kernel defconfig, debugging, preemption, and very noticeable speedups/ debugging

Gennady Kupava gb at bsdmn.com
Thu Jan 14 09:56:21 CET 2010


Hi, Martin,

Here is the test results for .29 with drm and .32:

http://www.bsdmn.com/openmoko/debugtest2/resultsnew.txt

It look like .29 with drm is a bit faster than defconfig

.32 do not have debug options turned on, so almost everywhere it is on
the par with defconfig, most places it is slightly better.

Gennady

В Втр, 12/01/2010 в 13:47 +0100, Martin Jansa пишет:
> On Tue, Jan 12, 2010 at 01:59:28PM +0300, Gennady Kupava wrote:
> > hi all,
> 
> Hi Gennady,
> 
> > here is test results for individual options many people wanted:
> > 
> > http://www.bsdmn.com/openmoko/debugoptstesting/
> 
> Wow, great numbers!, thanks for those test... I'm looking forward for
> kernel guys decision and I'll apply it to SHR-U as soon as they decide.
> 
> As you already have this testsuite, can you please try it with those 3
> kernels (first one 2.6.29-drm is the same as andy-tracking with DRM
> patches from Thomas Weiss - there shouldn't be much difference, and the
> .31, .32 is from corresponding openmoko kernel branches also with
> applied DRM patches and we would like to switch SHR-U to .32 as soon as
> suspend is resolved there..
> 
> http://build.shr-project.org/tests/mrmoku/kms/images/om-gta02/
> 
> or just latest .32 if you don't want to spend more time with it..
> http://build.shr-project.org/tests/mrmoku/kms/images/om-gta02/uImage-2.6.32-oe1+gitr0+5e5d518ecad86cc891d91c88d3b144eb30339bca-r4-om-gta02.bin
> 
> 
> Thanks a lot!
> 






More information about the openmoko-kernel mailing list