From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Ioannis Angelakopoulos <iangelak@fb.com>,
Josef Bacik <josef@toxicpanda.com>,
David Sterba <dsterba@suse.com>, Sasha Levin <sashal@kernel.org>,
clm@fb.com, linux-btrfs@vger.kernel.org
Subject: [PATCH AUTOSEL 6.0 27/46] btrfs: add lockdep annotations for transaction states wait events
Date: Tue, 11 Oct 2022 10:49:55 -0400 [thread overview]
Message-ID: <20221011145015.1622882-27-sashal@kernel.org> (raw)
In-Reply-To: <20221011145015.1622882-1-sashal@kernel.org>
From: Ioannis Angelakopoulos <iangelak@fb.com>
[ Upstream commit 3e738c531aad8caa7f3d20ab878a8a0d3574e730 ]
Add lockdep annotations for the transaction states that have wait
events;
1) TRANS_STATE_COMMIT_START
2) TRANS_STATE_UNBLOCKED
3) TRANS_STATE_SUPER_COMMITTED
4) TRANS_STATE_COMPLETED
The new macros introduced here to annotate the transaction states wait
events have the same effect as the generic lockdep annotation macros.
With the exception of the lockdep annotation for TRANS_STATE_COMMIT_START
the transaction thread has to acquire the lockdep maps for the
transaction states as reader after the lockdep map for num_writers is
released so that lockdep does not complain.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Ioannis Angelakopoulos <iangelak@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/ctree.h | 32 +++++++++++++++++++++++++
fs/btrfs/disk-io.c | 8 +++++++
fs/btrfs/transaction.c | 53 ++++++++++++++++++++++++++++++++++--------
3 files changed, 83 insertions(+), 10 deletions(-)
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index e886cf639c0f..f8172e269f03 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -1098,6 +1098,7 @@ struct btrfs_fs_info {
*/
struct lockdep_map btrfs_trans_num_writers_map;
struct lockdep_map btrfs_trans_num_extwriters_map;
+ struct lockdep_map btrfs_state_change_map[4];
#ifdef CONFIG_BTRFS_FS_REF_VERIFY
spinlock_t ref_verify_lock;
@@ -1181,6 +1182,13 @@ enum {
BTRFS_ROOT_RESET_LOCKDEP_CLASS,
};
+enum btrfs_lockdep_trans_states {
+ BTRFS_LOCKDEP_TRANS_COMMIT_START,
+ BTRFS_LOCKDEP_TRANS_UNBLOCKED,
+ BTRFS_LOCKDEP_TRANS_SUPER_COMMITTED,
+ BTRFS_LOCKDEP_TRANS_COMPLETED,
+};
+
/*
* Lockdep annotation for wait events.
*
@@ -1219,6 +1227,22 @@ enum {
#define btrfs_lockdep_release(owner, lock) \
rwsem_release(&owner->lock##_map, _THIS_IP_)
+/*
+ * Macros for the transaction states wait events, similar to the generic wait
+ * event macros.
