Regressions in 2.5.69
From: Andrew Morton <[email protected]>
Date: Fri, 16 May 2003 00:47:07 -0700
To: "Paolo Ciarrocchi" <[email protected]>
Subject: Re: [BENCHMARK AIM9] Regressions in 2.5.69
Quoted 18 lines
Quoted 10 lines · Paolo Ciarrocchi
Hi all/Andrew/Martin,
I noticed regression in a few tests,
I deleted the results of tests that don't show differences between the two kernel version.Hope it helps.
Ciao,
PaoloQuoted 6 lines
2.5.67
2.5.69creat-clo 10010 86.8132 86813.19 File Creations and Closes/second
creat-clo 10030 22.0339 22033.90 File Creations and Closes/second
^^^^BIG REGRESSION
Quoted 5 lines
I cannot repeat any of this. In fact 2.5.69-mm is a bit faster than 2.5.67.
I tested ext2 mainly. But a spot-check of creat-clo on reiserfs showed no regression either.
Ok, thanks Andrew.
I dunno what happened during the test.
I'll try again to run the test with both the kernel. I'll back to you if I'm able to reproduce those numbers.
Ciao,
Paolo
Signature · 5 lines
______________________________________________ http://www.linuxmail.org/ Now with e-mail forwarding for only US$5.95/yr Powered by Outblaze