| 00:26.34 | *** join/#arm-netbook orly_owl (~david@unaffiliated/orly-owl/x-3167833) |
| 00:52.43 | *** join/#arm-netbook freakazoid0223 (~matt@pool-173-75-233-172.phlapa.fios.verizon.net) |
| 01:09.05 | *** join/#arm-netbook ZaEarl (~malmrose@c-67-166-109-0.hsd1.ut.comcast.net) |
| 01:12.22 | *** join/#arm-netbook ZaEarl (~malmrose@c-67-166-109-0.hsd1.ut.comcast.net) |
| 01:41.40 | *** join/#arm-netbook vinifm (~vinix@177.98.172.134) |
| 01:47.07 | *** join/#arm-netbook ZaEarl (~malmrose@c-67-166-109-0.hsd1.ut.comcast.net) |
| 01:55.20 | *** join/#arm-netbook stefanro1 (~stefan@p57B95BF4.dip.t-dialin.net) |
| 02:02.15 | *** join/#arm-netbook Maqs (~marcus@marcus.eunomia.de) |
| 02:02.19 | *** join/#arm-netbook itamar_ (~itamar@cl-24.udi-01.br.sixxs.net) |
| 02:46.04 | *** join/#arm-netbook Mehhh (adabd66c@gateway/web/freenode/ip.173.171.214.108) |
| 03:19.04 | vinifm | hi |
| 03:36.36 | Sv | hi |
| 04:18.34 | *** join/#arm-netbook jquip (~johnny@106.211.151.90) |
| 04:25.40 | Mehhh | Can I just remove the firewire drivers before compile? |
| 04:25.44 | Mehhh | delete the folder |
| 04:43.37 | Mehhh | Wait, is there any reason some of these are in the tree in the first place? |
| 04:43.43 | Mehhh | besides that they came with mainline |
| 05:01.02 | *** join/#arm-netbook rm (rm@fsf/member/rm) |
| 05:20.41 | *** join/#arm-netbook Maqs (~marcus@marcus.eunomia.de) |
| 05:28.57 | Mehhh | is there any problem with gcc 4.8-ing this? |
| 05:52.52 | *** join/#arm-netbook rsalveti (~rsalveti@linaro/rsalveti) |
| 05:57.38 | *** join/#arm-netbook Ershov (~Thunderbi@89.163.11.110) |
| 06:28.48 | *** join/#arm-netbook jquip (~johnny@42.107.231.206) |
| 06:32.22 | *** join/#arm-netbook Quarx (~Quarx@109.120.33.184) |
| 07:11.38 | Mehhh | Is sunxi-bsp the better way to go? |
| 07:20.21 | libv | Mehhh: the better way to go for what? |
| 07:20.30 | libv | Mehhh: for getting a 3.8 kernel up and running? |
| 07:25.14 | *** join/#arm-netbook rellla (~rellla@p5B078561.dip0.t-ipconnect.de) |
| 07:26.00 | Mehhh | no, for the optimizations in the 4.7 |
| 07:26.05 | Mehhh | gcc |
| 07:26.32 | Mehhh | oh wait |
| 07:26.58 | Mehhh | sorry wrong thing, better way to go for managing the entire thing and creating the image |
| 07:27.06 | Mehhh | better question, which path do you take? :) |
| 07:28.46 | libv | i wrote down what i did in firststeps |
| 07:29.07 | libv | which i think by now you should've been able to execute several times over already |
| 07:29.11 | Mehhh | yes |
| 07:29.18 | Mehhh | did, then i saw that |
| 07:29.32 | Mehhh | i wondered the difference in management (as in, why) |
| 07:29.43 | *** join/#arm-netbook Ershov (~Thunderbi@89.163.11.110) |
| 07:30.08 | libv | Mehhh: does the difference matter to you once you have succeeded? |
| 07:30.37 | Mehhh | Not really |
| 07:30.37 | libv | the bsp thing is about convenience |
| 07:30.45 | Mehhh | oh, that's what i figured |
| 07:30.46 | Mehhh | gotcha |
| 07:30.47 | libv | which is important for some people |
| 07:31.06 | libv | for someone like me, having control over how to build the kernel, is quite important |
| 07:31.26 | libv | Mehhh: but again, why does all of this matter? |
| 07:31.31 | libv | Mehhh: matter to you personally |
| 07:31.40 | libv | you've been in this channel for quite a while now |
| 07:32.02 | libv | surely you've succeeded in getting your device up and running by now |
| 07:32.12 | Mehhh | yes, but i'm also OCD |
| 07:32.20 | Mehhh | and doing a writeup for other people not related to the channel |
| 07:32.41 | libv | checks RecentChanges in the wiki |
| 07:33.22 | libv | where is it? |
| 07:34.03 | Mehhh | In LibreOffice |
| 07:34.13 | Mehhh | i don't understand what you're asking |
| 07:34.20 | libv | with me it took a while until things were working for me |
| 07:34.35 | Mehhh | oh, haven't checked all that yet, just amde sure it boot |
| 07:34.49 | libv | first, because firststeps was not there yet, secondly because the psu i used was underpowered and the device was not stable at all |
| 07:34.51 | Mehhh | It's easier to go back and edit |
| 07:34.57 | Mehhh | ahhhh |
| 07:34.58 | Mehhh | mk802? |
| 07:35.05 | libv | no, mele |
| 07:35.25 | Mehhh | Oh gotcha, yeh i'm having to figure out certain portions for mk802 |
| 07:35.35 | libv | once the thing was working, i asked hno the right questions, and hno gave some stellar answers, and the A7HD tablet was supported in no time |
| 07:35.56 | Mehhh | Hell yeh |
| 07:36.08 | libv | now, what good is an office document going to do anyone? |
| 07:36.55 | Mehhh | I'm handing the project off to someone else, I just need this done so i can have an arm netbook completed before i go on vacation |
| 07:38.38 | libv | so have you actually been able to get anything real booted on this device? |
| 07:38.43 | Mehhh | yes |
| 07:39.04 | Mehhh | i have to fork this into my personal use and one for the project |
| 07:39.05 | libv | ok, so you have the .fex and uboot figured out |
| 07:39.09 | Mehhh | yes |
| 07:39.20 | libv | have you sent this in to the ml? |
| 07:39.23 | Mehhh | uboot was never a problem, i'm worried more about speed and support and the modules |
| 07:39.57 | Mehhh | for one specific device |
| 07:40.07 | libv | have you documented this device further on the wiki? |
| 07:40.13 | libv | or are you just using the mk802? |
| 07:40.56 | Mehhh | just the mk802 |
| 07:41.05 | Mehhh | internal lapdock mod |
| 07:41.42 | Mehhh | The project owner can't get his working (two mk802's), so i'm saying screw it and just DIY. |
| 07:42.17 | Mehhh | I'm just looking at anything extra that may get shoved in, and/or config support i want |
| 07:43.00 | libv | ok, why not just do what i have done on projects: hand over a known good sd card which should not be touched, store a dd-ed image somewhere, and point the person taking over to the FirstSteps howto for further info on how this was created |
| 07:43.27 | Mehhh | I'll most likely being doing the kernel updates at a later point |
| 07:43.28 | libv | if firststeps is not good enough, then fix it |
| 07:43.33 | Mehhh | i'll prob send him the sd card |
| 07:43.43 | Mehhh | hmmm, i may |
| 07:43.50 | libv | i personally cover more bases. |
