All of lore.kernel.org
 help / color / mirror / Atom feed
From: Luca Bortot <l.bortot@inet.it>
To: Francois Romieu <romieu@fr.zoreil.com>, netdev@oss.sgi.com
Subject: Re: jumbo on 8169
Date: Mon, 27 Dec 2004 15:00:02 +0100	[thread overview]
Message-ID: <41D01562.4090606@inet.it> (raw)
In-Reply-To: <20041227123136.GA25187@electric-eye.fr.zoreil.com>

Francois Romieu wrote:
> Luca Bortot <l.bortot@inet.it> :
>>To put it short, could you please give me a link or a hint or whatever 
>>to let me set jumbo frames on r8169 (I'm currently running kernel 
>>6.9.10/i386)?
> 
> 
> You can use any recent patch issued by Andrew Morton (-mm) or apply
> http://www.fr.zoreil.com/people/francois/misc/20041218-2.6.10-rc3-r8169.c-test.patch
> 
> I have not regenerated the whole patch against 2.6.10 yet. So if you want
> to apply the aforementionned patch on top of 2.6.10, you will have to revert
> (cd linux-2.6.10; patch -R -p1 -d. < ...) the attached patch first.
> 
> Please note that you will be limited to ~7000 bytes frames at most (but it
> is enough to make a noticeable difference).
> 
> Success/failure report + description of the hardware (lspci -vx/dmesg) will
> be welcome.

It did it as it should:

applied the patch, recompiled & reboot, could now run
ifconfig eth2 mtu 7000

hardware in short:
intel p3 800mhz
384mb ram
m/b QDI Advance 9
nic Hamlet HNNG32TX (realtek 8169 based)

running fedora core 3 / kernel 2.6.10 / NAPI enabled


I'm testing it together with a windows box (which is directly connected 
via a cross cable):
athlon XP 2600
1gb ram
m/b asus a7n8x
same nic
windows XP

based on a simple tcp test I made (writes zeroes to a socket in 32Kb 
blocks and prints the write speed), these are the results (win box cpu 
not reported - always under 10% load):

BEFORE PATCH (mtu 1500)
speed ~38 MB/s
cpu idle 10%
cpu system 90%

AFTER PATCH (mtu 7000)
speed ~45MB/s
cpu idle 40%
cpu system 60%


as requested, lspci /vx

00:0b.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8169 
Gigabit Ethernet (rev 10)
         Subsystem: Realtek Semiconductor Co., Ltd. RTL-8169 Gigabit 
Ethernet
         Flags: bus master, 66Mhz, medium devsel, latency 64, IRQ 5
         I/O ports at dc00 [size=256]
         Memory at e6603000 (32-bit, non-prefetchable) [size=256]
         Expansion ROM at e3000000 [disabled] [size=128K]
         Capabilities: [dc] Power Management version 2
00: ec 10 69 81 17 00 b0 02 10 00 00 02 08 40 00 00
10: 01 dc 00 00 00 30 60 e6 00 00 00 00 00 00 00 00
20: 00 00 00 00 00 00 00 00 00 00 00 00 ec 10 69 81
30: 00 00 00 e3 dc 00 00 00 00 00 00 00 05 01 20 40


Thanks for helping
Luca Bortot

       reply	other threads:[~2004-12-27 14:00 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <41CFF27A.2070008@inet.it>
     [not found] ` <20041227123136.GA25187@electric-eye.fr.zoreil.com>
2004-12-27 14:00   ` Luca Bortot [this message]
2004-12-27 16:38     ` jumbo on 8169 Francois Romieu
2004-12-28 10:13       ` Luca Bortot

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=41D01562.4090606@inet.it \
    --to=l.bortot@inet.it \
    --cc=netdev@oss.sgi.com \
    --cc=romieu@fr.zoreil.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.