linux-kernel - Re: bcache on XFS: metadata I/O (dirent I/O?) not getting cached at all?

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20190207092819.GY24140@tuebingen.mpg.de>
Date:   Thu, 7 Feb 2019 10:28:19 +0100
From:   Andre Noll <maan@...bingen.mpg.de>
To:     Coly Li <colyli@...e.de>
Cc:     Dave Chinner <david@...morbit.com>, Nix <nix@...eri.org.uk>,
        linux-bcache@...r.kernel.org, linux-xfs@...r.kernel.org,
        linux-kernel@...r.kernel.org, Christoph Hellwig <hch@....de>,
        axboe@...nel.dk
Subject: Re: bcache on XFS: metadata I/O (dirent I/O?) not getting cached at
 all?

On Thu, Feb 07, 10:27, Coly Li wrote

> If different file system handles metadata flags in unified ways, it is
> OK to me to change the code to: !(bio->bi_opf & (REQ_META |REQ_PRIO)).

Yes, that's the smallest fix that should also go into 4.19-stable.

In the long run, we should try to get rid of the 45 instances of
REQ_PRIO. Most users specify REQ_META | REQ_PRIO anyway, which leaves
only a few other instances to look at.

I think the one in submit_bh_wbc() of fs/buffer.c can just be removed
while block/cfq-iosched.c does not use REQ_META at all, so the simple
s/REQ_PRIO/REQ_META should be OK. drivers/staging/erofs/data.c is
also easy to fix.

Best
Andre
-- 
Max Planck Institute for Developmental Biology
Max-Planck-Ring 5, 72076 Tübingen, Germany. Phone: (+49) 7071 601 829
http://people.tuebingen.mpg.de/maan/

Download attachment "signature.asc" of type "application/pgp-signature" (196 bytes)