linux-kernel - Re: [PATCH v2 0/3] scsi: qedf: sanitise uaccess

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: <a44acf6edf8714f9e247e8de371d0f00521ec6b5.camel@redhat.com>
Date:   Mon, 31 Jul 2023 14:33:28 -0400
From:   Laurence Oberman <loberman@...hat.com>
To:     Oleksandr Natalenko <oleksandr@...hat.com>,
        linux-kernel@...r.kernel.org
Cc:     linux-scsi@...r.kernel.org, Saurav Kashyap <skashyap@...vell.com>,
        Johannes Thumshirn <Johannes.Thumshirn@....com>,
        David Laight <David.Laight@...LAB.COM>,
        GR-QLogic-Storage-Upstream@...vell.com,
        "James E.J. Bottomley" <jejb@...ux.ibm.com>,
        "Martin K. Petersen" <martin.petersen@...cle.com>,
        Jozef Bacik <jobacik@...hat.com>, Rob Evers <revers@...hat.com>
Subject: Re: [PATCH v2 0/3] scsi: qedf: sanitise uaccess

On Mon, 2023-07-31 at 10:40 +0200, Oleksandr Natalenko wrote:
> qedf driver, debugfs part of it specifically, touches __user pointers
> directly for printing out info to userspace via sprintf(), which may
> cause crash like this:
> 
> BUG: unable to handle kernel paging request at 00007ffd1d6b43a0
> IP: [<ffffffffaa7a882a>] string.isra.7+0x6a/0xf0
> Oops: 0003 [#1] SMP
> Call Trace:
>  [<ffffffffaa7a9f31>] vsnprintf+0x201/0x6a0
>  [<ffffffffaa7aa556>] sprintf+0x56/0x80
>  [<ffffffffc04227ed>] qedf_dbg_stop_io_on_error_cmd_read+0x6d/0x90
> [qedf]
>  [<ffffffffaa65bb2f>] vfs_read+0x9f/0x170
>  [<ffffffffaa65cb82>] SyS_pread64+0x92/0xc0
> 
> Avoid this by preparing the info in a kernel buffer first, either
> allocated on stack for small printouts, or via vmalloc() for big
> ones,
> and then copying it to the userspace properly.
> 
> Changes since v1 [1]:
> 
>   * use scnprintf() for on-stack buffers too
>   * adjust an on-stack buffer size in qedf_dbg_debug_cmd_read() to be
> a
>     multiple of 8, and also size it properly
>   * accumulate acks and reviews
> 
> [1]
> https://lore.kernel.org/lkml/20230728065819.139694-1-oleksandr@redhat.com/
> 
> Oleksandr Natalenko (3):
>   scsi: qedf: do not touch __user pointer in
>     qedf_dbg_stop_io_on_error_cmd_read() directly
>   scsi: qedf: do not touch __user pointer in
> qedf_dbg_debug_cmd_read()
>     directly
>   scsi: qedf: do not touch __user pointer in
> qedf_dbg_fp_int_cmd_read()
>     directly
> 
>  drivers/scsi/qedf/qedf_dbg.h     |  2 ++
>  drivers/scsi/qedf/qedf_debugfs.c | 35 +++++++++++++++++++-----------
> --
>  2 files changed, 23 insertions(+), 14 deletions(-)
> 
In case it's needed

For the series v2 against 6.5-rc3
Reviewed-by: Laurence Oberman <loberman@...hat.com>
Tested-by: Laurence Oberman <loberman@...hat.com>

Test notes

Linux segstorage5 6.5.0-rc3+

[root@...storage5 qedf]# cd host2
[root@...storage5 host2]# ls
clear_stats  debug  driver_stats  fp_int  io_trace  offload_stats 
stop_io_on_error

[root@...storage5 host2]# cat stop_io_on_error
false

[root@...storage5 host2]# cat fp_int

Fastpath I/O completions

#0: 844
#1: 990
#2: 1036
#3: 1116
#4: 953
#5: 822
#6: 882
#7: 1073
#8: 1030
#9: 992
#10: 789
#11: 705
#12: 490
#13: 532
#14: 646
#15: 705

[root@...storage5 host2]# cat debug
debug mask = 0x2