[PATCH] fixes for u-boot and kernel to handle GPS properly. Was: [Qtmoko] qtmoko v28, GPS issue, Qi & uboot

[Gennady Kupava]

Hi, all. 

Seem i fixed this one. Need fix for both kernel and u-boot. Kernel fix
is complete harmless, it will work under any conditions.

U-boot patch:
http://www.bsdmn.com/openmoko/gpsfix/uboot_do_not_touch_serials.patch
Idea of this patch is to keep u-boot from controlling serial's GPIO
configuration. It turned out that u-boot also touch GSM port
configuration on resume. This turns this off.

Kernel patch:
http://www.bsdmn.com/openmoko/gpsfix/kernel_Force-GPS-power-up-on-resume-if-it-were-powered-up-on-suspend.patch

I also did precompiled versions for whoose who using u-boot and qtmoko:

Compiled u-boot (242 + this fix):
http://www.bsdmn.com/openmoko/gpsfix/u-boot_242_gps.udfu
Compiled kernel (qtmoko29 + latest ts jitter + this fix):
http://www.bsdmn.com/openmoko/gpsfix/uImage-b6.bin
Kernel modules: http://www.bsdmn.com/openmoko/gpsfix/modules-b6.tar.gz

Extract kernel to /, do depmod -a 2.6.34b6-v28, flash kernel, u-boot and
it should work.

Regards,
Gennady.

В Срд, 27/10/2010 в 18:33 +0400, Gennady Kupava пишет: 
> Hi, Radek, David.
> 
> > > 5. really unsure: may be also something wrong with GPS. Seem it works
> > > well after boot, but after resume it had problems to get fix. But same
> > > were in .29.
> > 
> > It would be nice to have application which shows GPS sattelites for debugging 
> > this. In fact such app is already there [1] but works only with GPSD. I wanted 
> > to modify qtopia gps api so that it provides GPSD compatible socket so that 
> > gpsd apps like this one and e.g. tangogps can work out of the box.
> 
> Yes! Such application is really really missing. Something with just
> displaying some kind debugging info and satellites signal levels, like
> in my Garmin would be very useful, as it will help greatly. Is it
> getting fix or not? how long it will take? how much satellites visible?
> Current situation in nerongps (fix/not fix) is not really informative.
> Also it is great idea to support most common universal unix GPS
> interface.
> 
> > Could you try with Qi? I've not tried with qtmoko, but I have the same
> sympthoms with u-boot + shr.
> 
> Nice hint, thanks! If so fix for u-boot should be relatively simple.
> Something related to serials setup :) I'll try to fix this... later.
> 
> btw, from thread:
> 
> > "One difference between uboot and qi is that qi doesn't initialize the
> display."
> 
> NO! I am sure where is much more difference. I am sure main problem of u-boot were 
> 'too complicated to make changes, let's rewrite'. I am sure where is MORE difference, as 
> bootloader seem responsible to initial init of many openmoko's hardware.
> For example, with NOR u-boot, current in suspend in 10 times more than with NAND one with same
> kernel. Most voltages, memory timings, cpu frequency are set in bootloader and not touched by kernel 
> (it's just what i know). Serial ports seem setup also somehow related to bootloader.
> 
> Btw, as we have absolutely no WSODs no with u-boot (which inits display), and WSOD issue
> can be attributed to particular patches, where is no reason not to init display.
> 
> Regards,
> Gennady.
> 
>