| 07:43.50 | Mehhh | Didn't know if i was allowed to |
| 07:44.01 | libv | it's a wiki. |
| 07:44.11 | libv | if we do not like your changes, we fix them |
| 07:44.12 | Mehhh | I mean.... not in terms of possible but in terms of |
| 07:44.19 | Mehhh | favorable / un |
| 07:44.22 | Mehhh | ah |
| 07:44.23 | Mehhh | kk |
| 07:45.32 | Mehhh | the malis on the latest pull are the ones from the video, correct? |
| 07:46.08 | libv | what malis, what pull, what video? |
| 07:46.16 | Mehhh | your cube |
| 07:46.24 | Mehhh | @ cnx |
| 07:46.33 | libv | that runs on my lima driver code |
| 07:47.05 | Mehhh | I mean, is that included already in the latest pull / wiki |
| 07:47.15 | libv | latest pull of what? |
| 07:47.32 | Mehhh | i cloned the entire repo |
| 07:47.40 | libv | which repo? |
| 07:47.41 | Mehhh | i didn't know if the changes were private / testing or on there |
| 07:47.45 | Mehhh | the entire thing |
| 07:48.00 | libv | the entirety of github.com? |
| 07:48.11 | Mehhh | sunxi |
| 07:48.24 | Mehhh | github.com/linux-sunxi/ |
| 07:48.28 | Mehhh | cloned |
| 07:48.39 | Mehhh | all sub-dirs |
| 07:48.54 | libv | http://lmgtfy.com/?q=lima+driver |
| 07:49.02 | Mehhh | I know what it is |
| 07:49.17 | libv | you clearly don'T |
| 07:49.31 | libv | because your questions make absolutely no sense |
| 07:49.48 | Mehhh | I clearly do. I was asking it if was linked already |
| 07:50.03 | Mehhh | or sep |
| 07:50.11 | libv | it has always been separate |
| 07:50.23 | Mehhh | Cool, thank you for answering my question. |
| 07:50.35 | libv | it's not useful for normal users of sunxi |
| 07:50.43 | libv | and it is supposed to be much more general than just sunxi |
| 07:50.53 | libv | sunxi is just the perfect platform for lima development at this time |
| 07:51.00 | Mehhh | Look, I don't know if i said something or |
| 07:51.13 | Mehhh | phrased something incorrectly or in a way that isn't kosher. |
| 07:51.19 | libv | it is the way you ask questions |
| 07:51.20 | Mehhh | I'm just getting it done |
| 07:52.12 | libv | they seem to be without a clear goal, and are strangely phrased, and my requests for quantification gets more wishy-washy answers |
| 07:52.18 | Mehhh | If me being casual offends you, you don't have to respond. |
| 07:52.42 | Mehhh | I'm really not here to pick fights. |
| 07:53.30 | Mehhh | The goals aren't clear because I have two different sets of goals |
| 07:54.06 | Mehhh | Your driver is obviously nice, and works well. |
| 07:54.36 | Mehhh | Well, good enough for the acceleration that will be expected of the image. |
| 07:54.40 | libv | it's not even a driver yet. |
| 07:54.42 | Mehhh | and then some |
| 07:55.54 | Mehhh | Looks like one to me, hince this image vs the other two |
| 07:56.05 | Mehhh | I'm trying to pull a Ford here, not a Ferrari. |
| 07:56.45 | libv | in case you were not paying attention, the cube is running on limare, a special lib meant to gradually build up knowledge of the mali hw, there are some gles tests, and there are corresponding implementations using the limare library, which gets adapted accordingly |
| 07:57.08 | libv | q3a is not running on top of limare (yet) |
| 07:57.18 | libv | it runs on the arm binaries |
| 07:57.51 | Mehhh | i am paying attention |
| 07:57.53 | libv | the arm binaries are a complete opengles[12] driver stack |
| 07:58.22 | Mehhh | Nothing new I didn't read the past few days |
| 07:58.27 | Mehhh | Brb |
| 07:59.05 | libv | then the questions you've been asking just now were not in correlation with the knowledge you supposedly already acquired |
| 08:00.32 | Mehhh | You say automobile, I say car. |
| 08:00.50 | Mehhh | Doesn't matter |
| 08:01.20 | oliv3r | ... |
| 08:02.56 | libv | oliv3r: :) |
| 08:03.20 | oliv3r | americans |
| 08:04.10 | Mehhh | Socrates vs Euthyphro 2012. |
| 08:12.59 | *** join/#arm-netbook jquip (~johnny@sentry.hbcse.tifr.res.in) |
| 08:56.03 | Mehhh | Not sure who wrote the neon / vfp wiki notes but A+ |
| 08:59.41 | jquip | <PROTECTED> |
| 09:03.07 | libv | for anyone wanting to play with lima themselves, just pushed some things to https://gitorious.org/lima |
| 09:03.31 | *** join/#arm-netbook cheng_ (~cheng@175.139.154.152) |
| 09:10.30 | *** join/#arm-netbook popolon (~popolon@og-free.planet-service.fr) |
| 09:20.54 | *** join/#arm-netbook Ershov1 (~Thunderbi@89.163.11.110) |
| 09:29.45 | *** join/#arm-netbook Ershov (~Thunderbi@89.163.11.110) |
| 09:32.20 | *** part/#arm-netbook Ershov (~Thunderbi@89.163.11.110) |
| 10:08.01 | Mehhh | hmmmm, my mac is zero'd out |
| 10:08.29 | Mehhh | Is there any way to have it automatically detect within that script? |
| 10:10.30 | Mehhh | VS i am assuming a manual edit / rec |
| 10:13.58 | libv | you did not RTFM |
| 10:14.22 | libv | no, sorry |
| 10:14.33 | libv | what script? bsp one? |
| 10:15.09 | Mehhh | no, script.bin |
| 10:15.47 | libv | ? |
| 10:16.05 | libv | think about this for a minute |
| 10:16.33 | Mehhh | I did and found the solution |
| 10:48.25 | *** join/#arm-netbook tinti (~tinti@maxtrack-F4-0-3-gacc04.bhe.embratel.net.br) |
| 11:20.04 | *** join/#arm-netbook jquip (~johnny@sentry.hbcse.tifr.res.in) |
| 11:21.11 | jquip | Hola Amigos! I have a a10_mid + ft5x touchscreen tablet right here... modprobe is OK... but no touch? using def_config(mnemoc) |
| 11:22.09 | jquip | should i disable multitouch in the ft5x.h?? I thought it was resolved ?? |
| 11:22.31 | jquip | mnemoc: have tested 3.0.52 |
| 11:22.35 | jquip | :) |
| 11:22.47 | jquip | works well... |
| 11:23.45 | jquip | erm: HALP! |
| 11:26.30 | jquip | mnemoc: but thing though... var/log/kern.log spits hundreds of these : |
| 11:26.32 | jquip | <PROTECTED> |
| 11:26.34 | jquip | Jan 12 18:39:41 linaro-alip kernel: [ 498.540000] UMP<1>: No handler for IOCTL. cmd: 0x400c9007, arg: 0xbea6caf8 |
| 11:26.36 | jquip | Jan 12 18:39:41 linaro-alip kernel: [ 498.550000] UMP<1>: No handler for IOCTL. cmd: 0x400c9007, arg: 0xbea6caf8 |
| 11:26.38 | jquip | Jan 12 18:39:41 linaro-alip kernel: [ 498.560000] UMP<1>: No handler for IOCTL. cmd: 0x400c9007, arg: 0xbea6ae38 |
| 11:26.40 | jquip | Jan 12 18:39:41 linaro-alip kernel: [ 498.560000] UMP<1>: No handler for IOCTL. cmd: 0x400c9007, arg: 0xbea6ae38 |
