dwv.c: In function 'f':
dwv.c:2: internal compiler error: tree check: expected tree that contains 'decl
with visibility' structure, have 'parm_decl'  in declare_return_variable, at
tree-inline.c:1271

Curiously the compiler doesn't crash on x86-64 linux.  However
I see the same failing tree check for several Ada testcases on
x86-64 which pass happily on x86-32.

The relevant code snippet is:

    /* LLVM LOCAL begin */
#ifdef ENABLE_LLVM
    /* The return node can suddenly become multiple assignment.  Because of
this
     * conservatively don't consider this thing for SSA form anymore.
    */
    if (DECL_P (var))
      DECL_GIMPLE_FORMAL_TEMP_P (var) = 0;
#endif
    /* LLVM LOCAL end */

The check fails on the assignment.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 10:04:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 10:04:43 -0500
Subject: [LLVMbugs] [Bug 2292] tree check failure inlining pure function
In-Reply-To: <bug-2292-206@http.llvm.org/bugs/>
Message-ID: <200805061504.m46F4huI021080@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2292


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #1 from Duncan Sands <baldrick at free.fr>  2008-05-06 10:04:42 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062024.html


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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 11:19:44 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 11:19:44 -0500
Subject: [LLVMbugs] [Bug 2293] New: lli/llc crash with "Cannot promote to
	smaller type"
Message-ID: <bug-2293-206@http.llvm.org/bugs/>

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

           Summary: lli/llc crash with "Cannot promote to smaller type"
           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: jonas at mit.edu
                CC: llvmbugs at cs.uiuc.edu


The following code, 

; ModuleID = 'bugpoint-reduced-simplified.bc'

define i32 @main() {
entry:
        br label %forloop1_for_block
forloop1_for_block:             ; preds = %forloop1_for_block, %entry
        %forloop1_for_phi = phi i3 [ 0, %entry ], [ %nextvar,
%forloop1_for_block ]; <i3> [#uses=1]
        %nextvar = add i3 %forloop1_for_phi, 1          ; <i3> [#uses=1]
        br label %forloop1_for_block
}



causes lli and llc from svn trunk (r50721) to crash, with a message such as:
[jonas at hamiltonian swig.svn]$ lli bugpoint-reduced-simplified.bc 

lli: LegalizeDAG.cpp:4054:
llvm::SDOperand<unnamed>::SelectionDAGLegalize::PromoteOp(llvm::SDOperand):
Assertion `NVT > VT && MVT::isInteger(NVT) == MVT::isInteger(VT) && "Cannot
promote to smaller type!"' failed.

lli[0xcde54a]

lli[0xcde674]

/lib/libc.so.6[0x2b3b0be6e7d0]

/lib/libc.so.6(gsignal+0x35)[0x2b3b0be6e765]

/lib/libc.so.6(abort+0x110)[0x2b3b0be701c0]

/lib/libc.so.6(__assert_fail+0xef)[0x2b3b0be67c2f]

lli[0xa7d5d9]

lli[0xa8fa3b]

lli[0xa8fe5f]

lli(_ZN4llvm12SelectionDAG8LegalizeEv+0x4d)[0xa8ffe1]

lli(_ZN4llvm16SelectionDAGISel17CodeGenAndEmitDAGERNS_12SelectionDAGE+0xe3)[0x9e3df7]

lli(_ZN4llvm16SelectionDAGISel16SelectBasicBlockEPNS_10BasicBlockERNS_15MachineFunctionERNS_20FunctionLoweringInfoE+0xcb)[0x9fc159]

lli(_ZN4llvm16SelectionDAGISel13runOnFunctionERNS_8FunctionE+0x20e)[0x9fdf72]

lli[0x86ed41]

lli(_ZN4llvm13FPPassManager13runOnFunctionERNS_8FunctionE+0x129)[0xc79c8b]

lli(_ZN4llvm23FunctionPassManagerImpl3runERNS_8FunctionE+0x74)[0xc79eb2]

lli(_ZN4llvm19FunctionPassManager3runERNS_8FunctionE+0x86)[0xc7a016]

lli(_ZN4llvm3JIT16runJITOnFunctionEPNS_8FunctionE+0x75)[0x96d753]

lli(_ZN4llvm3JIT20getPointerToFunctionEPNS_8FunctionE+0x224)[0x96dad2]

lli(_ZN4llvm3JIT11runFunctionEPNS_8FunctionERKSt6vectorINS_12GenericValueESaIS4_EE+0x83)[0x96dba9]

lli(_ZN4llvm15ExecutionEngine17runFunctionAsMainEPNS_8FunctionERKSt6vectorISsSaISsEEPKPKc+0x45a)[0x9659ec]

lli(main+0x5f0)[0x802d5a]

/lib/libc.so.6(__libc_start_main+0xf4)[0x2b3b0be5ab44]

lli[0x8022b9]

Aborted (core dumped)



This is on x86_64 linux (ubuntu). I think it might have something to do with
using a phi node with a sub-32 bit type, as virtually identical code with an
i32 does not crash. 

[thanks, LLVM is awesome]


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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 14:58:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 14:58:30 -0500
Subject: [LLVMbugs] [Bug 2101] clang assert in bitfield initializer codegen
In-Reply-To: <bug-2101-206@http.llvm.org/bugs/>
Message-ID: <200805061958.m46JwUse000997@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2101


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #2 from Devang Patel <dpatel at apple.com>  2008-05-06 14:58:29 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080505/005597.html


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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 17:14:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 17:14:23 -0500
Subject: [LLVMbugs] [Bug 2294] New: possible integer codegen bug
Message-ID: <bug-2294-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2294

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


Using recent svn on x86-32 the code below behaves differently at -O0 and -O1. 
Clearly there's an integer overflow happening but the behavior difference is
still evident when -fwrapv is specified, so this seems like a bug.

#include <stdio.h>

short g_2 = 1;
int volatile g_5 = 0;

void func_1(void) ;
void func_1(void) 
{ 
  int tmp ;
  while (g_2 > 0) {
    if (!g_2) {
      tmp = g_5;
    }
    g_2++;
  }
}

int main(void) 
{ 
  func_1();
  printf("%d\n", g_2);
  return (0);
}


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

From bugzilla-daemon at cs.uiuc.edu  Tue May  6 18:57:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 6 May 2008 18:57:00 -0500
Subject: [LLVMbugs] [Bug 2286] FrontendObjC/nonfragile-ivar.m fails on
	x86-32 linux
In-Reply-To: <bug-2286-206@http.llvm.org/bugs/>
Message-ID: <200805062357.m46Nv0La010155@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2286


Bill Wendling <isanbard at gmail.com> changed:

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




--- Comment #8 from Bill Wendling <isanbard at gmail.com>  2008-05-06 18:56:59 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062087.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062088.html


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

From bugzilla-daemon at cs.uiuc.edu  Wed May  7 01:56:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 May 2008 01:56:37 -0500
Subject: [LLVMbugs] [Bug 2288] LLVM type size doesn't match GCC type size!
In-Reply-To: <bug-2288-206@http.llvm.org/bugs/>
Message-ID: <200805070656.m476ubaO026061@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2288


Tanya Lattner <tonic at nondot.org> changed:

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




--- Comment #4 from Tanya Lattner <tonic at nondot.org>  2008-05-07 01:56:37 ---
Reverted patch:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062108.html


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

From bugzilla-daemon at cs.uiuc.edu  Wed May  7 17:01:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 May 2008 17:01:38 -0500
Subject: [LLVMbugs] [Bug 2290] llvm-gcc doesn't bootstrap on arm-qemu (linux)
In-Reply-To: <bug-2290-206@http.llvm.org/bugs/>
Message-ID: <200805072201.m47M1cIQ014446@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2290


Nick Lewycky <nlewycky at google.com> changed:

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




--- Comment #2 from Nick Lewycky <nlewycky at google.com>  2008-05-07 17:01:37 ---
All fixed. Thanks!


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

From bugzilla-daemon at cs.uiuc.edu  Wed May  7 17:25:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 May 2008 17:25:53 -0500
Subject: [LLVMbugs] [Bug 2025] llvm-gcc4.2 won' t build if using LLVM
 release build and boostrap enabled
In-Reply-To: <bug-2025-206@http.llvm.org/bugs/>
Message-ID: <200805072225.m47MPr9c015267@zion.cs.uiuc.edu>

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


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #14 from Devang Patel <dpatel at apple.com>  2008-05-07 17:25:53 ---
"Fixed"
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062134.html


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

From bugzilla-daemon at cs.uiuc.edu  Wed May  7 17:41:31 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 May 2008 17:41:31 -0500
Subject: [LLVMbugs] [Bug 2295] New: incorrect optimization (instcombine)
Message-ID: <bug-2295-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2295

           Summary: incorrect optimization (instcombine)
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: scott.llvm at h4ck3r.net
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1621)
 --> (http://llvm.org/bugs/attachment.cgi?id=1621)
.ll to demonstrate mis-optimization

Using LLVM 2.2:

llvm-as -f x.ll
opt -instcombine -f x.bc -o x.opt.bc
lli x.bc
lli x.opt.bc

The first outputs (as expected):
90 (Z)
and the second (optimized) incorrectly outputs:
33 (!)

I was unable to get bugpoint to work, but the attached file has been manually
minimized (73 lines). Happens with "opt -std-compile-opts" also, but "opt
-instcombine" is all that's necessary.

It appears two different pointers into allocated memory are being conflated.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May  7 19:36:59 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 7 May 2008 19:36:59 -0500
Subject: [LLVMbugs] [Bug 2295] incorrect optimization (instcombine)
In-Reply-To: <bug-2295-206@http.llvm.org/bugs/>
Message-ID: <200805080036.m480axaJ019166@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2295


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-05-07 19:36:59 ---
Bill already fixed this bug awhile back.  Please verify when you get a chance,
thanks!


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 00:33:57 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 00:33:57 -0500
Subject: [LLVMbugs] [Bug 2296] New: Instcombine shouldn' t turn
 x86_sse2_storel_dq into a full-width store
Message-ID: <bug-2296-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2296

           Summary: Instcombine shouldn't turn x86_sse2_storel_dq into a
                    full-width store
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Obvious bug: visitCallInst shouldn't turn x86_sse2_storel_dq into a store
because it's not equivalent to a full-width store.

Trivial patch, but I lost track of which testcase ran into this issue.  Putting
here so it won't get lost.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 01:31:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 01:31:37 -0500
Subject: [LLVMbugs] [Bug 2283] Problem compiling some GCC regtests at -O0
In-Reply-To: <bug-2283-206@http.llvm.org/bugs/>
Message-ID: <200805080631.m486Vbc0029253@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2283


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #1 from Duncan Sands <baldrick at free.fr>  2008-05-08 01:31:36 ---
The compile test was smashing the stack, fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062119.html

The execute test works for me.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 01:58:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 01:58:12 -0500
Subject: [LLVMbugs] [Bug 2297] New: Over-aggressive dead store elimination
 in instcombine
Message-ID: <bug-2297-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2297

           Summary: Over-aggressive dead store elimination in instcombine
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


int a() {
  int len;
  char* x = malloc(10);
  x[1] = 0;
  x[0] = 1;
  len =  strlen(x);
  x[0] = 0;
  b(x);
  return len;
}

A bit of a manufactured testcase, but it doesn't do anything especially
unusual.

Output from llvm-gcc -c -emit-llvm | opt -mem2reg -simplifycfg -instcombine |
llvm-dis:
; ModuleID = '<stdin>'
target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
target triple = "i386-pc-linux-gnu"

define i32 @a() nounwind  {
entry:
        %tmp1 = call i8* @malloc( i32 10 ) nounwind             ; <i8*>
[#uses=4]
        %tmp3 = getelementptr i8* %tmp1, i32 1          ; <i8*> [#uses=1]
        store i8 0, i8* %tmp3, align 1
        %tmp7 = call i32 @strlen( i8* %tmp1 ) nounwind readonly                
; <i32> [#uses=1]
        store i8 0, i8* %tmp1, align 1
        %tmp11 = call i32 (...)* @b( i8* %tmp1 ) nounwind               ; <i32>
[#uses=0]
        ret i32 %tmp7
}

declare i8* @malloc(i32) nounwind 

declare i32 @strlen(i8*) nounwind readonly 

declare i32 @b(...)

This is clearly not equivalent to the original program: the output depends on
whatever random junk happens to be in x[0] after malloc returns.

Another issue, which I haven't quite figured out: running the same program
through llvm-gcc -c -emit-llvm | opt -mem2reg -instcombine | llvm-dis produces
the following:

; ModuleID = '<stdin>'
target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
target triple = "i386-pc-linux-gnu"

define i32 @a() nounwind  {
entry:
        %tmp1 = call i8* @malloc( i32 10 ) nounwind             ; <i8*>
[#uses=4]
        %tmp3 = getelementptr i8* %tmp1, i32 1          ; <i8*> [#uses=1]
        store i8 0, i8* %tmp3, align 1
        store i8 0, i8* %tmp1, align 1
        %tmp11 = call i32 (...)* @b( i8* %tmp1 ) nounwind               ; <i32>
[#uses=0]
        br label %return

return:         ; preds = %entry
        %tmp7 = call i32 @strlen( i8* %tmp1 ) nounwind readonly                
; <i32> [#uses=1]
        ret i32 %tmp7
}

declare i8* @malloc(i32) nounwind 

declare i32 @strlen(i8*) nounwind readonly 

declare i32 @b(...)

It's clearly wrong to reorder the strlen call in that way.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 02:12:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 02:12:05 -0500
Subject: [LLVMbugs] [Bug 2107] Poor lowering for _mm_loadl_epi64
In-Reply-To: <bug-2107-206@http.llvm.org/bugs/>
Message-ID: <200805080712.m487C5U1030475@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2107


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

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




--- Comment #19 from Evan Cheng <evan.cheng at apple.com>  2008-05-08 02:12:05 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062137.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062144.html


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 03:00:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 03:00:09 -0500
Subject: [LLVMbugs] [Bug 2298] New: Missed optimization: should instcombine
 icmp ~A, ~B to icmp B, A
Message-ID: <bug-2298-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2298

           Summary: Missed optimization: should instcombine icmp ~A, ~B to
                    icmp B, A
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Pretty simple bug: int a(int a, int b) {return ~a < ~b;} currently outputs the
following:
define i32 @a(i32 %a, i32 %b) nounwind  {
entry:
        %tmp1not = xor i32 %a, -1               ; <i32> [#uses=1]
        %tmp2not = xor i32 %b, -1               ; <i32> [#uses=1]
        %tmp3 = icmp slt i32 %tmp1not, %tmp2not         ; <i1> [#uses=1]
        %tmp34 = zext i1 %tmp3 to i32           ; <i32> [#uses=1]
        ret i32 %tmp34
}

It should output the following:
define i32 @a(i32 %a, i32 %b) nounwind  {
entry:
        %tmp3 = icmp slt i32 %b, %a             ; <i1> [#uses=1]
        %tmp34 = zext i1 %tmp3 to i32           ; <i32> [#uses=1]
        ret i32 %tmp34
}

Not an especially useful instcombine, but it's extremely cheap to implement.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 12:38:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 12:38:30 -0500
Subject: [LLVMbugs] [Bug 2297] Over-aggressive dead store elimination in
	instcombine
In-Reply-To: <bug-2297-206@http.llvm.org/bugs/>
Message-ID: <200805081738.m48HcUjr001713@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2297


Chris Lattner <sabre at nondot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
          Component|new bugs                    |Scalar Optimizations
           Keywords|                            |miscompilation
            Product|new-bugs                    |libraries
         Resolution|                            |FIXED
   Target Milestone|---                         |2.3
            Version|unspecified                 |trunk




--- Comment #2 from Chris Lattner <sabre at nondot.org>  2008-05-08 12:38:30 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062159.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062160.html

Thanks Eli!  I appreciate the nice small testcase,

-Chris


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 16:11:35 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 16:11:35 -0500
Subject: [LLVMbugs] [Bug 2299] New: -anders-aa and -gvn marks values as
 undefined incorrectly
Message-ID: <bug-2299-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2299

           Summary: -anders-aa and -gvn marks values as undefined
                    incorrectly
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1622)
 --> (http://llvm.org/bugs/attachment.cgi?id=1622)
testcase

Using SVN r50866, opt -ander-aa -gvn marks some values as undefined, but that
is not correct:

This portion of IR:
        %buf2 = bitcast [8 x i32]* %buf12 to i8*                ; <i8*>
[#uses=1]
        %tmp4 = call i64 @fread( i8* noalias  %buf2, i64 32, i64 1,
%struct.FILE* noalias  %tmp1 ) nounwind             ; <i64> [#uses=0]
        %tmp7 = load i32* %buf12.sub, align 4           ; <i32> [#uses=1]
        %tmp9 = malloc i8, i32 %tmp7            ; <i8*> [#uses=1]

Gets turned into:

        %buf2 = bitcast [8 x i32]* %buf12 to i8*                ; <i8*>
[#uses=1]
        %tmp4 = call i64 @fread( i8* noalias  %buf2, i64 32, i64 1,
%struct.FILE* noalias  %tmp1 ) nounwind             ; <i64> [#uses=0]
        %tmp9 = malloc i8, i32 undef            ; <i8*> [#uses=1]


To reproduce:
../../../Debug/bin/opt -anders-aa -gvn z.bc -o -|llvm-dis

-basic-aa works, does it get confused by the noalias mark?
FWIW the original C code had no aliases:
 ...
 fread(buf,sizeof(buf),1,f);
 use( malloc(*(int*)buf) );
...


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 16:23:26 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 16:23:26 -0500
Subject: [LLVMbugs] [Bug 2300] New: Wrong indices in GPE with pointer
 arithmetic and uintptr_t
Message-ID: <bug-2300-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2300

           Summary: Wrong indices in GPE with pointer arithmetic and
                    uintptr_t
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: matthieu at uiuc.edu
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1625)
 --> (http://llvm.org/bugs/attachment.cgi?id=1625)
the test case

llvm back-end build: RELEASE_2_2
llvm-gcc front-end: 4.2 (downloaded from the website)

Overview:

When doing some specific pointer arithmetic, the emitted code from llvm-gcc is
wrong when compiling the test case with an optimization level different than
-O0.
Looking at the IR, the indexes used in GPE instructions are wrong.

To reproduce the bug:
Download the attached test case and compile the code with llvm-gcc with
different optimizations level. Whenever it is different than -O0, the result is
wrong.

Expected result:
"Diff is 16384" (without the quotes)

Obtained result:
"Diff is 16384" when compiled with -O0
"Diff is 4294955008" when using higher levels

Additional information:

The IR I obtained at O2:

define i32 @main(i32 %argc, i8** %argv) nounwind  {
entry:
        %tmp124 = malloc [65536 x i8]           ; <[65536 x i8]*> [#uses=3]
        %ctg2 = getelementptr [65536 x i8]* %tmp124, i32 0, i32 -12288         
; <i8*> [#uses=1]
        %tmp9 = ptrtoint i8* %ctg2 to i32               ; <i32> [#uses=1]
        %tmp1314 = ptrtoint [65536 x i8]* %tmp124 to i32                ; <i32>
[#uses=1]
        %tmp15 = sub i32 %tmp9, %tmp1314                ; <i32> [#uses=1]
        %tmp18 = tail call i32 (i8*, ...)* @printf( i8* noalias  getelementptr
([12 x i8]* @.str, i32 0, i32 0), i32 %tmp15 ) nounwind          ; <i32>
[#uses=0]
        free [65536 x i8]* %tmp124
        ret i32 0
}

the second index for %ctg2 seems really bad.

Note as well that changing the cast (uintptr_t) to (char *) seems to remove the
problem.

Using opt with -std-compile-opts on the unoptimized IR generate the correct
result.

Is there a way to know which passes are actually used at -O2? I would like to
run bugpoint but using -std-compile-opts does not generate a wrong result so it
seems the bug is either from an additional pass or from the front-end.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May  8 17:27:59 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 8 May 2008 17:27:59 -0500
Subject: [LLVMbugs] [Bug 2300] Wrong indices in GPE with pointer arithmetic
 and uintptr_t
In-Reply-To: <bug-2300-206@http.llvm.org/bugs/>
Message-ID: <200805082227.m48MRxeO012741@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2300


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #1 from Dan Gohman <gohman at apple.com>  2008-05-08 17:27:59 ---
This looks like the same bug as in PR2295. This testcase also
passes with LLVM trunk, while it fails with LLVM 2.2.

Some background for debugging these kinds of bugs:
The llvm-gcc front-end behaves differently at -O2 than at -O0,
and this is often why compiling with -O0 and optimizing the
result with opt -std-compile-opts produces different results.
To get past this, compile with -O2 (or whatever) and add
-mllvm -disable-llvm-optzns. This will run the front-end as if
it were an -O2 compile, but the LLVM optimizations won't run,
so you can get a bitcode file that you can use with
opt -std-compile-opts with a better chance of reproducing the
problem.

