00:01.26 | *** join/#arm-netbook naobsd (7c294571@gateway/web/freenode/ip.124.41.69.113) |
00:15.26 | Turl | I'm failing completely today, I cannot even get uboot going on SD |
01:18.46 | CIA-91 | rhombus-tech: Jean-Marc master * r96b26b8e0724 /allwinner_a10/orders/Jeanmarc.mdwn: |
02:40.20 | *** join/#arm-netbook ZaEarl (~malmrose@173-11-125-162-SFBA.hfc.comcastbusiness.net) |
03:06.07 | t0dbld|work | anyone in here read Chinese ? |
03:10.52 | hipboi__ | me |
03:11.15 | hipboi__ | i am good at Chinese |
03:16.48 | t0dbld|work | nice |
03:17.00 | t0dbld|work | i could use some help translating a document |
03:17.36 | t0dbld|work | if i DB it and give you link think you could help me ? |
03:23.07 | hipboi__ | a document? |
03:23.11 | hipboi__ | how long |
03:23.19 | hipboi__ | if two long, i don't have time |
03:23.34 | hipboi__ | if some sentens i can translate soon |
03:24.03 | hipboi__ | i have time when building |
03:25.20 | t0dbld|work | ok |
03:25.26 | t0dbld|work | https://www.dropbox.com/s/y9gfie8ku0mfxju/crane_fex_guide.doc |
03:25.30 | t0dbld|work | you dont have to do it now |
03:25.38 | t0dbld|work | i jsut need help so i can better understand it |
03:25.42 | t0dbld|work | mostly display stuff |
03:25.58 | t0dbld|work | i dont know what the values = for alot of stuff |
03:26.23 | t0dbld|work | as i am using a tablet not a mele i need to change some things |
03:28.08 | nikk | t0dbld|work what's wrong with your fex? which conditions it doesn't suite? |
03:28.32 | t0dbld|work | well for 1 i have an internal display |
03:28.45 | t0dbld|work | i already changed ram and few other things |
03:29.09 | t0dbld|work | the display is huge issue once i get that i can pick away at others |
03:29.30 | t0dbld|work | but i need to stop having to use an external jsut to know that it even booted while i make other changes |
03:50.37 | t0dbld|work | i feel like the tools are failing me |
03:51.19 | t0dbld|work | i turn to fex , make one change and back to bin and i loose all screen it seems maybe even boot i havnt checked logs |
03:56.37 | *** join/#arm-netbook Eigen (~Eigen@S0106b8c75dc9a198.vn.shawcable.net) |
04:07.35 | t0dbld | err jsut not understanding whats going on |
04:34.36 | nikk | t0dbld it's better not to change [working] .bin sections. |
04:35.12 | nikk | only make changes, if something does not work |
04:36.57 | t0dbld | it doesnt work |
04:37.01 | t0dbld | thats what im fixing |
04:37.06 | t0dbld | im getting some where now |
04:37.57 | nikk | LCD? did it work before? find that firmware, take .bin from there, and use it |
04:38.22 | t0dbld | nikk: it never worked with linux this is an android tablet |
04:38.35 | t0dbld | im booting gentoo and ubuntu |
04:38.45 | nikk | so, it's a question of driver, not bin... |
04:38.51 | t0dbld | no its not |
04:39.12 | nikk | bin is just a database of parameters, not more |
04:39.30 | nikk | and driver use it, no? |
04:39.41 | t0dbld | right and by default its only configuring the parameters for hdmi out |
04:39.48 | nikk | ?????? |
04:40.21 | t0dbld | a mele doesnt have an internal screen |
04:40.27 | t0dbld | its set to use hdmi out |
04:40.46 | t0dbld | or you can change it the same way i am to use vga |
04:40.54 | t0dbld | im changing it to use internal |
04:41.38 | nikk | it's al about kernel/driver. in 2.6 it was kernel option to use lcd ly default for exampl |
04:42.00 | t0dbld | thanks for you help nikk |
04:42.09 | t0dbld | im all set over here thanks |
04:42.12 | nikk | bin probably doesn't provide order |
04:42.41 | *** join/#arm-netbook cromo (~cromo@139-255-251-94.net.stream.pl) |
05:42.19 | *** join/#arm-netbook hno (~hno@squid/developer/hno) |
05:54.20 | hipboi__ | t0dbld: there is an English version of the fex config manual |
05:58.37 | t0dbld | haha any idea where i may obtain it hipboi__ ? |
05:58.49 | hipboi__ | i will find it |
05:59.11 | hipboi__ | i remember i have someone translated it |
06:00.33 | t0dbld | that would be sweet |
06:01.02 | t0dbld | i got screen to light up now but got a ways to go and would rather know exactly ratehr than guess :-) |
06:11.40 | hipboi__ | http://dl.dropbox.com/u/65312725/Crane_Fex_Guide%20v0.4%20%EF%BC%88EN%29.pdf |
06:12.09 | hipboi__ | t0dbld: can you add a link on the wiki about the document? |
06:12.27 | rm | Error (404) |
06:12.39 | hipboi__ | uploading |
06:12.41 | hipboi__ | sorry |
06:12.52 | rm | and you have a double width ( in there |
06:15.30 | t0dbld | hipboi__: can you share that so i can download a local copy ? |
06:15.34 | rm | loaded now, but it's kinda cryptic, I am not the target audience most likely :P |
06:15.52 | t0dbld | nm got it |
06:16.01 | hipboi__ | no |
06:16.10 | hipboi__ | it's open |
06:18.39 | *** join/#arm-netbook ZaEarl (~malmrose@173-11-125-162-SFBA.hfc.comcastbusiness.net) |
06:20.23 | DonkeyHotei | ZaEarl: http://dl.dropbox.com/u/65312725/Crane_Fex_Guide%20v0.4%20%EF%BC%88EN%29.pdf |
06:21.16 | ZaEarl | very nice |
06:28.50 | nikk | even better view it on google https://docs.google.com/open?id=0B8j-5CDP-TreVXRZaXRBbzRVajg |
06:40.40 | nikk | but it's not complete reference ;-) for example, i have [target] section with option "power_start" = 1 ;-) |
06:42.38 | nikk | there is nothing about it on page 6 |
06:43.10 | *** join/#arm-netbook eflatun (~eflatun@46.1.136.80) |
06:45.40 | *** join/#arm-netbook rell01 (~chatzilla@p5B07809D.dip0.t-ipconnect.de) |
06:47.24 | hno | t0dbld, do you have a working android image for your device? If you do then there is a system.bin in there,.You can reverse it to fex with mnemoc tools. |
06:48.56 | t0dbld | hno: yeah thats what i have been doing |
06:49.04 | t0dbld | i got internal screen to light up now |
06:49.28 | DonkeyHotei | you don't need to reverse it, because you can unimg the livesuit rom |
06:49.31 | hno | if you have a working system.bin for android then the same should work fine for Linux. |
06:50.00 | hno | DonkeyHotei, to change anything you need to reverse it. |
06:50.53 | DonkeyHotei | no, just grab the unimg'ed fex, edit, and "script" |
06:52.05 | hno | DonkeyHotei, if the Android image contains the fex yes. Many do not. |
06:52.48 | hno | don't know why they started to put the fex in android images lately. It's not needed, but very nice. |
06:53.53 | t0dbld | got board with ubuntu ... and im out of sdcards ... time to see if i can get plasma active to boot :-) |
06:54.56 | DonkeyHotei | nand boot would be real progress, and nand boot with brom -> straight to uboot is the holy grail |
06:55.44 | t0dbld | DonkeyHotei: than why dont you get on it ? so far all i here out of you is you telling other people shit , I have yet to here you contribute one useful thought or idea to anyone , no one is stoping you from working or contributing |
06:56.10 | t0dbld | isntead you argue with people and tell them what they should do |
06:56.18 | t0dbld | well have at it hoss |
06:57.24 | DonkeyHotei | if that's truly your perception, nothing i ever do will change it |
06:57.57 | t0dbld | sure it will contribute something , do one of the things you keep telling everyone to do |
06:58.06 | t0dbld | if you are such an expert on it than lead by example |
06:58.16 | t0dbld | otherwise enjoy the show |
07:08.43 | t0dbld | goign to bed good night all |
07:09.14 | hno | relax everyone. |
07:10.25 | hno | asks t0dbld|sleep to use IRC away status rather than nick changes. |
07:12.51 | hno | DonkeyHotei, getting NAND support should mostly be a matter of retrofitting the NAND support code. Have not looked at it yet with USB being first priority. |
07:13.08 | hno | you are very welcome to dig in and get your hands dirty if you like. |
07:13.20 | DonkeyHotei | thanx |
07:13.32 | DonkeyHotei | usb meaning usb boot? |
07:14.08 | hno | usb meaning USB console. And maybe after that USB host support if it's not too far off. |
07:15.02 | DonkeyHotei | ah |
07:15.39 | DonkeyHotei | but serial console, if i could get it to work, should suffice for the time being |
07:16.53 | hno | yes, serial console is a lot better, but usb console is much easier to get going. |
07:17.06 | hno | to the user.. in software serial is much easier. |
07:18.41 | DonkeyHotei | yes, well, it's all relative |
07:21.06 | hno | DonkeyHotei, what device do you have? |
07:22.14 | DonkeyHotei | ZaTab, and not in front of me; i was working on it at the ZaReason shop, and i can again when i am next there, which can't be until at least later this week |
07:25.01 | hno | hopes the ZaTab have UART & JTAG pads internally. |
07:34.23 | DonkeyHotei | probably, but not practical to use for this |
07:34.41 | DonkeyHotei | the sd interface is the best bet for now |
07:37.16 | hno | DonkeyHotei, yes, but hard if you also need to use an sdcard. |
07:38.18 | DonkeyHotei | yes, but we already know sd boot works, so we don't really need serial as badly when booting from sd |
07:38.49 | DonkeyHotei | it would be pretty useful for nand |
07:43.22 | hno | Looks to be the same device as Visture V3 / ELIJAH U9 Phantom / ... |
07:43.49 | *** join/#arm-netbook eflatun_ (~eflatun@95.173.235.208) |
07:46.03 | hno | DonkeyHotei, working on initial NAND support is a bit difficult without it. |
07:47.33 | *** join/#arm-netbook Gumboot (~gumboot@rev.bovine.muck.net.nz) |
07:49.00 | DonkeyHotei | that's what i've been saying all along |
07:57.21 | CIA-91 | rhombus-tech: Francois-Henry master * r20030a245e42 /allwinner_a10/orders/FrancoisM__95__.mdwn: |
08:04.57 | *** join/#arm-netbook hipboi (~hipboi@221.4.213.95) |
08:27.11 | *** join/#arm-netbook eflatun_ (~eflatun@95.173.235.208) |
