From 22e757a49cf010703fcb9c9b4ef793248c39b0c2 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Tue, 2 Sep 2014 12:12:51 +1000 Subject: [PATCH 1/7] xfs: don't dirty buffers beyond EOF generic/263 is failing fsx at this point with a page spanning EOF that cannot be invalidated. The operations are: 1190 mapwrite 0x52c00 thru 0x5e569 (0xb96a bytes) 1191 mapread 0x5c000 thru 0x5d636 (0x1637 bytes) 1192 write 0x5b600 thru 0x771ff (0x1bc00 bytes) where 1190 extents EOF from 0x54000 to 0x5e569. When the direct IO write attempts to invalidate the cached page over this range, it fails with -EBUSY and so any attempt to do page invalidation fails. The real question is this: Why can't that page be invalidated after it has been written to disk and cleaned? Well, there's data on the first two buffers in the page (1k block size, 4k page), but the third buffer on the page (i.e. beyond EOF) is failing drop_buffers because it's bh->b_state == 0x3, which is BH_Uptodate | BH_Dirty. IOWs, there's dirty buffers beyond EOF. Say what? OK, set_buffer_dirty() is called on all buffers from __set_page_buffers_dirty(), regardless of whether the buffer is beyond EOF or not, which means that when we get to ->writepage, we have buffers marked dirty beyond EOF that we need to clean. So, we need to implement our own .set_page_dirty method that doesn't dirty buffers beyond EOF. This is messy because the buffer code is not meant to be shared and it has interesting locking issues on the buffer dirty bits. So just copy and paste it and then modify it to suit what we need. Note: the solutions the other filesystems and generic block code use of marking the buffers clean in ->writepage does not work for XFS. It still leaves dirty buffers beyond EOF and invalidations still fail. Hence rather than play whack-a-mole, this patch simply prevents those buffers from being dirtied in the first place. cc: Signed-off-by: Dave Chinner Reviewed-by: Brian Foster Signed-off-by: Dave Chinner --- fs/xfs/xfs_aops.c | 61 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 61 insertions(+) diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 11e9b4caa54f..b984647c24db 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -1753,11 +1753,72 @@ xfs_vm_readpages( return mpage_readpages(mapping, pages, nr_pages, xfs_get_blocks); } +/* + * This is basically a copy of __set_page_dirty_buffers() with one + * small tweak: buffers beyond EOF do not get marked dirty. If we mark them + * dirty, we'll never be able to clean them because we don't write buffers + * beyond EOF, and that means we can't invalidate pages that span EOF + * that have been marked dirty. Further, the dirty state can leak into + * the file interior if the file is extended, resulting in all sorts of + * bad things happening as the state does not match the underlying data. + * + * XXX: this really indicates that bufferheads in XFS need to die. Warts like + * this only exist because of bufferheads and how the generic code manages them. + */ +STATIC int +xfs_vm_set_page_dirty( + struct page *page) +{ + struct address_space *mapping = page->mapping; + struct inode *inode = mapping->host; + loff_t end_offset; + loff_t offset; + int newly_dirty; + + if (unlikely(!mapping)) + return !TestSetPageDirty(page); + + end_offset = i_size_read(inode); + offset = page_offset(page); + + spin_lock(&mapping->private_lock); + if (page_has_buffers(page)) { + struct buffer_head *head = page_buffers(page); + struct buffer_head *bh = head; + + do { + if (offset < end_offset) + set_buffer_dirty(bh); + bh = bh->b_this_page; + offset += 1 << inode->i_blkbits; + } while (bh != head); + } + newly_dirty = !TestSetPageDirty(page); + spin_unlock(&mapping->private_lock); + + if (newly_dirty) { + /* sigh - __set_page_dirty() is static, so copy it here, too */ + unsigned long flags; + + spin_lock_irqsave(&mapping->tree_lock, flags); + if (page->mapping) { /* Race with truncate? */ + WARN_ON_ONCE(!PageUptodate(page)); + account_page_dirtied(page, mapping); + radix_tree_tag_set(&mapping->page_tree, + page_index(page), PAGECACHE_TAG_DIRTY); + } + spin_unlock_irqrestore(&mapping->tree_lock, flags); + __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); + } + return newly_dirty; +} + const struct address_space_operations xfs_address_space_operations = { .readpage = xfs_vm_readpage, .readpages = xfs_vm_readpages, .writepage = xfs_vm_writepage, .writepages = xfs_vm_writepages, + .set_page_dirty = xfs_vm_set_page_dirty, .releasepage = xfs_vm_releasepage, .invalidatepage = xfs_vm_invalidatepage, .write_begin = xfs_vm_write_begin, From 85e584da3212140ee80fd047f9058bbee0bc00d5 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 2 Sep 2014 12:12:52 +1000 Subject: [PATCH 2/7] xfs: don't zero partial page cache pages during O_DIRECT writes xfs is using truncate_pagecache_range to invalidate the page cache during DIO reads. This is different from the other filesystems who only invalidate pages during DIO writes. truncate_pagecache_range is meant to be used when we are freeing the underlying data structs from disk, so it will zero any partial ranges in the page. This means a DIO read can zero out part of the page cache page, and it is possible the page will stay in cache. buffered reads will find an up to date page with zeros instead of the data actually on disk. This patch fixes things by using invalidate_inode_pages2_range instead. It preserves the page cache invalidation, but won't zero any pages. [dchinner: catch error and warn if it fails. Comment.] cc: stable@vger.kernel.org Signed-off-by: Chris Mason Reviewed-by: Dave Chinner Reviewed-by: Brian Foster Reviewed-by: Christoph Hellwig Signed-off-by: Dave Chinner --- fs/xfs/xfs_file.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 076b1708d134..827cfb2451b1 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -296,7 +296,16 @@ xfs_file_read_iter( xfs_rw_iunlock(ip, XFS_IOLOCK_EXCL); return ret; } - truncate_pagecache_range(VFS_I(ip), pos, -1); + + /* + * Invalidate whole pages. This can return an error if + * we fail to invalidate a page, but this should never + * happen on XFS. Warn if it does fail. + */ + ret = invalidate_inode_pages2_range(VFS_I(ip)->i_mapping, + pos >> PAGE_CACHE_SHIFT, -1); + WARN_ON_ONCE(ret); + ret = 0; } xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL); } From 834ffca6f7e345a79f6f2e2d131b0dfba8a4b67a Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Tue, 2 Sep 2014 12:12:52 +1000 Subject: [PATCH 3/7] xfs: don't zero partial page cache pages during O_DIRECT writes Similar to direct IO reads, direct IO writes are using truncate_pagecache_range to invalidate the page cache. This is incorrect due to the sub-block zeroing in the page cache that truncate_pagecache_range() triggers. This patch fixes things by using invalidate_inode_pages2_range instead. It preserves the page cache invalidation, but won't zero any pages. cc: stable@vger.kernel.org Signed-off-by: Dave Chinner Reviewed-by: Brian Foster Reviewed-by: Christoph Hellwig Signed-off-by: Dave Chinner --- fs/xfs/xfs_file.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 827cfb2451b1..19917faac2d2 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -644,7 +644,15 @@ xfs_file_dio_aio_write( pos, -1); if (ret) goto out; - truncate_pagecache_range(VFS_I(ip), pos, -1); + /* + * Invalidate whole pages. This can return an error if + * we fail to invalidate a page, but this should never + * happen on XFS. Warn if it does fail. + */ + ret = invalidate_inode_pages2_range(VFS_I(ip)->i_mapping, + pos >> PAGE_CACHE_SHIFT, -1); + WARN_ON_ONCE(ret); + ret = 0; } /* From 7d4ea3ce63a6bc532abb334c469c18481798af8c Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Tue, 2 Sep 2014 12:12:53 +1000 Subject: [PATCH 4/7] xfs: use ranged writeback and invalidation for direct IO Now we are not doing silly things with dirtying buffers beyond EOF and using invalidation correctly, we can finally reduce the ranges of writeback and invalidation used by direct IO to match that of the IO being issued. Bring the writeback and invalidation ranges back to match the generic direct IO code - this will greatly reduce the perturbation of cached data when direct IO and buffered IO are mixed, but still provide the same buffered vs direct IO coherency behaviour we currently have. Signed-off-by: Dave Chinner Reviewed-by: Brian Foster Reviewed-by: Christoph Hellwig Signed-off-by: Dave Chinner --- fs/xfs/xfs_file.