Bug 15496
| Summary: | 2.6.34-rc1 crashes early while startup | ||
|---|---|---|---|
| Product: | Platform Specific/Hardware | Reporter: | Thomas Meyer (thomas) |
| Component: | i386 | Assignee: | platform_i386 |
| Status: | CLOSED CODE_FIX | ||
| Severity: | normal | CC: | akpm, maciej.rutecki, rjw |
| Priority: | P1 | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Kernel Version: | 2.6.34-rc1 | Subsystem: | |
| Regression: | Yes | Bisected commit-id: | |
| Bug Depends on: | |||
| Bug Blocks: | 15310 | ||
| Attachments: |
config for 2.6.34-rc1
boot log with option "vga=0xf06" screenshot "normal boot" kobject is not initialised warning, then OOPS |
||
Created attachment 25435 [details]
boot log with option "vga=0xf06"
Created attachment 25436 [details]
screenshot "normal boot"
Should block metabug 15310. https://bugzilla.kernel.org/show_bug.cgi?id=15310 Commit a3d3203e4bb40f253b1541e310dc0f9305be7c84 still fails to boot with same config. We really need to see the first screenful of that crash output so we can see where it crashed, please. Perhaps you can play with the "boot_delay" kernel boot parameter to slow the output down, then photograph it just before the oops starts to scroll off the screen? Perhaps you could also explore the earlyprintk support (Documentation/kernel-parameters.txt) or netconsole (Documentation/networking/netconsole.txt). Thanks. (In reply to comment #5) > We really need to see the first screenful of that crash output so we can see > where it crashed, please. > > Perhaps you can play with the "boot_delay" kernel boot parameter to slow the > output down, then photograph it just before the oops starts to scroll off the > screen? > > Perhaps you could also explore the earlyprintk support > (Documentation/kernel-parameters.txt) or netconsole > (Documentation/networking/netconsole.txt). eth0 seems to get initializied later, so no console log via netconsole. the parameter boot_delay (e.g. boot_delay=4000) has no influence on the output speed of the vga console! Created attachment 25631 [details]
kobject is not initialised warning, then OOPS
this patch seems to fix the oops: http://lkml.org/lkml/2010/3/26/41 May the bugs https://bugzilla.kernel.org/show_bug.cgi?id=15600 https://bugzilla.kernel.org/show_bug.cgi?id=15518 are also fixed by above patch? Handled-By : Yinghai Lu <yinghai@kernel.org> Patch : https://patchwork.kernel.org/patch/88437/ It's possible that these bugs are also fixed by the patch, but it would be good to verify that. Fixed by commit c967da6a0ba837f762042e931d4afcf72045547c . |
Created attachment 25433 [details] config for 2.6.34-rc1 Screenshots to follow. config attached.