https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c2 Markus Elfring changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW InfoProvider|Markus.Elfring@web.de | --- Comment #2 from Markus Elfring 2011-03-16 21:59:24 UTC --- Created an attachment (id=419819) --> (http://bugzilla.novell.com/attachment.cgi?id=419819) small make log I do not see that a core file is generated in my situation. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c4 --- Comment #4 from Markus Elfring 2011-03-16 22:34:21 UTC --- Created an attachment (id=419825) --> (http://bugzilla.novell.com/attachment.cgi?id=419825) small core dump I'm sorry that I forgot to switch off the file size limitation temporarily. Sonne:/usr/src/linux-2.6.38.y # file core* core: ELF 64-bit LSB core file x86-64, version 1 (SYSV), SVR4-style, from 'p gcc -Wp,-MD,s' Do you need the dump in variants for different version combinations of the generation environment? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c6 --- Comment #6 from Markus Elfring 2011-03-17 06:37:48 UTC --- (In reply to comment #3) I can not generate a back-trace myself by the tool "GNU debugger" at the moment because of the issue "GDB or dependency python not properly configured?" (bug #677225). -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c8 Jiri Slaby changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO InfoProvider| |Markus.Elfring@web.de --- Comment #8 from Jiri Slaby 2011-03-17 09:05:19 UTC --- (In reply to comment #7)

Created an attachment (id=419870) --> (http://bugzilla.novell.com/attachment.cgi?id=419870) [details] /usr/src/linux-2.6.38.y/scripts/basic/fixdep

Can you reproduce the program crash that I reported here also on your test systems?

No. It dies in: => 0x00000000006022d0 <+0>: movd %esi,%xmm1 What processor is that? (attach /proc/cpuinfo). I doubt it's something older than Pentium III? You definitely should have XMM registers. So I see no reason for it to die. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c10 Jiri Slaby changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO InfoProvider| |Markus.Elfring@web.de --- Comment #10 from Jiri Slaby 2011-03-17 09:48:55 UTC --- Well, about the generated exception, the manual says: #UD If CR0.EM[bit 2] = 1. (XMM register operations only) if CR4.OSFXSR[bit 9] = 0. (XMM register operations only) if CPUID.01H:EDX.SSE2[bit 26] = 0. #NM If CR0.TS[bit 3] = 1. CR0.EM should be 0 -- I assume you don't run a virtual machine. CR4.OSFXSR should be 1 -- you have fxsr capability, so the kernel should set it to 1. CPUID.01H:EDX.SSE2 should be 1 -- the capability is listed in the flags above. #NM should not matter. The kernel should just switch the context and kick the process to continue. Is there something in dmesg when this happens? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c12 Markus Elfring changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW InfoProvider|Markus.Elfring@web.de | --- Comment #12 from Markus Elfring 2011-03-17 10:53:51 UTC --- (In reply to comment #10) Yes. Sonne:/usr/src/linux-2.6.38.y # dmesg .. [ 1856.591497] fixdep[5047]: segfault at 6022d0 ip 00000000006022d0 sp 00007fff846555d8 error 15 in fixdep[602000+1000] -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c14 Jiri Slaby changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO InfoProvider| |Markus.Elfring@web.de --- Comment #14 from Jiri Slaby 2011-03-17 12:38:55 UTC --- (In reply to comment #12)

[ 1856.591497] fixdep[5047]: segfault at 6022d0 ip 00000000006022d0 sp 00007fff846555d8 error 15 in fixdep[602000+1000]

Ah, let's revoke what I wrote. error 15h (10101b) means: 1 The fault was caused by a page-level protection violation. 0 The access causing the fault was a read. 1 The access causing the fault originated when the processor was executing in user mode. 0 The fault was not caused by reserved bit violation. 1 The fault was caused by an instruction fetch. This means the code for strchr was mapped without X bit. Does your gdb already work? I need output of /proc/<PID>/maps obtained by: gdb --args ./scripts/basic/fixdep scripts/basic/.fixdep.d scripts/basic/fixdep 'gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/basic/fixdep scripts/basic/fixdep.c ' In GDB type: (gdb) run (gdb) info proc process SOME_PID In shell: $ cat /proc/SOME_PID/maps > maps And attach maps. (In reply to comment #13)

Which source files are required for the generation of the program "fixdep"?

scripts/basic/fixdep.c

I had no problems with the compilation of Linux 2.6.37.2 by the tool "GCC 4.5.2" on my own computer (hardware/physical machine). Was anything changed in the meantime that is relevant to the reported software situation?

What's the meantime? What versions do you have now? BTW while Richard is here -- is there a way how to dig out the shared libs map from a core file? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c16 Jiri Slaby changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO InfoProvider| |Markus.Elfring@web.de --- Comment #16 from Jiri Slaby 2011-03-17 19:21:44 UTC --- (In reply to comment #15)

Example: http://forums.opensuse.org/english/get-technical-help-here/install-boot-logi...

- I do not know at the moment when I will be able to update some remaining packages from the available repositories.

Is there a bug entry somewhere here? I cannot deduce anything from that example... I'm running 11.4 on several machine and have no problem with that.

- Will I need to install an alternative python version in my directory "/usr/local"?

Ok, probably the strace output would be enough. Try that instead of gdb --args.

Do compiler versions matter here?

It should not. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c18 Jiri Slaby changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO InfoProvider| |Markus.Elfring@web.de --- Comment #18 from Jiri Slaby 2011-03-18 08:43:39 UTC --- (In reply to comment #17)

How do you think about the following excerpt?