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 19917faac2d2..de5368c803f9 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -291,7 +291,7 @@ xfs_file_read_iter( if (inode->i_mapping->nrpages) { ret = filemap_write_and_wait_range( VFS_I(ip)->i_mapping, - pos, -1); + pos, pos + size - 1); if (ret) { xfs_rw_iunlock(ip, XFS_IOLOCK_EXCL); return ret; @@ -303,7 +303,8 @@ xfs_file_read_iter( * happen on XFS. Warn if it does fail. */ ret = invalidate_inode_pages2_range(VFS_I(ip)->i_mapping, - pos >> PAGE_CACHE_SHIFT, -1); + pos >> PAGE_CACHE_SHIFT, + (pos + size - 1) >> PAGE_CACHE_SHIFT); WARN_ON_ONCE(ret); ret = 0; } @@ -641,7 +642,7 @@ xfs_file_dio_aio_write( if (mapping->nrpages) { ret = filemap_write_and_wait_range(VFS_I(ip)->i_mapping, - pos, -1); + pos, pos + count - 1); if (ret) goto out; /* @@ -650,7 +651,8 @@ xfs_file_dio_aio_write( * happen on XFS. Warn if it does fail. */ ret = invalidate_inode_pages2_range(VFS_I(ip)->i_mapping, - pos >> PAGE_CACHE_SHIFT, -1); + pos >> PAGE_CACHE_SHIFT, + (pos + count - 1) >> PAGE_CACHE_SHIFT); WARN_ON_ONCE(ret); ret = 0; } From ca446d880c399bb31301e7d8eefbd7fe3c504c4e Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Tue, 2 Sep 2014 12:12:53 +1000 Subject: [PATCH 5/7] xfs: don't log inode unless extent shift makes extent modifications The file collapse mechanism uses xfs_bmap_shift_extents() to collapse all subsequent extents down into the specified, previously punched out, region. This function performs some validation, such as whether a sufficient hole exists in the target region of the collapse, then shifts the remaining exents downward. The exit path of the function currently logs the inode unconditionally. While we must log the inode (and abort) if an error occurs and the transaction is dirty, the initial validation paths can generate errors before the transaction has been dirtied. This creates an unnecessary filesystem shutdown scenario, as the caller will cancel a transaction that has been marked dirty. Modify xfs_bmap_shift_extents() to OR the logflags bits as modifications are made to the inode bmap. Only log the inode in the exit path if logflags has been set. This ensures we only have to cancel a dirty transaction if modifications have been made and prevents an unnecessary filesystem shutdown otherwise. Signed-off-by: Brian Foster Reviewed-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Dave Chinner --- fs/xfs/libxfs/xfs_bmap.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c index de2d26d32844..86df952d3e24 100644 --- a/fs/xfs/libxfs/xfs_bmap.c +++ b/fs/xfs/libxfs/xfs_bmap.c @@ -5424,7 +5424,7 @@ xfs_bmap_shift_extents( struct xfs_bmap_free *flist, int num_exts) { - struct xfs_btree_cur *cur; + struct xfs_btree_cur *cur = NULL; struct xfs_bmbt_rec_host *gotp; struct xfs_bmbt_irec got; struct xfs_bmbt_irec left; @@ -5435,7 +5435,7 @@ xfs_bmap_shift_extents( int error = 0; int i; int whichfork = XFS_DATA_FORK; - int logflags; + int logflags = 0; xfs_filblks_t blockcount = 0; int total_extents; @@ -5478,16 +5478,11 @@ xfs_bmap_shift_extents( } } - /* We are going to change core inode */ - logflags = XFS_ILOG_CORE; if (ifp->if_flags & XFS_IFBROOT) { cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork); cur->bc_private.b.firstblock = *firstblock; cur->bc_private.b.flist = flist; cur->bc_private.b.flags = 0; - } else { - cur = NULL; - logflags |= XFS_ILOG_DEXT; } /* @@ -5545,11 +5540,14 @@ xfs_bmap_shift_extents( blockcount = left.br_blockcount + got.br_blockcount; xfs_iext_remove(ip, *current_ext, 1, 0); + logflags |= XFS_ILOG_CORE; if (cur) { error = xfs_btree_delete(cur, &i); if (error) goto del_cursor; XFS_WANT_CORRUPTED_GOTO(i == 1, del_cursor); + } else { + logflags |= XFS_ILOG_DEXT; } XFS_IFORK_NEXT_SET(ip, whichfork, XFS_IFORK_NEXTENTS(ip, whichfork) - 1); @@ -5575,6 +5573,7 @@ xfs_bmap_shift_extents( got.br_startoff = startoff; } + logflags |= XFS_ILOG_CORE; if (cur) { error = xfs_bmbt_update(cur, got.br_startoff, got.br_startblock, @@ -5582,6 +5581,8 @@ xfs_bmap_shift_extents( got.br_state); if (error) goto del_cursor; + } else { + logflags |= XFS_ILOG_DEXT; } (*current_ext)++; @@ -5597,6 +5598,7 @@ xfs_bmap_shift_extents( xfs_btree_del_cursor(cur, error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR); - xfs_trans_log_inode(tp, ip, logflags); + if (logflags) + xfs_trans_log_inode(tp, ip, logflags); return error; } From 1669a8ca2105968f660cf7d84ba38fd18075cd99 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Tue, 2 Sep 2014 12:12:53 +1000 Subject: [PATCH 6/7] xfs: xfs_file_collapse_range is delalloc challenged If we have delalloc extents on a file before we run a collapse range opertaion, we sync the range that we