Skip to contentSkip to conversation

Introduce a mmap sem to deal with some mmap issues

PATCH Superseded Josef Bacik 3 participants 8 messages Dec 14, 2020 – Dec 23, 2020 Btrfs file system

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

cover Introduce a mmap sem to deal with some mmap issues

Hello,

Both Filipe and I have found different issues that result in the same thing, we need to be able to exclude mmap from happening in certain scenarios. The specifics are well described in the commit logs, but generally there's 2 issues

1) dedupe needs to validate that pages match, and since the validation is done outside of the extent lock we can race with mmap and dedupe pages that do not match.
2) We can deadlock in certain low metadata scenarios where we need to flush an ordered extent, but can't because mmap is holding the page lock.

These issues exist for remap and fallocate, so add an i_mmap_sem to allow us to disallow mmap in these cases. I'm still waiting on xfstests to finish with this, but 2 hours in and no lockdep or deadlocks. Thanks,

Josef Bacik (4):
btrfs: add a i_mmap_lock to our inode
btrfs: cleanup inode_lock/inode_unlock uses
btrfs: exclude mmaps while doing remap
btrfs: exclude mmap from happening during all fallocate operations

 fs/btrfs/btrfs_inode.h   |  1 +
 fs/btrfs/ctree.h         |  1 +
 fs/btrfs/delayed-inode.c |  4 ++--
 fs/btrfs/file.c          | 20 ++++++++++----------
 fs/btrfs/inode.c         | 10 ++++++++++
 fs/btrfs/ioctl.c         | 26 +++++++++++++-------------
 fs/btrfs/reflink.c       | 30 ++++++++++++++++++++++++------
 fs/btrfs/relocation.c    |  4 ++--
 8 files changed, 63 insertions(+), 33 deletions(-)
Signature · 1 line
2.26.2
#2 lore raw

PATCH 2/4 btrfs: cleanup inode_lock/inode_unlock uses

A few places we intermix btrfs_inode_lock with a inode_unlock, and some places we just use inode_lock/inode_unlock instead of btrfs_inode_lock. None of these places are using this incorrectly, but as we adjust some of these callers it would be nice to keep everything consistent, so convert everybody to use btrfs_inode_lock/btrfs_inode_unlock.

---
 fs/btrfs/delayed-inode.c |  4 ++--
 fs/btrfs/file.c          | 18 +++++++++---------
 fs/btrfs/ioctl.c         | 26 +++++++++++++-------------
 fs/btrfs/reflink.c       |  4 ++--
 fs/btrfs/relocation.c    |  4 ++--
 5 files changed, 28 insertions(+), 28 deletions(-)
fs/btrfs/delayed-inode.c +2 -2
index 70c0340d839c..7532229e7efb 100644
@@ -1588,8 +1588,8 @@ bool btrfs_readdir_get_delayed_items(struct inode *inode,
 	 * We can only do one readdir with delayed items at a time because of
 	 * item->readdir_list.
 	 */
-	inode_unlock_shared(inode);
-	inode_lock(inode);
+	btrfs_inode_unlock(inode, BTRFS_ILOCK_SHARED);
+	btrfs_inode_lock(inode, 0);
 
 	mutex_lock(&delayed_node->mutex);
 	item = __btrfs_first_delayed_insertion_item(delayed_node);
fs/btrfs/file.c +9 -9
index 0e41459b8de6..ed633ae87e06 100644
@@ -2131,7 +2131,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	if (ret)
 		goto out;
 
-	inode_lock(inode);
+	btrfs_inode_lock(inode, 0);
 
 	atomic_inc(&root->log_batch);
 
@@ -2163,7 +2163,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	 */
 	ret = start_ordered_ops(inode, start, end);
 	if (ret) {
-		inode_unlock(inode);
+		btrfs_inode_unlock(inode, 0);
 		goto out;
 	}
 
@@ -2259,7 +2259,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	 * file again, but that will end up using the synchronization
 	 * inside btrfs_sync_log to keep things safe.
 	 */
-	inode_unlock(inode);
+	btrfs_inode_unlock(inode, 0);
 
 	if (ret != BTRFS_NO_LOG_SYNC) {
 		if (!ret) {
@@ -2289,7 +2289,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 
 out_release_extents:
 	btrfs_release_log_ctx_extents(&ctx);
-	inode_unlock(inode);
+	btrfs_inode_unlock(inode, 0);
 	goto out;
 }
 
@@ -2872,7 +2872,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	if (ret)
 		return ret;
 
-	inode_lock(inode);
+	btrfs_inode_lock(inode, 0);
 	ino_size = round_up(inode->i_size, fs_info->sectorsize);
 	ret = find_first_non_hole(BTRFS_I(inode), &offset, &len);
 	if (ret < 0)
@@ -2912,7 +2912,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 		truncated_block = true;
 		ret = btrfs_truncate_block(BTRFS_I(inode), offset, 0, 0);
 		if (ret) {
-			inode_unlock(inode);
+			btrfs_inode_unlock(inode, 0);
 			return ret;
 		}
 	}
@@ -3013,7 +3013,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 				ret = ret2;
 		}
 	}
-	inode_unlock(inode);
+	btrfs_inode_unlock(inode, 0);
 	return ret;
 }
 