08:55.27 | *** join/#arm-netbook hipboi (~hipboi@221.4.213.95) |
09:10.07 | naobsd | does anyone try libstagefright interface in ffmpeg on Linux on A10 device? |
09:30.34 | *** join/#arm-netbook eflatun (~eflatun@95.173.235.208) |
09:31.34 | Turl | hmm |
09:31.54 | Turl | so the source kernel we have uses mali r2p3 |
09:32.08 | Turl | my android blobs are r2p4 |
09:44.48 | mnemoc | Turl: does the patch version break compatibility? |
09:47.51 | Turl | mnemoc: if kernel and userland mismatch version it doesn't work afaik |
09:48.36 | mnemoc | is there a mali version with can use in kernel, android AND xorg at the same time? |
09:49.00 | Turl | as long as you get all the pieces on the same version |
09:49.02 | Turl | sure |
09:49.03 | Turl | :P |
09:49.14 | mnemoc | considering what you have seen available |
09:49.49 | Turl | well my tab has r2p4 on its stock android |
09:49.55 | Turl | probably the old ubuntu image has it too |
09:50.20 | mnemoc | if you confirm we can set the kernel version to r2p4 and make everyone happy |
09:50.46 | Turl | I heard the new ubuntu has r2p4-0rel1 :P |
09:51.03 | mnemoc | and that counts as "mismatch"> |
09:51.03 | mnemoc | ? |
09:51.13 | Turl | I don't know tbh |
09:51.19 | mnemoc | zenitraM: ? |
09:52.01 | Turl | btw batt life on this thing is nice |
09:52.07 | Turl | I went to sleep with 29% |
09:52.11 | Turl | wake with the same :P |
09:52.31 | mnemoc | it has an awesome standby mode indeed |
09:53.45 | mnemoc | the onda vi10 elite my wife uses can be a week on the shelf without losing battery |
09:54.33 | Turl | video decoding is decent too |
09:54.52 | Turl | I watched like 1:30 of video on max brightness |
09:55.13 | Turl | and I was ~47% |
09:55.22 | Turl | so 18%, not bad |
09:55.35 | Turl | the one thing I hate is charging though :) |
09:55.53 | Turl | charging over USB is no good :P |
09:56.20 | mnemoc | 0.5mA? |
09:56.22 | Turl | a bit more |
09:56.30 | Turl | 800mA for sure, maybe more |
09:56.37 | Turl | it's a motorola wall charger |
09:57.03 | Turl | 0.5mA won't get you much far btw :P |
09:57.16 | mnemoc | 0.5A :p |
09:57.41 | Turl | it brought a wall charger, I think 2A |
09:57.47 | Turl | with a round plug for a special charging port |
09:58.01 | Turl | but it has a US plug, need to get an adapter |
09:58.21 | mnemoc | my question was more on the line, are you sure the OTG of your tablet can take more than 500mA ? |
09:59.10 | mnemoc | I don't know much about electronics but I assume they have to do some extra stuff to allow that |
09:59.39 | mnemoc | if not, why adding a DC connector at all |
10:00.46 | nikk | my new CM9 fail on 3.0.31+ kernel... It was so earlier, but after replacing libs in egl by ones from stock fw or 4.0.3 source (don't remember which ones) it was working... Now that trick doesn't work... http://pastebin.com/AM1AuikK |
10:01.12 | *** join/#arm-netbook rell01 (~chatzilla@p5B07809D.dip0.t-ipconnect.de) |
10:02.08 | Turl | mnemoc: haven't checked really |
10:02.08 | mnemoc | nikk: uhm... the v2 branch doesn't have any extra change on that area unless it comes from the kernel-common merges... which I doubt |
10:02.18 | Turl | mnemoc: the tab has a port for host and another for 'client' btw :P |
10:02.36 | nikk | mnemoc it's not kernel issue. it was working on 3.0.31+ two days ago ;-) |
10:02.59 | nikk | mnemoc maybe, it's .config related... |
10:03.28 | Turl | gralloc failing nikk? W/libEGL (10203): 0: eglInitialize(0x1) failed (EGL_BAD_ALLOC) |
10:03.47 | nikk | Turl yup. doesn't work ) |
10:05.19 | nikk | 1st time it was the same situation, but zenitraM suggested me replacing libs in egl with stock ones, and after CM9 booted on 3.0.31 |
10:09.10 | Turl | nikk: can you strings *|grep r2p ? |
10:09.15 | Turl | on the stock libs |
10:11.26 | nikk | 1.4 Linux-r2p4-01dev0 |
10:13.27 | Turl | and they worked with the v2 branch? weird |
10:13.50 | mnemoc | so the 'p' is not that importnatn? |
10:15.33 | nikk | Turl: http://pastebin.com/uUZe3kvV |
10:16.38 | nikk | Turl almost the same info in libs from common section of your git |
10:18.58 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
10:19.10 | nikk | Turl, i'm not sure, which one I used that time, maybe it was libs from 4.0.3 source: http://pastebin.com/9iZ2ANQy |
10:19.12 | mnemoc | wb gimli |
10:19.33 | gimli | :) |
10:19.55 | gimli | a few hours left and then off o linuxtag in berlin :D |
10:24.53 | hipboi | mnemoc: do you see my mail? |
10:26.04 | naobsd | nikk: please make sure all mali related files are taken from same source/rom |
10:27.07 | naobsd | mali.ko ump.ko /system/lib/egl/lib*mali.so /system/lib/hw/gralloc.sun4i.so /system/lib/lib{Mali,UMP}.so |
10:27.18 | nikk | naobsd drivers compiled from source, there only r2p1 r2p2 r2p3 versions |
10:27.43 | naobsd | ah, sure, modversion must be different... :( |
10:28.33 | naobsd | did you try mali source from ARM? |
10:28.35 | nikk | but they loaded fine... and some combination was working on 3.031+ and CM9 |
10:29.17 | mnemoc | hipboi: yes, and replied :) |
10:29.25 | naobsd | hmm |
10:29.53 | naobsd | where is mali source in under lichee... |
10:29.59 | nikk | naobsd not yet, iin modules |
10:30.05 | mnemoc | hipboi: the difference between this one and the one from 2.6.36 is quite minimal... https://github.com/amery/linux-allwinner/commit/90ee8bd51532912a9bd8771d5250e1c708f435cb |
10:30.09 | naobsd | I see |
10:30.23 | hipboi | mnemoc: ok |
10:30.35 | mnemoc | hipboi: but I'm really thankful for sharing it |
10:30.57 | hipboi | : ) |
10:31.17 | mnemoc | hipboi: the feeling of been working in old code is ugly, this confirmed that at least on sata we are not behind :) |
10:32.37 | mnemoc | hipboi: considering the sun5i code already leaked, and if the problem with sharing is because of the human and bw resources involved, I would gladly do as middleman if you decide I can get read access to your git :) |
10:33.05 | naobsd | nikk: I compared DX910-SW-99002-r2p2-03rel0 in lichee and source from ARM. arch-ca8-virtex820-m400-1 need to be copied but others are same. |
10:33.06 | hipboi | hi |
10:33.23 | hipboi | i think it will be released to your friend |
10:33.37 | hipboi | who do the A13 project |
10:34.01 | mnemoc | tsvetan, from olimex |
10:34.19 | hipboi | yes |
10:34.45 | naobsd | try r2p4-02rel0 or rel1 to use r2p4-01dev0 |
10:35.03 | mnemoc | what kernel version are you working there currently? still 3.0 or got an insight of what android 5.0 will use? |
10:36.06 | naobsd | if I can get some time, I'll try myself too |
10:37.04 | hipboi | 3.3, currently |
10:42.56 | Turl | mnemoc: kernel version is mostly irrelevant for android |
10:44.30 | Turl | zomg :D |
10:44.35 | Turl | got it booting with my kernel :D |
10:44.44 | mnemoc | Turl: what changed? |
10:45.28 | Turl | mnemoc: the repackage method |
10:45.30 | Turl | lol |
10:47.36 | mnemoc | doh |
10:50.18 | Turl | <4>[ 279.280000] mali: disagrees about version of symbol module_layout |
10:50.27 | Turl | o.O I built the mali module fresh |
10:57.07 | Turl | hm now it loads |
10:57.08 | Turl | <4>[ 58.730000] Mali: ERR: /root/zatab/linux-allwinner/modules/mali/DX910-SW-99002-r2p4-02rel0/src/devicedrv/mali/common/mali_kernel_GP2.c |
10:57.09 | Turl | <4>[ 58.800000] maligp_core_version_legal() 441 |
10:57.09 | Turl | <4>[ 58.800000] Error: reading this from maligp version register: 0x0 |
11:01.37 | mnemoc | scratches his head |
11:04.34 | hno | The Ubuntu image released by hipboi have r2p4 mali libraries, hopefully for sun4i. The current lichee BSP package we have access to have r2p3 libraries. |
11:05.32 | Turl | hno: android roms have r2p4 too |
11:07.17 | nikk | Turl, so, I need recompile kernel modules with DX910-SW-99002-r2p4-02rel1 and it will work? |
11:07.18 | mnemoc | Turl: but from your paste their aren't happy with what we have |
11:07.51 | nikk | Turl, or i's CM9 source code issue? |
11:08.32 | nikk | gralloc must be updated too? |
11:09.53 | nikk | cause they have kernel modules, gralloc for android and UMP userspace drivers: http://www.malideveloper.com/developer-resources/drivers/open-source-mali-gpus-android-gralloc-module.php |
11:10.39 | Turl | <PROTECTED> |
11:10.40 | Turl | hm |
11:11.31 | mnemoc | are you thinking in lying? :P |
11:11.42 | hno | it failed to access mali registers? |
11:12.25 | hno | that check should be if the hardware matches the driver. faking it is most likely a bad idea. |
11:12.56 | hno | i.e. that you use MALI400 driver on MALI400 and not MALI200. |
11:13.15 | mnemoc | but they seem to want even the exact patch level... |
11:13.24 | hno | in this register? |
11:13.42 | mnemoc | to let the userspace lib work |
11:14.16 | hno | they want matching kernel driver & userspace. But not down to patchlevel. They have major versin number in the kernel driver that is verified. |
11:16.02 | mnemoc | ah, ok. |
11:19.47 | naobsd | Turl: did you use arch-ca8-virtex820-m400-1? |
11:19.51 | Turl | yeah |
11:20.17 | mnemoc | in that case, can we get a single version moved from modules/mali to drivers/gpu/ ? |
11:25.14 | mnemoc | nice new video, http://www.youtube.com/watch?v=KZVgKu6v808 .. the engineerguy explains the accelerometer |
11:30.05 | nikk | naobsd you were right, wi-fi works with manual loaded module on lib, which try to load realtek, but it was needed to replace wpa_supplicant to prebuilt one ;-) |
11:31.22 | nikk | naobsd i have weird behaviour with sun4i-ir module loaded, CM9 think it have hard keyboard and soft one didn't pop-up ;-) unloading IR module fix that |
