Subject: Re: about TRIM/DISCARD support and barriers
From: James Bottomley <James.Bottomley@HansenPartnership.com>
To: David Woodhouse <dwmw2@infradead.org>
Cc: Matthew Wilcox <matthew@wil.cx>, Tejun Heo <teheo@suse.de>,
       Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
       Nick Piggin <npiggin@suse.de>, Jens Axboe <jens.axboe@oracle.com>,
       IDE/ATA development list <linux-ide@vger.kernel.org>,
       Jeff Garzik <jeff@garzik.org>, Dongjun Shin <djshin90@gmail.com>,
       chris.mason@oracle.com
In-Reply-To: <1227447584.4901.405.camel@macbook.infradead.org>
References: <4928E010.4090801@kernel.org> <4929023C.2060302@suse.de>
	 <20081123123514.GI5707@parisc-linux.org>
	 <1227447584.4901.405.camel@macbook.infradead.org>
Content-Type: text/plain
Date: Mon, 24 Nov 2008 07:52:56 +0900
Message-Id: <1227480776.25499.3.camel@localhost.localdomain>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1188
Lines: 30

On Sun, 2008-11-23 at 13:39 +0000, David Woodhouse wrote:
> > We don't attempt to put non-contiguous ranges into a single TRIM yet.
> 
> We don't even merge contiguous ranges -- I still need to fix the
> elevators to stop writes crossing writes,

I don't think we want to do that ... it's legal if the write isn't a
barrier and it will inhibit merging.  That may be just fine for a SSD,
but it's not for spinning media since they get better performance out of
merged writes.

>  before we can stop discards
> from also being barriers. (Discards are just writes, for the purpose of
> that conversation).

Perhaps they shouldn't be ... they have different characteristics.  For
instance, a discard may cross a read or write that has no sectors in
common with it; a discard may be merged as a non contiguous range
(assuming the drive supports multiple ranges), etc.

I think it might be better to give it its own type for the elevators.

James


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