From: linas@austin.ibm.com (Linas Vepstas)
To: Arnd Bergmann <arnd@arndb.de>
Cc: cbe-oss-dev@ozlabs.org, netdev@vger.kernel.org,
Jeff Garzik <jeff@garzik.org>
Subject: Re: [Cbe-oss-dev] [PATCH 12/15] spidernet: increase the NAPI weight
Date: Thu, 14 Jun 2007 17:08:29 -0500 [thread overview]
Message-ID: <20070614220829.GX4397@austin.ibm.com> (raw)
In-Reply-To: <200706132249.51894.arnd@arndb.de>
On Wed, Jun 13, 2007 at 10:49:51PM +0200, Arnd Bergmann wrote:
> On Wednesday 13 June 2007, Jeff Garzik wrote:
> > > +/* We really really want to empty the ring buffer every time,
> > > + * so as to avoid the RX ram full bug. So set te napi wieght
> > > + * to the ring size.
> > > + */
> > > +#define SPIDER_NET_NAPI_WEIGHT SPIDER_NET_RX_DESCRIPTORS_DEFAULT
> >
> > I don't see why spider_net should have a different NAPI weight from
> > other drivers
It was a lame attempt to try to trick napi into draining the entire
RX queue in one go, with the goal of avoiding the dreaded rx ram full.
I'm not sure it made much of a difference, so we can let this slide.
> Would it help to do it the other way round, as in
No, that would shorten the RX queue, thus making it more likely
to overflow. At gigabit speeds, its petty easy to fill this thing
up multiple times per jiffy. The driver should continue to operate
either way, but the larger queue should keep it from being a busy
beaver.
--linas
next prev parent reply other threads:[~2007-06-14 22:08 UTC|newest]
Thread overview: 69+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-06-07 19:17 [PATCH 0/18] spidernet driver bug fixes Linas Vepstas
2007-06-07 19:20 ` [PATCH 1/18] spidernet: skb used after netif_receive_skb Linas Vepstas
2007-06-07 19:22 ` [PATCH 2/18] spidernet: checksum and ethtool Linas Vepstas
2007-06-07 19:24 ` [PATCH 3/18] spidernet: beautify error messages Linas Vepstas
2007-06-07 19:25 ` [PATCH 4/18] spidernet: move a block of code around Linas Vepstas
2007-06-07 19:27 ` [PATCH 5/18] spidernet: zero out a pointer Linas Vepstas
2007-06-07 19:29 ` [PATCH 6/18] spidernet: null out skb pointer after its been used Linas Vepstas
2007-06-07 19:33 ` [PATCH 7/18] spidernet: Don't terminate the RX ring Linas Vepstas
2007-06-07 19:35 ` [PATCH 8/18] spidernet: enhance the dump routine Linas Vepstas
2007-06-07 19:39 ` [PATCH 9/18] spidernet: reset the card when an rxramfull is seen Linas Vepstas
2007-06-07 19:41 ` [PATCH 10/18] spidernet: service TX later Linas Vepstas
2007-06-07 19:43 ` [PATCH 11/18] spidernet: increase the NAPI weight Linas Vepstas
2007-06-07 19:45 ` [PATCH 12/18] spidernet: don't flag rare packets as bad packets Linas Vepstas
2007-06-07 19:51 ` [PATCH 13/18] spidernet: Cure RX ram full bug Linas Vepstas
2007-06-07 19:53 ` [PATCH 14/18] spidernet: silence the ramfull messages Linas Vepstas
2007-06-07 19:55 ` [PATCH 15/18] spidernet: minor RX optimization Linas Vepstas
2007-06-07 19:57 ` [PATCH 16/18] spidernet: fix misnamed flag Linas Vepstas
2007-06-07 20:01 ` [PATCH 17/18] spidernet: turn off descriptor chain end interrupt Linas Vepstas
