[PATCH AUTOSEL 6.6 044/294] btrfs: make btrfs_discard_workfn() block_group ref explicit

[PATCH AUTOSEL 6.6 044/294] btrfs: make btrfs_discard_workfn() block_group ref explicit

[Sasha Levin]

From: Boris Burkov <boris@bur.io>

[ Upstream commit 895c6721d310c036dcfebb5ab845822229fa35eb ]

Currently, the async discard machinery owns a ref to the block_group
when the block_group is queued on a discard list. However, to handle
races with discard cancellation and the discard workfn, we have a
specific logic to detect that the block_group is *currently* running in
the workfn, to protect the workfn's usage amidst cancellation.

As far as I can tell, this doesn't have any overt bugs (though
finish_discard_pass() and remove_from_discard_list() racing can have a
surprising outcome for the caller of remove_from_discard_list() in that
it is again added at the end).

But it is needlessly complicated to rely on locking and the nullity of
discard_ctl->block_group. Simplify this significantly by just taking a
refcount while we are in the workfn and unconditionally drop it in both
the remove and workfn paths, regardless of if they race.

Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/btrfs/discard.c | 34 ++++++++++++++++------------------
 1 file changed, 16 insertions(+), 18 deletions(-)

diff --git a/fs/btrfs/discard.c b/fs/btrfs/discard.c
index 944a7340f6a44..3981c941f5b55 100644
--- a/fs/btrfs/discard.c
+++ b/fs/btrfs/discard.c
@@ -167,13 +167,7 @@ static bool remove_from_discard_list(struct btrfs_discard_ctl *discard_ctl,
 	block_group->discard_eligible_time = 0;
 	queued = !list_empty(&block_group->discard_list);
 	list_del_init(&block_group->discard_list);
-	/*
-	 * If the block group is currently running in the discard workfn, we
-	 * don't want to deref it, since it's still being used by the workfn.
-	 * The workfn will notice this case and deref the block group when it is
-	 * finished.
-	 */
-	if (queued && !running)
+	if (queued)
 		btrfs_put_block_group(block_group);
 
 	spin_unlock(&discard_ctl->lock);
@@ -260,9 +254,10 @@ static struct btrfs_block_group *peek_discard_list(
 			block_group->discard_cursor = block_group->start;
 			block_group->discard_state = BTRFS_DISCARD_EXTENTS;
 		}
-		discard_ctl->block_group = block_group;
 	}
 	if (block_group) {
+		btrfs_get_block_group(block_group);
+		discard_ctl->block_group = block_group;
 		*discard_state = block_group->discard_state;
 		*discard_index = block_group->discard_index;
 	}
@@ -493,9 +488,20 @@ static void btrfs_discard_workfn(struct work_struct *work)
 
 	block_group = peek_discard_list(discard_ctl, &discard_state,
 					&discard_index, now);
-	if (!block_group || !btrfs_run_discard_work(discard_ctl))
+	if (!block_group)
 		return;
+	if (!btrfs_run_discard_work(discard_ctl)) {
+		spin_lock(&discard_ctl->lock);
+		btrfs_put_block_group(block_group);
+		discard_ctl->block_group = NULL;
+		spin_unlock(&discard_ctl->lock);
+		return;
+	}
 	if (now < block_group->discard_eligible_time) {
+		spin_lock(&discard_ctl->lock);
+		btrfs_put_block_group(block_group);
+		discard_ctl->block_group = NULL;
+		spin_unlock(&discard_ctl->lock);
 		btrfs_discard_schedule_work(discard_ctl, false);
 		return;
 	}
@@ -547,15 +553,7 @@ static void btrfs_discard_workfn(struct work_struct *work)
 	spin_lock(&discard_ctl->lock);
 	discard_ctl->prev_discard = trimmed;
 	discard_ctl->prev_discard_time = now;
-	/*
-	 * If the block group was removed from the discard list while it was
-	 * running in this workfn, then we didn't deref it, since this function
-	 * still owned that reference. But we set the discard_ctl->block_group
-	 * back to NULL, so we can use that condition to know that now we need
-	 * to deref the block_group.
-	 */
-	if (discard_ctl->block_group == NULL)
-		btrfs_put_block_group(block_group);
+	btrfs_put_block_group(block_group);
 	discard_ctl->block_group = NULL;
 	__btrfs_discard_schedule_work(discard_ctl, now, false);
 	spin_unlock(&discard_ctl->lock);
-- 
2.39.5

[PATCH AUTOSEL 6.6 045/294] btrfs: avoid linker error in btrfs_find_create_tree_block()

[Sasha Levin]

From: Mark Harmstone <maharmstone@fb.com>

[ Upstream commit 7ef3cbf17d2734ca66c4ed8573be45f4e461e7ee ]

The inline function btrfs_is_testing() is hardcoded to return 0 if
CONFIG_BTRFS_FS_RUN_SANITY_TESTS is not set. Currently we're relying on
the compiler optimizing out the call to alloc_test_extent_buffer() in
btrfs_find_create_tree_block(), as it's not been defined (it's behind an
 #ifdef).

