Segmentation fault happen when I was compiling 3.8.5 kernel

[Jan Kohoutek]

/usr/src/linux-3.8.5$ sudo make
make[1]: Nothing to be done for `all'.
make[1]: Nothing to be done for `relocs'.
  CHK include/generated/uapi/linux/version.h
  CHK include/generated/utsrelease.h
  CALL scripts/checksyscalls.sh
  CHK include/generated/compile.h
  CC arch/x86/kernel/cpu/capflags.o
  CC arch/x86/kernel/cpu/perf_event.o
arch/x86/kernel/cpu/perf_event.c: In function ‘get_segment_base’:
arch/x86/kernel/cpu/perf_event.c:1887:1: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <file:///usr/share/doc/gcc-4.7/README.Bugs> for instructions.
Preprocessed source stored into /tmp/ccj6Z7AH.out file, please attach this to your bugreport.
make[3]: *** [arch/x86/kernel/cpu/perf_event.o] Error 1
make[2]: *** [arch/x86/kernel/cpu] Error 2
make[1]: *** [arch/x86/kernel] Error 2
make: *** [arch/x86] Error 2

ProblemType: Crash
DistroRelease: Ubuntu 13.04
Package: cpp-4.7 4.7.2-23ubuntu1
ProcVersionSignature: Ubuntu 3.8.0-16.26-generic 3.8.5
Uname: Linux 3.8.0-16-generic x86_64
NonfreeKernelModules: zfs zunicode zavl zcommon znvpair
ApportVersion: 2.9.2-0ubuntu5
Architecture: amd64
Date: Fri Apr 5 10:11:37 2013
ExecutablePath: /usr/lib/gcc/x86_64-linux-gnu/4.7/cc1
InstallationDate: Installed on 2013-02-05 (58 days ago)
InstallationMedia: Ubuntu 12.10 "Quantal Quetzal" - Release amd64 (20121017.5)
MarkForUpload: True
SourcePackage: gcc-4.7
UpgradeStatus: No upgrade log present (probably fresh install)

[Doko-v (doko-v)]

Created attachment 29808
preprocessed source

lowering to -O1 works around it. r197184 does work, r197476 not.

$ gcc -c -g -O2 perf_event.i arch/x86/kernel/cpu/perf_event.c: In function 'get_segment_base':
arch/x86/kernel/cpu/perf_event.c:1887:1: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.

(gdb) bt
#0 0x0000000000bd312d in ?? ()
#1 0x0000000000bd305e in ?? ()
#2 0x0000000000bd2c6b in ?? ()
#3 0x0000000000bd357d in ?? ()
#4 0x0000000000bd51b0 in deps_analyze_insn(deps_desc*, rtx_def*) ()
#5 0x0000000000bd5603 in sched_analyze(deps_desc*, rtx_def*, rtx_def*) ()
#6 0x0000000000760d12 in sched_rgn_compute_dependencies(int) ()
#7 0x00000000007635f2 in schedule_insns() ()
#8 0x0000000000763a41 in ?? ()
#9 0x000000000070b778 in execute_one_pass(opt_pass*) ()
#10 0x000000000070bad5 in execute_pass_list(opt_pass*) ()
#11 0x000000000070bae7 in execute_pass_list(opt_pass*) ()
#12 0x000000000070bae7 in execute_pass_list(opt_pass*) ()
#13 0x00000000007e5439 in tree_rest_of_compilation(tree_node*) ()
#14 0x0000000000592afa in ?? ()
#15 0x000000000059434c in cgraph_optimize() ()
#16 0x00000000005947fa in cgraph_finalize_compilation_unit() ()
#17 0x00000000004d44e8 in c_write_global_declarations() ()
#18 0x000000000079edce in toplev_main(int, char**) ()
#19 0x00007ffff670dea5 in __libc_start_main () from /lib/x86_64-linux-gnu/libc.so.6
#20 0x00000000004c3351 in _start ()

[Doko-v (doko-v)]

seen on x86_64-linux-gnu

[Abel-i (abel-i)]

See http://gcc.gnu.org/ml/gcc-patches/2013-04/msg00003.html. Either the backport of 56077 should be reverted or, if the ICE is coming from a different place, Maxim's patch is applied instead.

[Doko-v (doko-v)]

Reverting the backport for 56077 lets the file compile

[Matthias Klose (doko)]

fixed in 4.7.2-23ubuntu2, will need some time to build on all architectures, get from -proposed for x86 in 2h.

[Abel-i (abel-i)]

Sure, then I'll revert the patch after testing -- Richi, are you fine with that?

[Doko-v (doko-v)]

backporting r190733 fixes the issue too, testsuite still running

[Rguenther-suse (rguenther-suse)]

"doko at gcc dot gnu.org" <email address hidden> wrote:

>
>http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56848
>
>--- Comment #5 from Matthias Klose <doko at gcc dot gnu.org> 2013-04-05
>12:30:38 UTC ---
>backporting r190733 fixes the issue too, testsuite still running

Please revert for 4.7.3 and re-apply for 4.7.3 with the additional backport.

[Jakub-gcc (jakub-gcc)]

(In reply to comment #6)
> "doko at gcc dot gnu.org" <email address hidden> wrote:
> Please revert for 4.7.3 and re-apply for 4.7.3 with the additional backport.

I guess you mean revert for 4.7.3 and re-apply for 4.7.4 with the additional backport ;)

[Doko-v (doko-v)]

the backport of r190733 doesn't show any regressions on x86_64-linux-gnu

[Ebotcazou (ebotcazou)]

This also affects Qemu.

[Ebotcazou (ebotcazou)]

Apparently the automatic xref didn't work, here is the commit:
  http://gcc.gnu.org/ml/gcc-cvs/2013-04/msg00244.html

[Launchpad Janitor (janitor)]

This bug was fixed in the package gcc-4.7 - 4.7.2-23ubuntu2

---------------
gcc-4.7 (4.7.2-23ubuntu2) raring; urgency=low

  * Fix PR middle-end/56848, reverting the fix for PR middle-end/56077.
    LP: #1164886.
 -- Matthias Klose <email address hidden> Fri, 05 Apr 2013 12:13:56 +0200

[Rguenth (rguenth)]

Fixed.

[Jackie-rosen (jackie-rosen)]

*** Bug 260998 has been marked as a duplicate of this bug. ***
Seen from the domain http://volichat.com
Page where seen: http://volichat.com/adult-chat-rooms
Marked for reference. Resolved as fixed @bugzilla.