| 11:26.42 | jquip | Jan 12 18:39:41 linaro-alip kernel: [ 498.570000] UMP<1>: No handler for IOCTL. cmd: 0x400c9007, arg: 0xbea6caf8 |
| 11:26.44 | jquip | Jan 12 18:39:41 linaro-alip kernel: [ 498.580000] UMP<1>: No handler for IOCTL. cmd: 0x400c9007, arg: 0xbea6caf8 |
| 11:26.48 | RaYmAn | please stop spamming |
| 11:26.48 | jquip | Jan 12 18:39:41 linaro-alip kernel: [ 498.590000] UMP<1>: No handler for IOCTL. cmd: 0x400c9007, arg: |
| 11:27.03 | RaYmAn | one line would clearly have been enough :P |
| 11:27.15 | jquip | sorry |
| 11:27.18 | jquip | yes |
| 11:27.20 | RaYmAn | however, it sounds like you're using wrong kernel version vs drivers |
| 11:27.53 | jquip | now that would be highly unlikely... 3.0.52 both... |
| 11:28.19 | RaYmAn | the UMP log thing means *userspace* graphics drivers try to talk to the kernel in a way that the kernel doesn't know (essentially) |
| 11:28.54 | jquip | ahhh! interesting... go on... |
| 11:29.48 | RaYmAn | that's about it. Wrong version userspace drivers vs kernel drivers. |
| 11:38.50 | libv | jquip: standard kernel should have r3p0 mali |
| 11:39.27 | Mehhh | Does neon work with the default lin tc w/hf ? |
| 11:39.50 | Mehhh | armhf + neon |
| 11:39.54 | jquip | libv: but but .. mnemoc gave me this to test http://dl.linux-sunxi.org/amery/sunxi-3.0/20121206-1739/ |
| 11:40.43 | jquip | and this was yesterdays... |
| 11:40.55 | jquip | 3.0.52 |
| 11:41.42 | libv | strings -n12 /usr/lib/libMali.so | grep BUILD |
| 11:41.50 | libv | jquip: do not paste the result. |
| 11:42.11 | libv | use the thing between your ears to interprete the result |
| 11:46.16 | jquip | ahahah : alrite :) i just pasted them to emphasize .. meh goodness... |
| 11:48.01 | Mehhh | mehhh goodness |
| 11:55.00 | rm | http://www.buyincoins.com/new_en/details/rikomagic-mk802-ii-android-4-0-mini-1080p-pc-wifi-player-google-dongle-hd-tv-box-product-20383.html |
| 11:55.08 | rm | "OS: Android 4.0 / Can also run desktop versions of Linux from a bootable SD Card." |
| 11:57.40 | *** join/#arm-netbook ssspiff (~wim@103.4-78-194.adsl-static.isp.belgacom.be) |
| 11:57.41 | *** join/#arm-netbook ssspiff (~wim@unaffiliated/yukito) |
| 11:58.32 | Mehhh | Only to the buyer's remorse when they try to jam an SD card into the USB port or MicroSD slot. |
| 12:01.01 | mnemoc | jquip: the hwpack came with the wrong libs? |
| 12:01.07 | *** join/#arm-netbook Ershov (~Thunderbi@89.163.11.110) |
| 12:01.31 | mnemoc | jquip: or you added the wrong libs yourself? |
| 12:03.50 | libv | rm: i wonder why AMD Zxxx is always being mentioned these days |
| 12:03.59 | libv | for the gpu |
| 12:10.55 | jquip | mnemoc: i did an rsync -ar rootfs/* .. but I did use a rootfs that was might have been spurious... |
| 12:14.46 | jquip | will test on a fresh rootfs.. |
| 12:16.19 | mnemoc | the media create script in sunxi-bsp isn't working? |
| 12:19.52 | *** join/#arm-netbook Quarx (~Quarx@109.120.33.184) |
| 12:21.09 | *** join/#arm-netbook focus_it (~focus_it@cpc7-dals17-2-0-cust182.hari.cable.virginmedia.com) |
| 12:23.45 | focus_it | when is mk802 quad core A31 coming out? :-) |
| 12:25.20 | focus_it | using AXP221 power management chip preferably |
| 12:25.35 | jquip | mnemoc: oh shoot.. i must've missed a step or two.. will get back ... |
| 12:25.37 | mnemoc | how can we know? :) |
| 12:25.48 | mnemoc | focus_it: --^ |
| 12:26.40 | mnemoc | focus_it: I'm still trying to get a reply about buying a devkit for allwinner's cortex-a7... I suppose normal manufacturers might have the same problem |
| 12:26.58 | mnemoc | unless allwinner/wits itself is designing the PCBs and making the software |
| 12:27.10 | focus_it | http://www.gizmochina.com/2012/12/05/ampe-a10-tablet-tear-down-with-allwinner-a31-quad-core-processor/ tear down of A31 quad core |
| 12:27.45 | mnemoc | focus_it: can you add a page in linux-sunxi.org for the A31 ? |
| 12:28.04 | focus_it | eek how? |
| 12:28.43 | mnemoc | it's a wiki... register and write :) |
| 12:29.02 | focus_it | will try |
| 12:29.45 | mnemoc | it's pretty sad they abandoned mali |
| 12:30.03 | focus_it | mnemoc: how come other companies have already built A31 while you can't get dev kit??? |
| 12:30.04 | Mehhh | Do we know for certain that's the case in the A40? |
| 12:31.03 | RaYmAn | focus_it: I suspect it's a lot easier to get it if you produce stuff next door or similar |
| 12:31.04 | mnemoc | Mehhh: there is no such thing as the A40 |
| 12:31.09 | RaYmAn | rahter than across the internets |
| 12:31.35 | RaYmAn | mnemoc: what do they use isntead of mali? |
| 12:31.51 | mnemoc | Mehhh: there was tons of "speculation" about the sun6i. device manufacturers made up the A40 and A15X to refer to the "unnamed" sun6i chip |
| 12:32.16 | Mehhh | ah ah |
| 12:32.30 | mnemoc | RaYmAn: PowerVR SGX544 |
| 12:32.40 | focus_it | RaYmAn: I agree - face to face is very important in China. I advise everyone, if you want to do business in China, make a point of visiting the factory before drawing up biz plans |
| 12:34.25 | mnemoc | even lkcl ended up getting the A10 PCB designed by wits |
| 12:34.57 | mnemoc | with A31 might be the same |
| 12:35.25 | mnemoc | so no need for manufacturers to have a devkit |
| 12:35.39 | mnemoc | they just pay allwinner/wits to do all the work |
| 12:36.48 | focus_it | mnemoc: that sounds interesting - so any idea how much money they charge - is it like high price or a typical CN design house price? |
| 12:37.35 | mnemoc | I guess it's competitive with other chinese design houses |
| 12:37.48 | RaYmAn | focus_it: there's probably a reason 95% of a10 devices are identical |
| 12:37.50 | mnemoc | but the key for allwinner is cheap production |
| 12:38.11 | mnemoc | they only "adjust" the reference design |
| 12:38.34 | mnemoc | the only real diff between tablets is the choice of ts |
| 12:39.01 | *** join/#arm-netbook rz2k (~rz2k@95-25-166-111.broadband.corbina.ru) |
| 12:39.15 | Mehhh | Are their SoCs pin /bga compatible ? |
| 12:39.23 | focus_it | RaYmAn: cubies schematic and lkcl schematics and wits schematics all similar - with same repeated 'mistakes'. I now understand!! |
| 12:40.02 | mnemoc | Mehhh: no |
| 12:40.06 | Mehhh | rats |
