| 00:01.24 | *** join/#arm-netbook vgrade2 (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 00:03.57 | sepp | lundman: thanks a lot |
| 00:06.24 | lundman | np |
| 00:06.44 | lundman | I have the mele right in front of me, on the desk with two monitors :) |
| 00:11.41 | hno | PhantomsDad, https://github.com/hno/uboot-allwinner/blob/sun4i/include/configs/sun4i.h#L154 |
| 00:13.28 | lundman | one of the issues with ICS4 is you can't change resolution |
| 00:13.45 | lundman | but I did get googleplay on there, and bubbleupnp to test video playback |
| 00:15.04 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 00:15.05 | PhantomsDad | hno, Yea I saw that. But I'm confused. Unless my math is wrong, 544KB with 512 byte sectors is sector 2116, which puts it inside the 1st partition (assuming 1st part starts at sector 2048)? |
| 00:16.59 | PhantomsDad | Doh! my math is wrong. |
| 00:17.05 | PhantomsDad | lol |
| 00:17.05 | hno | (544+128) * 2 = 1344. Still a good marin to 2048 |
| 00:17.36 | hno | s/rin/rgin/ |
| 00:18.23 | PhantomsDad | hno, OK so the build instructions should advise users to begin by erasing sectors 1-2047 I think. |
| 00:19.08 | steev | hmm, Turl, there's an issue when compiling the kernel with the android toolchain (namely that it doesn't report itself as __linux__, iirc it seems to treat most ifdefs as if it were BSD) |
| 00:20.54 | steev | need something akin to https://github.com/ahmedammar/linux-imx6q-ipukms/commit/b6ea125d7ac37a6774cd3b8d764d0a8e8b061ac7 |
| 00:22.50 | hno | PhantomsDad, only environment may need to be erased, and only if there is a valid u-boot environment already saved there. |
| 00:23.10 | PhantomsDad | hno, the boot.scr is plain text file? |
| 00:23.29 | hno | yes, or uboot image if you like. It should understand both. |
| 00:23.49 | steev | hno: has anyone considered moboot for the allwinners? |
| 00:24.01 | naobsd | echo __linux__ | arm-linux-androideabi-4.4.x/bin/arm-linux-androideabi-gcc -E - |
| 00:24.01 | hno | steev, not that I know of. |
| 00:24.08 | naobsd | is "1" |
| 00:24.10 | PhantomsDad | hno, thanks |
| 00:24.25 | hno | PhantomsDad, added the storage map to wiki page. |
| 00:24.26 | naobsd | I guess you're using arm-eabi-x.x.x |
| 00:24.51 | Turl | steev: you're using a 'bare metal' toolchain |
| 00:24.59 | Turl | that's why I ended up disabling DRM :) |
| 00:25.22 | Turl | you need to use a linux toolchain, that'll set __linux__ |
| 00:25.24 | steev | i've never heard the android toolchain called "bare metal" before |
| 00:25.34 | Turl | steev: well, it's arm-eabi- |
| 00:25.37 | Turl | not arm-linux-eabi |
| 00:25.51 | Turl | you're not using arm-linux-androideabi either :p |
| 00:25.52 | naobsd | arm-linux-androideabi-4.4.x is default toolchain on (at least) 4.0.4_r1.2 |
| 00:26.03 | Turl | yeah but kernels build with arm-eabi- |
| 00:26.27 | steev | so how do i tell the cm build system to use arm-linux-androideabi-4.4.x |
| 00:26.36 | Turl | steev: you don't |
| 00:26.39 | Turl | steev: just disable DRM |
| 00:26.43 | Turl | you don't need it on android anyway |
| 00:26.44 | Turl | :P |
| 00:26.52 | steev | it's your defconfig! :P |
| 00:26.58 | naobsd | CROSS_COMPILE= |
| 00:26.59 | Turl | steev: I disabled it |
| 00:27.05 | steev | hrm |
| 00:27.23 | steev | corsair approved my rma, so that's nice, considering the ram was just over a year old |
| 00:27.32 | Turl | steev: https://github.com/amery/linux-allwinner/commit/88aee3c84f03b4aeec43b9860e6d68b494826682 |
| 00:27.36 | steev | probably be faster to just buy more |
| 00:28.05 | steev | ah crap |
| 00:28.10 | steev | pebcak |
| 00:28.19 | steev | i'd switched the kernel repo to a different branch |
| 00:28.37 | naobsd | I'm looking wiki... |
| 00:28.39 | naobsd | hmm |
| 00:28.58 | naobsd | someone need CWM for A10 LCD less box |
| 00:29.14 | naobsd | please ask me |
| 00:29.24 | steev | naobsd: ask you? |
| 00:29.42 | naobsd | sorry |
| 00:29.42 | Turl | LCD less? o.O |
| 00:29.49 | Turl | how do you use it? :P |
| 00:29.49 | steev | Turl: maybe hdmi only? |
| 00:29.53 | naobsd | hmm |
| 00:29.56 | naobsd | sorry |
| 00:30.16 | naobsd | if someone need CWM for A10 w/o touchscreen |
| 00:30.30 | steev | isn't that any of them? |
| 00:30.37 | steev | doesn't have cwm touch on his ly-f1 |
| 00:30.39 | naobsd | I can make CWM with keyboard/mouse control |
| 00:30.43 | Turl | I have CWM with and without touchscreen for my tablet :) |
| 00:30.44 | naobsd | grr |
| 00:30.53 | steev | i want cwm touch |
| 00:30.55 | naobsd | if someone need CWM for A10 w/o physical buttons |
| 00:30.59 | Turl | (physical buttons and touch) |
| 00:31.12 | Turl | steev: the CWM touch code is closed though :( |
| 00:31.13 | naobsd | my brain is dead |
| 00:31.14 | steev | my vol - key is.... kinda sunken in |
| 00:31.20 | steev | Turl: i figured as much |
| 00:31.45 | Turl | I could build one for you one of these days I guess |
| 00:31.59 | steev | no rush, i'm actually back on stock at the moment |
| 00:32.23 | Turl | I didn't like the idea of using stock much :P |
| 00:32.44 | steev | well my rom still isn't quite working right :P |
| 00:32.44 | Turl | I heard many privacy horror stories regarding chinese shipping software |
| 00:32.52 | *** join/#arm-netbook sepp_ (~seb@port-92-206-92-64.dynamic.qsc.de) |
| 00:33.04 | steev | as soon as i tilt it, the gsensor kicks in and decides up is down |
| 00:33.12 | steev | i've made the change but haven't tested |
| 00:33.15 | Turl | steev: that's an easy fix |
| 00:33.31 | Turl | you need to enable one or more board defines => bam fixed :) |
| 00:33.50 | steev | yeah, still not entirely sure which sensor is actually used on my board |
| 00:33.59 | Turl | steev: https://github.com/allwinner-dev-team/android_device_allwinner_zatab/blob/ics/BoardConfig.mk#L25 |
| 00:34.22 | Turl | steev: run getevent on adb shell |
| 00:34.30 | steev | oh, hm |
| 00:34.34 | Turl | it'll be pretty obvious reading the first output lines |
| 00:35.32 | steev | cool |
| 00:35.41 | lundman | cool, getevent |
| 00:35.59 | Turl | you can getevent /dev/.... too to just get events from a single device |
| 00:36.13 | steev | mma7660, hv_keypad, ft5x_ts, sun4i-keyboard, and axp20-supplyer |
| 00:36.27 | steev | i know it's the mma7660, but why 3 different keyboards? |
| 00:37.00 | Turl | sun4i one is the keys on your device |
| 00:37.01 | steev | and with that listing, do i nekd ts driver? |
| 00:37.04 | steev | hacked* |
| 00:37.06 | Turl | axp20 handles the power button |
| 00:37.22 | Turl | no idea what's hv_keypad |
| 00:37.38 | Turl | I've seen a driver for ft5x, I think it's the same quarx has |
| 00:37.44 | steev | ah, cool |
| 00:37.50 | Turl | his works, so I suppose it would |
| 00:38.02 | Turl | https://github.com/allwinner-dev-team/android_device_allwinner_common/blob/master/libsensors/Android.mk |
| 00:38.08 | steev | i'll have to use this when my new tablet shows up |
| 00:38.14 | Turl | the revert/invert flags are the same |
| 00:39.24 | lundman | i could try ubuntu on the mele, but the image is very old |
| 00:40.46 | naobsd | hv_keypad is for touch key on TS |
| 00:40.56 | naobsd | on some tablet. it's minor |
| 00:42.56 | steev | ugh, my laptop can't handle being only 4G :/ |
| 00:43.20 | hno | A10 have many keyboard interfaces. |
| 00:43.51 | hno | LRADC, Key matrix, PS2, USB, twi to various controllers, usb, ... |
| 00:44.03 | hno | and GPIO as well. |
| 00:46.28 | *** join/#arm-netbook vgrade2 (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 00:57.43 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 01:03.35 | CIA-126 | rhombus-tech: Naoki master * r6d8b36517421 /allwinner_a10/hacking_the_mele_a1000/android.mdwn: I don't use current kernel and Quarx2k's work yet |
| 01:06.17 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 01:10.02 | naobsd | mmm "I'm not using..." my english is bad :( |
| 01:11.16 | lundman | all english is bad :) |
| 01:11.22 | lundman | terrible language! |
| 01:12.23 | naobsd | :) |
| 01:12.36 | naobsd | what is good language? binary? ;) |
| 01:13.16 | *** join/#arm-netbook killerkink (b637f91b@gateway/web/freenode/ip.182.55.249.27) |
| 01:13.38 | lundman | swedish? japanese? |
| 01:14.19 | lundman | wonder if I can steal the video-settings program from mele android 2.3 and use with ICS4 |
| 01:17.23 | *** join/#arm-netbook vgrade2 (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 01:20.59 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 01:32.31 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 01:33.55 | *** join/#arm-netbook kaspter (~kaspter@116.231.122.26) |
| 01:44.50 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 01:48.43 | steev | Turl: oh bah i just noticed i'm using the allwinner-dev-team kernel not amery's |
| 01:48.54 | steev | stupid 4G of ram causing my system to be super slow |
| 01:49.29 | steev | what branch should i be using? |
| 01:49.33 | steev | the commit doesn't say |
| 01:50.24 | Turl | my tree also has the commit steev |
| 01:50.46 | Turl | https://github.com/allwinner-dev-team/linux-allwinner/commits/allwinner-3.0 |
| 01:51.05 | Turl | if you want to use amery's use the 3.0 -v2 branch |
| 01:52.02 | Turl | good night all, I'm off to learning algebra :P |
| 01:52.07 | steev | hrm i'm using your tree, allwinner-v3.0 |
| 01:52.38 | Turl | fetch and checkout again? :) |
| 01:54.19 | steev | yeah gonna try that |
