From: Benjamin Marzinski <bmarzins@redhat.com>
To: Mikulas Patocka <mpatocka@redhat.com>,
Mike Snitzer <snitzer@kernel.org>,
Heinz Mauelshagen <mauelshagen@redhat.com>
Cc: dm-devel@lists.linux.dev
Subject: [PATCH 2/3] dm-delay: change locking to avoid contention
Date: Tue, 7 May 2024 17:16:24 -0400 [thread overview]
Message-ID: <20240507211626.26706-3-bmarzins@redhat.com> (raw)
In-Reply-To: <20240507211626.26706-1-bmarzins@redhat.com>
The delayed_bios list is protected by one mutex shared by all dm-delay
devices. This mutex must be held whenever a bio is added or expired bios
are removed from the list. Since a large number of expired bios could
be on the list, flush_delayed_bios() can schedule while holding the
mutex. This means a flush_delayed_bios() call on any dm-delay device can
slow down delay_map() calls on any other dm-delay device.
To keep dm-delay devices from slowing each other down and keep
processing delay bios from slowing adding delayed bios, the global mutex
has been removed, and each dm-delay device now has two locks.
delayed_bios_lock is a spinlock that must be held whenever the
delayed_bios list is accessed. process_bios_lock is a mutex that must be
held whenever a process has temporarily pulled bios off the delayed_bios
list to check which ones should be processed. It must be held until all
the bios that won't be processed are returned to the list. This is what
flush_delayed_bios() now does. The mutex is necessary to guarantee that
delay_presuspend() sees the entire list of delayed bios when it calls
flush_delayed_bios().
Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com>
---
drivers/md/dm-delay.c | 36 +++++++++++++++++++++++-------------
1 file changed, 23 insertions(+), 13 deletions(-)
diff --git a/drivers/md/dm-delay.c b/drivers/md/dm-delay.c
index 584e3b10802d..b1829c106a37 100644
--- a/drivers/md/dm-delay.c
+++ b/drivers/md/dm-delay.c
@@ -29,6 +29,8 @@ struct delay_class {
struct delay_c {
struct timer_list delay_timer;
struct mutex timer_lock;
+ struct mutex process_bios_lock; /* hold while removing bios to be processed from list */
+ spinlock_t delayed_bios_lock; /* hold on all accesses to delayed_bios list */
struct workqueue_struct *kdelayd_wq;
struct work_struct flush_expired_bios;
struct list_head delayed_bios;
@@ -49,8 +51,6 @@ struct dm_delay_info {
unsigned long expires;
};
-static DEFINE_MUTEX(delayed_bios_lock);
-
static void handle_delayed_timer(struct timer_list *t)
{
struct delay_c *dc = from_timer(dc, t, delay_timer);
@@ -89,12 +89,16 @@ static void flush_delayed_bios(struct delay_c *dc, bool flush_all)
{
struct dm_delay_info *delayed, *next;
struct bio_list flush_bio_list;
+ LIST_HEAD(local_list);
unsigned long next_expires = 0;
bool start_timer = false;
bio_list_init(&flush_bio_list);
- mutex_lock(&delayed_bios_lock);
- list_for_each_entry_safe(delayed, next, &dc->delayed_bios, list) {
+ mutex_lock(&dc->process_bios_lock);
+ spin_lock(&dc->delayed_bios_lock);
+ list_replace_init(&dc->delayed_bios, &local_list);
+ spin_unlock(&dc->delayed_bios_lock);
+ list_for_each_entry_safe(delayed, next, &local_list, list) {
cond_resched();
if (flush_all || time_after_eq(jiffies, delayed->expires)) {
