linux-kernel - Re: [PATCH] iov_iter: separate direction from flavour

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20210704172948.GA1730187@roeck-us.net>
Date:   Sun, 4 Jul 2021 10:29:48 -0700
From:   Guenter Roeck <linux@...ck-us.net>
To:     Al Viro <viro@...iv.linux.org.uk>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
        David Sterba <dsterba@...e.com>,
        Miklos Szeredi <miklos@...redi.hu>,
        Anton Altaparmakov <anton@...era.com>,
        David Howells <dhowells@...hat.com>,
        Matthew Wilcox <willy@...radead.org>,
        Pavel Begunkov <asml.silence@...il.com>
Subject: Re: [PATCH] iov_iter: separate direction from flavour

Hi,

On Thu, Apr 22, 2021 at 02:50:39PM -0400, Al Viro wrote:
> Instead of having them mixed in iter->type, use separate ->iter_type
> and ->data_source (u8 and bool resp.)  And don't bother with (pseudo-)
> bitmap for the former - microoptimizations from being able to check
> if the flavour is one of two values are not worth the confusion for
> optimizer.  It can't prove that we never get e.g. ITER_IOVEC | ITER_PIPE,
> so we end up with extra headache.
> 
> Signed-off-by: Al Viro <viro@...iv.linux.org.uk>

This patch results in the following runtime warning on nommu systems.

[    8.567154] Run /init as init process
[    8.572112] ------------[ cut here ]------------
[    8.572248] WARNING: CPU: 0 PID: 1 at lib/iov_iter.c:468 iov_iter_init+0x35/0x58
[    8.572484] CPU: 0 PID: 1 Comm: init Not tainted 5.13.0-09606-g303392fd5c16 #1
[    8.572695] Hardware name: MPS2 (Device Tree Support)
[    8.573278] [<2100ae75>] (unwind_backtrace) from [<2100a2bb>] (show_stack+0xb/0xc)
[    8.573594] [<2100a2bb>] (show_stack) from [<2100da03>] (__warn+0x5f/0x80)
[    8.573738] [<2100da03>] (__warn) from [<2100da55>] (warn_slowpath_fmt+0x31/0x60)
[    8.573886] [<2100da55>] (warn_slowpath_fmt) from [<210d8e1d>] (iov_iter_init+0x35/0x58)
[    8.574044] [<210d8e1d>] (iov_iter_init) from [<21059cab>] (vfs_read+0x89/0xc6)
[    8.574191] [<21059cab>] (vfs_read) from [<2105d92b>] (read_code+0x15/0x2e)
[    8.574329] [<2105d92b>] (read_code) from [<21085a8d>] (load_flat_file+0x341/0x4f0)
[    8.574481] [<21085a8d>] (load_flat_file) from [<21085e03>] (load_flat_binary+0x47/0x2dc)
[    8.574639] [<21085e03>] (load_flat_binary) from [<2105d581>] (bprm_execve+0x1fd/0x32c)
[    8.574797] [<2105d581>] (bprm_execve) from [<2105dbb3>] (kernel_execve+0xa3/0xac)
[    8.574947] [<2105dbb3>] (kernel_execve) from [<211e7095>] (kernel_init+0x31/0xb0)
[    8.575099] [<211e7095>] (kernel_init) from [<2100814d>] (ret_from_fork+0x11/0x24)
[    8.575287] Exception stack(0x21429fb0 to 0x21429ff8)
[    8.575433] 9fa0:                                     00000000 00000000 00000000 00000000
[    8.575593] 9fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[    8.575743] 9fe0: 00000000 00000000 00000000 00000000 00000000 00000000
[    8.575933] ---[ end trace ba15568c05035a77 ]---

This is with qemu's mps2-an385 emulation and and mps2_defconfig.
The same warning is also observed with m68k and mcf5208evb,
though the traceback isn't as nice.

WARNING: CPU: 0 PID: 1 at lib/iov_iter.c:468 0x40135e4e
...
Call Trace:
        [<402b0f42>] 0x402b0f42
 [<402b0fea>] 0x402b0fea
 [<40135e4e>] 0x40135e4e
 [<40135e4e>] 0x40135e4e
 [<4009c610>] 0x4009c610
...

Reverting this patch fixes the problem for both mps2-an385 and mcf5208evb.

Guenter