| 01:54.26 | steev | i see the log entry |
| 01:55.22 | steev | none of them seem to have it set, so wth, ill look into it, go enjoy algebra |
| 01:57.45 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 02:09.37 | *** join/#arm-netbook vgrade2 (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 02:14.37 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 02:15.12 | *** join/#arm-netbook dfletcher (~fletch@108-196-222-251.lightspeed.sntcca.sbcglobal.net) |
| 02:15.40 | *** join/#arm-netbook dfletcher (~fletch@drupal.org/user/72475/view) |
| 02:16.18 | CIA-126 | rhombus-tech: Gary master * r6fdfafba1875 /allwinner_a10/hacking_the_mele_a1000/Building_Debian_From_Source_Code_for_Mele.mdwn: |
| 02:31.39 | *** join/#arm-netbook vgrade2 (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 02:48.41 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 02:52.35 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 03:00.23 | *** join/#arm-netbook vgrade2 (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 03:05.31 | *** join/#arm-netbook lerc (~quassel@121-74-238-5.telstraclear.net) |
| 03:05.56 | *** join/#arm-netbook rm___ (rm@fsf/member/rm) |
| 03:06.43 | *** join/#arm-netbook Brandon15811 (~Brandon15@ec2linux.exaktop.tk) |
| 03:11.40 | lundman | Hmm, not entirely sure what fs they used in the mele-1.6.img file. the boot and uImage is easy to get out, but then rest.. |
| 03:13.05 | *** join/#arm-netbook tallnerd1985 (~tallnerd1@33.sub-166-250-0.myvzw.com) |
| 03:21.13 | steev | Turl: hm, okay, something really weird is going on, i've clobbered my build, as well as removed the out directory completely. it still attempts to build mali, and even though it's not set in any defconfigs, if i go into out/target/product/a710/obj/KERNEL_OBJ/ and then grep DRM .config (nothing shows), however grep MALI .config shows CONFIG_GPU_MALI=y |
| 03:29.37 | *** join/#arm-netbook tallnerd1985_ (~tallnerd1@53.sub-166-250-3.myvzw.com) |
| 03:50.22 | steev | Turl: i think i figured out what the issue is... just gotta test my theory |
| 03:51.40 | naobsd | http://forum.xda-developers.com/showpost.php?p=27039990&postcount=669 for someone who want to use a key to enter recovery while power on |
| 03:54.31 | naobsd | on A10 tablets |
| 03:55.13 | naobsd | (not for A10 boxes... it uses sun4i-keyboard interface) |
| 03:56.40 | naobsd | (for A10 boxes, please add a switch and modify u-boot to use it;) |
| 03:58.05 | steev | the rtl8192cu folder is looking really suspect |
| 04:07.09 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
| 04:52.40 | *** join/#arm-netbook naGt (~nacho@107.177.218.87.dynamic.jazztel.es) |
| 05:04.00 | *** join/#arm-netbook kaspter (~kaspter@108.171.248.180) |
| 05:04.41 | ZaEarl | naobsd, so i can edit sys_config1.fex and pick which key starts up in recovery? |
| 05:05.34 | lundman | guess I just have to open the mele to be able to be of any use |
| 05:06.56 | ZaEarl | like I say, if you can't open it, you don't own it. |
| 05:07.40 | lundman | I was just putting it off, thinking there would be more binaries :) |
| 05:07.54 | lundman | but if I dont have uart, I cant play with u-boot etc |
| 05:08.14 | lundman | i added ext4 and zfs to u-boot, so I want to test that no |
| 05:08.22 | steev | Turl: in case you get bored, http://pastebin.com/JNcZtPLv i can't figure out where the issue comes from. not sure what is causing all those gcc: not found, or why it keeps trying to build the mali stuff. the kernel config is the sun4i_crane_defconfig, kernel is amery/linux-allwinner, branch is allwinner-v3.0-android-v2 |
| 05:08.42 | steev | (but ftr, it happens with allwinner-dev-team allwinner-3.0 as well) |
| 05:09.42 | steev | gonna try with V=1 |
| 05:09.56 | naobsd | ZaEarl: I just explained general thing. I don't know detail about your device. I can't answer. you can try. |
| 05:10.00 | lundman | looks like mismatched " somewhere near $directory |
| 05:10.46 | steev | lundman: yeah well, you ever tried grepping for " in the kernel sources? |
| 05:11.13 | lundman | I would suspect the Makefile |
| 05:11.24 | ZaEarl | naobsd, I saw the explanation for what to do, but not where to make the edits. |
| 05:11.32 | naobsd | steev: remove space in CROSS_COMPILE="(here) |
| 05:11.59 | steev | naobsd: (here)? |
| 05:11.59 | lundman | or that one |
| 05:13.03 | steev | i'm not sure where that cross_compile is pulled from exactly |
| 05:13.15 | naobsd | steev: I can see space just after =" but it may not be a problem... I'm not sure |
| 05:14.43 | lundman | run "env" and check its clean I guess |
| 05:14.49 | steev | it might be a problem due to the quotes, which i'm assuming are there in case of spaces in the path name, but i doubt it. i've been looking |
| 05:18.29 | naobsd | ZaEarl: please edit (source of) script.bin in nanda. |
| 05:19.50 | naobsd | it was mentioned in old post, sorry |
| 05:21.23 | steev | i'm gonna test something |
| 05:23.21 | naobsd | http://forum.xda-developers.com/showpost.php?p=27039990&postcount=669 step 7 is added |
| 05:24.07 | steev | nope. i'm using the kernel config, and even wifi setup from allwinner/common, just like zatab is |
| 05:26.11 | steev | heck the only difference between mine and the zatab one is that their gsensor is bma250, mine is mma7660 |
| 05:31.47 | *** join/#arm-netbook ZaEarl (~malmrose@173-11-125-162-SFBA.hfc.comcastbusiness.net) |
| 05:32.52 | lundman | heh it was nothing to open, i should not have put it off for so long |
| 05:36.02 | steev | okay, giving up for the night, even just trying to cross compile the kernel itself is bombing. night all |
| 06:17.02 | *** join/#arm-netbook ZaEarl (~malmrose@173-11-125-162-SFBA.hfc.comcastbusiness.net) |
| 06:35.05 | *** join/#arm-netbook tallnerd1985 (~tallnerd1@c-24-21-67-138.hsd1.wa.comcast.net) |
| 06:35.41 | hno | lundman, yes the mele is very hacking friendly. |
| 06:36.04 | tallnerd1985 | Any experience with mk802 yet? |
| 06:36.24 | hno | lundman, did a little HW modification to mine: http://www.henriknordstrom.net/code/A10/uart-module.jpg |
| 06:37.26 | lundman | hmm why did you do that? |
| 06:37.59 | hno | because I alays use the console, and my power cable was a mess. |
| 06:38.35 | lundman | although I am tempted to try to find seomwhere to permanently attach the console, drilla hole in the plastic |
| 06:39.34 | lundman | or I'll just run the cable out the side, meh :) |
| 06:40.19 | lundman | but with uart I believe I can play with u-boot more, and tackle the sata part as well, no? |
| 06:43.01 | *** join/#arm-netbook rell01 (~chatzilla@p5B07801C.dip0.t-ipconnect.de) |
| 06:45.15 | hno | lundman, u-boot yes, sata no. |
| 06:45.30 | hno | unless you are trying to add sata to uboot |
| 06:46.05 | lundman | yes |
| 06:46.25 | hno | then serial console is very much needed |
| 06:46.29 | lundman | work is slow, other kernel work is stopped. might as well help here |
| 06:46.44 | hno | nice |
| 06:46.53 | hno | very welcome |
| 06:47.00 | lundman | serial console is the uart connector yes? I noticed the image also mention the usb connector, but I assume that is host |
| 06:48.12 | lundman | wait, today is flag day? damnit |
| 06:48.18 | *** join/#arm-netbook rell01_ (~chatzilla@p5B07807A.dip0.t-ipconnect.de) |
| 06:53.02 | hno | lundman, correct. The 4 pin connector you see connected in my picture is the UART / serial console. The 5-pin connector just next to the LAN port is the USB UTG port for USB device mode (connecting to PC) |
| 06:54.54 | lundman | good to confirm. I bought the usb thing from Cubie, but left it at home, so I'll give it a whirl tomorrow |
| 06:55.26 | hno | usb uart? |
| 06:55.41 | lundman | ye |
| 06:56.11 | hno | good. The little extra board you see on my picture is also an USB UART if that was not obvious. |
| 06:56.13 | lundman | http://www.aliexpress.com/snapshot/106555554.html |
| 06:56.41 | lundman | yeah I gathered as such. Not the same spec since you have to get power from somewhere else |
| 06:57.10 | hno | lundman, I don't. It's powering the mele. |
| 06:58.13 | lundman | oh power goes in that direction |
| 07:00.42 | hno | yes. I just connect a single USB cable to the board and I have serial console + power. |
| 07:01.04 | lundman | not entirely insane indeed |
| 07:01.05 | hno | but nowdays I usually have two USB cables, one to the OTG port as well. |
| 07:04.15 | lundman | guess I should sneak home early, so I can tape the flag up in the window |
| 07:05.53 | hno | lundman, where are you located? |
| 07:06.35 | lundman | tokyo |
| 07:14.04 | CIA-126 | rhombus-tech: Booshi master * r9300c6e08035 /allwinner_a10/orders/booshi444.mdwn: |
| 07:14.07 | *** join/#arm-netbook hipboi (~hipboi@221.4.213.95) |
| 07:15.17 | Turl | steev: sure your .config doesn't have them in? |
| 07:26.06 | mnemoc | Turl: moin, can you test: https://github.com/amery/linux-allwinner/commit/88aee3c84f03b4aeec43b9860e6d68b494826682#commitcomment-1421350 ? |
| 07:26.46 | Turl | mnemoc: I'd very much rather not touch include/drm/drm.h |
| 07:26.57 | Turl | mnemoc: it's non-allwinerized upstream code :) |
| 07:28.26 | mnemoc | Turl: this is about mali support, not allwinner's |
| 07:30.34 | mnemoc | Turl: but been optional (you said it's only for X11 support) probably adding a "depends on" and "default m" in the drm/mali's Kconfig is better idea |
| 07:31.16 | Turl | mnemoc: it'd break the same if you build radeon DRM on x86 with a toolchain that doesn't -D__linux__ |
