[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <69c51e23-48b9-4672-e559-d2e257ade29f@huawei.com>
Date: Wed, 18 Aug 2021 12:29:10 +0100
From: John Garry <john.garry@...wei.com>
To: "Martin K. Petersen" <martin.petersen@...cle.com>
CC: <tyreld@...ux.ibm.com>, <mpe@...erman.id.au>,
<benh@...nel.crashing.org>, <paulus@...ba.org>,
<jejb@...ux.ibm.com>, <linux-scsi@...r.kernel.org>,
<linuxppc-dev@...ts.ozlabs.org>, <linux-kernel@...r.kernel.org>,
<hare@...e.de>, <bvanassche@....org>, <hch@....de>,
<linux-next@...r.kernel.org>, <sfr@...b.auug.org.au>
Subject: Re: [PATCH] scsi: ibmvfc: Stop using scsi_cmnd.tag
Hi Martin,
>
>> Use scsi_cmd_to_rq(scsi_cmnd)->tag in preference to scsi_cmnd.tag.
>
> Applied to 5.15/scsi-staging and rebased for bisectability.
>
Thanks, and sorry for the hassle. But I would still like the maintainers
to have a look, as I was curious about current usage of scsi_cmnd.tag in
that driver.
> Just to be picky it looks like there's another scsi_cmmd tag lurking in
> qla1280.c but it's sitting behind an #ifdef DEBUG_QLA1280.
>
That driver does not even compile with DEBUG_QLA1280 set beforehand.
I'll fix that up and send as separate patches in case you want to
shuffle the tag patch in earlier, which is prob not worth the effort.
I've done a good few more x86 randconfigs and tried to audit the code
for more references, so hopefully that's the last.
Thanks
Powered by blists - more mailing lists