11:32.29 | *** join/#arm-netbook hp_ (~kvirc@118-168-205-168.dynamic.hinet.net) |
11:37.35 | naobsd | hmm I can use both nano and realtek with same wpa_supplicant...ah I guess your realtek module support wifi-direct? |
11:38.09 | naobsd | I didn't try wifi-direct capable one. I guess it needs wpa_supplicant_8 |
11:38.51 | naobsd | and no idea about IR |
11:39.12 | naobsd | it was worked on my CM9 as a keyboard |
11:39.30 | nikk | naobsd yup, it was wpa_supplicant_8 with wi-fi direct enabled :) |
11:39.41 | naobsd | keyboard? as same as other input device like as buttons |
11:39.50 | nikk | and yes, ir works as keyboard ) |
11:40.17 | nikk | and even types numbers in imput fields |
11:40.23 | nikk | input |
11:40.57 | nikk | and CM9 think it's hard keyboard, and disables onscreen one |
11:41.20 | naobsd | it's strange "think it's hard keyboard" |
11:41.35 | nikk | and i can switch this in keyboard swither back to soft, but only in tablet interface |
11:41.53 | Turl | mnemoc: I managed to kill the kernel with a module trying to print debug info :) |
11:41.54 | naobsd | I didn't try new kernel source and quarx's CM9 yet ;) |
11:42.01 | Turl | this thing seems to lack a watchdog :P |
11:42.18 | mnemoc | ouch |
11:42.28 | naobsd | btw which device has nano wifi? only NOVO7 Advanced? ;) |
11:43.00 | nikk | probably so, and it works only on 20 mbit ) |
11:43.11 | naobsd | I want to throw away nano from window! |
11:43.29 | nikk | thought about usb one on rtl8192cu / su ) |
11:43.53 | naobsd | I can use 8192cu USB dongle |
11:43.58 | naobsd | fortunately ;) |
11:44.26 | nikk | i finded one external dongle, but it's on su version, different driver i think |
11:44.55 | naobsd | but no USB typeA port on NOVO7 Adv :( |
11:45.20 | naobsd | not nice to use convert cable ;) |
11:45.56 | nikk | yup. in my case it doesn't matter as i plan to put it in the car as permanent install ))) |
11:46.19 | naobsd | i'll leave here soon sorry |
11:46.28 | naobsd | cya |
11:46.37 | nikk | cya |
11:47.05 | hno | The A10 do have a watchdog of some kind. But normally not used. |
11:47.49 | hno | Turl ^ |
11:49.17 | hno | noticed it's existence in the u-boot reset code where it's used to reset the CPU. |
11:50.07 | Turl | hno: :) |
11:55.37 | lundman | :) |
11:57.36 | nikk | hehe, 4 of 6 parts in "modules" failed to build on mac ))) |
12:00.56 | Turl | woot it works :D |
12:01.01 | Turl | I was missing the pmu stuff |
12:02.30 | hno | Found the watchdog in the kernel sources as well, part of the standby timer. Briefly used during resume from standby in case dram training fails. |
12:02.37 | nikk | Turl what's work exactly? |
12:02.46 | Turl | nikk: screen with new mali stuff |
12:02.48 | Turl | :D |
12:03.02 | nikk | Turl I also recompiled mine ;-) not loaded yet |
12:03.04 | hno | Turl, android or Linux? |
12:03.26 | nikk | gralloc old one? |
12:03.34 | nikk | only driver changed? |
12:03.42 | Turl | hno: android |
12:03.47 | Turl | nikk: source built gralloc |
12:04.14 | mnemoc | Turl: anything to push yet? :) |
12:04.23 | nikk | and kernel part? new? version of gralloc? |
12:06.04 | Turl | mnemoc: I'll push updated mali in a sec |
12:06.31 | mnemoc | \p/ |
12:07.01 | nikk | DX910-SW-99005-r2p4-02rel1 and DX910-SW-99002-r2p4-02rel1 ? |
12:08.20 | Turl | beware big commit :) https://github.com/allwinner-dev-team/linux-allwinner/commit/4a39143017011eb064aded134e49ca68e6875bc9.patch |
12:08.22 | Turl | mnemoc: ^ |
12:12.16 | nikk | hm, i made a diff with CM9 gralloc and DX910-SW-99005-r2p4-02rel1, difference only in makefile ))) |
12:12.48 | nikk | so it's definitely r2p4-02rel1 was there already |
12:12.53 | Turl | yeah we took the gralloc source from there |
12:14.53 | mnemoc | Turl: no deleting of old versions? |
12:15.11 | Turl | mnemoc: I'd rather keep commits atomic |
12:16.02 | Turl | what's a typical touchscreen on this things? |
12:20.56 | mnemoc | ft5x and goodix? |
12:24.31 | mnemoc | we may try to mainline those and axp now for 3.5 |
12:25.24 | naobsd | I can get some more time now :) |
12:25.50 | naobsd | Turl: what's PMU? I can see "Mali-400 PMU" under arch -ca8-virtex820-m400-1 |
12:27.15 | nikk | it's mali 400 4sure ) |
12:27.57 | naobsd | and it seems USING_MALI_PMU is 1 on my tree... |
12:28.07 | Turl | power management unit is my guess |
12:28.29 | naobsd | sorry, what I asked is what you missed |
12:28.37 | Turl | oh |
12:28.47 | Turl | naobsd: the "platform" stuff |
12:28.55 | Turl | compare the two makefiles |
12:31.26 | naobsd | ah |
12:35.15 | naobsd | I noticed there is only "default" under mali/platform in r2p4-02 |
12:36.26 | *** join/#arm-netbook cromo (~cromo@139-255-251-94.net.stream.pl) |
12:36.50 | Turl | yeah |
12:40.25 | DonkeyHotei | [Thu 2012-05-17 01:08:51 AM PDT] <hipboi> after i commit tonight, you can input in u-boot console <-------- still no commit |
12:40.36 | nikk | hm, not in my case. recompiled kernel and modules with DX910-SW-99002-r2p4-02rel1, but same: W/libEGL ( 249): 0: eglInitialize(0x1) failed (EGL_BAD_ALLOC) |
12:44.40 | nikk | heh, it doesn't understand PLATFORM? |
12:45.55 | nikk | it's from strings output on loaded one... |
12:48.02 | naobsd | hmm. now I compared r2p3-01rel0 ARM <-> allwinner |
12:48.09 | naobsd | I can see some differences... |
12:48.41 | naobsd | (I was compared r2p2 but it seems it might not be good, sorry :) |
12:49.00 | mnemoc | you have to compare the exact same release |
12:49.40 | naobsd | what I said is : I compared r2p2 ARM and r2p2 allwinner and now I compared r2p3 ARM and r2p3 allwinner |
12:50.18 | Turl | mnemoc: so I've got Goodix-TS-board-3 hmm |
12:50.30 | naobsd | why I compared r2p2 is : I alread have r2p2 ARM source for my AML8726-M device |
12:50.47 | naobsd | and I was lazy |
12:50.50 | Turl | mnemoc: no sauce for Goodix-TS-board-3? :( |
12:51.39 | Turl | "goodix_touch_3F.ko" |
12:51.44 | *** join/#arm-netbook hipboi_ (~hipboi@123.89.119.201) |
12:52.16 | Turl | hey hipboi_ |
12:53.15 | DonkeyHotei | [Thu 2012-05-17 01:08:51 AM PDT] <hipboi> after i commit tonight, you can input in u-boot console <-------- still no commit |
12:53.17 | DonkeyHotei | hipboi: ^ |
12:53.22 | DonkeyHotei | hipboi_: ^ |
12:53.24 | nikk | Turl: what I'm doing wrong? compiled and loaded 3.0.31 with r2p4, but now mali.ko wont load: bad address ;-) |
12:53.43 | Turl | check dmesg nikk |
12:54.14 | nikk | http://pastebin.com/Vn5iaqGs |
12:54.54 | Turl | yeah nikk you're missing the pmu stuff |
12:54.55 | nikk | but ump loaded fine ))) |
12:55.06 | Turl | ump != pmu |
12:55.13 | Turl | and yeah ump loads fine regardless |
12:55.14 | Turl | :P |
12:55.33 | naobsd | TARGET_PLATFORM=mali400-pmu |
12:56.43 | naobsd | for compiling mali |
12:56.49 | nikk | Turl: aout goodix, mine (v1) also doesn't work on 3.0.31 with script.bin from 4.0.3. Driver requires another section for it. |
13:00.35 | nikk | naobsd: TARGET_PLATFORM=mali400-pmu USING_MMU=1 USING_UMP=1 USING_PMM=1 BUILD=release CONFIG=ca8-virtex820-m400-1 ... yup? |
13:02.51 | naobsd | nikk: I think so but not tried yet myself |
13:04.27 | naobsd | hmm what's this... _mali_osk_malloc -> _mali_osk_valloc / _mali_osk_free -> _mali_osk_vfree in mali_ukk_mem.c |
13:06.41 | Turl | hipboi_: any chance you got drivers/input/touchscreen/goodix_touch_3F.mod.c ? :P |
13:07.19 | Turl | license=GPL |
13:07.20 | Turl | description=Goodix Touchscreen Driver |
13:07.20 | Turl | depends=ft5x_ts |
13:07.25 | Turl | weird depends |
13:07.28 | Turl | but it's GPL :) |
13:07.42 | naobsd | nikk: sorry. platform/default/mali_platform.c is different between ARM and Allwinner |
13:08.35 | *** join/#arm-netbook eFfeM_work (~frans@D4B268BA.static.ziggozakelijk.nl) |
13:11.54 | naobsd | nikk: please use modules/mali/DX910-SW-99002-r2p3-01rel0/src/devicedrv/mali/platform/default/mali_platform.c in allwinner source |
13:12.40 | naobsd | I checked mali.ko in NOVO7 Adv1 rom and I noticed mali_platform_power_mode_change() is not empty |
13:13.19 | naobsd | platform/mali400-pmu/mali_platform.c has emtpy mali_platform_power_mode_change() so it's not the file compiled |
13:13.50 | naobsd | s/compiled/actually used/ |
13:14.36 | nikk | why not PMU version? |
13:15.11 | naobsd | "default" and "mali400-pmu" are just a directory name. |
13:15.26 | naobsd | codes in .c is important ;) |
13:16.26 | naobsd | default/mali_platform.c in ARM r2p4 is empty, but in Allwinner r2p3 is not empty |
13:17.45 | mnemoc | can that part be moved to arch/arn/mach-sun4i ? |
13:17.51 | nikk | but source differ too: mach/sys_config.h against mach/script_v2.h and second one have power control ) |
13:18.22 | naobsd | hmm |
13:18.36 | Turl | fuuu touchscreen |
13:18.47 | Turl | y u no have source D: |
13:19.01 | naobsd | I don't understand detail yet so mali400-pmu may work... it also has some code in mali_platform_init( |
13:20.37 | naobsd | mmm where is script_v2.h ... |
13:20.43 | nikk | but in general they are look the same |
13:21.02 | mnemoc | arch/arm/mach-sun4i/include/mach |
13:21.53 | nikk | in DX910-SW-99002-r2p3-01rel0/src/devicedrv/mali/platform/mali400-pmu/mali_platform.c |
13:22.07 | naobsd | I can't find script_v2.h in allwinner leaded 3.0.8 |
13:22.11 | naobsd | leaked |
13:22.22 | Turl | naobsd: it isn't there |
13:22.32 | Turl | kill those two _v2 headers |
13:22.34 | nikk | no, it's in amery's git ))) |
13:22.35 | mnemoc | script_v2.h is lichee-2.6.36 |
13:22.50 | mnemoc | sys_config.h is lichee-3.0.8 |
13:22.50 | nikk | merged to 3.0.31 ))) |
13:23.02 | mnemoc | they cleaned that that part |
