From 78f707bfc723552e8309b7c38a8d0cc51012e813 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 17 Feb 2009 13:59:08 +0100 Subject: block: revert part of 18ce3751ccd488c78d3827e9f6bf54e6322676fb The above commit added WRITE_SYNC and switched various places to using that for committing writes that will be waited upon immediately after submission. However, this causes a performance regression with AS and CFQ for ext3 at least, since sync_dirty_buffer() will submit some writes with WRITE_SYNC while ext3 has sumitted others dependent writes without the sync flag set. This causes excessive anticipation/idling in the IO scheduler because sync and async writes get interleaved, causing a big performance regression for the below test case (which is meant to simulate sqlite like behaviour). ---- test case ---- int main(int argc, char **argv) { int fdes, i; FILE *fp; struct timeval start; struct timeval end; struct timeval res; gettimeofday(&start, NULL); for (i=0; i --- fs/buffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 665d446b25bc..62b57e330b69 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -3108,7 +3108,7 @@ int sync_dirty_buffer(struct buffer_head *bh) if (test_clear_buffer_dirty(bh)) { get_bh(bh); bh->b_end_io = end_buffer_write_sync; - ret = submit_bh(WRITE_SYNC, bh); + ret = submit_bh(WRITE, bh); wait_on_buffer(bh); if (buffer_eopnotsupp(bh)) { clear_buffer_eopnotsupp(bh); -- cgit v1.2.2 From 1cf6e7d83bf334cc5916137862c920a97aabc018 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Wed, 18 Feb 2009 14:48:18 -0800 Subject: mm: task dirty accounting fix YAMAMOTO-san noticed that task_dirty_inc doesn't seem to be called properly for cases where set_page_dirty is not used to dirty a page (eg. mark_buffer_dirty). Additionally, there is some inconsistency about when task_dirty_inc is called. It is used for dirty balancing, however it even gets called for __set_page_dirty_no_writeback. So rather than increment it in a set_page_dirty wrapper, move it down to exactly where the dirty page accounting stats are incremented. Cc: YAMAMOTO Takashi Signed-off-by: Nick Piggin Acked-by: Peter Zijlstra Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 665d446b25bc..ff4d1cdd779b 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -777,6 +777,7 @@ static int __set_page_dirty(struct page *page, __inc_zone_page_state(page, NR_FILE_DIRTY); __inc_bdi_stat(mapping->backing_dev_info, BDI_RECLAIMABLE); + task_dirty_inc(current); task_io_account_write(PAGE_CACHE_SIZE); } radix_tree_tag_set(&mapping->page_tree, -- cgit v1.2.2 From a8e7d49aa7be728c4ae241a75a2a124cdcabc0c5 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 19 Mar 2009 11:32:05 -0700 Subject: Fix race in create_empty_buffers() vs __set_page_dirty_buffers() Nick Piggin noticed this (very unlikely) race between setting a page dirty and creating the buffers for it - we need to hold the mapping private_lock until we've set the page dirty bit in order to make sure that create_empty_buffers() might not build up a set of buffers without the dirty bits set when the page is dirty. I doubt anybody has ever hit this race (and it didn't solve the issue Nick was looking at), but as Nick says: "Still, it does appear to solve a real race, which we should close." Acked-by: Nick Piggin Signed-off-by: Linus Torvalds --- fs/buffer.c | 23 +++++++++++------------ 1 file changed, 11 insertions(+), 12 deletions(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 9f697419ed8e..891e1c78e4f1 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -760,15 +760,9 @@ EXPORT_SYMBOL(mark_buffer_dirty_inode); * If warn is true, then emit a warning if the page is not uptodate and has * not been truncated. */ -static int __set_page_dirty(struct page *page, +static void __set_page_dirty(struct page *page, struct address_space *mapping, int warn) { - if (unlikely(!mapping)) - return !TestSetPageDirty(page); - - if (TestSetPageDirty(page)) - return 0; - spin_lock_irq(&mapping->tree_lock); if (page->mapping) { /* Race with truncate? */ WARN_ON_ONCE(warn && !PageUptodate(page)); @@ -785,8 +779,6 @@ static int __set_page_dirty(struct page *page, } spin_unlock_irq(&mapping->tree_lock); __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); - - return 1; } /* @@ -816,6 +808,7 @@ static int __set_page_dirty(struct page *page, */ int __set_page_dirty_buffers(struct page *page) { + int newly_dirty; struct address_space *mapping = page_mapping(page); if (unlikely(!mapping)) @@ -831,9 +824,12 @@ int __set_page_dirty_buffers(struct page *page) bh = bh->b_this_page; } while (bh != head); } + newly_dirty = !TestSetPageDirty(page); spin_unlock(&mapping->private_lock); - return __set_page_dirty(page, mapping, 1); + if (newly_dirty) + __set_page_dirty(page, mapping, 1); + return newly_dirty; } EXPORT_SYMBOL(__set_page_dirty_buffers); @@ -1262,8 +1258,11 @@ void mark_buffer_dirty(struct buffer_head *bh) return; } - if (!test_set_buffer_dirty(bh)) - __set_page_dirty(bh->b_page, page_mapping(bh->b_page), 0); + if (!test_set_buffer_dirty(bh)) { + struct page *page = bh->b_page; + if (!TestSetPageDirty(page)) + __set_page_dirty(page, page_mapping(page), 0); + } } /* -- cgit v1.2.2 From 585d3bc06f4ca57f975a5a1f698f65a45ea66225 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Wed, 25 Feb 2009 10:44:19 +0100 Subject: fs: move bdev code out of buffer.c Move some block device related code out from buffer.c and put it in block_dev.c. I'm trying to move non-buffer_head code out of buffer.c Signed-off-by: Al Viro --- fs/buffer.c | 145 ------------------------------------------------------------ 1 file changed, 145 deletions(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 891e1c78e4f1..a2fd743d97cb 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -165,151 +165,6 @@ void end_buffer_write_sync(struct buffer_head *bh, int uptodate) put_bh(bh); } -/* - * Write out and wait upon all the dirty data associated with a block - * device via its mapping. Does not take the superblock lock. - */ -int sync_blockdev(struct block_device *bdev) -{ - int ret = 0; - - if (bdev) - ret = filemap_write_and_wait(bdev->bd_inode->i_mapping); - return ret; -} -EXPORT_SYMBOL(sync_blockdev); - -/* - * Write out and wait upon all dirty data associated with this - * device. Filesystem data as well as the underlying block - * device. Takes the superblock lock. - */ -int fsync_bdev(struct block_device *bdev) -{ - struct super_block *sb = get_super(bdev); - if (sb) { - int res = fsync_super(sb); - drop_super(sb); - return res; - } - return sync_blockdev(bdev); -} - -/** - * freeze_bdev -- lock a filesystem and force it into a consistent state - * @bdev: blockdevice to lock - * - * This takes the block device bd_mount_sem to make sure no new mounts - * happen on bdev until thaw_bdev() is called. - * If a superblock is found on this device, we take the s_umount semaphore - * on it to make sure nobody unmounts until the snapshot creation is done. - * The reference counter (bd_fsfreeze_count) guarantees that only the last - * unfreeze process can unfreeze the frozen filesystem actually when multiple - * freeze requests arrive simultaneously. It counts up in freeze_bdev() and - * count down in thaw_bdev(). When it becomes 0, thaw_bdev() will unfreeze - * actually. - */ -struct super_block *freeze_bdev(struct block_device *bdev) -{ - struct super_block *sb; - int error = 0; - - mutex_lock(&bdev->bd_fsfreeze_mutex); - if (bdev->bd_fsfreeze_count > 0) { - bdev->bd_fsfreeze_count++; - sb = get_super(bdev); - mutex_unlock(&bdev->bd_fsfreeze_mutex); - return sb; - } - bdev->bd_fsfreeze_count++; - - down(&bdev->bd_mount_sem); - sb = get_super(bdev); - if (sb && !