[Bug 956051] Re: libc6 crash while running 'xm'

[jwestfall]

I am running into this issue as well on a 12.04 domU with a AMD Opteron
6272 based box.

[14553.237586] apport[18522] trap invalid opcode ip:7f1d632625fc
sp:7ffff8701390 error:0 in libm-2.15.so[7f1d63220000+f9000]

Debugging info

(gdb) exec-file python-dbg
(gdb) run
Starting program: /usr/bin/python-dbg 
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Python 2.7.3 (default, Apr 20 2012, 22:01:19) 
[GCC 4.6.3] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> print test

Program received signal SIGILL, Illegal instruction.
__ieee754_exp_fma4 (x=-0.5) at ../sysdeps/ieee754/dbl-64/e_exp.c:63
63	../sysdeps/ieee754/dbl-64/e_exp.c: No such file or directory.
(gdb) bt
#0  __ieee754_exp_fma4 (x=-0.5) at ../sysdeps/ieee754/dbl-64/e_exp.c:63
#1  0x000000000058509f in ?? ()
#2  0x00000000009bde80 in ?? ()
#3  0x0000000100c52a10 in ?? ()
#4  0x0000000000417380 in ?? ()
#5  0x00000000009d2970 in ?? ()
#6  0x0000000000000000 in ?? ()
(gdb) info address __ieee754_exp_fma4
Symbol "__ieee754_exp_fma4" is a function at address 0x7ffff6cc35f0.
(gdb) disassemble 0x7ffff6cc35f0,+30
Dump of assembler code from 0x7ffff6cc35f0 to 0x7ffff6cc360e:
   0x00007ffff6cc35f0 <__ieee754_exp_fma4+0>:	push   %rbp
   0x00007ffff6cc35f1 <__ieee754_exp_fma4+1>:	mov    %rsp,%rbp
   0x00007ffff6cc35f4 <__ieee754_exp_fma4+4>:	and    $0xffffffffffffffe0,%rsp
   0x00007ffff6cc35f8 <__ieee754_exp_fma4+8>:	add    $0x10,%rsp
=> 0x00007ffff6cc35fc <__ieee754_exp_fma4+12>:	vmovsd %xmm0,-0x20(%rsp)
   0x00007ffff6cc3602 <__ieee754_exp_fma4+18>:	mov    -0x20(%rsp),%rax
   0x00007ffff6cc3607 <__ieee754_exp_fma4+23>:	mov    %rax,%rcx
   0x00007ffff6cc360a <__ieee754_exp_fma4+26>:	shr    $0x20,%rcx

vmovsd is an avx instruction, but avx isn't supported according to cpu
flags.

# grep flags /proc/cpuinfo | head -1
flags		: fpu de tsc msr pae cx8 cmov pat clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt lm rep_good nopl aperfmperf pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt aes hypervisor lahf_lm cmp_legacy extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch xop fma4 perfctr_core arat cpb

perhaps eglibc is assuming if fma4 is supported, avx must be too.

thanks
jim

-- 
You received this bug notification because you are a member of Ubuntu
Foundations Bugs, which is subscribed to eglibc in Ubuntu.
https://bugs.launchpad.net/bugs/956051

Title:
  libc6 crash while running 'xm'

Status in “eglibc” package in Ubuntu:
  Confirmed

Bug description:
  Ubuntu Version: 12.04 (precise)

  libc6: 2.15-0ubuntu5
  xen-hypervisor-4.1-amd64: 4.1.2-2ubuntu2
  linux-image-generic: 3.2.0.18.20

  Everytime I try to run 'xm list' on a new AMD Opteron 6274 system (Bulldozer architecture) I'm getting a crash.
  I couldn't even file a bug report through "apport-bug".
  Repeating the same procedure with Ubuntu Oneiric showed no problems at all.

  dmesg showed following messages:

  [  479.891581] xm[1746] trap invalid opcode ip:7fd684f8a5fc sp:7fff3ebb3c70 error:0 in libm-2.15.so[7fd684f48000+f9000]
  [  479.937611] apport[1758] trap invalid opcode ip:7fd6cb9ea5fc sp:7fff9bbc06b0 error:0 in libm-2.15.so[7fd6cb9a8000+f9000]
  [  479.937641] Process 1758(apport) has RLIMIT_CORE set to 1
  [  479.937645] Aborting core

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/eglibc/+bug/956051/+subscriptions