There's also -fdebug-pass-arguments, to print out the
command-line arguments for the passes llvm-gcc runs.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  9 00:21:24 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 May 2008 00:21:24 -0500
Subject: [LLVMbugs] [Bug 2298] Missed optimization: should instcombine icmp
 ~A, ~B to icmp B, A
In-Reply-To: <bug-2298-206@http.llvm.org/bugs/>
Message-ID: <200805090521.m495LOJP024749@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2298


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-09 00:21:23 ---
This triggered a quite a few times in multisource across several different
programs:
****    %tmp177.i = icmp slt i32 %tmp176.i17, -1                ; <i1>
[#uses=1]
...
****    %tmp42 = icmp ugt i32 %tmp35, %tmp40            ; <i1> [#uses=1]
****    %tmp27 = icmp ugt i32 %tmp23, %tmp26            ; <i1> [#uses=1]
****    %tmp33 = icmp ugt i32 %tmp30, %tmp32            ; <i1> [#uses=1]
...
****    %tmp23 = icmp ugt i32 %tmp17, %tmp22            ; <i1> [#uses=1]

****    %tmp18 = icmp ugt i32 %tmp13, %tmp17            ; <i1> [#uses=1]
....
****    %tmp42 = icmp ult i32 %tmp41, -4                ; <i1> [#uses=1]
...
****    %tmp24 = icmp ugt i32 %tmp14, %tmp22            ; <i1> [#uses=1]
****    %tmp37 = icmp ugt i32 %tmp29, %tmp35            ; <i1> [#uses=1]
****    %tmp27 = icmp ugt i32 %tmp21, %tmp26            ; <i1> [#uses=1]
****    %tmp41 = icmp ugt i32 %tmp32, %tmp39            ; <i1> [#uses=1]


I also implemented the '-x == -y' xform, which is only valid for equality
operators:
**** ---        %tmp14 = icmp eq i32 %tmp, -4           ; <i1> [#uses=1]
**** ---        %tmp436751 = icmp eq i32 %tmp441.rec, -2                ; <i1>
[#uses=1]

this seems to come from loop optimizations or something.  Lots of things that
use ".rec" as a suffix.

Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062193.html

Thanks!


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  9 07:59:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 May 2008 07:59:16 -0500
Subject: [LLVMbugs] [Bug 2301] New: Internal compiler error when doing
 inline optimization in llvm-gcc4.2-2.2
Message-ID: <bug-2301-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2301

           Summary: Internal compiler error when doing inline optimization
                    in llvm-gcc4.2-2.2
           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: mikael.lepisto at tut.fi
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1626)
 --> (http://llvm.org/bugs/attachment.cgi?id=1626)
Test code for reproducing the error.

When compiling file without -fno-inline switch, internal compiler error
occures.

e.g. this one works:

elhigu at mr-lenovo:~/Downloads/bugzillatesting$
/home/elhigu/stow_repo/llvm-gcc4.2-2.2.sources/bin/gcc --emit-llvm -O5
-ffreestanding -fno-unit-at-a-time -c -c -DNDEBUG -DHOST_EXAMPLE_CODE=0
-DARCH_IS_32BIT -DARCH_IS_GENERIC stripped_motion_est_2.c  -fno-inline

This does not:

elhigu at mr-lenovo:~/Downloads/bugzillatesting$
/home/elhigu/stow_repo/llvm-gcc4.2-2.2.sources/bin/gcc --emit-llvm -O5
-ffreestanding -fno-unit-at-a-time -c -c -DNDEBUG -DHOST_EXAMPLE_CODE=0
-DARCH_IS_32BIT -DARCH_IS_GENERIC stripped_motion_est_2.c
stripped_motion_est_2.c: In function ???SomeFunction???:
stripped_motion_est_2.c:18: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.
elhigu at mr-lenovo:~/Downloads/bugzillatesting$

More specific test run with gdb and cc1:

elhigu at mr-lenovo:~/Downloads/bugzillatesting$ gdb --args
/home/elhigu/stow_repo/llvm-gcc4.2-2.2.sources/libexec/gcc/i686-pc-linux-gnu/4.2.1/cc1
-quiet  stripped_motion_est_2.c -emit-llvm-bc -o stripped_motion_est_2.o -quiet
-dumpbase stripped_motion_est_2.c -auxbase stripped_motion_est_2 -O2 
-ffreestanding -fno-unit-at-a-time
GNU gdb 6.8-debian
Copyright (C) 2008 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i486-linux-gnu"...
(gdb) run
Starting program:
/home/elhigu/stow_repo/llvm-gcc4.2-2.2.sources/libexec/gcc/i686-pc-linux-gnu/4.2.1/cc1
-quiet stripped_motion_est_2.c -emit-llvm-bc -o stripped_motion_est_2.o -quiet
-dumpbase stripped_motion_est_2.c -auxbase stripped_motion_est_2 -O2
-ffreestanding -fno-unit-at-a-time
[Thread debugging using libthread_db enabled]
[New Thread 0xb7cc76c0 (LWP 1473)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb7cc76c0 (LWP 1473)]
optimize_inline_calls (fn=0xb7beaf00) at
../../llvm-gcc4.2-2.2.source/gcc/tree-inline.c:1248
1248              if (!DECL_P (base_m))
Current language:  auto; currently c
(gdb) backtrace
#0  optimize_inline_calls (fn=0xb7beaf00) at
../../llvm-gcc4.2-2.2.source/gcc/tree-inline.c:1248
#1  0x080bb19a in tree_rest_of_compilation (fndecl=0xb7beaf00) at
../../llvm-gcc4.2-2.2.source/gcc/tree-optimize.c:449
#2  0x08056cb1 in c_expand_body (fndecl=0xb7beaf00) at
../../llvm-gcc4.2-2.2.source/gcc/c-decl.c:7141
#3  0x0836ed2d in cgraph_expand_function (node=0xb7bf8e0c) at
../../llvm-gcc4.2-2.2.source/gcc/cgraphunit.c:1305
#4  0x083703b5 in cgraph_assemble_pending_functions () at
../../llvm-gcc4.2-2.2.source/gcc/cgraphunit.c:376
#5  0x08370115 in cgraph_finalize_function (decl=0xb7beaf00, nested=false) at
../../llvm-gcc4.2-2.2.source/gcc/cgraphunit.c:511
#6  0x08057060 in finish_function () at
../../llvm-gcc4.2-2.2.source/gcc/c-decl.c:7110
#7  0x08099a65 in c_parser_declaration_or_fndef (parser=0xb7bf3000,
fndef_ok=true, empty_ok=true, nested=false, start_attr_ok=<value optimized
out>, foreach_elem=0x0)
    at ../../llvm-gcc4.2-2.2.source/gcc/c-parser.c:1621
#8  0x080a40d8 in c_parser_external_declaration (parser=0xb7bf3000) at
../../llvm-gcc4.2-2.2.source/gcc/c-parser.c:1334
#9  0x080a4a58 in c_parse_file () at
../../llvm-gcc4.2-2.2.source/gcc/c-parser.c:1220
#10 0x0808b99a in c_common_parse_file (set_yydebug=0) at
../../llvm-gcc4.2-2.2.source/gcc/c-opts.c:1284
#11 0x082ca782 in toplev_main (argc=14, argv=0xbff93154) at
../../llvm-gcc4.2-2.2.source/gcc/toplev.c:1108
#12 0x080aa302 in main (argc=0, argv=0xb7bdff60) at
../../llvm-gcc4.2-2.2.source/gcc/llvm-main.cpp:39
(gdb)


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  9 08:02:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 May 2008 08:02:11 -0500
Subject: [LLVMbugs] [Bug 2302] New: Incorrect optimization with recursive
	call to main
Message-ID: <bug-2302-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2302

           Summary: Incorrect optimization with recursive call to main
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


#include <stdio.h>
#include <stdlib.h>
int x;
int main() {
if (++x > 10) exit(0);
else main();
printf("foo");
return 0;
}

This program should not print any text; compiled with llvm-gcc -O3, it does.

AFAIK, this is perfectly legal C code (although not legal C++).


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  9 09:03:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 May 2008 09:03:48 -0500
Subject: [LLVMbugs] [Bug 2303] New: instcombine crash
Message-ID: <bug-2303-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2303

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


Created an attachment (id=1627)
 --> (http://llvm.org/bugs/attachment.cgi?id=1627)
testcase .ll

This breaks the llvm-gcc build.

$ opt -instcombine y.bc -disable-output
opt[0x9d4d92]
opt[0x9d4eb8]
/lib/libc.so.6[0x7f5c707cd100]
opt(_ZNK4llvm5Value10getValueIDEv+0xc)[0x6bab06]
opt(_ZNK4llvm11Instruction9getOpcodeEv+0x15)[0x6babe1]
opt(_ZN4llvm10BranchInst7classofEPKNS_11InstructionE+0x15)[0x6fa749]
opt(_ZN4llvm8isa_implINS_10BranchInstENS_14TerminatorInstEEEbRKT0_+0x15)[0x711abb]
opt(_ZN4llvm13isa_impl_wrapINS_10BranchInstEKNS_14TerminatorInstES3_E4doitERS3_+0x15)[0x711ad5]
opt(_ZN4llvm11isa_impl_clINS_14TerminatorInstEE3isaINS_10BranchInstEEEbRKS1_+0x15)[0x711aef]
opt(_ZN4llvm11isa_impl_clIPNS_14TerminatorInstEE3isaINS_10BranchInstEEEbS2_+0x15)[0x711b09]
opt(_ZN4llvm3isaINS_10BranchInstEPNS_14TerminatorInstEEEbRKT0_+0x18)[0x711b26]
opt(_ZN4llvm8dyn_castINS_10BranchInstEPNS_14TerminatorInstEEENS_10cast_rettyIT_T0_E8ret_typeERKS6_+0x15)[0x732c75]
opt[0x768f56]
opt[0x79f55a]
opt[0x79fdd4]
opt(_ZN4llvm13FPPassManager13runOnFunctionERNS_8FunctionE+0x127)[0x968e3f]
opt(_ZN4llvm13FPPassManager11runOnModuleERNS_6ModuleE+0x52)[0x968fac]
opt(_ZN4llvm13MPPassManager11runOnModuleERNS_6ModuleE+0xec)[0x968ae4]
opt(_ZN4llvm15PassManagerImpl3runERNS_6ModuleE+0x74)[0x968c74]
opt(_ZN4llvm11PassManager3runERNS_6ModuleE+0x21)[0x968cd7]
opt(main+0xa1b)[0x6c88cb]
/lib/libc.so.6(__libc_start_main+0xf4)[0x7f5c707b91c4]
opt(cos+0x119)[0x6ba309]
Segmentation fault


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  9 10:13:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 May 2008 10:13:58 -0500
Subject: [LLVMbugs] [Bug 2303] instcombine crash
In-Reply-To: <bug-2303-206@http.llvm.org/bugs/>
Message-ID: <200805091513.m49FDw41018172@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2303


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #2 from Duncan Sands <baldrick at free.fr>  2008-05-09 10:13:58 ---
Fixed here:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080505/062204.html


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

From bugzilla-daemon at cs.uiuc.edu  Fri May  9 18:01:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 9 May 2008 18:01:30 -0500
Subject: [LLVMbugs] [Bug 1337] Remove duplicate (explicit) checks in test
	cases.
In-Reply-To: <bug-1337-206@http.llvm.org/bugs/>
Message-ID: <200805092301.m49N1Ujj031847@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1337


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #1 from Dan Gohman <gohman at apple.com>  2008-05-09 18:01:29 ---
The last of the trailing &&'s were removed a while ago. And there don't
appear to be any remaining tests still using duplicated script lines for
the purpose of ensuring that each command succeeds.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 10 03:39:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 May 2008 03:39:07 -0500
Subject: [LLVMbugs] [Bug 2304] New: CHECKER: false positive on uninitialized
	warning
Message-ID: <bug-2304-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2304

           Summary: CHECKER: false positive on uninitialized warning
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Using SVN r4294967285:

$ ~/llvm-svn/llvm/tools/clang/utils/ccc-analyzer xtest.c -c
xtest.c
ANALYZE: xtest.c foo
xtest.c:10:2: warning: [CHECKER] Branch condition evaluates to an uninitialized
value.
        if(pt->partno == 1 || !found) {
        ^                     ~~~~~~
ANALYZE: xtest.c bar
1 diagnostic generated.

$ cat xtest.c
struct x {
        int partno;
};
int foo(struct x *pt)
{
        int found;
        if(pt->partno != 1) {
                found = 0;
        }
        if(pt->partno == 1 || !found) {
                return 1;
        }
        return 0;
}

int bar(int partno)
{
        int found;
        if(partno != 1) {
                found = 0;
        }
        if(partno == 1 || !found) {
                return 1;
        }
        return 0;
}

The HTML output produced says that the branch if(pt->partno...) is untaken.
Notice that if I use a struct it gives a warning, if I don't use the struct it
doesn't give a warning. found is not used uninitialized, because when 
partno is 1 the !found condition is not evaluated anymore.
Also if partno was 1 on the first if, it still is that on the 2nd one.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 10 03:49:51 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 May 2008 03:49:51 -0500
Subject: [LLVMbugs] [Bug 2305] New: CHECKER: false positive on uninitialized
 warning ( path-sensitiveness)
Message-ID: <bug-2305-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2305

           Summary: CHECKER: false positive on uninitialized warning (path-
                    sensitiveness)
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1628)
 --> (http://llvm.org/bugs/attachment.cgi?id=1628)
scan-build report

Using SVN r50942:

$ ~/llvm-svn/llvm/tools/clang/utils/ccc-analyzer ytest.c -c
ytest.c
ANALYZE: ytest.c foo
ytest.c:11:3: warning: [CHECKER] Dereference of undefined value.
                sections[0]=0;
                ^~~~~~~~~~~~~
1 diagnostic generated.

int foo(int dend, int* dst)
{
        int *pehdr = 0, *sections;
        if(dend>0xf8) {
                pehdr = &dst[dend];
                while(pehdr>dst) {
                        sections = pehdr;
                }
        }
        if(pehdr) {
                sections[0]=0;
        }
        return 0;
}

If pehdr is non-NULL, then pehdr is dst + dend, dend is >0xf8 => (pehdr>dst) is
true. The checker says the loop condition is false.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 10 04:18:04 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 May 2008 04:18:04 -0500
Subject: [LLVMbugs] [Bug 2306] New: CHECKER: path sensitivity
Message-ID: <bug-2306-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2306

           Summary: CHECKER: path sensitivity
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: edwintorok at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1629)
 --> (http://llvm.org/bugs/attachment.cgi?id=1629)
checker html report

Using SVN r50942:

/home/edwin/llvm-svn/llvm/tools/clang/utils/ccc-analyzer -c path_sens.c
path_sens.c
ANALYZE: path_sens.c foo
path_sens.c:10:11: warning: [CHECKER] Dereference of null pointer.
                return *bar;
                        ^~~
1 diagnostic generated.

int foo(int left, int right)
{
        int x=5;
        int *bar = 0;
        while(left<=right) {
                left++;
                bar = &x;
        }
        if(left<=right) {
                return *bar;
        }
        return 0;
}

See the attached html report, it says that the while() condition is false, and
then the if() condition  is true.
That can't happen, after the while (left<=right) is always false, hence the if
is never entered.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 10 05:35:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 May 2008 05:35:09 -0500
Subject: [LLVMbugs] [Bug 2307] New: switch without body crashes code
	generator
Message-ID: <bug-2307-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2307

           Summary: switch without body crashes code generator
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


The following crashes codegen:

void f() {
  switch(42) { }
}


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 10 05:50:17 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 May 2008 05:50:17 -0500
Subject: [LLVMbugs] [Bug 2308] New: invalid memory read/write with clang
 serialized input
Message-ID: <bug-2308-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2308

           Summary: invalid memory read/write with clang serialized input
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Bitcode Reader
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1630)
 --> (http://llvm.org/bugs/attachment.cgi?id=1630)
valgrind dump

running "valgrind clang --test-pickling ./test/Serialization/complex.c" gives
invalid memory reads/writes in the valgrind report.
clang/test/Serialization/stmt_exprs.c produces similar report.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 10 19:12:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 May 2008 19:12:55 -0500
Subject: [LLVMbugs] [Bug 2309] New: codegening static structure inside
 functions triggers assertion
Message-ID: <bug-2309-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2309

           Summary: codegening static structure inside functions triggers
                    assertion
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Windows NT
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1632)
 --> (http://llvm.org/bugs/attachment.cgi?id=1632)
test case

clang -emit-llvm triggers the following assertion with the code attached:

clang: Constants.cpp:381: llvm::ConstantStruct::ConstantStruct(const
llvm::StructType*, const std::vector<llvm::Constant*,
std::allocator<llvm::Constant*> >&): 
Assertion `V.size() == T->getNumElements() && "Invalid initializer vector for
constant structure"' failed.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 10 19:15:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 10 May 2008 19:15:22 -0500
Subject: [LLVMbugs] [Bug 2310] New: codegening structure variable
 definition+initialization triggers assertion
Message-ID: <bug-2310-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2310

           Summary: codegening structure variable definition+initialization
                    triggers assertion
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: nunoplopes at sapo.pt
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1633)
 --> (http://llvm.org/bugs/attachment.cgi?id=1633)
test case

The attached code produces the following assertion (with -emit-llvm):
clang: CodeGenFunction.h:183: llvm::Value* clang::CodeGen::LValue::getAddress()
const: Assertion `isSimple()' failed.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 11 08:30:02 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 May 2008 08:30:02 -0500
Subject: [LLVMbugs] [Bug 2311] New: LLVM codegen unnecessarily realigns
 stack for x86 SSE code
Message-ID: <bug-2311-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2311

           Summary: LLVM codegen unnecessarily realigns stack for x86 SSE
                    code
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


#include <emmintrin.h>

typedef short vSInt16 __attribute__ ((__vector_size__ (16)));

static const vSInt16 a = {- 22725, - 12873, - 22725, - 12873, - 22725, - 12873,
- 22725, - 12873};;

vSInt16 madd(vSInt16 b)
{
    return _mm_madd_epi16(a, b);
}

Generated code:
madd:
        pushl   %ebp
        movl    %esp, %ebp
        andl    $-16, %esp
        movaps  .LCPI1_0, %xmm1
        pmaddwd %xmm1, %xmm0
        movl    %ebp, %esp
        popl    %ebp
        ret

The stack alignment is unnecessary because nothing is spilled onto the stack.

Probably not a big issue, but it looks silly.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 11 09:29:06 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 May 2008 09:29:06 -0500
Subject: [LLVMbugs] [Bug 2311] LLVM codegen unnecessarily realigns stack for
 x86 SSE code
In-Reply-To: <bug-2311-206@http.llvm.org/bugs/>
Message-ID: <200805111429.m4BET6OF013730@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2311


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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |evan.cheng at apple.com,
                   |                            |asl at math.spbu.ru
             Status|NEW                         |RESOLVED
         Resolution|                            |LATER




--- Comment #1 from Anton Korobeynikov <asl at math.spbu.ru>  2008-05-11 09:29:06 ---
Unfortunately, we need to know about stack alignment too early, before RA runs.
At that point we don't know, whether there will be vector spill, or not. Stack
realignment logic is overly conservative here, but otherwise we can produce
unaligned loads/stores.

Fixing this will require some huge RA changes.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 11 09:53:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 May 2008 09:53:39 -0500
Subject: [LLVMbugs] [Bug 2312] New: GVN is buggy
Message-ID: <bug-2312-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2312

           Summary: GVN is buggy
           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
                CC: resistor at mac.com, llvmbugs at cs.uiuc.edu


Created an attachment (id=1634)
 --> (http://llvm.org/bugs/attachment.cgi?id=1634)
Bytecode in question

Consider attached testcase.

