linux-kernel - Re: [PATCH] scsi: do not requeue requests unaligned with device sector size

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: <1f44fb00-78c6-fb5c-9169-6aeeaf4c06ef@suse.de>
Date:   Wed, 21 Dec 2016 09:09:22 +0100
From:   Hannes Reinecke <hare@...e.de>
To:     Christoph Hellwig <hch@...radead.org>,
        Mauricio Faria de Oliveira <mauricfo@...ux.vnet.ibm.com>
Cc:     jejb@...ux.vnet.ibm.com, martin.petersen@...cle.com,
        linux-scsi@...r.kernel.org, linux-block@...r.kernel.org,
        linux-kernel@...r.kernel.org, gpiccoli@...ux.vnet.ibm.com
Subject: Re: [PATCH] scsi: do not requeue requests unaligned with device
 sector size

On 12/21/2016 08:50 AM, Christoph Hellwig wrote:
> On Tue, Dec 20, 2016 at 12:02:27AM -0200, Mauricio Faria de Oliveira wrote:
>> When a SCSI command (e.g., read operation) is partially completed
>> with good status and residual bytes (i.e., not all the bytes from
>> the specified transfer length were transferred) the SCSI midlayer
>> will update the request/bios with the completed bytes and requeue
>> the request in order to complete the remainder/pending bytes.
>>
>> However, when the device sector size is greater than the 512-byte
>> default/kernel sector size, alignment restrictions and validation
>> apply (both to the starting logical block address, and the number
>> of logical blocks to transfer) -- values must be multiples of the
>> device sector size, otherwise the kernel fails the request in the
>> preparation stage (e.g., sd_setup_read_write_cmnd() at sd.c file):
> 
> How do you even get an unaligned residual count?  Except for SES
> processor devices (which will only issue BLOCK_PC commands) this is
> not allowed by SPC:
> 
> "The residual count shall be reported in bytes if the peripheral device
>  type in the destination target descriptor is 03h (i.e., processor device),
>  and in destination device blocks for all other device type codes."

Which actually would be pretty much my objection, too.

This would only be applicable for 512e drives, where we _might_ end up
with a residual smaller than the physical sector size.
But that should be handled by firmware; after all, that's what the 'e'
implies, right?

Cheers,

Hannes
-- 
Dr. Hannes Reinecke		   Teamlead Storage & Networking
hare@...e.de			               +49 911 74053 688
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: F. Imendörffer, J. Smithard, J. Guild, D. Upmanyu, G. Norton
HRB 21284 (AG Nürnberg)