Add a stub version of alloc_test_extent_buffer() to avoid linker errors
on non-standard optimization levels. This problem was seen on GCC 14
with -O0 and is helps to see symbols that would be otherwise optimized
out.

Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Mark Harmstone <maharmstone@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/btrfs/extent_io.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index b2ae50dcca0fe..ed08d8e5639f5 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -3565,10 +3565,10 @@ struct extent_buffer *find_extent_buffer(struct btrfs_fs_info *fs_info,
 	return eb;
 }
 
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 struct extent_buffer *alloc_test_extent_buffer(struct btrfs_fs_info *fs_info,
 					u64 start)
 {
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 	struct extent_buffer *eb, *exists = NULL;
 	int ret;
 
@@ -3604,8 +3604,11 @@ struct extent_buffer *alloc_test_extent_buffer(struct btrfs_fs_info *fs_info,
 free_eb:
 	btrfs_release_extent_buffer(eb);
 	return exists;
-}
+#else
+	/* Stub to avoid linker error when compiled with optimizations turned off. */
+	return NULL;
 #endif
+}
 
 static struct extent_buffer *grab_extent_buffer(
 		struct btrfs_fs_info *fs_info, struct page *page)
-- 
2.39.5

[PATCH AUTOSEL 6.6 046/294] btrfs: run btrfs_error_commit_super() early

[Sasha Levin]

From: Qu Wenruo <wqu@suse.com>

[ Upstream commit df94a342efb451deb0e32b495d1d6cd4bb3a1648 ]

[BUG]
Even after all the error fixes related the
"ASSERT(list_empty(&fs_info->delayed_iputs));" in close_ctree(), I can
still hit it reliably with my experimental 2K block size.

[CAUSE]
In my case, all the error is triggered after the fs is already in error
status.

I find the following call trace to be the cause of race:

           Main thread                       |     endio_write_workers
---------------------------------------------+---------------------------
close_ctree()                                |
|- btrfs_error_commit_super()                |
|  |- btrfs_cleanup_transaction()            |
|  |  |- btrfs_destroy_all_ordered_extents() |
|  |     |- btrfs_wait_ordered_roots()       |
|  |- btrfs_run_delayed_iputs()              |
|                                            | btrfs_finish_ordered_io()
|                                            | |- btrfs_put_ordered_extent()
|                                            |    |- btrfs_add_delayed_iput()
|- ASSERT(list_empty(delayed_iputs))         |
   !!! Triggered !!!

The root cause is that, btrfs_wait_ordered_roots() only wait for
ordered extents to finish their IOs, not to wait for them to finish and
removed.

[FIX]
Since btrfs_error_commit_super() will flush and wait for all ordered
extents, it should be executed early, before we start flushing the
workqueues.

And since btrfs_error_commit_super() now runs early, there is no need to
run btrfs_run_delayed_iputs() inside it, so just remove the
btrfs_run_delayed_iputs() call from btrfs_error_commit_super().

Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/btrfs/disk-io.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 2387210231f23..1e1650012606e 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -4313,6 +4313,14 @@ void __cold close_ctree(struct btrfs_fs_info *fs_info)
 	/* clear out the rbtree of defraggable inodes */
 	btrfs_cleanup_defrag_inodes(fs_info);
 
+	/*
+	 * Handle the error fs first, as it will flush and wait for all ordered
+	 * extents.  This will generate delayed iputs, thus we want to handle
+	 * it first.
+	 */
+	if (unlikely(BTRFS_FS_ERROR(fs_info)))
+		btrfs_error_commit_super(fs_info);
+
 	/*
 	 * Wait for any fixup workers to complete.
 	 * If we don't wait for them here and they are still running by the time
@@ -4402,9 +4410,6 @@ void __cold close_ctree(struct btrfs_fs_info *fs_info)
 			btrfs_err(fs_info, "commit super ret %d", ret);
 	}
 