Running opt -gvn results to bytecode, which produces different (and invalid!)
results.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 11 11:42:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 May 2008 11:42:09 -0500
Subject: [LLVMbugs] [Bug 2313] New: ARM: Rd and Rm registers should be
 different in MUL on ARMv5
Message-ID: <bug-2313-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2313

           Summary: ARM: Rd and Rm registers should be different in MUL on
                    ARMv5
           Product: libraries
           Version: trunk
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: ARM
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: jhw at conjury.org
                CC: llvmbugs at cs.uiuc.edu


The ARMv6 CPU allows the Rd and Rm registers to be the same in the MUL
instruction, but this isn't allowed on, for example the ARMv5 CPU.  As a
result, GAS will issue complaints about LLVM-generated ARM code.  Here's a
sample:

sample.s: Assembler messages:
sample.s:59: Rd and Rm should be different in mul
sample.s:63: Rd and Rm should be different in mul
sample.s:104: Rd and Rm should be different in mul
sample.s:108: Rd and Rm should be different in mul

These messages came from assembling the following LLVM-generated code:

        .globl  rand_r
        .align  2
rand_r:
        ldr r3, [r0]
        ldr r2, .LCPI3_0
        cmp r3, #0
        moveq r3, r2
        ldr r2, .LCPI3_1
        umull r1, r2, r3, r2
        sub r1, r3, r2
        add r2, r2, r1, lsr #1
        mov r2, r2, lsr #16
        mov r1, #197, 30 @ 788
        orr r1, r1, #2, 22 @ 2048
        mul r1, r2, r1
        ldr r12, .LCPI3_2
        mul r2, r2, r12          ; <------------ Here is line 59.
        sub r3, r3, r2
        mov r2, #167
        orr r2, r2, #65, 24 @ 16640
        mul r3, r3, r2            ; <------------ Here is line 63.
        sub r3, r3, r1
        sub r2, r3, #6, 2 @ -2147483647
        cmp r3, #0
        movge r2, r3
        str r2, [r0]
        bic r0, r2, #2, 2 @ -2147483648
        bx lr
.LBB3_1:
        .LCPI3_0:
        .long   123459876

        .LCPI3_1:
        .long   110892733

        .LCPI3_2:
        .long   127773

Yes, this really is the rand_r() function from the standard C library.

This code is fine for ARMv6, but it doesn't work for the ARMv5 CPU I'm playing
with, so here is the bug report.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 11 11:59:21 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 May 2008 11:59:21 -0500
Subject: [LLVMbugs] [Bug 2314] New: Poor codegen of select + vector ops
Message-ID: <bug-2314-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2314

           Summary: Poor codegen of select + vector ops
           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: resistor at mac.com
                CC: llvmbugs at cs.uiuc.edu


This message to the list seems to have gotten lost:

Hello LLVM-ers, 

I'm now working on writing LLVM IR backend for my SIMD language MUDA. 

http://lucille.sourceforge.net/muda/

LLVM is so nice, I observed LLVM do good SIMD optimization 
during initial my MUDA -> LLVM IR experience. 

but I found, for x86 backend, "select" instruction is mapped to branch, 
not triple of and*/andn*/or* (or blend* in SSE4 case), 
as described in lib/Target/X86/README-SSE.txt. 

Is there any (technical) reason that current(I'm using LLVM trunk) LLVM x86
backend 
doesn't do such a optimization? 

Is it easy for the LLVM newbie to write a patch to emit 
and/andn/or x86 instruction for "select" LLVM instruction? 
If so, where should I look at to start to hack it? 

Thanks in advance, 

Syoyo 



FYI, I wrote a following portable vectorized max() function 

;; max.ll 
define <4xfloat> @muda_maxf4(<4xfloat> %a, <4xfloat> %b) 
{ 

        ;; extract 
        %a0 = extractelement <4xfloat> %a, i32 0 
        %a1 = extractelement <4xfloat> %a, i32 1 
        %a2 = extractelement <4xfloat> %a, i32 2 
        %a3 = extractelement <4xfloat> %a, i32 3 

        %b0 = extractelement <4xfloat> %b, i32 0 
        %b1 = extractelement <4xfloat> %b, i32 1 
        %b2 = extractelement <4xfloat> %b, i32 2 
        %b3 = extractelement <4xfloat> %b, i32 3 

        ;; c[N] = a[N] > b[N] 
        %c0 = fcmp ogt float %a0, %b0 
        %c1 = fcmp ogt float %a1, %b1 
        %c2 = fcmp ogt float %a2, %b2 
        %c3 = fcmp ogt float %a3, %b3 

        ;; if %c[N] == 1 then %a[N] else %b[N] 

        %r0 = select i1 %c0, float %a0, float %b0 
        %r1 = select i1 %c1, float %a1, float %b1 
        %r2 = select i1 %c2, float %a2, float %b2 
        %r3 = select i1 %c3, float %a3, float %b3 

        ;; pack 

        %tmp0 = insertelement <4xfloat> undef, float %r0, i32 0 
        %tmp1 = insertelement <4xfloat> %tmp0, float %r1, i32 1 
        %tmp2 = insertelement <4xfloat> %tmp1, float %r2, i32 2 
        %r    = insertelement <4xfloat> %tmp2, float %r3, i32 3 

        ret <4xfloat> %r 
} 

and got following assembly code, 

$ llvm-as max.ll; opt -std-compile-opts max.bc -f | llc -march=x86 -mcpu=penryn
-f 

        .text 
        .align  16 
        .globl  muda_maxf4 
        .type   muda_maxf4, at function 
muda_maxf4: 
        extractps       $3, %xmm1, %xmm2 
        extractps       $3, %xmm0, %xmm3 
        ucomiss %xmm2, %xmm3 
        ja      .LBB1_2 # 
.LBB1_1:        # 
        movaps  %xmm2, %xmm3 
.LBB1_2:        # 
        extractps       $1, %xmm1, %xmm2 
        extractps       $1, %xmm0, %xmm4 
        ucomiss %xmm2, %xmm4 
        ja      .LBB1_4 # 
.LBB1_3:        # 
        movaps  %xmm2, %xmm4 
.LBB1_4:        # 
        movss   %xmm4, %xmm2 
        unpcklps        %xmm3, %xmm2 
        extractps       $2, %xmm1, %xmm3 
        extractps       $2, %xmm0, %xmm4 
        ucomiss %xmm3, %xmm4 
        ja      .LBB1_6 # 
.LBB1_5:        # 
        movaps  %xmm3, %xmm4 
.LBB1_6:        # 
        ucomiss %xmm1, %xmm0 
        ja      .LBB1_8 # 
.LBB1_7:        # 
        movaps  %xmm1, %xmm0 
.LBB1_8:        # 
        unpcklps        %xmm4, %xmm0 
        unpcklps        %xmm2, %xmm0 
        ret 
        .size   muda_maxf4, .-muda_maxf4 

where I expected to get one of following 

- cmpps + andps + andnps + orps 
- cmpps + blendps(in SSE4 case) 
- maxps


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 11 12:17:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 11 May 2008 12:17:38 -0500
Subject: [LLVMbugs] [Bug 2315] New: Missing break in LegalizeOp case
Message-ID: <bug-2315-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2315

           Summary: Missing break in LegalizeOp case
           Product: libraries
           Version: 2.2
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: critical
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: giuma.cordes at gmail.com
                CC: llvmbugs at cs.uiuc.edu


There is a missing break in lib/CodeGen/SelectionDAG/LegalizeDAG.cpp, line
4002, causing an erroneous fallthrough to the next case.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 08:02:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 08:02:45 -0500
Subject: [LLVMbugs] [Bug 2264] internal compiler error
In-Reply-To: <bug-2264-206@http.llvm.org/bugs/>
Message-ID: <200805121302.m4CD2jaC009200@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2264


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #8 from Duncan Sands <baldrick at free.fr>  2008-05-12 08:02:44 ---
Fixed differently here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080512/062263.html


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 10:07:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 10:07:11 -0500
Subject: [LLVMbugs] [Bug 2310] codegening structure variable
 definition+initialization triggers assertion
In-Reply-To: <bug-2310-206@http.llvm.org/bugs/>
Message-ID: <200805121507.m4CF7BRk012796@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2310


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-12 10:07:11 ---
Fixed in r50966.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 10:43:20 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 10:43:20 -0500
Subject: [LLVMbugs] [Bug 2277] llc -enable-correct-eh-support: " Not all
 catch info was assigned to a landing pad!"' failed
In-Reply-To: <bug-2277-206@http.llvm.org/bugs/>
Message-ID: <200805121543.m4CFhKk7013851@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2277


Duncan Sands <baldrick at free.fr> changed:

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




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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 11:08:33 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 11:08:33 -0500
Subject: [LLVMbugs] [Bug 2307] switch without body crashes code generator
In-Reply-To: <bug-2307-206@http.llvm.org/bugs/>
Message-ID: <200805121608.m4CG8Xog014718@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2307


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-12 11:08:32 ---
Fixed in r50968.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 12:33:33 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 12:33:33 -0500
Subject: [LLVMbugs] [Bug 2308] invalid memory read/write with clang
	serialized input
In-Reply-To: <bug-2308-206@http.llvm.org/bugs/>
Message-ID: <200805121733.m4CHXXU8017694@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2308


Ted Kremenek <kremenek at apple.com> changed:

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




--- Comment #2 from Ted Kremenek <kremenek at apple.com>  2008-05-12 12:33:32 ---
Fixed in r50976:

 
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080512/005676.html

Diagnosis:

Incorrect serialization of "decl context" in ScopedDecl.  Essentially a local
variable on the stack was being registered with the deserializer instead of the
DeclCtx member variable of the ScopedDecl object.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 13:19:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 13:19:43 -0500
Subject: [LLVMbugs] [Bug 2316] New: Suspected misoptimization
Message-ID: <bug-2316-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2316

           Summary: Suspected misoptimization
           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: sanxiyn at gmail.com
                CC: llvmbugs at cs.uiuc.edu


SVN r50980.

llvm-gcc -O2 and -O3 produce different results for the attached file.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 13:30:36 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 13:30:36 -0500
Subject: [LLVMbugs] [Bug 2316] Suspected misoptimization
In-Reply-To: <bug-2316-206@http.llvm.org/bugs/>
Message-ID: <200805121830.m4CIUacC019789@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2316


Seo Sanghyeon <sanxiyn at gmail.com> changed:

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




--- Comment #2 from Seo Sanghyeon <sanxiyn at gmail.com>  2008-05-12 13:30:35 ---
Sorry for the noise.

*** This bug has been marked as a duplicate of bug 2312 ***


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 13:32:36 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 13:32:36 -0500
Subject: [LLVMbugs] [Bug 2263] clang doesn't pick up type attribute on
	argument
In-Reply-To: <bug-2263-206@http.llvm.org/bugs/>
Message-ID: <200805121832.m4CIWaL9019886@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2263


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-12 13:32:35 ---
Nate fixed this:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080505/005658.html

Testcase here:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080512/005681.html


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 14:06:06 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 14:06:06 -0500
Subject: [LLVMbugs] [Bug 2317] New: new vicmp/vfcmp constant exprs not being
 read/ written correctly?
Message-ID: <bug-2317-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2317

           Summary: new vicmp/vfcmp constant exprs not being read/written
                    correctly?
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Bitcode Reader
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: natebegeman at mac.com
                CC: llvmbugs at cs.uiuc.edu


Attached testcase does not round-trip through llvm-as / llvm-dis


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 14:51:52 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 14:51:52 -0500
Subject: [LLVMbugs] [Bug 2315] Missing break in LegalizeOp case
In-Reply-To: <bug-2315-206@http.llvm.org/bugs/>
Message-ID: <200805121951.m4CJpqAY022596@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2315


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #2 from Dan Gohman <gohman at apple.com>  2008-05-12 14:51:47 ---
Thanks! This patch is now applied on trunk and the 2.3 release branch:
http://llvm.org/viewvc/llvm-project?rev=50967&view=rev
http://llvm.org/viewvc/llvm-project?rev=50989&view=rev


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 15:22:51 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 15:22:51 -0500
Subject: [LLVMbugs] [Bug 2318] New: Instruction Combining fails assertion
	check
Message-ID: <bug-2318-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2318

           Summary: Instruction Combining fails assertion check
           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: lefever at crhc.uiuc.edu
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1640)
 --> (http://llvm.org/bugs/attachment.cgi?id=1640)
Bugpoint reduced bytecode

The attached bytecode causes an assertion failure in opt when you run:

opt bugpoint-reduced-simplified.bc -instcombine

The assertion failure message is:

opt: /home/local/lefever/llvm/080509/llvm/include/llvm/Support/Casting.h:199:
typename llvm::cast_retty<To, From>::ret_type llvm::cast(const Y&) [with X =
llvm::ConstantInt, Y = llvm::Value*]: Assertion `isa<X>(Val) && "cast<Ty>()
argument of incompatible type!"' failed.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 16:43:57 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 16:43:57 -0500
Subject: [LLVMbugs] [Bug 2319] New: Call result #2 has unhandled type i32
Message-ID: <bug-2319-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2319

           Summary: Call result #2 has unhandled type i32
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: dpatel at apple.com
                CC: llvmbugs at cs.uiuc.edu


Attached test case causes following failure on darwin-x86.

Call result #2 has unhandled type i32
0   llc                                 0x005c2fe1
_ZNSt8_Rb_treeIN4llvm3sys4PathES2_St9_IdentityIS2_ESt4lessIS2_ESaIS2_EE13insert_uniqueERKS2_
+ 7441
1   libSystem.B.dylib                   0x930bd97b _sigtramp + 43
2   ???                                 0xffffffff 0x0 + 4294967295
3   libSystem.B.dylib                   0x93136782 raise + 26
4   libSystem.B.dylib                   0x93145d3f abort + 73
5   llc                                 0x00217be1
_ZN4llvm15BitstreamReader18ReadBlockInfoBlockEv + 3729
6   llc                                 0x0002e68f
_ZN4llvm2cl3optINS_12PluginLoaderELb0ENS0_6parserISsEEED0Ev + 165023
7   llc                                 0x0003b10f
_ZN4llvm2cl3optINS_12PluginLoaderELb0ENS0_6parserISsEEED0Ev + 216863
8   llc                                 0x0004dcb1
_ZN4llvm2cl3optINS_12PluginLoaderELb0ENS0_6parserISsEEED0Ev + 293569
9   llc                                 0x0025c189
_ZSt6removeIN9__gnu_cxx17__normal_iteratorIPPN4llvm6SDNodeESt6vectorIS4_SaIS4_EEEES4_ET_SA_SA_RKT0_
+ 70345
10  llc                                 0x002a4d27
_ZSt6removeIN9__gnu_cxx17__normal_iteratorIPPN4llvm6SDNodeESt6vectorIS4_SaIS4_EEEES4_ET_SA_SA_RKT0_
+ 368231
11  llc                                 0x00294d7a
_ZSt6removeIN9__gnu_cxx17__normal_iteratorIPPN4llvm6SDNodeESt6vectorIS4_SaIS4_EEEES4_ET_SA_SA_RKT0_
+ 302778
12  llc                                 0x00259aca
_ZSt6removeIN9__gnu_cxx17__normal_iteratorIPPN4llvm6SDNodeESt6vectorIS4_SaIS4_EEEES4_ET_SA_SA_RKT0_
+ 60426
13  llc                                 0x002aa524
_ZSt6removeIN9__gnu_cxx17__normal_iteratorIPPN4llvm6SDNodeESt6vectorIS4_SaIS4_EEEES4_ET_SA_SA_RKT0_
+ 390756
14  llc                                 0x002ab3ac
_ZSt6removeIN9__gnu_cxx17__normal_iteratorIPPN4llvm6SDNodeESt6vectorIS4_SaIS4_EEEES4_ET_SA_SA_RKT0_
+ 394476
15  llc                                 0x002fe33e _ZN4llvm12AtomicSDNodeD1Ev +
1518
16  llc                                 0x00336320 _ZN4llvm12AtomicSDNodeD1Ev +
230864
17  llc                                 0x0033880e _ZN4llvm12AtomicSDNodeD1Ev +
240318
18  llc                                 0x00075b32
_ZN4llvm16SelectionDAGISel14SelectRootInitEv + 82
19  llc                                 0x00554ea6 _ZN4llvm12FunctionPassD1Ev +
22214
20  llc                                 0x005554c4 _ZN4llvm12FunctionPassD1Ev +
23780
21  llc                                 0x00555646 _ZN4llvm12FunctionPassD1Ev +
24166
22  llc                                 0x00003054 _mh_execute_header + 8276
23  llc                                 0x00002136 _mh_execute_header + 4406
24  ???                                 0x00000004 0x0 + 4
Abort trap


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 18:43:26 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 18:43:26 -0500
Subject: [LLVMbugs] [Bug 2320] New: clang crashes on nested enum
Message-ID: <bug-2320-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2320

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


enum e0 {E0 = sizeof(enum e0 { E1 })};

Run with clang -fsyntax-only:
clang: /home/eli/llvm/tools/clang/lib/Sema/SemaDecl.cpp:1850: virtual void
clang::Sema::ActOnEnumBody(clang::SourceLocation, void*, void**, unsigned int):
Assertion `!Enum->isDefinition() && "Enum redefinitions can't reach here"'
failed.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 18:50:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 18:50:29 -0500
Subject: [LLVMbugs] [Bug 2321] New: clang crashes redeclaring union tag as
	enum
Message-ID: <bug-2321-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2321

           Summary: clang crashes redeclaring union tag as enum
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


union u0;
enum u0 { U0A };

Crashes with assertion:
clang: /home/eli/llvm/include/llvm/Support/Casting.h:199: typename
llvm::cast_retty<To, From>::ret_type llvm::cast(const Y&) [with X =
clang::EnumDecl, Y = clang::Decl*]: Assertion `isa<X>(Val) && "cast<Ty>()
argument of incompatible type!"' failed.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 19:20:13 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 19:20:13 -0500
Subject: [LLVMbugs] [Bug 2322] New: clang overflows the stack with many case
	statements
Message-ID: <bug-2322-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2322

           Summary: clang overflows the stack with many case statements
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase from GCC testsuite:

#define LIM1(x) x##0: x##1: x##2: x##3: x##4: x##5: x##6: x##7: x##8: x##9: 
#define LIM2(x) LIM1(x##0) LIM1(x##1) LIM1(x##2) LIM1(x##3) LIM1(x##4) \
                LIM1(x##5) LIM1(x##6) LIM1(x##7) LIM1(x##8) LIM1(x##9)
#define LIM3(x) LIM2(x##0) LIM2(x##1) LIM2(x##2) LIM2(x##3) LIM2(x##4) \
                LIM2(x##5) LIM2(x##6) LIM2(x##7) LIM2(x##8) LIM2(x##9)
#define LIM4(x) LIM3(x##0) LIM3(x##1) LIM3(x##2) LIM3(x##3) LIM3(x##4) \
                LIM3(x##5) LIM3(x##6) LIM3(x##7) LIM3(x##8) LIM3(x##9)
#define LIM5(x) LIM4(x##0) LIM4(x##1) LIM4(x##2) LIM4(x##3) LIM4(x##4) \
                LIM4(x##5) LIM4(x##6) LIM4(x##7) LIM4(x##8) LIM4(x##9)
#define LIM6(x) LIM5(x##0) LIM5(x##1) LIM5(x##2) LIM5(x##3) LIM5(x##4) \
                LIM5(x##5) LIM5(x##6) LIM5(x##7) LIM5(x##8) LIM5(x##9)
#define LIM7(x) LIM6(x##0) LIM6(x##1) LIM6(x##2) LIM6(x##3) LIM6(x##4) \
                LIM6(x##5) LIM6(x##6) LIM6(x##7) LIM6(x##8) LIM6(x##9)

void q19_func (long i)
{
  switch (i) {
    LIM5 (case 1)
      break;
  }
}

clang crashes with a stack overflow because it recursively constructs the case
statements.