| 07:32.18 | mnemoc | and why do you want to build it off-tree? |
| 07:32.52 | mnemoc | s/off-tree/manually/ |
| 07:33.00 | Turl | hm? |
| 07:33.29 | Turl | my point was, if I'm building for android and with a bare metal toolchain, you don't need the DRM stuff at all |
| 07:33.59 | Turl | if you're building your kernel for your X11-able mele you'll enable it and use a -linux- toolchain most likely |
| 07:35.03 | hno | mnemoc, why do you want to import off-tree modules which is not scheduled for mainlining and which have delicate version dependencies on other components into the kernel tree? |
| 07:36.01 | Turl | if you want to use a bare metal one *and* build DRM just do hax and add -D__linux__ to your cflags :) |
| 07:38.11 | mnemoc | hno: only mali... |
| 07:39.41 | hno | yes, but why? |
| 07:40.23 | mnemoc | hno: to stop people from assuming the ./build.sh are the right way of building the kernel |
| 07:40.50 | mnemoc | mali is the only 3rd-party whihc is actually interesting to maintain |
| 07:41.00 | hno | and instead end up with a mess from people needing different mali versions? |
| 07:41.28 | mnemoc | I was told that version is good for everyone |
| 07:43.49 | mnemoc | everything was done under that statement, with people using it without problems in android ics and X11 |
| 07:45.35 | mnemoc | if that assertion was false I have no problem in reverting mali support to it's previous state |
| 07:47.14 | mnemoc | the ticket is still open :) |
| 07:49.37 | mnemoc | should have kept this in a branch longer :-/ |
| 07:52.06 | Turl | has anyone noticed that, other than for 'fantasy', all of the other governors perform badly? |
| 07:58.05 | killerkink | turl: performance and interactive work |
| 07:58.34 | killerkink | turl: others seem like junk. ondemand need to tune the upper/lower limit |
| 07:59.18 | Turl | of course well, performance works :P |
| 07:59.27 | Turl | killerkink: 'fantasy' is good |
| 07:59.28 | mnemoc | 60MHz ftw! |
| 07:59.31 | Turl | ondemand was crap :P |
| 08:00.02 | killerkink | turl: the battery on fatasy in ics seem to be higher. not sure if its the cpu though |
| 08:00.07 | Turl | mnemoc: my phone can do 19Mhz, problem? ;) |
| 08:00.18 | mnemoc | :) |
| 08:00.30 | Turl | killerkink: yeah fantasy ramps up to max apparently |
| 08:00.40 | Turl | (talking from seeing it working, haven't seen the code) |
| 08:00.57 | Turl | maximum interactivity :D |
| 08:01.42 | killerkink | still ics seem to comsume more batt with the same cpu setting |
| 08:02.08 | Turl | probably the mali drains more power compared to the A8 |
| 08:02.19 | Turl | ics uses the gpu more heavily |
| 08:02.42 | *** join/#arm-netbook ka6sox (ka6sox@nasadmin/ka6sox) |
| 08:03.06 | Mazon | does anyone know if the mele units are able to run 24/7 debian (or other non-android)? |
| 08:03.30 | killerkink | some users claim that stock ics run on lower batt |
| 08:03.36 | specing | All machines are able to run 24/7 |
| 08:03.40 | Turl | Mazon: I've heard so |
| 08:04.04 | mnemoc | why shouldn't? |
| 08:04.08 | Mazon | I just saw a page which stated some instability issues |
| 08:04.09 | Turl | killerkink: if I were to believe everything users claim I'd be crazy or dead by now :) |
| 08:04.31 | killerkink | turl: just saying. :) |
| 08:04.45 | killerkink | turl: i personally dont believe it |
| 08:04.49 | Turl | but my battery life is OK |
| 08:05.12 | Turl | I mean, I pretty much usage-raped the table two days and it still had some battery left :P |
| 08:05.13 | killerkink | turl: its ok. but it does seem ics take a bigger hit on the batt |
| 08:05.22 | Turl | tablet* |
| 08:05.51 | killerkink | turl: well, for you with bigger batt capacity, why would you care?:) |
| 08:06.36 | Turl | haha, 8Ah gets you far along :) |
| 08:07.51 | Turl | chinese batteries aren't known for their awesome longevity though, let's see how much this one lasts |
| 08:08.49 | mnemoc | they are usually refurbished |
| 08:09.07 | Turl | refurb cells? o.O |
| 08:09.45 | mnemoc | recycling is supposed to be a good thing :) |
| 08:10.36 | Turl | yeah but.. I didn't know you could refurbish batt cells :P |
| 08:11.53 | Turl | well cya later, idk why I'm not studying still |
| 08:12.16 | Turl | I didn't wake up @4AM to drink pepsi and check on IRC lol |
| 08:12.41 | mnemoc | I mean they pick "still-good" ones from other batteries to make "new" ones |
| 08:14.21 | *** join/#arm-netbook eebrah (~Ibrahim_K@41-139-199-130.safaricombusiness.co.ke) |
| 08:17.17 | mnemoc | hno: so you confirm we can't bless r2p4 and trash the rest and need to return to the multiversion model? |
| 08:20.41 | killerkink | turl: damn, mine is not even half of yours. :) |
| 08:22.55 | mnemoc | killerkink: get a zatab ;-) |
| 08:25.06 | killerkink | is zatab ips? |
| 08:25.24 | killerkink | mnemoc: then everyone will develop on the same tab. no fun at all |
| 08:25.48 | mnemoc | good point |
| 08:39.25 | Turl | yes IPS |
| 08:40.37 | killerkink | turl: does it have "brightness" or "darkness" issue? |
| 08:40.46 | Turl | hm? |
| 08:41.08 | killerkink | does it go dark when set to the lower brightness level? |
| 08:41.13 | killerkink | lowest |
| 08:41.21 | Turl | like, dark fully dark backlight off? |
| 08:41.57 | mnemoc | that bug is annoying, yes |
| 08:42.47 | Turl | mine doesn't |
| 08:42.54 | Turl | I mean it goes pretty dark |
| 08:42.57 | Turl | but it's still legible |
| 08:43.01 | Turl | not like backlight turned off |
| 08:43.08 | Turl | https://github.com/allwinner-dev-team/android_device_allwinner_common/blob/master/liblights/lights.c#L77 |
| 08:43.15 | Turl | I saw that the other day |
| 08:43.24 | Turl | comment is by allwinner |
| 08:43.38 | mnemoc | what's the lower brightness % you can actually see on your device? |
| 08:43.40 | Turl | you can prolly bump the value and forget about the bug :) |
| 08:43.56 | Turl | 0? idk, it's a slider :P |
| 08:44.38 | killerkink | yeah. that code in the brightness prevent you from going too lower |
| 08:44.43 | mnemoc | the a10 tablets I have here (well... my daughters') use their factory image... and go fully dark at like 80% |
| 08:44.48 | killerkink | some of the a10 tablet suffer from this |
| 08:45.04 | Turl | 80%? o.O |
| 08:45.09 | Turl | that's.. unexpected |
| 08:45.32 | Turl | that's a full bug :P |
| 08:45.37 | killerkink | mnemoc: which brand/model. I know some of teclast a10 has this issue |
| 08:45.44 | Turl | I can understand if it goes off at say 1%, and can be software hotfixed |
| 08:45.53 | Turl | but 80% = dead is unacceptable |
| 08:46.18 | killerkink | in fact, a10t only have like 4/5 discrete brightness settings due to this |
| 08:46.20 | mnemoc | fy-l1/2.3 and onda vi10 elite/4.0 |
| 08:47.02 | killerkink | mnemoc: interesting, i tot only teclast has it. the chinese user were bashing teclast over it at their forum at some point in time |
| 08:47.37 | mnemoc | it might be fixed in newer firmwares, but I can't update them |
| 08:47.54 | Turl | <PROTECTED> |
| 08:47.54 | Turl | <PROTECTED> |
| 08:48.02 | Turl | that's inherited from the platform on my tablets |
| 08:48.09 | Turl | so yeah, I can go pretty low :) |
| 08:48.46 | Turl | that's from a 0-255 value iirc |
| 08:49.09 | killerkink | mnemoc: for teclast a10t, i heard its an hardware issue. |
| 08:49.28 | mnemoc | ouch |
| 08:49.28 | Turl | mnemoc: why can't you upgrade them? |
| 08:49.52 | mnemoc | 1) no windows, 2) can't trash the state of their games |
| 08:50.08 | Turl | 1) windows? o.O |
| 08:50.11 | Turl | mnemoc: you can recompile a fixed lighs hal and replace it on the devices too |
| 08:50.58 | mnemoc | Turl: I was talkinag about reflashing the devices with the latest firmwares |
| 08:51.17 | Turl | yeah, I still don't get why you need windows |
| 08:51.18 | Turl | :P |
| 08:51.30 | Turl | to upgrade you pop a zip on your sdcard/nand and flash it |
| 08:51.35 | mnemoc | Turl: livesuit doesn't work on qemu |
| 08:51.56 | Turl | ah, livestuff |
| 08:52.02 | Turl | mnemoc: it works on virtualbox btw |
| 08:52.08 | Turl | I unbricked my tab twice with it |
| 08:52.12 | mnemoc | oracle software, no way |
| 08:52.29 | Turl | mnemoc: haha |
| 08:52.29 | mnemoc | I might try phoenixcard |
| 08:52.38 | Turl | phoenixwut? |
| 08:52.54 | Turl | mnemoc: btw, I prefer the oracle sw to using vmware :P |
| 08:53.04 | mnemoc | but still have the risk of trashing the state of the games they play there |
| 08:53.20 | mnemoc | Turl: I won't use any of those |
| 08:53.41 | Turl | qemu was painfully slow and hard to set up last I checked |
| 08:54.15 | mnemoc | qemu/kvm is quite decent.... but I would only need it to run phoenixcard |
| 08:54.50 | mnemoc | will wait until the summer when they have finished all the games and don't care about starting from zero again |
| 08:55.13 | Turl | you know you can backup the stuff right? :) |
| 08:55.37 | mnemoc | trusted titanium once when updating my phone |
| 08:55.40 | mnemoc | total failure |
| 08:55.45 | Turl | titanium is crap :P |
| 08:55.48 | Turl | I meant a real backup |
| 08:56.16 | Turl | think more like 'a / and /home tarball' |
| 08:56.54 | lundman | :) |
| 08:57.09 | mnemoc | that requires understanding how android stores apps and their data... and not really interesting at the moment |
| 08:57.23 | lundman | but oracle didnt write virtualbox, just bought it :) |
