linux-kernel - Re: block: Don't invalidate pagecache for invalid falloc modes

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]

Date:   Wed, 11 Oct 2023 17:08:21 -0400
From:   Mike Snitzer <snitzer@...nel.org>
To:     Jens Axboe <axboe@...nel.dk>
Cc:     Sarthak Kukreti <sarthakkukreti@...omium.org>,
        linux-block@...r.kernel.org, linux-kernel@...r.kernel.org,
        Bart Van Assche <bvanassche@....org>,
        "Darrick J. Wong" <darrick.wong@...cle.com>,
        stable@...r.kernel.org, "Darrick J . Wong" <djwong@...nel.org>,
        Christoph Hellwig <hch@....de>
Subject: Re: block: Don't invalidate pagecache for invalid falloc modes

On Wed, Oct 11 2023 at  4:53P -0400,
Jens Axboe <axboe@...nel.dk> wrote:

> On 10/11/23 2:50 PM, Mike Snitzer wrote:
> > On Wed, Oct 11 2023 at  4:20P -0400,
> > Jens Axboe <axboe@...nel.dk> wrote:
> > 
> >> On 10/11/23 2:12 PM, Sarthak Kukreti wrote:
> >>> Only call truncate_bdev_range() if the fallocate mode is
> >>> supported. This fixes a bug where data in the pagecache
> >>> could be invalidated if the fallocate() was called on the
> >>> block device with an invalid mode.
> >>
> >> Fix looks fine, but would be nicer if we didn't have to duplicate the
> >> truncate_bdev_range() in each switch clause. Can we check this upfront
> >> instead?
> > 
> > No, if you look at the function (rather than just the patch in
> > isolation) we need to make the call for each case rather than collapse
> > to a single call at the front (that's the reason for this fix, because
> > otherwise the default: error case will invalidate the page cache too).
> 
> Yes that part is clear, but it might look cleaner to check a valid mask
> first rather than have 3 duplicate calls.

OK.
 
> > Just so you're aware, I also had this feedback that shaped the patch a
> > bit back in April:
> > https://listman.redhat.com/archives/dm-devel/2023-April/053986.html
> > 
> >> Also, please wrap commit messages at 72-74 chars.
> > 
> > Not seeing where the header should be wrapped.  You referring to the
> > Fixes: line?  I've never seen those wrapped.
> 
> I'm referring to the commit message itself.

Ah, you'd like lines extended because they are too short.