public inbox for linux-bcache@vger.kernel.org
 help / color / mirror / Atom feed
From: Hannes Reinecke <hare@suse.de>
To: Coly Li <colyli@suse.de>, linux-bcache@vger.kernel.org
Cc: linux-block@vger.kernel.org
Subject: Re: [PATCH 02/14] bcache: explicitly make cache_set only have single cache
Date: Mon, 17 Aug 2020 08:11:28 +0200	[thread overview]
Message-ID: <aaf411c3-b58d-91b8-188e-54856ae795e7@suse.de> (raw)
In-Reply-To: <20200815041043.45116-3-colyli@suse.de>

On 8/15/20 6:10 AM, Coly Li wrote:
> Currently although the bcache code has a framework for multiple caches
> in a cache set, but indeed the multiple caches never completed and users
> use md raid1 for multiple copies of the cached data.
> 
> This patch does the following change in struct cache_set, to explicitly
> make a cache_set only have single cache,
> - Change pointer array "*cache[MAX_CACHES_PER_SET]" to a single pointer
>    "*cache".
> - Remove pointer array "*cache_by_alloc[MAX_CACHES_PER_SET]".
> - Remove "caches_loaded".
> 
> Now the code looks as exactly what it does in practic: only one cache is
> used in the cache set.
> 
> Signed-off-by: Coly Li <colyli@suse.de>
> ---
>   drivers/md/bcache/alloc.c  |  2 +-
>   drivers/md/bcache/bcache.h |  8 +++-----
>   drivers/md/bcache/super.c  | 19 ++++++++-----------
>   3 files changed, 12 insertions(+), 17 deletions(-)
> 
> diff --git a/drivers/md/bcache/alloc.c b/drivers/md/bcache/alloc.c
> index 4493ff57476d..3385f6add6df 100644
> --- a/drivers/md/bcache/alloc.c
> +++ b/drivers/md/bcache/alloc.c
> @@ -501,7 +501,7 @@ int __bch_bucket_alloc_set(struct cache_set *c, unsigned int reserve,
>   
>   	bkey_init(k);
>   
> -	ca = c->cache_by_alloc[0];
> +	ca = c->cache;
>   	b = bch_bucket_alloc(ca, reserve, wait);
>   	if (b == -1)
>   		goto err;

Maybe drop variable 'ca' altogether?

> diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
> index 5ff6e9573935..aa112c1adba1 100644
> --- a/drivers/md/bcache/bcache.h
> +++ b/drivers/md/bcache/bcache.h
> @@ -519,9 +519,7 @@ struct cache_set {
>   
>   	struct cache_sb		sb;
>   
> -	struct cache		*cache[MAX_CACHES_PER_SET];
> -	struct cache		*cache_by_alloc[MAX_CACHES_PER_SET];
> -	int			caches_loaded;
> +	struct cache		*cache;
>   
>   	struct bcache_device	**devices;
>   	unsigned int		devices_max_used;
> @@ -808,7 +806,7 @@ static inline struct cache *PTR_CACHE(struct cache_set *c,
>   				      const struct bkey *k,
>   				      unsigned int ptr)
>   {
> -	return c->cache[PTR_DEV(k, ptr)];
> +	return c->cache;
>   }
>   
>   static inline size_t PTR_BUCKET_NR(struct cache_set *c,
> @@ -890,7 +888,7 @@ do {									\
>   /* Looping macros */
>   
>   #define for_each_cache(ca, cs, iter)					\
> -	for (iter = 0; ca = cs->cache[iter], iter < (cs)->sb.nr_in_set; iter++)
> +	for (iter = 0; ca = cs->cache, iter < 1; iter++)
>   
>   #define for_each_bucket(b, ca)						\
>   	for (b = (ca)->buckets + (ca)->sb.first_bucket;			\

I guess 'for_each_cache' can be killed, too.

> diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
> index 7057ec48f3d1..e9ccfa17beb8 100644
> --- a/drivers/md/bcache/super.c
> +++ b/drivers/md/bcache/super.c
> @@ -1675,7 +1675,7 @@ static void cache_set_free(struct closure *cl)
>   	for_each_cache(ca, c, i)
>   		if (ca) {
>   			ca->set = NULL;
> -			c->cache[ca->sb.nr_this_dev] = NULL;
> +			c->cache = NULL;
>   			kobject_put(&ca->kobj);
>   		}
>   
> @@ -2166,7 +2166,7 @@ static const char *register_cache_set(struct cache *ca)
>   
>   	list_for_each_entry(c, &bch_cache_sets, list)
>   		if (!memcmp(c->sb.set_uuid, ca->sb.set_uuid, 16)) {
> -			if (c->cache[ca->sb.nr_this_dev])
> +			if (c->cache)
>   				return "duplicate cache set member";
>   
>   			if (!can_attach_cache(ca, c))
> @@ -2216,14 +2216,11 @@ static const char *register_cache_set(struct cache *ca)
>   
>   	kobject_get(&ca->kobj);
>   	ca->set = c;
> -	ca->set->cache[ca->sb.nr_this_dev] = ca;
> -	c->cache_by_alloc[c->caches_loaded++] = ca;
> +	ca->set->cache = ca;
>   
> -	if (c->caches_loaded == c->sb.nr_in_set) {
> -		err = "failed to run cache set";
> -		if (run_cache_set(c) < 0)
> -			goto err;
> -	}
> +	err = "failed to run cache set";
> +	if (run_cache_set(c) < 0)
> +		goto err;
>   
>   	return NULL;
>   err:
> @@ -2240,8 +2237,8 @@ void bch_cache_release(struct kobject *kobj)
>   	unsigned int i;
>   
>   	if (ca->set) {
> -		BUG_ON(ca->set->cache[ca->sb.nr_this_dev] != ca);
> -		ca->set->cache[ca->sb.nr_this_dev] = NULL;
> +		BUG_ON(ca->set->cache != ca);
> +		ca->set->cache = NULL;
>   	}
>   
>   	free_pages((unsigned long) ca->disk_buckets, ilog2(meta_bucket_pages(&ca->sb)));
> 
Cheers,

Hannes
-- 
Dr. Hannes Reinecke            Teamlead Storage & Networking
hare@suse.de                               +49 911 74053 688
SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer

  reply	other threads:[~2020-08-17  6:11 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-15  4:10 [PATCH 00/14] bcache: remove multiple caches code framework Coly Li
2020-08-15  4:10 ` [PATCH 01/14] bcache: remove 'int n' from parameter list of bch_bucket_alloc_set() Coly Li
2020-08-17  6:08   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 02/14] bcache: explicitly make cache_set only have single cache Coly Li
2020-08-17  6:11   ` Hannes Reinecke [this message]
2020-08-15  4:10 ` [PATCH 03/14] bcache: remove for_each_cache() Coly Li
2020-08-17  6:13   ` Hannes Reinecke
2020-08-22 11:40     ` Coly Li
2020-08-15  4:10 ` [PATCH 04/14] bcache: add set_uuid in struct cache_set Coly Li
2020-08-17  6:15   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 05/14] bcache: only use block_bytes() on struct cache Coly Li
2020-08-17  6:16   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 06/14] bcache: remove useless alloc_bucket_pages() Coly Li
2020-08-17  6:17   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 07/14] bcache: remove useless bucket_pages() Coly Li
2020-08-17  6:18   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 08/14] bcache: only use bucket_bytes() on struct cache Coly Li
2020-08-17  6:19   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 09/14] bcache: avoid data copy between cache_set->sb and cache->sb Coly Li
2020-08-17  6:22   ` Hannes Reinecke
2020-08-17  6:30     ` Coly Li
2020-08-15  4:10 ` [PATCH 10/14] bcache: don't check seq numbers in register_cache_set() Coly Li
2020-08-17  6:23   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 11/14] bcache: remove can_attach_cache() Coly Li
2020-08-17  6:24   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 12/14] bcache: check and set sync status on cache's in-memory super block Coly Li
2020-08-17  6:25   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 13/14] bcache: remove embedded struct cache_sb from struct cache_set Coly Li
2020-08-17  6:26   ` Hannes Reinecke
2020-08-15  4:10 ` [PATCH 14/14] bcache: move struct cache_sb out of uapi bcache.h Coly Li
2020-08-17  6:36   ` Hannes Reinecke
2020-08-17  7:10     ` Coly Li
2020-08-17 10:28   ` Christoph Hellwig
2020-08-17 11:48     ` Coly 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=aaf411c3-b58d-91b8-188e-54856ae795e7@suse.de \
    --to=hare@suse.de \
    --cc=colyli@suse.de \
    --cc=linux-bcache@vger.kernel.org \
    --cc=linux-block@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