I want to report a btrfs bug that happened this morning on kernel 3.2.1. I was copying a +5 GB file between two external usb harddisks, one formated with btrfs (the source) and the other with ext4 (the destination).
In the meantime I decided to compile the 3.3-rc1 kernel with the output dir on the ext4 external harddisk. After some time I started to get many I/O errors coming from the ext4 drive (sdc1) on dmesg, e.g.: Buffer I/O error on device sdc1, logical block 3084285 After that it looks like the ext4 HD was unmounted automatically and the btrfs HD made this BUG: usb 2-1: USB disconnect, device number 2 EXT4-fs error (device sdc1): ext4_find_entry:935: inode #109051908: comm bash: reading directory lblock 0 EXT4-fs error (device sdc1): ext4_find_entry:935: inode #109051905: comm bash: reading directory lblock 0 EXT4-fs error (device sdc1): ext4_find_entry:935: inode #2: comm umount: reading directory lblock 0 lost page write due to I/O error on sdb1 lost page write due to I/O error on sdb1 lost page write due to I/O error on sdb1 btrfs: 1 errors while writing supers ------------[ cut here ]------------ kernel BUG at /home/mafra/linux-2.6/fs/btrfs/disk-io.c:2835! invalid opcode: 0000 [#1] PREEMPT SMP CPU 2 Modules linked in: loop af_packet snd_pcm_oss snd_mixer_oss microcode arc4 sr_mod cdrom sg ath9k mac80211 uvcvideo i2c_i801 ath9k_common ath9k_hw ath cfg80211 sony_laptop rfkill radeon ttm drm_kms_helper drm i2c_algo_bit Pid: 17579, comm: umount Not tainted 3.2.1+ #68 Sony Corporation VPCEB4X1E/VAIO RIP: 0010:[<ffffffff811b4689>] [<ffffffff811b4689>] write_all_supers+0x879/0x880 RSP: 0018:ffff88011198dca8 EFLAGS: 00010292 RAX: 000000000000003a RBX: ffff8800b4b378b8 RCX: 00000000000010a5 RDX: 00000000000004c2 RSI: 0000000000000046 RDI: ffffffff81b0c880 RBP: ffff88011198dd08 R08: 0000000000000000 R09: 0000000000000000 R10: 000000000000000a R11: 0000000000000001 R12: ffff88011198dcd0 R13: ffff8800b4b378b8 R14: ffff8800acf88000 R15: ffff8801046a7400 FS: 00007f89586087e0(0000) GS:ffff880117c80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00007f8957d47fc0 CR3: 000000011199e000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process umount (pid: 17579, threadinfo ffff88011198c000, task ffff880106eb4940) Stack: ffff8800b4b37840 0000000000000000 0000000100000000 ffff880100000001 0000000000000001 ffff8800b4b378b8 ffff88011198dd08 ffff88011263fe80 ffff8801046a7400 ffff880105618540 ffff88011198dd78 ffff88011263fef8 Call Trace: [<ffffffff811b469e>] write_ctree_super+0xe/0x10 [<ffffffff811b9778>] btrfs_commit_transaction+0x6a8/0x870 [<ffffffff811b8a8a>] ? join_transaction+0x24a/0x280 [<ffffffff810611f0>] ? wake_up_bit+0x40/0x40 [<ffffffff81195a57>] btrfs_sync_fs+0x47/0x70 [<ffffffff811029be>] __sync_filesystem+0x5e/0x90 [<ffffffff81102a53>] sync_filesystem+0x43/0x60 [<ffffffff810dc226>] generic_shutdown_super+0x36/0xe0 [<ffffffff810dc361>] kill_anon_super+0x11/0x20 [<ffffffff810dc715>] deactivate_locked_super+0x45/0x80 [<ffffffff810dd115>] deactivate_super+0x45/0x60 [<ffffffff810f753d>] mntput_no_expire+0x9d/0xf0 [<ffffffff810f837a>] sys_umount+0x6a/0x380 [<ffffffff814e6e7b>] system_call_fastpath+0x16/0x1b Code: 33 00 e9 7e fd ff ff 44 89 ce 48 c7 c7 80 a4 78 81 31 c0 e8 04 ea 32 00 0f 0b 8b 75 b8 48 c7 c7 80 a4 78 81 31 c0 e8 f1 e9 32 00 <0f> 0b 0f 1f 44 00 00 55 48 89 f7 89 d6 48 89 e5 e8 72 f7 ff ff RIP [<ffffffff811b4689>] write_all_supers+0x879/0x880 RSP <ffff88011198dca8> ---[ end trace 37039e5aca9c365c ]--- SysRq : Emergency Remount R/O I will move to testing 3.3-rc1 now, is this bug already fixed there? -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html