13:23.11 | mnemoc | nikk: eh? it shouldn't be in v2 |
13:23.14 | nikk | so, definitely we must use default not pmu one ) |
13:23.28 | naobsd | nikk: mali_platform_power_mode_change() in default/mali_platform.c is NOT empty |
13:24.39 | naobsd | both _init() are similar so it will work after init |
13:25.14 | nikk | this one still uses script_v2 ) https://github.com/amery/linux-allwinner/blob/allwinner-v3.0-android-v2/modules/mali/DX910-SW-99002-r2p3-01rel0/src/devicedrv/mali/platform/mali400-pmu/mali_platform.c |
13:25.31 | mnemoc | uh |
13:25.59 | mnemoc | but that's not the one used by the build_crane script, is it? |
13:26.16 | naobsd | but mali400-mtu one may not work after suspend...or it may work because mali400-pmu one may not sleep ;) |
13:27.01 | naobsd | TARGET_PLATFORM=mali400-pmu is not defined so I think default is compiled in allwinner build |
13:27.10 | nikk | yup |
13:27.53 | mnemoc | does mali provide a method to do that platform thing in arch/arm/mach-sun4i/ instead of inside the driver dir? |
13:28.46 | naobsd | actually mali/ump can be compiled outside of kernel source tree |
13:34.55 | mnemoc | that doesn't mean it's desirable to keep it that way |
13:35.21 | mnemoc | unless we really can't find a version usable for android and linux people |
13:37.07 | nikk | mnemoc current release must work for all of us, i hope ;) |
13:37.19 | mnemoc | :D |
13:40.30 | nikk | Turl, i think specific modules loading must go from ramdisk to /system/etc/init.d/ in CM9, so it can be easily customizable |
13:41.16 | nikk | system/etc/init.d # ls |
13:41.16 | nikk | 00banner 03nanowifi 05bluetooth |
13:41.16 | nikk | 03accelerometer 03touchkeys 07gps |
13:41.16 | nikk | 03irda 04dvb 90userinit |
13:44.45 | *** join/#arm-netbook kaspter (~kaspter@116.237.61.223) |
13:47.18 | mnemoc | uh, iMX6 devkit on sale.... http://boundarydevices.com/products-2/sabre-lite-imx6-sbc/ |
13:47.56 | mnemoc | I thought it was utterly postponed |
13:48.32 | Turl | ha |
13:48.36 | Turl | quad core A9 |
13:48.54 | mnemoc | and sata |
13:49.01 | Turl | and 300US :P |
13:50.11 | mnemoc | I would gladly spend the same 300US in a sun6i devkit ;-) |
13:55.06 | Turl | winnering today, CM9 built boot.img and it works :) |
13:55.07 | nikk | hm, recompiled, now it's loading with new release, but still no display: http://pastebin.com/n7qaBN7R |
13:55.21 | A0Sheds | I'm waiting to see how the Freescale imx6 parts will be available this fall. By the looks of things they will keep their prices artificially high |
13:56.36 | A0Sheds | they delayed launch from June until August for the imx6, so I'm not sure what those board have in them for parts since they went back for another repsin on the silicone |
13:57.18 | A0Sheds | the boundary devices imx6 boards have the buggy parts, but they don't mention that in the data sheet |
13:58.16 | A0Sheds | there has been an imx6 dev board out for the past few months available through Arrow |
13:59.20 | A0Sheds | it was $220 last month, it looks like the price jumped to $300 now |
13:59.47 | mnemoc | :( |
14:00.03 | A0Sheds | NIT6X i.MX6 QUAD CORE DEVELOPMENT KIT |
14:00.33 | A0Sheds | the funny thing is that Arrow didn't have any part number in their system for the boards |
14:00.47 | A0Sheds | NIT6XW AD CORE DEVELOPMENT KIT - WIFI |
14:00.50 | Turl | I want to hit someone :P kernel works but I don't have touch driver :( |
14:01.04 | A0Sheds | NIT6X1024X600 AY WITH CAPACITIVE MULTI-TOUCH |
14:01.23 | A0Sheds | is what they gave me |
14:01.55 | nikk | Turl mine don't ( |
14:08.24 | nikk | Turl loaded all libs from youth "common" device tree, no bad words, but: http://pastebin.com/MwyN1pnr ... Do you also have "D/DisplayDispatcher( 227): hw_get display module Failed!" |
14:09.49 | Turl | nikk: I added display literally two minutes ago on common |
14:09.50 | Turl | :P |
14:09.59 | Turl | so build it haha |
14:10.13 | *** join/#arm-netbook QingPei (~qingpei@123.118.106.179) |
14:10.32 | nikk | which one? it's all in frameworks, no? |
14:11.00 | Turl | nikk: https://github.com/allwinner-dev-team/android_device_allwinner_common/tree/master/display |
14:11.05 | naobsd | nikk: about "_mali_surface_destroy_callback" error, I guess you used binary blobs in leaked android 4.0.3 |
14:12.19 | naobsd | I can't find any reference to it in any released ROM for real tablet |
14:13.02 | nikk | naobsd yes, it was previous try, they "somehow: worked before, now i switched to others |
14:15.59 | nikk | heh, GitError: allwinner-dev-team/android_device_allwinner_common rev-list ('^639428d41eaa7a391987a70cd9dbb68fef9dd2a3', 'HEAD', '--'): fatal: bad object 639428d41eaa7a391987a70cd9dbb68fef9dd2a3 |
14:16.06 | nikk | what does it mean? ) |
14:16.31 | nikk | it's repo sync |
14:17.55 | Turl | idk :/ |
14:27.15 | hp_ | argg.. when i 7z x mele image.. my laptop gets soo hot.. it crashes |
14:27.27 | mnemoc | o_O |
14:27.41 | mnemoc | hp laptop? :) |
14:27.43 | hp_ | i'll have to try my virtual machine :) |
14:27.47 | hp_ | mnemoc: dell :) |
14:27.53 | mnemoc | :) |
14:28.07 | Turl | acer ftw |
14:28.22 | Turl | I can poke memory to make my fans spin unthrottled :D |
14:28.49 | hp_ | Turl good thing i have a virtual machine to unpack the image :) |
14:29.12 | traeak | hp_: laptops are pretty poorly designed in general. core i5/i7 should never be in a laptop |
14:29.13 | hp_ | mnemoc: just wondering did you see the android 4.0 release :) |
14:29.28 | hp_ | traeak: that is probably true :) |
14:29.29 | nikk | my mac goes hot even with 8GB RAM ) |
14:29.30 | traeak | hp_: does it matter ? :-p |
14:29.56 | traeak | hp_: i usualy crash laptops becuase the ram overheats...the cpu throttles at 99C, the ram rams to 130C and crashes |
14:30.05 | traeak | ramps even |
14:30.22 | mnemoc | hp_: I got both tarballs from qware if you mean that |
14:30.25 | hp_ | traeak: i think my laptop with linux is soo old i need to clean the fan |
14:30.28 | traeak | alienware: crap, toshiba: crap, dell: crap |
14:30.31 | nikk | device/allwinner/common/: leaving master; does not track upstream |
14:30.32 | nikk | device/allwinner/zatab/: leaving ics; does not track upstream |
14:30.32 | nikk | error: device/allwinner/common/: allwinner-dev-team/android_device_allwinner_common checkout 639428d41eaa7a391987a70cd9dbb68fef9dd2a3 |
14:30.40 | hp_ | mnemoc: just wondering if you got it booted |
14:30.44 | nikk | what's all this? |
14:31.02 | traeak | i'm lazy waiting for a video driver to drop in my lap |
14:31.10 | mnemoc | hp_: I'm not really interested in android |
14:31.13 | traeak | actually coding up a storm on work |
14:31.29 | mnemoc | hp_: poke Turl and nikk for that |
14:31.37 | hp_ | mnemoc: and ubuntu? |
14:31.40 | nikk | traeak: which one? |
14:32.06 | mnemoc | hp_: my userspace is ubuntu, yes, but headless |
14:32.20 | nikk | hp_ yes it boots ) even works sometimez ))) |
14:32.28 | hp_ | nikk :) |
14:32.44 | hp_ | mnemoc: i am interested in ubuntu also need an extra desktop :) |
14:32.54 | traeak | nikk: framebuffer would be sufficient |
14:33.16 | nikk | hp_ which your prefer? 4.0.3 chinese or not working CM9? ;-) |
14:33.43 | hp_ | nikk chinese and then set to english? |
14:33.48 | nikk | yup |
14:34.22 | nikk | sometimez very funny translations (to russian) ))) |
14:38.26 | nikk | <PROTECTED> |
14:43.29 | CIA-91 | rhombus-tech: lkcl master * r0d586e539b6b /allwinner_a10/hacking_the_mele_a1000.mdwn: |
14:43.31 | *** join/#arm-netbook okapi (~opk@ip-178-200-55-252.unitymediagroup.de) |
15:02.44 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
15:02.52 | *** join/#arm-netbook cromo (~cromo@139-255-251-94.net.stream.pl) |
15:09.02 | hp_ | nikk cool to hear that the chinese android 4.0 is working hehe :) |
15:11.27 | mnemoc | not a bad idea to make ITX-alike arm boards... but arm11? :( |
15:11.53 | A0Sheds | maybe they want slow |
15:12.39 | traeak | arm11 hehe |
15:12.51 | traeak | arm11 belongs on sodimm sized dealies |
15:13.32 | traeak | and for arm to compete head to head with intel i think they realy need that a15, and need it to be cheap |
15:13.38 | A0Sheds | how many ARM devices can actually be sensibly used for mini-itx? not many |
15:13.40 | traeak | or else mips64 needs to be properly resurrected |
15:14.00 | traeak | i suspect mips doesn't have a clue yet |
15:14.02 | mnemoc | the a10 would fit nicely in a desktopish board |
15:14.29 | traeak | a10 is still too slow for general purpose computing. probably more than fine for POS and very specific applications |
15:14.48 | A0Sheds | a10 has no PCI/PCIe |
15:15.10 | mnemoc | that's a good point |
15:15.12 | traeak | sata IMHO is more important than pci/pci-e |
15:15.15 | mnemoc | sata but not pcie |
15:15.17 | nikk | guys, don't foget the price ))) |
15:15.26 | A0Sheds | if it wasn't for Marvell I'd like to make and ARM + AMD GPU board |
15:15.38 | mnemoc | nikk: don't forget about the monthly power bill |
15:15.49 | traeak | i certainly hope the lima driver kicks ass |
15:16.03 | traeak | i hope it stil applies to the new mali t6xx or whatever it is thye are calling it |
15:16.05 | A0Sheds | I just saying that mini-itx would need a PCI slot, so it could be the same footprint -PCI |
15:16.42 | traeak | mini-itx would be a mistake for ARM stuff i think. mini-itx is dinosaur from x86 pc age |
15:17.09 | A0Sheds | I'm still searching for an ARM SOC worthy for making board |
15:17.11 | traeak | "how can we jam a bunch of crap together with a 150W power envelope?" |
15:17.41 | A0Sheds | still waiting to hear back from Renesas as well |
