From: Coly Li colyli@suse.de
stable inclusion from stable-v5.10.150 commit c263516c2c20df9c29f33baeb4a817af7212fb69 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I6D0XA
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=...
--------------------------------
[ Upstream commit d2d05b88035d2d51a5bb6c5afec88a0880c73df4 ]
Inside set_at_max_writeback_rate() the calculation in following if() check is wrong, if (atomic_inc_return(&c->idle_counter) < atomic_read(&c->attached_dev_nr) * 6)
Because each attached backing device has its own writeback thread running and increasing c->idle_counter, the counter increates much faster than expected. The correct calculation should be, (counter / dev_nr) < dev_nr * 6 which equals to, counter < dev_nr * dev_nr * 6
This patch fixes the above mistake with correct calculation, and helper routine idle_counter_exceeded() is added to make code be more clear.
Reported-by: Mingzhe Zou mingzhe.zou@easystack.cn Signed-off-by: Coly Li colyli@suse.de Acked-by: Mingzhe Zou mingzhe.zou@easystack.cn Link: https://lore.kernel.org/r/20220919161647.81238-6-colyli@suse.de Signed-off-by: Jens Axboe axboe@kernel.dk Signed-off-by: Sasha Levin sashal@kernel.org Signed-off-by: Jialin Zhang zhangjialin11@huawei.com --- drivers/md/bcache/writeback.c | 73 +++++++++++++++++++++++++---------- 1 file changed, 52 insertions(+), 21 deletions(-)
diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c index 3e879e985373..83dcdd85abc0 100644 --- a/drivers/md/bcache/writeback.c +++ b/drivers/md/bcache/writeback.c @@ -157,6 +157,53 @@ static void __update_writeback_rate(struct cached_dev *dc) dc->writeback_rate_target = target; }
+static bool idle_counter_exceeded(struct cache_set *c) +{ + int counter, dev_nr; + + /* + * If c->idle_counter is overflow (idel for really long time), + * reset as 0 and not set maximum rate this time for code + * simplicity. + */ + counter = atomic_inc_return(&c->idle_counter); + if (counter <= 0) { + atomic_set(&c->idle_counter, 0); + return false; + } + + dev_nr = atomic_read(&c->attached_dev_nr); + if (dev_nr == 0) + return false; + + /* + * c->idle_counter is increased by writeback thread of all + * attached backing devices, in order to represent a rough + * time period, counter should be divided by dev_nr. + * Otherwise the idle time cannot be larger with more backing + * device attached. + * The following calculation equals to checking + * (counter / dev_nr) < (dev_nr * 6) + */ + if (counter < (dev_nr * dev_nr * 6)) + return false; + + return true; +} + +/* + * Idle_counter is increased every time when update_writeback_rate() is + * called. If all backing devices attached to the same cache set have + * identical dc->writeback_rate_update_seconds values, it is about 6 + * rounds of update_writeback_rate() on each backing device before + * c->at_max_writeback_rate is set to 1, and then max wrteback rate set + * to each dc->writeback_rate.rate. + * In order to avoid extra locking cost for counting exact dirty cached + * devices number, c->attached_dev_nr is used to calculate the idle + * throushold. It might be bigger if not all cached device are in write- + * back mode, but it still works well with limited extra rounds of + * update_writeback_rate(). + */ static bool set_at_max_writeback_rate(struct cache_set *c, struct cached_dev *dc) { @@ -167,21 +214,8 @@ static bool set_at_max_writeback_rate(struct cache_set *c, /* Don't set max writeback rate if gc is running */ if (!c->gc_mark_valid) return false; - /* - * Idle_counter is increased everytime when update_writeback_rate() is - * called. If all backing devices attached to the same cache set have - * identical dc->writeback_rate_update_seconds values, it is about 6 - * rounds of update_writeback_rate() on each backing device before - * c->at_max_writeback_rate is set to 1, and then max wrteback rate set - * to each dc->writeback_rate.rate. - * In order to avoid extra locking cost for counting exact dirty cached - * devices number, c->attached_dev_nr is used to calculate the idle - * throushold. It might be bigger if not all cached device are in write- - * back mode, but it still works well with limited extra rounds of - * update_writeback_rate(). - */ - if (atomic_inc_return(&c->idle_counter) < - atomic_read(&c->attached_dev_nr) * 6) + + if (!idle_counter_exceeded(c)) return false;
if (atomic_read(&c->at_max_writeback_rate) != 1) @@ -195,13 +229,10 @@ static bool set_at_max_writeback_rate(struct cache_set *c, dc->writeback_rate_change = 0;
/* - * Check c->idle_counter and c->at_max_writeback_rate agagain in case - * new I/O arrives during before set_at_max_writeback_rate() returns. - * Then the writeback rate is set to 1, and its new value should be - * decided via __update_writeback_rate(). + * In case new I/O arrives during before + * set_at_max_writeback_rate() returns. */ - if ((atomic_read(&c->idle_counter) < - atomic_read(&c->attached_dev_nr) * 6) || + if (!idle_counter_exceeded(c) || !atomic_read(&c->at_max_writeback_rate)) return false;