| 12:40.24 | mnemoc | they don't do odd variants |
| 12:41.01 | mnemoc | the chips have very distintive targets, and so pin compatibility isn't interesting |
| 12:41.02 | Mehhh | I've been noticing that. I was looking for a stick verson that was slightly longer with a more narrow PCB |
| 12:41.05 | Mehhh | didn't find it |
| 12:41.19 | mnemoc | the A10s was made for that |
| 12:41.53 | Mehhh | Yeh, all those PCBs looked to be =/= |
| 12:42.19 | *** join/#arm-netbook XenGi (~XenGi@2a02:748:a800:149:154:158:199:1) |
| 12:42.24 | Mehhh | I think the smaller ones were telechips, amlogic, etc |
| 12:42.28 | mnemoc | BGA336 14x14mm |
| 12:42.36 | mnemoc | the allwinner A10s I mean |
| 12:42.38 | *** join/#arm-netbook opm_70 (~opm@157.Red-80-33-235.staticIP.rima-tde.net) |
| 12:42.50 | Mehhh | Does the S hold any gains over the non? |
| 12:43.09 | mnemoc | http://www.allwinnertech.com/en/product/A31.html |
| 12:43.36 | mnemoc | Mehhh: smaller size probably counts as a gain for the dongle market |
| 12:43.54 | Mehhh | Oh, it's physically smaller. Score. |
| 12:44.22 | focus_it | wow! it has 4K video! |
| 12:45.14 | focus_it | PMIC AXP221 has 21 channels of managed power!! sheesh! |
| 12:47.40 | focus_it | grumble - it has ms os support - probable ms will want everything locked out for Linux uses |
| 12:48.27 | focus_it | I wish lkcl with his FSF endorsed CPU - more than ever its needed |
| 12:50.49 | *** join/#arm-netbook Jef91 (~Instantbi@bodhilinux/team/Jef91) |
| 12:51.16 | Jef91 | anyone around know a good HOWTO for getting an alternative OS booting on the new chrome book? |
| 12:52.06 | Mehhh | the acer c7? |
| 12:52.19 | mnemoc | probably the exynos |
| 12:52.30 | mnemoc | considering this is an ARM-centric channel |
| 12:53.25 | Jef91 | yea that one mnemoc |
| 12:53.28 | Jef91 | the new samsung one |
| 12:54.58 | Mehhh | Posted in main ;) |
| 12:55.36 | mnemoc | Jef91: RaYmAn is (afaik) who has played more with that one here |
| 12:55.51 | Jef91 | kk mnemoc |
| 12:56.24 | RaYmAn | Jef91: checkout Olof Johansson's g+ posts: https://plus.google.com/109993695638569781190/posts |
| 12:58.26 | *** join/#arm-netbook Ershov (~Thunderbi@89.163.11.110) |
| 13:07.17 | *** join/#arm-netbook popolon (~popolon@og-free.planet-service.fr) |
| 13:08.55 | focus_it | mnemoc: done http://linux-sunxi.org/A31#External_links |
| 13:10.13 | mnemoc | focus_it: what about adding more details/specs about the soc too? |
| 13:10.55 | focus_it | hmm.. i haven't any at the moment |
| 13:11.02 | antonl | Hi guys. About exynos-based chromebook: what works in linux installed that way and what doesn't? 3d, sound? |
| 13:11.26 | mnemoc | focus_it: the allwinner's product page has info... |
| 13:11.33 | focus_it | mnemoc: sorry wrong reply! I were referring to A31! |
| 13:11.36 | RaYmAn | antonl: yes |
| 13:12.44 | focus_it | mnemoc: confused - more info about which soc? |
| 13:13.11 | mnemoc | A31 |
| 13:13.19 | antonl | Any hope of completely replacing the chromeos with linux? Is internal SSD much faster than a class 10 SD card? |
| 13:13.37 | *** join/#arm-netbook Ershov1 (~Thunderbi@89.163.11.110) |
| 13:13.52 | focus_it | mnemoc: the two links there provide the only information I have - is there more that can be found from Allwinner site? |
| 13:14.09 | mnemoc | focus_it: never mind |
| 13:14.17 | focus_it | yes |
| 13:14.20 | focus_it | :-) |
| 13:14.23 | mnemoc | focus_it: I was asking you to add information to the wiki, not links |
| 13:14.28 | mnemoc | but don't worry |
| 13:14.58 | focus_it | love to as I get prepare more stuff |
| 13:15.33 | RaYmAn | antonl: hm, ssd should be pretty fast, but I haven't run any proper tests myself. You can easily replace default os with linux - you can also boot directly off SD or USB for dev/multiboot |
| 13:17.12 | antonl | so, the bootloader is unlocked? I remember many problems of now-sold AC100, regarding bootloader issues. |
| 13:19.06 | RaYmAn | yes, it's unlocked. You need to enable dev mode to boot alt. os, but it's easy |
| 13:21.20 | *** join/#arm-netbook vinifm (~vinix@187.58.77.182) |
| 13:22.50 | RaYmAn | antonl: incidentally, it's also pretty much unbrickable (without hardware mods). You can always boot a chromeos recovery image and reflash the entire device |
| 13:23.44 | Jef91 | hey RaYmAn do you know if there are any pre-built images for the chromebook yet? |
| 13:30.38 | antonl | So, currently we are bound to the kernel distributed in the official chreomeos image? What's with hw video decoding? |
| 13:31.05 | RaYmAn | no, once in dev mode you can flash/boot whatever kernel you want (it just needs to be signed with the dev key) |
| 13:32.09 | rastko | rz2k: Where should I post a patch for ili210x touchscreen? |
| 13:32.26 | RaYmAn | I don't think hw video decoding is there yet, but I don't know |
| 13:32.36 | Jef91 | Alrighty RaYmAn. It always confuses me why people don't just create a .img file of their working bootable SD card instead of making people follow long HOWTO |
| 13:32.37 | Jef91 | s |
| 13:33.00 | RaYmAn | probably because it's rather easy to do and distributing a fairly large image is a pain? :) |
| 13:33.29 | RaYmAn | there is uhm, a project where you download & run a script and it prepares the chromebook and installs linux |
| 13:33.45 | *** join/#arm-netbook Ershov (~Thunderbi@89.163.11.110) |
| 13:34.02 | Jef91 | RaYmAn: not really. I host images for serveral ARM devices all on source forge free of charge |
| 13:34.10 | antonl | So, have samsung released the kernel sources that were used to build the chromebook kernel? |
| 13:34.39 | RaYmAn | of course |
| 13:34.50 | RaYmAn | Jef91: http://chromeos-cr48.blogspot.dk/ |
| 13:35.21 | RaYmAn | antonl: git.chromium.org - chromiumos/third_party/kernel.git |
| 13:36.33 | Jef91 | What is the latest samsung chromebook called RaYmAn? |
| 13:37.04 | RaYmAn | ARM Chromebook? xe303<something> is the modle nr |
| 13:37.06 | RaYmAn | model* |
| 13:40.03 | antonl | GPU is Mali 604? |
| 13:46.10 | RaYmAn | yes |
| 14:05.43 | *** join/#arm-netbook nemik_ (~cyanact@c-67-162-60-211.hsd1.il.comcast.net) |
| 14:06.01 | libv | so has anyone tried building/running the freshly pushed lima code yet? |
| 14:10.40 | libv | aw :( |
