linux-kernel - Re: [PATCH-for-4.6 1/6] target: Add target_alloc

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: <1452676969.27508.60.camel@haakon3.risingtidesystems.com>
Date:	Wed, 13 Jan 2016 01:22:49 -0800
From:	"Nicholas A. Bellinger" <nab@...ux-iscsi.org>
To:	Christoph Hellwig <hch@....de>
Cc:	"Nicholas A. Bellinger" <nab@...erainc.com>,
	target-devel <target-devel@...r.kernel.org>,
	linux-scsi <linux-scsi@...r.kernel.org>,
	lkml <linux-kernel@...r.kernel.org>,
	Sagi Grimberg <sagig@...lanox.com>,
	Hannes Reinecke <hare@...e.de>,
	Andy Grover <agrover@...hat.com>,
	Vasu Dev <vasu.dev@...ux.intel.com>, Vu Pham <vu@...lanox.com>
Subject: Re: [PATCH-for-4.6 1/6] target: Add target_alloc_session() helper
 function

On Wed, 2016-01-13 at 09:17 +0100, Christoph Hellwig wrote:
> On Wed, Jan 13, 2016 at 12:09:26AM -0800, Nicholas A. Bellinger wrote:
> > > I'd much rather have the fabrics drivers call transport_alloc_session_tags
> > > directly from the callback than growing even more arguments here.
> > > 
> > 
> > What's the point..?
> > 
> > Only vhost-scsi currently needs to allocate extra resources for
> > se_session percpu-ida tag pre-allocation.
> 
> The point is that allocating "tags" (or better command structures) isn't
> fundamentally related to allocating a session, and in fact not even
> used by many drivers.

Huh..?  It's used by iscsi-target, ib_isert, tcm_qla2xxx, vhost-scsi and
tcm_fc.

That's the majority of the drivers people care about.

>  A flow where drivers that need these "tags"
> explicitly allocate them is a lot more obvious than forcing more
> arguments to a function that already has a lot and isn't related to
> these "tags" otherwise.

The goal is to get to the point where all fabric drivers use tag
pre-allocation, and not vice-versa.

That said, I think it makes sense to leave it as is instead of adding
new callbacks for iscsi-target, ib_isert, and tcm_fc to just handle
se_session tag pre-allocation.