| 02:28.45 | *** join/#arm-netbook Baybal (~baybal@S0106001e58392615.vs.shawcable.net) |
| 10:46.43 | *** join/#arm-netbook lkcl (~lkcl@nat65.mia.three.co.uk) |
| 11:25.50 | *** join/#arm-netbook SQlvpapir (~teis@188.177.95.62) |
| 16:55.32 | *** join/#arm-netbook ibot (ibot@rikers.org) |
| 16:55.32 | *** topic/#arm-netbook is ARM netbook GNU/Linux hacking - logs: http://ibot.rikers.org/%23arm-netbook |
| 19:04.32 | lkcl | could someone look at http://piratepad.net/LPEw9nHASt and let me know if i got the sequence right? |
| 19:07.11 | *** join/#arm-netbook SQlvpapir (~teis@188.177.95.62) |
| 19:10.28 | fjp | is completely lost with that; sorry |
| 19:10.49 | fjp | BTW, did you notice that D-I was running with framebuffer? |
| 19:11.35 | fjp | I'm currently thinking how to shape the installation. |
| 19:19.09 | SQlvpapir | does the hardware support adjusting backlight brightness? |
| 19:45.55 | lkcl | fjp: no problem |
| 19:46.25 | lkcl | SQlvpapir: yep. it's on PWM. depending on the duty cycle, the LED backlight is bright or darker. quite straightforward. |
| 20:59.34 | lkcl | fjp: just created drivers/w1/w1_ds2431_ctpc89e.c |
| 20:59.42 | lkcl | in branch ct-pc89e/2.6.24.7/lkcl |
| 20:59.58 | fjp | cool |
| 21:02.06 | lkcl | i have nooo idea if it's safe |
| 21:02.34 | fjp | Did you put a HUGE warning in the Kconfig descr? |
| 21:05.33 | lkcl | naah. i just did the reverse-engineering, the key parts - read and write using GPIO Bank N, Pin 4 |
| 21:06.27 | fjp | I'll leave it into your branch until it's tested. |
| 21:06.56 | lkcl | *lol* |
| 21:07.01 | lkcl | thaaanks |
| 21:07.14 | lkcl | i get to blow up my machine's eeprom, it's okay, right? :) |
| 21:07.35 | fjp | Why? Anybody can use your branch if they want... |
| 21:07.47 | lkcl | they've put code into the kernel which goes directly to shutdown if the eeprom doesn't have certain values in it |
| 21:07.59 | lkcl | fucking fucking stupid |
| 21:08.14 | fjp | But exposing others to something that may not be safe and isn't tested is not very responsible IMO. |
| 21:08.25 | lkcl | so if the eeprom gets overwritten |
| 21:08.28 | lkcl | yes i know. |
| 21:08.41 | lkcl | it means that i have to risk using this *sigh* |
| 21:08.58 | fjp | Wait a bit for review. |
| 21:09.53 | lkcl | yehh |
| 21:09.57 | lkcl | noo! |
| 21:09.59 | lkcl | whyy? |
| 21:10.04 | lkcl | take a riiiisk, why nooot |
| 21:10.05 | lkcl | ha ha |
| 21:10.11 | fjp | Up to yu :-) |
| 21:10.27 | lkcl | neh, i know. i'll wait. |
| 21:12.40 | fjp | 's working on an extremely minimal ramdisk that will load the actual D-I image from /dev/mmcblk1p1 or /dev/mmcblk1 |
| 21:15.22 | lkcl | yaay, good stuff. i want to borrow that, for use on an odroid i've got sitting here. same tricks needed. |
| 21:17.48 | fjp | -rw-r--r-- 1 root root 66539 Mar 21 22:17 ramdisk.gz |
| 21:18.30 | fjp | We should be able to piggyback that onto any kernel :-) |
| 21:21.18 | lkcl | yaay |
| 21:21.18 | fjp | http://paste.debian.net/65257/ |
| 21:21.58 | fjp | All the magic is in /init (C proggie) |
| 21:22.04 | lkcl | smart. |
| 21:22.22 | fjp | Not really. Simply stolen from joeyh. |
| 21:24.45 | fjp | The real nice part is that as long as VFAT and mmcblk support are compiled in, it's completely independent of the kernel. |
| 21:25.27 | fjp | So I won't have to: build kernel - build D-I boot image for kernel - build kernel bootpImage including D-I image |
| 21:25.39 | fjp | That would be a total pain. |
| 21:26.14 | mnemoc | prefer early inits to be shell scripts |
| 21:26.47 | fjp | Sure. But requires more support. |
| 21:26.56 | fjp | Now I can do without busybox even. |
| 21:27.25 | mnemoc | yes, i rely in quite extense bb support |
| 21:27.34 | fjp | When the real D-I image has been loaded everything is shell again. (Well, most things). |
| 21:28.35 | fjp | Wow. It seems to work. |
| 21:28.48 | mnemoc | \o/ |
| 21:29.04 | fjp | At least, I get exactly the expected errors when I don't have the actual D-I image on the flash card. |
| 21:31.13 | lkcl | very cool |
| 23:01.23 | fjp | Yay! Brilliant! |
| 23:03.40 | fjp | root initrd is getting loaded and D-I starts |
| 23:07.20 | fjp | Hmm. Major kernel errors when running DHCP... |
| 23:07.34 | fjp | And system frozen. |
| 23:08.00 | fjp | Out of memory? |
| 23:11.03 | fjp | Doubtful. Let's try static IP. |
| 23:12.40 | fjp | BUG: spinlock recursion on CPU#0 |
| 23:13.04 | fjp | Happens too when I configure static IP from netcfg. |
| 23:25.35 | fjp | Same if I configure the network completely manually and then try a wget :-( |
| 23:31.26 | fjp | Question is, why do things work correctly when I boot the same kernel with the ChiTech initrd? |