From: Shaohua Li <shli@kernel.org>
To: linux-raid@vger.kernel.org
Cc: neilb@suse.de, axboe@kernel.dk, dan.j.williams@intel.com,
shli@fusionio.com
Subject: [patch 4/9 v2] raid5: remove some device_lock locking places
Date: Tue, 19 Jun 2012 09:59:45 +0800 [thread overview]
Message-ID: <20120619020116.612175829@kernel.org> (raw)
In-Reply-To: 20120619015941.891400722@kernel.org
[-- Attachment #1: raid5-remove-some-lock.patch --]
[-- Type: text/plain, Size: 7128 bytes --]
With per-stripe lock and bi_phys_segments lockless, we can safely remove some
locking places of device_lock.
stripe ->read, ->toread ... are protected by per-stripe lock. Accessing bio
list of the stripe is always serialized by this lock. If bio in ->read,
->toread ... list are shared by multiple stripes, there are two protections:
1. bi_phys_segments acts as a reference count
2. traverse the list uses r5_next_bio, which makes traverse never access bio
not belonging to the stripe
Let's have an example:
| stripe1 | stripe2 | stripe3 |
...bio1......|bio2|bio3|....bio4.....
stripe2 has 4 bios, when it's finished, it will decrement bi_phys_segments for
all bios, but only end_bio for bio2 and bio3. bio1->bi_next still points to
bio2, but this doesn't matter. When stripe1 is finished, it will not touch bio2
because of r5_next_bio check. Next time stripe1 will end_bio for bio1 and
stripe3 will end_bio bio4.
before add_stripe_bio() addes a bio to a stripe, we already increament the bio
bi_phys_segments, so don't worry other stripes release the bio.
Signed-off-by: Shaohua Li <shli@fusionio.com>
---
drivers/md/raid5.c | 51 +++++++++++++++------------------------------------
1 file changed, 15 insertions(+), 36 deletions(-)
Index: linux/drivers/md/raid5.c
===================================================================
--- linux.orig/drivers/md/raid5.c 2012-06-13 14:08:04.000000000 +0800
+++ linux/drivers/md/raid5.c 2012-06-13 14:12:23.989189085 +0800
@@ -747,15 +747,13 @@ static void ops_complete_biofill(void *s
{
struct stripe_head *sh = stripe_head_ref;
struct bio *return_bi = NULL;
- struct r5conf *conf = sh->raid_conf;
int i;
pr_debug("%s: stripe %llu\n", __func__,
(unsigned long long)sh->sector);
/* clear completed biofills */
- spin_lock_irq(&conf->device_lock);
- spin_lock(&sh->stripe_lock);
+ spin_lock_irq(&sh->stripe_lock);
for (i = sh->disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
@@ -781,8 +779,7 @@ static void ops_complete_biofill(void *s
}
}
}
- spin_unlock(&sh->stripe_lock);
- spin_unlock_irq(&conf->device_lock);
+ spin_unlock_irq(&sh->stripe_lock);
clear_bit(STRIPE_BIOFILL_RUN, &sh->state);
return_io(return_bi);
@@ -794,7 +791,6 @@ static void ops_complete_biofill(void *s
static void ops_run_biofill(struct stripe_head *sh)
{
struct dma_async_tx_descriptor *tx = NULL;
- struct r5conf *conf = sh->raid_conf;
struct async_submit_ctl submit;
int i;
@@ -805,12 +801,10 @@ static void ops_run_biofill(struct strip
struct r5dev *dev = &sh->dev[i];
if (test_bit(R5_Wantfill, &dev->flags)) {
struct bio *rbi;
- spin_lock_irq(&conf->device_lock);
- spin_lock(&sh->stripe_lock);
+ spin_lock_irq(&sh->stripe_lock);
dev->read = rbi = dev->toread;
dev->toread = NULL;
- spin_unlock(&sh->stripe_lock);
- spin_unlock_irq(&conf->device_lock);
+ spin_unlock_irq(&sh->stripe_lock);
while (rbi && rbi->bi_sector <
dev->sector + STRIPE_SECTORS) {
tx = async_copy_data(0, rbi, dev->page,
@@ -1146,14 +1140,12 @@ ops_run_biodrain(struct stripe_head *sh,
if (test_and_clear_bit(R5_Wantdrain, &dev->flags)) {
struct bio *wbi;
- spin_lock_irq(&sh->raid_conf->device_lock);
- spin_lock(&sh->stripe_lock);
+ spin_lock_irq(&sh->stripe_lock);
chosen = dev->towrite;
dev->towrite = NULL;
BUG_ON(dev->written);
wbi = dev->written = chosen;
- spin_unlock(&sh->stripe_lock);
- spin_unlock_irq(&sh->raid_conf->device_lock);
+ spin_unlock_irq(&sh->stripe_lock);
while (wbi && wbi->bi_sector <
dev->sector + STRIPE_SECTORS) {
@@ -2340,8 +2332,7 @@ static int add_stripe_bio(struct stripe_
(unsigned long long)sh->sector);
- spin_lock_irq(&conf->device_lock);
- spin_lock(&sh->stripe_lock);
+ spin_lock_irq(&sh->stripe_lock);
if (forwrite) {
bip = &sh->dev[dd_idx].towrite;
if (*bip == NULL && sh->dev[dd_idx].written == NULL)
@@ -2375,8 +2366,7 @@ static int add_stripe_bio(struct stripe_