15:18.30 | A0Sheds | http://am.renesas.com/applications/automotive/cis/cis_highend/rcar_h1/index.jsp |
15:18.31 | traeak | sadly the dingoo guys are probably representative of the chinese in general. the dingoo took off because it was hackable. the company never really understood why the dingoo succeeded and proceded to trash their advantage |
15:20.00 | A0Sheds | most customers in China we talked to were not too interested in an A10 card, they wanted something more recent with more performance A9 or more cores |
15:20.18 | *** join/#arm-netbook steev_ (u1383@gentoo/developer/steev) |
15:20.43 | traeak | A0Sheds: for what types of applications? just automotive? |
15:20.54 | A0Sheds | granted that a big piece they are missing is the software |
15:21.22 | A0Sheds | they think in terms of Android or Windows only |
15:21.33 | A0Sheds | there needs to be Linux + desktop support |
15:21.55 | traeak | yup another reason they don't get it |
15:21.59 | A0Sheds | traeak, applications for what? |
15:22.25 | traeak | A0Sheds: customers in china, any specific uses? or just in general SOC ? |
15:22.57 | A0Sheds | for tablets, netbooks, etc etc |
15:23.06 | A0Sheds | even for embedded |
15:23.16 | traeak | google did most of the work for android so in that way they can do the least amount of investment with a fairly controlled system |
15:23.58 | traeak | linux is scary because it's way too much like the american wild west |
15:24.08 | A0Sheds | if only Ubuntu didn't break everything they touch it might have come in handy |
15:24.52 | A0Sheds | linux isn't scary so much as that they have no software engineers to support it |
15:25.13 | A0Sheds | the oems just want a turn key package for making a product |
15:25.40 | A0Sheds | schematics, gerbers, BOM, OS image |
15:25.47 | *** join/#arm-netbook Hexxeh (u1532@gateway/web/irccloud.com/x-yysxxhqujzzqqqea) |
15:26.53 | traeak | hardware guys like hardware, software is scary to them |
15:27.03 | traeak | google really is a godsend for them |
15:27.11 | traeak | google android i mean |
15:27.22 | hno | DonkeyHotei, what input problem with u-boot console? |
15:27.26 | mnemoc | that in theory should make them more fond to be free software frienly... |
15:27.40 | traeak | mnemoc: in theory, except they still want competitive advantage |
15:27.58 | A0Sheds | the various manufacturers just have PCB stuffers and injection molding machines, not a software team |
15:28.06 | mnemoc | hardware guys should look at a hardware competitive advantage... |
15:28.21 | traeak | that's an ugly race to the bottom as well though |
15:28.37 | traeak | microsoft set up a pretty unrealistic situation |
15:28.43 | DonkeyHotei | hno: he said he was going to commit the serial input patch, but http://git.hands.com/?p=u-boot.git;a=shortlog;h=refs/heads/lichee-dev is unchanged |
15:28.56 | hno | a patch correcting what? |
15:28.57 | traeak | where software costs for OS, etc is several multiples of the hardware cost |
15:29.23 | A0Sheds | the large OEM"s with x86 products are told by Wintel to not make ARM netbooks |
15:29.25 | traeak | business people salivate over that |
15:30.20 | traeak | like dell/hp puts together some system for 250USD or whatever, then sells you a full windows system that adds 1200USD or whatever to the price tag |
15:30.52 | DonkeyHotei | hno: uboot ignores serial input over sd |
15:31.06 | traeak | serial input over sd ? huh? |
15:31.12 | nikk | A0Sheds ppl need office and IE, not arm netbook ) |
15:31.47 | A0Sheds | and now thanks to corruption once again projects like the $30 tablet in India have stalled |
15:31.57 | hno | DonkeyHotei, sounds as the port is routed wrongly. u-boot do not use script.bin but program the port controller directly. |
15:32.21 | traeak | 30usd is IMHO a bit too aggressive |
15:32.40 | A0Sheds | Winders is cheap in China and bootlegged everywhere |
15:32.41 | DonkeyHotei | hno: this is established. the patch he was going to commit would fix that |
15:33.10 | DonkeyHotei | he had it locally, but never committed it |
15:35.08 | traeak | A0Sheds: that works okay for PCs but for tablets, some laptops, these embedded systems that crap won't fly |
15:35.10 | A0Sheds | http://www.mydigitalfc.com/news/pantel-unveil-low-cost-tablet-sim-slot-187 this one is $65 US |
15:35.19 | traeak | and windows8 on arm isn't going to be piratable |
15:36.06 | traeak | and of course MS refusing to release visual express 2012 with support for anything but "metro" what the hell are they thinking? |
15:37.12 | traeak | A0Sheds: has india actually ever really released any fully integrated electronics products into the world market before? |
15:37.56 | A0Sheds | traeak, they are focusing on their own internal market |
15:38.38 | A0Sheds | and their market is growing larger than the market in the US |
15:38.38 | mnemoc | traeak: I have two notion ink adam tablets |
15:39.09 | traeak | those are made in india? |
15:39.13 | mnemoc | yes |
15:40.08 | traeak | hmm |
15:41.35 | DonkeyHotei | page doesn't load |
15:42.56 | traeak | http://liliputing.com/2012/05/via-apc-a-49-android-computer-with-an-arm11-cpu.html |
15:44.12 | A0Sheds | did VIA buy Wondermedia? |
15:44.31 | traeak | i think they spun them off or something |
15:45.05 | A0Sheds | http://www.wondermedia.com.tw/en/products/platform/soc/wm8950/index.jsp and they pick the arm11 soc |
15:45.16 | traeak | more via cluelessness |
15:45.34 | A0Sheds | yeah, they can't help themselves |
15:46.10 | traeak | and their a9 doesn't have sata or pcie |
15:51.42 | CIA-91 | rhombus-tech: Hermann master * r2abc9a5701be /allwinner_a10/orders/.mdwn: |
15:51.59 | DonkeyHotei | steev_: do you by any chance know anything about goodix_touch_3F? |
15:52.15 | steev_ | DonkeyHotei: zilch actually |
15:52.48 | DonkeyHotei | steev_: thought i'd ask, because i found a forum post addressed to you asking to recompile it |
15:52.59 | steev_ | DonkeyHotei: that isn't me |
15:53.29 | steev_ | sadly i don't have the time, nor inclination to save my nick on every android forum out there |
15:53.35 | mnemoc | there are others called steev too : |
15:53.51 | DonkeyHotei | lol |
15:54.05 | steev_ | i'm steev@gentoo |
15:55.19 | mnemoc | DonkeyHotei: the old kernel had a drivers/input/touchscreen/f3x-ts.c which doesn't exist in the new... might be related? |
15:56.14 | DonkeyHotei | mnemoc: apparently ft5x_ts is a module dependency |
15:57.41 | ZaEarl | That was fast. I thought it would take longer. http://googleblog.blogspot.com/2012/05/weve-acquired-motorola-mobility.html |
15:59.03 | mnemoc | fast? it was like 6M to be authorized |
15:59.52 | hno | DonkeyHotei, odd.. the changes for SD port serial looks fine to me. But maybe also need to change pullup or disable SD function, not sure. |
16:00.26 | mnemoc | hno: nand support in your tree comes before or after usb? |
16:04.59 | hno | mnemoc, after. |
16:09.36 | DonkeyHotei | hno: where? |
16:10.01 | A0Sheds | I bet that the freescale imx6 parts will be really expensive >$50, they won't let the price fall low in order to protect board makers selling $300 boards/modules |
16:10.53 | traeak | A0Sheds: historically i think that's true. only high volume orders |
16:23.44 | A0Sheds | it's funny how VIA thinks their card will compete |
16:23.59 | A0Sheds | it's crap and they see how it also needs software |
16:24.12 | A0Sheds | sorry, they don't see |
16:24.30 | A0Sheds | it's not 2008 |
16:25.29 | CIA-91 | rhombus-tech: phil master * rc2e4ba9ea0a8 /index/ (discussion.html discussion.mdwn): make the index/discussion page use markdown |
16:25.50 | A0Sheds | they didn't even learn from their junk arm11 netbooks |
16:25.57 | mnemoc | gimli: hi, are you subscribed to arm-netbook ML? if not, what should I answer regarding the source availablity of your xbmc port? |
16:27.02 | gimli | i'm not subscribed. as soon i have the HW decoder running and integrated, i'll got to push it to my github repo |
16:27.14 | gimli | until that point nothing will be availabe |
16:27.46 | mnemoc | gimli: thanks :) |
16:28.06 | gimli | if someone wanna talk to me, he can find me here on the irc |
16:28.50 | mnemoc | ok |
16:29.38 | A0Sheds | http://apc.io/about/ |
16:30.31 | traeak | the 13.5W under load looks scary |
16:30.56 | mnemoc | my thinkpad consumes 17W... |
16:31.20 | mnemoc | with ssd and 11.6 display |
16:33.50 | CIA-91 | rhombus-tech: hands master * rc7eb02dd0d11 /index/discussion.mdwn: tidy the markup on index/discussion a bit |
16:35.13 | *** join/#arm-netbook t0dbld|work (~t0dbld@unaffiliated/t0dbld) |
16:35.46 | *** join/#arm-netbook rell01 (~chatzilla@p5B07809D.dip0.t-ipconnect.de) |
16:37.48 | A0Sheds | http://apc.io/blog/2012/05/22/hello-world-2/ |
16:38.46 | A0Sheds | lets see if they take the comment down or actually respond |
16:39.20 | zenitraM | i don't see any comment |
16:39.23 | mnemoc | 0 Comments |
16:39.32 | mnemoc | awaiting moderation probably |
16:40.10 | A0Sheds | Why didnt you make an ARM Coretx A9 board since that would actually be marketable with an actual chance of sucess? You could have used the wn8950 http://www.wondermedia.com.tw/en/products/platform/soc/wm8950/index.jsp and offered 1GB DDR3 for a few $$ more and gotten hundreds of people to help support it with software for no cost to VIA. Why does VIA always miss the boat with x86 and now all over again with ARM? |
16:40.38 | mnemoc | :) |
16:40.51 | traeak | A0Sheds: hopefully it allows some company with a clue to do what's right and make via go out of business |
16:41.07 | A0Sheds | they are like vampires |
