IRC log for #arm-netbook on 20100327

00:58.50*** join/#arm-netbook SQlvpapir (~teis@188.177.95.62)
09:59.16*** join/#arm-netbook lkcl (~lkcl@nat66.mia.three.co.uk)
10:11.32fjpCool. Got enlightenment running with Lenny.
10:28.12fjpBut we really need the graphics accelleration...
11:13.14lkclfjp: yep on the graphics acceleration.  that'll need the mer-smartq 2.6.32 kernel really, first - i don't fancy back-porting chronolytic's work to 2.6.24 yukk
11:18.31fjpHmm:
11:18.36fjpMy tree/the 2.6.32 one is
11:18.36fjpgit://gitorious.org/mcuelenaere-smartq-kernel/mcuelenaere-smartq-kernel.git
11:18.36fjpRegards,
11:18.37fjpMaurus Cuelenaere
11:19.13fjphttp://lists.infradead.org/pipermail/libertas-dev/2009-October/002797.html
11:19.50fjplkcl: Had a chance to look at poweroff again?
11:20.22lkclnot yet.  it really is that simple, btw.  pull that gpio low, machine powers off.
11:20.34lkcli did it by mistake.
11:20.43lkclso i know it works
11:21.25fjpIt's the most problematic issue we have ATM IMO.
11:22.27lkclwhere are the s3c_gpio_cfgpin calls in smdk6410_machine_init() ?
11:24.13fjpI don't understand what you're getting at.
11:25.16fjpWe added some in ct89e_hard_reset
11:25.31lkclthere are some gpio inits needed at the beginning of smdk6410_machine_init()
11:26.02lkclone of them is to pull sound "off"
11:26.06fjppc89e_machine_init I think you mean?
11:26.09lkclthe others i haven't deciphered yet
11:26.12lkclohh, ok
11:26.34fjpWe have:
11:26.37fjp+       /* CT-PC89E gpio init */
11:26.37fjp+       s3c_gpio_cfgpin(0x1AD, 1);
11:26.37fjp+       s3c_gpio_setpin(0x1AD, 0);
11:26.38fjp+
11:26.39fjp+       s3c_gpio_cfgpin(0x201, 1);
11:26.39fjp+       s3c_gpio_pullup(0x201, 0);
11:26.41fjp+       s3c_gpio_setpin(0x201, 1);
11:26.41fjp+
11:26.44fjp+       s3c_gpio_cfgpin(0x200, 1);
11:26.44fjp+       s3c_gpio_pullup(0x200, 0);
11:26.46fjp+       s3c_gpio_setpin(0x200, 1);
11:28.45lkclyep that's the ones.
11:28.47lkclwhew ok :)
11:29.34fjpwonders if hard_reset gets called during poweroff
11:30.29lkclthey _completely_ messed with the ordering, removed a ton of essential shutdown code, and replaced it with a hard poweroff
11:30.51lkclthey made a complete mess of the shutdown process
11:31.08lkclso, sync of filesystems won't be called.... etc. etc.
11:31.35lkcldarn it - i'm looking for my file which has all the list of gpio functions in it
11:32.35lkclok got it.
11:32.37lkclright
11:32.38lkclit's GPK12
11:32.43lkclnot GPH8
11:33.09lkclok actually there's two
11:33.21lkclone is GPH8 followed by FPF14
11:33.25lkclthe other is GPK12
11:33.44fjpThe first combi we have in hard_reset now.
11:34.04fjpWe don't have GPK12 yet AFAIK/IIRC
11:34.13lkcllet me make the file available...
11:34.25lkclhttp://lkcl.net/arm_systems/CT-PC89E/S3C_VA_GPIO
11:34.25fjpNah, I can't read it anyway.
11:35.05lkclno as in the file with the s3c_gpio_xxxxx functions i've reverse-engineered, not the vmlinux.lst
11:35.37lkclin it, you can see that there are no less than _three_ "shutdown" procedures.
11:35.59lkclone involves GPK12, in two different ways
11:36.08lkclthe other involves GPH8 and GPF14.
11:36.56lkclthe "safe" one is GPK12 cfgpinn=1, pullup=1, setpin=0
11:37.01fjpAnd they probably ripped out the watchdog stuff?
11:37.12lkclthe "unsafe" one is GPK12 pullup=0, cfgpin=1, setpin=0
11:37.32lkclfjp: the watchdog stuff is yes, a) entirely unused b) entirely unused
11:37.43lkcland the one marked "arm_machine_restart" is:
11:37.57lkclcfgpin GPH8 =1
11:38.03lkclpullup GPH8 = 0
11:38.10lkclsetpin GPH8=0
11:38.14fjpYes, I see it.
11:38.18lkclcfgpin GPF14=0
11:38.22lkclpullup GPF14=0
11:38.24lkclok
11:39.49lkclso i reckon if you add the GPK12 stuff from the "safe" shutdown, the machine will power off nicely.
11:39.54lkclit takes a second or two
11:40.25fjpAnd add that in the reset hook so that it gets done before the standard watchdog stuff.
11:41.14fjpI'll give that a try.
11:42.09fjpShould I drop the GPH8/GPF14 stuff we have there now?
11:43.33fjpthinks it would be nice to have an index of GPIO functions that have been identified.
11:44.41lkcldon't honestly know.
11:45.11lkclyes, very much so.  it's what we did on the HTC handhelds, create a wiki table, put their uses.
11:50.37fjpYay! Works.
11:52.34fjpBoth 'poweroff' and 'reboot' now poweroff, but that's fine by me.
11:53.38fjpI would say setting backlight on poweroff is not relevant as it powers off anyway. So let's assume the GPH8 stuff isn't needed either.
11:53.52fjpDo you know what GPH8 controls?
12:14.59lkclnaah, haven't got a clue
12:15.27lkcli really need an updated version of IDApro, one that recognises 0x7N opcodes
12:15.55lkclexcellent
12:25.17lkclwe'll find it eventually.
12:25.22lkclit may be the WIFI.
12:25.30lkcli haven't decoded WIFI poweron/off yet
12:27.49lkclyep it's probably wifi
12:29.47lkcls3c_gpio_cfgpin(S3C_GPH0, 1);
12:29.49lkclhmmm
12:29.59lkclohh . no that's 8
12:30.02lkclGPH8
12:34.04lkcland GPH6
12:36.17lkclfjp: yep, GPH8 is power-on/off, and i suspect GPH6 is reset
12:36.37lkclhttp://lkcl.net/arm_systems/CT-PC89E/S3C_VA_GPIO
12:52.01lkclbtw the FSF is now aware of chitech's GPL violation.  it happened by mistake.   i've warned chitech in advance that they can avoid a serious shit-storm by complying with the GPL.
12:52.26lkclalso i've contacted bruce perens and he liked the idea of doing a "Gnu Netbook"
15:23.25SQlvpapirfor a moment I thought the netbook was fried by the coins used for counterbalancing the heavy screen, then I realized the AC/DC inverter was sneaking out of the US plug adapter. *phew*
17:11.47mnemochi, to try to recognize it in a bag full of ac adapters, can you tell me the "specs" of the adapter of the chitech and if it has any "brand" on it?
20:43.55*** join/#arm-netbook bjdooks (~ben@trinity.fluff.org)
23:13.04*** join/#arm-netbook lkcl (~lkcl@nat66.mia.three.co.uk)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.