Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*\[PATCH\s+2\.6\.12\.1\s+5\/12\]\s+S2io\:\s+Performance\s+improvements\s*$/: 42 ]

Total 42 documents matching your query.

1. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "Ravinandan Arakali" <ravinandan.arakali@xxxxxxxxxxxx>
Date: Tue, 2 Aug 2005 16:13:57 -0700
Hi Christoph, Following is SGI's stand on this issue: SGI recommends that customers use the -sn2 kernel. This is the kernel that is installed by our factory when we ship systems. The -sn2 kernel is a
/archives/netdev/2005-08/msg00006.html (9,328 bytes)

2. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "'Christoph Hellwig'" <hch@xxxxxxxxxxxxx>
Date: Wed, 3 Aug 2005 00:26:45 +0100
The my argument is wrong for SuSE ;-) This still needs to be a runtime switch though, not just for this reason. Platform ifdefs are not the way to go.
/archives/netdev/2005-08/msg00007.html (8,931 bytes)

3. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: Prarit Bhargava <prarit@xxxxxxx>
Date: Wed, 03 Aug 2005 08:48:30 -0400
On Tue, Aug 02, 2005 at 04:13:57PM -0700, Ravinandan Arakali wrote: Hi Christoph, Following is SGI's stand on this issue: SGI recommends that customers use the -sn2 kernel. This is the kernel that is
/archives/netdev/2005-08/msg00008.html (8,658 bytes)

4. [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: raghavendra.koushik@xxxxxxxxxxxx
Date: Thu, 7 Jul 2005 15:27:41 -0700 (PDT)
This patch relates to mostly performance related changes. 1. Fixed incorrect computation of PANIC level in rx_buffer_level(). 2. Removed unnecessary PIOs(read/write of tx_traffic_int and rx_traffic_
/archives/netdev/2005-07/msg00093.html (15,338 bytes)

5. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: Arthur Kepner <akepner@xxxxxxx>
Date: Thu, 7 Jul 2005 16:15:20 -0700 (PDT)
I thought that an mmiowb() was called for here (to order the PIO writes above more cheaply than doing the readq()). I posted a patch like this some time ago: http://marc.theaimsgroup.com/?l=linux-ne
/archives/netdev/2005-07/msg00103.html (10,817 bytes)

6. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "Raghavendra Koushik" <raghavendra.koushik@xxxxxxxxxxxx>
Date: Thu, 7 Jul 2005 18:06:19 -0700
On an Altix machine I believe the readq was necessary to flush the PIO writes. How long did you run the tests? I had seen in long duration tests that an occasional write (TXDL control word and the a
/archives/netdev/2005-07/msg00108.html (13,236 bytes)

7. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxxxxx>
Date: Thu, 07 Jul 2005 20:00:34 -0700 (PDT)
wmb() does no such thing. It only has influence on load and store instructions done by the local processor, it has no effect on what the PCI bus may do with PIO writes (ie. post them). If you need a
/archives/netdev/2005-07/msg00109.html (10,210 bytes)

8. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: Jeff Garzik <jgarzik@xxxxxxxxx>
Date: Thu, 07 Jul 2005 23:08:02 -0400
David S. Miller wrote: If you need a PIO to complete in a specific order, you have to read it back. If you need PIO operations to occur Correct. A PCI read is the only way to ensure that all the CPU/
/archives/netdev/2005-07/msg00110.html (10,838 bytes)

9. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: Arthur Kepner <akepner@xxxxxxx>
Date: Fri, 8 Jul 2005 08:31:21 -0700 (PDT)
The most recent tests I did used pktgen, and they ran for a total time of ~.5 hours (changing pkt_size every 30 seconds or so). The pktgen tests and other tests (like nttcp) have been run several tim
/archives/netdev/2005-07/msg00114.html (11,269 bytes)

10. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "Raghavendra Koushik" <raghavendra.koushik@xxxxxxxxxxxx>
Date: Fri, 8 Jul 2005 11:16:14 -0700
I'll include this fix in the next patch that incorporates any other review comments coming my way.. Thanks for pointing it out. -Koushik
/archives/netdev/2005-07/msg00116.html (12,498 bytes)

11. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "Ravinandan Arakali" <ravinandan.arakali@xxxxxxxxxxxx>
Date: Fri, 8 Jul 2005 11:17:29 -0700
Thanks for pointing that out. We will wait for any other comments on our 12 patches. If there are no other, will send out a patch13 to include the mmiowb() change. Thanks, Ravi --Original Message--
/archives/netdev/2005-07/msg00117.html (12,211 bytes)

12. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Tue, 12 Jul 2005 21:27:54 +0100
this enabled it only on kernel that are built to only run on SN2 hardware, which is completely useless in practice. Besides that defining a CONFIG_ symbol from source files is a big no-go. What exac
/archives/netdev/2005-07/msg00129.html (10,604 bytes)

13. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxxxxx>
Date: Tue, 12 Jul 2005 13:34:04 -0700 (PDT)
Yes, do this in the Kconfig file instead.
/archives/netdev/2005-07/msg00130.html (10,656 bytes)

14. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "Leonid Grossman" <leonid.grossman@xxxxxxxxxxxx>
Date: Tue, 12 Jul 2005 13:56:55 -0700
In short, this is one of the ASIC modes where headers and payload are separated by the hardware, and placed in separate receive buffers. (More details are in the ASIC programming manual that is post
/archives/netdev/2005-07/msg00132.html (11,311 bytes)

15. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "Ravinandan Arakali" <ravinandan.arakali@xxxxxxxxxxxx>
Date: Tue, 12 Jul 2005 14:00:52 -0700
The two-buffer mode was added as a configurable option to Kconfig file several months ago. Hence the macro is CONFIG_2BUFF_MODE. The two-buffer receive mode involves two buffers (128 byte aligned) fo
/archives/netdev/2005-07/msg00133.html (11,121 bytes)

16. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxxxxx>
Date: Tue, 12 Jul 2005 14:04:11 -0700 (PDT)
We're saying that you should choose CONFIG_2BUFF_MODE, when CONFIG_IA64_SGI_SN2 is set, inside the Kconfig file using the "default" Kconfig directive. You should never change the setting of CONFIG_*
/archives/netdev/2005-07/msg00134.html (10,674 bytes)

17. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Tue, 12 Jul 2005 22:07:04 +0100
And that doesn't help either, CONFIG_IA64_SGI_SN2 isn't used in practice, any production Altix with 26 kernels runs CONFIG_IA64_GENERIC kernels. You have to make this run-time selectable.
/archives/netdev/2005-07/msg00135.html (11,142 bytes)

18. Re: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: Andi Kleen <ak@xxxxxxx>
Date: 12 Jul 2005 23:26:41 +0200
At least SLES9/ia64 has a SN2 kernel. -Andi
/archives/netdev/2005-07/msg00136.html (10,002 bytes)

19. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "Ravinandan Arakali" <ravinandan.arakali@xxxxxxxxxxxx>
Date: Tue, 12 Jul 2005 14:54:28 -0700
Okay, got it. Will submit patch on Kconfig file after the macro for SGI platforms is confirmed. Ravi --Original Message-- From: David S. Miller [mailto:davem@xxxxxxxxxxxxx] Sent: Tuesday, July 12, 20
/archives/netdev/2005-07/msg00137.html (10,653 bytes)

20. RE: [PATCH 2.6.12.1 5/12] S2io: Performance improvements (score: 1)
Author: "Ravinandan Arakali" <ravinandan.arakali@xxxxxxxxxxxx>
Date: Fri, 29 Jul 2005 09:37:55 -0700
We are trying to use the "default" directive in Kconfig. We tried using an unconditional directive(just to test it out) such as "default y" and a conditional one such as "default y if CONFIG_IA64_SG
/archives/netdev/2005-07/msg00209.html (10,999 bytes)


This search system is powered by Namazu