| 14:14.06 | Turl | libv: what's the status of it? :) |
| 14:14.52 | slapin | hi, all! |
| 14:15.02 | Turl | hi slapin |
| 14:16.03 | slapin | is there somewhare I can order chromebook? looks tasty... |
| 14:16.22 | slapin | or all are gone by now? |
| 14:16.40 | Turl | google play store? just a guess though |
| 14:16.56 | mnemoc | amazon |
| 14:17.02 | Turl | http://www.google.com/intl/en/chrome/devices/buynow.html#online |
| 14:17.46 | libv | Turl: it will spin the cube companion for you |
| 14:17.58 | libv | Turl: it should be easily buildable on a normal linux |
| 14:19.05 | Turl | talking about the cube, I saw some benchmarks on the wiki I think |
| 14:19.20 | Turl | it was like a lot of fps for 720p and like.. 2-3 for 1080p |
| 14:19.47 | Turl | libv: why is that? |
| 14:20.12 | libv | Turl: first off, 1080p is 2.25 the amount of data |
| 14:20.17 | libv | times even |
| 14:20.51 | libv | so the 40fps for normal quake already becomes like 18fps |
| 14:20.58 | Turl | hrm, not the cube but quake |
| 14:21.19 | libv | then, add in the fact that this data not only needs to go somewhere |
| 14:21.27 | libv | it needs to come from somewhere as well |
| 14:21.53 | libv | there's a load of texture lookups that need to be done for quake |
| 14:22.48 | Turl | 40 fps on 720p, 5 fps on 1080p |
| 14:23.00 | libv | so if you factor in that these texture lookups will be anywhere from 1 to 4pixels (even mipmapped) or even much more for a complex scene... |
| 14:23.34 | libv | and you're suddenly pulling in a lot of data across the memory subsystem |
| 14:23.37 | rz2k | <rastko> rz2k: Where should I post a patch for ili210x touchscreen? http://linux-sunxi.org/Submitting_patches |
| 14:24.13 | libv | 40 / (5 * 2.25) = 3.55 |
| 14:24.20 | libv | so we are actually doing slightly better than that |
| 14:24.51 | libv | also, the pp can only push out one pixel per Hz, and there is only 1pp available |
| 14:25.39 | Turl | libv: how much slower is it going to be on A13? |
| 14:25.59 | libv | depends more on clocking and on the memory subsystem than anything else |
| 14:26.25 | rastko | rz2k: ty |
| 14:26.27 | Turl | their memory subsystem was cut in half I think |
| 14:28.38 | libv | it's a mobile gpu, from back in the day when people were still selling predominantly 800x480 devices |
| 14:28.55 | libv | even the s2 had some meek resolution |
| 14:29.02 | Turl | libv: A13 has 16 bit SDRAM controller, A10 has 16/32 bit |
| 14:29.42 | Turl | libv: galaxy s 3 has a mali 400 |
| 14:29.49 | libv | heh, s2 came with 4xpp, and only had to use it to fill 800x480 |
| 14:30.24 | libv | we have a quarter of the pixel sucking and pushing power |
| 14:30.37 | libv | and a much worse memory controller no doubt |
| 14:31.01 | Turl | SGS3 has 720p screen |
| 14:31.23 | libv | so? |
| 14:31.26 | Turl | A1X tablets are in the >720p <1080p range usually |
| 14:31.31 | libv | still 4 PPs at 1 pixel per Hz |
| 14:31.32 | Turl | so yeah we have way less :) |
| 14:32.29 | libv | sgs3 is only clocked marginally higher |
| 14:33.51 | Mehhh | A10 = 4x M4 @ 250mhz ? |
| 14:34.23 | libv | Mehhh: once again you failed at reading |
| 14:34.39 | slapin | do anybody here have technical contacts at Omnivision? |
| 14:35.24 | *** join/#arm-netbook Jef91 (~Instantbi@bodhilinux/team/Jef91) |
| 14:35.35 | Turl | libv: so 1 pixel per Hz you say, that's like 320MPixel/s right? |
| 14:36.15 | libv | Turl: yeah, but it will not be fully occupied |
| 14:36.27 | libv | and there is a really good chance that it is simply bandwidth starved |
| 14:37.35 | Turl | yeah |
| 14:38.31 | Turl | libv: on an ideal world it'd be like 40% used only when painting a 1080p output |
| 14:39.21 | libv | Turl: i am not sure how MPixels are measured here, because the shaders for q3a are several instructions long |
| 15:43.14 | *** join/#arm-netbook avernos (~avernos@unaffiliated/avernos) |
| 16:12.54 | *** join/#arm-netbook avernos (~avernos@111.192.249.62) |
| 16:12.54 | *** join/#arm-netbook avernos (~avernos@unaffiliated/avernos) |
| 17:06.01 | *** join/#arm-netbook penguin42 (~dg@tu006.demon.co.uk) |
| 17:09.35 | *** join/#arm-netbook MMlosh (~MMlosh@2001:470:6f:23:705d:dfc0:738:8316) |
| 17:17.58 | *** join/#arm-netbook Jef91 (8a579be5@gateway/web/freenode/ip.138.87.155.229) |
| 17:31.46 | rz2k | libv: interestingly, new odroid boards have mali-400 mp4 clocked higher like the sgs3 too. |
| 17:34.06 | *** join/#arm-netbook ssspiff (~wim@d8D862466.access.telenet.be) |
| 17:34.07 | *** join/#arm-netbook ssspiff (~wim@unaffiliated/yukito) |
| 17:50.52 | *** join/#arm-netbook rellla (~Thunderbi@p4FE5632A.dip0.t-ipconnect.de) |
| 18:09.59 | *** join/#arm-netbook voronaam (~avorona@159.153.138.99) |
| 18:22.37 | *** join/#arm-netbook freakazoid0223 (~matt@pool-173-75-233-172.phlapa.fios.verizon.net) |
| 18:23.59 | *** join/#arm-netbook hg_5 (~chatzilla@unaffiliated/hg-5/x-8664886) |
| 18:40.45 | *** join/#arm-netbook vinifm (~vini@177.19.132.23) |
| 18:45.33 | vinifm | i2c client driver need to configure clk? |
| 18:51.07 | *** join/#arm-netbook madmalkav (~Hamlet@188.86.253.69) |
| 18:52.10 | madmalkav | hi |
| 18:52.53 | madmalkav | Allwinner a31 with quad a7 cores? what in the **** hell is an A7 arm core? how it compares with cortex a8, a9 and a15? |
| 18:53.13 | mnemoc | madmalkav: the a7 is the little brother of the a15 |
| 18:53.23 | madmalkav | oh, cool |
| 18:53.28 | madmalkav | haven't heard about that |
| 18:53.45 | mnemoc | it's a low power companion |
| 18:53.54 | mnemoc | same as the a5 |
| 18:53.56 | madmalkav | a15 for servers, a7 for mobile devices? |
| 18:54.09 | mnemoc | no. they are supposed to live in the same chip |
| 18:54.29 | mnemoc | switching from a15 to a7 when the load is low |
| 18:54.37 | madmalkav | aaah, yeah, I read about that |
| 18:54.45 | madmalkav | so, this is a cpu only made with those? |
| 18:54.47 | madmalkav | weird |
| 18:55.03 | mnemoc | the chinese found a cheap way of making "quad cores" |
| 18:55.12 | mnemoc | good for marketing |
| 18:55.32 | madmalkav | lkol |
| 18:56.14 | Turl | mnemoc: the new allwinner is big.LITTLE? o.O |
| 18:56.46 | mnemoc | no. it's only little |
| 18:56.54 | Turl | o.o |
| 18:58.01 | mnemoc | intented for cheap mobile... so it kind of makes sense |