+ */
+#define btrfs_might_wait_for_state(owner, i) \
+ do { \
+ rwsem_acquire(&owner->btrfs_state_change_map[i], 0, 0, _THIS_IP_); \
+ rwsem_release(&owner->btrfs_state_change_map[i], _THIS_IP_); \
+ } while (0)
+
+#define btrfs_trans_state_lockdep_acquire(owner, i) \
+ rwsem_acquire_read(&owner->btrfs_state_change_map[i], 0, 0, _THIS_IP_)
+
+#define btrfs_trans_state_lockdep_release(owner, i) \
+ rwsem_release(&owner->btrfs_state_change_map[i], _THIS_IP_)
+
/* Initialization of the lockdep map */
#define btrfs_lockdep_init_map(owner, lock) \
do { \
@@ -1226,6 +1250,14 @@ enum {
lockdep_init_map(&owner->lock##_map, #lock, &lock##_key, 0); \
} while (0)
+/* Initialization of the transaction states lockdep maps. */
+#define btrfs_state_lockdep_init_map(owner, lock, state) \
+ do { \
+ static struct lock_class_key lock##_key; \
+ lockdep_init_map(&owner->btrfs_state_change_map[state], #lock, \
+ &lock##_key, 0); \
+ } while (0)
+
static inline void btrfs_wake_unfinished_drop(struct btrfs_fs_info *fs_info)
{
clear_and_wake_up_bit(BTRFS_FS_UNFINISHED_DROPS, &fs_info->flags);
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 811d743e26e6..68c6cb4e9283 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2992,6 +2992,14 @@ void btrfs_init_fs_info(struct btrfs_fs_info *fs_info)
btrfs_lockdep_init_map(fs_info, btrfs_trans_num_writers);
btrfs_lockdep_init_map(fs_info, btrfs_trans_num_extwriters);
+ btrfs_state_lockdep_init_map(fs_info, btrfs_trans_commit_start,
+ BTRFS_LOCKDEP_TRANS_COMMIT_START);
+ btrfs_state_lockdep_init_map(fs_info, btrfs_trans_unblocked,
+ BTRFS_LOCKDEP_TRANS_UNBLOCKED);
+ btrfs_state_lockdep_init_map(fs_info, btrfs_trans_super_committed,
+ BTRFS_LOCKDEP_TRANS_SUPER_COMMITTED);
+ btrfs_state_lockdep_init_map(fs_info, btrfs_trans_completed,
+ BTRFS_LOCKDEP_TRANS_COMPLETED);
INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots);
INIT_LIST_HEAD(&fs_info->space_info);
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 44e47db4c8e8..d3576f84020d 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -550,6 +550,7 @@ static void wait_current_trans(struct btrfs_fs_info *fs_info)
refcount_inc(&cur_trans->use_count);
spin_unlock(&fs_info->trans_lock);
+ btrfs_might_wait_for_state(fs_info, BTRFS_LOCKDEP_TRANS_UNBLOCKED);
wait_event(fs_info->transaction_wait,
cur_trans->state >= TRANS_STATE_UNBLOCKED ||
TRANS_ABORTED(cur_trans));
@@ -868,6 +869,15 @@ static noinline void wait_for_commit(struct btrfs_transaction *commit,
u64 transid = commit->transid;
bool put = false;
+ /*
+ * At the moment this function is called with min_state either being
+ * TRANS_STATE_COMPLETED or TRANS_STATE_SUPER_COMMITTED.
+ */
+ if (min_state == TRANS_STATE_COMPLETED)
+ btrfs_might_wait_for_state(fs_info, BTRFS_LOCKDEP_TRANS_COMPLETED);
+ else
+ btrfs_might_wait_for_state(fs_info, BTRFS_LOCKDEP_TRANS_SUPER_COMMITTED);
+
while (1) {
wait_event(commit->commit_wait, commit->state >= min_state);
if (put)
@@ -1980,6 +1990,7 @@ void btrfs_commit_transaction_async(struct btrfs_trans_handle *trans)
* Wait for the current transaction commit to start and block
* subsequent transaction joins
*/
+ btrfs_might_wait_for_state(fs_info, BTRFS_LOCKDEP_TRANS_COMMIT_START);
wait_event(fs_info->transaction_blocked_wait,
cur_trans->state >= TRANS_STATE_COMMIT_START ||
TRANS_ABORTED(cur_trans));
@@ -2137,12 +2148,12 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
ktime_t interval;
ASSERT(refcount_read(&trans->use_count) == 1);
+ btrfs_trans_state_lockdep_acquire(fs_info, BTRFS_LOCKDEP_TRANS_COMMIT_START);
/* Stop the commit early if ->aborted is set */
if (TRANS_ABORTED(cur_trans)) {
ret = cur_trans->aborted;
- btrfs_end_transaction(trans);
- return ret;
+ goto lockdep_trans_commit_start_release;
}
btrfs_trans_release_metadata(trans);
@@ -2159,10 +2170,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
* Any running threads may add more while we are here.