Not sure if this is worth fixing to work, but it would at least be nice if it
didn't crash.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 12 21:41:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 12 May 2008 21:41:29 -0500
Subject: [LLVMbugs] [Bug 2323] New: small input causes cc1 to use all RAM
Message-ID: <bug-2323-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2323

           Summary: small input causes cc1 to use all RAM
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


Using svn 50966 on Ubuntu Feisty on x86-32:

[regehr at babel tmp2]$ ulimit -v 1500000
[regehr at babel tmp2]$ llvm-gcc -Os -S small.c
terminate called after throwing an instance of 'std::bad_alloc'
  what():  std::bad_alloc
*** glibc detected *** /home/regehr/libexec/gcc/i686-pc-linux-gnu/4.2.1/cc1:
free(): invalid pointer: 0x08c67640 ***
llvm-gcc: Internal error: Aborted (program cc1)
Please submit a full bug report.
See <URL:http://developer.apple.com/bugreporter> for instructions.
[regehr at babel tmp2]$ cat small.c

static inline unsigned long int
mod_rhs (const long int rhs)
{
}

int g_66 = -1;
int g_79 = 0xF7D1D410;
int func_29 (int p_30, signed char p_32, unsigned int p_34);
int func_62 (unsigned char p_64, int p_65, int p_67, unsigned char p_68);
int
func_27 (int p_28)
{
  unsigned char l_368 = -1;
  p_28 = func_29 (l_368, p_28, 1);
}

int
func_29 (int p_30, signed char p_32, unsigned int p_34)
{
  if (0x5ADF851F)
    {
      int l_131 = -1;
      unsigned int l_139 = -2;
      short l_143 = 1;
      if (p_32)
        {
          func_12 (1 * p_34 >= 1 - 1);
        }
      else
        {
          g_79 = p_30;
        }
      for (p_34 = -9; p_34 >= 0; --p_34)
        {
          unsigned char l_142 = 4;
          if (p_30)
            {
            }
          else
            {
              return l_139;
            }
          l_131 =
            1 > 1 < p_32 !=
            (1 % mod_rhs (func_62 (l_142, l_142, l_143, p_30))) | 1 *
            p_30 + 1;
        }
    }
}
int
func_62 (unsigned char p_64, int p_65, int p_67, unsigned char p_68)
{
  for (g_66 = -19; g_66 == 0; ++g_66)
    {
    }
}


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 13 06:53:21 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 May 2008 06:53:21 -0500
Subject: [LLVMbugs] [Bug 2324] New: gvn non-local CSE miscompilation
Message-ID: <bug-2324-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2324

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


Created an attachment (id=1645)
 --> (http://llvm.org/bugs/attachment.cgi?id=1645)
testcase .ll

The recent gvn non-local CSE changes broke several tests from the
Ada testsuite.  I reduced one down to the attached (but still large)
testcase.  The visible problem is that gvn thinks that the two memcmp
calls have the same results and eliminates one, but this is bogus
because %var_ara_8 changes between the two calls.

Reproduce using:
  opt -gvn
and observe the transformation of the memcpy calls.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 13 13:56:54 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 May 2008 13:56:54 -0500
Subject: [LLVMbugs] [Bug 2325] New: clang crashes codegen'ing goto in
	compound statement
Message-ID: <bug-2325-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2325

           Summary: clang crashes codegen'ing goto in compound statement
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
int a() { return ({goto L; 0;}) && ({L: 5;});}

Crashes with clang -emit-llvm.

Apparently this is supposed to return 1, although I have no clue how this is
supposed to work.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 13 15:07:44 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 May 2008 15:07:44 -0500
Subject: [LLVMbugs] [Bug 2323] inline + tailduplicate infinite loop
In-Reply-To: <bug-2323-206@http.llvm.org/bugs/>
Message-ID: <200805132007.m4DK7iZZ001854@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2323


Dale Johannesen <dalej at apple.com> changed:

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




--- Comment #5 from Dale Johannesen <dalej at apple.com>  2008-05-13 15:07:44 ---
Fixed here.
http://llvm.org/viewvc/llvm-project?view=rev&revision=51063


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 13 16:26:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 13 May 2008 16:26:00 -0500
Subject: [LLVMbugs] [Bug 2324] gvn non-local CSE miscompilation
In-Reply-To: <bug-2324-206@http.llvm.org/bugs/>
Message-ID: <200805132126.m4DLQ0uN004334@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2324


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #1 from Owen Anderson <resistor at mac.com>  2008-05-13 16:25:59 ---
Fixed in r51069.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 14 11:36:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 May 2008 11:36:30 -0500
Subject: [LLVMbugs] [Bug 2326] New: ICE
Message-ID: <bug-2326-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2326

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


Using svn 51018 on Ubuntu Feisty on x86-32 I get this:

[regehr at babel tmp3]$ llvm-gcc small.c
cc1: SelectionDAG.cpp:1071: llvm::SDOperand
llvm::SelectionDAG::FoldSetCC(llvm::MVT::ValueType, llvm::SDOperand,
llvm::SDOperand, llvm::ISD::CondCode): Assertion
`!MVT::isInteger(N1.getValueType()) && "Illegal setcc for integer!"' failed.
small.c:55: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.





typedef signed char int8_t;
typedef short int int16_t;
typedef int int32_t;
typedef unsigned char uint8_t;
typedef unsigned short int uint16_t;
typedef unsigned int uint32_t;
int8_t g_39 = 0x23L;
uint16_t g_67 = 0x3DF5L;
uint8_t g_118 = 0x5DL;
int32_t func_4 ;
int32_t
func_59 (uint32_t p_60)
{
  int32_t l_61 = 0x86C3CE99L;
  uint32_t l_117 = -2L;
  int16_t l_127 = 4L;
  uint8_t l_165 = 1L;
  for (p_60 = -24;; ++p_60)
    {
      uint32_t l_108 = 0L;
      uint8_t l_125 = 0x9DL;
      int32_t l_132 = 0x41C1EEBAL;
      int32_t l_167 = 0L;
      if (g_39)
        {
          int8_t l_119 = 0xE6L;
          uint16_t l_163 = 1L;
          if (g_39)

              for (l_61 = 0;; --p_60)
                return g_67;

          else
            {
              uint16_t l_116 = 0xC021L;
              if (lshift_s_s
                  (l_116 <
                   rshift_s_u (func_15 (p_60 == l_108 & p_60 <= l_108, l_117),
                               p_60) < func_4, 0xF14100A9L))
                return g_118;
              for (l_125 = -26;; l_125 += 0)

              if (l_163)
                {
                  uint8_t l_164 = 0L;
                  for (l_164 = 9;; ++l_164)
                    return l_165;
                }
              else
                for (l_119 = -20;; l_119 += 5)
                  return l_167;
            }
        }
    }
}


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 14 12:41:51 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 May 2008 12:41:51 -0500
Subject: [LLVMbugs] [Bug 2327] New: Collector back end uses inconsistent
 code for finding frame offsets
Message-ID: <bug-2327-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2327

           Summary: Collector back end uses inconsistent code for finding
                    frame offsets
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: maxburke at gmail.com
                CC: llvmbugs at cs.uiuc.edu


I just noticed in the trunk that the code in
MachineCodeAnalysis::FindStackOffsets() uses a different algorithm for finding
stack offsets than what the X86 backend uses for its overridden
TargetRegisterInfo::getFrameIndexOffsets. Should
MachineCodeAnalysis::FindStackOffsets perhaps be calling
TargetRegisterInfo::getFrameIndexOffsets instead?


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 14 13:22:20 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 14 May 2008 13:22:20 -0500
Subject: [LLVMbugs] [Bug 2326] selectiondag ice
In-Reply-To: <bug-2326-206@http.llvm.org/bugs/>
Message-ID: <200805141822.m4EIMKLO018965@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2326


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #1 from Dan Gohman <gohman at apple.com>  2008-05-14 13:22:17 ---
Pretty testcase :-). Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080512/062433.html


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 02:43:57 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 02:43:57 -0500
Subject: [LLVMbugs] [Bug 2061] Remove the tail duplication pass for llvm-gcc
 & -std-compile-opts
In-Reply-To: <bug-2061-206@http.llvm.org/bugs/>
Message-ID: <200805150743.m4F7hvSB017964@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2061


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

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




--- Comment #3 from Evan Cheng <evan.cheng at apple.com>  2008-05-15 02:43:56 ---
Poof.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 13:36:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 13:36:29 -0500
Subject: [LLVMbugs] [Bug 2328] New: Missed instbombine of x/x
Message-ID: <bug-2328-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2328

           Summary: Missed instbombine of x/x
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


unsigned foo(const unsigned x) {  return (x / x); }

Trivial bug: instcombine should combine x/x to 1 if x is an integer.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 14:10:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 14:10:08 -0500
Subject: [LLVMbugs] [Bug 2329] New: Comparison of constant inttoptrs not
	folded
Message-ID: <bug-2329-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2329

           Summary: Comparison of constant inttoptrs not folded
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


>From GCC Bug 17141:

void abort (void);
struct a
{
  int i;
} *a;
int f(void)
{
  int *ii = &a->i;
  void *l;
  a->i = 1;
  if (*ii)
   l = &&l1;
  else
   l = &&l2;
  goto *l;
l1:
  return 0;
l2:
  abort ();
}

This should fold to almost nothing, but we end up with a branch on a
complicated constant expression:
br i1 icmp eq (i8* inttoptr (i32 1 to i8*), i8* inttoptr (i32 2 to i8*)), label
%l2, label %l1

"icmp eq (i8* inttoptr (i32 1 to i8*), i8* inttoptr (i32 2 to i8*))" should
fold to false.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 14:43:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 14:43:48 -0500
Subject: [LLVMbugs] [Bug 2330] New: Misc missed instcombines
Message-ID: <bug-2330-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2330

           Summary: Misc missed instcombines
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Not sure it's really worthwhile to file a bug for each of these, so I guess
I'll collect them here, and then decide what to do with them.  Should I file
separate bugs instead?

>From GCC Bug 31261:
int a(int a) {return (8 - a) & 7;}
The "8 - a" should be simplified to "-a".

>From GCC Bug 24696:
int
f (unsigned long a, unsigned long b, unsigned long c)
{
  return ((a & (c - 1)) != 0) || ((b & (c - 1)) != 0);
}
The two icmps should be combined.

>From GCC Bug 20192:
#define PMD_MASK    (~((1UL << 23) - 1))
void clear_pmd_range(unsigned long start, unsigned long end) 
{ 
    if (!(start & ~PMD_MASK) && !(end & ~PMD_MASK)) 
        f(); 
}
The two icmps should be combined.

>From GCC Bug 15350:
void bar (void);
void
foo (int a)
{
  if ((1 << a) & 1)
    bar ();
}
Should be combined to a != 0.

>From GCC Bug 15241:
unsigned int
foo (unsigned int a, unsigned int b)
{
  if (a <= 7 && b <= 7)
    baz ();
}
Should combine to "(a|b) <= 7".

>From GCC Bug 3756:
int
pn (int n)
{
  return (n >= 0 ? 1 : -1);
}
Should combine to (n >> 31) | 1.

>From GCC Bug 28685:
int test(int a, int b)
{
  int lt = a < b;
  int eq = a == b;

  return (lt || eq);
}
Should combine to "a <= b".


void a(int variable)
{
  if (variable == 4 || variable == 6)
    bar();
}
This should optimize to "if ((variable | 2) == 6)".


unsigned int f(unsigned int i, unsigned int n) {++i; if (i == n) ++i; return
i;}
unsigned int f2(unsigned int i, unsigned int n) {++i; i += i == n; return i;}
These should combine to the same thing.

>From GCC Bug 33512:
int f(int y, int x)
{
  return x & ((~x) | y);
}
Should combine to x & y.

>From GCC Bug 31096:
int f(int a, int b)
{
  return a * 10 == b * 10;
}
This can be optimized to a & 0x7FFFFFFF == b & 0x7FFFFFFF.

>From GCC Bug 15784:
#define abs(x) x>0?x:-x
int f(int x, int y)
{
  return (abs(x)) >= 0;
}
This should optimize to x == INT_MIN. (With -fwrapv.)

>From GCC Bug 14753:
void
rotate_cst (unsigned int a)
{
  a = (a << 10) | (a >> 22);
  if (a == 123)
    bar ();
}
void
minus_cst (unsigned int a)
{
  unsigned int tem;

  tem = 20 - a;
  if (tem == 5)
    bar ();
}
void
mask_gt (unsigned int a)
{
  /* This is equivalent to a > 15.  */
  if ((a & ~7) > 8)
    bar ();
}
void
rshift_gt (unsigned int a)
{
  /* This is equivalent to a > 23.  */
  if ((a >> 2) > 5)
    bar ();
}
All should simplify to a single comparison.

>From GCC Bug 32605:
int c(int* x) {return (char*)x+2 == (char*)x;}
Should combine to 0.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 21:06:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 21:06:00 -0500
Subject: [LLVMbugs] [Bug 2331] New: a patch for win32 build when there is no
 bison present
Message-ID: <bug-2331-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2331

           Summary: a patch for win32 build when there is no bison present
           Product: Build scripts
           Version: trunk
          Platform: PC
        OS/Version: Windows 2000
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Makefiles
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: humeafo at gmail.com
                CC: llvmbugs at cs.uiuc.edu


a patch for win32 build when there is no bison present
there is a mistake in the batch file which copy the needed files to the wrong
place


Index: D:/LLVM/llvm/win32/dobison.cmd

===================================================================

--- D:/LLVM/llvm/win32/dobison.cmd (revision 51136)

+++ D:/LLVM/llvm/win32/dobison.cmd (working copy)

@@ -22,6 +22,7 @@


 :nobison
 echo Bison not found.  Using pre-generated files.
-copy %~pn4.cpp.cvs %5\%3.cpp
+copy %~pn4.cpp.cvs .\%3.cpp
+copy %~pn4.h.cvs .\%3.h
 copy %~pn4.h.cvs %5\%3.h
 exit


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 21:19:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 21:19:53 -0500
Subject: [LLVMbugs] [Bug 2332] New: clang incorrectly rejects "a(a)int a;
 {a=10; return a; }"
Message-ID: <bug-2332-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2332

           Summary: clang incorrectly rejects "a(a)int a;{a=10;return a;}"
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Per summary, clang rejects the following code:
a(a)int a;{a=10;return a;}

both gcc and comeau accept this construct.  I haven't looked closely at the
standard, but I would be surprised if this weren't valid.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 21:29:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 21:29:12 -0500
Subject: [LLVMbugs] [Bug 2333] New: clang codegen crash with function
 declared without a prototype then defined
Message-ID: <bug-2333-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2333

           Summary: clang codegen crash with function declared without a
                    prototype then defined
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
int a();
int b()
{
  a(1);
}
int a(int b)
{
  return b;
}

crashes with clang -emit-llvm.

This in CodeGenFunction::GenerateCode because it adds a body to a function
declaration with an incomplete argument list.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 21:39:45 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 21:39:45 -0500
Subject: [LLVMbugs] [Bug 2334] New: clang crashes codegen' s conditional of
 compatible but different function pointers
Message-ID: <bug-2334-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2334

           Summary: clang crashes codegen's conditional of compatible but
                    different function pointers
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
int a();
int b(int);
int c() {return 1 ? a : b;}

crashes clang -emit-llvm.

Potential patch already sent to the list, but I want to make sure this doesn't
get lost.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 15 22:00:57 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 15 May 2008 22:00:57 -0500
Subject: [LLVMbugs] [Bug 2328] Missed instcombine of x/x
In-Reply-To: <bug-2328-206@http.llvm.org/bugs/>
Message-ID: <200805160300.m4G30vAa022838@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2328


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-05-15 22:00:57 ---
Implemented:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080512/062511.html


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 16 06:07:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 May 2008 06:07:22 -0500
Subject: [LLVMbugs] [Bug 2335] New: llvm-gcc generates llvm. memcpy
 intrinsics with -ffreestanding flag
Message-ID: <bug-2335-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2335

           Summary: llvm-gcc generates llvm.memcpy intrinsics with -
                    ffreestanding flag
           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: vjaaskel at cs.tut.fi
                CC: llvmbugs at cs.uiuc.edu


llvm-gcc turns some simple array initializations to llvm.memcpy intrinsics even
if the -ffreestanding flag is given. For example compiling the following code
with '-c --emit-llvm -O2 -ffreestanding' flags

   void foo() {
      char a[11] = "1234567890";
      bar(a);
   }

Initializes the array 'a' with:

   call void @llvm.memcpy.i32( i8* %a1, i8* getelementptr ([11 x i8]* @.str,
i32 0, i32 0), i32 11, i32 1 )


The llvm.memcpy.i32 intrinsic is lowered to a call to memcpy which is a problem
if memcpy is not available. I'm not sure what the correct behavior would be
here, but it doesn't seem right that memcpy is required for a simple array
initialization even if -ffreestanding flag is given.

This problem occurs at least with llvm-gcc-4.2.0 release for llvm 2.2.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 16 06:41:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 May 2008 06:41:16 -0500
Subject: [LLVMbugs] [Bug 2335] llvm-gcc generates llvm. memcpy intrinsics
 with -ffreestanding flag
In-Reply-To: <bug-2335-206@http.llvm.org/bugs/>
Message-ID: <200805161141.m4GBfGQ1014439@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2335


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #3 from Duncan Sands <baldrick at free.fr>  2008-05-16 06:41:15 ---
>From http://gcc.gnu.org/onlinedocs/gcc/Standards.html:

GCC requires the freestanding environment provide memcpy, memmove, memset and
memcmp. Finally, if __builtin_trap is used, and the target does not implement
the trap pattern, then GCC will emit a call to abort.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 16 22:05:59 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 16 May 2008 22:05:59 -0500
Subject: [LLVMbugs] [Bug 2336] New: llvm miscompiles tree-cfg.c
Message-ID: <bug-2336-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2336

           Summary: llvm miscompiles tree-cfg.c
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: rafael.espindola at gmail.com
        ReportedBy: rafael.espindola at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1647)
 --> (http://llvm.org/bugs/attachment.cgi?id=1647)
testcase

Bootstrap is broken on linux i686. The stage2 compiler segfaults. I have
reduced the problem to wrong code being produced for tree-cfg.c.

The attached tree-cfg.bc is correct. It was produced with -O0. Passing it by
"opt -std-compile-opts" produces an invalid output.

To reproduce the bug I am manually linking cc1. If someone wants I can upload
the ".o" files.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 04:16:49 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 04:16:49 -0500
Subject: [LLVMbugs] [Bug 2329] Comparison of constant inttoptrs not folded
In-Reply-To: <bug-2329-206@http.llvm.org/bugs/>
Message-ID: <200805170916.m4H9GnEv031980@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2329


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #3 from Nick Lewycky <nicholas at mxc.ca>  2008-05-17 04:16:47 ---
It now optimizes down to:

define i32 @f() nounwind  {
entry:
        %tmp1 = load %struct.a** @a, align 4            ; <%struct.a*>
[#uses=1]
        %tmp2 = getelementptr %struct.a* %tmp1, i32 0, i32 0            ;
<i32*> [#uses=1]
        store i32 1, i32* %tmp2, align 4
        ret i32 0
}


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 10:33:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 10:33:01 -0500
Subject: [LLVMbugs] [Bug 2272] Re-enable inlining at -O2 at above
In-Reply-To: <bug-2272-206@http.llvm.org/bugs/>
Message-ID: <200805171533.m4HFX1jt010894@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2272


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #6 from Duncan Sands <baldrick at free.fr>  2008-05-17 10:33:00 ---
I have re-enabled inlining at -O2, but with the same
threshold (100) used when optimizing for size:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080512/062578.html


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 11:10:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 11:10:07 -0500
Subject: [LLVMbugs] [Bug 2234] turn off gcc inliner from llvm-gcc
In-Reply-To: <bug-2234-206@http.llvm.org/bugs/>
Message-ID: <200805171610.m4HGA7mZ012038@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2234


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #16 from Duncan Sands <baldrick at free.fr>  2008-05-17 11:10:06 ---
The gcc early inliner has been turned off, along with the other early ipa
optimizations:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080512/062577.html
This may have broken attr(always_inline) - not sure.  If so, there should be
another PR for that.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 11:42:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 11:42:27 -0500
Subject: [LLVMbugs] [Bug 2337] New: SimplfyLibCalls should inline calls to
	abs
Message-ID: <bug-2337-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2337

           Summary: SimplfyLibCalls should inline calls to abs
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Per description, SimplfyLibCalls should simplify calls to abs.  This isn't
necessary for llvm-gcc, but it is useful for compilers that aren't clever
enough to fold calls to abs, like clang.

Simple testcase (in clang):
int a(int x) {return abs(x);}

This does bring up an interesting side-issue: can LLVM simplify calls to llabs?
llabs, as far as I can tell, only became a reserved external identifier in C99.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 13:06:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 13:06:34 -0500
Subject: [LLVMbugs] [Bug 2329] Comparison of constant inttoptrs not folded
In-Reply-To: <bug-2329-206@http.llvm.org/bugs/>
Message-ID: <200805171806.m4HI6Ygt015990@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2329


Nick Lewycky <nicholas at mxc.ca> changed:

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




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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 14:04:36 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 14:04:36 -0500
Subject: [LLVMbugs] [Bug 2338] New: instcombine isn't run after DSE
Message-ID: <bug-2338-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2338

           Summary: instcombine isn't run after DSE
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu


Surprisingly (to me at least) instcombine isn't run anytime late in the
optimizer pipeline.  This prevents llvm from eliminating the malloc/free in
this example, which requires dse to remove the store first.

void *malloc(__SIZE_TYPE__);
void free(void*);
int f(void)
{
  char *i = malloc(1);
  *i = 1;
  free (i);
}


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 14:11:22 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 14:11:22 -0500
Subject: [LLVMbugs] [Bug 2339] New: Instcoombine infinitely cycles
Message-ID: <bug-2339-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2339

           Summary: Instcoombine infinitely cycles
           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
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1648)
 --> (http://llvm.org/bugs/attachment.cgi?id=1648)
Bytecode in question

Consider attached testcase.