| 08:57.26 | Turl | mnemoc: nah, it doesn't |
| 08:58.17 | Turl | mnemoc: there's this thing called "nandroid backup" which is a tarball of each filesystem ;) |
| 08:58.41 | Turl | (or a dd-like image, if it's not a filesystem) |
| 08:59.00 | mnemoc | and then what's the point in updating if the restore will trash it? :p |
| 08:59.25 | Turl | mnemoc: you can just restore the equivalent of /home |
| 08:59.33 | Turl | aka /data :P |
| 08:59.44 | Turl | but that's not the point |
| 09:00.00 | mnemoc | is that compatible across OS versions? |
| 09:00.06 | Turl | the point is, if you f*ck your tablet putting a fixed lights hal, you can revert it :P |
| 09:00.25 | Turl | mnemoc: restoring /data? it can be |
| 09:00.34 | Turl | you need to keep 2 things in mind |
| 09:00.43 | Turl | system keys used, and android version |
| 09:00.44 | mnemoc | I don't use tablets, and my daughters really don't care much |
| 09:01.15 | mnemoc | Turl: what do you mean by system keys? |
| 09:01.21 | lundman | mind you, we have 2000 solaris servers, all being reinstalled with openindiana |
| 09:01.24 | lundman | no oracle here |
| 09:01.33 | Turl | mnemoc: apks are signed with keys |
| 09:01.35 | mnemoc | :D |
| 09:02.06 | mnemoc | Turl: so I need to backup the keys "repo" too? |
| 09:02.11 | Turl | mnemoc: the apk keys are checked against a database to grant permissions |
| 09:02.11 | Turl | no |
| 09:02.42 | Turl | if your backup says apk A has system key, "abcd" |
| 09:02.57 | Turl | yet your android system was signed with "zzzz" if will fail |
| 09:03.19 | mnemoc | uh |
| 09:03.39 | zenitraM | but that is only for system apks |
| 09:03.44 | Turl | (luckily? enough) chinese manufacturers don't even bother setting up a pair of keys and sign it with the dev aosp keys :P |
| 09:04.00 | Turl | zenitraM: user apks are checked too |
| 09:04.15 | zenitraM | but they aren't / shouldn't be signed with system keys |
| 09:04.17 | Turl | zenitraM: that's why you need to uninstall an apk and then install it again if you sign it with another key |
| 09:04.27 | Turl | zenitraM: indeed, they have another keys |
| 09:04.39 | zenitraM | that is because they check if the key of the newer version matches the older one |
| 09:04.39 | Turl | s/another/different/ |
| 09:04.53 | Turl | zenitraM: yeah |
| 09:05.07 | Turl | zenitraM: if you, say, tamper the apk on recovery, key check will fail too |
| 09:05.17 | zenitraM | anyway, backing an app is just backing all the data at /data/data/app/<package_name> and the apk at /data/app iirc |
| 09:05.51 | zenitraM | CM7 had a funny security issue where an app signed with system keys could have system permissions (superuser) without requiring them explicitly |
| 09:05.57 | zenitraM | and of course the system keys were public |
| 09:06.35 | mnemoc | hno: the a13 SDK doesn't come with the real .git dirs :( |
| 09:06.38 | Turl | zenitraM: you always need to request superuser permissions |
| 09:06.57 | mnemoc | hno: they are broken symlinks to a missing .repo |
| 09:06.59 | Turl | whatever key you use |
| 09:07.11 | zenitraM | it wasn't superuser.. but system apps are granted all priviledges and ignore the ones on the manifest |
| 09:07.16 | Turl | zenitraM: when you run "su <whatever>", that launches superuser |
| 09:07.42 | Turl | zenitraM: yeah, that issue existed and was fixed |
| 09:07.48 | Turl | zenitraM: it wasn't CM7 only though |
| 09:07.58 | zenitraM | it was all of android |
| 09:07.58 | mnemoc | /data/data/ sounds funky |
| 09:08.22 | Turl | zenitraM: the issue as I recall it was, that given a system-signed app "whatever.apk" |
| 09:08.32 | Turl | an user could create another apk with the same package name |
| 09:08.33 | zenitraM | also all wifi keys are stored at plaintext at /data/data/wifi/wpa_supplicant.conf |
| 09:08.44 | Turl | sign it with publicly available system keys |
| 09:08.58 | Turl | and install it normally, upgrading the system one |
| 09:09.09 | Turl | and get all the privs that way |
| 09:09.09 | zenitraM | mmm it could be that one then |
| 09:09.39 | Turl | zenitraM: that file is hardly publicly accesible though |
| 09:10.09 | Turl | zenitraM: pidgin stores all your passwords on plaintext on a xml file on ~/.pidgin/ |
| 09:10.13 | zenitraM | well, true, only by su |
| 09:10.40 | zenitraM | but it is useful to know, especially when there is a unfixed bug in wpa_supplicant which deletes the eduroam WiFi keys |
| 09:10.40 | Turl | s/.pidgin/.purple/ |
| 09:11.01 | Turl | zenitraM: it was fixed |
| 09:11.04 | Turl | on CM at least |
| 09:11.37 | zenitraM | Turl: it still deletes them for me |
| 09:11.46 | Turl | what are you running? |
| 09:12.04 | zenitraM | on both cm7 nightly on my desire and cm9 and aokp on my nexus |
| 09:12.29 | Turl | well no idea re. AOKP |
| 09:12.47 | Turl | but I'm pretty sure it was fixed |
| 09:13.03 | zenitraM | i had to whip a script that sets the password manually using wpa_cli and run it from tasker |
| 09:13.28 | zenitraM | idk.. at least for me it isn't. maybe there's something else more on my uni's routers |
| 09:16.31 | Turl | zenitraM: http://review.cyanogenmod.com/#/c/9259/1 |
| 09:18.01 | zenitraM | Turl: http://code.google.com/p/cyanogenmod/issues/detail?id=2479 <-- comments by me commenting it wasn't fixed for me after it was pushed |
| 09:18.36 | zenitraM | might not be the same bug though |
| 09:20.08 | zenitraM | mine drops while roaming between different routers |
| 09:21.08 | Turl | zenitraM: are you using EAP? |
| 09:21.24 | zenitraM | yup |
| 09:23.51 | Turl | it seems the wifi stuff got rewritten on the [HC, ICS] timeframe |
| 09:24.30 | zenitraM | no, that was while still on GB |
| 09:24.51 | Turl | yeah, but I mean, the code got moved around post-gb |
| 09:24.57 | zenitraM | two or three nightly builds after the fix was pushed |
| 09:25.04 | zenitraM | mmm idk |
| 09:28.17 | Turl | zenitraM: all the config writes have been moved to wifi/java/android/net/wifi/WifiConfigStore.java |
| 09:28.34 | Turl | I suppose you could put some logging on there if you want to debug this |
| 09:29.28 | zenitraM | i'll try when i have time... |
| 09:29.34 | zenitraM | although the key is removed by: |
| 09:29.35 | zenitraM | V/WifiStateTracker( 182): Changing supplicant state: ASSOCIATED ==> DISCONNECTED |
| 09:29.35 | zenitraM | I/wpa_supplicant( 4596): CTRL-EVENT-DISCONNECTED - Disconnect event - remove keys |
| 09:29.36 | zenitraM | V/WifiMonitor( 182): Event [CTRL-EVENT-DISCONNECTED - Disconnect event - remove keys] |
| 09:29.48 | zenitraM | or so it seems |
| 09:31.31 | Turl | that's unrelated |
| 09:31.42 | Turl | that calls wpa_clear_keys |
| 09:31.44 | Turl | <PROTECTED> |
| 09:31.44 | Turl | <PROTECTED> |
| 09:31.45 | Turl | <PROTECTED> |
| 09:31.45 | Turl | void wpa_clear_keys(struct wpa_supplicant *wpa_s, const u8 *addr) |
| 09:32.34 | zenitraM | i'll have to debug it then |
| 09:34.08 | *** join/#arm-netbook Quarx (~Quarx@46.233.229.38) |
| 10:21.13 | *** join/#arm-netbook kaspter (~kaspter@116.231.122.26) |
| 10:22.07 | mnemoc | Turl: sun5i sources come with a drivers/input/touchscreen/gt818_ts.h ... may that be yours? |
| 10:22.36 | mnemoc | .c |
| 10:25.20 | mnemoc | and drivers/input/touchscreen/goodix_touch.c |
| 10:29.57 | mnemoc | hno: arch/arm/mach-sun5i/ comes with a binary built-in.o! |
| 10:30.14 | mnemoc | hno: no sources at all |
| 10:34.56 | zenitraM | as Eva said, "SDK from Allwinner is complied with GPL." |
| 10:35.03 | zenitraM | she probably meant.. "is compiled with GPL." |
| 10:36.15 | mnemoc | this people is incredible... |
| 10:37.14 | zenitraM | are those sources already uploaded somewhere? |
| 10:37.38 | mnemoc | we have a leaked snapshot from december |
| 10:38.07 | mnemoc | but still, WTF |
| 10:38.10 | zenitraM | i mean the ones from tsvetan |
| 10:38.31 | mnemoc | nope, from Qware |
| 10:39.22 | zenitraM | so they just threw an .o to a client... |
| 10:41.44 | mnemoc | yes |
| 10:42.47 | zenitraM | *facepalm* |
| 10:43.08 | mnemoc | at least we got some drivers... |
| 10:54.45 | CIA-126 | rhombus-tech: master * r6be08309aef5 /allwinner_a10/orders/.mdwn: |
| 11:01.00 | *** join/#arm-netbook killerkink (af9cdba0@gateway/web/freenode/ip.175.156.219.160) |
| 11:12.06 | *** join/#arm-netbook eflatun (~eflatun@46.1.118.211) |
| 11:12.09 | hipboi | mnemoc: zenitraM: see ? |
| 11:12.18 | hipboi | what i have said before? |
| 11:12.33 | hipboi | they don't care the f**ing GPL |
| 11:14.00 | mnemoc | hipboi: sad |
| 11:15.03 | mnemoc | hipboi: everything would be far easier if they play by the rules.... and they would win a lot from it |
| 11:15.35 | hipboi | you can not imagin the market guys' world |
| 11:15.40 | mnemoc | hipboi: but thanks to mr Tom Cubie the sources we already have are having their licenses fixed :) |
| 11:16.10 | mnemoc | btw, is cedarx IP owned by allwinner or by someone else? |
| 11:16.20 | hipboi | i am not sure |
| 11:16.22 | lundman | woo |
| 11:16.39 | hipboi | by allwinner, i think |
| 11:16.50 | hipboi | it's the core technogoly of allwinner |
| 11:17.14 | mnemoc | hipboi: having that open sources we can add support on every video processing library... ffmpeg, vlc, gstreamer, .... |
| 11:17.35 | hipboi | since it's core technology |
| 11:17.42 | hipboi | it's very sensitive |
| 11:17.46 | mnemoc | but the magic is in the chip |
| 11:17.58 | mnemoc | not in the library |
