From: David Miller <davem@davemloft.net>
To: bcrl@kvack.org
Cc: shemminger@osdl.org, mchan@broadcom.com,
herbert@gondor.apana.org.au, hadi@cyberus.ca,
jesse.brandeburg@intel.com, auke-jan.h.kok@intel.com,
netdev@vger.kernel.org
Subject: Re: [PATCH] [e1000]: Remove unnecessary tx_lock
Date: Tue, 08 Aug 2006 15:06:07 -0700 (PDT) [thread overview]
Message-ID: <20060808.150607.10247901.davem@davemloft.net> (raw)
In-Reply-To: <20060808170432.GD27383@kvack.org>
From: Benjamin LaHaise <bcrl@kvack.org>
Date: Tue, 8 Aug 2006 13:04:32 -0400
> Maybe the way NETDEV_TX_BUSY is implemented is wrong -- that is, it should
> be possible to return a result saying "we sent the packet, but the queue is
> now full". That would eliminate the atomic op that netif_queue_stop()
> performs and allow higher layers to merge the necessary barrier on the full
> case with the op on the tx lock. That way we could have lockless queue
> handling within the driver. This might need start/stop sequence counters,
> though.
The driver ->hard_start_xmit() method is invoked with the queue
unlocked, so this kind of scheme would not be workable.
While the queue is unlocked, another cpu could empty entries in the
TX queue making it not-full, which invalidates this "queue is now
full" return value the driver just gave.
We don't do things the way we do it now for fun, it's just the most
reasonable scheme we've come up with given the locking constraints.
next prev parent reply other threads:[~2006-08-08 22:06 UTC|newest]
Thread overview: 85+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-08-03 13:15 [PATCH] [e1000]: Remove unnecessary tx_lock jamal
2006-08-03 14:02 ` Herbert Xu
2006-08-03 14:24 ` jamal
2006-08-03 16:36 ` Brandeburg, Jesse
2006-08-03 18:05 ` Michael Chan
2006-08-03 22:08 ` jamal
2006-08-04 0:09 ` Michael Chan
2006-08-04 1:10 ` Herbert Xu
2006-08-04 8:37 ` Herbert Xu
2006-08-04 10:10 ` Herbert Xu
2006-08-04 10:16 ` jamal
2006-08-04 10:25 ` Herbert Xu
2006-08-04 10:45 ` jamal
2006-08-05 23:04 ` jamal
2006-08-05 23:06 ` Herbert Xu
2006-08-05 23:21 ` jamal
2006-08-05 23:30 ` Herbert Xu
2006-08-05 16:45 ` jamal
2006-08-04 17:12 ` Stephen Hemminger
2006-08-04 17:28 ` Michael Chan
2006-08-04 18:08 ` Stephen Hemminger
2006-08-04 23:31 ` David Miller
2006-08-05 16:56 ` jamal
2006-08-05 23:05 ` Herbert Xu
2006-08-05 23:17 ` jamal
2006-08-05 23:19 ` Herbert Xu
2006-08-05 23:36 ` jamal
2006-08-06 2:51 ` Herbert Xu
2006-08-06 7:14 ` Edgar E. Iglesias
2006-08-06 7:24 ` Herbert Xu
2006-08-06 7:30 ` Edgar E. Iglesias
2006-08-06 7:26 ` David Miller
2006-08-06 7:36 ` Herbert Xu
2006-08-06 8:06 ` Edgar E. Iglesias
2006-08-06 8:27 ` Herbert Xu
2006-08-06 9:03 ` Edgar E. Iglesias
2006-08-06 9:10 ` Herbert Xu
2006-08-06 9:18 ` Edgar E. Iglesias
2006-08-06 8:35 ` David Miller
2006-08-06 12:24 ` jamal
2006-08-06 12:33 ` jamal
2006-08-06 23:16 ` Jesse Brandeburg
2006-08-07 12:50 ` jamal
2006-08-07 15:21 ` Edgar E. Iglesias
2006-08-07 15:40 ` jamal
2006-08-07 15:59 ` Edgar E. Iglesias
2006-08-07 16:31 ` Jamal Hadi Salim
2006-08-07 17:04 ` Edgar E. Iglesias
2006-08-07 18:00 ` jamal
2006-08-07 18:47 ` Edgar E. Iglesias
2006-08-07 19:03 ` jamal
2006-08-07 19:14 ` Edgar E. Iglesias
2006-08-07 19:34 ` jamal
2006-08-07 20:28 ` Edgar E. Iglesias
2006-08-08 0:52 ` jamal
2006-08-07 20:53 ` Brandeburg, Jesse
2006-08-08 1:07 ` jamal
2006-08-07 23:23 ` Herbert Xu
2006-08-07 23:35 ` Brandeburg, Jesse
2006-08-07 23:40 ` Herbert Xu
2006-08-07 16:29 ` Edgar E. Iglesias
2006-08-07 16:36 ` jamal
2006-08-06 19:22 ` jamal
2006-08-08 1:19 ` jamal
2006-08-08 1:22 ` Herbert Xu
2006-08-08 1:33 ` jamal
2006-08-08 2:17 ` Herbert Xu
2006-08-08 3:10 ` jamal
2006-08-08 12:21 ` jamal
2006-08-08 12:39 ` Herbert Xu
2006-08-06 17:20 ` Michael Chan
2006-08-06 23:04 ` Herbert Xu
2006-08-07 3:56 ` Michael Chan
2006-08-07 4:21 ` Herbert Xu
2006-08-08 17:04 ` Benjamin LaHaise
2006-08-08 22:06 ` David Miller [this message]
2006-08-08 23:21 ` Benjamin LaHaise
2006-08-09 0:25 ` Herbert Xu
2006-08-09 1:25 ` Benjamin LaHaise
2006-08-04 1:16 ` jamal
2006-08-04 1:18 ` Herbert Xu
2006-08-04 1:25 ` jamal
2006-08-04 4:06 ` Michael Chan
2006-08-03 22:06 ` jamal
-- strict thread matches above, loose matches on Subject: below --
2006-08-08 5:43 Brandeburg, Jesse
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=20060808.150607.10247901.davem@davemloft.net \
--to=davem@davemloft.net \
--cc=auke-jan.h.kok@intel.com \
--cc=bcrl@kvack.org \
--cc=hadi@cyberus.ca \
--cc=herbert@gondor.apana.org.au \
--cc=jesse.brandeburg@intel.com \
--cc=mchan@broadcom.com \
--cc=netdev@vger.kernel.org \
--cc=shemminger@osdl.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;
as well as URLs for NNTP newsgroup(s).