Typical output from opt -instcombine -debug is:

IC: Old =       mul i32 %tmp162163, ptrtoint (i32* getelementptr (i32* null,
i32 1) to i32)             ; <i32>:4 [#uses=1]
    New =       add i32 %tmp164, mul (i32 ptrtoint (i32* getelementptr (i32*
null, i32 1) to i32), i32 2)               ; <i32>:<badref> [#uses=0]
IC: Old =       add i32 %tmp164, mul (i32 ptrtoint (i32* getelementptr (i32*
null, i32 1) to i32), i32 2)               ; <i32>:4 [#uses=1]
    New =       mul i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to
i32), %tmp164165             ; <i32>:<badref> [#uses=0]


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 17:03:20 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 17:03:20 -0500
Subject: [LLVMbugs] [Bug 2340] New: SimplifyCFG doesn' t proerly clean up
 trivially dead instructions
Message-ID: <bug-2340-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2340

           Summary: SimplifyCFG doesn't proerly clean up trivially dead
                    instructions
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1649)
 --> (http://llvm.org/bugs/attachment.cgi?id=1649)
Testcase

To reproduce: run "opt -simplifycfg" on the attached testcase. Code after
simplifycfg:

define i32 @x(i32 %x) {
entry:
        %call = call i32 (...)* @foo( )         ; <i32> [#uses=0]
        %cond = icmp eq i32 %x, 9               ; <i1> [#uses=0]
        %call9 = call i32 (...)* @bar( )                ; <i32> [#uses=0]
        ret i32 0
}

The icmp is unused; SimplifyCFG should have eliminated it when it eliminated
the branch that used it.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 17 23:12:02 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 17 May 2008 23:12:02 -0500
Subject: [LLVMbugs] [Bug 2339] Instcombine infinitely cycles
In-Reply-To: <bug-2339-206@http.llvm.org/bugs/>
Message-ID: <200805180412.m4I4C2Z9003880@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2339


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #3 from Chris Lattner <sabre at nondot.org>  2008-05-17 23:12:01 ---
Thanks, patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080512/062585.html


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 18 18:50:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 18 May 2008 18:50:32 -0500
Subject: [LLVMbugs] [Bug 2341] New: memcmp() optimization is erroneous
Message-ID: <bug-2341-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2341

           Summary: memcmp() optimization is erroneous
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: MacOS X
            Status: NEW
          Severity: critical
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: scottm at aero.org
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1651)
 --> (http://llvm.org/bugs/attachment.cgi?id=1651)
memcmp() optimization bug test case

After a long haitus on an internal work task, I'm (yet again) building llvm
from bootstrap (Release native compiled llvm -> llvm gcc 4.2 -> Release
llvm-gcc compiled llvm). Found a really good bug that I've been able to narrow
down to a test case.

The offending assembly code is:

        movw    _.str-"L4$pb"(%esi), %ax
        cmpw    %ax, _buf-"L4$pb"+4(%esi)

which is comparing two 16-bit words, not two 32-bit words. This allows things
like "code" and "cond" to be memcmp()-ed as equal. This is clearly erroneous.

This is TOT code.

To reproduce: llvm-c++ -O3 testcase.cpp -lstdc++
Output produced is: was_token(buf + 4, 4) = 1
Output should be: was_token(buf + 4, 4) = 0


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 01:46:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 01:46:32 -0500
Subject: [LLVMbugs] [Bug 2342] New: Can't assemble the output of llc
Message-ID: <bug-2342-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2342

           Summary: Can't assemble the output of llc
           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: viridia at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1653)
 --> (http://llvm.org/bugs/attachment.cgi?id=1653)
example bitcode file

I've attached an example .bc file showing the problem. When I attempt to
generate an assembly file via llc, and then attempt to pass it to gcc or as, it
gets a large number of errors complaining about "extra junk on line"


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 04:34:42 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 04:34:42 -0500
Subject: [LLVMbugs] [Bug 2341] memcmp() optimization is erroneous
In-Reply-To: <bug-2341-206@http.llvm.org/bugs/>
Message-ID: <200805190934.m4J9YgLf000671@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2341


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #2 from Duncan Sands <baldrick at free.fr>  2008-05-19 04:34:41 ---
Fixed in r51240.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 12:03:20 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 12:03:20 -0500
Subject: [LLVMbugs] [Bug 2343] New: llvm-gcc bootstrap is broken
Message-ID: <bug-2343-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2343

           Summary: llvm-gcc bootstrap is broken
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: critical
          Priority: P2
         Component: Common Code Generator Code
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: asl at math.spbu.ru
                CC: evan.cheng at apple.com, llvmbugs at cs.uiuc.edu,
                    nlewycky at google.com


Created an attachment (id=1655)
 --> (http://llvm.org/bugs/attachment.cgi?id=1655)
Function is question

It seems, that recent taildup changes revealed some subtle bugs not known
before. Consider attached .bc file. The C snippet in question is:

        tree cases = get_cases_for_edge (e, stmt);

        /* If we have a list of cases associated with E, then use it
           as it's a lot faster than walking the entire case vector.  */
        if (cases)
          {
            edge e2 = find_edge (e->src, dest);
            tree last, first;

            first = cases;
            while (cases)
              {
                last = cases;
                CASE_LABEL (cases) = label;
                cases = TREE_CHAIN (cases);
              }

            /* If there was already an edge in the CFG, then we need
               to move all the cases associated with E to E2.  */
             if (e2)
              {
                tree cases2 = get_cases_for_edge (e2, stmt);

                TREE_CHAIN (last) = TREE_CHAIN (cases2);
                TREE_CHAIN (cases2) = first;
              }
          }

>From this we can see, that last is the 'if (e2)' is always non-NULL. The LLVM
IR is also fine (only part showed):

bb392:          ; preds = %bb379, %bb319
        %last.0 = phi %struct.tree_node* [ undef, %bb319 ], [ %cases.0, %bb379
]                ; <%struct.tree_node*> [#use$
        %cases.0 = phi %struct.tree_node* [ %tmp314, %bb319 ], [ %tmp391,
%bb379 ]              ; <%struct.tree_node*> [#use$
        %tmp394 = icmp eq %struct.tree_node* %cases.0, null             ; <i1>
[#uses=1]
        br i1 %tmp394, label %bb397, label %bb326

This is header of the loop and jump from %bb319 is possible iff %tmp314 is
non-null, thus loop executes at least one time and %last.0 is properly
initialized (thus - non null).

Let's look into generated assembler:
        movl    %ecx, %ebx
        testl   %ebx, %ebx
        jne     .LBB1_47        # bb326
.LBB1_52:       # bb397
        testl   %eax, %eax
        je      .LBB1_80        # bb695
.LBB1_53:       # bb402
        movl    %esi, 4(%esp)
        movl    %eax, (%esp)
        call    get_cases_for_edge
        movl    (%eax), %esi
        movl    %esi, (%ebx)

here %last.0 is %ebx and we saw that after call %ebx is always zero! I suppose,
that codedegen 'merged' %last.0 and %cases.0 somehow (or some codegen passes)


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 15:50:26 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 15:50:26 -0500
Subject: [LLVMbugs] [Bug 2344] New: -pre-RA-sched=list-td generates bad code
Message-ID: <bug-2344-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2344

           Summary: -pre-RA-sched=list-td generates bad code
           Product: new-bugs
           Version: unspecified
          Platform: Other
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: Bart.Coppens at elis.ugent.be
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1656)
 --> (http://llvm.org/bugs/attachment.cgi?id=1656)
Testcase showing the faulty behaviour

The attached file seems to cause list-td to generate bad code. Compiled with
llc without list-td, it produces a fine result, but with list-td, the result
seems wrong.

Excpected behaviour:
Since toBool is false in '@mystery' upon the call from main, all useless_...
values should be irrelevant to the eventual outcome. Since notToBool is true,
the code at the end should properly write '13' to %tmp, read the same value
from %tmp again and return it after sign-extending it.

Actual behaviour:
Returns '13' when compiled with llc. Returns '-65' (presumably it's just a
random value on the stack) when -pre-RA-sched=list-td is enabled.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 15:57:25 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 15:57:25 -0500
Subject: [LLVMbugs] [Bug 2345] New: ICE from svn 51266
Message-ID: <bug-2345-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2345

           Summary: ICE from svn 51266
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: regehr at cs.utah.edu
                CC: llvmbugs at cs.uiuc.edu


This is using LLVM for x86-32 on Ubuntu Feisty.

I hesitate to even report this bug because the crash is on bogus code -- I was
using LLVM for x86 to (mis)compile some C that contains AVR-specific inline
asm.  But anyway:

[regehr at babel tmp6]$  llvm-gcc -S -O0 small.c 
cc1: SelectionDAGISel.cpp:4184: void
llvm::SelectionDAGLowering::visitInlineAsm(llvm::CallSite): Assertion
`(OpInfo.ConstraintType == TargetLowering::C_RegisterClass ||
OpInfo.ConstraintType == TargetLowering::C_Register) && "Unknown constraint
type!"' failed.
small.c:6: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.

[regehr at babel tmp6]$ cat small.c
int *foo;

void hello (void)
{
  __asm__ ("lpm %0, Z"::"z"(foo));
}


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 16:39:13 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 16:39:13 -0500
Subject: [LLVMbugs] [Bug 2342] Can't assemble the output of llc
In-Reply-To: <bug-2342-206@http.llvm.org/bugs/>
Message-ID: <200805192139.m4JLdDFK023277@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2342


Dale Johannesen <dalej at apple.com> changed:

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




--- Comment #1 from Dale Johannesen <dalej at apple.com>  2008-05-19 16:39:13 ---
This should be fixed, at least for all currently supported Darwin targets.
http://llvm.org/viewvc/llvm-project?view=rev&revision=51277


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 17:13:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 17:13:03 -0500
Subject: [LLVMbugs] [Bug 2346] New: Crash in LICM: Assertion `LoadValue && "
 No store through the pointer found!"'
Message-ID: <bug-2346-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2346

           Summary: Crash in LICM: Assertion `LoadValue && "No store through
                    the pointer found!"'
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1657)
 --> (http://llvm.org/bugs/attachment.cgi?id=1657)
Testcase

Attached testcase crashes with "opt -licm".


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 17:22:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 17:22:29 -0500
Subject: [LLVMbugs] [Bug 2334] clang crashes codegen' s conditional of
 compatible but different function pointers
In-Reply-To: <bug-2334-206@http.llvm.org/bugs/>
Message-ID: <200805192222.m4JMMTH4024695@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2334


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-19 17:22:28 ---
Fixed in r51202.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 19 17:24:21 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 19 May 2008 17:24:21 -0500
Subject: [LLVMbugs] [Bug 2294] possible integer codegen bug in
	index-set-split
In-Reply-To: <bug-2294-206@http.llvm.org/bugs/>
Message-ID: <200805192224.m4JMOLNH024785@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2294


Devang Patel <dpatel at apple.com> changed:

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




--- Comment #2 from Devang Patel <dpatel at apple.com>  2008-05-19 17:24:20 ---
Fixed.

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062645.html


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 07:51:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 07:51:05 -0500
Subject: [LLVMbugs] [Bug 2333] clang codegen crash with function declared
 without a prototype then defined
In-Reply-To: <bug-2333-206@http.llvm.org/bugs/>
Message-ID: <200805201251.m4KCp5Z6010006@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2333


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-20 07:51:02 ---
"Fixed" by r51312.  (It's really a workaround, but oh well... fixing it to work
properly isn't a correctness bug.)


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 07:56:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 07:56:07 -0500
Subject: [LLVMbugs] [Bug 2332] clang incorrectly rejects "a(a)int a; {a=10;
	return a; }"
In-Reply-To: <bug-2332-206@http.llvm.org/bugs/>
Message-ID: <200805201256.m4KCu7Yr010159@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2332


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #3 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-20 07:56:07 ---
Fixed by r51311.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 07:59:36 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 07:59:36 -0500
Subject: [LLVMbugs] [Bug 2079] clang asset: initializer with more or less
 braces than default
In-Reply-To: <bug-2079-206@http.llvm.org/bugs/>
Message-ID: <200805201259.m4KCxab2010285@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2079


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-20 07:59:36 ---
Both tests now do the right thing; fixed by enabling SemaInit.

If anyone thinks we should be warning for either of these constructs, please
file a bug.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 08:08:52 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 08:08:52 -0500
Subject: [LLVMbugs] [Bug 2018] Crash with void compound literal
In-Reply-To: <bug-2018-206@http.llvm.org/bugs/>
Message-ID: <200805201308.m4KD8qBr010594@zion.cs.uiuc.edu>

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


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-20 08:08:51 ---
Fixed by adding some more checks to Sema.  (r51300 and r51301).


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 17:06:17 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 17:06:17 -0500
Subject: [LLVMbugs] [Bug 2346] Crash in LICM: Assertion `LoadValue && " No
 store through the pointer found!"'
In-Reply-To: <bug-2346-206@http.llvm.org/bugs/>
Message-ID: <200805202206.m4KM6HCv009781@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2346


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-20 17:06:17 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062706.html


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 17:42:56 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 17:42:56 -0500
Subject: [LLVMbugs] [Bug 2347] New: Clang segfault
Message-ID: <bug-2347-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2347

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


On this:

void f (unsigned int m)
{
  extern int e[2][m];

  e[0][0] = 0;
}


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 18:00:54 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 18:00:54 -0500
Subject: [LLVMbugs] [Bug 2348] New: Initializer bug
Message-ID: <bug-2348-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2348

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


The following is fine and clang rejects it:

static struct { int z; } s[2];
int *t = &(*s).z;


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 18:06:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 18:06:07 -0500
Subject: [LLVMbugs] [Bug 2349] New: Initializer bug/2
Message-ID: <bug-2349-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2349

           Summary: Initializer bug/2
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: NetBSD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: neil at daikokuya.co.uk
                CC: llvmbugs at cs.uiuc.edu


The below is invalid; clang accepts it.  It also avoid clang's
returning-address-of-local warning, which it perhaps should trigger.

short *a (void)
{
  short int b;
  static short *bp = &b;

  return bp;
}


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 22:39:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 22:39:28 -0500
Subject: [LLVMbugs] [Bug 2349] Initializer bug/2
In-Reply-To: <bug-2349-206@http.llvm.org/bugs/>
Message-ID: <200805210339.m4L3dSkX019522@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2349


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-20 22:39:27 ---
Fixed; thanks for the bug report!


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 20 22:42:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 20 May 2008 22:42:39 -0500
Subject: [LLVMbugs] [Bug 2348] Initializer bug
In-Reply-To: <bug-2348-206@http.llvm.org/bugs/>
Message-ID: <200805210342.m4L3gdAL019634@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2348


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-20 22:42:39 ---
Fixed; thanks for the bug report!


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 21 00:34:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 May 2008 00:34:19 -0500
Subject: [LLVMbugs] [Bug 2347] Clang segfault
In-Reply-To: <bug-2347-206@http.llvm.org/bugs/>
Message-ID: <200805210534.m4L5YJoC022449@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2347


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-21 00:34:19 ---
Fixed in r51363.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 21 07:09:43 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 May 2008 07:09:43 -0500
Subject: [LLVMbugs] [Bug 2350] New: Assertion failure
Message-ID: <bug-2350-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2350

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


Compiling

void f (const void);

assertion "NumParams == getNumParams() && "Parameter count mismatch!"" failed:
file "Decl.cpp", line 482, function "void
clang::FunctionDecl::setParams(clang::ParmVarDecl**, unsigned int)"
Abort trap


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 21 09:57:29 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 May 2008 09:57:29 -0500
Subject: [LLVMbugs] [Bug 2351] New: crash in llc with debug info
Message-ID: <bug-2351-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2351

           Summary: crash in llc with debug info
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1660)
 --> (http://llvm.org/bugs/attachment.cgi?id=1660)
Testcase

Per description, attached testcase crashes running llc over it.  (The testcase
was generated by clang, and reduced by bugpoint.)

I don't know if the attached bc file is valid; I haven't looked closely, and I
know next to nothing about the debugging info format.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 21 17:04:46 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 May 2008 17:04:46 -0500
Subject: [LLVMbugs] [Bug 2352] New: extern vla incorrectly accepted
Message-ID: <bug-2352-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2352

           Summary: extern vla incorrectly accepted
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Forked from Bug 2347

void f (unsigned int m)
{
  extern int e[2][m];

  e[0][0] = 0;
}

Should print a diagnostic; this violates the constraint in C99 6.7.5.2p2.

Related bug: Bug 2044.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 21 17:38:09 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 May 2008 17:38:09 -0500
Subject: [LLVMbugs] [Bug 2343] llvm-gcc bootstrap is broken
In-Reply-To: <bug-2343-206@http.llvm.org/bugs/>
Message-ID: <200805212238.m4LMc9K6030677@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2343


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

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




--- Comment #12 from Evan Cheng <evan.cheng at apple.com>  2008-05-21 17:38:08 ---
Looks like a pre-Evan coalescer bug. Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062735.html

Now the code snippet looks like this:
        movl    %edx, %edi
        movl    %ecx, %edx
        testl   %edx, %edx
        jne     .LBB1_47        # bb326
.LBB1_52:       # bb397
        testl   %eax, %eax
        je      .LBB1_80        # bb695
.LBB1_53:       # bb402
        movl    %esi, 4(%esp)
        movl    %eax, (%esp)
        call    get_cases_for_edge
        movl    (%eax), %esi
        movl    %esi, (%edi)

This correctly saves the previous iteration of "case" into "last" (%edi) and
used it as store address. Please verify this fixes Linux i386 llvm-gcc
bootstrapping.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 21 18:51:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 May 2008 18:51:08 -0500
Subject: [LLVMbugs] [Bug 2353] New: Disabling gcc inliner causes WebKit
 performance regression
Message-ID: <bug-2353-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2353

           Summary: Disabling gcc inliner causes WebKit performance
                    regression
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Interprocedural Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: evan.cheng at apple.com
                CC: llvmbugs at cs.uiuc.edu


Duncan's patch 51314 regressed WebKit performance by ~5%. I've reverted it for
now 51397.

This probably means there were llvm optimizer deficiencies that was covered up
by running the gcc inliner. We need to identify and fix them.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 21 23:47:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 21 May 2008 23:47:19 -0500
Subject: [LLVMbugs] [Bug 2267] llvm-gcc no longer accepts some assembly:
 Invalid or unsupported inline assembly!
In-Reply-To: <bug-2267-206@http.llvm.org/bugs/>
Message-ID: <200805220447.m4M4lJwC008620@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2267


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #8 from Chris Lattner <sabre at nondot.org>  2008-05-21 23:47:18 ---
Fixed, patch here: 
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062746.html
Testcase here: test/CodeGen/X86/asm-indirect-mem.ll


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 22 04:23:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 May 2008 04:23:32 -0500
Subject: [LLVMbugs] [Bug 2350] Assertion failure
In-Reply-To: <bug-2350-206@http.llvm.org/bugs/>
Message-ID: <200805220923.m4M9NWmK026192@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2350


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-22 04:23:31 ---
Fixed in r51424.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 22 06:50:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 May 2008 06:50:27 -0500
Subject: [LLVMbugs] [Bug 2336] llvm miscompiles tree-cfg.c
In-Reply-To: <bug-2336-206@http.llvm.org/bugs/>
Message-ID: <200805221150.m4MBoR0F030456@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2336


Rafael ??vila de Esp??ndola <rafael.espindola at gmail.com> changed:

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




--- Comment #2 from Rafael ??vila de Esp??ndola <rafael.espindola at gmail.com>  2008-05-22 06:50:26 ---
This has been fixed. I just recompiled the attached testcase for this bug and
relinked. I will try a full bootstrap on Friday.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 22 06:56:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 22 May 2008 06:56:48 -0500
Subject: [LLVMbugs] [Bug 2354] New: readdir broken with
 -D_FILE_OFFSET_BITS=64 ( asm on func discarded)
Message-ID: <bug-2354-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2354

           Summary: readdir broken with -D_FILE_OFFSET_BITS=64 (asm on func
                    discarded)
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: parser
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Per description; simple real testcase:
#include <stdio.h>
#include <dirent.h>
int main() {        
struct dirent *dp;
DIR *dirp = opendir(".");
dp = readdir(dirp);
puts(dp->d_name);
}

Should list the first file in the current dir. With "clang -emit-llvm
-D_FILE_OFFSET_BITS=64", the first 8 letters get cut off.  The issue is the
following declaration:

extern struct dirent *readdir (DIR *__dirp) __asm__ ("" "readdir64")
     __attribute__ ((__nonnull__ (1)));

clang's parser is lazy, so it decides to throw away the __asm__ bit.  (The
relevant code is in Parser::ParseInitDeclaratorListAfterFirstDeclarator.)  This
makes everything break, of course.

