http://bugzilla.suse.com/show_bug.cgi?id=1029318 http://bugzilla.suse.com/show_bug.cgi?id=1029318#c2 David Mair changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |IN_PROGRESS --- Comment #2 from David Mair --- I'm working on it. It is fair to call it a bug in crash (it should be fixed there) though it's a bug in the embedded gdb version in the crash source tarball (there's an upstream gdb patch that fixes it but has not been incorporated into the crash tarball upstream). That said, I understand the error became visible due to the C compiler being changed in the Factory base packages, note: not the compiler version number but the compiler model was changed. IOW, while it can be fixed by patching the gdb in the crash tarball, I believe it can also be fixed by reverting the C compiler in Factory to the one previously used. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=1029318 http://bugzilla.suse.com/show_bug.cgi?id=1029318#c4 --- Comment #4 from David Mair --- The intended fix to this problem report is the following: https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commitdiff;h=7546... Thanks to ptesarik for the link. Note the patch is a fix for gdb not crash and the crash source tarball contains a gdb source tarball of choice embedded and with it's own patches for application after the gdb source is expanded, i.e. source tarball for A contains source tarball for B and patches/builds it's own needs. The patching of Factory crash to fix crash's own gdb problems is a bit untidy and I wouldn't call my status finished. I want to run my approach by a friend with relevant experience for a review before I make the commit. There's a good chance it will be this week judging from the fact that the patch above is fully usable. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=1029318 http://bugzilla.suse.com/show_bug.cgi?id=1029318#c6 Michal Nowak changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |dmair@suse.com Flags| |needinfo?(dmair@suse.com) --- Comment #6 from Michal Nowak --- Is this related to this bug and the crash module, or is it something else? # echo c > /proc/sysrq-trigger ... [ 58.500686] Kernel panic - not syncing: Fatal exception [ 58.501643] Kernel Offset: 0x28000000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff) [ 58.503213] Rebooting in 90 seconds.. ... (More in https://openqa.opensuse.org/tests/382343/file/serial0.txt.) As a result vmcore was not created: https://openqa.opensuse.org/tests/382343#step/crash/63. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=1029318 http://bugzilla.suse.com/show_bug.cgi?id=1029318#c8 --- Comment #8 from David Mair --- (In reply to Michal Nowak from comment #6)

Is this related to this bug and the crash module, or is it something else?

# echo c > /proc/sysrq-trigger ... [ 58.500686] Kernel panic - not syncing: Fatal exception [ 58.501643] Kernel Offset: 0x28000000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff) [ 58.503213] Rebooting in 90 seconds.. ... (More in https://openqa.opensuse.org/tests/382343/file/serial0.txt.)

As a result vmcore was not created: https://openqa.opensuse.org/tests/382343#step/crash/63.

Taking my point about the (only case I found of the) crashing function during the coredump being in drivers/tty then I wonder if the problem is because you seem to have the serial console setup for the boot and crash kernel then the crash is handled by drivers/tty. Is there any chance the host has only one serial console and you didn't actually intend for it to be used as the console in one of the cases: crash kernel or coredump. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=1029318 http://bugzilla.suse.com/show_bug.cgi?id=1029318#c10 Michal Nowak changed: What |Removed |Added ---------------------------------------------------------------------------- Status|IN_PROGRESS |RESOLVED Resolution|--- |FIXED Flags|needinfo?(dmair@suse.com) | --- Comment #10 from Michal Nowak --- I seems to me that crash build fine in TW and openQA is not affected. I guess we can close this. -- You are receiving this mail because: You are on the CC list for the bug.