(sb->s_flags & MS_RDONLY)) { - sb->s_frozen = SB_FREEZE_WRITE; - smp_wmb(); - - __fsync_super(sb); - - sb->s_frozen = SB_FREEZE_TRANS; - smp_wmb(); - - sync_blockdev(sb->s_bdev); - - if (sb->s_op->freeze_fs) { - error = sb->s_op->freeze_fs(sb); - if (error) { - printk(KERN_ERR - "VFS:Filesystem freeze failed\n"); - sb->s_frozen = SB_UNFROZEN; - drop_super(sb); - up(&bdev->bd_mount_sem); - bdev->bd_fsfreeze_count--; - mutex_unlock(&bdev->bd_fsfreeze_mutex); - return ERR_PTR(error); - } - } - } - - sync_blockdev(bdev); - mutex_unlock(&bdev->bd_fsfreeze_mutex); - - return sb; /* thaw_bdev releases s->s_umount and bd_mount_sem */ -} -EXPORT_SYMBOL(freeze_bdev); - -/** - * thaw_bdev -- unlock filesystem - * @bdev: blockdevice to unlock - * @sb: associated superblock - * - * Unlocks the filesystem and marks it writeable again after freeze_bdev(). - */ -int thaw_bdev(struct block_device *bdev, struct super_block *sb) -{ - int error = 0; - - mutex_lock(&bdev->bd_fsfreeze_mutex); - if (!bdev->bd_fsfreeze_count) { - mutex_unlock(&bdev->bd_fsfreeze_mutex); - return -EINVAL; - } - - bdev->bd_fsfreeze_count--; - if (bdev->bd_fsfreeze_count > 0) { - if (sb) - drop_super(sb); - mutex_unlock(&bdev->bd_fsfreeze_mutex); - return 0; - } - - if (sb) { - BUG_ON(sb->s_bdev != bdev); - if (!(sb->s_flags & MS_RDONLY)) { - if (sb->s_op->unfreeze_fs) { - error = sb->s_op->unfreeze_fs(sb); - if (error) { - printk(KERN_ERR - "VFS:Filesystem thaw failed\n"); - sb->s_frozen = SB_FREEZE_TRANS; - bdev->bd_fsfreeze_count++; - mutex_unlock(&bdev->bd_fsfreeze_mutex); - return error; - } - } - sb->s_frozen = SB_UNFROZEN; - smp_wmb(); - wake_up(&sb->s_wait_unfrozen); - } - drop_super(sb); - } - - up(&bdev->bd_mount_sem); - mutex_unlock(&bdev->bd_fsfreeze_mutex); - return 0; -} -EXPORT_SYMBOL(thaw_bdev); - /* * Various filesystems appear to want __find_get_block to be non-blocking. * But it's the page lock which protects the buffers. To get around this, -- cgit v1.2.2 From a64c8610bd3b753c6aff58f51c04cdf0ae478c18 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Fri, 27 Mar 2009 22:14:10 -0400 Subject: block_write_full_page: Use synchronous writes for WBC_SYNC_ALL writebacks When doing synchronous writes because wbc->sync_mode is set to WBC_SYNC_ALL, send the write request using WRITE_SYNC, so that we don't unduly block system calls such as fsync(). Signed-off-by: "Theodore Ts'o" Acked-by: Jan Kara --- fs/buffer.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 891e1c78e4f1..e7ebd95e0c68 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -1714,6 +1714,7 @@ static int __block_write_full_page(struct inode *inode, struct page *page, struct buffer_head *bh, *head; const unsigned blocksize = 1 << inode->i_blkbits; int nr_underway = 0; + int write_op = (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE); BUG_ON(!PageLocked(page)); @@ -1805,7 +1806,7 @@ static int __block_write_full_page(struct inode *inode, struct page *page, do { struct buffer_head *next = bh->b_this_page; if (buffer_async_write(bh)) { - submit_bh(WRITE, bh); + submit_bh(write_op, bh); nr_underway++; } bh = next; @@ -1859,7 +1860,7 @@ recover: struct buffer_head *next = bh->b_this_page; if (buffer_async_write(bh)) { clear_buffer_dirty(bh); - submit_bh(WRITE, bh); + submit_bh(write_op, bh); nr_underway++; } bh = next; -- cgit v1.2.2 From 47e4491b40df73c3b117e3d80b31b5b512a4b19f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 1 Apr 2009 07:07:16 -0400 Subject: Cleanup after commit 585d3bc06f4ca57f975a5a1f698f65a45ea66225 fsync_bdev() export and a bunch of stubs for !CONFIG_BLOCK case had been left behind Signed-off-by: Al Viro --- fs/buffer.