Bug found while trying to build ruby with clang.  And it took quite a while to
figure out what was going on...


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 05:27:50 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 05:27:50 -0500
Subject: [LLVMbugs] [Bug 2355] New: llvm-gcc fails to generate_muldi3.o
Message-ID: <bug-2355-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2355

           Summary: llvm-gcc fails to generate_muldi3.o
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-gcc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rafael.espindola at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1664)
 --> (http://llvm.org/bugs/attachment.cgi?id=1664)
testcase

The attached file is a reduced testcase:

./cc1 test.i
 __muldi3cc1: ../../llvm-gcc-4.2/gcc/llvm-convert.cpp:888: llvm::Value*
TreeToLLVM::Emit(tree_node*, const MemRef*): Assertion
`(isAggregateTreeType(((exp)->common.type)) == (DestLoc != 0) || ((enum
tree_code) (exp)->common.code) == MODIFY_EXPR) && "Didn't pass DestLoc to an
aggregate expr, or passed it to scalar!"' failed.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 05:34:02 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 05:34:02 -0500
Subject: [LLVMbugs] [Bug 2355] llvm-gcc fails to generate_muldi3.o
In-Reply-To: <bug-2355-206@http.llvm.org/bugs/>
Message-ID: <200805231034.m4NAY29w029517@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2355


Rafael ??vila de Esp??ndola <rafael.espindola at gmail.com> changed:

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




--- Comment #1 from Rafael ??vila de Esp??ndola <rafael.espindola at gmail.com>  2008-05-23 05:34:02 ---
Fixed by a "svn up" on the llvm-gcc directory. Sorry about that.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 19:06:21 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 19:06:21 -0500
Subject: [LLVMbugs] [Bug 2356] New: Crash in llc with asm (regression)
Message-ID: <bug-2356-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2356

           Summary: Crash in llc with asm (regression)
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Transformation Utilities
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: dalej at apple.com
                CC: llvmbugs at cs.uiuc.edu


Attached file crashes (just feed into llc).  Recent regression involving asm. 
Reduced from gcc.c-torture/execute/20030222-1.c.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 19:20:26 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 19:20:26 -0500
Subject: [LLVMbugs] [Bug 2351] crash in llc with debug info
In-Reply-To: <bug-2351-206@http.llvm.org/bugs/>
Message-ID: <200805240020.m4O0KQNn021307@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2351


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #6 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-23 19:20:25 ---
Okay, fine; the only interesting thing here is the enhancement request for a
debug info validator.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 19:25:23 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 19:25:23 -0500
Subject: [LLVMbugs] [Bug 2357] New: declaration error,
	from self included file
Message-ID: <bug-2357-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2357

           Summary: declaration error, from self included file
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: preprocessor
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: csaba.hruska at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Hi!
Here is my new bugreport :)
gcc accepts it. Its from xvid library, but it is heavily stripped down to be
simple and clean.
it throws error with clang svn revision 51520.

there is a file attched: qpel.c
It includes itself, and there are a nasty define and ifdef hack in it.

PS:
It's not my code :) its from blender modeler program.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 20:42:26 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 20:42:26 -0500
Subject: [LLVMbugs] [Bug 2358] New: SCCP incorrectly folds cttz(undef) to
	undef
Message-ID: <bug-2358-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2358

           Summary: SCCP incorrectly folds cttz(undef) to undef
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:

target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128"
target triple = "i686-pc-linux-gnu"

define i32 @x(i32 %b) {
entry:
  %val = call i32 @llvm.cttz.i32(i32 undef)
  ret i32 %val
}

declare i32 @llvm.cttz.i32(i32)

SCCP folds x to ret i32 undef; this is wrong because all of the bits except for
the bottom 6 must be zero.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 21:16:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 21:16:47 -0500
Subject: [LLVMbugs] [Bug 2359] New: wrong instcombine float comparison
	simplification
Message-ID: <bug-2359-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2359

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


The fcmp is folded to "true" by instcombine.  Reduced by bugpoint
from ACATS c34002a, which recently started failing.

define i1 @f(i8* %x) {
entry:
        %tmp462 = load i8* %x, align 1          ; <i8> [#uses=1]
        %tmp462463 = sitofp i8 %tmp462 to float         ; <float> [#uses=1]
        %tmp464 = fcmp ugt float %tmp462463, 0x47EFFFFFE0000000         ; <i1>
[#uses=1]
        ret i1 %tmp464
}


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 23:00:03 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 23:00:03 -0500
Subject: [LLVMbugs] [Bug 2358] SCCP incorrectly folds cttz(undef) to undef
In-Reply-To: <bug-2358-206@http.llvm.org/bugs/>
Message-ID: <200805240400.m4O403Jn019727@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2358


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-23 23:00:03 ---
Fixed, nice catch!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062893.html


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 23 23:07:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 23 May 2008 23:07:01 -0500
Subject: [LLVMbugs] [Bug 2359] wrong instcombine float comparison
	simplification
In-Reply-To: <bug-2359-206@http.llvm.org/bugs/>
Message-ID: <200805240407.m4O471hk019997@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2359


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-23 23:07:00 ---
Fixed.  ACATS rocks.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062894.html

-Chris


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 24 00:12:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 May 2008 00:12:19 -0500
Subject: [LLVMbugs] [Bug 2331] a patch for win32 build when there is no
	bison present
In-Reply-To: <bug-2331-206@http.llvm.org/bugs/>
Message-ID: <200805240512.m4O5CJek022413@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2331


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-24 00:12:18 ---
Applied this patch, please verify that it works for you:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062897.html


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 24 00:19:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 May 2008 00:19:12 -0500
Subject: [LLVMbugs] [Bug 2248] Clang cannot compile its preprocessed output
 with same flags
In-Reply-To: <bug-2248-206@http.llvm.org/bugs/>
Message-ID: <200805240519.m4O5JChA022658@zion.cs.uiuc.edu>

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


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-24 00:19:11 ---
I don't see this.  Maybe it was already fixed?


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 24 01:12:44 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 May 2008 01:12:44 -0500
Subject: [LLVMbugs] [Bug 2248] Clang cannot compile its preprocessed output
 with same flags
In-Reply-To: <bug-2248-206@http.llvm.org/bugs/>
Message-ID: <200805240612.m4O6Ciho024321@zion.cs.uiuc.edu>

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


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

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




--- Comment #2 from Neil Booth <neil at daikokuya.co.uk>  2008-05-24 01:12:43 ---
Nope, still broken.  Here's the cut-down preprocessed file.

# 1 "/tmp/bug.c" 1
typedef __builtin_va_list va_list;
void d (int tag, ...)
{
  va_list ap;

  __builtin_stdarg_start((ap), (tag));
  __builtin_va_end (ap);
}

$ clang -pedantic-errors /tmp/bug.i
/tmp/bug.i:7:3: error: implicit declaration of function
'__builtin_stdarg_start' is invalid in C99
  __builtin_stdarg_start((ap), (tag));
  ^
1 diagnostic generated.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 24 07:37:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 May 2008 07:37:34 -0500
Subject: [LLVMbugs] [Bug 2360] New: typedefed function prototype bug
Message-ID: <bug-2360-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2360

           Summary: typedefed function prototype bug
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: csaba.hruska at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1667)
 --> (http://llvm.org/bugs/attachment.cgi?id=1667)
buggy source file

Clang segfaults for this file.

if the b is the firts in the declaration then it compiles.
it also compiles with gcc.
i used clang svn revision 51543.

code:

typedef void fn_t();

fn_t a,b;

void b()
{
}


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 24 08:47:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 May 2008 08:47:01 -0500
Subject: [LLVMbugs] [Bug 2361] New: Undiagnosed file-scope VLA
Message-ID: <bug-2361-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2361

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


Clang happily accepts

int i; int c[][i];


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 24 08:47:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 May 2008 08:47:58 -0500
Subject: [LLVMbugs] [Bug 2362] New: Duplicate diagnostic
Message-ID: <bug-2362-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2362

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


int i; int c[][i] = { 0 };

$ clang /tmp/bug.c
/tmp/bug.c:1:16: error: variable-sized object may not be initialized
int i; int c[][i] = { 0 };
               ^
/tmp/bug.c:1:16: error: variable-sized object may not be initialized
int i; int c[][i] = { 0 };
               ^
2 diagnostics generated.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 24 13:12:17 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 24 May 2008 13:12:17 -0500
Subject: [LLVMbugs] [Bug 2363] New: DenseMap won't work with const key,
	integer value
Message-ID: <bug-2363-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2363

           Summary: DenseMap won't work with const key, integer value
           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: viridia at gmail.com
                CC: llvmbugs at cs.uiuc.edu


DenseMap<const Ptr *, int> doesn't work for two reasons:

1) On line 221, the placement new fails because the argument is const.
2) DenseMap<int>::isPod is undefined.

For the former, I'm not sure what to do. For the latter, well, you can't define
DenseMapInfo<int> because there are no available integer sentinel values for
getEmptyKey() - you can't assume -1 is an invalid key. Probably the correct
thing to do is to define a separate type traits class for just those traits
needed by values (not keys). (Or define a whole type traits system, which might
not be a bad idea, it would clean up the current container implementations
quite a bit.)


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 08:22:55 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 08:22:55 -0500
Subject: [LLVMbugs] [Bug 2362] Duplicate diagnostic
In-Reply-To: <bug-2362-206@http.llvm.org/bugs/>
Message-ID: <200805251322.m4PDMtmn022593@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2362


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-25 08:22:54 ---
Fixed in r51555.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 09:04:31 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 09:04:31 -0500
Subject: [LLVMbugs] [Bug 2151] clang -fsyntax-only runs forever on provided
	input
In-Reply-To: <bug-2151-206@http.llvm.org/bugs/>
Message-ID: <200805251404.m4PE4V1R023709@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2151


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #7 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-25 09:04:30 ---
Fixed in r51556.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 09:25:26 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 09:25:26 -0500
Subject: [LLVMbugs] [Bug 1961] Test failures with clang objc tests
In-Reply-To: <bug-1961-206@http.llvm.org/bugs/>
Message-ID: <200805251425.m4PEPQhs024356@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1961


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-25 09:25:26 ---
This seems to have gotten fixed at some point. (Although these tests aren't run
without manually enabling them.)


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 14:06:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 14:06:48 -0500
Subject: [LLVMbugs] [Bug 2364] New: likely integer codegen bug
Message-ID: <bug-2364-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2364

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


Using svn 51558 on Ubuntu Feisty on x86-32 this code is miscompiled:

int func_6 (void)
{
  int i = 0;
  char x;
  for (x = 0;
       (x != 12);
       x -= 4) {
    i++;
  }
  return i;
}

llvm-gcc -fwrapv -O1 -S -o - small.c -Wall -fomit-frame-pointer

gives:

func_6:
        movl    $253, %eax
        ret

I believe the correct result is 61.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 15:56:46 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 15:56:46 -0500
Subject: [LLVMbugs] [Bug 2329] Comparison of constant inttoptrs not folded
In-Reply-To: <bug-2329-206@http.llvm.org/bugs/>
Message-ID: <200805252056.m4PKuksq002447@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2329


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #6 from Nick Lewycky <nicholas at mxc.ca>  2008-05-25 15:56:45 ---
Patch checked in.

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062906.html


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 16:53:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 16:53:58 -0500
Subject: [LLVMbugs] [Bug 2033] corrupted CallGraph
In-Reply-To: <bug-2033-206@http.llvm.org/bugs/>
Message-ID: <200805252153.m4PLrw7x003963@zion.cs.uiuc.edu>

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


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #1 from Nick Lewycky <nicholas at mxc.ca>  2008-05-25 16:53:58 ---
I'm no longer able to reproduce this bug.


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

From neil at daikokuya.co.uk  Sun May 25 17:21:46 2008
From: neil at daikokuya.co.uk (Neil Booth)
Date: Mon, 26 May 2008 07:21:46 +0900
Subject: [LLVMbugs] [Bug 2364] New: likely integer codegen bug
In-Reply-To: <bug-2364-206@http.llvm.org/bugs/>
References: <bug-2364-206@http.llvm.org/bugs/>
Message-ID: <20080525222146.GA10026@daikokuya.co.uk>

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

> http://llvm.org/bugs/show_bug.cgi?id=2364
> 
>            Summary: likely integer codegen bug
>            Product: new-bugs
>            Version: unspecified
>           Platform: PC
>         OS/Version: Windows XP
>             Status: NEW
>           Severity: normal
>           Priority: P2
>          Component: new bugs
>         AssignedTo: unassignedbugs at nondot.org
>         ReportedBy: regehr at cs.utah.edu
>                 CC: llvmbugs at cs.uiuc.edu
> 
> 
> Using svn 51558 on Ubuntu Feisty on x86-32 this code is miscompiled:
> 
> int func_6 (void)
> {
>   int i = 0;
>   char x;
>   for (x = 0;
>        (x != 12);
>        x -= 4) {
>     i++;
>   }
>   return i;
> }
> 
> llvm-gcc -fwrapv -O1 -S -o - small.c -Wall -fomit-frame-pointer
> 
> gives:
> 
> func_6:
>         movl    $253, %eax
>         ret
> 
> I believe the correct result is 61.

This looks like undefined behaviour (integer overflow).

Neil.

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 18:44:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 18:44:07 -0500
Subject: [LLVMbugs] [Bug 2364] signedness error in scalar-evolution
	HowFarToZero
In-Reply-To: <bug-2364-206@http.llvm.org/bugs/>
Message-ID: <200805252344.m4PNi7IB007282@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2364


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #2 from Nick Lewycky <nicholas at mxc.ca>  2008-05-25 18:44:06 ---
Fixed. Patch here:

http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080519/062907.html


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 19:22:52 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 19:22:52 -0500
Subject: [LLVMbugs] [Bug 2365] New: Tutorial examples broken in 2.2
Message-ID: <bug-2365-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2365

           Summary: Tutorial examples broken in 2.2
           Product: new-bugs
           Version: unspecified
          Platform: All
        OS/Version: All
            Status: NEW
          Keywords: documentation
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: markleone at gmail.com
                CC: llvmbugs at cs.uiuc.edu


A couple of changes in 2.2 appear to have broken the examples in the tutorial
(llvm.org/docs/tutorial).  E.g. in the first example IRBuilder must be replaced
with LLVMBuilder and BasicBlock::create() must be replaced with new
BasicBlock().

(Apologies if documentation issues are not appropriate for bugzilla.)

Mark Leone


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 21:16:19 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 21:16:19 -0500
Subject: [LLVMbugs] [Bug 2365] Tutorial examples broken in 2.2
In-Reply-To: <bug-2365-206@http.llvm.org/bugs/>
Message-ID: <200805260216.m4Q2GJ9l011369@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2365


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |WONTFIX




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-25 21:16:19 ---
The tutorial isn't outdated; it's actually too new.  Try running the examples
against a trunk or LLVM 2.3 (which is currently in the middle of prerelease
testing, and is scheduled to be released June 4).


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 22:34:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 22:34:05 -0500
Subject: [LLVMbugs] [Bug 2366] New: clamav miscompiles
Message-ID: <bug-2366-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2366

           Summary: clamav miscompiles
           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: nicholas at mxc.ca
                CC: llvmbugs at cs.uiuc.edu


$ ./clamscan.llc --debug --verbose -d${PROJ_SRC_DIR}/dbdir -r
${PROJ_SRC_DIR}/inputs/*

  [...]

  Initialising AC pattern matcher of root[%d]

  cli_initroots: Initializing BM tables of root[%d]

  %s loaded

  Segmentation fault (core dumped)

versus CBE: [...]

  Initialising AC pattern matcher of root[%d]

  cli_initroots: Initializing BM tables of root[%d]

  %s loaded

  cli_loadmd5: Initializing MD5 list structure

  [etc, no crash]

The core is not entirely useful:
(gdb) bt
#0  0xf7dde6a5 in readdir () from /lib/i686/cmov/libc.so.6
#1  0x080b4d4c in cli_loaddbdir ()
#2  0x080b3434 in cli_load ()
#3  0xffb41083 in ?? ()
#4  0x00000000 in ?? ()

Further, bugpoint is useless. It reduces down to br/br/ret void.


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

From bugzilla-daemon at cs.uiuc.edu  Sun May 25 22:46:13 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sun, 25 May 2008 22:46:13 -0500
Subject: [LLVMbugs] [Bug 2367] New: MultiSource/Applications/d/make_parser
	miscompiles
Message-ID: <bug-2367-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2367

           Summary: MultiSource/Applications/d/make_parser miscompiles
           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: nicholas at mxc.ca
                CC: llvmbugs at cs.uiuc.edu


$ ./make_dparser.llc -v ${PROJ_SRC_DIR}/grammar.g
Segmentation fault

make_dparser.cbe works. Here's the stacktrace:
(gdb) bt
#0  0x080637c3 in exhaustive_parse ()
#1  0x08057b04 in main ()

Bugpoint is useless here, producing br/br/ret void again.

That said, bugpoint-opt and bugpoint-llvm-ld can't find any problem, so it's
probably in the backend.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 02:07:27 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 02:07:27 -0500
Subject: [LLVMbugs] [Bug 2368] New: lcssa passes take 50% of the time in
	large testcase
Message-ID: <bug-2368-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2368

           Summary: lcssa passes take 50% of the time in large testcase
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: resistor at mac.com, llvmbugs at cs.uiuc.edu


Original testcase at http://gcc.gnu.org/bugzilla/attachment.cgi?id=14134, from
gcc bug at http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33237.

Running opt -std-compile-opts over the attachment results in two lcssa passes
that take up half the time spent in opt.  (Report below.)

===-------------------------------------------------------------------------===
                      ... Pass execution timing report ...
===-------------------------------------------------------------------------===
  Total Execution Time: 50.9271 seconds (51.2564 wall clock)

   ---User Time---   --System Time--   --User+System--   ---Wall Time---  ---
Name ---
  14.8209 ( 29.2%)   0.2360 ( 77.6%)  15.0569 ( 29.5%)  15.0543 ( 29.3%) 
Global Value Numbering
  12.6607 ( 25.0%)   0.0000 (  0.0%)  12.6607 ( 24.8%)  12.6599 ( 24.6%) 
Loop-Closed SSA Form Pass
  11.4527 ( 22.6%)   0.0000 (  0.0%)  11.4527 ( 22.4%)  11.4535 ( 22.3%) 
Loop-Closed SSA Form Pass
   1.5200 (  3.0%)   0.0000 (  0.0%)   1.5200 (  2.9%)   1.5184 (  2.9%) 
Simplify the CFG
   1.4920 (  2.9%)   0.0000 (  0.0%)   1.4920 (  2.9%)   1.4928 (  2.9%) 
Simplify the CFG
   1.2200 (  2.4%)   0.0000 (  0.0%)   1.2200 (  2.3%)   1.2207 (  2.3%)  Loop
Invariant Code Motion
   1.1120 (  2.1%)   0.0000 (  0.0%)   1.1120 (  2.1%)   1.1147 (  2.1%) 
Simplify the CFG
[snip]


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 04:20:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 04:20:07 -0500
Subject: [LLVMbugs] [Bug 1682] make test failures on Solaris (SPARC)
In-Reply-To: <bug-1682-206@http.llvm.org/bugs/>
Message-ID: <200805260920.m4Q9K7up031819@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1682


Gabor Greif <ggreif at gmail.com> changed:

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




--- Comment #15 from Gabor Greif <ggreif at gmail.com>  2008-05-26 04:20:05 ---
Nice work Eli!

make report is clean now :-)


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 05:31:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 05:31:16 -0500
Subject: [LLVMbugs] [Bug 2368] lcssa passes take 50% of the time in large
	testcase
In-Reply-To: <bug-2368-206@http.llvm.org/bugs/>
Message-ID: <200805261031.m4QAVGjN001238@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2368


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #4 from Owen Anderson <resistor at mac.com>  2008-05-26 05:31:15 ---
I'm closing this bug as fixed.  Feel free to file others about other parts of
LLVM that are slow on this case.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 05:35:28 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 05:35:28 -0500
Subject: [LLVMbugs] [Bug 2004] missed instcombine with trunc+sext
In-Reply-To: <bug-2004-206@http.llvm.org/bugs/>
Message-ID: <200805261035.m4QAZSVT001360@zion.cs.uiuc.edu>

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


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-26 05:35:25 ---
Fixed by the recent signbit enhancements to instcombine.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 09:33:54 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 09:33:54 -0500
Subject: [LLVMbugs] [Bug 2369] New: scalarrepl overly aggressive with large
	arrays
Message-ID: <bug-2369-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2369

           Summary: scalarrepl overly aggressive with large arrays
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
#include <string.h>
void memtest1 (char* dst, char* src)
{
  char temp[100];
  memcpy (&temp, src, 100);
  memcpy (dst, &temp, 100); 
}

The generated assembly for X86 is about 400 instructions, which is clearly
ridiculous.  The issue is that scalarrepl is turning this into 100 loads, all
of which get spilled to the stack, then 100 stores, all of which have to be
reloaded from the stack.

