- 03 Aug, 2016 30 commits
-
-
Darrick J. Wong authored
Now that the generic btree code supports overlapping intervals, plug in the rmap btree to this functionality. We will need it to find potential left neighbors in xfs_rmap_{alloc,free} later in the patch set. Signed-off-by:Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by:
Dave Chinner <dchinner@redhat.com> Signed-off-by:
Dave Chinner <david@fromorbit.com>
-
Darrick J. Wong authored
Originally-From: Dave Chinner <dchinner@redhat.com> Implement the generic btree operations needed to manipulate rmap btree blocks. This is very similar to the per-ag freespace btree implementation, and uses the AGFL for allocation and freeing of blocks. Adapt the rmap btree to store owner offsets within each rmap record, and to handle the primary key being redefined as the tuple [agblk, owner, offset]. The expansion of the primary key is crucial to allowing multiple owners per extent. [darrick: adapt the btree ops to deal with offsets] [darrick: remove init_rec_from_key] [darrick: move unwritten bit to rm_offset] Signed-off-by:
-
Darrick J. Wong authored
Originally-From: Dave Chinner <dchinner@redhat.com> The rmap btree is allocated from the AGFL, which means we have to ensure ENOSPC is reported to userspace before we run out of free space in each AG. The last allocation in an AG can cause a full height rmap btree split, and that means we have to reserve at least this many blocks *in each AG* to be placed on the AGFL at ENOSPC. Update the various space calculation functions to handle this. Also, because the macros are now executing conditional code and are called quite frequently, convert them to functions that initialise variables in the struct xfs_mount, use the new variables everywhere and document the calculations better. [darrick.wong@oracle.com: don't reserve blocks if !rmap] [dchinner@redhat.com: update m_ag_max_usable after growfs] Signed-off-by:
-
Darrick J. Wong authored
The rmap btrees will use the AGFL as the block allocation source, so we need to ensure that the transaction reservations reflect the fact this tree is modified by allocation and freeing. Hence we need to extend all the extent allocation/free reservations used in transactions to handle this. Note that this also gets rid of the unused XFS_ALLOCFREE_LOG_RES macro, as we now do buffer reservations based on the number of buffers logged via xfs_calc_buf_res(). Hence we only need the buffer count calculation now. [darrick: use rmap_maxlevels when calculating log block resv] Signed-off-by:
Brian Foster <bfoster@redhat.com> Signed-off-by:
-
Darrick J. Wong authored
Originally-From: Dave Chinner <dchinner@redhat.com> Now we can read and write rmap btree blocks, we can add support to the growfs code to initialise new rmap btree blocks. [darrick.wong@oracle.com: fill out the rmap offset fields] Signed-off-by:
-
Darrick J. Wong authored
Originally-From: Dave Chinner <dchinner@redhat.com> Now we have all the surrounding call infrastructure in place, we can start filling out the rmap btree implementation. Start with the on-disk btree format; add everything needed to read, write and manipulate rmap btree blocks. This prepares the way for adding the btree operations implementation. [darrick: record owner and offset info in rmap btree] [darrick: fork, bmbt and unwritten state in rmap btree] [darrick: flags are a separate field in xfs_rmap_irec] [darrick: calculate maxlevels separately] [darrick: move the 'unwritten' bit into unused parts of rm_offset] Signed-off-by:
-
Darrick J. Wong authored
Originally-From: Dave Chinner <dchinner@redhat.com> Add the stubs into the extent allocation and freeing paths that the rmap btree implementation will hook into. While doing this, add the trace points that will be used to track rmap btree extent manipulations. [darrick.wong@oracle.com: Extend the stubs to take full owner info.] Signed-off-by:
-
Darrick J. Wong authored
