linux-kernel - Re: [PATCH] libata: remove unlock+relock cycle in ata_scsi

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] [day] [month] [year] [list]

Message-ID: <1290010276.4736.51.camel@mulgrave.site>
Date:	Wed, 17 Nov 2010 10:11:16 -0600
From:	James Bottomley <James.Bottomley@...e.de>
To:	Christoph Hellwig <hch@...radead.org>
Cc:	Jeff Garzik <jeff@...zik.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	linux-ide@...r.kernel.org, linux-scsi@...r.kernel.org,
	LKML <linux-kernel@...r.kernel.org>, Tejun Heo <tj@...nel.org>
Subject: Re: [PATCH] libata: remove unlock+relock cycle in ata_scsi_queuecmd

On Wed, 2010-11-17 at 10:47 -0500, Christoph Hellwig wrote:
> > Right ... I couldn't persuade anyone else to do it, so I'll probably end
> > up coding it.  It looks like the serial number zero check might be
> > bogus.  If so I'll remove it, then push the serial number acquisition
> > down only into the locked queuecommand of only those drivers that
> > actually use it (which were listed in the quoted email, and which might
> > have a big impetus to remove it if the use is trivial).  Then we can
> > begin unwinding the locking.
> 
> In dc395x, eata_pio, in2000, lpfc, megaraid_legacy, megaraid_mbox,
> megaraid_sas, mesh, ncr53c8xx, qla1280, qla4xxx, wd33c93 and i2o_scsi
> use it in printks only, and all this can be safely removed.
> 
> Except for EH that only leaves dpt_i2o, eata, mpt2sas, u14-34f, mptscsi
> as non-trivial users.  If you sort out the EH side I'll volunteer to
> kill the printks and add private serial numbers to the five drivers
> mentioned above.

You're on (you were cc'd on the patch) ... I was planning just to move
the scsi_cmnd_get_serial() call into the locked queuecommand routine,
but a more permanent solution might be better.

James


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/