-	if (BTRFS_FS_ERROR(fs_info))
-		btrfs_error_commit_super(fs_info);
-
 	kthread_stop(fs_info->transaction_kthread);
 	kthread_stop(fs_info->cleaner_kthread);
 
@@ -4541,10 +4546,6 @@ static void btrfs_error_commit_super(struct btrfs_fs_info *fs_info)
 	/* cleanup FS via transaction */
 	btrfs_cleanup_transaction(fs_info);
 
-	mutex_lock(&fs_info->cleaner_mutex);
-	btrfs_run_delayed_iputs(fs_info);
-	mutex_unlock(&fs_info->cleaner_mutex);
-
 	down_write(&fs_info->cleanup_work_sem);
 	up_write(&fs_info->cleanup_work_sem);
 }
-- 
2.39.5

[PATCH AUTOSEL 6.6 047/294] btrfs: fix non-empty delayed iputs list on unmount due to async workers

[Sasha Levin]

From: Filipe Manana <fdmanana@suse.com>

[ Upstream commit cda76788f8b0f7de3171100e3164ec1ce702292e ]

At close_ctree() after we have ran delayed iputs either explicitly through
calling btrfs_run_delayed_iputs() or later during the call to
btrfs_commit_super() or btrfs_error_commit_super(), we assert that the
delayed iputs list is empty.

We have (another) race where this assertion might fail because we have
queued an async write into the fs_info->workers workqueue. Here's how it
happens:

1) We are submitting a data bio for an inode that is not the data
   relocation inode, so we call btrfs_wq_submit_bio();

2) btrfs_wq_submit_bio() submits a work for the fs_info->workers queue
   that will run run_one_async_done();

3) We enter close_ctree(), flush several work queues except
   fs_info->workers, explicitly run delayed iputs with a call to
   btrfs_run_delayed_iputs() and then again shortly after by calling
   btrfs_commit_super() or btrfs_error_commit_super(), which also run
   delayed iputs;

4) run_one_async_done() is executed in the work queue, and because there
   was an IO error (bio->bi_status is not 0) it calls btrfs_bio_end_io(),
   which drops the final reference on the associated ordered extent by
   calling btrfs_put_ordered_extent() - and that adds a delayed iput for
   the inode;

5) At close_ctree() we find that after stopping the cleaner and
   transaction kthreads the delayed iputs list is not empty, failing the
   following assertion:

      ASSERT(list_empty(&fs_info->delayed_iputs));

Fix this by flushing the fs_info->workers workqueue before running delayed
iputs at close_ctree().

David reported this when running generic/648, which exercises IO error
paths by using the DM error table.

Reported-by: David Sterba <dsterba@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/btrfs/disk-io.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 1e1650012606e..34a30d61b470c 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -4341,6 +4341,19 @@ void __cold close_ctree(struct btrfs_fs_info *fs_info)
 	 */
 	btrfs_flush_workqueue(fs_info->delalloc_workers);
 
+	/*
+	 * We can have ordered extents getting their last reference dropped from
+	 * the fs_info->workers queue because for async writes for data bios we
+	 * queue a work for that queue, at btrfs_wq_submit_bio(), that runs
+	 * run_one_async_done() which calls btrfs_bio_end_io() in case the bio
+	 * has an error, and that later function can do the final
+	 * btrfs_put_ordered_extent() on the ordered extent attached to the bio,
+	 * which adds a delayed iput for the inode. So we must flush the queue
+	 * so that we don't have delayed iputs after committing the current
+	 * transaction below and stopping the cleaner and transaction kthreads.
+	 */
+	btrfs_flush_workqueue(fs_info->workers);
+
 	/*
 	 * When finishing a compressed write bio we schedule a work queue item
 	 * to finish an ordered extent - btrfs_finish_compressed_write_work()
-- 
2.39.5

[PATCH AUTOSEL 6.6 048/294] btrfs: get zone unusable bytes while holding lock at btrfs_reclaim_bgs_work()

[Sasha Levin]

From: Filipe Manana <fdmanana@suse.com>

[ Upstream commit 1283b8c125a83bf7a7dbe90c33d3472b6d7bf612 ]

At btrfs_reclaim_bgs_work(), we are grabbing a block group's zone unusable
bytes while not under the protection of the block group's spinlock, so
this can trigger race reports from KCSAN (or similar tools) since that
field is typically updated while holding the lock, such as at
__btrfs_add_free_space_zoned() for example.

