similar to: [LLVMdev] llvm-gcc Bootstrap Errors

I'm getting the following when building llvm-gcc with an optimized set of LLVM libraries: /ptmp/dag/build.llvm-gcc-4.2.trunk.official.opt/x86_64-unknown-linux-gnu/./gcc/xgcc -B/ptmp/dag/build.llvm-gcc-4.2.trunk.official.opt/x86_64-unknown-linux-gnu/./gcc/ -B/cray/iss/compiler/cost/tools/llvm-tools/llvm/install.trunk.official.opt/x86_64-unknown-linux-gnu/x86_64-unknown-linux-gnu/bin/

On Wednesday 10 December 2008 12:18, Török Edwin wrote: > On 2008-12-10 19:54, David Greene wrote: > > I'm almost finished with the BuildBot validation config, but I'm having > > trouble convincing it to compiler llvm-gcc-4.2. > > > > Here's the relevant portion of the config: > > > > f.addStep(SVN(workdir="llvm-gcc-4.2/src",

Got a strange problem. With our modified llvm here,llvm-gcc won't bootstrap. It fails compiling unwind-dw2.c during translation from the gcc IR to llvm. It fails with -O0 -emit-llvm so according to the docs, this is strictly a frontend bug. A run of delta produced an 18 line testcase with nothing remarkable in it. However, it bootstraps fine using the official llvm sources. Note that the

I've had failing llvm-gcc bootstraps on x86_64 for about a week now. It falls over when trying to link makedepend with a stage1 compiler: [x86_64-off-opt]: libcpp.a(charset.o): In function `init_iconv_desc': [x86_64-off-opt]: (.text+0xef7): undefined reference to `alloca' [And hundreds more of similar nature] Supposedly gcc doesn't inline alloca if -ansi or -fno-builtin is

On Apr 19, 2007, at 2:37 PM, David Greene wrote: > Got a strange problem. With our modified llvm here,llvm-gcc won't > bootstrap. It fails compiling unwind-dw2.c during translation from > the gcc IR to llvm. It fails with -O0 -emit-llvm so according to > the docs, this is strictly a frontend bug. A run of delta produced > an 18 line testcase with nothing remarkable in it.

I'm almost finished with the BuildBot validation config, but I'm having trouble convincing it to compiler llvm-gcc-4.2. Here's the relevant portion of the config: f.addStep(SVN(workdir="llvm-gcc-4.2/src", mode='clobber', baseURL='http://llvm.org/svn/llvm-project/llvm-gcc-4.2/', defaultBranch='trunk')) checking="" if optimized ==

Yes. On Mon, Nov 10, 2008 at 4:45 PM, David Greene <dag at cray.com> wrote: > On Monday 10 November 2008 15:28, Bill Wendling wrote: >> Hi Daniel, >> >> I see that you're doing a Debug build. Could you do a "Release" or >> "Release-Asserts" build instead? This will catch errors which >> developers won't normally see. > > I

On Monday 10 November 2008 16:50, Daniel Berlin wrote: > Yes. Cool. I'd like to look at installing something here. You offered to walk us through how to do this. I'll see if I can install buildbot here and then mail you back when I'm ready to go. -Dave > On Mon, Nov 10, 2008 at 4:45 PM, David Greene <dag at

Forgot to CC the list. I'm looking into it. -Dave ---------- Forwarded Message ---------- Subject: Re: [LLVMdev] Bootstrap Failure Date: Thursday 12 November 2009 15:50 From: David Greene <dag at cray.com> To: Bill Wendling <wendling at apple.com> On Thursday 12 November 2009 15:49, you wrote: > These are the likely culprits. David, it looks

>> These results have far fewer failures than svn-trunk, and are also >> comparable to bootstrapping with gcc-4.6.2, summarized here: >> http://paste.lisp.org/display/126363 >> (Unfortunately, I no longer have the whole build/test log for the gcc46 bootstrap.) >> This consistency between different bootstraps of the release gives me >> some hope that g++-4.0.1 is