For the rmap btree to work, we have to feed the extent owner information to the the allocation and freeing functions. This information is what will end up in the rmap btree that tracks allocated extents. While we technically don't need the owner information when freeing extents, passing it allows us to validate that the extent we are removing from the rmap btree actually belonged to the owner we expected it to belong to. We also define a special set of owner values for internal metadata that would otherwise have no owner. This allows us to tell the difference between metadata owned by different per-ag btrees, as well as static fs metadata (e.g. AG headers) and internal journal blocks. There are also a couple of special cases we need to take care of - during EFI recovery, we don't actually know who the original owner was, so we need to pass a wildcard to indicate that we aren't checking the owner for validity. We also need special handling in growfs, as we "free" the space in the last AG when extending it, but because it's new space it has no actual owner... While touching the xfs_bmap_add_free() function, re-order the parameters to put the struct xfs_mount first. Extend the owner field to include both the owner type and some sort of index within the owner. The index field will be used to support reverse mappings when reflink is enabled. When we're freeing extents from an EFI, we don't have the owner information available (rmap updates have their own redo items). xfs_free_extent therefore doesn't need to do an rmap update. Make sure that the log replay code signals this correctly. This is based upon a patch originally from Dave Chinner. It has been extended to add more owner information with the intent of helping recovery operations when things go wrong (e.g. offset of user data block in a file). [dchinner: de-shout the xfs_rmap_*_owner helpers] [darrick: minor style fixes suggested by Christoph Hellwig] Signed-off-by:
-
Darrick J. Wong authored
Originally-From: Dave Chinner <dchinner@redhat.com> XFS reserves a small amount of space in each AG for the minimum number of free blocks needed for operation. Adding the rmap btree increases the number of reserved blocks, but it also increases the complexity of the calculation as the free inode btree is optional (like the rmbt). Rather than calculate the prealloc blocks every time we need to check it, add a function to calculate it at mount time and store it in the struct xfs_mount, and convert the XFS_PREALLOC_BLOCKS macro just to use the xfs-mount variable directly. Signed-off-by:
-
Darrick J. Wong authored
Originally-From: Dave Chinner <dchinner@redhat.com> The rmap btree will require the same stats as all the other generic btrees, so add all the code for that now. Signed-off-by:
-
Darrick J. Wong authored
Originally-From: Dave Chinner <dchinner@redhat.com> Add new per-ag rmap btree definitions to the per-ag structures. The rmap btree will sit in the empty slots on disk after the free space btrees, and hence form a part of the array of space management btrees. This requires the definition of the btree to be contiguous with the free space btrees. Signed-off-by:
-
Darrick J. Wong authored
By my calculations, a 1,073,741,824 block AG with a 1k block size can attain a maximum height of 9. Assuming a record size of 24 bytes, a key/ptr size of 44 bytes, and half-full btree nodes, we'd need 53,687,092 blocks for the records and ~6 million blocks for the keys. That requires a btree of height 9 based on the following derivation: Block size = 1024b sblock CRC header = 56b == 1024-56 = 968 bytes for tree data rmapbt record = 24b == 40 records per leaf block rmapbt ptr/key = 44b == 22 ptr/keys per block Worst case, each block is half full, so 20 records and 11 ptrs per block. 1073741824 rmap records / 20 records per block == 53687092 leaf blocks 53687092 leaves / 11 ptrs per block == 4880645 level 1 blocks == 443695 level 2 blocks == 40336 level 3 blocks == 3667 level 4 blocks == 334 level 5 blocks == 31 level 6 blocks == 3 level 7 blocks == 1 level 8 block Signed-off-by:
-
Darrick J. Wong authored
Add a couple of tracepoints for the deferred extent free operation and a site for injecting errors while finishing the operation. This makes it easier to debug deferred ops and test log redo. Signed-off-by:
-
Darrick J. Wong authored
Refactor the EFI intent item recovery (and cancellation) functions into a general function that scans the AIL and an intent item type specific handler. Move the function that recovers a single EFI item into the extent free item code. We'll want the generalized function when we start wiring up more redo item types. Furthermore, ensure that log recovery only replays the redo items that were in the AIL prior to recovery by checking the item LSN against the largest LSN seen during log scanning. As written this should never happen, but we can be defensive anyway. Signed-off-by:
-
Darrick J. Wong authored
Mechanical change of flist/free_list to dfops, since they're now deferred ops, not just a freeing list. Signed-off-by:
-
Darrick J. Wong authored
Drop the compatibility shims that we were using to integrate the new deferred operation mechanism into the existing code. No new code. Signed-off-by:
-
Darrick J. Wong authored
Restructure everything that used xfs_bmap_free to use xfs_defer_ops instead. For now we'll just remove the old symbols and play some cpp magic to make it work; in the next patch we'll actually rename everything. Signed-off-by:
-
Darrick J. Wong authored
Connect the xfs_defer mechanism with the pieces that we'll need to handle deferred extent freeing. We'll wire up the existing code to our new deferred mechanism later. Signed-off-by:
-
Darrick J. Wong authored
Replace structure typedefs with struct xfs_foo_* in the EFI/EFD handling code in preparation to move it over to deferred ops. Signed-off-by:
Christoph Hellwig <hch@lst.de> Signed-off-by:
-
Darrick J. Wong authored
Add tracepoints for the internals of the deferred ops mechanism and tracepoint classes for clients of the dops, to make debugging easier. Signed-off-by:
-
Darrick J. Wong authored
All the code around struct xfs_bmap_free basically implements a deferred operation framework through which we can roll transactions (to unlock buffers and avoid violating lock order rules) while managing all the necessary log redo items. Previously we only used this code to free extents after some sort of mapping operation, but with the advent of rmap and reflink, we suddenly need to do more than that. With that in mind, xfs_bmap_free really becomes a deferred ops control structure. Rename the structure and move the deferred ops into their own file to avoid further bloating of the bmap code. Signed-off-by:
-
Darrick J. Wong authored
Refactor the btree_change_owner function into a more generic apparatus which visits all blocks in a btree. We'll use this in a subsequent patch for counting btree blocks for AG reservations. Signed-off-by:
-
Darrick J. Wong authored
Create a function to enable querying of btree records mapping to a range of keys. This will be used in subsequent patches to allow querying the reverse mapping btree to find the extents mapped to a range of physical blocks, though the generic code can be used for any range query. The overlapped query range function needs to use the btree get_block helper because the root block could be an inode, in which case bc_bufs[nlevels-1] will be NULL. Signed-off-by:
-
Darrick J. Wong authored
On a filesystem with both reflink and reverse mapping enabled, it's possible to have multiple rmap records referring to the same blocks on disk. When overlapping intervals are possible, querying a classic btree to find all records intersecting a given interval is inefficient because we cannot use the left side of the search interval to filter out non-matching records the same way that we can use the existing btree key to filter out records coming after the right side of the search interval. This will become important once we want to use the rmap btree to rebuild BMBTs, or implement the (future) fsmap ioctl. (For the non-overlapping case, we can perform such queries trivially by starting at the left side of the interval and walking the tree until we pass the right side.) Therefore, extend the btree code to come closer to supporting intervals as a first-class record attribute. This involves widening the btree node's key space to store both the lowest key reachable via the node pointer (as the btree does now) and the highest key reachable via the same pointer and teaching the btree modifying functions to keep the highest-key records up to date. This behavior can be turned on via a new btree ops flag so that btrees that cannot store overlapping intervals don't pay the overhead costs in terms of extra code and disk format changes. When we're deleting a record in a btree that supports overlapped interval records and the deletion results in two btree blocks being joined, we defer updating the high/low keys until after all possible joining (at higher levels in the tree) have finished. At this point, the btree pointers at all levels have been updated to remove the empty blocks and we can update the low and high keys. When we're doing this, we must be careful to update the keys of all node pointers up to the root