| 11:18.48 | mnemoc | open sourcing them will only benefit devices with allwinner chips |
| 11:19.06 | *** join/#arm-netbook eflatun (~eflatun@46.1.118.211) |
| 11:19.18 | mnemoc | from what i've seen the libraries are pretty dumb, only passing stuff through |
| 11:19.51 | zenitraM | exactly |
| 11:20.04 | zenitraM | and there is definetly market for an usable, cheap as hell, realtime h264 encoder |
| 11:48.34 | *** join/#arm-netbook eflatun_ (~eflatun@46.1.31.203) |
| 11:48.58 | *** join/#arm-netbook RITRedbeard (~RITReadbe@c-68-37-165-37.hsd1.nj.comcast.net) |
| 12:03.06 | *** join/#arm-netbook eebrah (~Ibrahim_K@41-139-199-130.safaricombusiness.co.ke) |
| 12:08.03 | *** join/#arm-netbook RITRedbeard (~RITReadbe@c-68-37-165-37.hsd1.nj.comcast.net) |
| 12:11.18 | *** join/#arm-netbook MMlosh (~MMlosh@2001:718:1e03:5168:4526:d463:2a4:b3c2) |
| 12:41.32 | *** join/#arm-netbook eflatun (~eflatun@46.1.31.203) |
| 12:44.55 | *** join/#arm-netbook Gujs (~Gujs@cpe-92-37-87-208.dynamic.amis.net) |
| 12:48.57 | *** join/#arm-netbook kaspter (~kaspter@218.82.120.78) |
| 12:53.02 | *** join/#arm-netbook eflatun (~eflatun@46.1.31.203) |
| 13:04.48 | *** join/#arm-netbook QingPei (~qingpei@inet-hqmc03-o.oracle.com) |
| 13:10.50 | mnemoc | uh... google posting job descriptions for hardware engineers... (for mountain view, CA) |
| 13:13.15 | CIA-126 | rhombus-tech: lkcl master * r67cfbfcb40e4 /allwinner_a10/orders/ (not_valid/twoflower.mdwn shantak.mdwn twoflower.mdwn): order processing cleanup |
| 13:13.58 | CIA-126 | rhombus-tech: lkcl master * r790e62ee9a46 /allwinner_a10/orders/ (.mdwn catoff.mdwn): order processing cleanup |
| 13:17.59 | *** join/#arm-netbook PhantomsDad (~PhantomsD@c-69-255-33-53.hsd1.md.comcast.net) |
| 13:18.00 | *** join/#arm-netbook PhantomsDad (~PhantomsD@konversation/developer/PhantomsDad) |
| 13:47.20 | *** join/#arm-netbook RITRedbeard (~RITReadbe@c-68-37-165-37.hsd1.nj.comcast.net) |
| 14:10.51 | traeak | ugh california |
| 14:11.23 | steev | Turl: i'm 100% positive, however once the defconfig gets copied into obj/KERNEL_OBJS CONFIG_GPU_MALI is turned on |
| 14:13.41 | steev | http://pastebin.com/bFMzsMU0 |
| 14:14.05 | steev | Turl: i'm not doing anything special to the kernel at all. i'm even using the zatab_defconfig in my board.mk |
| 14:19.10 | steev | i've tried blowing away the kernel and obj/KERNEL_OBJS directories multiple times, i'm not doing anything "fancy" to the build system at all, quite literally it's . build/envsetup.sh && brunch a710 |
| 14:29.02 | CIA-126 | rhombus-tech: Gary master * r2f2b51bf2b68 /allwinner_a10/hacking_the_mele_a1000/Building_Debian_From_Source_Code_for_Mele.mdwn: Mentione Adesso wireless keyboard and trackball. |
| 14:32.41 | CIA-126 | rhombus-tech: Gary master * rc24fed408988 /allwinner_a10/A10_register_guide.mdwn: Minor spelling and correct subpage Page Location. |
| 14:45.19 | *** join/#arm-netbook killerkink (af9cdba0@gateway/web/freenode/ip.175.156.219.160) |
| 14:46.16 | killerkink | just discover that for 3.0.31(wip mali branch), tablet will reboot when going to deep suspend. anyone using this branch? |
| 14:51.37 | killerkink | turl: which git branch are you running on now? |
| 14:54.13 | Turl | steev: paste arch/arm/Kconfig plz |
| 14:56.15 | Turl | mnemoc: interesting.. any chance you found an archive with just the driver files floating on the interwebz? :) |
| 14:56.48 | Turl | I'm kinda glad I can type with two fingers, lets me type with frozen hands :D |
| 14:58.59 | Turl | killerkink: allwinner-3.0 and the /cpufreq wip one |
| 14:59.45 | killerkink | Turl: i am going to try allwinner-v3.0-android-v2 first |
| 15:00.14 | libv | Turl: i would think https://github.com/amery/allwinner-a10-video/ |
| 15:00.35 | Turl | libv: hm? |
| 15:00.36 | killerkink | Turl: have to try to suspend without connecting the usb? |
| 15:00.44 | libv | for cedarx |
| 15:00.57 | Turl | libv: I meant the touch ones he found |
| 15:01.01 | libv | ah, ok |
| 15:01.33 | Turl | killerkink: your tablet reboots, or is it just your userspace? |
| 15:01.59 | robws | are there any restrictions/rules on what pins can be remapped on the A10 ? For example, would I be able to remap the hdmi port into a usb port or uart ? |
| 15:02.13 | Turl | (btw, I've never seen the A10 reboot so far, the watchdog is off/unimplemented or something |
| 15:02.14 | Turl | ) |
| 15:02.15 | killerkink | Turl: once it going into full suspend, it will not wake up and after a short while reboot |
| 15:03.04 | Turl | killerkink: reboot as in full or 'soft' reboot? |
| 15:03.26 | killerkink | after from the boot up screen not boot animation |
| 15:03.43 | killerkink | i guess that the "full reboot"? |
| 15:03.48 | Turl | weird |
| 15:03.55 | Turl | my a10 never ever rebooted so far :P |
| 15:04.05 | Turl | and I've run buggy kernel code on it |
| 15:04.06 | killerkink | yeah, it only on v.3.0.31 |
| 15:04.19 | Turl | I have to long press power to kill it then start it manually |
| 15:04.50 | killerkink | i am running the default sun4i_crane config |
| 15:05.16 | killerkink | one of the modules that is put to sleep must be causing it! |
| 15:07.38 | Turl | possibly |
| 15:08.04 | Turl | killerkink: do you have any working, from source kernel? |
| 15:09.04 | killerkink | Turl: allwinner-v3.0-android-v2 is the same but i am using the modules from wip/mali |
| 15:09.24 | killerkink | 3.0.8 stock is working. |
| 15:09.30 | killerkink | but i wonder if it could be userspace |
| 15:09.34 | Turl | killerkink: the mali stuff was merged on allwinner-v3.0-android-v2 |
| 15:09.55 | killerkink | turl: since when? it wasnt a short while back |
| 15:10.08 | Turl | killerkink: like, yesterday or the day before |
| 15:10.21 | killerkink | oh |
| 15:10.27 | killerkink | yeah. just saw the commit |
| 15:10.50 | killerkink | but i dont think my branch has been updated |
| 15:11.02 | killerkink | i fork it before the merge |
| 15:11.20 | traeak | anyone have a clue about the performance of these brazos 2.0 cpus from amd ? |
| 15:11.54 | Turl | killerkink: what tab do you have btw? |
| 15:11.59 | killerkink | p76ti |
| 15:11.59 | Turl | or what's your device if it's not a tab? |
| 15:12.25 | killerkink | its a 800x480 512mb tab |
| 15:12.48 | Turl | is that the one with the hugely broken backlight? :P |
| 15:12.50 | killerkink | i will have to go back to the old mali and try first |
| 15:13.02 | killerkink | turl: no, my backlight is fine. :) |
| 15:13.22 | killerkink | but its has one end with a bad view angle |
| 15:14.35 | Turl | so your backlight doesn't turn off when you lower brightness? |
| 15:14.40 | killerkink | no |
| 15:14.54 | Turl | it must be some other tab then :) |
| 15:15.03 | killerkink | teclast a10t |
| 15:15.07 | killerkink | mine is teclast p76ti |
| 15:15.20 | killerkink | a10t has that issue |
| 15:16.58 | traeak | mnemoc: ouch the new amd "e" series processors are exactly the same as the previous with some mhz tweaks, but a better southbridge (includes usb3) |
| 15:17.21 | traeak | and everyone wonders why netbooks aren't selling...no compelling reasons for anyone to upgrade |
| 15:18.18 | Turl | traeak: that's not an argument |
| 15:18.27 | Turl | traeak: look at iphone 4 and 4S, yet they sell them |
| 15:18.32 | mnemoc | add crappy keyboards and crappy displays |
| 15:19.15 | mnemoc | Turl: Apple fanboys buy whatever Apple launches, they need no reason |
| 15:19.30 | traeak | Turl: apple can market bricks and they woudl sell like hotcakes it's not a good comparison |
| 15:19.43 | Turl | haha |
| 15:19.49 | Turl | SGS2 -> SGS3 |
| 15:19.52 | traeak | well here's the other deal |
| 15:19.54 | Turl | like that one better? :P |
| 15:20.08 | traeak | these stupid |"ultrabook frizbees" tried to copy apple to make lots of money |
| 15:20.13 | traeak | they are tanking from what i understand |
| 15:20.34 | Turl | the intel ones were selling good I thought? |
| 15:20.49 | Turl | I'd buy one if I had the money, my core2duo has served me well already :P |
| 15:21.07 | traeak | super thin for real expensive? |
| 15:21.26 | Turl | real expensive? |
| 15:21.35 | Turl | the prices were about the same as a notebook |
| 15:21.41 | killerkink | turl: i guess it not the mali driver. i switch to r2p3...and its the same |
| 15:21.53 | mnemoc | Turl: notebooks are overpriced |
| 15:21.59 | killerkink | going to have to go back to stock kernel |
| 15:22.05 | Turl | mnemoc: of course, but they've always been compared to desktops |
| 15:22.14 | Turl | they take more design cost $$ |
| 15:22.26 | Turl | as you have to fit components on a small case and make it light |
| 15:22.31 | zenitraM | ultrabooks are like 400 less than their macbook air equivalents. that's enough for me |
| 15:22.46 | traeak | and we dont' care if a netbook is x86 or not honestly :-p |
| 15:23.01 | Turl | *mac* are like 200% of their pc counterparts :P |
| 15:23.01 | traeak | ultrabooks are like 800USD and up |
| 15:23.26 | traeak | i was seriously eyeing a timelineX a couple years back |
| 15:23.45 | mnemoc | they are not worth more than $500 |
| 15:24.09 | Turl | traeak: the timeline series is nice |
| 15:24.18 | Turl | I remember when they launched and promised 8-12h |
| 15:24.26 | Turl | or 24 or w/e don0t remember :P |
| 15:24.42 | Turl | the standard those days was like 1-2h batt life :P |
| 15:24.53 | steev | Turl: http://paste.lugons.org/show/1935/ |
| 15:25.14 | Turl | woot! lodgeit :D |
