linux/fs, branch v4.4-rc3

Merge tag 'nfs-for-4.4-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

2015-11-28T01:22:47Z

Pull NFS client bugfixes from Trond Myklebust: "Highlights include: Stable patches: - Fix a NFSv4 callback identifier leak that was also causing client crashes - Fix NFSv4 callback decoding issues when incoming requests are truncated - Don't declare the attribute cache valid when we call nfs_update_inode with an empty attribute structure. - Resend LAYOUTGET when there is a race that changes the seqid Bugfixes: - Fix a number of issues with the NFSv4.2 CLONE ioctl() - Properly set NFS v4.2 NFSDBG_FACILITY - NFSv4 referrals are broken; Cleanup FATTR4_WORD0_FS_LOCATIONS after decoding success - Use sliding delay when LAYOUTGET gets NFS4ERR_DELAY - Ensure that attrcache is revalidated after a SETATTR" * tag 'nfs-for-4.4-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: nfs4: resend LAYOUTGET when there is a race that changes the seqid nfs: if we have no valid attrs, then don't declare the attribute cache valid nfs: ensure that attrcache is revalidated after a SETATTR nfs4: limit callback decoding to received bytes nfs4: start callback_ident at idr 1 nfs: use sliding delay when LAYOUTGET gets NFS4ERR_DELAY NFS4: Cleanup FATTR4_WORD0_FS_LOCATIONS after decoding success NFS: Properly set NFS v4.2 NFSDBG_FACILITY nfs: reduce the amount of ifdefs for v4.2 in nfs4file.c nfs: use btrfs ioctl defintions for clone nfs: allow intra-file CLONE nfs: offer native ioctls even if CONFIG_COMPAT is set nfs: pass on count for CLONE operations

Merge branch 'for-linus-4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs

2015-11-27T23:45:45Z

Pull btrfs fixes from Chris Mason: "This has Mark Fasheh's patches to fix quota accounting during subvol deletion, which we've been working on for a while now. The patch is pretty small but it's a key fix. Otherwise it's a random assortment" * 'for-linus-4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: btrfs: fix balance range usage filters in 4.4-rc btrfs: qgroup: account shared subtree during snapshot delete Btrfs: use btrfs_get_fs_root in resolve_indirect_ref btrfs: qgroup: fix quota disable during rescan Btrfs: fix race between cleaner kthread and space cache writeout Btrfs: fix scrub preventing unused block groups from being deleted Btrfs: fix race between scrub and block group deletion btrfs: fix rcu warning during device replace btrfs: Continue replace when set_block_ro failed btrfs: fix clashing number of the enhanced balance usage filter Btrfs: fix the number of transaction units needed to remove a block group Btrfs: use global reserve when deleting unused block group after ENOSPC Btrfs: tests: checking for NULL instead of IS_ERR() btrfs: fix signed overflows in btrfs_sync_file

nfs4: resend LAYOUTGET when there is a race that changes the seqid

2015-11-25T20:32:13Z

pnfs_layout_process will check the returned layout stateid against what the kernel has in-core. If it turns out that the stateid we received is older, then we should resend the LAYOUTGET instead of falling back to MDS I/O. Signed-off-by: Jeff Layton Cc: stable@vger.kernel.org # 3.18+ Signed-off-by: Trond Myklebust

nfs: if we have no valid attrs, then don't declare the attribute cache valid

2015-11-25T20:31:49Z

If we pass in an empty nfs_fattr struct to nfs_update_inode, it will (correctly) not update any of the attributes, but it then clears the NFS_INO_INVALID_ATTR flag, which indicates that the attributes are up to date. Don't clear the flag if the fattr struct has no valid attrs to apply. Reviewed-by: Steve French Signed-off-by: Jeff Layton Cc: stable@vger.kernel.org Signed-off-by: Trond Myklebust

nfs: ensure that attrcache is revalidated after a SETATTR

2015-11-25T20:24:30Z

If we get no post-op attributes back from a SETATTR operation, then no attributes will of course be updated during the call to nfs_update_inode. We know however that the attributes are invalid at that point, since we just changed some of them. At the very least, the ctime will be bogus. If we get no post-op attributes back on the call, mark the attrcache invalid to reflect that fact. Reviewed-by: Steve French Signed-off-by: Jeff Layton Signed-off-by: Trond Myklebust

btrfs: fix balance range usage filters in 4.4-rc

2015-11-25T13:27:33Z

There's a regression in 4.4-rc since commit bc3094673f22 (btrfs: extend balance filter usage to take minimum and maximum) in that existing (non-ranged) balance with -dusage=x no longer works; all chunks are skipped. After staring at the code for a while and wondering why a non-ranged balance would even need min and max thresholds (..which then were not set correctly, leading to the bug) I realized that the only problem was the fact that the filter functions were named wrong, thanks to patching copypasta. Simply renaming both functions lets the existing btrfs-progs call balance with -dusage=x and now the non-ranged filter function is invoked, properly using only a single chunk limit. Signed-off-by: Holger Hoffstätte Fixes: bc3094673f22 ("btrfs: extend balance filter usage to take minimum and maximum") Reviewed-by: Filipe Manana Signed-off-by: Chris Mason