*/
ret = btrfs_run_delayed_refs(trans, 0);
- if (ret) {
- btrfs_end_transaction(trans);
- return ret;
- }
+ if (ret)
+ goto lockdep_trans_commit_start_release;
}
btrfs_create_pending_block_groups(trans);
@@ -2191,10 +2200,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
if (run_it) {
ret = btrfs_start_dirty_block_groups(trans);
- if (ret) {
- btrfs_end_transaction(trans);
- return ret;
- }
+ if (ret)
+ goto lockdep_trans_commit_start_release;
}
}
@@ -2209,6 +2216,9 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
if (trans->in_fsync)
want_state = TRANS_STATE_SUPER_COMMITTED;
+
+ btrfs_trans_state_lockdep_release(fs_info,
+ BTRFS_LOCKDEP_TRANS_COMMIT_START);
ret = btrfs_end_transaction(trans);
wait_for_commit(cur_trans, want_state);
@@ -2222,6 +2232,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
cur_trans->state = TRANS_STATE_COMMIT_START;
wake_up(&fs_info->transaction_blocked_wait);
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_COMMIT_START);
if (cur_trans->list.prev != &fs_info->trans_list) {
enum btrfs_trans_state want_state = TRANS_STATE_COMPLETED;
@@ -2323,6 +2334,16 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
wait_event(cur_trans->writer_wait,
atomic_read(&cur_trans->num_writers) == 1);
+ /*
+ * Make lockdep happy by acquiring the state locks after
+ * btrfs_trans_num_writers is released. If we acquired the state locks
+ * before releasing the btrfs_trans_num_writers lock then lockdep would
+ * complain because we did not follow the reverse order unlocking rule.
+ */
+ btrfs_trans_state_lockdep_acquire(fs_info, BTRFS_LOCKDEP_TRANS_COMPLETED);
+ btrfs_trans_state_lockdep_acquire(fs_info, BTRFS_LOCKDEP_TRANS_SUPER_COMMITTED);
+ btrfs_trans_state_lockdep_acquire(fs_info, BTRFS_LOCKDEP_TRANS_UNBLOCKED);
+
/*
* We've started the commit, clear the flag in case we were triggered to
* do an async commit but somebody else started before the transaction
@@ -2332,6 +2353,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
if (TRANS_ABORTED(cur_trans)) {
ret = cur_trans->aborted;
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_UNBLOCKED);
goto scrub_continue;
}
/*
@@ -2466,6 +2488,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
mutex_unlock(&fs_info->reloc_mutex);
wake_up(&fs_info->transaction_wait);
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_UNBLOCKED);
ret = btrfs_write_and_wait_transaction(trans);
if (ret) {
@@ -2497,6 +2520,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
*/
cur_trans->state = TRANS_STATE_SUPER_COMMITTED;
wake_up(&cur_trans->commit_wait);
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_SUPER_COMMITTED);
btrfs_finish_extent_commit(trans);
@@ -2510,6 +2534,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
*/
cur_trans->state = TRANS_STATE_COMPLETED;
wake_up(&cur_trans->commit_wait);
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_COMPLETED);
spin_lock(&fs_info->trans_lock);
list_del_init(&cur_trans->list);
@@ -2538,7 +2563,10 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
unlock_reloc:
mutex_unlock(&fs_info->reloc_mutex);
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_UNBLOCKED);
scrub_continue:
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_SUPER_COMMITTED);
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_COMPLETED);
btrfs_scrub_continue(fs_info);
cleanup_transaction:
btrfs_trans_release_metadata(trans);
@@ -2556,6 +2584,11 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
btrfs_lockdep_release(fs_info, btrfs_trans_num_extwriters);
btrfs_lockdep_release(fs_info, btrfs_trans_num_writers);
goto cleanup_transaction;
+
+lockdep_trans_commit_start_release:
+ btrfs_trans_state_lockdep_release(fs_info, BTRFS_LOCKDEP_TRANS_COMMIT_START);
+ btrfs_end_transaction(trans);
+ return ret;
}
/*
--
2.35.1
next prev parent reply other threads:[~2022-10-11 14:53 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-11 14:49 [PATCH AUTOSEL 6.0 01/46] arm64: dts: qcom: sdm845: narrow LLCC address space Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 02/46] ARM: dts: imx6: delete interrupts property if interrupts-extended is set Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 03/46] ARM: dts: imx7d-sdb: config the max pressure for tsc2046 Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 04/46] soc: mediatek: Let PMIC Wrapper and SCPSYS depend on OF Sasha Levin
2022-10-11 22:49 ` Jean Delvare