instead of stopping at the first set of keys that don't need updating. This is because it's possible for a single deletion to cause joining of multiple levels of tree, and so we need to update everything going back to the root. The diff_two_keys functions return < 0, 0, or > 0 if key1 is less than, equal to, or greater than key2, respectively. This is consistent with the rest of the kernel and the C library. In btree_updkeys(), we need to evaluate the force_all parameter before running the key diff to avoid reading uninitialized memory when we're forcing a key update. This happens when we've allocated an empty slot at level N + 1 to point to a new block at level N and we're in the process of filling out the new keys. Signed-off-by:
-
Darrick J. Wong authored
Add some function pointers to bc_ops to get the btree keys for leaf and node blocks, and to update parent keys of a block. Convert the _btree_updkey calls to use our new pointer, and modify the tree shape changing code to call the appropriate get_*_keys pointer instead of _btree_copy_keys because the overlapping btree has to calculate high key values. Signed-off-by:
-
Darrick J. Wong authored
When a btree block has to be split, we pass the new block's ptr from xfs_btree_split() back to xfs_btree_insert() via a pointer parameter; however, we pass the block's key through the cursor's record. It is a little weird to "initialize" a record from a key since the non-key attributes will have garbage values. When we go to add support for interval queries, we have to be able to pass the lowest and highest keys accessible via a pointer. There's no clean way to pass this back through the cursor's record field. Therefore, pass the key directly back to xfs_btree_insert() the same way that we pass the btree_ptr. As a bonus, we no longer need init_rec_from_key and can drop it from the codebase. Signed-off-by:
-
Darrick J. Wong authored
If we make the inode root block of a btree unfull by expanding the root, we must set *stat to 1 to signal success, rather than leaving it uninitialized. Signed-off-by:
-
Darrick J. Wong authored
When we're deleting realtime extents, we need to lock the summary inode in case we need to update the summary info to prevent an assert on the rsumip inode lock on a debug kernel. While we're at it, fix the locking annotations so that we avoid triggering lockdep warnings. Signed-off-by:
-
Darrick J. Wong authored
Apparently cris doesn't require structure stride to align with the largest type in the struct, so list[0] isn't at offset 4 like it is everywhere else. Fix this... insofar as existing XFSes on cris are screwed. Signed-off-by:
-
Darrick J. Wong authored
When we're iterating inode xattrs by handle, we have to copy the cursor back to userspace so that a subsequent invocation actually retrieves subsequent contents. Signed-off-by:
-
- 22 Jul, 2016 2 commits
-
-
Dave Chinner authored
-
Dave Chinner authored
Been around for long enough now, hasn't caused any regression test failures in the past 3 months, so it's time to make it a fully supported feature. Signed-off-by:
Eric Sandeen <sandeen@redhat.com> Signed-off-by:
-
- 21 Jul, 2016 5 commits
-
-
Dave Chinner authored
In xfs_finish_page_writeback(), we have a loop that looks like this: do { if (off < bvec->bv_offset) goto next_bh; if (off > end) break; bh->b_end_io(bh, !error); next_bh: off += bh->b_size; } while ((bh = bh->b_this_page) != head); The b_end_io function is end_buffer_async_write(), which will call end_page_writeback() once all the buffers have marked as no longer under IO. This issue here is that the only thing currently protecting both the bufferhead chain and the page from being reclaimed is the PageWriteback state held on the page. While we attempt to limit the loop to just the buffers covered by the IO, we still read from the buffer size and follow the next pointer in the bufferhead chain. There is no guarantee that either of these are valid after the PageWriteback flag has been cleared. Hence, loops like this are completely unsafe, and result in use-after-free issues. One such problem was caught by Calvin Owens with KASAN: ..... INFO: Freed in 0x103fc80ec age=18446651500051355200 cpu=2165122683 pid=-1 free_buffer_head+0x41/0x90 __slab_free+0x1ed/0x340 kmem_cache_free+0x270/0x300 free_buffer_head+0x41/0x90 try_to_free_buffers+0x171/0x240 xfs_vm_releasepage+0xcb/0x3b0 try_to_release_page+0x106/0x190 