btrfs: qgroup: account shared subtree during snapshot delete

2015-11-25T13:27:33Z

Commit 0ed4792 ('btrfs: qgroup: Switch to new extent-oriented qgroup mechanism.') removed our qgroup accounting during btrfs_drop_snapshot(). Predictably, this results in qgroup numbers going bad shortly after a snapshot is removed. Fix this by adding a dirty extent record when we encounter extents during our shared subtree walk. This effectively restores the functionality we had with the original shared subtree walking code in 1152651 (btrfs: qgroup: account shared subtrees during snapshot delete). The idea with the original patch (and this one) is that shared subtrees can get skipped during drop_snapshot. The shared subtree walk then allows us a chance to visit those extents and add them to the qgroup work for later processing. This ultimately makes the accounting for drop snapshot work. The new qgroup code nicely handles all the other extents during the tree walk via the ref dec/inc functions so we don't have to add actions beyond what we had originally. Signed-off-by: Mark Fasheh Signed-off-by: Chris Mason

Btrfs: use btrfs_get_fs_root in resolve_indirect_ref

2015-11-25T13:22:08Z

The backref code will look up the fs_root we're trying to resolve our indirect refs for, unfortunately we use btrfs_read_fs_root_no_name, which returns -ENOENT if the ref is 0. This isn't helpful for the qgroup stuff with snapshot delete as it won't be able to search down the snapshot we are deleting, which will cause us to miss roots. So use btrfs_get_fs_root and send false for check_ref so we can always get the root we're looking for. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Mark Fasheh Signed-off-by: Chris Mason

btrfs: qgroup: fix quota disable during rescan

2015-11-25T13:22:08Z

There's a race condition that leads to a NULL pointer dereference if you disable quotas while a quota rescan is running. To fix this, we just need to wait for the quota rescan worker to actually exit before tearing down the quota structures. Signed-off-by: Justin Maggard Signed-off-by: Chris Mason

Btrfs: fix race between cleaner kthread and space cache writeout

2015-11-25T13:22:08Z

When a block group becomes unused and the cleaner kthread is currently running, we can end up getting the current transaction aborted with error -ENOENT when we try to commit the transaction, leading to the following trace: [59779.258768] WARNING: CPU: 3 PID: 5990 at fs/btrfs/extent-tree.c:3740 btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs]() [59779.272594] BTRFS: Transaction aborted (error -2) (...) [59779.291137] Call Trace: [59779.291621] [] dump_stack+0x4e/0x79 [59779.292543] [] warn_slowpath_common+0x9f/0xb8 [59779.293435] [] ? btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs] [59779.295000] [] warn_slowpath_fmt+0x48/0x50 [59779.296138] [] ? write_one_cache_group.isra.32+0x77/0x82 [btrfs] [59779.297663] [] btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs] [59779.299141] [] commit_cowonly_roots+0x1de/0x261 [btrfs] [59779.300359] [] btrfs_commit_transaction+0x4c4/0x99c [btrfs] [59779.301805] [] btrfs_sync_fs+0x145/0x1ad [btrfs] [59779.302893] [] sync_filesystem+0x7f/0x93 (...) [59779.318186] ---[ end trace 577e2daff90da33a ]--- The following diagram illustrates a sequence of steps leading to this problem: CPU 1 CPU 2 adds bg A to list fs_info->unused_bgs adds bg B to list fs_info->unused_bgs cleaner kthread delete_unused_bgs() sees bg A in list fs_info->unused_bgs btrfs_start_transaction() deletes bg A update_block_group(bg C) --> adds bg C to list fs_info->unused_bgs deletes bg B sees bg C in the list fs_info->unused_bgs btrfs_remove_chunk(bg C) btrfs_remove_block_group(bg C) --> checks if the block group is in a dirty list, and because it isn't now, it does nothing --> the block group item is deleted from the extent tree --> adds bg C to list transaction->dirty_bgs some task calls btrfs_commit_transaction(t N + 1) commit_cowonly_roots() btrfs_write_dirty_block_groups() --> sees bg C in cur_trans->dirty_bgs --> calls write_one_cache_group() which returns -ENOENT because it did not find the block group item in the extent tree --> transaction aborte with -ENOENT because write_one_cache_group() returned that error So fix this by adding a block group to the list of dirty block groups before adding it to the list of unused block groups. This happened on a stress test using fsstress plus concurrent calls to fallocate 20G and truncate (releasing part of the space allocated with fallocate). Signed-off-by: Filipe Manana Signed-off-by: Chris Mason