linux-kernel - Re: [PATCH] scsi: Let scsi_execute_cmd() mark args->sshdr as invalid

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: <73fd2741-3730-ca1d-7e23-0bf9cf10f423@suse.com>
Date:   Sun, 21 May 2023 07:23:56 +0200
From:   Juergen Gross <jgross@...e.com>
To:     "Martin K. Petersen" <martin.petersen@...cle.com>,
        John Garry <john.g.garry@...cle.com>
Cc:     linux-kernel@...r.kernel.org, linux-scsi@...r.kernel.org,
        "James E.J. Bottomley" <jejb@...ux.ibm.com>, stable@...r.kernel.org
Subject: Re: [PATCH] scsi: Let scsi_execute_cmd() mark args->sshdr as invalid

On 21.05.23 03:19, Martin K. Petersen wrote:
> 
> John,
> 
>> @Martin, Do you have any preference for what we do now? This code
>> which does not check for error and does not pre-zero sshdr is
>> longstanding, so I am not sure if Juergen's change is required for for
>> v6.4. I'm thinking to fix callers for v6.5 and also maybe change the
>> API, as I described.
> 
> As I alluded to in the tracing thread, I'd like to see SK/ASC/ASCQ being
> generally available in the scsi_cmnd results instead of all this sense
> buffer and sense header micromanagement in every caller. That's a pretty
> heavy lift, though.
> 
> Short term we need all callers to be fixed up. I'm not a particularly
> big fan of scsi_execute_cmd() zeroing something being passed in. I
> wonder if it would be worth having a DECLARE_SENSE_HEADER()?

sshdr is output only data, so setting it before returning seems to be a
sensible thing to do.

Letting the callers do that is kind of a layering violation IMHO, as this
would spread the knowledge that scsi_execute_cmd() isn't setting its output
data always.

In the end its your decision, of course.


Juergen

Download attachment "OpenPGP_0xB0DE9DD628BF132F.asc" of type "application/pgp-keys" (3099 bytes)

Download attachment "OpenPGP_signature" of type "application/pgp-signature" (496 bytes)