From: Chuck Lever III <chuck.lever@oracle.com>
To: Dave Chinner <david@fromorbit.com>
Cc: Linux NFS Mailing List <linux-nfs@vger.kernel.org>,
netdev <netdev@vger.kernel.org>, "tgraf@suug.ch" <tgraf@suug.ch>,
Jeff Layton <jlayton@redhat.com>
Subject: Re: [PATCH RFC 29/30] NFSD: Convert the filecache to use rhashtable
Date: Thu, 23 Jun 2022 23:59:45 +0000 [thread overview]
Message-ID: <EDD9404B-ACBA-4284-8AFC-8AB4536481A3@oracle.com> (raw)
In-Reply-To: <20220623223320.GG1098723@dread.disaster.area>
> On Jun 23, 2022, at 6:33 PM, Dave Chinner <david@fromorbit.com> wrote:
>
> On Thu, Jun 23, 2022 at 05:27:20PM +0000, Chuck Lever III wrote:
>> Also I just found Neil's nice rhashtable explainer:
>>
>> https://lwn.net/Articles/751374/
>>
>> Where he writes that:
>>
>>> Sometimes you might want a hash table to potentially contain
>>> multiple objects for any given key. In that case you can use
>>> "rhltables" — rhashtables with lists of objects.
>>
>> I believe that is the case for the filecache. The hash value is
>> computed based on the inode pointer, and therefore there can be more
>> than one nfsd_file object for a particular inode (depending on who
>> is opening and for what access). So I think filecache needs to use
>> rhltable, not rhashtable. Any thoughts from rhashtable experts?
>
> Huh, I assumed the file cache was just hashing the whole key so that
> every object in the rht has it's own unique key and hash and there's
> no need to handle multiple objects per key...
>
> What are you trying to optimise by hashing only the inode *pointer*
> in the nfsd_file object keyspace?
Well, this design is inherited from the current filecache
implementation.
It assumes that all nfsd_file objects that refer to the same
inode will always get chained into the same bucket. That way:
506 static void
507 __nfsd_file_close_inode(struct inode *inode, unsigned int hashval,
508 struct list_head *dispose)
509 {
510 struct nfsd_file *nf;
511 struct hlist_node *tmp;
512
513 spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
514 hlist_for_each_entry_safe(nf, tmp, &nfsd_file_hashtbl[hashval].nfb_head, nf_node) {
515 if (inode == nf->nf_inode)
516 nfsd_file_unhash_and_release_locked(nf, dispose);
517 }
518 spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
519 }
nfsd_file_close_inode() can lock one hash bucket and just
walk that hash chain to find all the nfsd_file's associated
with a particular in-core inode.
Actually I don't think there's any other reason to keep that
hashing design, but Jeff can confirm that.
So I guess we could use rhltable and keep the nfsd_file items
for the same inode on the same hash list? I'm not sure it's
worth the trouble: this part of filecache isn't really on the
hot path.
--
Chuck Lever
next prev parent reply other threads:[~2022-06-23 23:59 UTC|newest]
Thread overview: 51+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-22 14:12 [PATCH RFC 00/30] Overhaul NFSD filecache Chuck Lever
2022-06-22 14:12 ` [PATCH RFC 01/30] NFSD: Report filecache LRU size Chuck Lever
2022-06-22 14:12 ` [PATCH RFC 02/30] NFSD: Report count of calls to nfsd_file_acquire() Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 03/30] NFSD: Report count of freed filecache items Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 04/30] NFSD: Report average age of " Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 05/30] NFSD: Add nfsd_file_lru_dispose_list() helper Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 06/30] NFSD: Refactor nfsd_file_gc() Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 07/30] NFSD: Refactor nfsd_file_lru_scan() Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 08/30] NFSD: Report the number of items evicted by the LRU walk Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 09/30] NFSD: Record number of flush calls Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 10/30] NFSD: Report filecache item construction failures Chuck Lever
2022-06-22 14:13 ` [PATCH RFC 11/30] NFSD: Zero counters when the filecache is re-initialized Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 12/30] NFSD: Hook up the filecache stat file Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 13/30] NFSD: WARN when freeing an item still linked via nf_lru Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 14/30] NFSD: Trace filecache LRU activity Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 15/30] NFSD: Leave open files out of the filecache LRU Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 16/30] NFSD: Fix the filecache LRU shrinker Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 17/30] NFSD: Never call nfsd_file_gc() in foreground paths Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 18/30] NFSD: No longer record nf_hashval in the trace log Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 19/30] NFSD: Remove lockdep assertion from unhash_and_release_locked() Chuck Lever
2022-06-22 14:14 ` [PATCH RFC 20/30] NFSD: nfsd_file_unhash can compute hashval from nf->nf_inode Chuck Lever
2022-06-22 14:15 ` [PATCH RFC 21/30] NFSD: Refactor __nfsd_file_close_inode() Chuck Lever
2022-06-22 14:15 ` [PATCH RFC 22/30] NFSD: nfsd_file_hash_remove can compute hashval Chuck Lever
2022-06-22 14:15 ` [PATCH RFC 23/30] NFSD: Remove nfsd_file::nf_hashval Chuck Lever
2022-06-22 14:15 ` [PATCH RFC 24/30] NFSD: Remove stale comment from nfsd_file_acquire() Chuck Lever
2022-06-22 14:15 ` [PATCH RFC 25/30] NFSD: Clean up "open file" case in nfsd_file_acquire() Chuck Lever
2022-06-22 14:15 ` [PATCH RFC 26/30] NFSD: Document nfsd_file_cache_purge() API contract Chuck Lever
2022-06-22 14:15 ` [PATCH RFC 27/30] NFSD: Replace the "init once" mechanism Chuck Lever
2022-06-22 14:15 ` [PATCH RFC 28/30] NFSD: Set up an rhashtable for the filecache Chuck Lever
2022-06-23 22:56 ` Al Viro
2022-06-23 23:51 ` Chuck Lever III
2022-06-24 0:14 ` Chuck Lever III
2022-06-24 0:29 ` Al Viro
2022-06-22 14:15 ` [PATCH RFC 29/30] NFSD: Convert the filecache to use rhashtable Chuck Lever
2022-06-23 0:38 ` Dave Chinner
2022-06-23 0:58 ` Chuck Lever III
2022-06-23 17:27 ` Chuck Lever III
2022-06-23 22:33 ` Dave Chinner
2022-06-23 23:59 ` Chuck Lever III [this message]
2022-06-22 14:16 ` [PATCH RFC 30/30] NFSD: Clean up unusued code after rhashtable conversion Chuck Lever
2022-06-22 18:36 ` [PATCH RFC 00/30] Overhaul NFSD filecache Wang Yugui
2022-06-22 19:04 ` Chuck Lever III
2022-06-22 19:59 ` Chuck Lever III
2022-06-23 9:02 ` Wang Yugui
2022-06-23 16:44 ` Chuck Lever III
2022-06-23 17:51 ` Wang Yugui
2022-06-24 15:30 ` Chuck Lever III
2022-06-23 0:21 ` Dave Chinner
2022-06-23 1:01 ` Chuck Lever III
2022-06-23 20:27 ` Frank van der Linden
2022-06-28 17:57 ` Chuck Lever III
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=EDD9404B-ACBA-4284-8AFC-8AB4536481A3@oracle.com \
--to=chuck.lever@oracle.com \
--cc=david@fromorbit.com \
--cc=jlayton@redhat.com \
--cc=linux-nfs@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=tgraf@suug.ch \
/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).