@@ -3378,7 +3378,7 @@ static long btrfs_fallocate(struct file *file, int mode,
 
 	if (mode & FALLOC_FL_ZERO_RANGE) {
 		ret = btrfs_zero_range(inode, offset, len, mode);
-		inode_unlock(inode);
+		btrfs_inode_unlock(inode, 0);
 		return ret;
 	}
 
@@ -3488,7 +3488,7 @@ static long btrfs_fallocate(struct file *file, int mode,
 	unlock_extent_cached(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
 			     &cached_state);
 out:
-	inode_unlock(inode);
+	btrfs_inode_unlock(inode, 0);
 	/* Let go of our reservation. */
 	if (ret != 0 && !(mode & FALLOC_FL_ZERO_RANGE))
 		btrfs_free_reserved_data_space(BTRFS_I(inode), data_reserved,
fs/btrfs/ioctl.c +13 -13
index dde49a791f3e..becb5d9ae3ce 100644
@@ -226,7 +226,7 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 	if (ret)
 		return ret;
 
-	inode_lock(inode);
+	btrfs_inode_lock(inode, 0);
 	fsflags = btrfs_mask_fsflags_for_type(inode, fsflags);
 	old_fsflags = btrfs_inode_flags_to_fsflags(binode->flags);
 
@@ -353,7 +353,7 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
  out_end_trans:
 	btrfs_end_transaction(trans);
  out_unlock:
-	inode_unlock(inode);
+	btrfs_inode_unlock(inode, 0);
 	mnt_drop_write_file(file);
 	return ret;
 }
@@ -449,7 +449,7 @@ static int btrfs_ioctl_fssetxattr(struct file *file, void __user *arg)
 	if (ret)
 		return ret;
 
-	inode_lock(inode);
+	btrfs_inode_lock(inode, 0);
 
 	old_flags = binode->flags;
 	old_i_flags = inode->i_flags;
@@ -501,7 +501,7 @@ static int btrfs_ioctl_fssetxattr(struct file *file, void __user *arg)
 		inode->i_flags = old_i_flags;
 	}
 
-	inode_unlock(inode);
+	btrfs_inode_unlock(inode, 0);
 	mnt_drop_write_file(file);
 
 	return ret;
@@ -1010,7 +1010,7 @@ static noinline int btrfs_mksubvol(const struct path *parent,
 out_dput:
 	dput(dentry);
 out_unlock:
-	inode_unlock(dir);
+	btrfs_inode_unlock(dir, 0);
 	return error;
 }
 
@@ -1602,7 +1602,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			ra_index += cluster;
 		}
 
-		inode_lock(inode);
+		btrfs_inode_lock(inode, 0);
 		if (IS_SWAPFILE(inode)) {
 			ret = -ETXTBSY;
 		} else {
@@ -1611,13 +1611,13 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			ret = cluster_pages_for_defrag(inode, pages, i, cluster);
 		}
 		if (ret < 0) {
-			inode_unlock(inode);
+			btrfs_inode_unlock(inode, 0);
 			goto out_ra;
 		}
 
 		defrag_count += ret;
 		balance_dirty_pages_ratelimited(inode->i_mapping);
-		inode_unlock(inode);
+		btrfs_inode_unlock(inode, 0);
 
 		if (newer_than) {
 			if (newer_off == (u64)-1)
@@ -1665,9 +1665,9 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 
 out_ra:
 	if (do_compress) {
-		inode_lock(inode);
+		btrfs_inode_lock(inode, 0);
 		BTRFS_I(inode)->defrag_compress = BTRFS_COMPRESS_NONE;
-		inode_unlock(inode);
+		btrfs_inode_unlock(inode, 0);
 	}
 	if (!file)
 		kfree(ra);
@@ -3083,9 +3083,9 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 		goto out_dput;
 	}
 
-	inode_lock(inode);
+	btrfs_inode_lock(inode, 0);
 	err = btrfs_delete_subvolume(dir, dentry);
-	inode_unlock(inode);
+	btrfs_inode_unlock(inode, 0);
 	if (!err) {
 		fsnotify_rmdir(dir, dentry);
 		d_delete(dentry);
@@ -3094,7 +3094,7 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 out_dput:
 	dput(dentry);
 out_unlock_dir:
-	inode_unlock(dir);
+	btrfs_inode_unlock(dir, 0);
 free_subvol_name:
 	kfree(subvol_name_ptr);
 free_parent:
fs/btrfs/reflink.c +2 -2
index b03e7891394e..97f58b15eba2 100644
@@ -816,7 +816,7 @@ loff_t btrfs_remap_file_range(struct file *src_file, loff_t off,
 		return -EINVAL;
 
 	if (same_inode)
-		inode_lock(src_inode);
+		btrfs_inode_lock(src_inode, 0);
 	else
 		lock_two_nondirectories(src_inode, dst_inode);
 
@@ -832,7 +832,7 @@ loff_t btrfs_remap_file_range(struct file *src_file, loff_t off,
 
 out_unlock:
 	if (same_inode)
-		inode_unlock(src_inode);
+		btrfs_inode_unlock(src_inode, 0);
 	else
 		unlock_two_nondirectories(src_inode, dst_inode);
 
fs/btrfs/relocation.c +2 -2
index 19b7db8b2117..3fa4f7ba8da3 100644
@@ -2553,7 +2553,7 @@ static noinline_for_stack int prealloc_file_extent_cluster(
 	if (ret)
 		return ret;
 
