linux-kernel - Re: [GIT PULL] vfs fixes

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: <ZWH2kSJXcXEohpyd@telecaster>
Date:   Sat, 25 Nov 2023 05:28:49 -0800
From:   Omar Sandoval <osandov@...ndov.com>
To:     Christian Brauner <brauner@...nel.org>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        Omar Sandoval <osandov@...com>,
        David Howells <dhowells@...hat.com>,
        linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [GIT PULL] vfs fixes

On Sat, Nov 25, 2023 at 02:10:52PM +0100, Christian Brauner wrote:
> On Fri, Nov 24, 2023 at 10:25:15AM -0800, Linus Torvalds wrote:
> > On Fri, 24 Nov 2023 at 02:28, Christian Brauner <brauner@...nel.org> wrote:
> > >
> > > * Fix a bug introduced with the iov_iter rework from last cycle.
> > >
> > >   This broke /proc/kcore by copying too much and without the correct
> > >   offset.
> > 
> > Ugh. I think the whole /proc/kcore vmalloc handling is just COMPLETELY broken.
> 
> Ugh, I didn't even look at that closely because the fix was obviously
> correct for that helper alone. Let's try and just return zeroed memory
> like you suggested in your last mail before we bother fixing any of
> this.
> 
> Long-term plan, it'd be nice to just get distros to start turning
> /proc/kcore off. Maybe I underestimate legitimate users but this
> requires CAP_SYS_RAW_IO so it really can only be useful to pretty
> privileged stuff anyway.

drgn needs /proc/kcore for debugging the live kernel, which is a very
important use case for lots of our users. And it does in fact read from
KCORE_VMALLOC segments, which is why I found and fixed this bug. I'm
happy to clean up this code, although it's a holiday weekend here so I
won't get to it immediately of course. But please don't rip this out.

Omar