2022-10-16 14:46 ` Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 05/46] arm64: dts: qcom: sc7280-idp: correct ADC channel node name and unit address Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 06/46] ARM: dts: imx6q: add missing properties for sram Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 07/46] ARM: dts: imx6dl: " Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 08/46] ARM: dts: imx6qp: " Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 09/46] ARM: dts: imx6sl: " Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 10/46] ARM: dts: imx6sll: " Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 11/46] ARM: dts: imx6sx: " Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 12/46] ARM: dts: imx6sl: use tabs for code indent Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 13/46] ARM: dts: imx6sx-udoo-neo: don't use multiple blank lines Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 14/46] kselftest/arm64: Fix validatation termination record after EXTRA_CONTEXT Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 15/46] kselftest/arm64: Allow larger buffers in get_signal_context() Sasha Levin
2022-10-11 15:04 ` Mark Brown
2022-10-13 17:58 ` Sasha Levin
2022-10-13 18:02 ` Mark Brown
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 16/46] sparc: Fix the generic IO helpers Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 17/46] arm64: atomics: remove LL/SC trampolines Sasha Levin
2022-10-12 8:55 ` Catalin Marinas
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 18/46] arm64: run softirqs on the per-CPU IRQ stack Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 19/46] arm64: dts: imx8mm-kontron: Use the VSELECT signal to switch SD card IO voltage Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 20/46] arm64: dts: imx8ulp: no executable source file permission Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 21/46] arm64: dts: imx8mq-librem5: Add bq25895 as max17055's power supply Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 22/46] ARM: orion: fix include path Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 23/46] btrfs: dump extra info if one free space cache has more bitmaps than it should Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 24/46] btrfs: add macros for annotating wait events with lockdep Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 25/46] btrfs: add lockdep annotations for num_writers wait event Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 26/46] btrfs: add lockdep annotations for num_extwriters " Sasha Levin
2022-10-11 14:49 ` Sasha Levin [this message]
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 28/46] btrfs: add lockdep annotations for pending_ordered " Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 29/46] btrfs: change the lockdep class of free space inode's invalidate_lock Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 30/46] btrfs: add lockdep annotations for the ordered extents wait event Sasha Levin
2022-10-11 14:49 ` [PATCH AUTOSEL 6.0 31/46] btrfs: scrub: properly report super block errors in system log Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 32/46] btrfs: scrub: try to fix super block errors Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 33/46] btrfs: get rid of block group caching progress logic Sasha Levin
2022-10-11 23:46 ` Omar Sandoval
2022-10-13 17:55 ` Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 34/46] btrfs: don't print information about space cache or tree every remount Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 35/46] btrfs: call __btrfs_remove_free_space_cache_locked on cache load failure Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 36/46] btrfs: check superblock to ensure the fs was not modified at thaw time Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 37/46] btrfs: remove the unnecessary result variables Sasha Levin
2022-10-12 11:54 ` David Sterba
2022-10-13 17:56 ` Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 38/46] btrfs: introduce BTRFS_QGROUP_RUNTIME_FLAG_CANCEL_RESCAN Sasha Levin
2022-10-12 12:56 ` David Sterba
2022-10-12 23:12 ` Qu Wenruo
2022-10-13 17:56 ` Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 39/46] btrfs: add KCSAN annotations for unlocked access to block_rsv->full Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 40/46] btrfs: separate out the eb and extent state leak helpers Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 41/46] btrfs: relax block-group-tree feature dependency checks Sasha Levin
2022-10-12 13:01 ` David Sterba
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 42/46] arm64: dts: uniphier: Add USB-device support for PXs3 reference board Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 43/46] ARM: 9233/1: stacktrace: Skip frame pointer boundary check for call_with_stack() Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 44/46] ARM: 9234/1: stacktrace: Avoid duplicate saving of exception PC value Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 45/46] ARM: 9242/1: kasan: Only map modules if CONFIG_KASAN_VMALLOC=n Sasha Levin
2022-10-11 14:50 ` [PATCH AUTOSEL 6.0 46/46] selftests/cpu-hotplug: Use return instead of exit Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20221011145015.1622882-27-sashal@kernel.org \
--to=sashal@kernel.org \
--cc=clm@fb.com \
--cc=dsterba@suse.com \
--cc=iangelak@fb.com \
--cc=josef@toxicpanda.com \
--cc=linux-btrfs@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).