-	inode_lock(&inode->vfs_inode);
+	btrfs_inode_lock(&inode->vfs_inode, 0);
 	for (nr = 0; nr < cluster->nr; nr++) {
 		start = cluster->boundary[nr] - offset;
 		if (nr + 1 < cluster->nr)
@@ -2571,7 +2571,7 @@ static noinline_for_stack int prealloc_file_extent_cluster(
 		if (ret)
 			break;
 	}
-	inode_unlock(&inode->vfs_inode);
+	btrfs_inode_unlock(&inode->vfs_inode, 0);
 
 	if (cur_offset < prealloc_end)
 		btrfs_free_reserved_data_space_noquota(inode->root->fs_info,
Signature · 1 line
2.26.2
#3 lore raw

PATCH 3/4 btrfs: exclude mmaps while doing remap

In reply to Josef Bacik

Darrick reported a potential issue to me where we could allow mmap writes after validating a page range matched in the case of dedupe. Generally we rely on lock page -> lock extent with the ordered flush to protect us, but this is done after we check the pages because we use the generic helpers, so we could modify the page in between doing the check and locking the range.

There also exists a deadlock, as described by Filipe

"""
When cloning a file range, we lock the inodes, flush any delalloc within the respective file ranges, wait for any ordered extents and then lock the file ranges in both inodes. This means that right after we flush delalloc and before we lock the file ranges, memory mapped writes can come in and dirty pages in the file ranges of the clone operation.

Most of the time this is harmless and causes no problems. However, if we are low on available metadata space, we can later end up in a deadlock when starting a transaction to replace file extent items. This happens if when allocating metadata space for the transaction, we need to wait for the async reclaim thread to release space and the reclaim thread needs to flush delalloc for the inode that got the memory mapped write and has its range locked by the clone task.

Basically what happens is the following:

1) A clone operation locks inodes A and B, flushes delalloc for both inodes in the respective file ranges and waits for any ordered extents in those ranges to complete;

2) Before the clone task locks the file ranges, another task does a memory mapped write (which does not lock the inode) for one of the inodes of the clone operation. So now we have a dirty page in one of the ranges used by the clone operation;

3) The clone operation locks the file ranges for inodes A and B;

4) Later, when iterating over the file extents of inode A, the clone task attempts to start a transaction. There's not enough available free metadata space, so the async reclaim task is started (if not running already) and we wait for someone to wake us up on our reservation ticket;

5) The async reclaim task is not able to release space by any other means and decides to flush delalloc for the inode of the clone operation;

6) The workqueue job used to flush the inode blocks when starting delalloc for the inode, since the file range is currently locked by the clone task;

7) But the clone task is waiting on its reservation ticket and the async reclaim task is waiting on the flush job to complete, which can't progress since the clone task has the file range locked. So unless some other task is able to release space, for example an ordered extent for some other inode completes, we have a deadlock between all these tasks;

When this happens stack traces like the following showup in dmesg/syslog:

