Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*3c59x\s+\(was\s+Route\s+cache\s+performance\s+under\s+stress\)\s*$/: 18 ]

Total 18 documents matching your query.

1. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: xxxxxxxx>
Date: Tue, 10 Jun 2003 18:15:07 +0200 (CEST)
and it
/archives/netdev/2003-06/msg00269.html (11,462 bytes)

2. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: tescu <bogdan.costescu@xxxxxxxxxxxxxxxxxxxxx>
Date: Tue, 10 Jun 2003 18:20:29 +0200
hange t
/archives/netdev/2003-06/msg00271.html (11,421 bytes)

3. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: xxxxxxxxxxxxxxxxxxxx>
Date: Tue, 10 Jun 2003 12:23:42 -0400
test s
/archives/netdev/2003-06/msg00272.html (10,971 bytes)

4. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: xxxxxxxx>
Date: Tue, 10 Jun 2003 18:45:03 +0200 (CEST)
if they
/archives/netdev/2003-06/msg00277.html (12,121 bytes)

5. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Miller" <davem@xxxxxxxxxx>
Date: Tue, 10 Jun 2003 18:49:49 +0200
nterrup
/archives/netdev/2003-06/msg00278.html (10,725 bytes)

6. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: 明 <yoshfuji@xxxxxxxxxxxxxx>
Date: Wed, 11 Jun 2003 11:54:34 +0200
racoon/
/archives/netdev/2003-06/msg00332.html (10,970 bytes)

7. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: ak@xxxxxxx>
Date: Wed, 11 Jun 2003 12:05:20 +0200
s. --ro
/archives/netdev/2003-06/msg00333.html (12,285 bytes)

8. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: der <lpetande@xxxxxxxxxx>
Date: Wed, 11 Jun 2003 12:38:33 +0200
r IP" b
/archives/netdev/2003-06/msg00334.html (12,275 bytes)

9. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: lf Baechle <ralf@xxxxxxxxxxxxxx>
Date: Wed, 11 Jun 2003 08:08:00 -0400 (EDT)
. -Andi
/archives/netdev/2003-06/msg00339.html (13,079 bytes)

10. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Bogdan Costescu <bogdan.costescu@xxxxxxxxxxxxxxxxxxxxx>
Date: Tue, 10 Jun 2003 18:15:07 +0200 (CEST)
I did try about 2 years ago and converted the driver to MMIO. I wasn't able to see _any_ kind of improvement and I was using it in parallel computation where latency counts. I have to say though that
/archives/netdev/2003-06/msg01168.html (11,500 bytes)

11. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Andi Kleen <ak@xxxxxxx>
Date: Tue, 10 Jun 2003 18:20:29 +0200
Can't you just wrap it in a few macros and offer a config for those who want the best performance and a runtime test for the others? Then switch between PIO and mmio dynamically. Even runtime test s
/archives/netdev/2003-06/msg01170.html (11,555 bytes)

12. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Jeff Garzik <jgarzik@xxxxxxxxx>
Date: Tue, 10 Jun 2003 12:23:42 -0400
I prefer a compile-time test. But yes, this is what several other net drivers do: offer a config option for MMIO (or PIO), and the default is MMIO unless that is known to be unsafe on certain boards
/archives/netdev/2003-06/msg01171.html (11,115 bytes)

13. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Bogdan Costescu <bogdan.costescu@xxxxxxxxxxxxxxxxxxxxx>
Date: Tue, 10 Jun 2003 18:45:03 +0200 (CEST)
I don't really want to sound like defending the 3c59x driver, but... The 3c90x driver released by 3Com uses some mechanism "similar" to NAPI which is based on the on-board timer; these timer interrup
/archives/netdev/2003-06/msg01176.html (12,159 bytes)

14. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Andi Kleen <ak@xxxxxxx>
Date: Tue, 10 Jun 2003 18:49:49 +0200
You can play some tricks with the driver to make eth_type_trans disappear from the profiles. This usually helps a lot because it avoids one full "fetch from cache cold memory" roundtrip per packet,
/archives/netdev/2003-06/msg01177.html (10,859 bytes)

15. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>
Date: Wed, 11 Jun 2003 11:54:34 +0200
Andi! Interesting. Can we get into details? Cheers. --ro
/archives/netdev/2003-06/msg01231.html (11,116 bytes)

16. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Andi Kleen <ak@xxxxxxx>
Date: Wed, 11 Jun 2003 12:05:20 +0200
eth_type_trans checks the ethernet protocol ID and sets the broadcast/multicast/ unicast L2 type. Some NICs have bits in the RX descriptor for most of them. They have a "packet is TCP or UDP or IP" b
/archives/netdev/2003-06/msg01232.html (12,465 bytes)

17. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>
Date: Wed, 11 Jun 2003 12:38:33 +0200
Thanks! Yes. Like to give this a try when I got chance. It should be something for the driver authors. Any patch handy? e1000? Cheers. --ro
/archives/netdev/2003-06/msg01233.html (12,483 bytes)

18. Re: 3c59x (was Route cache performance under stress) (score: 1)
Author: Jamal Hadi <hadi@xxxxxxxxxxxxxxxx>
Date: Wed, 11 Jun 2003 08:08:00 -0400 (EDT)
I have done prefetching experiments with a NAPIezed sb1250.c driver on MIPS. I never got rid of eth_type_trans() just prefetched skb->data a few lines before calling it. I did see eth_type_trans() al
/archives/netdev/2003-06/msg01238.html (13,287 bytes)


This search system is powered by Namazu