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index a2fd743d97cb..b71e52925c83 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -3281,7 +3281,6 @@ EXPORT_SYMBOL(cont_write_begin); EXPORT_SYMBOL(end_buffer_read_sync); EXPORT_SYMBOL(end_buffer_write_sync); EXPORT_SYMBOL(file_fsync); -EXPORT_SYMBOL(fsync_bdev); EXPORT_SYMBOL(generic_block_bmap); EXPORT_SYMBOL(generic_cont_expand_simple); EXPORT_SYMBOL(init_buffer); -- cgit v1.2.2 From e3a7cca1ef4c1af9b0acef9bd66eff6582a737b5 Mon Sep 17 00:00:00 2001 From: Edward Shishkin Date: Tue, 31 Mar 2009 15:19:39 -0700 Subject: vfs: add/use account_page_dirtied() Add a helper function account_page_dirtied(). Use that from two callsites. reiser4 adds a function which adds a third callsite. Signed-off-by: Edward Shishkin Cc: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index a2fd743d97cb..73abe6d8218c 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -621,14 +621,7 @@ static void __set_page_dirty(struct page *page, spin_lock_irq(&mapping->tree_lock); if (page->mapping) { /* Race with truncate? */ WARN_ON_ONCE(warn && !PageUptodate(page)); - - if (mapping_cap_account_dirty(mapping)) { - __inc_zone_page_state(page, NR_FILE_DIRTY); - __inc_bdi_stat(mapping->backing_dev_info, - BDI_RECLAIMABLE); - task_dirty_inc(current); - task_io_account_write(PAGE_CACHE_SIZE); - } + account_page_dirtied(page, mapping); radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); } -- cgit v1.2.2 From c2ec175c39f62949438354f603f4aa170846aabb Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 31 Mar 2009 15:23:21 -0700 Subject: mm: page_mkwrite change prototype to match fault Change the page_mkwrite prototype to take a struct vm_fault, and return VM_FAULT_xxx flags. There should be no functional change. This makes it possible to return much more detailed error information to the VM (and also can provide more information eg. virtual_address to the driver, which might be important in some special cases). This is required for a subsequent fix. And will also make it easier to merge page_mkwrite() with fault() in future. Signed-off-by: Nick Piggin Cc: Chris Mason Cc: Trond Myklebust Cc: Miklos Szeredi Cc: Steven Whitehouse Cc: Mark Fasheh Cc: Joel Becker Cc: Artem Bityutskiy Cc: Felix Blyakher Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 73abe6d8218c..6d51a3da362c 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -2313,9 +2313,10 @@ int block_commit_write(struct page *page, unsigned from, unsigned to) * unlock the page. */ int -block_page_mkwrite(struct vm_area_struct *vma, struct page *page, +block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, get_block_t get_block) { + struct page *page = vmf->page; struct inode *inode = vma->vm_file->f_path.dentry->d_inode; unsigned long end; loff_t size; @@ -2340,6 +2341,9 @@ block_page_mkwrite(struct vm_area_struct *vma, struct page *page, ret = block_commit_write(page, 0, end); out_unlock: + if (ret) + ret = VM_FAULT_SIGBUS; + unlock_page(page); return ret; } -- cgit v1.2.2 From 56a76f8275c379ed73c8a43cfa1dfa2f5e9cfa19 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 31 Mar 2009 15:23:23 -0700 Subject: fs: fix page_mkwrite error cases in core code and btrfs page_mkwrite is called with neither the page lock nor the ptl held. This means a page can be concurrently truncated or invalidated out from underneath it. Callers are supposed to prevent truncate races themselves, however previously the only thing they can do in case they hit one is to raise a SIGBUS. A sigbus is wrong for the case that the page has been invalidated or truncated within i_size (eg. hole punched). Callers may also have to perform memory allocations in this path, where again, SIGBUS would be wrong. The previous patch ("mm: page_mkwrite change prototype to match fault") made it possible to properly specify errors. Convert the generic buffer.c code and btrfs to return sane error values (in the case of page removed from pagecache, VM_FAULT_NOPAGE will cause the fault handler to exit without doing anything, and the fault will be retried properly). This fixes core code, and converts btrfs as a template/example. All other filesystems defining their own page_mkwrite should be fixed in a similar manner. Acked-by: Chris Mason