lkml.org 
[lkml]   [1998]   [Aug]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: kernel compile time comparison (2.0 vs 2.1 with 64MB)
Date
Hi mingo,

On Mon, 03 Aug 1998, MOLNAR Ingo wrote:
....
>> > this is not necessarily true for all systems. Although the system in
>> > question was a UP box, it's i think worth to mention that compilation
>> > speeds on SMP improve measurably with 2.1.
>>
>> Because of finer-grained kernel locking, yes. [...]
>
>simply because we schedule better.

Well, I tend to believe statements like this when I see numbers. What exactly
do you mean by "schedule better" and how do you measure this?

>> > also, compilation _can_ be 'kernel-bound' on UP too, just try compiling
>> > Linux on a say 4MB box. (or 8MB)
>>
>> Bad example: then you are not kernel bound, but memory (swap) bound. And
>> 2.0.x is noticeably *faster* than 2.1.x in low memory machines.
>
>hm, and where is the swapping logic? in the kernel ...

When you begin swapping, you are really measuring disk i/o performance. Before
you begin swapping, you are really measuring CPU performance.

In low memory machines, 2.1.x begins swapping earlier than 2.0.x, so even
though 2.1.x may have better "swapping logic", it performs worse (there was
plenty of data to support this statement lately on the list).

In machines with sufficient memory, no swapping occurs, so even though 2.1.x
may have better "swapping logic", it performs no better than 2.0.x (*WRT
swapping*).

This is the typical case where more means less: more swap heuristics and more
code mean less performance.

>we do not depend on the speed of the kernel, we depend on the decisions
>made by the kernel. (or rather, kernel developers ;)
>

Contrarily to popular belief, kernel developers are not Gods. They sometimes
make the wrong decisions.

It is however noticeably harder to make them accept this simple fact. ;)

Cheers,
--
Andrew D. Balsa
andrebalsa@altern.org


-
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.altern.org/andrebalsa/doc/lkml-faq.html

\
 
 \ /
  Last update: 2005-03-22 13:43    [W:0.047 / U:0.052 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site