Linux-NFS Archive on lore.kernel.org
 help / color / Atom feed
From: Trond Myklebust <trondmy@hammerspace.com>
To: "bcodding@redhat.com" <bcodding@redhat.com>
Cc: "linux-nfs@vger.kernel.org" <linux-nfs@vger.kernel.org>
Subject: Re: [PATCH v2 00/16] Readdir enhancements
Date: Wed, 4 Nov 2020 17:04:43 +0000
Message-ID: <50bf1f5c78d16d1018741febf822a00142a07b5b.camel@hammerspace.com> (raw)
In-Reply-To: <FEA7C67B-4091-4797-B05E-D762F0D0B3A1@redhat.com>

Hi Ben

Thanks for the review and the testing!

On Wed, 2020-11-04 at 11:14 -0500, Benjamin Coddington wrote:
> Hi Trond, these look great!
> 
> I'm doing some comparison testing before/after this set, and I'm
> getting
> into some memory pressure on a client with 4G ram listing 1.5M
> dentries 
> with
> 12 char filenames.
> 
> It looks like before this set, the readdir code was a bit more
> resilient 
> in
> the face of memory pressure, and I'm wondering if we've dropped a
> call 
> to
> mark_page_accessed().
> 
> * Ben adds:
> 
> @@ -460,7 +461,8 @@ static int nfs_readdir_search_array(struct 
> nfs_readdir_descriptor *desc)
>                  desc->last_cookie = array->last_cookie;
>                  desc->current_index += array->size;
>                  desc->page_index++;
> -       }
> +       } else
> +               mark_page_accessed(desc->page);
>          kunmap_atomic(array);
>          return status;
>   }
> 
> .. no, that's not any better.  I'm still getting evicted pages (or,
> at
> least, low-indexed pages that don't have PageUptodate() set), which 
> makes
> it nearly impossible to finish listing this directory because we just
> keep
> invalidating the mapping.
> 

You're right that I had screwed up the page access marking in the
previous patchsets. I believe this should be fixed in v3 by the
conversion to use grab_cache_page(), which calls find_or_create_page()
and should therefore do the right thing with the FGP_ACCESSED flag.

I believe the reason why your patch above fails to fully correct the
issue is because we always want to mark the page as accessed if we've
scanned it.

-- 
Trond Myklebust
Linux NFS client maintainer, Hammerspace
trond.myklebust@hammerspace.com



  reply index

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-03 15:33 trondmy
2020-11-03 15:33 ` [PATCH v2 01/16] NFS: Ensure contents of struct nfs_open_dir_context are consistent trondmy
2020-11-03 15:33   ` [PATCH v2 02/16] NFS: Clean up readdir struct nfs_cache_array trondmy
2020-11-03 15:33     ` [PATCH v2 03/16] NFS: Clean up nfs_readdir_page_filler() trondmy
2020-11-03 15:33       ` [PATCH v2 04/16] NFS: Clean up directory array handling trondmy
2020-11-03 15:33         ` [PATCH v2 05/16] NFS: Don't discard readdir results trondmy
2020-11-03 15:33           ` [PATCH v2 06/16] NFS: Remove unnecessary kmap in nfs_readdir_xdr_to_array() trondmy
2020-11-03 15:33             ` [PATCH v2 07/16] NFS: Replace kmap() with kmap_atomic() in nfs_readdir_search_array() trondmy
2020-11-03 15:33               ` [PATCH v2 08/16] NFS: Simplify struct nfs_cache_array_entry trondmy
2020-11-03 15:33                 ` [PATCH v2 09/16] NFS: Support larger readdir buffers trondmy
2020-11-03 15:33                   ` [PATCH v2 10/16] NFS: More readdir cleanups trondmy
2020-11-03 15:33                     ` [PATCH v2 11/16] NFS: nfs_do_filldir() does not return a value trondmy
2020-11-03 15:33                       ` [PATCH v2 12/16] NFS: Reduce readdir stack usage trondmy
2020-11-03 15:33                         ` [PATCH v2 13/16] NFS: Cleanup to remove nfs_readdir_descriptor_t typedef trondmy
2020-11-03 15:33                           ` [PATCH v2 14/16] NFS: Allow the NFS generic code to pass in a verifier to readdir trondmy
2020-11-03 15:33                             ` [PATCH v2 15/16] NFS: Handle NFS4ERR_NOT_SAME and NFSERR_BADCOOKIE from readdir calls trondmy
2020-11-03 15:33                               ` [PATCH v2 16/16] NFS: Improve handling of directory verifiers trondmy
2020-11-04 10:12                                 ` Mkrtchyan, Tigran
2020-11-04 17:21                                   ` Trond Myklebust
2020-11-03 15:55           ` [PATCH v2 05/16] NFS: Don't discard readdir results Benjamin Coddington
2020-11-03 16:14             ` Trond Myklebust
2020-11-04 16:14 ` [PATCH v2 00/16] Readdir enhancements Benjamin Coddington
2020-11-04 17:04   ` Trond Myklebust [this message]
2020-11-04 17:19     ` Benjamin Coddington

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=50bf1f5c78d16d1018741febf822a00142a07b5b.camel@hammerspace.com \
    --to=trondmy@hammerspace.com \
    --cc=bcodding@redhat.com \
    --cc=linux-nfs@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

Linux-NFS Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-nfs/0 linux-nfs/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-nfs linux-nfs/ https://lore.kernel.org/linux-nfs \
		linux-nfs@vger.kernel.org
	public-inbox-index linux-nfs

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-nfs


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git