 INFO: task kworker/u16:11:1810830 blocked for more than 120 seconds.
       Tainted: G    B   W         5.10.0-rc4-btrfs-next-73 #1
 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 task:kworker/u16:11  state:D stack:    0 pid:1810830 ppid:     2 flags:0x00004000
 Workqueue: btrfs-flush_delalloc btrfs_work_helper [btrfs]
 Call Trace:
  __schedule+0x5d1/0xcf0
  schedule+0x45/0xe0
  lock_extent_bits+0x1e6/0x2d0 [btrfs]
  ? finish_wait+0x90/0x90
  btrfs_invalidatepage+0x32c/0x390 [btrfs]
  ? __mod_memcg_state+0x8e/0x160
  __extent_writepage+0x2d4/0x400 [btrfs]
  extent_write_cache_pages+0x2b2/0x500 [btrfs]
  ? lock_release+0x20e/0x4c0
  ? trace_hardirqs_on+0x1b/0xf0
  extent_writepages+0x43/0x90 [btrfs]
  ? lock_acquire+0x1a3/0x490
  do_writepages+0x43/0xe0
  ? __filemap_fdatawrite_range+0xa4/0x100
  __filemap_fdatawrite_range+0xc5/0x100
  btrfs_run_delalloc_work+0x17/0x40 [btrfs]
  btrfs_work_helper+0xf1/0x600 [btrfs]
  process_one_work+0x24e/0x5e0
  worker_thread+0x50/0x3b0
  ? process_one_work+0x5e0/0x5e0
  kthread+0x153/0x170
  ? kthread_mod_delayed_work+0xc0/0xc0
  ret_from_fork+0x22/0x30
 INFO: task kworker/u16:1:2426217 blocked for more than 120 seconds.
       Tainted: G    B   W         5.10.0-rc4-btrfs-next-73 #1
 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 task:kworker/u16:1   state:D stack:    0 pid:2426217 ppid:     2 flags:0x00004000
 Workqueue: events_unbound btrfs_async_reclaim_metadata_space [btrfs]
 Call Trace:
  __schedule+0x5d1/0xcf0
  ? kvm_clock_read+0x14/0x30
  ? wait_for_completion+0x81/0x110
  schedule+0x45/0xe0
  schedule_timeout+0x30c/0x580
  ? _raw_spin_unlock_irqrestore+0x3c/0x60
  ? lock_acquire+0x1a3/0x490
  ? try_to_wake_up+0x7a/0xa20
  ? lock_release+0x20e/0x4c0
  ? lock_acquired+0x199/0x490
  ? wait_for_completion+0x81/0x110
  wait_for_completion+0xab/0x110
  start_delalloc_inodes+0x2af/0x390 [btrfs]
  btrfs_start_delalloc_roots+0x12d/0x250 [btrfs]
  flush_space+0x24f/0x660 [btrfs]
  btrfs_async_reclaim_metadata_space+0x1bb/0x480 [btrfs]
  process_one_work+0x24e/0x5e0
  worker_thread+0x20f/0x3b0
  ? process_one_work+0x5e0/0x5e0
  kthread+0x153/0x170
  ? kthread_mod_delayed_work+0xc0/0xc0
  ret_from_fork+0x22/0x30
(...)
several other tasks blocked on inode locks held by the clone task below
(...)
 RIP: 0033:0x7f61efe73fff
 Code: Unable to access opcode bytes at RIP 0x7f61efe73fd5.
 RSP: 002b:00007ffc3371bbe8 EFLAGS: 00000202 ORIG_RAX: 000000000000013c
 RAX: ffffffffffffffda RBX: 00007ffc3371bea0 RCX: 00007f61efe73fff
 RDX: 00000000ffffff9c RSI: 0000560fbd604690 RDI: 00000000ffffff9c
 RBP: 00007ffc3371beb0 R08: 0000000000000002 R09: 0000560fbd5d75f0
 R10: 0000560fbd5d81f0 R11: 0000000000000202 R12: 0000000000000002
 R13: 000000000000000b R14: 00007ffc3371bea0 R15: 00007ffc3371beb0
 task: fdm-stress        state:D stack:    0 pid:2508234 ppid:2508153 flags:0x00004000
 Call Trace:
  __schedule+0x5d1/0xcf0
  ? _raw_spin_unlock_irqrestore+0x3c/0x60
  schedule+0x45/0xe0
  __reserve_bytes+0x4a4/0xb10 [btrfs]
  ? finish_wait+0x90/0x90
  btrfs_reserve_metadata_bytes+0x29/0x190 [btrfs]
  btrfs_block_rsv_add+0x1f/0x50 [btrfs]
  start_transaction+0x2d1/0x760 [btrfs]
  btrfs_replace_file_extents+0x120/0x930 [btrfs]
  ? lock_release+0x20e/0x4c0
  btrfs_clone+0x3e4/0x7e0 [btrfs]
  ? btrfs_lookup_first_ordered_extent+0x8e/0x100 [btrfs]
  btrfs_clone_files+0xf6/0x150 [btrfs]
  btrfs_remap_file_range+0x324/0x3d0 [btrfs]
  do_clone_file_range+0xd4/0x1f0
  vfs_clone_file_range+0x4d/0x230
  ? lock_release+0x20e/0x4c0
  ioctl_file_clone+0x8f/0xc0
  do_vfs_ioctl+0x342/0x750
  __x64_sys_ioctl+0x62/0xb0
  do_syscall_64+0x33/0x80
  entry_SYSCALL_64_after_hwframe+0x44/0xa9

"""

Fix both of these issues by excluding mmaps from happening we are doing any sort of remap, which prevents this race completely.

---
 fs/btrfs/reflink.c | 30 ++++++++++++++++++++++++------
 1 file changed, 24 insertions(+), 6 deletions(-)
fs/btrfs/reflink.c +24 -6
index 97f58b15eba2..7df2182c98d4 100644
@@ -587,6 +587,20 @@ static void btrfs_double_extent_lock(struct inode *inode1, u64 loff1,
 	lock_extent(&BTRFS_I(inode2)->io_tree, loff2, loff2 + len - 1);
 }
 
+static void btrfs_double_mmap_lock(struct inode *inode1, struct inode *inode2)
+{
+	if (inode1 < inode2)
+		swap(inode1, inode2);
+	down_write(&BTRFS_I(inode1)->i_mmap_lock);
+	down_write_nested(&BTRFS_I(inode2)->i_mmap_lock, SINGLE_DEPTH_NESTING);
+}
+
+static void btrfs_double_mmap_unlock(struct inode *inode1, struct inode *inode2)
+{
+	up_write(&BTRFS_I(inode1)->i_mmap_lock);
+	up_write(&BTRFS_I(inode2)->i_mmap_lock);
+}
+
 static int btrfs_extent_same_range(struct inode *src, u64 loff, u64 len,
 				   struct inode *dst, u64 dst_loff)
 {
@@ -815,10 +829,12 @@ loff_t btrfs_remap_file_range(struct file *src_file, loff_t off,
 	if (remap_flags & ~(REMAP_FILE_DEDUP | REMAP_FILE_ADVISORY))
 		return -EINVAL;
 