David Greene wrote: > I'm getting the following when building llvm-gcc with an optimized set of > LLVM libraries: > > /ptmp/dag/build.llvm-gcc-4.2.trunk.official.opt/x86_64-unknown-linux-gnu/./gcc/xgcc > -B/ptmp/dag/build.llvm-gcc-4.2.trunk.official.opt/x86_64-unknown-linux-gnu/./gcc/ >

On Monday 10 November 2008 15:28, Bill Wendling wrote: > Hi Daniel, > > I see that you're doing a Debug build. Could you do a "Release" or > "Release-Asserts" build instead? This will catch errors which > developers won't normally see. I never knew about buildbot until just now. This could really help with the validation proposal I just posted. Can

On Fri, Dec 16, 2011 at 12:17:32PM -0500, David Fang wrote: >>> These results have far fewer failures than svn-trunk, and are also >>> comparable to bootstrapping with gcc-4.6.2, summarized here: >>> http://paste.lisp.org/display/126363 >>> (Unfortunately, I no longer have the whole build/test log for the gcc46 bootstrap.) >>> This consistency between

On 2008-12-10 19:54, David Greene wrote: > I'm almost finished with the BuildBot validation config, but I'm having > trouble convincing it to compiler llvm-gcc-4.2. > > Here's the relevant portion of the config: > > f.addStep(SVN(workdir="llvm-gcc-4.2/src", mode='clobber', > baseURL='http://llvm.org/svn/llvm-project/llvm-gcc-4.2/',

On 2008-12-10 22:17, David Greene wrote: > On Wednesday 10 December 2008 12:18, Török Edwin wrote: > >> On 2008-12-10 19:54, David Greene wrote: >> >>> I'm almost finished with the BuildBot validation config, but I'm having >>> trouble convincing it to compiler llvm-gcc-4.2. >>> >>> Here's the relevant portion of the config:

On Wednesday 10 October 2007 12:53, Chris Lattner wrote: > On Mon, 8 Oct 2007, Evan Cheng wrote: > > I am turning 32-bit add into 32-bit LEA on x86-64 but that's > > perfectly legal. Both > > leal (%esi,%edi), %eax > > leal (%rsi,%rdi), %eax > > are legal assembly. It's just the former requires a 67H prefix due to > > the 32-bit address size. >

Here's the .bc file -------------- next part -------------- A non-text attachment was scrubbed... Name: testcase.bc Type: application/octet-stream Size: 1328 bytes Desc: not available URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20091112/79af2e36/attachment.obj> -------------- next part -------------- -bw On Nov 12, 2009, at 1:52 PM, David Greene wrote: > Forgot to

On Oct 8, 2007, at 12:50 PM, David Greene wrote: > I've been unable to bootstrap llvm-gcc-4.0 or -4.2 for x86_64 > for some weeks now. The current problem is this: > > [x86_64-mod-dbg]: ./xgcc -B./ > -B/install.modified.debug/x86_64-unknown-linux-gnu/x86_64-unknown- > linux-gnu/bin/ > -isystem /install.modified.debug/x86_64-unknown-linux-gnu/x86_64- >

On Mon, 8 Oct 2007, Evan Cheng wrote: > I am turning 32-bit add into 32-bit LEA on x86-64 but that's > perfectly legal. Both > leal (%esi,%edi), %eax > leal (%rsi,%rdi), %eax > are legal assembly. It's just the former requires a 67H prefix due to > the 32-bit address size. > > This does point to a performance problem in the 3 address conversion > code (which I

I'm still trying to debug the validator buildbot. I'm getting this error when building some variants of llvm-gcc-4.2: /bin/sh ../../src/gcc/mkconfig.sh bconfig.h gcc -c -g -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings -Wold-style-definition -Wmissing-format-attribute -fno-common