Signed-off-by: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 6d51a3da362c..0c14f8d52ee5 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -2320,7 +2320,7 @@ block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, struct inode *inode = vma->vm_file->f_path.dentry->d_inode; unsigned long end; loff_t size; - int ret = -EINVAL; + int ret = VM_FAULT_NOPAGE; /* make the VM retry the fault */ lock_page(page); size = i_size_read(inode); @@ -2340,10 +2340,14 @@ block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, if (!ret) ret = block_commit_write(page, 0, end); -out_unlock: - if (ret) - ret = VM_FAULT_SIGBUS; + if (unlikely(ret)) { + if (ret == -ENOMEM) + ret = VM_FAULT_OOM; + else /* -ENOSPC, -EIO, etc */ + ret = VM_FAULT_SIGBUS; + } +out_unlock: unlock_page(page); return ret; } -- cgit v1.2.2 From 327c0e968645f2601a43f5ea7c19c7b3a5fa0a34 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Tue, 31 Mar 2009 15:23:31 -0700 Subject: vmscan: fix it to take care of nodemask try_to_free_pages() is used for the direct reclaim of up to SWAP_CLUSTER_MAX pages when watermarks are low. The caller to alloc_pages_nodemask() can specify a nodemask of nodes that are allowed to be used but this is not passed to try_to_free_pages(). This can lead to unnecessary reclaim of pages that are unusable by the caller and int the worst case lead to allocation failure as progress was not been make where it is needed. This patch passes the nodemask used for alloc_pages_nodemask() to try_to_free_pages(). Reviewed-by: KOSAKI Motohiro Acked-by: Mel Gorman Signed-off-by: KAMEZAWA Hiroyuki Cc: Rik van Riel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 0c14f8d52ee5..c77b848c3d43 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -290,7 +290,7 @@ static void free_more_memory(void) &zone); if (zone) try_to_free_pages(node_zonelist(nid, GFP_NOFS), 0, - GFP_NOFS); + GFP_NOFS, NULL); } } -- cgit v1.2.2 From c2d7543851849a6923680cdd7e1047ed1a84a1c5 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Tue, 31 Mar 2009 15:23:46 -0700 Subject: filesystem freeze: allow SysRq emergency thaw to thaw frozen filesystems Now that the filesystem freeze operation has been elevated to the VFS, and is just an ioctl away, some sort of safety net for unintentionally frozen root filesystems may be in order. The timeout thaw originally proposed did not get merged, but perhaps something like this would be useful in emergencies. For example, freeze /path/to/mountpoint may freeze your root filesystem if you forgot that you had that unmounted. I chose 'j' as the last remaining character other than 'h' which is sort of reserved for help (because help is generated on any unknown character). I've tested this on a non-root fs with multiple (nested) freezers, as well as on a system rendered unresponsive due to a frozen root fs. [randy.dunlap@oracle.com: emergency thaw only if CONFIG_BLOCK enabled] Signed-off-by: Eric Sandeen Cc: Takashi Sato Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index c77b848c3d43..f5f8b15a6e40 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -547,6 +547,39 @@ repeat: return err; } +void do_thaw_all(unsigned long unused) +{ + struct super_block *sb; + char b[BDEVNAME_SIZE]; + + spin_lock(&sb_lock); +restart: + list_for_each_entry(sb, &super_blocks, s_list) { + sb->s_count++; + spin_unlock(&sb_lock); + down_read(&sb->s_umount); + while (sb->s_bdev && !thaw_bdev(sb->s_bdev, sb)) + printk(KERN_WARNING "Emergency Thaw on %s\n", + bdevname(sb->s_bdev, b)); + up_read(&sb->s_umount); + spin_lock(&sb_lock); + if (__put_super_and_need_restart(sb)) + goto restart; + } + spin_unlock(&sb_lock); + printk(KERN_WARNING "Emergency Thaw complete\n"); +} + +/** + * emergency_thaw_all -- forcibly thaw every frozen filesystem + * + * Used for emergency unfreeze of all filesystems via SysRq + */ +void emergency_thaw_all(void) +{ + pdflush_operation(do_thaw_all, 0); +} + /** * sync_mapping_buffers - write out & wait upon a mapping's "associated" buffers * @mapping: the mapping which wants those buffers written -- cgit v1.2.2 From 97f76d3d197f201ac8a8a3ced5b8fef81568e50e Mon Sep 17 00:00:00 2001 From: Nikanth Karthikesan Date: Thu, 2 Apr 2009 16:56:46 -0700 Subject: vfs: check bh->b_blocknr only if BH_Mapped is set Check bh->b_blocknr only if BH_Mapped is set. akpm: I doubt if b_blocknr is ever uninitialised here, but it could conceivably cause a problem if we're doing a lookup for block zero. Signed-off-by: Nikanth Karthikesan Cc: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index f5f8b15a6e40..2963858f0f31 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -199,13 +199,13 @@ __find_get_block_slow(struct block_device *bdev, sector_t block) head = page_buffers(page); bh = head; do { - if (bh->b_blocknr == block) { + if (!buffer_mapped(bh)) + all_mapped = 0; + else if (bh->b_blocknr == block) { ret = bh; get_bh(bh); goto out_unlock; } - if (!buffer_mapped(bh)) - all_mapped = 0; bh = bh->b_this_page; } while (bh != head); -- cgit v1.2.2 From 9cf6b720f84d6999ff9a514d0a939dd183846aaf Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 6 Apr 2009 14:48:03 +0200 Subject: block: fsync_buffers_list() should use SWRITE_SYNC_PLUG Then it can submit all the buffers without unplugging for each one. We will kick off the pending IO if we come across a new address space. Signed-off-by: Jens Axboe Signed-off-by: Linus Torvalds --- fs/buffer.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 5d55a896ff78..43afaa5d6901 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -737,7 +737,7 @@ static int fsync_buffers_list(spinlock_t *lock, struct list_head *list) { struct buffer_head *bh; struct list_head tmp; - struct address_space *mapping; + struct address_space *mapping, *prev_mapping = NULL; int err = 0, err2; INIT_LIST_HEAD(&tmp); @@ -762,7 +762,18 @@ static int fsync_buffers_list(spinlock_t *lock, struct list_head *list) * contents - it is a noop if I/O is still in * flight on potentially older contents. */ - ll_rw_block(SWRITE_SYNC, 1, &bh); + ll_rw_block(SWRITE_SYNC_PLUG, 1, &bh); + + /* + * Kick off IO for the previous mapping. Note + * that we will not run the very last mapping, + * wait_on_buffer() will do that for us + * through sync_buffer(). + */ + if (prev_mapping && prev_mapping != mapping) + blk_run_address_space(prev_mapping); + prev_mapping = mapping; + brelse(bh); spin_lock(lock); } @@ -2957,12 +2968,13 @@ void ll_rw_block(int rw, int nr, struct buffer_head *bhs[]) for (i = 0; i < nr; i++) { struct buffer_head *bh = bhs[i]; - if (rw == SWRITE || rw == SWRITE_SYNC) + if (rw == SWRITE || rw == SWRITE_SYNC || rw == SWRITE_SYNC_PLUG) lock_buffer(bh); else if (!trylock_buffer(bh)) continue; - if (rw == WRITE || rw == SWRITE || rw == SWRITE_SYNC) { + if (rw == WRITE || rw == SWRITE || rw == SWRITE_SYNC || + rw == SWRITE_SYNC_PLUG) { if (test_clear_buffer_dirty(bh)) { bh->b_end_io = end_buffer_write_sync; get_bh(bh); -- cgit v1.2.2 From 1aa2a7cc6fd7b5c86681a6ae9dfd1072c261a435 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 6 Apr 2009 14:48:08 +0200 Subject: block: switch sync_dirty_buffer() over to WRITE_SYNC We should now have the logic in place to handle this properly without regressing on the write performance, so re-enable the sync writes. Signed-off-by: Jens Axboe Signed-off-by: Linus Torvalds --- fs/buffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/buffer.c') diff --git a/fs/buffer.c b/fs/buffer.c index 43afaa5d6901..6e35762b6169 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -3010,7 +3010,7 @@ int sync_dirty_buffer(struct buffer_head *bh) if (test_clear_buffer_dirty(bh)) { get_bh(bh); bh->b_end_io = end_buffer_write_sync; - ret = submit_bh(WRITE, bh); + ret = submit_bh(WRITE_SYNC, bh); wait_on_buffer(bh); if (buffer_eopnotsupp(bh)) { clear_buffer_eopnotsupp(bh); -- cgit v1.2.2