-	if (same_inode)
-		btrfs_inode_lock(src_inode, 0);
-	else
+	if (same_inode) {
+		btrfs_inode_lock(src_inode, BTRFS_ILOCK_MMAP);
+	} else {
 		lock_two_nondirectories(src_inode, dst_inode);
+		btrfs_double_mmap_lock(src_inode, dst_inode);
+	}
 
 	ret = btrfs_remap_file_range_prep(src_file, off, dst_file, destoff,
 					  &len, remap_flags);
@@ -831,10 +847,12 @@ loff_t btrfs_remap_file_range(struct file *src_file, loff_t off,
 		ret = btrfs_clone_files(dst_file, src_file, off, len, destoff);
 
 out_unlock:
-	if (same_inode)
-		btrfs_inode_unlock(src_inode, 0);
-	else
+	if (same_inode) {
+		btrfs_inode_unlock(src_inode, BTRFS_ILOCK_MMAP);
+	} else {
+		btrfs_double_mmap_unlock(src_inode, dst_inode);
 		unlock_two_nondirectories(src_inode, dst_inode);
+	}
 
 	return ret < 0 ? ret : len;
 }
Signature · 1 line
2.26.2
#6 lore raw

PATCH 4/4 btrfs: exclude mmap from happening during all fallocate operations

In reply to Josef Bacik

There's a small window where a deadlock can happen between fallocate and mmap. This is described in detail by Filipe

"""
When doing a fallocate operation we lock the inode, flush delalloc within the target range, wait for any ordered extents to complete and then lock the file range. Before we lock the range and after we flush delalloc, there is a time window where another task can come in and do a memory mapped write for a page within the fallocate range.

This means that after fallocate locks the range, there can be a dirty page in the range. More often than not, this does not cause any problem. The exception is when we are low on available metadata space, because an fallocate operation needs to start a transaction while holding the file range locked, either through btrfs_prealloc_file_range() or through the call to btrfs_fallocate_update_isize(). If that's the case, we can end up in a deadlock. The following list of steps explains how that happens:

1) A fallocate operation starts, locks the inode, flushes delalloc in the range and waits for ordered extents in the range to complete;

2) Before the fallocate task locks the file range, another task does a memory mapped write for a page in the fallocate target range. This is possible since memory mapped writes do not (and can not) lock the inode;

3) The fallocate task locks the file range. At this point there is one dirty page in the range (due to the memory mapped write);

4) When the fallocate task attempts to start a transaction, it blocks when attempting to reserve metadata space, since we are low on available metadata space. Before blocking (wait on its reservation ticket), it starts the async reclaim task (if not running already);

5) The async reclaim task is not able to release space through any other means, so it decides to flush delalloc for inodes with dirty pages. It finds that the inode used in the fallocate operation has a dirty page and therefore queues a job (fs_info->flushs_workers workqueue) to flush delalloc for that inode and waits on that job to complete;

6) The flush job blocks when attempting to lock the file range because it is currently locked by the fallocate task;

7) The fallocate task keeps waiting for its metadata reservation, waiting for a wakeup on its reservation ticket. The async reclaim task is waiting on the flush job, which in turn is waiting for locking the file range that is currently locked by the fallocate task. So unless some other task is able to release enough metadata space, for example an ordered extent for some other inode completes, we end up in a deadlock between all these tasks.

When this happens stack traces like the following showup in dmesg/syslog:

