On 2017/6/9 上午8:59, tang.junhui@xxxxxxxxxx wrote:
> Hello Coly,
>
>
> ＞ First we need to know why
> ＞ when writeback cashe is full but sectors_to_gc ＜ 0
>
> Some times, the writeback cashe is not full, we just wanna execute
>
> this commands to get more available buckets, but this command
>
> "echo 1 ＞ trigger_gc" does not works, this is the issue that the patch
>
> fixed.
Oh I see the point! Yes it's useful.
A question is, this patch changes expected behavior of trigger_gc. This
modification may change I/O load to existing installations. How about
add a new sysfs entry like trigger_gc_force ? And add comments in source
code to explain this change is necessary.
IMHO this is quite an important fix, thanks for your effort :-)
Coly
>
> 原始邮件
> *发件人：*＜i@xxxxxxx＞;
> *收件人：*唐军辉10074136;＜kent.overstreet@xxxxxxxxx＞;
> ＜bcache@xxxxxxxxxxxxxxxxxx＞;
> *抄送人：*＜linux-bcache@xxxxxxxxxxxxxxx＞;
> *日 期 ：*2017年06月08日 23:54
> *主 题 ：**Re: [PATCH] bcache: gc does not work when triggering by
> manualcommand*
>
>
> On 2017/6/8 下午2:39, tang.junhui@xxxxxxxxxx wrote:
> ＞ From: Tang Junhui ＜tang.junhui@xxxxxxxxxx＞
> ＞
> ＞ I try to execute the following command to trigger gc thread:
> ＞ [root@localhost internal]# echo 1 ＞ trigger_gc
> ＞ But it does not work, I debug the code in gc_should_run(), It works only
> ＞ if in invalidating or sectors_to_gc ＜ 0. So set sectors_to_gc to -1 to
> ＞ meet the condition when we trigger gc by manual command.
> ＞
> ＞ Signed-off-by: Tang Junhui ＜tang.junhui@xxxxxxxxxx＞
>
> This fix may make things more complicated. First we need to know why
> when writeback cashe is full but sectors_to_gc ＜ 0 ? If we don't find
> out the root reason, this fix is not convinced.
>
> Thanks.
>
> Coly
>
>
> ＞ ---
> ＞ drivers/md/bcache/btree.h | 6 ++++++
> ＞ drivers/md/bcache/sysfs.c | 2 +-
> ＞ 2 files changed, 7 insertions(+), 1 deletion(-)
> ＞
> ＞ diff --git a/drivers/md/bcache/btree.h b/drivers/md/bcache/btree.h
> ＞ index 9b80417..bab07d4 100644
> ＞ --- a/drivers/md/bcache/btree.h
> ＞ +++ b/drivers/md/bcache/btree.h
> ＞ @@ -263,6 +263,12 @@ static inline void wake_up_gc(struct cache_set *c)
> ＞ wake_up(&c-＞gc_wait);
> ＞ }
> ＞
> ＞ +static inline void wake_up_gc_force(struct cache_set *c)
> ＞ +{
> ＞ + atomic_set(&c-＞sectors_to_gc, -1);
> ＞ + wake_up(&c-＞gc_wait);
> ＞ +}
> ＞ +
> ＞ #define MAP_DONE 0
> ＞ #define MAP_CONTINUE 1
> ＞
> ＞ diff --git a/drivers/md/bcache/sysfs.c b/drivers/md/bcache/sysfs.c
> ＞ index b3ff57d..5fdb86c 100644
> ＞ --- a/drivers/md/bcache/sysfs.c
> ＞ +++ b/drivers/md/bcache/sysfs.c
> ＞ @@ -615,7 +615,7 @@ STORE(__bch_cache_set)
> ＞ }
> ＞
> ＞ if (attr == &sysfs_trigger_gc)
> ＞ - wake_up_gc(c);
> ＞ + wake_up_gc_force(c);
> ＞
> ＞ if (attr == &sysfs_prune_cache) {
> ＞ struct shrink_control sc;
> ＞
>
>
>
--
Coly Li
--
To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