if (sector >= sh->dev[dd_idx].sector + STRIPE_SECTORS)
set_bit(R5_OVERWRITE, &sh->dev[dd_idx].flags);
}
- spin_unlock(&sh->stripe_lock);
- spin_unlock_irq(&conf->device_lock);
+ spin_unlock_irq(&sh->stripe_lock);
pr_debug("added bi b#%llu to stripe s#%llu, disk %d.\n",
(unsigned long long)(*bip)->bi_sector,
@@ -2392,8 +2382,7 @@ static int add_stripe_bio(struct stripe_
overlap:
set_bit(R5_Overlap, &sh->dev[dd_idx].flags);
- spin_unlock(&sh->stripe_lock);
- spin_unlock_irq(&conf->device_lock);
+ spin_unlock_irq(&sh->stripe_lock);
return 0;
}
@@ -2443,8 +2432,7 @@ handle_failed_stripe(struct r5conf *conf
rdev_dec_pending(rdev, conf->mddev);
}
}
- spin_lock_irq(&conf->device_lock);
- spin_lock(&sh->stripe_lock);
+ spin_lock_irq(&sh->stripe_lock);
/* fail all writes first */
bi = sh->dev[i].towrite;
sh->dev[i].towrite = NULL;
@@ -2506,8 +2494,7 @@ handle_failed_stripe(struct r5conf *conf
bi = nextbi;
}
}
- spin_unlock(&sh->stripe_lock);
- spin_unlock_irq(&conf->device_lock);
+ spin_unlock_irq(&sh->stripe_lock);
if (bitmap_end)
bitmap_endwrite(conf->mddev->bitmap, sh->sector,
STRIPE_SECTORS, 0, 0);
@@ -2713,8 +2700,7 @@ static void handle_stripe_clean_event(st
struct bio *wbi, *wbi2;
int bitmap_end = 0;
pr_debug("Return write for disc %d\n", i);
- spin_lock_irq(&conf->device_lock);
- spin_lock(&sh->stripe_lock);
+ spin_lock_irq(&sh->stripe_lock);
wbi = dev->written;
dev->written = NULL;
while (wbi && wbi->bi_sector <
@@ -2729,8 +2715,7 @@ static void handle_stripe_clean_event(st
}
if (dev->towrite == NULL)
bitmap_end = 1;
- spin_unlock(&sh->stripe_lock);
- spin_unlock_irq(&conf->device_lock);
+ spin_unlock_irq(&sh->stripe_lock);
if (bitmap_end)
bitmap_endwrite(conf->mddev->bitmap,
sh->sector,
@@ -3188,8 +3173,7 @@ static void analyse_stripe(struct stripe
/* Now to look around and see what can be done */
rcu_read_lock();
- spin_lock_irq(&conf->device_lock);
- spin_lock(&sh->stripe_lock);
+ spin_lock_irq(&sh->stripe_lock);
for (i=disks; i--; ) {
struct md_rdev *rdev;
sector_t first_bad;
@@ -3335,8 +3319,7 @@ static void analyse_stripe(struct stripe
do_recovery = 1;
}
}
- spin_unlock(&sh->stripe_lock);
- spin_unlock_irq(&conf->device_lock);
+ spin_unlock_irq(&sh->stripe_lock);
if (test_bit(STRIPE_SYNCING, &sh->state)) {
/* If there is a failed device being replaced,
* we must be recovering.
@@ -4127,9 +4110,7 @@ static void make_request(struct mddev *m
if (!plugged)
md_wakeup_thread(mddev->thread);
- spin_lock_irq(&conf->device_lock);
remaining = raid5_dec_bi_active_stripes(bi);
- spin_unlock_irq(&conf->device_lock);
if (remaining == 0) {
if ( rw == WRITE )
@@ -4509,9 +4490,7 @@ static int retry_aligned_read(struct r5
release_stripe(sh);
handled++;
}
- spin_lock_irq(&conf->device_lock);
remaining = raid5_dec_bi_active_stripes(raid_bio);
- spin_unlock_irq(&conf->device_lock);
if (remaining == 0)
bio_endio(raid_bio, 0);
if (atomic_dec_and_test(&conf->active_aligned_reads))
next prev parent reply other threads:[~2012-06-19 1:59 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-06-19 1:59 [patch 0/9 v2] raid5: improve write performance for fast storage Shaohua Li
2012-06-19 1:59 ` [patch 1/9 v2] raid5: use wake_up_all for overlap waking Shaohua Li
2012-06-19 1:59 ` [patch 2/9 v2] raid5: add a per-stripe lock Shaohua Li
2012-06-19 1:59 ` [patch 3/9 v2] raid5: lockless access raid5 overrided bi_phys_segments Shaohua Li
2012-06-21 1:56 ` Dan Williams
2012-06-19 1:59 ` Shaohua Li [this message]
2012-06-19 1:59 ` [patch 5/9 v2] raid5: reduce chance release_stripe() taking device_lock Shaohua Li
2012-06-21 0:56 ` Dan Williams
2012-06-21 1:34 ` Shaohua Li
2012-06-19 1:59 ` [patch 6/9 v2] md: personality can provide unplug private data Shaohua Li
2012-06-19 1:59 ` [patch 7/9 v2] raid5: make_request use batch stripe release Shaohua Li
2012-06-19 1:59 ` [patch 8/9 v2] raid5: raid5d handle stripe in batch way Shaohua Li
2012-06-19 1:59 ` [patch 9/9 v2] raid5: create multiple threads to handle stripes Shaohua Li
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=20120619020116.612175829@kernel.org \
--to=shli@kernel.org \
--cc=axboe@kernel.dk \
--cc=dan.j.williams@intel.com \
--cc=linux-raid@vger.kernel.org \
--cc=neilb@suse.de \
--cc=shli@fusionio.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.