 INFO: task kworker/u16:11:1810830 blocked for more than 120 seconds.
       Tainted: G    B   W         5.10.0-rc4-btrfs-next-73 #1
 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 task:kworker/u16:11  state:D stack:    0 pid:1810830 ppid:     2 flags:0x00004000
 Workqueue: btrfs-flush_delalloc btrfs_work_helper [btrfs]
 Call Trace:
  __schedule+0x5d1/0xcf0
  schedule+0x45/0xe0
  lock_extent_bits+0x1e6/0x2d0 [btrfs]
  ? finish_wait+0x90/0x90
  btrfs_invalidatepage+0x32c/0x390 [btrfs]
  ? __mod_memcg_state+0x8e/0x160
  __extent_writepage+0x2d4/0x400 [btrfs]
  extent_write_cache_pages+0x2b2/0x500 [btrfs]
  ? lock_release+0x20e/0x4c0
  ? trace_hardirqs_on+0x1b/0xf0
  extent_writepages+0x43/0x90 [btrfs]
  ? lock_acquire+0x1a3/0x490
  do_writepages+0x43/0xe0
  ? __filemap_fdatawrite_range+0xa4/0x100
  __filemap_fdatawrite_range+0xc5/0x100
  btrfs_run_delalloc_work+0x17/0x40 [btrfs]
  btrfs_work_helper+0xf1/0x600 [btrfs]
  process_one_work+0x24e/0x5e0
  worker_thread+0x50/0x3b0
  ? process_one_work+0x5e0/0x5e0
  kthread+0x153/0x170
  ? kthread_mod_delayed_work+0xc0/0xc0
  ret_from_fork+0x22/0x30
 INFO: task kworker/u16:1:2426217 blocked for more than 120 seconds.
       Tainted: G    B   W         5.10.0-rc4-btrfs-next-73 #1
 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 task:kworker/u16:1   state:D stack:    0 pid:2426217 ppid:     2 flags:0x00004000
 Workqueue: events_unbound btrfs_async_reclaim_metadata_space [btrfs]
 Call Trace:
  __schedule+0x5d1/0xcf0
  ? kvm_clock_read+0x14/0x30
  ? wait_for_completion+0x81/0x110
  schedule+0x45/0xe0
  schedule_timeout+0x30c/0x580
  ? _raw_spin_unlock_irqrestore+0x3c/0x60
  ? lock_acquire+0x1a3/0x490
  ? try_to_wake_up+0x7a/0xa20
  ? lock_release+0x20e/0x4c0
  ? lock_acquired+0x199/0x490
  ? wait_for_completion+0x81/0x110
  wait_for_completion+0xab/0x110
  start_delalloc_inodes+0x2af/0x390 [btrfs]
  btrfs_start_delalloc_roots+0x12d/0x250 [btrfs]
  flush_space+0x24f/0x660 [btrfs]
  btrfs_async_reclaim_metadata_space+0x1bb/0x480 [btrfs]
  process_one_work+0x24e/0x5e0
  worker_thread+0x20f/0x3b0
  ? process_one_work+0x5e0/0x5e0
  kthread+0x153/0x170
  ? kthread_mod_delayed_work+0xc0/0xc0
  ret_from_fork+0x22/0x30
(...)
several tasks waiting for the inode lock held by the fallocate task below
(...)
 RIP: 0033:0x7f61efe73fff
 Code: Unable to access opcode bytes at RIP 0x7f61efe73fd5.
 RSP: 002b:00007ffc3371bbe8 EFLAGS: 00000202 ORIG_RAX: 000000000000013c
 RAX: ffffffffffffffda RBX: 00007ffc3371bea0 RCX: 00007f61efe73fff
 RDX: 00000000ffffff9c RSI: 0000560fbd5d90a0 RDI: 00000000ffffff9c
 RBP: 00007ffc3371beb0 R08: 0000000000000001 R09: 0000000000000003
 R10: 0000560fbd5d7ad0 R11: 0000000000000202 R12: 0000000000000001
 R13: 000000000000005e R14: 00007ffc3371bea0 R15: 00007ffc3371beb0
 task:fdm-stress        state:D stack:    0 pid:2508243 ppid:2508153 flags:0x00000000
 Call Trace:
  __schedule+0x5d1/0xcf0
  ? _raw_spin_unlock_irqrestore+0x3c/0x60
  schedule+0x45/0xe0
  __reserve_bytes+0x4a4/0xb10 [btrfs]
  ? finish_wait+0x90/0x90
  btrfs_reserve_metadata_bytes+0x29/0x190 [btrfs]
  btrfs_block_rsv_add+0x1f/0x50 [btrfs]
  start_transaction+0x2d1/0x760 [btrfs]
  btrfs_replace_file_extents+0x120/0x930 [btrfs]
  ? btrfs_fallocate+0xdcf/0x1260 [btrfs]
  btrfs_fallocate+0xdfb/0x1260 [btrfs]
  ? filename_lookup+0xf1/0x180
  vfs_fallocate+0x14f/0x440
  ioctl_preallocate+0x92/0xc0
  do_vfs_ioctl+0x66b/0x750
  ? __do_sys_newfstat+0x53/0x60
  __x64_sys_ioctl+0x62/0xb0
  do_syscall_64+0x33/0x80
  entry_SYSCALL_64_after_hwframe+0x44/0xa9

"""

Fix this by disallowing mmaps from happening while we're doing any of the fallocate operations on this inode.

---
 fs/btrfs/file.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
fs/btrfs/file.c +6 -6
index ed633ae87e06..d573ffcbd626 100644
@@ -2872,7 +2872,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	if (ret)
 		return ret;
 
-	btrfs_inode_lock(inode, 0);
+	btrfs_inode_lock(inode, BTRFS_ILOCK_MMAP);
 	ino_size = round_up(inode->i_size, fs_info->sectorsize);
 	ret = find_first_non_hole(BTRFS_I(inode), &offset, &len);
 	if (ret < 0)
@@ -2912,7 +2912,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 		truncated_block = true;
 		ret = btrfs_truncate_block(BTRFS_I(inode), offset, 0, 0);
 		if (ret) {
-			btrfs_inode_unlock(inode, 0);
+			btrfs_inode_unlock(inode, BTRFS_ILOCK_MMAP);
 			return ret;
 		}
 	}
@@ -3013,7 +3013,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 				ret = ret2;
 		}
 	}
-	btrfs_inode_unlock(inode, 0);
+	btrfs_inode_unlock(inode, BTRFS_ILOCK_MMAP);
 	return ret;
 }
 
@@ -3336,7 +3336,7 @@ static long btrfs_fallocate(struct file *file, int mode,
 			return ret;
 	}
 
-	btrfs_inode_lock(inode, 0);
+	btrfs_inode_lock(inode, BTRFS_ILOCK_MMAP);
 
 	if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + len > inode->i_size) {
 		ret = inode_newsize_ok(inode, offset + len);
@@ -3378,7 +3378,7 @@ static long btrfs_fallocate(struct file *file, int mode,
 
 	if (mode & FALLOC_FL_ZERO_RANGE) {
 		ret = btrfs_zero_range(inode, offset, len, mode);
-		btrfs_inode_unlock(inode, 0);
+		btrfs_inode_unlock(inode, BTRFS_ILOCK_MMAP);
 		return ret;
 	}
 
