Yury O. Tabolin wrote:
> Hi, so, I tried the kernels. First I tried to upgrade kernel to
> 2.6.32-45. Nothing has changed, the rate is same low.
Good, thanks for confirming.
[...]
> Then I installed 3.2.18-1~bpo60+1 kernel from squeeze-backports.
> Segfault disappeared. The rate increased slightly, but still
Ok, so now there are two different tasks to work on:
i) applying the improvement that eliminated the segfault to
squeeze
ii) fixing the rest of the speed regression relative to etch for
future kernels.
For (i): could you try 2.6.37-2 and 2.6.38-rc6 from
snapshot.debian.org and let us know how it goes?
For (ii): if you get a change to try 3.4.y from experimental, that
would be very useful. The only packages from outside squeeze aside
from the kernel itself that should be needed to test it are linux-base
and initramfs-tools, both of which can be found in squeeze-backports.
If the speed regression is still present, we can get help from
upstream. If it's fixed, we can try to find the patch that fixes it
and apply it to the wheezy and squeeze kernels. So either result is
progress.
Thanks again and hope that helps,
Jonathan
--
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20120620032255.GC2961@burratino