| 15:25.17 | traeak | well acer killt the timelinex so they could re-release them as ultrabooks for 300USD markup |
| 15:25.19 | Turl | I love that paste script |
| 15:25.25 | Turl | too bad the main site shut down :( |
| 15:26.04 | Turl | steev: looks good to me |
| 15:26.09 | Turl | steev: what's the issue then? |
| 15:26.24 | steev | Turl: it tries to build the mali module, and dies, see paste from earlier |
| 15:26.30 | steev | s/earlier/last night/ |
| 15:27.16 | traeak | actually usb3.0 is a big deal, along with hdmi |
| 15:27.20 | Turl | steev: I see a paste with greps only |
| 15:27.25 | traeak | okay back to work |
| 15:27.26 | Turl | got the actual error? |
| 15:27.44 | Turl | mnemoc: any chance you grabbed the driver files? :) |
| 15:27.47 | mnemoc | large external storage over usb2 sucks badly |
| 15:27.52 | mnemoc | Turl: 1m |
| 15:28.02 | Turl | mnemoc: I have a 1TB HDD |
| 15:28.06 | Turl | over USB |
| 15:28.12 | steev | 12:08 AM <steev> Turl: in case you get bored, http://pastebin.com/JNcZtPLv i can't figure out where the issue comes from. not sure what is causing all those gcc: not found, or why it keeps trying to build the mali stuff. the kernel config is the sun4i_crane_defconfig, kernel is amery/linux-allwinner, branch is allwinner-v3.0-android-v2 |
| 15:28.14 | mnemoc | Turl: I'm pushing a new wip/ |
| 15:28.15 | Turl | it's ok to mv the big stuff when your HDD fills up |
| 15:28.42 | steev | Turl: since then i switched to allwinner-dev-team/linux-allwinner and using the zatab_defconfig, still same issue |
| 15:28.59 | Turl | steev: there's something really fishy there |
| 15:29.07 | Turl | steev: your gcc fails continuously |
| 15:29.30 | Turl | steev: what make version are you running? |
| 15:29.53 | steev | Turl: 3.82 (gentoo default) |
| 15:30.04 | killerkink | turl: 3.0.8 is fine for me |
| 15:30.23 | Turl | steev: try with 3.81 |
| 15:30.24 | killerkink | so i guess there is something in 3.0.31 that is not working for me |
| 15:30.24 | killerkink | p76ti |
| 15:30.31 | Turl | I've seen people complaining about 82 |
| 15:30.54 | Turl | killerkink: source 3.0.8 as in the qware leak, or a blob you got? |
| 15:31.17 | killerkink | qware leak |
| 15:31.33 | steev | Turl: gentoo patches that issue |
| 15:31.53 | Turl | killerkink: know how to use git bisect? |
| 15:32.02 | Turl | steev: still, you don't miss much trying do you? |
| 15:32.24 | steev | installing now |
| 15:32.25 | Turl | you shouldn't be getting all those 'gcc: error: directory: No such file or directory' |
| 15:32.28 | killerkink | turl: no, i am very new to git |
| 15:32.32 | steev | i know i shouldn't |
| 15:32.49 | mnemoc | steev: -j1? |
| 15:32.58 | killerkink | let me try to compare .config first |
| 15:33.00 | traeak | problem i've had with usb3.0 on a laptop is that it doesn't provide enough power to run the external usb3.0 drive |
| 15:33.01 | steev | mnemoc: tried -j1 |
| 15:33.06 | mnemoc | :< |
| 15:33.16 | steev | i'll give make 3.81 a shot, doesn't take long to install |
| 15:33.58 | Turl | killerkink: http://ivanz.com/2009/03/27/git-bisect-the-awesome-way-to-find-the-mr-bug-commit/ |
| 15:35.10 | Turl | traeak: my usb2 port on my laptop runs a usb2 hdd on it's own.. |
| 15:35.20 | traeak | Turl: usb2 IS NOT usb3 |
| 15:35.28 | Turl | traeak: but HDD tech is the same |
| 15:35.28 | traeak | Turl: that same drive runs fine off usb2 |
| 15:35.39 | Turl | and if anything usb3 provides more power over the wire |
| 15:36.11 | traeak | SUPPOSED to sure, what i'm sayign is that usb3 seems to still be unstable |
| 15:36.32 | killerkink | turl: thanks, will try it later. but you are not having that issue. i wonder why. |
| 15:36.51 | killerkink | so, it may not be a regression? |
| 15:41.14 | mnemoc | Turl: found the branch? |
| 15:42.41 | Turl | this? https://github.com/amery/linux-allwinner/commits/wip/lichee-3.0.8-sun5i/import |
| 15:42.46 | Turl | if so then I did :P |
| 15:43.28 | mnemoc | ehm... I didn't want to be that obvius... yet |
| 15:45.53 | Turl | oh, sorry :\ |
| 15:46.06 | Turl | driver looks promising, let me build it |
| 15:47.17 | Turl | the other module prints version as 801NI though |
| 15:47.20 | Turl | not 818 |
| 15:48.16 | mnemoc | I'm not touching anything... only making a non destructive merge with our lichee-3.0.8-sun4i branch |
| 15:48.53 | steev | Turl: 3.81 still same |
| 15:49.30 | steev | http://pastebin.com/9bbb9YG4 |
| 15:50.14 | mnemoc | steev: have you built the kernel manually in that amachine? |
| 15:50.49 | steev | mnemoc: nope, when i tried it was same error (using the android cross compiler) |
| 15:50.58 | steev | but |
| 15:50.59 | killerkink | turl: is pmem important? |
| 15:51.05 | steev | let me try with my hardfloat cross compiler |
| 15:51.06 | steev | killerkink: yes |
| 15:51.27 | killerkink | thats what i thought? |
| 15:51.41 | killerkink | its not in 3.0.31? |
| 15:51.59 | killerkink | or different directory from 3.0.8? |
| 15:52.05 | steev | mmm |
| 15:52.18 | mnemoc | pmem was deprectated by android short after 3.0.8 |
| 15:52.20 | killerkink | in 3.0.8, its in drivers/misc |
| 15:52.23 | steev | oh |
| 15:52.31 | killerkink | i see |
| 15:52.32 | steev | i didn't know about the deprecation |
| 15:52.40 | killerkink | ok |
| 15:52.41 | steev | what's the new hotness mnemoc ? |
| 15:53.08 | mnemoc | they removed it from android-3.0 entirely |
| 15:53.21 | Turl | pmem is used by qualcomm for sure, but I'm doubting you need it for other chips nowadays |
| 15:53.23 | killerkink | http://groups.google.com/group/android-kernel/browse_thread/thread/e8c1d42d5c67f7b5? |
| 15:53.26 | mnemoc | I believe they don't need it anymore.... it was very hackish |
| 15:53.34 | killerkink | http://groups.google.com/group/android-kernel/browse_thread/thread/e8c1d42d5c67f7b5 |
| 15:55.12 | steev | huh, didn't know there was an android-kernel group |
| 15:55.30 | mnemoc | probably linaro people |
| 15:58.14 | killerkink | turl: i think i will stick with 3.0.8 for now. all the issues that i am facing is giving me a headache |
| 15:58.25 | killerkink | will check you .31 once i have a stable rom |
| 15:58.33 | killerkink | s/you/out |
| 15:58.38 | killerkink | s/you/out/ |
| 16:00.12 | *** join/#arm-netbook eflatun (~eflatun@46.1.118.35) |
| 16:03.26 | killerkink | turl: thanks for your advice. going off |
| 16:03.30 | libv | after having googled for the cedarX binaries, and subsequently falsely pasted to Turl, i had a brief glance at them |
| 16:03.30 | killerkink | bye all |
| 16:04.13 | libv | if anyone is willing to invest several manmonths into REing this, send them to me, i'll be glad to give some pointers and directions |
| 16:04.19 | mnemoc | libv: that code I have on github is sample code and ducumentation to use the cedarx .a libraries, not the library sources |
| 16:04.48 | mnemoc | libv: sepp_ already started REing it afaik |
| 16:04.56 | libv | cool |
| 16:06.12 | *** join/#arm-netbook RITRedbeard (~RITReadbe@c-68-37-165-37.hsd1.nj.comcast.net) |
| 16:06.25 | libv | i just msged him to see where he's at :) |
| 16:08.06 | steev | Turl: mnemoc: http://pastebin.com/YQYKnVZ0 |
| 16:09.06 | mnemoc | steev: eh?? on my repo or Turl's? |
| 16:09.07 | Turl | steev: there's some piece on your stack that's failing hard :P |
| 16:09.32 | Turl | steev: what's your host GCC? |
| 16:09.52 | steev | not using my host gcc, but it's currently 4.7.0 |
| 16:10.04 | Turl | steev: it uses the host gcc |
| 16:10.04 | steev | cross compile version is 4.5.3 |
| 16:10.09 | Turl | to build the menuconfig stuff and such |
| 16:10.14 | steev | neg |
| 16:10.32 | steev | make ARCH=arm CROSS_COMPILE=blah zatab_defconfig uses the cross compiler to do it |
| 16:10.57 | Turl | steev: and the cross compiled code runs on your x86 machine to track config deps and show UI? :) |
| 16:11.06 | mnemoc | steev: can you reset to c6564c6c2d9115d15016a1634c580bc7496d35d5 to discard mali problems? |
| 16:11.19 | steev | mnemoc: i'll give it a shot |
| 16:11.51 | steev | HEAD is now at c6564c6 Merge branch 'upstream-android-3.0' into allwinner-v3.0-android-v2 |
| 16:12.31 | mnemoc | that's right before I merged the mali in-treeing branch |
| 16:12.36 | steev | okay i need to bisect from c6564c..HEAD |
| 16:12.50 | steev | because that builds fine |
| 16:13.07 | Turl | steev: are you using a weirdo shell by any chance? |
| 16:13.18 | steev | if bash is weird, sure |
| 16:13.23 | Turl | bash is ok |
| 16:14.04 | steev | i'll look through the commits from c6564c6 and see what comes up |
| 16:14.25 | Turl | #error "No supported mali core defined" |
| 16:14.41 | Turl | my bet is that it's not even exporting the config options |
| 16:15.05 | steev | it's only 19 patches |
| 16:15.26 | mnemoc | steev: more than the patches itself I believe it's something in mali MAkefile |
| 16:15.34 | steev | mnemoc: i'm sure it is :) |
| 16:15.49 | steev | specially since i've never gotten mali to build here |
| 16:15.57 | mnemoc | damn mnemoc for merging that after only testing ubuntu+gcc 4.6 |
| 16:16.12 | *** join/#arm-netbook RITRedbeard (~RITReadbe@c-68-37-165-37.hsd1.nj.comcast.net) |
| 16:17.05 | *** join/#arm-netbook mpthompson (~IceChat77@c-50-131-39-81.hsd1.ca.comcast.net) |
| 16:17.23 | Turl | steev: https://github.com/amery/linux-allwinner/blob/allwinner-v3.0-android-v2/drivers/gpu/mali/Makefile |
| 16:17.41 | Turl | but, idk what might be wrong on such a simple makefile |