| 18:58.13 | mnemoc | it should be better than a dual a9 |
| 18:58.26 | Turl | power wise, or perf wise? |
| 18:58.31 | mnemoc | both |
| 18:58.43 | mnemoc | less power, better perf |
| 18:58.45 | madmalkav | what's the status on linux support for the powervr gpu it uses? |
| 18:58.51 | mnemoc | no idea |
| 18:59.03 | Turl | it's all proprietary |
| 18:59.21 | Turl | remember the intel poulsbo fiasco? |
| 19:00.07 | Turl | well, that GPU was a PowerVR |
| 19:02.09 | mnemoc | doesn't TI use SGX544 in their omap5 too? |
| 19:02.14 | madmalkav | but there is a good propietary driver for linux ? |
| 19:03.00 | mnemoc | madmalkav: gpu drivers are very platform specific. so, no. |
| 19:03.13 | Turl | TI (usually?) uses PVR too, yeah |
| 19:03.35 | Turl | madmalkav: there's "good" support for android usually |
| 19:03.47 | madmalkav | fuck android |
| 19:06.39 | Turl | madmalkav: http://libreplanet.org/wiki/Group:PowerVR_drivers |
| 19:07.24 | madmalkav | so, we are as well as with the mali |
| 19:07.26 | madmalkav | :/ |
| 19:07.39 | *** join/#arm-netbook tinti (~tinti@bhe201062162119.res-com.wayinternet.com.br) |
| 19:07.58 | Turl | we have libv working on mali |
| 19:07.59 | mnemoc | no. for mali there are linux drivers and a RE project with progress |
| 19:08.05 | Turl | and the proprietary ones kinda work |
| 19:08.25 | Turl | I think there's no powervr proprietary driver for X/non-android fb |
| 19:08.51 | madmalkav | "kinda", that's the problem. |
| 19:09.01 | Turl | well it's better than no support at all :) |
| 19:09.02 | madmalkav | Beagleboard cannot run X? seems like it uses a powervr gpu |
| 19:10.24 | Turl | you can run X without any GPU |
| 19:10.33 | madmalkav | XD |
| 19:10.47 | madmalkav | ok, I meant "Accelerated X" |
| 19:11.07 | Turl | the mali X driver doesn't accelerate anything, correct me if I'm wrong |
| 19:11.23 | Turl | I think it just lets you use GL on top of X |
| 19:11.29 | Turl | rz2k: ^ |
| 19:11.36 | rz2k | yeah, it is standard dri2 pass thru |
| 19:11.58 | rz2k | hooks for 2d acceleration are there, but they do nothing |
| 19:12.08 | rz2k | we need g2d stuff be populated there |
| 19:12.23 | Turl | madmalkav: http://blog.qt.digia.com/blog/2009/11/20/building-qt-to-make-use-of-the-beagle-boards-sgx-gpu/ |
| 19:12.32 | Turl | apparently TI releases a SDK with blobs |
| 19:13.00 | Turl | they seem to run qt without X? :p |
| 19:13.17 | Turl | "deploy it and Qt to your board, and after shutting down X, run the example with |
| 19:13.18 | Turl | the following arguments:" |
| 19:13.19 | Turl | :P |
| 19:13.42 | mnemoc | blobs that probably only work on TI hardware |
| 19:16.15 | rz2k | qt can run on gles2 |
| 19:16.19 | rz2k | or XlibGL |
| 19:20.47 | rz2k | techn_: |
| 19:20.56 | techn_ | rz2k: |
| 19:22.05 | rz2k | regarding todays discussion in ML: r3p1 x11 libs seem to be doing better than r3p1 as they have needed depends and use libxcb (x server C bindings), might be that our dri2swapbuffers stuff fixed too |
| 19:22.19 | rz2k | they still miss that symbol though |
| 19:22.46 | rz2k | interesting thing is that I was able to launch linaros default XBMC with r3p1 x11 libs |
| 19:22.57 | techn_ | yeah.. It would be interesting to see benchmark results with r3p1 too :) |
| 19:23.13 | rz2k | I was thinking about that too |
| 19:23.27 | rz2k | maybe framebuffer ones get some fixes too |
| 19:24.03 | techn_ | what I remember.. there was some task scheduling fixes @ driver level |
| 19:24.30 | techn_ | but who knows.. could be even worse :p |
| 19:26.04 | rz2k | there's new dma stuff there atleast https://github.com/linux-sunxi/mali-400-kernel-drivers/commit/a6dea94fe2da2cf1ce58828a7361ee063d09f0bd#L37R1 |
| 19:26.54 | techn_ | regarding sata performance.. working dma? |
| 19:28.09 | techn_ | and can it be applied to mali too? |
| 19:28.35 | hno | techn_? |
| 19:30.36 | techn_ | hno: that guy @ ML said that if cpu clocks are at 1Ghz sata gives 50MB/s.. otherwise 16MB/s |
| 19:32.28 | *** join/#arm-netbook lerc (~quassel@121-74-245-234.telstraclear.net) |
| 19:53.05 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
| 19:59.42 | *** join/#arm-netbook pwhalen (~paul@CPE001310360dac-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 20:11.04 | *** join/#arm-netbook penguin42 (~dg@tu006.demon.co.uk) |
| 20:11.04 | *** join/#arm-netbook lundman (~Accident@194.109.206.190) |
| 20:11.04 | *** join/#arm-netbook robclark (~robclark@ec2-50-19-223-142.compute-1.amazonaws.com) |
| 20:11.04 | *** join/#arm-netbook Brandon15811_ (~Brandon15@ec2linux.exaktop.tk) |
| 20:12.24 | hno | techn_, that applies to anything that does CPU bursts and then wait for something. |
| 20:12.46 | hno | even a little wait makes ondemand think you don't really need cpu power. |
| 20:25.46 | *** join/#arm-netbook ared_ (~ared@host222-179-dynamic.2-87-r.retail.telecomitalia.it) |
| 20:28.10 | *** join/#arm-netbook ZaEarl (~malmrose@c-67-172-235-190.hsd1.ut.comcast.net) |
| 20:28.50 | ZaEarl | how's our a13 kernel? |
| 20:29.14 | *** join/#arm-netbook ared__ (~ared@host222-179-dynamic.2-87-r.retail.telecomitalia.it) |
| 20:30.55 | *** join/#arm-netbook vinifm (~vini@177.19.132.23) |
| 20:30.56 | techn_ | ZaEarl: what you mean? |
| 20:31.20 | ZaEarl | does the code in github for the kernel run well on some a13 tablets? |
| 20:31.29 | techn_ | for me it atleast boots.. although I havent tried .52 kernel |
| 20:31.47 | techn_ | also touchscreen works |
| 20:32.10 | techn_ | and I have that a13_mid tablet |
| 20:34.05 | *** join/#arm-netbook ared_ (~ared@host253-179-dynamic.2-87-r.retail.telecomitalia.it) |
| 20:36.20 | rz2k | damn, I have it too, still didnt even tried to boot it |
| 20:36.40 | rz2k | only found UART |
| 20:49.22 | hno | ZaEarl, the A13 kernel also runs fine on Olimex A13 OLinuXino boards. |
| 20:54.59 | *** join/#arm-netbook orly_owl (~david@unaffiliated/orly-owl/x-3167833) |
| 20:57.35 | *** join/#arm-netbook hark (~hark@w.v89.eu) |
| 20:57.38 | techn_ | libv: No framebuffer libraries exist for UNKNOWN_VERSION, armhf. |
| 20:57.48 | *** part/#arm-netbook hark (~hark@w.v89.eu) |
| 20:57.48 | techn_ | with make framebuffer |
| 21:02.07 | rz2k | techn_: load mali module |
