[Gta04-owner] Fwd: [beagleboard] 1ghz news?

[Dr. H. Nikolaus Schaller]

Hi Neil,
great progress!

Am 28.07.2013 um 01:20 schrieb NeilBrown:

> On Thu, 25 Jul 2013 08:32:20 +0200 "Dr. H. Nikolaus Schaller"
> <hns at goldelico.com> wrote:
> 
>> 
>> Am 24.07.2013 um 15:57 schrieb Dr. H. Nikolaus Schaller:
>> 
>>> may be of interest for you:
>>> 
>>> again some progress on 1 GHz kernels (except for USB stability).
>> 
>> And someone has written a blog about this 1 GHz DM3730CBP100 issue:
>> 
>> http://teknoman117.wordpress.com/2013/07/24/beagleboard-xm-1-ghz-operation-safely-and-reliably/
>> 
>> One statement is very important for us as well and it shows how dramatic the situation is with newer kernels (and we did not really notice since all devices were tested with the 2.6.32 hw-validation kernel):
>> 
>> "the last time it was safe was kernel 3.0.28 back in early 2012"
>> 
>> So we should also work on:
>> 
>> http://projects.goldelico.com/p/gta04-kernel/issues/440/
>> 
> 
> If anyone wants to look into this, I've updated my 'mainline' branch to
> 3.11-rc2 (or close to that).

yes, definitively!

> Lots of things work.  The most obvious problem is that wifi is interfering
> with suspend/resume.  If you remove all the wifi modules, suspend/resume
> seems to work OK.  With wifi loaded, resume doesn't.
> 
> I didn't rebase my branch at all this time, just added more stuff on top.  So
> there is a reasonable change that it will merge with the "neil-plus" branch
> without too much pain.

I did the full merge [1], and only needed minimal changes to make it work on my environment.

1. it appears that video/omap-panel-tpo-td028ttec1.h is missing in the tree so I applied Marek's patch again
2. I see the same issue with the 8 MB omapfb allocation as before - but I could detect that it is somehow related to take the bootargs provided by U-Boot (the official u-boot provides more arguments than those set with CONFIG_CMDLINE)
3. the kernel still hangs after 'reboot' or 'reboot -f' (it is not dependent on USB which I though first to be the problem, does not appear after 'powerdown' or battery removal).

I have collected all known issues [2].

> 
> BTW my board runs at 1GHz *without* the adaptive-body-bias stuff.  However
> maybe it isn't "safe" ... i.e. presumably it can overheat or something.  I
> don't run anything that would keep it at 1GHz more than a few seconds at a
> time.
> 
> NeilBrown
> 


We are now for the first time at the cutting edge of Kernels for the GTA04. The next phase will be to go further, i.e. lead the official kernels. Not kidding :)
This means upstreaming what we have newly developed and add in the future... Openpandora.org and Beagleboard.org are already in this "contributing" mode.

BR and thanks for your great work to bring us so far! An I hope there will be as much support by you and others in the future as well, because with a leading kernel maintained on kernel.org we have better chances to get people interested in what we do.

Nikolaus

[1]: http://git.goldelico.com/?p=gta04-kernel.git;a=shortlog;h=refs/heads/neil-3.1x-plus
[2]: http://projects.goldelico.com/p/gta04-kernel/issues/label/1504/open/