@@ -3488,7 +3488,7 @@ static long btrfs_fallocate(struct file *file, int mode,
 	unlock_extent_cached(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
 			     &cached_state);
 out:
-	btrfs_inode_unlock(inode, 0);
+	btrfs_inode_unlock(inode, BTRFS_ILOCK_MMAP);
 	/* Let go of our reservation. */
 	if (ret != 0 && !(mode & FALLOC_FL_ZERO_RANGE))
 		btrfs_free_reserved_data_space(BTRFS_I(inode), data_reserved,
Signature · 1 line
2.26.2
#1 lore raw

PATCH 1/4 btrfs: add a i_mmap_lock to our inode

In reply to Josef Bacik

We need to be able to exclude page_mkwrite from happening concurrently with certain operations. To facilitate this, add a i_mmap_lock to our inode, down_read() it in our mkwrite, and add a new ILOCK flag to indicate that we want to take the i_mmap_lock as well. I used pahole to check the size of the btrfs_inode, the sizes are as follows

no lockdep:
before: 1120 (3 per 4k page)
after: 1160 (3 per 4k page)

lockdep:
before: 2072 (1 per 4k page)
after: 2224 (1 per 4k page)

We're slightly larger but it doesn't change how many objects we can fit per page.

---
 fs/btrfs/btrfs_inode.h |  1 +
 fs/btrfs/ctree.h       |  1 +
 fs/btrfs/inode.c       | 10 ++++++++++
 3 files changed, 12 insertions(+)
fs/btrfs/btrfs_inode.h +1 -0
index d9bf53d9ff90..66cec5c13396 100644
@@ -220,6 +220,7 @@ struct btrfs_inode {
 	/* Hook into fs_info->delayed_iputs */
 	struct list_head delayed_iput;
 
+	struct rw_semaphore i_mmap_lock;
 	struct inode vfs_inode;
 };
 
fs/btrfs/ctree.h +1 -0
index 3935d297d198..e2bebb3318cc 100644
@@ -3147,6 +3147,7 @@ extern const struct iomap_dio_ops btrfs_dio_ops;
 /* Inode locking type flags, by default the exclusive lock is taken */
 #define BTRFS_ILOCK_SHARED	(1U << 0)
 #define BTRFS_ILOCK_TRY 	(1U << 1)
+#define BTRFS_ILOCK_MMAP	(1U << 2)
 
 int btrfs_inode_lock(struct inode *inode, unsigned int ilock_flags);
 void btrfs_inode_unlock(struct inode *inode, unsigned int ilock_flags);
fs/btrfs/inode.c +10 -0
index 070716650df8..a5ee4d45ee02 100644
@@ -101,6 +101,7 @@ static void __endio_write_update_ordered(struct btrfs_inode *inode,
  * BTRFS_ILOCK_SHARED - acquire a shared lock on the inode
  * BTRFS_ILOCK_TRY - try to acquire the lock, if fails on first attempt
  *		     return -EAGAIN
+ * BTRFS_ILOCK_MMAP - acquire a write lock on the i_mmap_lock
  */
 int btrfs_inode_lock(struct inode *inode, unsigned int ilock_flags)
 {
@@ -121,6 +122,8 @@ int btrfs_inode_lock(struct inode *inode, unsigned int ilock_flags)
 		}
 		inode_lock(inode);
 	}
+	if (ilock_flags & BTRFS_ILOCK_MMAP)
+		down_write(&BTRFS_I(inode)->i_mmap_lock);
 	return 0;
 }
 
@@ -132,6 +135,8 @@ int btrfs_inode_lock(struct inode *inode, unsigned int ilock_flags)
  */
 void btrfs_inode_unlock(struct inode *inode, unsigned int ilock_flags)
 {
+	if (ilock_flags & BTRFS_ILOCK_MMAP)
+		up_write(&BTRFS_I(inode)->i_mmap_lock);
 	if (ilock_flags & BTRFS_ILOCK_SHARED)
 		inode_unlock_shared(inode);
 	else
@@ -8344,6 +8349,7 @@ vm_fault_t btrfs_page_mkwrite(struct vm_fault *vmf)
 
 	ret = VM_FAULT_NOPAGE; /* make the VM retry the fault */
 again:
+	down_read(&BTRFS_I(inode)->i_mmap_lock);
 	lock_page(page);
 	size = i_size_read(inode);
 
@@ -8367,6 +8373,7 @@ vm_fault_t btrfs_page_mkwrite(struct vm_fault *vmf)
 		unlock_extent_cached(io_tree, page_start, page_end,
 				     &cached_state);
 		unlock_page(page);
+		up_read(&BTRFS_I(inode)->i_mmap_lock);
 		btrfs_start_ordered_extent(ordered, 1);
 		btrfs_put_ordered_extent(ordered);
 		goto again;
@@ -8424,6 +8431,7 @@ vm_fault_t btrfs_page_mkwrite(struct vm_fault *vmf)
 	BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->root->last_log_commit;
 
 	unlock_extent_cached(io_tree, page_start, page_end, &cached_state);
+	up_read(&BTRFS_I(inode)->i_mmap_lock);
 
 	btrfs_delalloc_release_extents(BTRFS_I(inode), PAGE_SIZE);
 	sb_end_pagefault(inode->i_sb);
@@ -8432,6 +8440,7 @@ vm_fault_t btrfs_page_mkwrite(struct vm_fault *vmf)
 
