[ARM] Regression between r215896 and r215960

İsmail Dönmez ismail at donmez.ws
Wed Aug 20 05:06:43 PDT 2014


Hmm looking at the build bot r215843 was the last successful one
and r215870 was the first failing. There are three suspect commits on this
range:

r215846
r215862
r215863

CC'ing Saleem Abdulrasool who authered those three commits. Maybe he has an
idea.

Thanks.



On Wed, Aug 20, 2014 at 8:31 AM, Renato Golin <renato.golin at linaro.org>
wrote:

> Just found out that this is a bug in the clang/llvm compiled with that
> particular version, as opposed to that clang/llvm version compiled
> with an older clang. This will make finding the bug a lot harder... :(
>
> We may have to revert whatever patch broke it before we can actually fix
> it.
>
> cheers,
> --renato
>
> On 19 August 2014 22:12, Renato Golin <renato.golin at linaro.org> wrote:
> > Hi Ismail,
> >
> > Our buildbot picked it up:
> >
> >
> http://lab.llvm.org:8011/builders/clang-native-arm-cortex-a15-self-host/builds/1985
> >
> > I'm bisecting right now.
> >
> > cheers,
> > --renato
> >
> > On 19 August 2014 14:31, İsmail Dönmez <ismail at donmez.ws> wrote:
> >> Hi,
> >>
> >> To reproduce, get http://www.cs.utah.edu/~aek/code/card.cpp and
> >>
> >> [~]> clang++ -O1 card.cpp -o card
> >> clang-3.6: ../lib/Support/IntervalMap.cpp:138: IdxPair
> >> llvm::IntervalMapImpl::distribute(unsigned int, unsigned int, unsigned
> int,
> >> const unsigned int *, unsigned int *, unsigned int, bool): Assertion
> `Sum ==
> >> Elements + Grow && "Bad distribution sum"' failed.
> >> Stack dump:
> >> 0.      Program arguments: /havana/dist/llvm/bin/clang-3.6 -cc1 -triple
> >> armv7-unknown-linux-gnueabihf -emit-obj -disable-free -main-file-name
> >> card.cpp -mrelocation-model static -mdisable-fp-elim -fmath-errno
> >> -masm-verbose -mconstructor-aliases -fuse-init-array -target-cpu
> cortex-a8
> >> -target-abi aapcs-linux -mfloat-abi hard -dwarf-column-info
> -resource-dir
> >> /havana/dist/llvm/bin/../lib/clang/3.6.0 -internal-isystem
> >> /usr/lib/gcc/arm-linux-gnueabihf/4.9/../../../../include/c++/4.9
> >> -internal-isystem
> >>
> /usr/lib/gcc/arm-linux-gnueabihf/4.9/../../../../include/c++/4.9/arm-linux-gnueabihf
> >> -internal-isystem
> >>
> /usr/lib/gcc/arm-linux-gnueabihf/4.9/../../../../include/c++/4.9/backward
> >> -internal-isystem
> >>
> /usr/lib/gcc/arm-linux-gnueabihf/4.9/../../../../include/arm-linux-gnueabihf/c++/4.9
> >> -internal-isystem /usr/local/include -internal-isystem
> >> /havana/dist/llvm/bin/../lib/clang/3.6.0/include
> -internal-externc-isystem
> >> /usr/include/arm-linux-gnueabihf -internal-externc-isystem /include
> >> -internal-externc-isystem /usr/include -O1 -fdeprecated-macro
> >> -fdebug-compilation-dir /home/ismail -ferror-limit 19 -fmessage-length
> 104
> >> -mstackrealign -fno-signed-char -fobjc-runtime=gcc -fcxx-exceptions
> >> -fexceptions -fdiagnostics-show-option -fcolor-diagnostics -o
> >> /havana/tmp/card-13c953.o -x c++ card.cpp
> >> 1.      <eof> parser at end of file
> >> 2.      Code generation
> >> 3.      Running pass 'Function Pass Manager' on module 'card.cpp'.
> >> 4.      Running pass 'Greedy Register Allocator' on function
> >> '@_Z1T1vS_RfRS_'
> >> clang-3.6: error: unable to execute command: Aborted
> >> clang-3.6: error: clang frontend command failed due to signal (use -v
> to see
> >> invocation)
> >> clang version 3.6.0 (215976)
> >> Target: armv7l-unknown-linux-gnueabihf
> >> Thread model: posix
> >> clang-3.6: note: diagnostic msg: PLEASE submit a bug report to
> >> http://llvm.org/bugs/ and include the crash backtrace, preprocessed
> source,
> >> and associated run script.
> >> clang-3.6: note: diagnostic msg:
> >> ********************
> >>
> >> PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT:
> >> Preprocessed source(s) and associated run script(s) are located at:
> >> clang-3.6: note: diagnostic msg: /havana/tmp/card-e32b8e.cpp
> >> clang-3.6: note: diagnostic msg: /havana/tmp/card-e32b8e.sh
> >> clang-3.6: note: diagnostic msg:
> >>
> >> ********************
> >>
> >> This is on a native ARMv7 host (BeagleBone Black) and r215896 works
> fine.
> >>
> >>
> >> _______________________________________________
> >> cfe-commits mailing list
> >> cfe-commits at cs.uiuc.edu
> >> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
> >>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20140820/2d0785ce/attachment.html>


More information about the cfe-commits mailing list