How to prove no receive live_lock in my platform?
From: JoohyeonPark (sunset85_at_intizen.com)
Date: 08/12/04
- Next message: D. Alvarado: "ProFTP time out length"
- Previous message: PenguinsAnonymous_at_NotaChance.com: "Re: Thinkpad wireless masquerade"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: 11 Aug 2004 17:17:09 -0700
Hi
I'm working on IBM750fx customized board.
But I couldn't watch network performance increase in my board with
2.4 kenel
& NAPI patch & e100 driver.
The same code worked well in Intel Machine( PentimIII, Athlon ) and
about
150% performance increase occurred with respect to Interrupt mode.
I inspected network overrun in my platform and Intel platform.
About 200~300 overruns occurred in my platform, while 1000000~2000000
overrun occurred in Intel platform.
But My manager told me rhat this overrun problem was not reason but
results.
And he wanted me what caused no increase of performace in my
platform.
Are there other points in code which can prove no performance
increase in
my platform?
I would be very glad to receive your sincere reply.
Thanks.
- Next message: D. Alvarado: "ProFTP time out length"
- Previous message: PenguinsAnonymous_at_NotaChance.com: "Re: Thinkpad wireless masquerade"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|