16:41.13 | A0Sheds | they never die |
16:41.39 | A0Sheds | they take missteps all the time and still keep on going |
16:41.51 | A0Sheds | they were way ahead with low power x86 |
16:41.59 | traeak | "were" being key |
16:42.23 | mnemoc | I really like the idea of reusing ATX/ITX chassis, but arm11 and no sata??? come on |
16:42.24 | A0Sheds | I tried several times to work with them, they just don't listen |
16:42.43 | A0Sheds | they don't get it |
16:43.28 | A0Sheds | I sat there in meetings with them and I'd go over again and again how to do something |
16:43.45 | t0dbld|work | ok so who wanted to know about the android side of the boot ? :-) |
16:43.53 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
16:43.56 | A0Sheds | they would have their agenda from up high and go out and fail |
16:44.41 | nikk | t0dbld|work: me |
16:44.42 | A0Sheds | granted Intel wasn't above board with their tactics against VIA and AMD |
16:44.57 | A0Sheds | but they insist on selling old slow ARM designs |
16:45.39 | traeak | and wonderful via chrome :-p |
16:46.17 | nikk | Turl don't you have gapps signing problem on CM? |
16:46.17 | t0dbld|work | nikk: well im about to decompile its .bin files ... it is jsut using a uboot a drv ini scripts , couple of axf , |
16:46.37 | Turl | nikk: haven't flashed CM yet |
16:46.52 | Turl | I have no touchscreen ~.~ |
16:47.20 | nikk | Turl use mouse and keyboard ) |
16:47.43 | Turl | nikk: I don't have either :( |
16:47.53 | nikk | heh ) |
16:47.55 | Turl | my mouse is broken :P |
16:48.10 | t0dbld|work | lol hes from Arengitna ... and we havnt shipped those to him ^ |
16:48.10 | t0dbld|work | :-P |
16:48.17 | mnemoc | Turl: mail Tom, he might get you the driver |
16:48.18 | Turl | ha |
16:48.28 | Turl | mnemoc: do you have his email? |
16:48.33 | mnemoc | everything in .ar is broken anyway :p |
16:48.37 | t0dbld|work | turl you delted a driver ? |
16:48.42 | t0dbld|work | my tab is untouched |
16:48.49 | t0dbld|work | so i can grab what you need |
16:48.59 | mnemoc | Turl: mr.hipboi gmail com |
16:49.03 | Turl | t0dbld|work: no, the new kernel lacks it |
16:49.08 | Turl | mnemoc: thanks :) |
16:49.09 | t0dbld|work | oh |
16:49.16 | t0dbld|work | you got your kernel booting ? |
16:49.37 | mnemoc | t0dbld|work: he even has mali running ;-) |
16:50.10 | nikk | start buiding CM from scratch again. Still fighting with GL |
16:50.12 | t0dbld|work | well thats good , and im thinking after a few hours of looking here we should beable to write / boot of of nand |
16:50.17 | Turl | haha my mouse kinda sorta works again :D |
16:50.44 | naobsd | hmm my amlogic android tv box runs CM9 now... |
16:50.44 | t0dbld|work | let me start decompiling some of this |
16:50.56 | Turl | t0dbld|work: can you open your cam |
16:51.08 | Turl | and get me which sensors is it using? |
16:51.14 | Turl | so I can enable the right kernel parts |
16:51.27 | nikk | looks like hacking contest )))) |
16:52.15 | nikk | Turl you can make as many modules as you want, even all ))) |
16:52.38 | Turl | nikk: but you cannot load all of them simultaneously :) |
16:53.23 | nikk | Turl :) but you can choose and apply without mofify ramdisk. thanks init.d |
16:53.28 | naobsd | amlogic kernel stop USB power when suspending... then I can't wakeup tv box with USB mouse ;) |
16:53.56 | nikk | naobsd as A10 one, at least at older kernels |
16:54.07 | Turl | config file has it t0dbld :P found it |
16:54.20 | Turl | nikk: yeah but I want them all built into the kernel :) |
16:54.25 | t0dbld|work | ok |
16:54.32 | t0dbld|work | good i can go back to decompiling than |
16:55.44 | naobsd | nikk: I noticed soft keyboard is disabled if I connect USB keyboard. but softkey can be enabled via keyboard icon on status bar |
16:56.12 | nikk | naobsd yup, but not on phone UI ;-) |
16:56.23 | naobsd | ah, I see |
16:56.30 | nikk | i run on 160dpi and it was no possible to do so |
16:57.42 | naobsd | I added dirty hack to enable tablet ui on 800x480 160dpi (only on SystemUI) |
16:59.21 | naobsd | hmm this box has no power key... I can't get power dialog :( |
17:00.37 | naobsd | my answer: don't mind |
17:00.46 | naobsd | sleepy... |
17:01.09 | *** join/#arm-netbook gimli_ (~gimli@chello062178208178.1.15.univie.teleweb.at) |
17:01.09 | *** join/#arm-netbook gimli_ (~gimli@xbmc/staff/gimli) |
17:01.44 | nikk | naobsd do you have link? |
17:01.53 | *** join/#arm-netbook gimli_ (~gimli@chello062178208178.1.15.univie.teleweb.at) |
17:01.54 | *** join/#arm-netbook gimli_ (~gimli@xbmc/staff/gimli) |
17:02.21 | naobsd | nikk: sorry, what link? |
17:02.33 | nikk | systemui hack |
17:04.12 | naobsd | nikk: https://github.com/naobsd/cm_frameworks_base/compare/ics...ics-naobsd please refer 1st and 2nd change |
17:04.17 | naobsd | I have to say it's not nice at all ;) |
17:05.13 | naobsd | 2nd change has a lot of resource file but they are same as sw600 resource. I noticed symlink is enough for them ;) |
17:05.21 | nikk | naobsd, cool, it needs image resizing also, heh ;-) |
17:05.39 | naobsd | I'll redo 2nd change |
17:06.21 | naobsd | I'll go to bed soon |
17:06.21 | naobsd | 2AM now :( |
17:07.03 | naobsd | good night |
17:08.04 | nikk | nite |
17:09.32 | Turl | email sent :) |
17:11.01 | traeak | interesting |
17:11.26 | traeak | cross compiling the kernel with gcc-4.7 i didn't get the same inlining errors that i got with doing a native compile |
17:11.56 | traeak | i'd really prefer doing a qemu build though...have to screw with setting that crap up |
17:13.58 | traeak | i lied, it still crashes and burns, i didn't compile the modules first time out |
17:14.34 | nikk | it doesn't always work on mac also ;-) |
17:15.02 | mnemoc | traeak: tried just removing the `inline` from the headers? |
17:15.35 | traeak | mnemoc: getting there...my first question is: how does the rest of the kernel deal with what the code is trying to do? |
17:15.48 | traeak | mnemoc: or better question: what's the "blessed way" to fix this? |
17:16.09 | mnemoc | traeak: `static inline` |
17:16.43 | mnemoc | traeak: but those functions are too large anyway... it's very unlikely older compilers are actually inlining them |
17:17.00 | *** join/#arm-netbook guiguid (~guiguid@lns-bzn-48f-62-147-252-138.adsl.proxad.net) |
17:17.37 | mnemoc | traeak: Documentation/SubmittingPatches has some stuff on that |
17:17.44 | traeak | actualy they are one liners in the body |
17:18.11 | mnemoc | traeak: ah, ok. different functions that what I got on the ticket then |
17:18.40 | mnemoc | oneliners go in a .h or *the* .c that uses them as static inline |
17:18.49 | traeak | mnemoc: nah those functions are single lines of code (although i dunno ifthey expand out to many |
17:19.11 | traeak | probably should just go in hte header considering 3 different '.c' files need them |
17:19.12 | t0dbld|work | hmm god damn magic.bin segfaults |
17:19.41 | mnemoc | traeak: yes |
17:19.59 | traeak | what'st eh difference between stuff in drivers/media/video/sun4i_csi/csi1 and drivers/media/video/sun4i_csi/csi0 |
17:20.02 | traeak | note the cs1 vs cs0 |
17:20.37 | mnemoc | those are i2s drivers iirc, no idea what's the diff |
17:21.15 | t0dbld|work | ok got a file thats 512bytes called magic.bin any ideas guys ? |
17:21.43 | mnemoc | a key? |
17:21.58 | t0dbld|work | yeah only thing i can think of |
17:22.03 | traeak | don't worry my attempts at being lazy are failing, not a problem |
17:22.04 | t0dbld|work | guess i can try a hex editor |
17:38.49 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
17:44.29 | traeak | mnemoc: lovely whoever did this csi code hacked it up and copied shit around everywhere |
17:45.22 | mnemoc | :) |
17:45.40 | mnemoc | don't look at the disp drivers, they even wrap (twice) malloc |
17:46.01 | traeak | so basically having not done a lot of kernel development it's a pita to figure out what's the best way to fix this crap |
17:46.18 | traeak | well no ernel development honestly :-p |
17:49.55 | traeak | how would i find out if what i'm hacking on is a WIP or "done" ? |
17:50.29 | mnemoc | it's commercial-quality code :p |
17:53.24 | *** join/#arm-netbook burndtjamb (~rapefruit@CPE-72-131-120-166.wi.res.rr.com) |
17:53.37 | mnemoc | the new drivers they made (sunxi) are much better than the old ones |
17:57.34 | nikk | mnemoc: if i comment all CONFIG_HAS_EARLYSUSPEND if defs in IR driver, it will not be powered of on suspend? |
17:57.56 | nikk | powered off = disabled |
17:58.04 | mnemoc | i think so |
17:58.08 | mnemoc | test :) |
17:58.12 | nikk | ok, thats what i need |
17:58.23 | nikk | i tested it on 2.6 and it worked ;-) |
17:59.14 | mnemoc | :) |
17:59.44 | mnemoc | adding a CONFIG_ would be nicer than commenting them out |
17:59.59 | mnemoc | or even a module_param |
18:01.10 | A0Sheds | comment already deleted |
18:01.14 | t0dbld|work | anyone in here got an arm decompiler ? |
18:01.23 | nikk | mnemoc: i have little patch for sun4i-ir.c, lines 501/502 must be commented. it doesn't affect original IR, but with my creative sb live ir it fixes weird behaviour all events reported twice. original remote still works fine. |
18:02.27 | traeak | mnemoc: what's the best way to drop a patch? |
18:02.49 | traeak | mnemoc: http://pastie.org/3951221 probably not the best way i would guess |
18:03.24 | mnemoc | personally I prefer the good old git-format-patch attached on mail |
18:03.39 | mnemoc | because they come with description and authorship |
18:04.11 | traeak | git-format-patch? ugh (now have to learn more about git hehe) |