Fix this by grabbing the zone unusable bytes while we are still in the
critical section holding the block group's spinlock, which is right above
where we are currently grabbing it.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/btrfs/block-group.c | 18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

diff --git a/fs/btrfs/block-group.c b/fs/btrfs/block-group.c
index 434cf3d5f4cf1..226e6434a58a9 100644
--- a/fs/btrfs/block-group.c
+++ b/fs/btrfs/block-group.c
@@ -1885,6 +1885,17 @@ void btrfs_reclaim_bgs_work(struct work_struct *work)
 			up_write(&space_info->groups_sem);
 			goto next;
 		}
+
+		/*
+		 * Cache the zone_unusable value before turning the block group
+		 * to read only. As soon as the block group is read only it's
+		 * zone_unusable value gets moved to the block group's read-only
+		 * bytes and isn't available for calculations anymore. We also
+		 * cache it before unlocking the block group, to prevent races
+		 * (reports from KCSAN and such tools) with tasks updating it.
+		 */
+		zone_unusable = bg->zone_unusable;
+
 		spin_unlock(&bg->lock);
 
 		/*
@@ -1900,13 +1911,6 @@ void btrfs_reclaim_bgs_work(struct work_struct *work)
 			goto next;
 		}
 
-		/*
-		 * Cache the zone_unusable value before turning the block group
-		 * to read only. As soon as the blog group is read only it's
-		 * zone_unusable value gets moved to the block group's read-only
-		 * bytes and isn't available for calculations anymore.
-		 */
-		zone_unusable = bg->zone_unusable;
 		ret = inc_block_group_ro(bg, 0);
 		up_write(&space_info->groups_sem);
 		if (ret < 0)
-- 
2.39.5

[PATCH AUTOSEL 6.6 049/294] btrfs: send: return -ENAMETOOLONG when attempting a path that is too long

[Sasha Levin]

From: Filipe Manana <fdmanana@suse.com>

[ Upstream commit a77749b3e21813566cea050bbb3414ae74562eba ]

When attempting to build a too long path we are currently returning
-ENOMEM, which is very odd and misleading. So update fs_path_ensure_buf()
to return -ENAMETOOLONG instead. Also, while at it, move the WARN_ON()
into the if statement's expression, as it makes it clear what is being
tested and also has the effect of adding 'unlikely' to the statement,
which allows the compiler to generate better code as this condition is
never expected to happen.

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/btrfs/send.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
index aa1e6d88a72c7..e2ead36e5be42 100644
--- a/fs/btrfs/send.c
+++ b/fs/btrfs/send.c
@@ -487,10 +487,8 @@ static int fs_path_ensure_buf(struct fs_path *p, int len)
 	if (p->buf_len >= len)
 		return 0;
 
-	if (len > PATH_MAX) {
-		WARN_ON(1);
-		return -ENOMEM;
-	}
+	if (WARN_ON(len > PATH_MAX))
+		return -ENAMETOOLONG;
 
 	path_len = p->end - p->start;
 	old_buf_len = p->buf_len;
-- 
2.39.5

[PATCH AUTOSEL 6.6 050/294] btrfs: zoned: exit btrfs_can_activate_zone if BTRFS_FS_NEED_ZONE_FINISH is set

[Sasha Levin]

From: Johannes Thumshirn <johannes.thumshirn@wdc.com>

[ Upstream commit 26b38e28162ef4ceb1e0482299820fbbd7dbcd92 ]

If BTRFS_FS_NEED_ZONE_FINISH is already set for the whole filesystem, exit
early in btrfs_can_activate_zone(). There's no need to check if
BTRFS_FS_NEED_ZONE_FINISH needs to be set if it is already set.

Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/btrfs/zoned.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c
index 197dfafbf4013..6ef0b47facbf3 100644
--- a/fs/btrfs/zoned.c
+++ b/fs/btrfs/zoned.c
@@ -2220,6 +2220,9 @@ bool btrfs_can_activate_zone(struct btrfs_fs_devices *fs_devices, u64 flags)
 	if (!btrfs_is_zoned(fs_info))
 		return true;
 
+	if (test_bit(BTRFS_FS_NEED_ZONE_FINISH, &fs_info->flags))
+		return false;
+
 	/* Check if there is a device with active zones left */
 	mutex_lock(&fs_info->chunk_mutex);
 	spin_lock(&fs_info->zone_active_bgs_lock);
-- 
2.39.5