| 21:02.16 | rz2k | unknownversion = ioctl failed |
| 21:02.59 | techn_ | it's loaded |
| 21:04.33 | techn_ | but version doesn't output anything |
| 21:12.49 | rz2k | interesting |
| 21:12.59 | rz2k | what setup do you have? |
| 21:24.18 | mnemoc | github finally added attachments for the issues |
| 21:30.28 | *** join/#arm-netbook popolon (~popolon@og-free.planet-service.fr) |
| 21:48.37 | Mehhh | which version does the bsp way pull in by default right now? |
| 21:49.46 | techn_ | .42 version if no one has updated to .52 |
| 21:50.21 | techn_ | check .gitmodules |
| 21:50.37 | Mehhh | i meant malis |
| 21:50.44 | Mehhh | sorry i should've said that |
| 21:51.06 | techn_ | r3p0 |
| 21:51.54 | Mehhh | Crap, but p1 is looking better according to rz? |
| 21:52.08 | Mehhh | Well, for X11 |
| 21:52.23 | techn_ | yes.. but we only have x11/framebuffer libs for r3p0 |
| 21:52.30 | techn_ | oh.. sorry.. r3p1 |
| 21:53.08 | Mehhh | I suppose a mix-and-match wouldn't turn out so well ;) |
| 21:53.58 | techn_ | it can be done but.. it's complicated |
| 21:54.17 | rz2k | someone with arm asm knowledge, please tell me what blx DRI2SwapBuffers; DRI2SwapBuffers: ADRL R12, 0xBD81C; LDR PC, [R12,#(DRI2SwapBuffers_ptr - 0xBD81C)]!; means exactly? |
| 21:54.38 | techn_ | you'll need kernel which has r3p1 mali module |
| 21:54.38 | rz2k | blx is branch-link-execute afaik |
| 21:55.55 | rz2k | 0xbd81c is _gles_bind_buffer+1 |
| 21:56.08 | Mehhh | Yeh, I guess i only had defaults on for that |
| 21:56.17 | rz2k | questions is how to get rid of dri2swapbuffers import :/ |
| 21:56.24 | Mehhh | yuep, p0 from the cnx video page :\ |
| 22:00.17 | Turl | rz2k: if you put it on a pastebin I'll let you know |
| 22:01.09 | ssvb | rz2k: this is likely a .plt section, which contains code thunks which do actual function calls via pointers |
| 22:02.37 | ssvb | rz2k: are you sure nobody is using this function and you really don't need it? |
| 22:03.30 | rz2k | we have dri2swapbuffers in xorg driver as malischeduleswap |
| 22:03.41 | rz2k | and dri2proto gets the pointer there |
| 22:03.54 | rz2k | and as I understood, it really works |
| 22:04.08 | rz2k | so import in library is useless |
| 22:05.10 | ssvb | what happens if you patch the binary and replace the "DRI2SwapBuffers" string with some other function name which is guaranteed to be available? |
| 22:05.11 | Turl | can't you just make a dummy definition/alias then? |
| 22:06.02 | ssvb | I'm not that very familiar with ELF format though, so don't know if it is going to work (maybe the function names must be sorted, etc.) |
| 22:07.41 | ssvb | is /etc/ld.so.preload with some dummy library which has an empty stub for this function not an option? |
| 22:07.42 | rz2k | I will find that out soon |
| 22:15.02 | ssvb | rz2k: or maybe try something like "gcc -fPIC -shared -o new.so old.so file_with_dummy_DRI2SwapBuffers_function.c" |
| 22:15.36 | mnemoc | http://www.allwinnertech.com/en/product/A20.html |
| 22:15.51 | mnemoc | this one has mali400 |
| 22:18.43 | mnemoc | "In addition, A20 is pin-compatiblewith its predecessor single-core A10" |
| 22:21.11 | Mehhh | omgggggggggg |
| 22:21.52 | Mehhh | reflow onto the A10 slot? ;0 |
| 22:22.22 | rz2k | lol |
| 22:22.52 | rz2k | mnemoc: create topic @ cubieboard ML about this |
| 22:22.57 | rz2k | maybe Tom will get more info |
| 22:23.00 | hno | A20? |
| 22:23.13 | mnemoc | allwinner published A31 and A20 today |
| 22:23.32 | mnemoc | A20 is not in the menu, but linked from the "application" pages |
| 22:23.50 | bfree | no sata in either that I see |
| 22:24.06 | hno | bfree, sata is not mentioned in A10 briefs either. |
| 22:25.07 | hno | Mail400 MP2. |
| 22:26.12 | mnemoc | pin compatibility implies feature compatibility |
| 22:26.34 | hno | And they link to Cubieboard and OLinuxIno |
| 22:26.55 | mnemoc | also the hackberry :p |
| 22:27.04 | Mehhh | LOL |
| 22:27.05 | hno | mnemoc, kind of implies feature compatibility, but maybe not all features. |
| 22:27.06 | Mehhh | it says it has |
| 22:27.09 | Mehhh | 512MB L2 cache |
| 22:27.12 | Mehhh | on the infographic |
| 22:27.52 | mnemoc | wiki wiki |
| 22:28.28 | hno | http://www.allwinnertech.com/en/apply/home.html#android-box |
| 22:28.40 | hno | mentions SATA. |
| 22:28.58 | hno | also says that A20 is quad-core, while A20 page says dual-core. |
| 22:29.36 | hno | so still some typos here and there, likely to stay for some years. |
| 22:31.54 | Mehhh | At least it doesn't say AMD Z130 GPU |
| 22:32.11 | Turl | Mehhh: intel will go bankrupt ;) |
| 22:32.21 | Mehhh | ;) |
| 22:32.39 | Mehhh | Atom by Atom, they will crumble? |
| 22:33.01 | Turl | hno: they also say "4Kx2K" and "2160P" on the others |
| 22:33.03 | mnemoc | first moore's law, then moore's company |
| 22:34.11 | hno | Turl, yes, and as we know what they mean with 2160P is 1080P 3D. |
| 22:34.41 | Turl | lol, "visual bombox" |
| 22:36.58 | hno | There also is an A12 it seems. http://www.allwinnertech.com/en/product/A12.html |
| 22:37.21 | hno | Or... page says A10s. |
| 22:38.30 | rz2k | looks @ mali.so disasm and wonders why the hell mali400 library has functions and stuff for mali200 and mali300(wtf?) |
| 22:39.00 | Mehhh | So, the A20 doubles the L1 per core, but keeps the L2 @ 256/ea? |
| 22:40.05 | rz2k | also, I'm starting to suspect that libraries can be used for different SoCs, as register addresses and stuff are in kernel module. |
| 22:40.27 | rz2k | calls libv to the rescue |
| 22:40.37 | hno | Mehhh, Unless it's shared L2. |
| 22:40.54 | Mehhh | True |
| 22:41.00 | mnemoc | Turl tested mali libs against a mali module with os-allocated memory and the libs crashed badly |
| 22:42.00 | rz2k | that hwmem stuff from ux500/snowball? |
| 22:42.46 | hno | Ok, 2160P vs 4Kx2K is explained. HD H.264 2160P video decoding / UHD H.264 4K ×2K @30fps video decoding |
| 22:42.50 | mnemoc | our libs, our kernel. but mali's config.h changed to use OS_MEMORY instead of a hardcoded address |
| 22:43.03 | Mehhh | Would I be looked down upon if I made an easier BSP / helper script for kernel / fs / tc / img creation? |
| 22:43.28 | hno | Mehhh, it's too complicated today? |