| 16:18.29 | mnemoc | probably not that one |
| 16:18.43 | mnemoc | but something inside |
| 16:18.56 | steev | i'm using allwinner-dev-team to use the zatab_defconfig |
| 16:19.05 | steev | i figure i'm close enough to a zatab :P |
| 16:19.10 | mnemoc | steev: copy the .config :p |
| 16:19.22 | steev | mnemoc: zatab_defconfig IS the .config |
| 16:24.27 | steev | ugh, cm9 doesn't run a mrproper on the kernel |
| 16:24.49 | Turl | steev: it dies if you don't :) |
| 16:24.59 | steev | right, that should be automated |
| 16:25.04 | Turl | forcing an mrproper is not a good idea on a developer machine :) |
| 16:25.33 | Turl | steev: can you build manually but add V=1 so we see what's being run? |
| 16:26.11 | steev | sure, one sec |
| 16:26.12 | *** join/#arm-netbook RITRedbeard (~RITReadbe@c-68-37-165-37.hsd1.nj.comcast.net) |
| 16:26.45 | steev | need to build the uImage first, i'm sure you don't want all that (it works fine), it's just the modules that fails |
| 16:28.19 | steev | Turl: would you prefer i use my cross compiler, or the android one (they both die the same) |
| 16:28.37 | Turl | same for me |
| 16:30.38 | steev | gonna take it a bit, i forgot i was updating |
| 16:30.53 | DonkeyHotei | i'm looking through aliexpress. is gc800 different from mali400? |
| 16:31.30 | mnemoc | all chars except the last 2 |
| 16:31.58 | DonkeyHotei | lol |
| 16:32.09 | DonkeyHotei | are they different gpu's? |
| 16:32.38 | steev | yes |
| 16:32.43 | steev | gc800 is a vivante afaik |
| 16:33.12 | steev | i have sources for that :( |
| 16:34.19 | specing | define 'sources' |
| 16:35.43 | steev | http://pastebin.com/dvtSx1kL |
| 16:35.47 | steev | specing: kernel and userspace |
| 16:35.53 | steev | drivers |
| 16:36.17 | Turl | isn't vivante the one on rpi? |
| 16:36.21 | steev | and no, i can't release them, unless i have some huge fit of OMGIHATEMYJOBANDWANTTOBEFIREDANDNEVERWORKASAPROGRAMMERAGAIN |
| 16:36.27 | Turl | that was so critisized because it's closed sauce? |
| 16:36.52 | steev | doesn't know of any userspace that's open source in ARM SoC land |
| 16:36.53 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
| 16:36.58 | PhantomsDad | watches venus transit the sun: http://venustransit.gsfc.nasa.gov/ |
| 16:37.24 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
| 16:37.35 | steev | Turl: pastebin above |
| 16:37.58 | steev | drivers/gpu/mali/mali *does* exist |
| 16:38.26 | libv | rpi is broadcoms own videocore 4 |
| 16:38.30 | specing | steev: You have the actual source code? |
| 16:38.32 | specing | Wow |
| 16:38.53 | steev | specing: fsl uses the vivante for the mx6 |
| 16:39.23 | libv | specing: similarly, i have the code to pvr, means that i will never ever get to do anything free software with it. nothing to be proud of. |
| 16:39.30 | specing | Well atleast you can do various builds and share them, right? |
| 16:39.43 | steev | the only thing we don't have sources to for the mx5{1,3} is the shader compilers |
| 16:39.47 | libv | luckily pvr is pointless anyway |
| 16:39.48 | steev | specing: not really, no |
| 16:39.54 | steev | i mean i could, sure |
| 16:39.57 | specing | You can't do that either? |
| 16:40.27 | steev | but i don't see the point in doing it, aside from if someone needs it for a very specific purpose, and it's not like people are going to be submitting patches ;) |
| 16:41.12 | specing | Are there armv7a-hf build of PVR drivers? |
| 16:41.32 | specing | steev: But would it be legal for you to do it? |
| 16:43.51 | Turl | steev: -DSVN_REV_STRING=\"-rUnversioned directory\" |
| 16:43.57 | Turl | directory" sounds familiar :) |
| 16:44.55 | mnemoc | should I revert all the mali-intreeing stuff? |
| 16:46.16 | Turl | not yet mnemoc |
| 16:46.53 | mnemoc | ok |
| 16:47.08 | mnemoc | "intreeing".... what a word |
| 16:47.35 | specing | I see my enthusiasm just hit a corporate wall right there |
| 16:48.44 | libv | specing: it probably depends on the individual terms of the nda, but it would be considered quite a grey area, and it would only serve to help the hw vendors with binary drivers |
| 16:48.48 | mnemoc | specing: steev works for genesi doing imx6 stuff.... with NDAs over NDAs and all that stuff |
| 16:50.23 | Turl | steev: try this http://paste.debian.net/173164/ |
| 16:51.38 | Turl | mnemoc: no touch luck :( |
| 16:57.13 | Turl | so, I managed to render my device unbootable, again ~.~ |
| 16:57.50 | steev | stil same :( |
| 16:57.55 | steev | meeting time though |
| 16:59.07 | Turl | fu livesuite time |
| 16:59.33 | Turl | is there any way to just livesuite the nanda? |
| 16:59.45 | steev | not that i know of |
| 17:00.25 | Turl | that damn thing corrupts or something |
| 17:01.41 | steev | specing: mnemoc: it's really not so bad wrt NDAs, for the most part my job ignores/avoids them. i only have to pay attention when i do something like update a package for our releases |
| 17:03.26 | tallnerd1985 | mnemoc: Hey buddy, is there anybody working on any builds for the mk802 or is it still too new? |
| 17:03.44 | CIA-126 | rhombus-tech: hno master * rc44e074101d7 /allwinner_a10/a10_image.mdwn: |
| 17:05.35 | ZaEarl | Turl, rebuild the livesuit image: http://forum.xda-developers.com/showpost.php?p=22397975&postcount=5 |
| 17:06.22 | CIA-126 | rhombus-tech: hno master * re435a4f310d5 /allwinner_a10/a10_image.mdwn: |
| 17:07.28 | mnemoc | tallnerd1985: not that I'm aware |
| 17:08.55 | *** join/#arm-netbook eflatun (~eflatun@46.1.118.35) |
| 17:09.52 | tallnerd1985 | Okay, thanks |
| 17:17.50 | steev | Turl: hang on, looks like i'm missing something somewhere, definitely see the issue |
| 17:18.32 | Turl | ZaEarl: doesn't look like of much use |
| 17:19.35 | Turl | who came up with the idea of having a bootloader that reads configs from a fat partition? |
| 17:20.00 | *** join/#arm-netbook DonkeyHotei (~danielg4@50-0-149-176.dsl.dynamic.sonic.net) |
| 17:20.04 | steev | Turl: it should support ext2 as well, just a matter of the command being changed from fatload to ext2load |
| 17:20.48 | steev | Turl: this is interesting, according to a grep, it SHOULD default to 0000 if no version is specified, so why does it set it to Unversioned directory |
| 17:20.50 | specing | bookmarks steev and libv if he ever lays his hands on a tablet with those chips ;) |
| 17:21.15 | Turl | steev: diff svn version? |
| 17:21.23 | Turl | I think I don't even have svn installed |
| 17:21.30 | libv | specing: i will not do anything with sgx anymore |
| 17:21.41 | specing | :( |
| 17:21.50 | steev | specing: well you can get the mx6 tablet dev board in july, but it will be ~1500-2k |
| 17:22.07 | specing | Not that rich |
| 17:22.24 | steev | i really wish i could help out with freedreno, but due to having the sources for the amd gpu.... :/ |
| 17:22.34 | specing | I have a 10" OMAP tablet though |
| 17:22.48 | specing | Lost interest in it due to no keyboard :S |
| 17:22.56 | specing | And having this A10 one |
| 17:24.11 | tallnerd1985 | How is the performance of these tablets compared to let's say a Tegra 2 reference tablet? |
| 17:26.39 | specing | define 'performance'? |
| 17:27.31 | tallnerd1985 | Not Quadrant or any crap like that but as far as navigation and fluidity of the UI |
| 17:27.46 | Turl | tallnerd1985: it's usable |
| 17:27.52 | *** join/#arm-netbook vgrade (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 17:32.37 | steev | Turl: got it |
| 17:32.48 | steev | nope, almost got it |
| 17:34.22 | steev | Turl: okay, it doesn't like the MODULE_VERSION=(SVN_REV_STRING) |
| 17:34.26 | steev | inside ump and mali |
| 17:35.43 | steev | replacing those with "0000" (quotes required), it builds |
| 17:37.45 | steev | now to test if that works for my hack job |
| 17:40.52 | steev | \o/ |
| 17:47.42 | steev | okay. |
| 17:48.42 | steev | Turl: http://pastebin.com/bxM5dCSq it's a huge hack but it works |
| 17:48.52 | steev | i'm sure there is a better way |
| 17:49.06 | steev | maybe i'd get something if i had git svn installed |
| 17:50.54 | Turl | steev: do you have svnversion on your path? |
| 17:51.04 | steev | nope |
| 17:51.15 | steev | svn i have, but i dunno what this svnversion bit is |
| 17:51.20 | steev | oh |
| 17:51.22 | steev | wait i do |
| 17:51.27 | steev | and that's exactly what it comes from |
| 17:51.34 | steev | steev@lumen /storage/allwinner/a710 $ svnversion |
| 17:51.34 | steev | Unversioned directory |
| 17:51.40 | steev | however |
| 17:52.32 | steev | <lots of scrolling back in my terminal> |
| 17:52.49 | steev | drivers/gpu/mali/ump/linux/ump_kernel_linux.c:447:1: error: expected ‘,’ or ‘;’ before numeric constant |
| 17:53.01 | steev | (ditto for mali_kernel_linux.c (except line 487) |
| 17:53.03 | steev | ) |
| 17:56.20 | Turl | steev: what if you do |
| 17:56.28 | Turl | DEFINES += -DSVN_REV_STRING="\"$(SVN_REV)\"" |
| 18:05.50 | *** join/#arm-netbook MaDMaLKaV (~IceChat9@188.86.253.69) |
| 18:14.37 | steev | wouldn't that just make it "Unversioned directory" ? |
| 18:14.55 | steev | i mean sure it's better than not compiling, but would think they'd want an actual # in there |
| 18:15.45 | steev | sadly, i'm still stock, so i can't even test what i just built :P |
| 18:15.54 | steev | i meant to reimage last night, but forgot |
| 18:21.53 | *** join/#arm-netbook Vaevictus_ (57427124@gateway/web/freenode/ip.87.66.113.36) |
