From: James Bottomley <James.Bottomley@steeleye.com>
To: David Brownell <david-b@pacbell.net>
Cc: Andrew Morton <akpm@digeo.com>,
"Adam J. Richter" <adam@yggdrasil.com>,
James.Bottomley@SteelEye.com, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] generic device DMA (dma_pool update)
Date: Tue, 31 Dec 2002 17:47:49 -0600 [thread overview]
Message-ID: <200212312347.gBVNlnK04199@localhost.localdomain> (raw)
In-Reply-To: Message from David Brownell <david-b@pacbell.net> of "Tue, 31 Dec 2002 15:23:43 PST." <3E1226FF.9010407@pacbell.net>
david-b@pacbell.net said:
> USB device drivers tend to either allocate and reuse one dma buffer
> (kmalloc/kfree usage pattern) or use dma mapping ... so far.
Please be careful with this in drivers. Coherent memory can be phenomenally
expensive to obtain on some hardware. Sometimes it has to be implemented by
turning off caching and globally flushing the tlb. Thus it really makes
sense most of the time for drivers to allocate all the coherent memory they
need initially and not have it go through the usual memory allocate/free cycle
except under extreme conditions. That's sort really what both pci_pool and
mempool aim for.
James
next prev parent reply other threads:[~2002-12-31 23:39 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-12-31 22:02 [PATCH] generic device DMA (dma_pool update) Adam J. Richter
2002-12-31 22:41 ` Andrew Morton
2002-12-31 23:23 ` David Brownell
2002-12-31 23:27 ` Andrew Morton
2002-12-31 23:44 ` David Brownell
2002-12-31 23:47 ` James Bottomley [this message]
2003-01-01 17:10 ` James Bottomley
2002-12-31 23:35 ` David Brownell
-- strict thread matches above, loose matches on Subject: below --
2003-01-02 22:07 Adam J. Richter
2003-01-03 0:20 ` Russell King
2003-01-03 4:50 ` David Brownell
2003-01-03 6:11 ` David Brownell
2003-01-03 6:46 ` David Brownell
2003-01-03 6:52 ` William Lee Irwin III
2003-01-02 17:04 Adam J. Richter
2003-01-02 4:13 Adam J. Richter
2003-01-02 16:41 ` James Bottomley
2003-01-02 18:26 ` David Brownell
2003-01-01 19:21 Adam J. Richter
2003-01-01 19:48 ` James Bottomley
2003-01-02 2:11 ` David Brownell
2003-01-01 0:02 Adam J. Richter
2002-12-31 23:38 Adam J. Richter
2002-12-27 21:40 [RFT][PATCH] generic device DMA implementation James Bottomley
2002-12-28 1:56 ` David Brownell
2002-12-30 23:11 ` [PATCH] generic device DMA (dma_pool update) David Brownell
2002-12-31 15:00 ` James Bottomley
2002-12-31 17:04 ` David Brownell
2002-12-31 17:23 ` James Bottomley
2002-12-31 18:11 ` David Brownell
2002-12-31 18:44 ` James Bottomley
2002-12-31 19:29 ` David Brownell
2002-12-31 19:50 ` James Bottomley
2002-12-31 21:17 ` David Brownell
2002-12-31 16:36 ` James Bottomley
2002-12-31 17:32 ` David Brownell
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=200212312347.gBVNlnK04199@localhost.localdomain \
--to=james.bottomley@steeleye.com \
--cc=adam@yggdrasil.com \
--cc=akpm@digeo.com \
--cc=david-b@pacbell.net \
--cc=linux-kernel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox