Re: buffers vs. pages vs. kernel speed

Dr. Werner Fink (
Fri, 13 Jun 1997 12:32:39 +0200

> Sorry. I usually only apply patches from certain people like David. For
> some reason I did not see David's comment on your additional patch. I can
> miss such things because I don't have the time to follow everything. So I
> promise not to report about problems in prepatches any more.

Pardon ... this was not what I want ... please continue with testing and
reporting. Please apply the one line patch I've send and make
stress tests ... like

while true; do make clean; make -j zImage; done

in two distinct kernel trees ... and untaring big archives from disk to
disk ... and making big files system with low memory (booting with
the kernel parameter mem=4MB) ... and ... and ...