On a side note, it might be worth optimizing code like this to memmove.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 09:57:25 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 09:57:25 -0500
Subject: [LLVMbugs] [Bug 2370] New: memmove(a,a,n) should be optimized out
Message-ID: <bug-2370-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2370

           Summary: memmove(a,a,n) should be optimized out
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Per description; testcase:

target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128"
target triple = "i686-pc-linux-gnu"

define void @x(i8* %a) {
entry:
        tail call void @llvm.memmove.i32( i8* %a, i8* %a, i32 100, i32 1 )
        ret void
}

This function does nothing. Pretty simple optimization, although likely not
useful very often.  Note that this transformation can be done slightly more
powerfully if alias analysis is available.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 10:04:49 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 10:04:49 -0500
Subject: [LLVMbugs] [Bug 2371] New: codegen should inline small memmoves
Message-ID: <bug-2371-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2371

           Summary: codegen should inline small memmoves
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
target triple = "i686-pc-linux-gnu"
        %struct.anon = type { i32, i32, i32 }

define void @a(%struct.anon* %a, %struct.anon* %b) nounwind  {
entry:
        %tmp2 = bitcast %struct.anon* %a to i8*         ; <i8*> [#uses=1]
        %tmp3 = bitcast %struct.anon* %b to i8*         ; <i8*> [#uses=1]
        tail call void @llvm.memmove.i32( i8* %tmp2, i8* %tmp3, i32 12, i32 4 )
        ret void
}

Generated assembly:
a:
        subl    $12, %esp
        movl    20(%esp), %ecx
        movl    %ecx, 4(%esp)
        movl    16(%esp), %ecx
        movl    %ecx, (%esp)
        movl    $12, 8(%esp)
        call    memmove
        addl    $12, %esp
        ret

For a memmove this short, it should just load the values from the first
pointer, then store them into the other pointer.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 13:49:41 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 13:49:41 -0500
Subject: [LLVMbugs] [Bug 2372] New: ice
Message-ID: <bug-2372-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2372

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


This is on svn 51560 for ia32 on Ubuntu Feisty.

[regehr at babel tmp10]$ llvm-gcc -O1 small.c
cc1: LiveVariables.cpp:143: void llvm::LiveVariables::HandleVirtRegUse(unsigned
int, llvm::MachineBasicBlock*, llvm::MachineInstr*): Assertion `MBB !=
MRI->getV
RegDef(reg)->getParent() && "Should have kill for defblock!"' failed.
small.c:30: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://developer.apple.com/bugreporter> for instructions.

[regehr at babel tmp10]$ cat small.c

typedef short int int16_t;
typedef int int32_t;
typedef unsigned char uint8_t;
typedef unsigned int uint32_t;
static inline int
lshift_s_s (int left, int right)
{
  if ((left < 0) || (right) || (right >= sizeof (int))
      || (left > (1 >> right)))
    return left;
  return left << right;
}

int32_t
func_3 (int16_t p_5, int16_t p_6)
{
  uint8_t l_127;
  for (1; l_127; --l_127)
    func_8 (((p_5
              && (func_8 ((func_8 (func_18 (p_5, p_5, 1), 1)), 1) *
                  (lshift_s_s (0x8CCBL, p_6))))), 1);
  ((rshift_s_s
    (func_8 (p_5, (func_18 (0xDA599ACBL, (func_18 (1, p_6, 1)), 1))), 1)));
}

int32_t
func_8 (uint32_t p_9, uint32_t p_10)
{
  return 1;
}

[regehr at babel tmp10]$


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 21:24:33 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 21:24:33 -0500
Subject: [LLVMbugs] [Bug 2373] New: Type size mismatch assertion during
 llvm-gcc compile on __powixf2
Message-ID: <bug-2373-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2373

           Summary: Type size mismatch assertion during llvm-gcc compile on
                    __powixf2
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Backend: X86
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: chandlerc at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1669)
 --> (http://llvm.org/bugs/attachment.cgi?id=1669)
Reduced testcase

This is specific to x86-64 and Linux AFAIK:

/home/chandlerc/code/compilers/build/llvm-gcc/./gcc/cc1 -fpreprocessed
testcase.i -march=k8 -o /dev/null
 foocc1: /home/chandlerc/code/compilers/llvm-gcc/gcc/llvm-types.cpp:84: const
llvm::Type* llvm_set_type(tree_node*, const llvm::Type*): Assertion
`(!(__extension__ ({ const tree __t = (Tr); if (tree_code_type[(int) (((enum
tree_code) (__t)->common.code))] != (tcc_type)) tree_class_check_failed (__t,
(tcc_type), "/home/chandlerc/code/compilers/llvm-gcc/gcc/llvm-types.cpp", 84,
__FUNCTION__); __t; })->type.size) || !Ty->isSized() || !isInt64((__extension__
({ const tree __t = (Tr); if (tree_code_type[(int) (((enum tree_code)
(__t)->common.code))] != (tcc_type)) tree_class_check_failed (__t, (tcc_type),
"/home/chandlerc/code/compilers/llvm-gcc/gcc/llvm-types.cpp", 84,
__FUNCTION__); __t; })->type.size), true) || getInt64((__extension__ ({ const
tree __t = (Tr); if (tree_code_type[(int) (((enum tree_code)
(__t)->common.code))] != (tcc_type)) tree_class_check_failed (__t, (tcc_type),
"/home/chandlerc/code/compilers/llvm-gcc/gcc/llvm-types.cpp", 84,
__FUNCTION__); __t; })->type.size), true) ==
getTargetData().getABITypeSizeInBits(Ty)) && "LLVM type size doesn't match GCC
type size!"' failed.

testcase.i: In function 'foo':
testcase.i:2: internal compiler error: Aborted

Duncan diagnosed this correctly, and I'm filing and attaching patches here.

Also, this originally showed up compiling libgcc2.c to __powixf2 in case
someone searches with those terms. The above and attached are reduced test
cases.


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

From bugzilla-daemon at cs.uiuc.edu  Mon May 26 22:34:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Mon, 26 May 2008 22:34:30 -0500
Subject: [LLVMbugs] [Bug 2204] clang: sizeof(int16_t) != 2
In-Reply-To: <bug-2204-206@http.llvm.org/bugs/>
Message-ID: <200805270334.m4R3YU9B029811@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2204


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-26 22:34:30 ---
Fixed in r51577.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 27 00:11:35 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 May 2008 00:11:35 -0500
Subject: [LLVMbugs] [Bug 2360] typedefed function prototype bug
In-Reply-To: <bug-2360-206@http.llvm.org/bugs/>
Message-ID: <200805270511.m4R5BZkZ032534@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2360


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-27 00:11:35 ---
Fixed in r51584.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 27 00:17:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 May 2008 00:17:32 -0500
Subject: [LLVMbugs] [Bug 2192] Struct initializer initlistexpr wrong from
	sema
In-Reply-To: <bug-2192-206@http.llvm.org/bugs/>
Message-ID: <200805270517.m4R5HW7W032694@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2192


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-27 00:17:32 ---
Fixed by the recent initializer changes.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 27 00:20:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 May 2008 00:20:32 -0500
Subject: [LLVMbugs] [Bug 1986] Incorrect type generated for anonymous struct
In-Reply-To: <bug-1986-206@http.llvm.org/bugs/>
Message-ID: <200805270520.m4R5KWDZ000348@zion.cs.uiuc.edu>

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


Eli Friedman <sharparrow1 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sharparrow1 at yahoo.com
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-27 00:20:32 ---


*** This bug has been marked as a duplicate of bug 1954 ***


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 27 00:27:04 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 May 2008 00:27:04 -0500
Subject: [LLVMbugs] [Bug 2374] New: sema calculates incorrect type for
 statement expression with label
Message-ID: <bug-2374-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2374

           Summary: sema calculates incorrect type for statement expression
                    with label
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Semantic Analyzer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Split off from PR2325.

Testcase:
int a() { return ({L:5;}); }

This should be equivalent to "return 5;", but clang gives an error.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 27 04:48:42 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 May 2008 04:48:42 -0500
Subject: [LLVMbugs] [Bug 2375] New: llc crashes with initializer with
	bitcast of struct
Message-ID: <bug-2375-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2375

           Summary: llc crashes with initializer with bitcast of struct
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1671)
 --> (http://llvm.org/bugs/attachment.cgi?id=1671)
Patch

Bug per description; testcase:

; ModuleID = '-'
target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128"
target triple = "i686-pc-linux-gnu"
        %union.anon = type [8 x i8]
@x = global %union.anon bitcast ({ i64 } { i64 4 } to %union.anon)

llc currently crashes trying to emit this initializer:
void llvm::AsmPrinter::printDataDirective(const llvm::Type*): Assertion `0 &&
"Can't handle printing this type of thing"' failed.

Fortunately, this is really easy to fix; patch attached.


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

From bugzilla-daemon at cs.uiuc.edu  Tue May 27 14:36:16 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Tue, 27 May 2008 14:36:16 -0500
Subject: [LLVMbugs] [Bug 2376] New: std:: vector subscript out of range
 error in LocalSpiller when rewriting function
Message-ID: <bug-2376-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2376

           Summary: std::vector subscript out of range error in LocalSpiller
                    when rewriting function
           Product: tools
           Version: trunk
          Platform: PC
        OS/Version: other
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llc
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: spudder at blueyonder.co.uk
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1672)
 --> (http://llvm.org/bugs/attachment.cgi?id=1672)
Bugpoint reduced testcase 

Hi,

Compiling the attached testcase with llc on Windows Vista - SVN revision 51603,
built using VS 2005 - triggers a debug assertion from inside the MSVC++
standard library when attempting to access an element of a vector that doesn't
exist.

Running llc through the debugger, the code triggering this error is in
VirtRegMap.cpp:InvalidateKills

[snip]

if (KillOps[Reg] == &MO) {
  RegKills.reset(Reg);
  KillOps[Reg] = NULL;
}

[/snip]

Here, the value of "Reg" is 1061, but the size of KillOps is only 112.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 28 00:31:11 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 May 2008 00:31:11 -0500
Subject: [LLVMbugs] [Bug 2370] memmove(a,a,n) should be optimized out
In-Reply-To: <bug-2370-206@http.llvm.org/bugs/>
Message-ID: <200805280531.m4S5VBwo019020@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2370


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-28 00:31:11 ---
Done:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/062970.html


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 28 01:10:48 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 May 2008 01:10:48 -0500
Subject: [LLVMbugs] [Bug 2215] ClamAV test reads files from .svn directory
In-Reply-To: <bug-2215-206@http.llvm.org/bugs/>
Message-ID: <200805280610.m4S6AmNT020286@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2215


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-28 01:10:48 ---
Fixed:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/062972.html


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 28 01:13:25 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 May 2008 01:13:25 -0500
Subject: [LLVMbugs] [Bug 2367] MultiSource/Applications/d/make_parser
	miscompiles
In-Reply-To: <bug-2367-206@http.llvm.org/bugs/>
Message-ID: <200805280613.m4S6DPZ5020364@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2367


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #2 from Nick Lewycky <nicholas at mxc.ca>  2008-05-28 01:13:24 ---
Hunh. And now I can't reproduce it.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 28 12:40:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 May 2008 12:40:39 -0500
Subject: [LLVMbugs] [Bug 2289] llc assertion failure: Multiple implicit_def
 defining same register?
In-Reply-To: <bug-2289-206@http.llvm.org/bugs/>
Message-ID: <200805281740.m4SHedJ5017541@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2289


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

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




--- Comment #1 from Evan Cheng <evan.cheng at apple.com>  2008-05-28 12:40:23 ---
Fixed.
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/062988.html


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 28 14:58:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 May 2008 14:58:34 -0500
Subject: [LLVMbugs] [Bug 2377] New: SCEV should be a loop pass
Message-ID: <bug-2377-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2377

           Summary: SCEV should be a loop pass
           Product: libraries
           Version: trunk
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Loop Optimizer
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: resistor at mac.com
                CC: llvmbugs at cs.uiuc.edu


Because ScalarEvolution is a FunctionPass, it kills the LoopPassManager when it
is run.  This creates overhead because it require the re-instantiation of the
LPM, as well as forcing LCSSA to be recomputed repeatedly.  As of right now, we
compute LCSSA no less than five times in -std-compile-opts, mostly because of
this.  If SCEV were a loop pass, this could probably be reduced to one or two.


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 28 17:28:20 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 May 2008 17:28:20 -0500
Subject: [LLVMbugs] [Bug 1538] llvm2cpp should be made a target of llc
In-Reply-To: <bug-1538-206@http.llvm.org/bugs/>
Message-ID: <200805282228.m4SMSKTK027122@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1538


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

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




--- Comment #6 from Anton Korobeynikov <asl at math.spbu.ru>  2008-05-28 17:28:19 ---
Yep!


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

From bugzilla-daemon at cs.uiuc.edu  Wed May 28 22:56:07 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Wed, 28 May 2008 22:56:07 -0500
Subject: [LLVMbugs] [Bug 2378] New: likely volatile bug
Message-ID: <bug-2378-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2378

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


This behavior is observed for svn 51560 for x86-32 on Ubuntu Feisty.

At -O0, -Os, and -O3 the attached program accesses volatile g_31 zero times. 
At -O1 and -O2 it reads from g_31 three times.

The full command line I'm using is

  llvm-gcc -fwrapv -O1 -g small.c

The problem is: I can't reduce this input any further without making the
behavior go away.  Since you have different (and hopefully better) testcase
reduction infrastructure, perhaps you'll have better luck.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 01:38:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 01:38:32 -0500
Subject: [LLVMbugs] [Bug 2379] New: va_arg broken with x86_fp80
Message-ID: <bug-2379-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2379

           Summary: va_arg broken with x86_fp80
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:

; ModuleID = '-'
target datalayout =
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
target triple = "i686-pc-linux-gnu"

define i32 @a(i32 %i, ...) {
entry:
        %l = alloca i8*
        %l1 = bitcast i8** %l to i8*
        call void @llvm.va_start( i8* %l1 )
        %xxx = va_arg i8** %l, x86_fp80
        %ret = va_arg i8** %l, i32
        call void @llvm.va_end( i8* %l1 )
        ret i32 %ret
}

declare void @llvm.va_start(i8*) nounwind 
declare void @llvm.va_end(i8*) nounwind 

define i32 @main() {
entry:
        %call = call i32 (i32, ...)* @a( i32 1, x86_fp80 1.0, i32 195948557 )
        %cmp = icmp ne i32 %call, 195948557
        br i1 %cmp, label %ifthen, label %ifend

ifthen:         ; preds = %entry
        call void @abort( )
        br label %ifend

ifend:          ; preds = %ifthen, %entry
        ret i32 0
}

declare void @abort()


Run through "llvm-as | lli", should not call abort(), but it does call abort()
on x86 linux.  And if you replace the references to x86_fp80 with references to
double, it does in fact work.

It appears that the implementation of va_arg is ignoring the alignment of
x86_fp80.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 02:21:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 02:21:47 -0500
Subject: [LLVMbugs] [Bug 2380] New: crash emitting debug info in llvm-gcc
	front-end
Message-ID: <bug-2380-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2380

           Summary: crash emitting debug info in llvm-gcc front-end
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1676)
 --> (http://llvm.org/bugs/attachment.cgi?id=1676)
Testcase

Testcase attached; run "llvm-gcc -g llvmgcccrash.i" to crash.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 03:40:17 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 03:40:17 -0500
Subject: [LLVMbugs] [Bug 2381] New: Nethack doesn't compile!!!
Message-ID: <bug-2381-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2381

           Summary: Nethack doesn't compile!!!
           Product: clang
           Version: unspecified
          Platform: Macintosh
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: LLVM Codegen
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: isanbard at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Nethack doesn't compile because of an unimplemented codegen feature:

$ cat testcase.i
struct bork {
  unsigned field : 1;
};
struct bork foo[] = {
  { 0 }
};
$ clang -x c -o foo.o -emit-llvm-bc testcase.i
testcase.i:5:5: warning: cannot codegen this bitfield initialization yet
  { 0 }
    ^
1 diagnostic generated.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 03:52:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 03:52:05 -0500
Subject: [LLVMbugs] [Bug 2382] New: llvm-ld drops attributes off defnitions
	of functions
Message-ID: <bug-2382-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2382

           Summary: llvm-ld drops attributes off defnitions of functions
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Essentially the issue is that when llvm-ld merges a function decl without
attributes with a function decl with attributes, it loses the attributes.

I'll attach a testcase in a bit; it's two files, and the test is to run
"llvm-ld first.bc second.bc", then check the definition of check0 to make sure
the attributes are preserved.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 04:52:47 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 04:52:47 -0500
Subject: [LLVMbugs] [Bug 2383] New: wrong code emitted for empty struct
	byval parameter
Message-ID: <bug-2383-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2383

           Summary: wrong code emitted for empty struct byval parameter
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


define i32 @a({}* byval %x, i32 %y) nounwind  {
entry:
        ret i32 %y
}

emits the following:
a:
        movl    8(%esp), %eax
        ret

whereas the equivalent C code compiled with gcc or llvm-gcc:
struct S94 { struct __attribute__((aligned (16))){}a;};
int a(struct S94 x, int y) {return y;}

emits the following:

a:
        movl    4(%esp), %eax
        ret

I think the LLVM assembly should compile down to the same thing; I'm not
completely sure, though.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 07:06:14 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 07:06:14 -0500
Subject: [LLVMbugs] [Bug 2384] New: Support for -w (suppress warnings)
 missing from clang
Message-ID: <bug-2384-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2384

           Summary: Support for -w (suppress warnings) missing from clang
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Basic
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Bug per summary; sometimes it can be useful to suppress all warnings. For
example, some gcc tests consider any diagnostic output a failure.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 10:37:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 10:37:05 -0500
Subject: [LLVMbugs] [Bug 2384] Support for -w (suppress warnings) missing
	from clang
In-Reply-To: <bug-2384-206@http.llvm.org/bugs/>
Message-ID: <200805291537.m4TFb5Ln000893@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2384


Chris Lattner <sabre at nondot.org> changed:

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




--- Comment #1 from Chris Lattner <sabre at nondot.org>  2008-05-29 10:37:04 ---
Implemented, thanks:
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20080526/005912.html


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 10:38:46 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 10:38:46 -0500
Subject: [LLVMbugs] [Bug 2380] crash emitting debug info in llvm-gcc
	front-end
In-Reply-To: <bug-2380-206@http.llvm.org/bugs/>
Message-ID: <200805291538.m4TFckI6000964@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2380


Duncan Sands <baldrick at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |baldrick at free.fr
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED




--- Comment #1 from Duncan Sands <baldrick at free.fr>  2008-05-29 10:38:46 ---
Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063043.html


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 14:49:13 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 14:49:13 -0500
Subject: [LLVMbugs] [Bug 2371] codegen should inline small memmoves
In-Reply-To: <bug-2371-206@http.llvm.org/bugs/>
Message-ID: <200805291949.m4TJnDFt009645@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2371


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #4 from Dan Gohman <gohman at apple.com>  2008-05-29 14:49:13 ---
Implemented here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063058.html


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 15:20:56 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 15:20:56 -0500
Subject: [LLVMbugs] [Bug 2378] likely volatile bug
In-Reply-To: <bug-2378-206@http.llvm.org/bugs/>
Message-ID: <200805292020.m4TKKuat010595@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2378


Duncan Sands <baldrick at free.fr> changed:

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




--- Comment #5 from Duncan Sands <baldrick at free.fr>  2008-05-29 15:20:56 ---
It works for the reporter now.