shrink_page_list+0x118e/0x1a10 shrink_inactive_list+0x42c/0xdf0 shrink_zone_memcg+0xa09/0xfa0 shrink_zone+0x2c3/0xbc0 ..... Call Trace: <IRQ> [<ffffffff81e8b8e4>] dump_stack+0x68/0x94 [<ffffffff8153a995>] print_trailer+0x115/0x1a0 [<ffffffff81541174>] object_err+0x34/0x40 [<ffffffff815436e7>] kasan_report_error+0x217/0x530 [<ffffffff81543b33>] __asan_report_load8_noabort+0x43/0x50 [<ffffffff819d651f>] xfs_destroy_ioend+0x3bf/0x4c0 [<ffffffff819d69d4>] xfs_end_bio+0x154/0x220 [<ffffffff81de0c58>] bio_endio+0x158/0x1b0 [<ffffffff81dff61b>] blk_update_request+0x18b/0xb80 [<ffffffff821baf57>] scsi_end_request+0x97/0x5a0 [<ffffffff821c5558>] scsi_io_completion+0x438/0x1690 [<ffffffff821a8d95>] scsi_finish_command+0x375/0x4e0 [<ffffffff821c3940>] scsi_softirq_done+0x280/0x340 Where the access is occuring during IO completion after the buffer had been freed from direct memory reclaim. Prevent use-after-free accidents in this end_io processing loop by pre-calculating the loop conditionals before calling bh->b_end_io(). The loop is already limited to just the bufferheads covered by the IO in progress, so the offset checks are sufficient to prevent accessing buffers in the chain after end_page_writeback() has been called by the the bh->b_end_io() callout. Yet another example of why Bufferheads Must Die. cc: <stable@vger.kernel.org> # 4.7 Signed-off-by:Calvin Owens <calvinowens@fb.com> Reviewed-by:
-
Dave Chinner authored
One of the problems we currently have with delayed logging is that under serious memory pressure we can deadlock memory reclaim. THis occurs when memory reclaim (such as run by kswapd) is reclaiming XFS inodes and issues a log force to unpin inodes that are dirty in the CIL. The CIL is pushed, but this will only occur once it gets the CIL context lock to ensure that all committing transactions are complete and no new transactions start being committed to the CIL while the push switches to a new context. The deadlock occurs when the CIL context lock is held by a committing process that is doing memory allocation for log vector buffers, and that allocation is then blocked on memory reclaim making progress. Memory reclaim, however, is blocked waiting for a log force to make progress, and so we effectively deadlock at this point. To solve this problem, we have to move the CIL log vector buffer allocation outside of the context lock so that memory reclaim can always make progress when it needs to force the log. The problem with doing this is that a CIL push can take place while we are determining if we need to allocate a new log vector buffer for an item and hence the current log vector may go away without warning. That means we canot rely on the existing log vector being present when we finally grab the context lock and so we must have a replacement buffer ready to go at all times. To ensure this, introduce a "shadow log vector" buffer that is always guaranteed to be present when we gain the CIL context lock and format the item. This shadow buffer may or may not be used during the formatting, but if the log item does not have an existing log vector buffer or that buffer is too small for the new modifications, we swap it for the new shadow buffer and format the modifications into that new log vector buffer. The result of this is that for any object we modify more than once in a given CIL checkpoint, we double the memory required to track dirty regions in the log. For single modifications then we consume the shadow log vectorwe allocate on commit, and that gets consumed by the checkpoint. However, if we make multiple modifications, then the second transaction commit will allocate a shadow log vector and hence we will end up with double the memory usage as only one of the log vectors is consumed by the CIL checkpoint. The remaining shadow vector will be freed when th elog item is freed. This can probably be optimised in future - access to the shadow log vector is serialised by the object lock (as opposited to the active log vector, which is controlled by the CIL context lock) and so we can probably free shadow log vector from some objects when the log item is marked clean on removal from the AIL. Signed-off-by:
-
Dave Chinner authored