18:04.33 | mnemoc | `git format-patch` turns a bunch of commits into a bunch of mails |
18:04.54 | mnemoc | but as mails are only nice if you use mutt :p |
18:05.01 | mnemoc | (not my case) |
18:05.34 | t0dbld|work | so is that a no :-) ? |
18:06.20 | mnemoc | as attachment or hanging from a http they are perfect :p |
18:07.37 | mnemoc | github encourages the fork+pull request model.... I don't like it much because that means there are hundred public repos which aren't really intended for public use... |
18:07.43 | CIA-91 | rhombus-tech: Thomas master * r14cb488aeab9 /allwinner_a10/orders/Thuys.mdwn: |
18:08.21 | mnemoc | traeak: If you give me the authoriship data and text for the commit I can take your diff without problems |
18:08.42 | nikk | fond that with hour experiments. diff -u is proper format? ) |
18:09.03 | mnemoc | yes |
18:09.41 | traeak | mnemoc: well i did'n't make a branch or anything to make the fixes :-p |
18:10.04 | mnemoc | `git format-patch` and paste the result |
18:11.32 | traeak | mnemoc: that doesn't do anything, git diff does...i'm guessing it means i should have branched, made the fix there and then that would allow me to run the format-patch |
18:11.36 | nikk | what address send it too? http://dl.dropbox.com/u/2258618/NOVO7A/sun4i-ir.patch |
18:12.42 | mnemoc | traeak: it creates .patch files for each local commit you have |
18:14.28 | mnemoc | looks evil... it would be nicer if you could add some text for the commit to explain it |
18:14.44 | mnemoc | (and use proper C comments) |
18:16.37 | mnemoc | or actually delete the offending lines if they are really duplicatedc |
18:18.51 | traeak | mnemoc: it's basically a bunch of code deletions, inline -> static inline and moving some functions betwen files |
18:19.07 | traeak | mnemoc: sorry haven't used git before, it seems to be a pita though :-p |
18:19.23 | nikk | mnemoc: my link updated also with proper comments) |
18:19.24 | mnemoc | traeak: ok, link? and authoriship data |
18:20.12 | mnemoc | nikk: have you traced where the other event comes from? |
18:20.25 | nikk | mnemoc traced about a hour |
18:20.38 | MaDMaLKaV | hi all |
18:20.49 | MaDMaLKaV | zenitraM , you still have spare cables for the serial ? |
18:21.27 | zenitraM | MaDMaLKaV: i only have the ones mnemoc sent me |
18:21.41 | *** part/#arm-netbook QingPei (~qingpei@123.118.106.179) |
18:21.59 | MaDMaLKaV | ok, mnemoc is the Coruña guy them :) |
18:22.08 | MaDMaLKaV | mnemoc , you still have spare cables ? |
18:22.36 | mnemoc | MaDMaLKaV: yes |
18:22.48 | nikk | there are arrays for longpress also, but with some remotes 1 cycle made two (double confirm code), not remember now full logic, but it doesn't affect original remote at all |
18:22.50 | mnemoc | bought 80.. use 6 |
18:22.54 | MaDMaLKaV | can you please mail me some ? And I pay you a beer next time I go to the north |
18:23.19 | mnemoc | MaDMaLKaV: give me your address on /q |
18:23.22 | MaDMaLKaV | k |
18:24.52 | zenitraM | mmm just found this |
18:24.53 | zenitraM | http://code.google.com/p/fbterm/ |
18:27.56 | A0Sheds | http://www.linuxfordevices.com/ has been left to die, and EET and Electronic Products don't have any real ARM SOC or board news |
18:28.23 | traeak | A0Sheds: shame, it was bought out and killt |
18:28.28 | traeak | linuxfor devices i mean |
18:30.31 | *** join/#arm-netbook cromo (~cromo@139-255-251-94.net.stream.pl) |
18:30.34 | mnemoc | good we still have armfordevices :) |
18:30.48 | mnemoc | http://armdevices.net/ |
18:31.03 | mnemoc | even if the guy is not very objective :p |
18:31.48 | nikk | how to tell output not zImage but bImage on command line? make -C /Volumes/CM9/linux-allwinner O=/Volumes/CM9/out/target/product/novo7a/obj/KERNEL_OBJ ARCH=arm CROSS_COMPILE="/Volumes/CM9/prebuilt/darwin-x86/ccache/ccache /Volumes/CM9/prebuilt/darwin-x86/toolchain/arm-eabi-4.4.3/bin/arm-eabi-" sun4i_ivi_defconfig |
18:32.09 | nikk | this output zImage and modules |
18:32.13 | A0Sheds | http://armdevices.net/category/chip-provider/boxchip/ why is allwinner referred to as Boxchip? |
18:34.56 | A0Sheds | http://renesasmobile.com/products/product-renesas-mobile-MP5232.pdf |
18:35.23 | mnemoc | A0Sheds: it's their old name |
18:36.04 | mnemoc | A0Sheds: and many chinese manufacturers still refer to their new chips as Boxchip A1x in the trade shows |
18:36.17 | mnemoc | the F20 (sun3i) was the last officially labeled as boxchip |
18:36.47 | mnemoc | nikk: bImage is generated using objcpy |
18:37.07 | CIA-91 | rhombus-tech: phil master * ra6a0df297a3c /index/discussion.mdwn: reply to include/configs/suni4.h query |
18:37.10 | CIA-91 | rhombus-tech: phil master * r83e0ce741265 /: Merge branch 'master' of ssh://git.hands.com/rhombus |
18:37.15 | mnemoc | nikk: it's an android thing, `make` won't do it for you |
18:37.39 | nikk | i see. it's in CM build scripts. thx |
18:38.25 | CIA-91 | rhombus-tech: hands master * r91787423cda4 /index/discussion.mdwn: fix <HR> |
18:42.32 | mnemoc | A0Sheds: if you like anti-apple rantings, that's your place ;-) |
18:42.54 | nikk | there is just ifdefs in build/core/tasks.kernel.mk: ifeq ($(BOARD_USES_UBOOT),true) - uImage, else - zImage ))) |
18:45.29 | nikk | A0Sheds boxchip was the brand of previous allwinners' SOCs, as I know. even A10 is "boxchip" if you want ) |
18:46.05 | mnemoc | nikk: have you seen A10 socs with boxchip printed? |
18:46.21 | nikk | mnemoc no ) |
18:47.48 | nikk | but it's allwinners trademark, no? ))) |
18:48.08 | mnemoc | yes... |
18:48.34 | mnemoc | they have so many brands... |
18:48.39 | nikk | as in iphone we have samsung, but it's apple A10 or so ) |
18:49.44 | nikk | now A10 is more popular TM than Boxchip I think ) |
18:50.37 | mnemoc | "allwinner" was mentioned in the frontpage of the site of the spanish national tv and on the blog of my bank :p |
18:51.02 | mnemoc | both wrongly assumed it was the name of the a10-based tv dongle |
18:51.05 | nikk | realy? cooool ) |
18:51.18 | mnemoc | yup :) |
18:51.35 | nikk | and many chineese thought it's more powerful Cortex than A8 and A9 ))) |
18:52.31 | mnemoc | :D |
18:55.16 | mnemoc | I guess the cortex-a9 will be named A15 to win from the confusion |
18:57.57 | zenitraM | okay, found on the buildroot a patch to directfb to make it use g2d |
18:58.17 | zenitraM | ...g2d support is patched on top of the vmware directfb driver |
18:58.27 | mnemoc | o_O |
18:59.19 | zenitraM | https://github.com/allwinner-ics/lichee_buildroot/blob/master/package/directfb/directfb-1.4.11-vmware.patch |
19:02.37 | mnemoc | and functions are stilll called vmware_... |
19:02.55 | zenitraM | even *new debug* messages output as [VMware] |
19:03.35 | zenitraM | also, |
19:03.36 | zenitraM | 2011 xxxx Electronics |
19:03.41 | zenitraM | @ F23 G2D driver |
19:03.51 | zenitraM | seems they were having some identity issues at the moment |
19:06.25 | mnemoc | the axp drivers are copyrighted to "krosspower" x-powers.net |
19:06.33 | mnemoc | but Tom confirmed they are the same |
19:08.08 | mnemoc | nikk: name and email? for the commit |
19:30.51 | CIA-91 | rhombus-tech: hands master * rc820e74549ba /sandbox.mdwn: |
19:32.15 | nikk | nikk tsynik@gmail.com |
19:32.40 | mnemoc | gitanes? |
19:32.44 | nikk | yup |
19:35.40 | mnemoc | pushed |
19:35.42 | mnemoc | :) |
19:36.19 | nikk | not many ppl use IR, but it may be handy with full remotes like mine ) |
19:37.19 | mnemoc | sure |
19:37.26 | nikk | stock ainol remote have ESC NEXT VOL and JOYSTICK ) only good for set-top box ) |
19:37.42 | nikk | no power, menu, etc ) |
19:45.02 | zenitraM | giggles at the fact that nobody has bothered to try if the only "built-in" input method for the device they're developing on works :) |
19:46.01 | mnemoc | has no idea where his mele remote is |
19:51.09 | *** join/#arm-netbook rell01 (~chatzilla@p4FE56290.dip0.t-ipconnect.de) |
19:51.21 | *** join/#arm-netbook cromo (~cromo@139-255-251-94.net.stream.pl) |
20:06.40 | CIA-91 | rhombus-tech: lkcl master * r36541885947d /allwinner_a10/orders/ (.mdwn dhlotter.mdwn sauc1er.html sauc1er.mdwn): order processing cleanup |
20:07.09 | CIA-91 | rhombus-tech: lkcl master * r96adbb9f87f6 /allwinner_a10/orders/fixmeyer__64__gmail.com.mdwn: order processing cleanup |
20:10.01 | Turl | I just noticed I have more ARM devices than I thought |
20:10.15 | CIA-91 | rhombus-tech: lkcl master * r9db529f6dad8 /local.css: css test |
20:11.14 | CIA-91 | rhombus-tech: lkcl master * rcd8a8eb66e49 /local.css: css test |
20:11.32 | specing | Turl: How many?? |
20:11.38 | CIA-91 | rhombus-tech: lkcl master * rb678049390b6 /local.css: css test |
20:11.52 | specing | I have 4 under my control :) |
20:11.58 | CIA-91 | rhombus-tech: lkcl master * r294023f2f878 /local.css: css test |
20:13.01 | Turl | specing: I have 3 phones, an ereader and a tablet :) that's 5 |
20:13.25 | Turl | possibly a 6th, old phone who reports TI on usb flash mode, maybe an early OMAP :) |
20:13.35 | specing | Turl: Do have bootloader access on all of them? |
20:13.56 | Turl | bootloader as in what? |
20:14.08 | specing | Ability to upload your own kernels |
20:14.36 | Turl | yes on the phones, yes on the tablet |
20:14.42 | Turl | no idea on the kindle, probably :P |
20:16.31 | specing | Turl: You have no ARM devkits? :D |
20:17.34 | Turl | nope |
20:18.07 | Turl | I have a MIPS router tho :P MIPS is seriously outnumbered on my house :P |
20:18.23 | specing | Everybody has one of those |
20:18.30 | specing | broadcom crap, mostly |
20:18.46 | Turl | atheros here |
