devzero@xxxxxx wrote: > Hello, > > i was testing btrfs a little bit for stability and i think i have found an issue. > > during some heavy activity (multiple processes writing/reading to compressed /btrfs at the same time), i did some snapshots in a loop (every 10 seconds). > > very soon, all filesystem activity stopped, i.e. i cannot read/write anyting from/to /btrfs anymore. > this happens with latest git-unstable with kernel 2.6.27 > nothing in dmesg. > > besides that, the system seems ok. > > as the system is a vmware virtual machine, i did a snapshot of the system state, so we can always go back to this state. > > if somebody want`s to analyse this, please tell me what to do. > Thank you for reporting this. Please try the attached patch. Yan Zheng --- diff -urp 1/fs/btrfs/transaction.c 2/fs/btrfs/transaction.c --- 1/fs/btrfs/transaction.c 2008-12-13 12:35:29.487886730 +0800 +++ 2/fs/btrfs/transaction.c 2008-12-21 19:09:09.000000000 +0800 @@ -804,7 +804,7 @@ static noinline int finish_pending_snaps parent_inode = pending->dentry->d_parent->d_inode; parent_root = BTRFS_I(parent_inode)->root; - trans = btrfs_start_transaction(parent_root, 1); + trans = btrfs_join_transaction(parent_root, 1); /* * insert the directory item -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
