bin/187266: clang 3.4 aborts when building math/py-numeric with -march=athlon64 on i386
Don Lewis
truckman at FreeBSD.org
Wed Mar 5 00:00:00 UTC 2014
>Number: 187266
>Category: bin
>Synopsis: clang 3.4 aborts when building math/py-numeric with -march=athlon64 on i386
>Confidential: no
>Severity: non-critical
>Priority: medium
>Responsible: freebsd-bugs
>State: open
>Quarter:
>Keywords:
>Date-Required:
>Class: sw-bug
>Submitter-Id: current-users
>Arrival-Date: Wed Mar 05 00:00:00 UTC 2014
>Closed-Date:
>Last-Modified:
>Originator: Don Lewis
>Release: FreeBSD 11.0-CURRENT i386
>Organization:
FreeBSD Project
>Environment:
System: FreeBSD scratch.catspoiler.org 11.0-CURRENT FreeBSD 11.0-CURRENT #71 r262666M: Sat Mar 1 16:42:54 PST 2014 dl at scratch.catspoiler.org:/usr/obj/usr/src/sys/GENERICSMB
Ports tree revision r346717.
py27-numeric-24.2_3
>Description:
When attempting to build math/py-numeric on a machine with an Athlon 64
CPU running in i386 mode, clang 3.4 aborts while compiling the dlaed3_()
function in the file dlapack_lite.c.
cc -DNDEBUG -O2 -pipe -march=athlon64 -fno-strict-aliasing -fPIC -IInclude -IPackages/FFT/Include -IPackages/RNG/Include -I/usr/local/include/python2.7 -c Src/dlapack_lite.c -o build/temp.freebsd-11.0-CURRENT-i386-2.7/Src/dlapack_lite.o
Instruction does not dominate all uses!
%arrayidx106 = getelementptr inbounds double* %dlamda, i32 %sub83
%bound1492 = icmp ule double* %arrayidx106, %scevgep473
Instruction does not dominate all uses!
%arrayidx106 = getelementptr inbounds double* %dlamda, i32 %sub83
%bound0491 = icmp ule double* %scevgep471, %arrayidx106
Broken module found, compilation aborted!
Stack dump:
0. Program arguments: /usr/bin/cc -cc1 -triple i386-unknown-freebsd11.0 -emit-obj -disable-free -main-file-name dlapack_lite.c -mrelocation-model pic -pic-level 2 -mdisable-fp-elim -relaxed-aliasing -masm-verbose -mconstructor-aliases -target-cpu athlon64 -coverage-file /usr/ports/math/py-numeric/work/Numeric-24.2/build/temp.freebsd-11.0-CURRENT-i386-2.7/Src/dlapack_lite.o -resource-dir /usr/bin/../lib/clang/3.4 -D NDEBUG -I Include -I Packages/FFT/Include -I Packages/RNG/Include -I /usr/local/include/python2.7 -O2 -fdebug-compilation-dir /usr/ports/math/py-numeric/work/Numeric-24.2 -ferror-limit 19 -fmessage-length 80 -mstackrealign -fobjc-runtime=gnustep -fdiagnostics-show-option -fcolor-diagnostics -vectorize-loops -vectorize-slp -o build/temp.freebsd-11.0-CURRENT-i386-2.7/Src/dlapack_lite.o -x c Src/dlapack_lite.c
1. <eof> parser at end of file
2. Code generation
3. Running pass 'Function Pass Manager' on module 'Src/dlapack_lite.c'.
4. Running pass 'Module Verifier' on function '@dlaed3_'
cc: error: unable to execute command: Abort trap (core dumped)
cc: error: clang frontend command failed due to signal (use -v to see invocation)
FreeBSD clang version 3.4 (tags/RELEASE_34/final 197956) 20140216
Target: i386-unknown-freebsd11.0
Thread model: posix
cc: note: diagnostic msg: PLEASE submit a bug report to http://llvm.org/bugs/ and include the crash backtrace, preprocessed source, and associated run script.
cc: note: diagnostic msg:
********************
PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT:
Preprocessed source(s) and associated run script(s) are located at:
cc: note: diagnostic msg: /tmp/dlapack_lite-26ef53.c
cc: note: diagnostic msg: /tmp/dlapack_lite-26ef53.sh
cc: note: diagnostic msg:
********************
error: command 'cc' failed with exit status 254
*** Error code 1
Stop.
make[1]: stopped in /usr/ports/math/py-numeric
*** Error code 1
Stop.
make: stopped in /usr/ports/math/py-numeric
The compilation succeeds if -march=athlon64 is *not* specified.
A bug report has been filed upstream:
<http://llvm.org/bugs/show_bug.cgi?id=19029>
>How-To-Repeat:
Attempt to compile dlapack_lite.c from math/py-numeric with clang 3.4
on i386 with the -march=athlon64 flag.
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted:
More information about the freebsd-bugs
mailing list