are going to collapse to convert delalloc extents in that region to real extents to simplify the shift operation. However, the shift operation then assumes that the extent list is not going to change as it iterates over the extent list moving things about. Unfortunately, this isn't true because we can't hold the ILOCK over all the operations. We can prevent new IO from modifying the extent list by holding the IOLOCK, but that doesn't prevent writeback from running.... And when writeback runs, it can convert delalloc extents is the range of the file prior to the region being collapsed, and this changes the indexes of all the extents in the file. That causes the collapse range operation to Go Bad. The right fix is to rewrite the extent shift operation not to be dependent on the extent list not changing across the entire operation, but this is a fairly significant piece of work to do. Hence, as a short-term workaround for the problem, sync the entire file before starting a collapse operation to remove all delalloc ranges from the file and so avoid the problem of concurrent writeback changing the extent list. Diagnosed-and-Reported-by: Brian Foster Signed-off-by: Dave Chinner Reviewed-by: Brian Foster Reviewed-by: Christoph Hellwig Signed-off-by: Dave Chinner --- fs/xfs/xfs_bmap_util.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c index 2f1e30d39a35..76b6a2910e40 100644 --- a/fs/xfs/xfs_bmap_util.c +++ b/fs/xfs/xfs_bmap_util.c @@ -1470,6 +1470,19 @@ xfs_collapse_file_space( start_fsb = XFS_B_TO_FSB(mp, offset + len); shift_fsb = XFS_B_TO_FSB(mp, len); + /* + * writeback the entire file to prevent concurrent writeback of ranges + * outside the collapsing region from changing the extent list. + * + * XXX: This is a temporary fix until the extent shift loop below is + * converted to use offsets and lookups within the ILOCK rather than + * carrying around the index into the extent list for the next + * iteration. + */ + error = filemap_write_and_wait(VFS_I(ip)->i_mapping); + if (error) + return error; + error = xfs_free_file_space(ip, offset, len); if (error) return error; From 41b9d7263ea1e270019c5d04fa0ab15db50b9725 Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Tue, 2 Sep 2014 12:12:53 +1000 Subject: [PATCH 7/7] xfs: trim eofblocks before collapse range xfs_collapse_file_space() currently writes back the entire file undergoing collapse range to settle things down for the extent shift algorithm. While this prevents changes to the extent list during the collapse operation, the writeback itself is not enough to prevent unnecessary collapse failures. The current shift algorithm uses the extent index to iterate the in-core extent list. If a post-eof delalloc extent persists after the writeback (e.g., a prior zero range op where the end of the range aligns with eof can separate the post-eof blocks such that they are not written back and converted), xfs_bmap_shift_extents() becomes confused over the encoded br_startblock value and fails the collapse. As with the full writeback, this is a temporary fix until the algorithm is improved to cope with a volatile extent list and avoid attempts to shift post-eof extents. Signed-off-by: Brian Foster Reviewed-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Dave Chinner --- fs/xfs/xfs_bmap_util.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c index 76b6a2910e40..1707980f9a4b 100644 --- a/fs/xfs/xfs_bmap_util.c +++ b/fs/xfs/xfs_bmap_util.c @@ -1471,8 +1471,10 @@ xfs_collapse_file_space( shift_fsb = XFS_B_TO_FSB(mp, len); /* - * writeback the entire file to prevent concurrent writeback of ranges - * outside the collapsing region from changing the extent list. + * Writeback the entire file and force remove any post-eof blocks. The + * writeback prevents changes to the extent list via concurrent + * writeback and the eofblocks trim prevents the extent shift algorithm + * from running into a post-eof delalloc extent. * * XXX: This is a temporary fix until the extent shift loop below is * converted to use offsets and lookups within the ILOCK rather than @@ -1482,6 +1484,11 @@ xfs_collapse_file_space( error = filemap_write_and_wait(VFS_I(ip)->i_mapping); if (error) return error; + if (xfs_can_free_eofblocks(ip, true)) { + error = xfs_free_eofblocks(mp, ip, false); + if (error) + return error; + } error = xfs_free_file_space(ip, offset, len); if (error)