xfsprogs source commit 4280e59dcbc4cd8e01585efe788a68eb378048e8 xfs_da3_split() has to handle all three versions of the directory/attribute btree structure. The attr tree is v1, the dir tre is v2 or v3. The main difference between the v1 and v2/3 trees is the way tree nodes are split - in the v1 tree we can require a double split to occur because the object to be inserted may be larger than the space made by splitting a leaf. In this case we need to do a double split - one to split the full leaf, then another to allocate an empty leaf block in the correct location for the new entry. This does not happen with dir (v2/v3) formats as the objects being inserted are always guaranteed to fit into the new space in the split blocks. Indeed, for directories they *may* be an extra block on this buffer pointer. However, it's guaranteed not to be a leaf block (i.e. a directory data block) - the directory code only ever places hash index or free space blocks in this pointer (as a cursor of sorts), and so to use it as a directory data block will immediately corrupt the directory. The problem is that the code assumes that there may be extra blocks that we need to link into the tree once we've split the root, but this is not true for either dir or attr trees, because the extra attr block is always consumed by the last node split before we split the root. Hence the linking in an extra block is always wrong at the root split level, and this manifests itself in repair as a directory corruption in a repaired directory, leaving the directory rebuild incomplete. This is a dir v2 zero-day bug - it was in the initial dir v2 commit that was made back in February 1998. Fix this by ensuring the linking of the blocks after the root split never tries to make use of the extra blocks that may be held in the cursor. They are held there for other purposes and should never be touched by the root splitting code. Signed-off-by:
-
Arnd Bergmann authored
We check IS_DAX(inode) before calling either xfs_file_dax_read or xfs_file_dax_write, and this will lead the call being optimized out at compile time when CONFIG_FS_DAX is disabled. However, the two functions are marked STATIC, so they become global symbols when CONFIG_XFS_DEBUG is set, leaving us with two unused global functions that call into an undefined function and a broken "allmodconfig" build: fs/built-in.o: In function `xfs_file_dax_read': fs/xfs/xfs_file.c:348: undefined reference to `dax_do_io' fs/built-in.o: In function `xfs_file_dax_write': fs/xfs/xfs_file.c:758: undefined reference to `dax_do_io' Marking the two functions 'static noinline' instead of 'STATIC' will let the compiler drop the symbols when there are no callers but avoid the implicit inlining. Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Fixes: 16d4d435 ("xfs: split direct I/O and DAX path") Reviewed-by:
-
Brian Foster authored
XFS has had scattered reports of delalloc blocks present at ->releasepage() time. This results in a warning with a stack trace similar to the following: ... Call Trace: [<ffffffffa23c5b8f>] dump_stack+0x63/0x84 [<ffffffffa20837a7>] warn_slowpath_common+0x97/0xe0 [<ffffffffa208380a>] warn_slowpath_null+0x1a/0x20 [<ffffffffa2326caf>] xfs_vm_releasepage+0x10f/0x140 [<ffffffffa218c680>] ? page_mkclean_one+0xd0/0xd0 [<ffffffffa218d3a0>] ? anon_vma_prepare+0x150/0x150 [<ffffffffa21521c2>] try_to_release_page+0x32/0x50 [<ffffffffa2166b2e>] shrink_active_list+0x3ce/0x3e0 [<ffffffffa21671c7>] shrink_lruvec+0x687/0x7d0 [<ffffffffa21673ec>] shrink_zone+0xdc/0x2c0 [<ffffffffa2168539>] kswapd+0x4f9/0x970 [<ffffffffa2168040>] ? mem_cgroup_shrink_node_zone+0x1a0/0x1a0 [<ffffffffa20a0d99>] kthread+0xc9/0xe0 [<ffffffffa20a0cd0>] ? kthread_stop+0x100/0x100 [<ffffffffa26b404f>] ret_from_fork+0x3f/0x70 [<ffffffffa20a0cd0>] ? kthread_stop+0x100/0x100 This occurs because it is possible for shrink_active_list() to send pages marked dirty to ->releasepage() when certain buffer_head threshold conditions are met. shrink_active_list() doesn't check the page dirty state apparently to handle an old ext3 corner case where in some cases clean pages would not have the dirty bit cleared, thus it is up to the filesystem to determine how to handle the page. XFS currently handles the delalloc case properly, but this behavior makes the warning spurious. Update the XFS ->releasepage() handler to explicitly skip dirty pages. Retain the existing delalloc/unwritten checks so we continue to warn if such buffers exist on clean pages when they shouldn't. Diagnosed-by:
-
- 20 Jul, 2016 3 commits
-
-
Dave Chinner authored
-
Dave Chinner authored
-
Dave Chinner authored
-