Unfortunately useless. I need the start and not of strace of make, but strace of fixdep. I.e.: strace -o fixdep_strace ./scripts/basic/fixdep scripts/basic/.fixdep.d scripts/basic/fixdep 'gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/basic/fixdep scripts/basic/fixdep.c ' and attach fixdep_strace.

Is the incomplete checking of return values a relevant implementation detail in the source file (e.g. function "do_config_file") for the reported situation?

I don't know. It might matter. Try to add the checking like: .. exit(2); } if (fstat(fd, &st)) { perror("fstat"); return; } if (st.st_size == 0) { .. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c Jiri Slaby changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO InfoProvider| |Markus.Elfring@web.de -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c23 --- Comment #23 from Markus Elfring 2011-03-25 13:13:56 UTC --- Created an attachment (id=421421) --> (http://bugzilla.novell.com/attachment.cgi?id=421421) small trace log This file is the result from the following command. Sonne:/usr/src/linux-2.6.38.y # strace -o /home/elfring/Projekte/fixdep_strace680120b.txt ./scripts/basic/fixdep scripts/basic/.fixdep.d scripts/basic/fixdep '/usr/bin/gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/basic/fixdep scripts/basic/fixdep.c' cmd_scripts/basic/fixdep := /usr/bin/gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/basic/fixdep scripts/basic/fixdep.c Segmentation fault (core dumped) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c25 --- Comment #25 from Markus Elfring 2011-03-29 14:27:09 UTC --- Created an attachment (id=421929) --> (http://bugzilla.novell.com/attachment.cgi?id=421929) valgrind log I can provide a bit more informations if I include debug data for the affected program explicitly. Sonne:/usr/src/linux-2.6.38.y # /usr/local/bin/gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -g -o scripts/basic/fixdep scripts/basic/fixdep.c Sonne:/usr/src/linux-2.6.38.y # valgrind --verbose --trace-children=yes --log-file=vg_log680120_%p.txt scripts/basic/fixdep scripts/basic/.fixdep.d scripts/basic/fixdep 'gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/basic/fixdep scripts/basic/fixdep.c' > scripts/basic/.fixdep.tmp Segmentation fault It seems that I can circumvent the unexpected behaviour with the following commands. Sonne:/usr/src/linux-2.6.38.y # /usr/bin/gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/basic/fixdep scripts/basic/fixdep.c Sonne:/usr/src/linux-2.6.38.y # scripts/basic/fixdep scripts/basic/.fixdep.d scripts/basic/fixdep 'gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/basic/fixdep scripts/basic/fixdep.c' > scripts/basic/.fixdep.tmp How do you think about this result? Why has the program that was generated by "my GCC 4.5.2" got problems? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120

https://bugzilla.novell.com/show_bug.cgi?id=680120#c27


--- Comment #27 from Markus Elfring  2011-03-29 14:46:21 UTC ---
(In reply to comment #26)

How do you think about the following result?

Sonne:/usr/src/linux-2.6.38.y # git pull
..
>From git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.38.y
   521cb40..cf6013b  master     -> origin/master
 * [new tag]         v2.6.38.2  -> v2.6.38.2
>From git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.38.y
 * [new tag]         v2.6.38.1  -> v2.6.38.1
Updating 521cb40..cf6013b
..
Sonne:/usr/src/linux-2.6.38.y # make CC:=/usr/bin/gcc V:=1 oldconfig
make -f scripts/Makefile.build obj=scripts/basic
  gcc -Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes
-Wstrict-prototypes -O2 -fomit-frame-pointer     -o scripts/basic/fixdep
scripts/basic/fixdep.c
/bin/sh: line 1:  2958 Segmentation fault      (core dumped)
scripts/basic/fixdep scripts/basic/.fixdep.d scripts/basic/fixdep 'gcc
-Wp,-MD,scripts/basic/.fixdep.d -Wall -Wmissing-prototypes -Wstrict-prototypes
-O2 -fomit-frame-pointer     -o scripts/basic/fixdep scripts/basic/fixdep.c  '
> scripts/basic/.fixdep.tmp
make[1]: *** [scripts/basic/fixdep] Error 139
make: *** [scripts_basic] Error 2

-- 
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c29 --- Comment #29 from Markus Elfring 2011-03-30 13:25:25 UTC --- (In reply to comment #28) Now I have installed a customised kernel 2.6.38.2 additionally that was generated by openSUSE's GCC. I find the reported errors strange since I could generate a Linux 2.6.37.2 with my GCC 4.5.2 (built on 2011-01-08) without problems on 2011-02-28. Can it be that similar issues might happen because of collaboration challenges with the software component "glibc (version 2.11.3-12.15.1)" since the recent distribution upgrade on my computer? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=680120 https://bugzilla.novell.com/show_bug.cgi?id=680120#c31 --- Comment #31 from Markus Elfring 2011-04-30 14:45:54 UTC --- (In reply to comment #28) Now I have installed a customised kernel 2.6.38.4 additionally that was generated by openSUSE's GCC. But I find the following reaction strange. Sonne:/usr/src/linux-2.6.38.y # git pull && make CC:=/usr/bin/gcc HOSTCC:=/usr/bin/gcc V:=1 xconfig Already up-to-date. make -f scripts/Makefile.build obj=scripts/basic rm -f .tmp_quiet_recordmcount mkdir -p include/linux include/config make -f scripts/Makefile.build obj=scripts/kconfig xconfig scripts/kconfig/qconf Kconfig make[1]: *** [xconfig] Segmentation fault make: *** [xconfig] Error 2 Sonne:/usr/src/linux-2.6.38.y # dmesg .. qconf[2346]: segfault at 63a5d0 ip 000000000063a5d0 sp 00007fff027c96d8 error 15 in qconf[638000+3000] Would you like to look into this situation? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.