[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <fb0efbd1-a54f-09d6-bd27-6f665b461e58@acm.org>
Date: Thu, 18 May 2023 12:54:27 -0700
From: Bart Van Assche <bvanassche@....org>
To: John Garry <john.g.garry@...cle.com>,
Juergen Gross <jgross@...e.com>,
"Martin K. Petersen" <martin.petersen@...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 5/18/23 03:57, John Garry wrote:
> I think it's better to fix up the callers.
+1
> Further to that, I dislike
> how we pass a pointer to this local sshdr structure. I would prefer if
> scsi_execute_cmd() could kmalloc() the mem for these buffers and the
> callers could handle free'ing them - I can put together a patch for
> that, to see what people think.
sizeof(struct scsi_sense_hdr) = 8. Using kmalloc() to allocate an eight
byte data structure sounds like overkill to me. Additionally, making
scsi_execute_cmd() allocate struct scsi_sense_hdr and letting the
callers free that data structure will make it harder to review whether
or not any memory leaks are triggered. No such review is necessary if
the scsi_execute_cmd() caller allocates that data structure on the stack.
Bart.
Powered by blists - more mailing lists