From: "J. Bruce Fields" <bfields@fieldses.org>
To: NeilBrown <neilb@suse.de>
Cc: linux-nfs@vger.kernel.org
Subject: Re: [PATCH 3/7] sunrpc/cache: change deferred-request hash table to use hlist.
Date: Tue, 21 Sep 2010 22:59:44 -0400 [thread overview]
Message-ID: <20100922025944.GC22021@fieldses.org> (raw)
In-Reply-To: <20100922025506.31745.9409.stgit@localhost.localdomain>
On Wed, Sep 22, 2010 at 12:55:07PM +1000, NeilBrown wrote:
> Being a hash table, hlist is the best option.
>
> There is currently some ugliness were we treat "->next == NULL" as
> a special case to avoid having to initialise the whole array.
> This change nicely gets rid of that case.
This (or some hopefully correct variant!) should already be applied;
could you check that it looks right?
--b.
>
> Signed-off-by: NeilBrown <neilb@suse.de>
> ---
> include/linux/sunrpc/cache.h | 2 +-
> net/sunrpc/cache.c | 22 ++++++++--------------
> 2 files changed, 9 insertions(+), 15 deletions(-)
>
> diff --git a/include/linux/sunrpc/cache.h b/include/linux/sunrpc/cache.h
> index 52a7d72..0349635 100644
> --- a/include/linux/sunrpc/cache.h
> +++ b/include/linux/sunrpc/cache.h
> @@ -133,7 +133,7 @@ struct cache_req {
> * delayed awaiting cache-fill
> */
> struct cache_deferred_req {
> - struct list_head hash; /* on hash chain */
> + struct hlist_node hash; /* on hash chain */
> struct list_head recent; /* on fifo */
> struct cache_head *item; /* cache item we wait on */
> void *owner; /* we might need to discard all defered requests
> diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c
> index a9e850e..466f65c 100644
> --- a/net/sunrpc/cache.c
> +++ b/net/sunrpc/cache.c
> @@ -506,13 +506,13 @@ EXPORT_SYMBOL_GPL(cache_purge);
>
> static DEFINE_SPINLOCK(cache_defer_lock);
> static LIST_HEAD(cache_defer_list);
> -static struct list_head cache_defer_hash[DFR_HASHSIZE];
> +static struct hlist_head cache_defer_hash[DFR_HASHSIZE];
> static int cache_defer_cnt;
>
> static void __unhash_deferred_req(struct cache_deferred_req *dreq)
> {
> list_del_init(&dreq->recent);
> - list_del_init(&dreq->hash);
> + hlist_del_init(&dreq->hash);
> cache_defer_cnt--;
> }
>
> @@ -521,9 +521,7 @@ static void __hash_deferred_req(struct cache_deferred_req *dreq, struct cache_he
> int hash = DFR_HASH(item);
>
> list_add(&dreq->recent, &cache_defer_list);
> - if (cache_defer_hash[hash].next == NULL)
> - INIT_LIST_HEAD(&cache_defer_hash[hash]);
> - list_add(&dreq->hash, &cache_defer_hash[hash]);
> + hlist_add_head(&dreq->hash, &cache_defer_hash[hash]);
> }
>
> static int setup_deferral(struct cache_deferred_req *dreq, struct cache_head *item)
> @@ -587,7 +585,7 @@ static int cache_wait_req(struct cache_req *req, struct cache_head *item)
> * to clean up
> */
> spin_lock(&cache_defer_lock);
> - if (!list_empty(&sleeper.handle.hash)) {
> + if (!hlist_unhashed(&sleeper.handle.hash)) {
> __unhash_deferred_req(&sleeper.handle);
> spin_unlock(&cache_defer_lock);
> } else {
> @@ -642,23 +640,19 @@ static void cache_revisit_request(struct cache_head *item)
> struct cache_deferred_req *dreq;
> struct list_head pending;
>
> - struct list_head *lp;
> + struct hlist_node *lp, *tmp;
> int hash = DFR_HASH(item);
>
> INIT_LIST_HEAD(&pending);
> spin_lock(&cache_defer_lock);
>
> - lp = cache_defer_hash[hash].next;
> - if (lp) {
> - while (lp != &cache_defer_hash[hash]) {
> - dreq = list_entry(lp, struct cache_deferred_req, hash);
> - lp = lp->next;
> + hlist_for_each_entry_safe(dreq, lp, tmp,
> + &cache_defer_hash[hash], hash)
> if (dreq->item == item) {
> __unhash_deferred_req(dreq);
> list_add(&dreq->recent, &pending);
> }
> - }
> - }
> +
> spin_unlock(&cache_defer_lock);
>
> while (!list_empty(&pending)) {
>
>
next prev parent reply other threads:[~2010-09-22 3:01 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-09-22 2:55 [PATCH 0/7] Assorted nfsd patches for 2.6.37 NeilBrown
2010-09-22 2:55 ` [PATCH 1/7] sunrpc: fix race in new cache_wait code NeilBrown
2010-09-22 17:50 ` J. Bruce Fields
2010-09-23 3:00 ` Neil Brown
2010-09-23 3:25 ` J. Bruce Fields
2010-09-23 14:46 ` J. Bruce Fields
2010-10-01 23:09 ` J. Bruce Fields
2010-10-02 0:12 ` Neil Brown
2010-09-22 2:55 ` [PATCH 2/7] sunrpc/cache: fix recent breakage of cache_clean_deferred NeilBrown
[not found] ` <20100922025506.31745.74964.stgit-bi+AKbBUZKY6gyzm1THtWbp2dZbC/Bob@public.gmane.org>
2010-09-22 18:27 ` J. Bruce Fields
2010-09-22 2:55 ` [PATCH 7/7] nfsd: allow deprecated interface to be compiled out NeilBrown
2010-09-22 2:55 ` [PATCH 5/7] sunrpc/cache: allow thread manager more control of whether threads can wait for upcalls NeilBrown
2010-09-22 18:36 ` J. Bruce Fields
2010-09-23 3:23 ` Neil Brown
2010-09-22 2:55 ` [PATCH 6/7] nfsd: formally deprecate legacy nfsd syscall interface NeilBrown
[not found] ` <20100922025507.31745.57024.stgit-bi+AKbBUZKY6gyzm1THtWbp2dZbC/Bob@public.gmane.org>
2010-09-22 3:10 ` J. Bruce Fields
2010-09-22 2:55 ` [PATCH 4/7] sunrpc/cache: centralise handling of size limit on deferred list NeilBrown
[not found] ` <20100922025507.31745.61919.stgit-bi+AKbBUZKY6gyzm1THtWbp2dZbC/Bob@public.gmane.org>
2010-09-22 18:31 ` J. Bruce Fields
2010-09-23 3:02 ` Neil Brown
2010-09-22 2:55 ` [PATCH 3/7] sunrpc/cache: change deferred-request hash table to use hlist NeilBrown
2010-09-22 2:59 ` J. Bruce Fields [this message]
2010-09-22 4:51 ` Neil Brown
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=20100922025944.GC22021@fieldses.org \
--to=bfields@fieldses.org \
--cc=linux-nfs@vger.kernel.org \
--cc=neilb@suse.de \
/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 NNTP newsgroup(s).