 out_unlock:
 	unlock_page(page);
+	up_read(&BTRFS_I(inode)->i_mmap_lock);
 out:
 	btrfs_delalloc_release_extents(BTRFS_I(inode), PAGE_SIZE);
 	btrfs_delalloc_release_space(BTRFS_I(inode), data_reserved, page_start,
@@ -8680,6 +8689,7 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 	INIT_LIST_HEAD(&ei->delalloc_inodes);
 	INIT_LIST_HEAD(&ei->delayed_iput);
 	RB_CLEAR_NODE(&ei->rb_node);
+	init_rwsem(&ei->i_mmap_lock);
 
 	return inode;
 }
Signature · 1 line
2.26.2
#4 lore raw

Re: [PATCH 3/4] btrfs: exclude mmaps while doing remap

In reply to Josef Bacik
Quoted 6 lines · Josef Bacik

Darrick reported a potential issue to me where we could allow mmap writes after validating a page range matched in the case of dedupe. Generally we rely on lock page -> lock extent with the ordered flush to protect us, but this is done after we check the pages because we use the generic helpers, so we could modify the page in between doing the check and locking the range.

FWIW I only found that via code inspection because Matthew Wilcox asked me about whether or not filesystems did the right thing. I wrote the attached fstest to try to demonstrate the problem on btrfs but it actually passes on xfs and btrfs. This means either that (a) btrfs is doing it right through some other means because I don't understand its locking or (b) the test is wrong.

The test /does/ explode as expected on ocfs2 because mmap io lol there.

--D

From: Darrick J. Wong <[email protected]>
Subject: [PATCH] generic: test file writers racing with FIDEDUPERANGE

Create a test to make sure that dedupe actually locks the file ranges correctly before starting the content comparison and keeps them locked until the operation completes.

---
 src/Makefile          |    2 
 src/deduperace.c      |  370 +++++++++++++++++++++++++++++++++++++++++++++++++
 tests/generic/949     |   51 +++++++
 tests/generic/949.out |    2 
 tests/generic/group   |    1 
 5 files changed, 425 insertions(+), 1 deletion(-)
 create mode 100644 src/deduperace.c
 create mode 100755 tests/generic/949
 create mode 100644 tests/generic/949.out
src/Makefile +1 -1
index 5b9781c5..2875e8ef 100644
@@ -21,7 +21,7 @@ TARGETS = dirstress fill fill2 getpagesize holes lstat64 \
 
 LINUX_TARGETS = xfsctl bstat t_mtab getdevicesize preallo_rw_pattern_reader \
 	preallo_rw_pattern_writer ftrunc trunc fs_perms testx looptest \
-	locktest unwritten_mmap bulkstat_unlink_test \
+	locktest unwritten_mmap bulkstat_unlink_test deduperace \
 	bulkstat_unlink_test_modified t_dir_offset t_futimens t_immutable \
 	stale_handle pwrite_mmap_blocked t_dir_offset2 seek_sanity_test \
 	seek_copy_test t_readdir_1 t_readdir_2 fsync-tester nsexec cloner \
src/deduperace.c new file +370 -0

370 lines not on this page · raw message

tests/generic/949 new file +51 -0

51 lines not on this page · raw message

tests/generic/949.out new file +2 -0
new file mode 100644
index 00000000..2998b46c
@@ -0,0 +1,2 @@
+QA output created by 949
+Silence is golden.
tests/generic/group +1 -0
index fe196f0a..364c011d 100644
@@ -628,6 +628,7 @@
 722 auto quick atime bigtime shutdown
 947 auto quick rw clone
 948 auto quick rw copy_range
+949 auto quick rw dedupe clone
 1200 auto quick swapext quota
 1201 auto quick swapext quota
 1202 auto quick swapext
#5 lore raw

Re: [PATCH 3/4] btrfs: exclude mmaps while doing remap

Quoted 8 lines · Darrick J. Wong
Quoted 6 lines · Josef Bacik

Darrick reported a potential issue to me where we could allow mmap writes after validating a page range matched in the case of dedupe. Generally we rely on lock page -> lock extent with the ordered flush to protect us, but this is done after we check the pages because we use the generic helpers, so we could modify the page in between doing the check and locking the range.

Quoted 9 lines

FWIW I only found that via code inspection because Matthew Wilcox asked me about whether or not filesystems did the right thing. I wrote the attached fstest to try to demonstrate the problem on btrfs but it actually passes on xfs and btrfs. This means either that (a) btrfs is doing it right through some other means because I don't understand its locking or (b) the test is wrong.

The test /does/ explode as expected on ocfs2 because mmap io lol there.

The problem exists, it's just the window is super narrow. If you put a schedule_timeout(HZ) right after btrfs_remap_file_range_prep() you'd hit the problem every time. We basically have to mkwrite right between doing the vfs_dedupe_file_range_compare() and calling btrfs_extent_same(), which is going to be hard to do. You aren't missing something, it's just hard to hit. Thanks,

Josef

#7 lore raw
In reply to Josef Bacik
Quoted 3 lines · Josef Bacik

These issues exist for remap and fallocate, so add an i_mmap_sem to allow us to disallow mmap in these cases. I'm still waiting on xfstests to finish with this, but 2 hours in and no lockdep or deadlocks. Thanks,

Any chance you could look into lifting it to the VFS and also convert ext4 and XFS to the common scheme?