| 18:25.42 | *** join/#arm-netbook rell01 (~chatzilla@p4FE56324.dip0.t-ipconnect.de) |
| 18:30.28 | *** join/#arm-netbook eflatun (~eflatun@46.1.118.35) |
| 18:34.38 | *** part/#arm-netbook QingPei (~qingpei@inet-hqmc03-o.oracle.com) |
| 18:49.34 | *** join/#arm-netbook tomekgno (~tomekgno@edw221.neoplus.adsl.tpnet.pl) |
| 19:02.10 | *** join/#arm-netbook jimyx17 (~jaime@95.77.223.87.dynamic.jazztel.es) |
| 19:02.46 | *** join/#arm-netbook ZaEarl (~malmrose@173-11-125-162-SFBA.hfc.comcastbusiness.net) |
| 19:15.10 | *** join/#arm-netbook RITRedbeard (~RITReadbe@c-68-37-165-37.hsd1.nj.comcast.net) |
| 19:24.42 | *** join/#arm-netbook ka6sox (ka6sox@nasadmin/ka6sox) |
| 19:45.10 | *** join/#arm-netbook RITRedbeard__ (~RITReadbe@c-68-37-165-37.hsd1.nj.comcast.net) |
| 19:55.26 | *** join/#arm-netbook Hexxeh (u1532@gateway/web/irccloud.com/x-rtloyuipeewrzxym) |
| 19:57.18 | *** join/#arm-netbook steev (u1383@gentoo/developer/steev) |
| 19:57.28 | *** join/#arm-netbook vgrade2 (~martinbro@cpc2-nrte22-2-0-cust128.8-4.cable.virginmedia.com) |
| 20:09.36 | *** join/#arm-netbook tomekgno (~tomekgno@aeaf65.neoplus.adsl.tpnet.pl) |
| 20:13.03 | *** join/#arm-netbook mpthompson (~IceChat77@c-50-131-39-81.hsd1.ca.comcast.net) |
| 20:28.15 | CIA-126 | rhombus-tech: lkcl master * r0148f77b7c37 /allwinner_a10/A10_register_guide.mdwn: whitespace cleanup |
| 20:28.16 | CIA-126 | rhombus-tech: lkcl master * r0967045fde24 /allwinner_a10/a10_image.mdwn: whitespace cleanup |
| 20:37.46 | MMlosh | are/will the alpha-s the first boards produced? |
| 20:48.00 | lkcl | MMlosh: being designed. |
| 20:50.13 | MMlosh | hmm.. I suppose there is no timeframe (and that it wouldn't make sense for one to exist) |
| 20:51.44 | mnemoc | "it will be done when it's done" (tm) |
| 20:55.12 | rm | I'd say just get a Mele or MK802 and start hacking |
| 20:55.27 | hno | mnemoc, how very GPL of them... |
| 20:56.19 | hno | rm, have it been confirmed the MK802 have serial console on the pads? |
| 20:58.01 | ZaEarl | I just noticed in the mailing list the SDK is said to be GPL'd. Has anybody posted it to github? |
| 20:58.57 | ManoftheSea | lkcl, got a little heated about samba? |
| 21:00.27 | mnemoc | ZaEarl: there isn't really much to get out of it |
| 21:01.04 | ZaEarl | mnemoc, wasn't it useful for progress on x11 drivers? |
| 21:01.18 | mnemoc | ZaEarl: I'll finish pushing the non-destructive merge of lichee-3.0.8-sun4i and lichee-3.0.8-sun5i tomorrow morning |
| 21:01.32 | mnemoc | ZaEarl: nothing we don't have already |
| 21:02.01 | rm | hno, don't know, and yes I just was going to mention "might want to hold off a bit on the mk802" |
| 21:03.00 | mnemoc | ZaEarl: I can mail you a file listing if you find something interesting |
| 21:03.22 | ZaEarl | I just remember reading the published drivers saying they needed to be compiled with the sdk. |
| 21:03.34 | ZaEarl | for the mali |
| 21:03.51 | ZaEarl | this one: http://www.malideveloper.com/developer-resources/drivers/open-source-mali-gpus-linux-exadri2-and-x11-display-drivers.php |
| 21:04.02 | mnemoc | ZaEarl: we are talking about different SDKs |
| 21:04.16 | hno | ZaEarl, that's the MALI DDK you are thinking of. |
| 21:04.30 | hno | ARM proprietary. |
| 21:04.39 | ZaEarl | my mistake |
| 21:05.26 | ZaEarl | all right, then, carry on. |
| 21:05.44 | mnemoc | ZaEarl: hipboi has access to it |
| 21:06.14 | mnemoc | ZaEarl: and he could compile the drivers for us |
| 21:06.33 | mnemoc | ZaEarl: but that requires time he doesn't seem to have at the moment |
| 21:07.08 | ZaEarl | that's no good, we need to be able to compile the drivers |
| 21:07.29 | mnemoc | mail ARM |
| 21:07.55 | mnemoc | it's ARM who restricts the access to the DDK |
| 21:08.31 | mnemoc | hipboi is the closest any of us we will ever be to the DDK |
| 21:10.54 | specing | Anyone up for raiding ARM headquarters this sunday? |
| 21:12.02 | specing | Hehe :) |
| 21:12.12 | traeak | does hipboi's face shine with the glory of it? |
| 21:14.47 | mnemoc | traeak: eh? |
| 21:16.05 | traeak | no biblical scholars here, eh? |
| 21:17.35 | mnemoc | having access to something people needs but you can't share wouldn't "glorify" me... it would be awfully frustrating |
| 21:18.05 | ManoftheSea | holiest of holies it isn't |
| 21:18.06 | *** join/#arm-netbook polvo_ (~chatzilla@94.177.23.95.dynamic.jazztel.es) |
| 21:18.22 | ManoftheSea | nor will gazing upon it strike you down |
| 21:24.46 | steev | it is extremely frustrating, speaking from experience |
| 21:35.58 | hno | mnemoc, why can't mali gpu support be builtt as module? |
| 21:36.53 | mnemoc | hno: eh? afaik it can only be built as module |
| 21:37.08 | hno | Right.. Kconfig is confusing. |
| 21:37.44 | mnemoc | iirc igloo people uses something like "Enable Mali Support" |
| 21:37.51 | mnemoc | bool |
| 21:38.52 | *** join/#arm-netbook polvo_ (~chatzilla@94.177.23.95.dynamic.jazztel.es) |
| 21:39.04 | mnemoc | because it's only used to reach the makefile |
| 21:40.05 | hno | Yes. Very confusing. And probably won't if you run "make modules". |
| 21:41.53 | Turl | who asked me the other day about wifi? |
| 21:42.16 | Turl | I'm trying the mainline drivers for realtek 8192cu but they don't seem able to pick up the card on the tab |
| 21:42.57 | hno | Turl, mailnline drivers reportedly also do not work on the mele. |
| 21:43.34 | Turl | the allwinner ones are full of fail :P |
| 21:43.44 | hno | ? |
| 21:43.47 | Turl | either that or the chip is full of fail |
| 21:44.04 | Turl | they drop my connection randomly, won't scan at random times |
| 21:48.43 | mnemoc | Turl: the a13 tree has some changes there, 1m |
| 21:49.54 | Turl | hm hno |
| 21:49.59 | Turl | if I load the old module |
| 21:50.09 | Turl | it enables some usb stuff |
| 21:50.24 | hno | old as in? |
| 21:51.45 | mnemoc | try the rtl8192cu from the wip sun5i branch |
| 21:51.52 | mnemoc | Turl: ---^ |
| 21:52.16 | Turl | hno: the _SW one |
| 21:53.20 | Turl | mnemoc: that's the driver I hate :) |
| 21:53.28 | Turl | I want to use the other one :P |
| 21:54.02 | Turl | I think I found what I was looking for |
| 22:11.08 | hno | wtf is drivers/drm/mali/? Does not appear to actually contain anything at all. |
| 22:11.25 | Turl | hno: X11 DRM stuff |
| 22:12.35 | mnemoc | it's required by the x11 driver |
| 22:14.00 | hno | Yes, but... there is no connection to MALI in there at all from what I can tell. |
| 22:14.24 | hno | looks like a dummy. |
| 22:15.37 | mnemoc | it has a companion dev in arch/arm/mach-sun4i/devices.c |
| 22:16.09 | mnemoc | it doesn't make sense at all... but seems to work :| |
| 22:18.37 | hno | Have anyone got X11 to work? |
| 22:19.31 | mnemoc | not me |
| 22:20.16 | mnemoc | https://github.com/amery/linux-allwinner/issues/19 contains some info |
| 22:21.32 | mnemoc | from that it seems no one has |
| 22:21.55 | mnemoc | but I remember people using mali libraries from hipboi's ubuntu |
| 22:22.09 | mnemoc | and reporting success here |
| 22:23.14 | mnemoc | if steev can't get his problem with in-tree mali solved I'll revert all that stuff |
| 22:23.36 | mnemoc | merging it was a mistake |
| 22:24.06 | mnemoc | should have kept it on a branch longer |
| 22:24.17 | mnemoc | anyhow, need to go to sleep |
| 22:24.21 | mnemoc | good night |
| 22:24.26 | Turl | mnemoc: he fixed it |
| 22:24.54 | Turl | he has svnrevision or some other weird binary on his path, that's why it doesn't fail for us |
| 22:25.04 | mnemoc | that's good news :) |
| 22:25.12 | mnemoc | now i can sleep in peace |
| 22:25.14 | zenitraM | we can't get X11 to work without the DDK or some heavy rebasing on the snowball code which doesn't use the ddk |
| 22:25.17 | Turl | now we need to rewrite the version thingy |
| 22:25.25 | Turl | so it doesn't choke |
| 22:25.31 | Turl | but that's it, the issue is identified |
| 22:29.46 | Turl | hno: well I got it loading :D |
| 22:29.53 | Turl | I'm missing the firmware though >.> |
| 22:41.51 | hno | mnemoc, have cleaned up the MALI integration a bit. |
| 22:43.10 | Turl | <6>[ 1502.330000] rtl8192cu: Board Type 0 |
| 22:43.10 | Turl | <6>[ 1502.350000] rtl8192cu: rx_max_size 15360, rx_urb_num 8, in_ep 1 |
| 22:43.10 | Turl | <7>[ 1502.360000] ieee80211 phy3: Selected rate control algorithm 'rtl_rc' |
| 22:43.13 | Turl | hno: it loaded :D |
| 22:43.39 | Turl | hno: scanning works :D awesome |
| 22:44.31 | hno | Turl, mainline driver? |
| 22:45.27 | Turl | hno: yup :) |
| 22:45.56 | hno | Nice! |
| 22:46.52 | Turl | I'll push the bit to get it working, sec |
| 22:53.25 | Turl | hno: https://github.com/allwinner-dev-team/linux-allwinner/commit/db6d9f61577b9214a7c9f002457dd003d047e8ea |
| 22:59.57 | hno | Hmm.. what is wrong if all I get from the kernel is "DRAM: 512" then complete silence.. |
| 23:01.14 | lundman | not if that is what you wanted! |
| 23:10.37 | hno | I would prefer a bit more.. |
| 23:10.40 | *** join/#arm-netbook ZaEarl_ (~malmrose@173-11-125-162-SFBA.hfc.comcastbusiness.net) |
| 23:21.11 | hno | Hm.. rootwait do not seem to do anything at all.. |
| 23:36.42 | hno | Ah, it's rootwait not rootwait=N |
| 23:45.17 | hno | Ok, fresh compiled kernel works better... except that 8250 and sunxi_uart seemt to both claim the uart. |
| 23:57.38 | *** join/#arm-netbook killerkink (dcff028b@gateway/web/freenode/ip.220.255.2.139) |
| 23:59.14 | hno | and without EARLY_PRINTK the console seems dead.. |
| 23:59.39 | hno | Hm... |
