Summary: | sys-fs/btrfs-progs-3.12-r1 fails test: btrfsck should have repaired the image | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Markus Meier <maekke> |
Component: | Current packages | Assignee: | Mike Gilbert <floppym> |
Status: | RESOLVED OBSOLETE | ||
Severity: | normal | CC: | slyfox |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- |
Description
Markus Meier
2013-12-24 08:38:45 UTC
I guess it's a sheevaplug. It has very fun unaligned accesses handling by default. Can you set 'echo 3 > /proc/cpu/alignment' and see if it reports violations? Getting gdb backtrace would be sure awesome. (In reply to Markus Meier from comment #0) > that happens on arm and is a regression to current stable: > The current stable ebuild has no test suite, so I would not really call this a "regression" (unless you ran some test manually). (In reply to Sergei Trofimovich from comment #1) > I guess it's a sheevaplug. It has very fun unaligned accesses handling by > default. > > Can you set 'echo 3 > /proc/cpu/alignment' and see if it reports violations? > Getting gdb backtrace would be sure awesome. Yep the issue seems to be alignment specific, as setting 3 to /proc/cpu/alignment "fixes" the test suite. [12394727.502275] Alignment trap: btrfsck (7745) PC=0x0003a8f8 Instr=0xe1c120d0 Address=0xbec5853c FSR 0x001 [12394728.617529] Alignment trap: btrfsck (7751) PC=0x0003a8f8 Instr=0xe1c120d0 Address=0xbeb93b7c FSR 0x001 I'll try to debug this with gdb. (In reply to Mike Gilbert from comment #2) > The current stable ebuild has no test suite, so I would not really call this > a "regression" (unless you ran some test manually). Of course you're right! I didn't check the old ebuild log and assumed there was a test suite... sys-fs/btrfs-progs-3.12-r1 not in portage anymore and over 2,5 years in btrfs development is a lot. Marking as OBSOLETE. |