Re: bw_tcp & bw_unix throughput loss kernels >= 2.3.99-pre7

From: Manfred Spraul (manfred@colorfullife.com)
Date: Sun Jun 25 2000 - 10:01:50 EST

Next message: Nils Philippsen: "X hangs with 2.2.0-test{1-acXX,2} kernels"
Previous message: Erick Kinnee: "UDMA(33)/(66) or What the heck is going on?"
In reply to: Mike Galbraith: "bw_tcp & bw_unix throughput loss kernels >= 2.3.99-pre7"
Next in thread: Artur Skawina: "Re: bw_tcp & bw_unix throughput loss kernels >= 2.3.99-pre7"
Reply: Artur Skawina: "Re: bw_tcp & bw_unix throughput loss kernels >= 2.3.99-pre7"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Mike Galbraith wrote:
>
> Greetings,
>
> Noting a drop in both bw_unix and bw_tcp lmbench throughput numbers
> (bw_unix: 130->57 MB/sec bw_tcp: 94->46 MB/sec) on my box while testing
> 2.4.0-test2, I did a binary search.. 2.3.99-pre7 is the switch point.
> Box is 128mb PIII/500.. config info attached.
>
> Anyone else see this?
>
I assume everyone sees this:

Linus enabled SLAB_POISON in pre7.

Perhaps someone should build a list of all debugging checks that are
enabled in 2.4 and disabled in 2.2, otherwise the benchmark results are
meaningless.

SLAB_DEBUG
WAITQUEUE_DEBUG
unlock_kernel() contains a debug check
what else?

-- Manfred

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/

Next message: Nils Philippsen: "X hangs with 2.2.0-test{1-acXX,2} kernels"
Previous message: Erick Kinnee: "UDMA(33)/(66) or What the heck is going on?"
In reply to: Mike Galbraith: "bw_tcp & bw_unix throughput loss kernels >= 2.3.99-pre7"
Next in thread: Artur Skawina: "Re: bw_tcp & bw_unix throughput loss kernels >= 2.3.99-pre7"
Reply: Artur Skawina: "Re: bw_tcp & bw_unix throughput loss kernels >= 2.3.99-pre7"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

This archive was generated by hypermail 2b29 : Mon Jun 26 2000 - 21:00:07 EST