In article <20051018132617.GB27100@artax.karlin.mff.cuni.cz>
pcah8322@artax.karlin.mff.cuni.cz wrote:

> I tested the speed of a kernel with L2 cache patch built by Tsutsui-san.
> Results show that the patch probably works as expected, thanks!

Thank you for your testing, too :-)

> here is the difference of dmesg, which shows one probably unrelated
> improvement: the setting for SCSI devices are better.

Ah, my tree has another workaround patch for on-board ahc,
so it isn't cache related. The problem is filed in kern/23276:
http://www.netbsd.org/cgi-bin/query-pr-single.pl?number=23276

My patch is different from the PR, but anyway we should have
some proper options to ignore device configurations set by BIOS
(which may be conservative, i.e. sync and some options are
 disabled because they are not needed to load bootloader or kernel)
and probe them by itself if the adapter doesn't have SEEPROM
to store own configurations.
---
Izumi Tsutsui