Bug 62111
Summary: | Oops at __btrfs_map_block+0x2fa/0x1000 | ||
---|---|---|---|
Product: | File System | Reporter: | Josh Boyer (jwboyer) |
Component: | btrfs | Assignee: | Josef Bacik (josef) |
Status: | RESOLVED OBSOLETE | ||
Severity: | normal | CC: | dsterba, kentobias |
Priority: | P1 | ||
Hardware: | All | ||
OS: | Linux | ||
URL: | https://bugzilla.redhat.com/show_bug.cgi?id=1011724 | ||
Kernel Version: | 3.10.11-100.fc18.x86_64 | Subsystem: | |
Regression: | No | Bisected commit-id: |
Description
Josh Boyer
2013-09-25 16:59:47 UTC
The stacktrace looks very similar to what I see with the failing btrfs/011 test. Not exactly from the same sources: (gdb) l *(__btrfs_map_block+0x2fa) 0x56d8a is in __btrfs_map_block (fs/btrfs/volumes.c:4066). warning: Source file is more recent than executable. 4061 } 4062 4063 if (!device->in_fs_metadata || 4064 device->is_tgtdev_for_dev_replace) 4065 continue; 4066 4067 if (device->total_bytes > device->bytes_used) 4068 total_avail = device->total_bytes - device->bytes_used; 4069 else 4070 total_avail = 0; Can you please describe how to reproduce it? Based on the attached dmesg in the original RHT bug, I'd say the drive was really going bad or something else has caused the kernel to have difficulties talking to it. It's filled with: [ 2577.487831] Write(10): 2a 00 00 dc 79 01 00 00 01 00 [ 2577.522818] sd 5:0:0:0: [sdc] [ 2577.522830] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE [ 2577.522836] sd 5:0:0:0: [sdc] [ 2577.522841] Sense Key : Aborted Command [current] [ 2577.522848] sd 5:0:0:0: [sdc] [ 2577.522853] Add. Sense: No additional sense information [ 2577.522858] sd 5:0:0:0: [sdc] CDB: [ 2577.522861] Write(10): 2a 00 00 dc 79 01 00 00 01 00 [ 2577.559657] sd 5:0:0:0: [sdc] [ 2577.559669] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE [ 2577.559675] sd 5:0:0:0: [sdc] [ 2577.559680] Sense Key : Aborted Command [current] [ 2577.559687] sd 5:0:0:0: [sdc] [ 2577.559729] Add. Sense: No additional sense information [ 2577.559735] sd 5:0:0:0: [sdc] CDB: [ 2577.559738] Read(10): 28 00 00 dc 79 02 00 00 01 00 [ 2577.559968] btrfs: i/o error at logical 13637785821184 on dev /dev/sdc, sector 115591184, root 7927, inode 14257, offset 16842752, length 4096, links 1 (path: Music/The Plimsouls/The Plimsouls/02 - Now.flac) [ 2577.594771] sd 5:0:0:0: [sdc] [ 2577.594783] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE [ 2577.594790] sd 5:0:0:0: [sdc] [ 2577.594794] Sense Key : Aborted Command [current] [ 2577.594801] sd 5:0:0:0: [sdc] [ 2577.594806] Add. Sense: No additional sense information [ 2577.594811] sd 5:0:0:0: [sdc] CDB: [ 2577.594814] Write(10): 2a 00 00 dc 79 02 00 00 01 00 Hello. The drive was indeed going bad. It has since been replaced. Worked like a charm. I don't know if that means it's still a bug or not. Seems the code should handle the situation. For the record, it was a 4 disk USB3 btrfs RAID10 set with no partition tables (i.e. /dev/sdc not /dev/sdc1). Thanks for your help and time. This is a semi-automated bugzilla cleanup, report is against an old kernel version. If the problem still happens, please open a new bug. Thanks. |