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

From bugzilla-daemon at cs.uiuc.edu  Thu May 29 23:14:38 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Thu, 29 May 2008 23:14:38 -0500
Subject: [LLVMbugs] [Bug 1338] When SVN lands, we should rename...
In-Reply-To: <bug-1338-206@http.llvm.org/bugs/>
Message-ID: <200805300414.m4U4Ec94026548@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1338


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

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




--- Comment #7 from Anton Korobeynikov <asl at math.spbu.ru>  2008-05-29 23:14:37 ---
Done in:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063049.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063050.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063054.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063055.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063057.html
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063071.html


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 30 05:11:36 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 May 2008 05:11:36 -0500
Subject: [LLVMbugs] [Bug 2385] New: Use of freed memory by lcssa
Message-ID: <bug-2385-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2385

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


Created an attachment (id=1679)
 --> (http://llvm.org/bugs/attachment.cgi?id=1679)
testcase .ll

Running opt -lcssa on the attached testcase causes valgrind to say:

Invalid write of size 8
   at 0x82850E: (anonymous
namespace)::LCSSA::GetValueForBlock(llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Instruction*, llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*> >&) (LCSSA.cpp:275)
   by 0x828506: (anonymous
namespace)::LCSSA::GetValueForBlock(llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Instruction*, llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*> >&) (LCSSA.cpp:275)
   by 0x828D45: (anonymous
namespace)::LCSSA::ProcessInstruction(llvm::Instruction*,
llvm::SmallVector<llvm::BasicBlock*, 8> const&) (LCSSA.cpp:190)
   by 0x828F3F: (anonymous
namespace)::LCSSA::runOnLoop(llvm::LoopBase<llvm::BasicBlock>*,
llvm::LPPassManager&) (LCSSA.cpp:132)
   by 0x89DB71: llvm::LPPassManager::runOnFunction(llvm::Function&)
(LoopPass.cpp:225)
   by 0x95E196: llvm::FPPassManager::runOnFunction(llvm::Function&)
(PassManager.cpp:1178)
   by 0x95E303: llvm::FPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1198)
   by 0x95DE3B: llvm::MPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1248)
   by 0x95DFCB: llvm::PassManagerImpl::run(llvm::Module&)
(PassManager.cpp:1322)
   by 0x95E02E: llvm::PassManager::run(llvm::Module&) (PassManager.cpp:1354)
   by 0x6B9470: main (opt.cpp:431)
 Address 0x61a5d60 is 488 bytes inside a block of size 1,024 free'd
   at 0x4C22B3E: operator delete[](void*) (vg_replace_malloc.c:364)
   by 0x82959D: llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*> >::grow(unsigned) (DenseMap.h:360)
   by 0x82960F: llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*>
>::InsertIntoBucket(llvm::DomTreeNodeBase<llvm::BasicBlock>* const&,
llvm::Value* const&, std::pair<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*>*) (DenseMap.h:242)
   by 0x8296E9: llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*>
>::FindAndConstruct(llvm::DomTreeNodeBase<llvm::BasicBlock>* const&)
(DenseMap.h:185)
   by 0x829710: llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*>
>::operator[](llvm::DomTreeNodeBase<llvm::BasicBlock>* const&) (DenseMap.h:189)
   by 0x828496: (anonymous
namespace)::LCSSA::GetValueForBlock(llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Instruction*, llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*> >&) (LCSSA.cpp:256)
   by 0x828506: (anonymous
namespace)::LCSSA::GetValueForBlock(llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Instruction*, llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*> >&) (LCSSA.cpp:275)
   by 0x828506: (anonymous
namespace)::LCSSA::GetValueForBlock(llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Instruction*, llvm::DenseMap<llvm::DomTreeNodeBase<llvm::BasicBlock>*,
llvm::Value*, llvm::DenseMapInfo<llvm::DomTreeNodeBase<llvm::BasicBlock>*>,
llvm::DenseMapInfo<llvm::Value*> >&) (LCSSA.cpp:275)
   by 0x828D45: (anonymous
namespace)::LCSSA::ProcessInstruction(llvm::Instruction*,
llvm::SmallVector<llvm::BasicBlock*, 8> const&) (LCSSA.cpp:190)
   by 0x828F3F: (anonymous
namespace)::LCSSA::runOnLoop(llvm::LoopBase<llvm::BasicBlock>*,
llvm::LPPassManager&) (LCSSA.cpp:132)
   by 0x89DB71: llvm::LPPassManager::runOnFunction(llvm::Function&)
(LoopPass.cpp:225)
   by 0x95E196: llvm::FPPassManager::runOnFunction(llvm::Function&)
(PassManager.cpp:1178)


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 30 06:48:31 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 May 2008 06:48:31 -0500
Subject: [LLVMbugs] [Bug 2381] Nethack doesn't compile!!!
In-Reply-To: <bug-2381-206@http.llvm.org/bugs/>
Message-ID: <200805301148.m4UBmVjr020400@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2381


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #4 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-30 06:48:29 ---
Fixed in r51763.

I managed to get NetHack working with a few source tweaks to work around the
lack of transparent_union support.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 30 06:48:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 May 2008 06:48:58 -0500
Subject: [LLVMbugs] [Bug 2309] Implement constant initializer codegen for
	bitfields
In-Reply-To: <bug-2309-206@http.llvm.org/bugs/>
Message-ID: <200805301148.m4UBmwPf020433@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2309


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #2 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-30 06:48:58 ---
Fixed in r51763.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 30 06:57:08 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 May 2008 06:57:08 -0500
Subject: [LLVMbugs] [Bug 2386] New: erraneous html with -emit-html
Message-ID: <bug-2386-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2386

           Summary: erraneous html with -emit-html
           Product: clang
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Basic
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: hs4233 at mail.mn-solutions.de
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1680)
 --> (http://llvm.org/bugs/attachment.cgi?id=1680)
the source code

Some instances of <tr> (for indented text) are emitted as tr>. I've attached:

* the source-code file
* a snapshot on how it looks on Konqueror (it looks the same on Firefox)
* the emitted HTML file

This happened with LLVM SVN 51761 and CFE SVN 51761.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 30 12:32:30 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 May 2008 12:32:30 -0500
Subject: [LLVMbugs] [Bug 2385] Use of freed memory by lcssa
In-Reply-To: <bug-2385-206@http.llvm.org/bugs/>
Message-ID: <200805301732.m4UHWUgt031412@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2385


Owen Anderson <resistor at mac.com> changed:

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




--- Comment #1 from Owen Anderson <resistor at mac.com>  2008-05-30 12:32:30 ---
Fixed in r51777.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 30 15:40:10 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 May 2008 15:40:10 -0500
Subject: [LLVMbugs] [Bug 1893] Incomplete array "int i[];
 " should implicitly have one element
In-Reply-To: <bug-1893-206@http.llvm.org/bugs/>
Message-ID: <200805302040.m4UKeAmR004185@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=1893


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-30 15:40:09 ---
Fixed in r51787.


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

From bugzilla-daemon at cs.uiuc.edu  Fri May 30 16:00:12 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Fri, 30 May 2008 16:00:12 -0500
Subject: [LLVMbugs] [Bug 2387] New: setcc gets promoted from i1 to i32,
 but select remains i1 and triggers assert
Message-ID: <bug-2387-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2387

           Summary: setcc gets promoted from i1 to i32, but select remains
                    i1 and triggers assert
           Product: new-bugs
           Version: unspecified
          Platform: Macintosh
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: scooter.phd at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1683)
 --> (http://llvm.org/bugs/attachment.cgi?id=1683)
test case _addvsi3.ll

Always seems to be one more bug between me and a working CellSPU backend.
Today, the bug appears to be a setcc/select promotion problem. Here's the DAG
dump from inside gdb:

SelectionDAG has 25 nodes:
  0x1b04bc0: ch = EntryToken 
  0x1b04c20: i8 = Constant <1>
  0x1b04cc0: i32 = Constant <4294967295>
    0x1b04bc0: <multiple use>
    0x1b04df0: i32 = Register  #1025
  0x1b04e20: i32,ch = CopyFromReg 0x1b04bc0, 0x1b04df0
    0x1b04bc0: <multiple use>
    0x1b04e80: i32 = Register  #1026
  0x1b04eb0: i32,ch = CopyFromReg 0x1b04bc0, 0x1b04e80
    0x1b04eb0: <multiple use>
    0x1b04e20: <multiple use>
  0x1b04f80: i32 = add 0x1b04eb0, 0x1b04e20
  0x1b05030: ch = setgt 
  0x1b051e0: ch = setlt 
      0x1b04eb0: <multiple use>
      0x1b04cc0: <multiple use>
      0x1b05030: <multiple use>
    0x1b05610: i32 = setcc 0x1b04eb0, 0x1b04cc0, 0x1b05030
      0x1b04f80: <multiple use>
      0x1b04e20: <multiple use>
      0x1b051e0: <multiple use>
    0x1b055a0: i32 = setcc 0x1b04f80, 0x1b04e20, 0x1b051e0
      0x1b04f80: <multiple use>
      0x1b04e20: <multiple use>
      0x1b05030: <multiple use>
    0x1b04f10: i32 = setcc 0x1b04f80, 0x1b04e20, 0x1b05030
  0x1b056d0: i8 = select 0x1b05610, 0x1b055a0, 0x1b04f10
[truncated at this point because here's the bug]

The setcc's have all been promoted to i32, but the select didn't. This triggers
an assert much later on in SelectionDAG::ComputeMaskedBits because the bit
widths are the wrong size (clearly!)

Looking for advice on where to start debugging this bug...

To reproduce:

llvm-as -o - _addvsi3.ll | llc -march=cellspu -o -


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 03:00:37 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 03:00:37 -0500
Subject: [LLVMbugs] [Bug 2388] New: indvars makes i1 indvar, LSR doesn't fix
Message-ID: <bug-2388-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2388

           Summary: indvars makes i1 indvar, LSR doesn't fix
           Product: libraries
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu


Consider:
void foo(int x) {
  int i;
  _Bool b = x < 0;
  for (i = 0; i < 100000; ++i) {
    bar(i, b);

    b = !b;
  }
}

This compiles into this entertaining code:

bb:             ; preds = %bb, %entry
        %i.0.reg2mem.0 = phi i32 [ 0, %entry ], [ %indvar.next, %bb ]          
; <i32> [#uses=3]
        %indvar23 = trunc i32 %i.0.reg2mem.0 to i1              ; <i1>
[#uses=1]
        %b.0.in.reg2mem.0 = add i1 %indvar23, %tmp1             ; <i1>
[#uses=1]
        %tmp34 = zext i1 %b.0.in.reg2mem.0 to i32               ; <i32>
[#uses=1]


This is really bad.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 03:36:00 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 03:36:00 -0500
Subject: [LLVMbugs] [Bug 2389] New: instcombine should turn add i1 -> xor i1
Message-ID: <bug-2389-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2389

           Summary: instcombine should turn add i1 -> xor i1
           Product: libraries
           Version: 1.0
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sabre at nondot.org
                CC: llvmbugs at cs.uiuc.edu


Silly testcase:

define i32 @test(i32 %r) {
        %s = icmp eq i32 %r, 0          ; <i1> [#uses=2]
        %t = add i1 %s, %s              ; <i1> [#uses=3]
        %u = zext i1 %t to i32          ; <i32> [#uses=2]
        br i1 %t, label %A, label %B

A:              ; preds = %0
        ret i32 %u

B:              ; preds = %0
        %v = select i1 %t, i32 %r, i32 %u               ; <i32> [#uses=1]
        ret i32 %v
}


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 03:59:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 03:59:05 -0500
Subject: [LLVMbugs] [Bug 2390] New: LLVM doesn' t match the x86 linux
 calling convention for see vectors
Message-ID: <bug-2390-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2390

           Summary: LLVM doesn't match the x86 linux calling convention for
                    see vectors
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:

#include <xmmintrin.h>
__m128i r(__m128 d1, __m128 d2, __m128 d3, __m128i r, int t, __m128i s) {return
r+s;}

llvm-gcc code:
r:
        pushl   %ebp
        movl    %esp, %ebp
        andl    $-16, %esp
        movaps  %xmm3, %xmm0
        paddq   24(%ebp), %xmm0
        movl    %ebp, %esp
        popl    %ebp
        ret

gcc code:
r:
        subl    $12, %esp
        movdqa  48(%esp), %xmm0
        paddq   16(%esp), %xmm0
        addl    $12, %esp
        ret

Obviously not going to work for calls between them.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 06:27:39 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 06:27:39 -0500
Subject: [LLVMbugs] [Bug 2391] New: extractvalue doesn
Message-ID: <bug-2391-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2391

           Summary: extractvalue doesn
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


%struct.anon = type { i32, i32 }
define i32 @a(%struct.anon %r,  double %s) nounwind  {
%tmp = extractvalue %struct.anon %r, 0
ret i32 %tmp
}


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 06:28:40 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 06:28:40 -0500
Subject: [LLVMbugs] [Bug 2391] extractvalue doesn
In-Reply-To: <bug-2391-206@http.llvm.org/bugs/>
Message-ID: <200805311128.m4VBSehP004466@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2391


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-31 06:28:39 ---
Bleh, accidentally pressed the file bug button too early; I'll try again.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 06:29:01 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 06:29:01 -0500
Subject: [LLVMbugs] [Bug 2392] New: extractvalue doesn
Message-ID: <bug-2392-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2392

           Summary: extractvalue doesn
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


%struct.anon = type { i32, i32 }
define i32 @a(%struct.anon %r,  double %s) nounwind  {
%tmp = extractvalue %struct.anon %r, 0
ret i32 %tmp
}


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 06:29:34 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 06:29:34 -0500
Subject: [LLVMbugs] [Bug 2392] extractvalue doesn
In-Reply-To: <bug-2392-206@http.llvm.org/bugs/>
Message-ID: <200805311129.m4VBTYZP004524@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2392


Eli Friedman <sharparrow1 at yahoo.com> changed:

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




--- Comment #1 from Eli Friedman <sharparrow1 at yahoo.com>  2008-05-31 06:29:34 ---
Sorry, I'm really not trying to do this.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 06:31:56 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 06:31:56 -0500
Subject: [LLVMbugs] [Bug 2393] New: extractvalue doesn't properly round-trip
 through bitcode
Message-ID: <bug-2393-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2393

           Summary: extractvalue doesn't properly round-trip through bitcode
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


%struct.anon = type { i32, i32 }
define i32 @a(%struct.anon %r,  double %s) nounwind  {
%tmp = extractvalue %struct.anon %r, 0
ret i32 %tmp
}

crashes run through "llvm-as | llvm-dis".  (Using SVN r51813.)

I know this is very new code, but I figured this report might be useful anyway.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 08:02:05 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 08:02:05 -0500
Subject: [LLVMbugs] [Bug 2394] New: llvm-gcc generates wrong code for load
 from bitfield in packed struct
Message-ID: <bug-2394-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2394

           Summary: llvm-gcc generates wrong code for load from bitfield in
                    packed struct
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


struct __attribute((packed)) x {int a : 24;};
int a(struct x* g) {return g->a;}

Code generated with llvm-gcc:
a:
        movl    4(%esp), %eax
        movl    (%eax), %eax
        shll    $8, %eax
        sarl    $8, %eax
        ret

This is broken because the last byte might not be legal to load.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 11:28:32 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 11:28:32 -0500
Subject: [LLVMbugs] [Bug 2395] New: crash in BasicAA with gep+bitcasts
Message-ID: <bug-2395-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2395

           Summary: crash in BasicAA with gep+bitcasts
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1685)
 --> (http://llvm.org/bugs/attachment.cgi?id=1685)
Testcase

Crash on attached testcase; to reproduce, run "opt -gvn" over the input.

The crash is in CheckGEPInstructions
(llvm/lib/Analysis/BasicAliasAnalysis.cpp:687). I haven't looked too closely.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 12:11:10 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 12:11:10 -0500
Subject: [LLVMbugs] [Bug 2389] instcombine should turn add i1 -> xor i1
In-Reply-To: <bug-2389-206@http.llvm.org/bugs/>
Message-ID: <200805311711.m4VHBAxg014043@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2389


Nick Lewycky <nicholas at mxc.ca> changed:

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




--- Comment #1 from Nick Lewycky <nicholas at mxc.ca>  2008-05-31 12:11:09 ---
Fixed. Patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080526/063184.html


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 13:12:58 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 13:12:58 -0500
Subject: [LLVMbugs] [Bug 2396] New: Miscompilation with stack realignment
Message-ID: <bug-2396-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2396

           Summary: Miscompilation with stack realignment
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=1686)
 --> (http://llvm.org/bugs/attachment.cgi?id=1686)
Testcase (IL)

Attached testcase requires the stack to be realigned on Linux-x86 because it
uses align 16 on an alloca.  This causes the code to be miscompiled to hang for
a while, then eventually crash.  If the align 16 is removed, the test finishes
almost instantly.

Crash occurs when compiled with "llvm-as | llc | gcc -x assembler -" on Ubuntu
Linux x86.


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 13:57:42 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 13:57:42 -0500
Subject: [LLVMbugs] [Bug 2397] New: gvn reads freed memory
Message-ID: <bug-2397-206@http.llvm.org/bugs/>

http://llvm.org/bugs/show_bug.cgi?id=2397

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


Created an attachment (id=1688)
 --> (http://llvm.org/bugs/attachment.cgi?id=1688)
testcase .ll

$ valgrind opt -gvn x.bc --disable-output

Invalid read of size 2
   at 0x82C4562: llvm::Value::getValueID() const (Value.h:212)
   by 0x82C45D6: bool llvm::isa_impl<llvm::Instruction,
llvm::Value>(llvm::Value const&) (Value.h:261)
   by 0x82C4CA4: llvm::isa_impl_wrap<llvm::Instruction, llvm::Value const,
llvm::Value const>::doit(llvm::Value const&) (Casting.h:71)
   by 0x82C4CBA: bool
llvm::isa_impl_cl<llvm::Value>::isa<llvm::Instruction>(llvm::Value const&)
(Casting.h:83)
   by 0x82E01C2: bool llvm::isa_impl_cl<llvm::Value
const>::isa<llvm::Instruction>(llvm::Value const&) (Casting.h:92)
   by 0x82E01D8: bool llvm::isa_impl_cl<llvm::Value
const*>::isa<llvm::Instruction>(llvm::Value const*) (Casting.h:101)
   by 0x82E01F0: bool llvm::isa<llvm::Instruction, llvm::Value
const*>(llvm::Value const* const&) (Casting.h:116)
   by 0x83079A6: llvm::StoreInst::classof(llvm::Value const*)
(Instructions.h:354)
   by 0x834EED0: bool llvm::isa_impl<llvm::StoreInst, llvm::Value>(llvm::Value
const&) (Casting.h:54)
   by 0x834EEE6: llvm::isa_impl_wrap<llvm::StoreInst, llvm::Value const,
llvm::Value const>::doit(llvm::Value const&) (Casting.h:71)
   by 0x834EEFC: bool
llvm::isa_impl_cl<llvm::Value>::isa<llvm::StoreInst>(llvm::Value const&)
(Casting.h:83)
   by 0x834EF12: bool
llvm::isa_impl_cl<llvm::Value*>::isa<llvm::StoreInst>(llvm::Value*)
(Casting.h:101)
 Address 0x6c7ace0 is 16 bytes inside a block of size 52 free'd
   at 0x68022EC: operator delete(void*) (vg_replace_malloc.c:342)
   by 0x858DA60: llvm::User::operator delete(void*) (Value.cpp:375)
   by 0x855F2BA: llvm::LoadInst::~LoadInst() (Instructions.h:216)
   by 0x82F2F16: llvm::iplist<llvm::Instruction,
llvm::ilist_traits<llvm::Instruction>
>::erase(llvm::ilist_iterator<llvm::Instruction>) (ilist.h:377)
   by 0x854EF26: llvm::Instruction::eraseFromParent() (Instruction.cpp:68)
   by 0x834CFFA: (anonymous namespace)::GVN::iterateOnFunction(llvm::Function&)
(GVN.cpp:1186)
   by 0x834D17F: (anonymous namespace)::GVN::runOnFunction(llvm::Function&)
(GVN.cpp:1120)
   by 0x856F59A: llvm::FPPassManager::runOnFunction(llvm::Function&)
(PassManager.cpp:1178)
   by 0x856F73B: llvm::FPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1198)
   by 0x856F221: llvm::MPPassManager::runOnModule(llvm::Module&)
(PassManager.cpp:1248)
   by 0x856F3D9: llvm::PassManagerImpl::run(llvm::Module&)
(PassManager.cpp:1322)
   by 0x856F42C: llvm::PassManager::run(llvm::Module&) (PassManager.cpp:1354)


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

From bugzilla-daemon at cs.uiuc.edu  Sat May 31 14:27:53 2008
From: bugzilla-daemon at cs.uiuc.edu (bugzilla-daemon at cs.uiuc.edu)
Date: Sat, 31 May 2008 14:27:53 -0500
Subject: [LLVMbugs] [Bug 2393] extractvalue doesn't properly round-trip
	through bitcode
In-Reply-To: <bug-2393-206@http.llvm.org/bugs/>
Message-ID: <200805311927.m4VJRrG5017866@zion.cs.uiuc.edu>

http://llvm.org/bugs/show_bug.cgi?id=2393


Dan Gohman <gohman at apple.com> changed:

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




--- Comment #1 from Dan Gohman <gohman at apple.com>  2008-05-31 14:27:53 ---
This is fixed in 51820-51823.


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