2007-06-07 20:05 ` [PATCH 18/18] spidernet: driver docmentation Linas Vepstas
2007-06-08 1:12 ` [Cbe-oss-dev] [PATCH 0/18] spidernet driver bug fixes Michael Ellerman
2007-06-08 17:06 ` Linas Vepstas
2007-06-08 17:20 ` Jeff Garzik
2007-06-11 18:14 ` [PATCH 0/15] " Linas Vepstas
2007-06-11 18:17 ` [PATCH 1/15] spidernet: null out skb pointer after its been used Linas Vepstas
2007-06-11 18:21 ` [PATCH 2/15] spidernet: Cure RX ram full bug Linas Vepstas
2007-06-11 18:23 ` [PATCH 3/15] spidernet: Don't terminate the RX ring Linas Vepstas
2007-06-11 18:26 ` [PATCH 4/15] spidernet: silence the ramfull messages Linas Vepstas
2007-06-13 20:12 ` Jeff Garzik
2007-06-14 22:29 ` Linas Vepstas
2007-06-14 23:12 ` [PATCH] spidernet: Replace literal with const Linas Vepstas
2007-07-02 12:37 ` Jeff Garzik
2007-06-11 18:29 ` [PATCH 5/15] spidernet: turn off descriptor chain end interrupt Linas Vepstas
2007-06-11 18:32 ` [PATCH 6/15] spidernet: skb used after netif_receive_skb Linas Vepstas
2007-06-11 18:35 ` [PATCH 7/15] spidernet: checksum and ethtool Linas Vepstas
2007-06-11 18:41 ` [PATCH 8/15] spidernet: beautify error messages Linas Vepstas
2007-06-13 20:15 ` Jeff Garzik
2007-06-11 18:48 ` [PATCH 9/15] spidernet: enhance the dump routine Linas Vepstas
2007-06-11 18:52 ` [PATCH 10/15] spidernet: invalidate unused pointer Linas Vepstas
2007-06-11 18:59 ` [PATCH 11/15] spidernet: service TX later Linas Vepstas
2007-06-11 19:02 ` [PATCH 12/15] spidernet: increase the NAPI weight Linas Vepstas
2007-06-13 20:14 ` Jeff Garzik
2007-06-13 20:49 ` [Cbe-oss-dev] " Arnd Bergmann
2007-06-14 22:08 ` Linas Vepstas [this message]
2007-06-11 19:05 ` [PATCH 13/15] spidernet: move a block of code around Linas Vepstas
2007-06-11 19:09 ` [PATCH 14/15] spidernet: fix misnamed flag Linas Vepstas
2007-06-11 19:12 ` [PATCH 15/15] spidernet: driver docmentation Linas Vepstas
2007-06-13 20:10 ` [PATCH 1/15] spidernet: null out skb pointer after its been used Jeff Garzik
2007-06-14 22:00 ` Linas Vepstas
2007-06-12 2:01 ` [PATCH 0/15] spidernet driver bug fixes Michael Ellerman
2007-06-12 23:00 ` Jeff Garzik
2007-06-12 23:32 ` Linas Vepstas
2007-06-13 0:04 ` Jeff Garzik
2007-06-13 16:14 ` Linas Vepstas
2007-06-13 18:51 ` Jeff Garzik
2007-06-13 19:01 ` [Cbe-oss-dev] " Segher Boessenkool
2007-06-13 19:02 ` Jeff Garzik
2007-06-13 20:52 ` Arnd Bergmann
2007-06-13 23:55 ` Michael Ellerman
2007-06-13 18:52 ` Jeff Garzik
2007-06-14 22:08 ` [Cbe-oss-dev] " David Woodhouse
2007-06-14 23:01 ` Jeff Garzik
2007-06-14 23:03 ` David Woodhouse
2007-06-14 23:04 ` Jeff Garzik
2007-06-14 23:07 ` David Woodhouse
2007-06-14 23:32 ` Michael Ellerman
2007-06-13 1:33 ` Michael Ellerman
2007-06-13 1:54 ` Jeff Garzik
2007-06-13 13:53 ` Michael Ellerman
2007-06-13 18:45 ` Jeff Garzik
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=20070614220829.GX4397@austin.ibm.com \
--to=linas@austin.ibm.com \
--cc=arnd@arndb.de \
--cc=cbe-oss-dev@ozlabs.org \
--cc=jeff@garzik.org \
--cc=netdev@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;
as well as URLs for NNTP newsgroup(s).