struct bio *bio = dm_bio_from_per_bio_data(delayed,
@@ -114,7 +118,10 @@ static void flush_delayed_bios(struct delay_c *dc, bool flush_all)
}
}
}
- mutex_unlock(&delayed_bios_lock);
+ spin_lock(&dc->delayed_bios_lock);
+ list_splice(&local_list, &dc->delayed_bios);
+ spin_unlock(&dc->delayed_bios_lock);
+ mutex_unlock(&dc->process_bios_lock);
if (start_timer)
queue_timeout(dc, next_expires);
@@ -128,13 +135,13 @@ static int flush_worker_fn(void *data)
while (!kthread_should_stop()) {
flush_delayed_bios(dc, false);
- mutex_lock(&delayed_bios_lock);
+ spin_lock(&dc->delayed_bios_lock);
if (unlikely(list_empty(&dc->delayed_bios))) {
set_current_state(TASK_INTERRUPTIBLE);
- mutex_unlock(&delayed_bios_lock);
+ spin_unlock(&dc->delayed_bios_lock);
schedule();
} else {
- mutex_unlock(&delayed_bios_lock);
+ spin_unlock(&dc->delayed_bios_lock);
cond_resched();
}
}
@@ -168,6 +175,7 @@ static void delay_dtr(struct dm_target *ti)
if (dc->worker)
kthread_stop(dc->worker);
+ mutex_destroy(&dc->process_bios_lock);
mutex_destroy(&dc->timer_lock);
kfree(dc);
@@ -227,6 +235,8 @@ static int delay_ctr(struct dm_target *ti, unsigned int argc, char **argv)
ti->private = dc;
INIT_LIST_HEAD(&dc->delayed_bios);
mutex_init(&dc->timer_lock);
+ mutex_init(&dc->process_bios_lock);
+ spin_lock_init(&dc->delayed_bios_lock);
dc->may_delay = true;
dc->argc = argc;
@@ -310,14 +320,14 @@ static int delay_bio(struct delay_c *dc, struct delay_class *c, struct bio *bio)
delayed->context = dc;
delayed->expires = expires = jiffies + msecs_to_jiffies(c->delay);
- mutex_lock(&delayed_bios_lock);
+ spin_lock(&dc->delayed_bios_lock);
if (unlikely(!dc->may_delay)) {
- mutex_unlock(&delayed_bios_lock);
+ spin_unlock(&dc->delayed_bios_lock);
return DM_MAPIO_REMAPPED;
}
c->ops++;
list_add_tail(&delayed->list, &dc->delayed_bios);
- mutex_unlock(&delayed_bios_lock);
+ spin_unlock(&dc->delayed_bios_lock);
if (delay_is_fast(dc))
wake_up_process(dc->worker);
@@ -331,9 +341,9 @@ static void delay_presuspend(struct dm_target *ti)
{
struct delay_c *dc = ti->private;
- mutex_lock(&delayed_bios_lock);
+ spin_lock(&dc->delayed_bios_lock);
dc->may_delay = false;
- mutex_unlock(&delayed_bios_lock);
+ spin_unlock(&dc->delayed_bios_lock);
if (!delay_is_fast(dc))
timer_delete(&dc->delay_timer);
--
2.45.0
next prev parent reply other threads:[~2024-05-07 21:16 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-05-07 21:16 [PATCH 0/3] dm-delay synchronization cleanup Benjamin Marzinski
2024-05-07 21:16 ` [PATCH 1/3] dm-delay: fix workqueue delay_timer race Benjamin Marzinski
2024-05-07 21:16 ` Benjamin Marzinski [this message]
2024-05-07 21:16 ` [PATCH 3/3] dm-delay: remove timer_lock Benjamin Marzinski
2024-05-09 14:24 ` [PATCH 0/3] dm-delay synchronization cleanup Mike Snitzer
2024-05-09 16:07 ` Benjamin Marzinski
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=20240507211626.26706-3-bmarzins@redhat.com \
--to=bmarzins@redhat.com \
--cc=dm-devel@lists.linux.dev \
--cc=mauelshagen@redhat.com \
--cc=mpatocka@redhat.com \
--cc=snitzer@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 read-only IMAP folder(s) and NNTP newsgroup(s).