20:19.39 | Turl | MIPS has the privilege of getting to play music on my big speakers |
20:20.04 | specing | I was thinking about buying a stm32 / lpc to do some basic ARM osdev/controlling |
20:20.24 | specing | Maybe I should rather invest my time into a FPGA |
20:21.02 | Turl | I want to have a look at the hardware side of things a bit, I bought an arduino to eventually do something :P |
20:21.42 | specing | blegh arduinos |
20:21.54 | Turl | funny stuff though, my breadboard got faster from china than my arduino <.< |
20:22.02 | specing | I bitbang UART on a ATtiny45 ;) |
20:22.56 | lkcl | specing: do you have a bumper-sticker with that on it? "I bitbang UARTs" ? |
20:23.13 | specing | Not yet, no |
20:23.14 | Turl | haha |
20:23.29 | lkcl | i'll get one made up, specially, just for you, eh? |
20:23.37 | specing | :) |
20:23.47 | lkcl | teehee |
20:24.38 | specing | I took a while to write it, too, because I kept using the wrong register to access port status /facepalm |
20:25.55 | Turl | imagine the day where your walls have nfc tags instead of light switches |
20:27.39 | Turl | specing: look, I've got an upgrade path :P http://hlt.media.mit.edu/?p=1229 |
20:34.23 | specing | Turl: Rather invest in an Xmega ;D |
20:36.03 | specing | they cost the same, if not less than regular AVRs |
20:52.03 | nikk | have 2+ MIPS ) |
20:52.16 | nikk | and OMAP ) |
20:54.46 | mnemoc | has a 10" mips64 (loongson) laptop ;-) |
20:55.03 | mnemoc | on the shelf since 2y ago... |
20:55.13 | nikk | oh, i fogot aout PowerVR ) |
20:56.36 | Turl | mnemoc: :) |
20:59.00 | traeak | mnemoc: you have too many toys :-p |
20:59.08 | traeak | mnemoc: not married i gather? (hehe) |
21:00.55 | mnemoc | i do... but from time to time I need to spend in "tools" to improve my professional skills ;-) |
21:02.19 | traeak | "tools" |
21:02.32 | mnemoc | :) |
21:36.42 | Turl | mnemoc: check this out :D https://github.com/amery/linux-allwinner/blob/allwinner-v3.0-android-v2/drivers/video/sun4i/A10_linux_display_driver_interface.xls |
21:38.04 | traeak | cool xls files in a linux kernel repo |
21:38.28 | Turl | xls files containing chinese text* |
21:38.55 | zenitraM | xls files containing chinese text and description of a linux kernel gpio* |
21:39.04 | zenitraM | s/gpio/ioctls/ |
21:39.29 | mnemoc | I don't plan to remove any of the few bits of "documentation" we have... even if in chinese or xls :p |
21:39.41 | steev_ | mnemoc++ |
21:39.48 | Turl | http://paste.ubuntu.com/1001880/ |
21:40.13 | Turl | half courtesy of an issue on the tracker :P |
21:40.43 | traeak | just rifled through some svn repo...some ofhte damn cpp files are utf16LE encoded |
21:41.08 | zenitraM | Turl: there's a cleaner version of that code by mnemoc |
21:43.24 | zenitraM | also, theres a fb_tool.dat or something like that executable that will toggle between displays |
21:55.49 | mnemoc | sunxi-tools should actually include a real tool for that... |
22:02.10 | steev_ | whoa |
22:02.34 | steev_ | their 3.0.8 kernel compiles without any changes/weird scripts being run? wtf |
22:02.35 | mnemoc | next weekend will do :) |
22:03.37 | mnemoc | sure |
22:03.59 | steev_ | i've never experienced a kernel from a vendor that built without issue before |
22:04.52 | mnemoc | this is a cool vendor :p |
22:05.14 | traeak | steev_: use gcc-4.7 i suspect that woudl crash and burn (if the same code was in 3.0.31) |
22:05.27 | Turl | it's so cool they distributed their build directories verbatim :P |
22:07.13 | ZaEarl | wouldn't it be cooler if a vendor actually had public github repos? ;) |
22:07.14 | mnemoc | and the kernel sources for a soc no manufacturer had received (sun5i/a12,a13) in the git history :p |
22:07.35 | mnemoc | ZaEarl: they do, for most of the android stuff |
22:07.48 | steev_ | stuff still needs to be worked on privately (first) |
22:07.51 | mnemoc | ZaEarl: https://github.com/allwinnerwk |
22:08.25 | ZaEarl | nice, I don't think i've seen that yet |
22:08.44 | steev_ | i definitely hadn't |
22:10.42 | CIA-91 | rhombus-tech: tzafrir master * r57ab932233b7 /iMX6.mdwn: a 'cpus' tag |
22:11.36 | ZaEarl | steev_, why does stuff need to be worked on pirvately? |
22:11.46 | CIA-91 | rhombus-tech: tzafrir master * ra980d715e5d9 /cpus.mdwn: |
22:12.21 | steev_ | ZaEarl: because you need to make sure it works before some retard on the internet grabs the sources, compiles it, and blows the fuses on their board (just an example) |
22:12.42 | steev_ | s/retard/idiot/ |
22:13.08 | CIA-91 | rhombus-tech: tzafrir master * r9c1360c42c89 /am335x.mdwn: a 'cpus' tag |
22:13.16 | steev_ | (and then calls/emails you and bitches you out for releasing shit software) |
22:15.29 | steev_ | look at the vast number of threads on android forums where people say, this is still a wip, <known issues are> and there are people who complain because <known issue> isn't working yet |
22:15.52 | mnemoc | :) |
22:18.12 | ZaEarl | I think the benefits outweigh the disadvantages. |
22:18.47 | mnemoc | not when the idiots are paid customers demanding every drop to work out of the box and immediatelly |
22:19.03 | mnemoc | paying* |
22:20.10 | A0Sheds | ZaEarl : are you guys working on an ARM netbook as well? |
22:20.30 | ZaEarl | A0Sheds, not at this time. |
22:20.50 | mnemoc | paying customers will be calling you for support all f* day (personal experience) and not letting you work :| |
22:21.13 | A0Sheds | ZaEarl: are you going to Computex next month? |
22:21.35 | steev_ | mnemoc: god i can relate |
22:22.07 | ZaEarl | no, we stick with the china fairs |
22:22.38 | A0Sheds | ZaEarl: which ones do you find the most worth attending? |
22:22.52 | steev_ | we gave 52 of our machines to Linaro/Canonical/Ubuntu developers. if you look in mainline, our board support file is (C) Linaro in the header. but they've done squat on it aside from taking our work and submitting it to the kernel |
22:23.04 | A0Sheds | Honk Kong Electronics or? |
22:23.14 | ZaEarl | Global Sources China Sourcing |
22:23.19 | A0Sheds | ah ok |
22:23.42 | hno | mnemoc, public repositories to not conflict with supported code drops to paying customers. |
22:24.05 | steev_ | hno: it doesn't matter if you say "this is not supported" |
22:24.15 | steev_ | customers expect support on ANYTHING you provide |
22:24.21 | mnemoc | hno: after certain (undefined) level of maturity :p |
22:26.42 | hno | steev_, not in my experience. Most paying customers know the distinction perfectly well. The problem is those who are not paying and tryign to leech their way. |
22:27.01 | mnemoc | hno: you have smart customers :) |
22:27.30 | steev_ | hno: pretty hard for people not to pay for our devices. |
22:28.18 | mnemoc | you probably don't get called when "the system doesn't work" because the servers room is under 1m of water or the whole f* city has a power outtage |
22:28.54 | steev_ | mnemoc: i don't get those calls either :P |
22:29.18 | mnemoc | me makes software for car wash and lube centers :p |
22:29.37 | steev_ | but i have gotten a *very* angry email from a customer because something didn't work when they installed a distro which shall not be named (and we didn't do that release at all) |
22:29.59 | mnemoc | :) |
22:32.04 | hno | steev_, and your response was? |
22:33.24 | steev_ | hno: i politely explained that i was not familiar with the distribution in question, had no contact with anyone involved with it, but after a quick look at their website, suggested they try the forums that the distro had available there. |
22:35.36 | hno | steev_, back to the source repositories topic. Having a public source repository is quite different from rekeasing firmware images. I have not met anyone who considers the public source repository to be the same as stable release. |
22:36.53 | hno | probably has a bit to do with the effort needed to make use of it. Building completely from source repository usually requires a fair bit of reading and preparation. |
22:37.50 | mnemoc | or "HELP!!!!!" mails :-/ |
22:39.24 | hno | HELP!!! mails on that topic usually get a RTFM back, with the exception of a paying customer needing help to build a supported code drop. |
22:39.53 | hno | those just get a RTM back. |
22:40.04 | mnemoc | :p |
22:40.05 | specing | Lol |
22:40.35 | specing | hno: Who do you work for? |
22:41.22 | hno | works for himself. |
22:42.55 | hno | have not had an employment in 7 years. |
22:45.02 | hno | or 12 depending on how you count, |
22:45.07 | traeak | hehe, but you get money |
22:45.24 | traeak | the company i work for has...3 full time employees |
22:45.25 | traeak | heh |
22:45.31 | hno | yes, i have customers |
22:45.38 | hno | but no boss |
22:45.50 | hno | other than myself |
22:46.13 | *** join/#arm-netbook faux123 (faux@macinyasha.net) |
22:51.54 | hno | 1992 I parttime worked on porting games from Amiga (motorola 68K assembly) to PC, while studying.1993-1995 I worked at the university while studying, 1996-2000 I was employed in a consulting business. 2000-2005 I managed a small company building Linux based firewalls and similar equipment. From 2005 I am just myself and a network of people I work with. |
22:54.29 | hno | and been using free software since 1992, and actively involved in free software projects since 1995. |
22:58.18 | ZaEarl | steev_, which kernel did you compile without trouble? |
23:05.41 | nikk | i think he mentioned qware with their old 4.0.3, i compiled it without any trouble, even config for mu device was there ;-) |
23:13.03 | ZaEarl | that was my guess, but I wanted to be sure |