linux-kernel - Re: [PATCH 17/23] proc: add a read_iter method to proc proc

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: <20200702051811.GB30361@lst.de>
Date:   Thu, 2 Jul 2020 07:18:11 +0200
From:   Christoph Hellwig <hch@....de>
To:     Al Viro <viro@...iv.linux.org.uk>
Cc:     Christoph Hellwig <hch@....de>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Luis Chamberlain <mcgrof@...nel.org>,
        Matthew Wilcox <willy@...radead.org>,
        Kees Cook <keescook@...omium.org>,
        Iurii Zaikin <yzaikin@...gle.com>,
        linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org
Subject: Re: [PATCH 17/23] proc: add a read_iter method to proc proc_ops

On Wed, Jul 01, 2020 at 10:27:51PM +0100, Al Viro wrote:
> On Wed, Jul 01, 2020 at 10:09:45PM +0200, Christoph Hellwig wrote:
> > This will allow proc files to implement iter read semantics.
> 
> *UGH*
> 
> You are introducing file_operations with both ->read() and ->read_iter();
> worse, in some cases they are not equivalent.  Sure, ->read() takes
> precedence right now, but...  why not a separate file_operations for
> ->read_iter-capable files?

I looked at that initially.  We'd need to more instances as there
already are two due to compat stuff.  If that is preferably I can
switch to that version.

> I really hate the fallbacks of that sort - they tend to be brittle
> as hell.  And while we are at it, I'm not sure that your iter_read() 
> has good cause to be non-static.

The other user of it is seq_file, which as-is should go away, but
will probably keep the occasional version of it in the caller.  I just
got really tired of reimplementing it a few times.