00:20 <steev> Wizzup: i've tried what was on the linux-exynos wiki, what comes in the linux-firmware repo, and the arch _cros.bin firmware

00:20 <steev> i can't wait to get to Austin tomorrow

00:20 <steev> all of them go tits up

00:20 <steev> and it doesn't help that the wifis here in vegas suck ass

00:20 <steev> but i won't be there til 4pm

00:21 <steev> conceivably, i could just use the rt3070

06:46 <sjoerd> steev: to get wifi working on mainline we need some regulator changes first at least

06:48 <javier__> Wizzup: sure, I'm ok if you create me an account

06:55 <javier__> steev: what sjoerd said, to make the wifi to work (or anything hooked to the MMC interface) the voltage skew between rails vdd_arm and vdd_int has to be kept between a minimum (350mV iirc)

06:56 <javier__> steev: take a look at the int-voltage-locker device node in chrome os arch/arm/boot/dts/exynos5422-peach-pi.dts

06:57 <javier__> steev: the corresponding driver is drivers/regulator/locker.c

06:57 <javier__> what this driver does is to always keep the voltage skew less than maximum-spread = <300000>

07:00 <javier__> steev: that approach was posted but upstream said that the regulator core has to be extended to support such a use case: https://lkml.org/lkml/2014/4/30/730

11:16 <Wizzup> javier__: ack

12:14 <tfiga> reading the above I get to a conclusion that the claims about a race between software engineers trying to design more and more generic software

12:14 <tfiga> and hardware engineers in turn designing more and more twisted hardware might not be exaggerated...

16:06 <si1v3r> tfiga: Java on a chip.......... nuff said.

16:06 <tfiga> :)