| 22:43.39 | mnemoc | Mehhh: why not helping to improve the current instead of making another? |
| 22:43.44 | Mehhh | not really, but in terms of selecting |
| 22:44.02 | Mehhh | i noticed the bsp didn't have the mk802 unless i moved the fex, etc |
| 22:44.02 | hno | selecting what? |
| 22:44.09 | Mehhh | modules without digging into cfg |
| 22:44.47 | Mehhh | or being able to have it git into an exact structure and choose the malis for instance on the fly |
| 22:45.08 | *** join/#arm-netbook alcides (~alcides@187.113.9.169) |
| 22:45.08 | *** join/#arm-netbook alcides (~alcides@unaffiliated/alcides) |
| 22:45.22 | Mehhh | errr dir / file tree, not structure |
| 22:45.56 | Mehhh | If it's set, i can release a python / qt4 script to the forums that uses the BSP |
| 22:46.11 | hno | nothing is set in stone. |
| 22:46.17 | Mehhh | Nice :) |
| 22:46.21 | Mehhh | Wasn't sure if against the rules |
| 22:46.59 | hno | do we have rules? |
| 22:47.08 | mnemoc | not that I know.... |
| 22:47.30 | Mehhh | If you have no rules, is that a rule? ::: brain explodes :::: |
| 22:47.54 | mnemoc | there is no rule about not having rules |
| 22:48.19 | hno | settles for we have no rules defined. |
| 22:59.21 | Mehhh | I think i found a bsp bug but i'll fix it. I don't think it's importing the mk802 fex on ./configure # |
| 22:59.38 | Mehhh | I could be wrong and just sleepy :P |
| 23:01.01 | mnemoc | that probably means sunxi-boards doesn't have an mk802.fex |
| 23:01.33 | Mehhh | weird, mine does |
| 23:02.10 | techn_ | a20 is most interesting of those releases :p |
| 23:02.36 | mnemoc | techn_: but hno is right. without sources, even if it's pin-compatible, it's useless |
| 23:02.43 | techn_ | if it's mostly a10/a13 compatible |
| 23:03.01 | mnemoc | sun5i SDKs are a pile of bins |
| 23:03.17 | mnemoc | nothing hints that will be improved in sun6i |
| 23:04.01 | mnemoc | sometimes I wish to be lawyer to be able to sue companies at will :| |
| 23:04.33 | Turl | mnemoc: well, on the other side they listed cubieboard, hackberry etc on their official site |
| 23:04.48 | mnemoc | Turl: yes, that was unexpected |
| 23:04.53 | techn_ | I think after success of this a10/a13 project, receiving sources wouldn't be that hard |
| 23:04.57 | Mehhh | Ok so if i clone the boards git i do see the mk802 fex but not if i do the bsp |
| 23:05.07 | Turl | the least they could do is hand tom the source if he does a cubieboard2 |
| 23:05.48 | techn_ | .. I also think this project is currently at start.. even though hw starts to be aging |
| 23:06.08 | mnemoc | Mehhh: that's a problem with git submodules, you marry a commit... not a branch :| |
| 23:06.27 | Mehhh | So i guess i'll just upload it to the bsp |
| 23:06.49 | mnemoc | Mehhh: cd sunxi-boards; git checkout master; git pull --rebase |
| 23:07.20 | techn_ | Mehhh: or make update :) |
| 23:07.26 | Mehhh | ;) |
| 23:07.47 | mnemoc | i'm not so sure `make update` does what's intended |
| 23:08.05 | mnemoc | hates git submodules |
| 23:08.17 | techn_ | it should checkout current head and rebases it |
| 23:08.25 | Mehhh | you can't have them git from each other? |
| 23:08.27 | hno | techn_, Olimex have never received usable SDK sources for A13. |
| 23:08.55 | Mehhh | I do that with my linux configs all the time on git. Use a git to pull a submod to another git |
| 23:08.58 | mnemoc | techn_: it most likely updates up to the commit it the submodule recort, not up to the head |
| 23:09.17 | Mehhh | ahh |
| 23:09.41 | Mehhh | I'm not trying to be a pain, just trying to streamline this |
| 23:10.31 | techn_ | mnemoc: yeah.. It could have some problems.. feel free to fix them ;) |
| 23:12.48 | Mehhh | So are the fex files pulled or kept sep in the bsp? |
| 23:13.11 | mnemoc | Mehhh: hold |
| 23:14.28 | techn_ | Mehhh: you'll need to add mk802 stuff to u-boot |
| 23:14.30 | mnemoc | I just pushed a sub-module update |
| 23:14.39 | Mehhh | awesome |
| 23:14.59 | Mehhh | oh yeh crap, you're right techn_ |
| 23:15.49 | hno | Mehhh, not hardto do. |
| 23:15.51 | rz2k | you are f@cking kidding me.. exynos default kernel has ours mali_drm |
| 23:15.58 | rz2k | 1:1 line |
| 23:15.59 | Mehhh | ROFL |
| 23:16.07 | rz2k | s/exynos/hardkernel/ |
| 23:16.09 | techn_ | mnemoc: how's your work.. when you'll have free time? ;) |
| 23:16.18 | mnemoc | when I quit |
| 23:16.33 | rz2k | the one that i've fixed looking at snowball old sources that were for 2.6 drm |
| 23:16.45 | techn_ | also hno seems AFK quite alot :p |
| 23:17.01 | hno | yes. not doing IRC at work. |
| 23:17.31 | hno | only evenings after kids @bed, or on my way to/from work. |
| 23:17.40 | hno | in principle. |
| 23:18.09 | mnemoc | envies the "at work"/"at home" difference |
| 23:18.20 | techn_ | hno: same |
| 23:18.57 | techn_ | mnemoc: I have opportunity to work @ home.. but just becouse of that I wont do it :p |
| 23:19.02 | Mehhh | The mk802ii should be compatible with mk802, correct? |
| 23:19.06 | Mehhh | uboot |
| 23:19.13 | Mehhh | so I could just clone it and rename |
| 23:19.47 | techn_ | Mehhh: mele is most likely more compatible.. or how much memory your device has? |
| 23:19.52 | Mehhh | 1GB |
| 23:20.06 | Mehhh | oh crap i forgot about that, so i'll have to make two new folders |
| 23:20.10 | *** join/#arm-netbook ared_ (~ared@host199-175-dynamic.4-87-r.retail.telecomitalia.it) |
| 23:20.17 | Mehhh | 512 / 1GB |
| 23:20.20 | mnemoc | Mehhh: use a10-meminfo, make a good script.fex, and I make the u-boot config for you |
| 23:20.55 | Mehhh | nice |
| 23:22.40 | mnemoc | techn_: the most annoying part in my case is the TZ diff... so my work "shift" is from 3pm to midnight... but assumed to have worked from 9am to 3pm too, and also expected to be avaialble some ours after their 6pm (my midnight) if a customer has troubles |
| 23:22.52 | Mehhh | https://github.com/maxnet/a10-meminfo right? |
| 23:22.56 | mnemoc | Mehhh: yes |
| 23:23.12 | mnemoc | s/ours/hours/ |
| 23:24.33 | hno | can't work@home. Nothing gets done. |
| 23:24.41 | Mehhh | ok so just compile from the device? |
| 23:24.48 | Mehhh | or is this meant over serial |
| 23:25.03 | mnemoc | Mehhh: run the static bin with su in android |
| 23:25.15 | Mehhh | oh, android. ok cool |
| 23:25.53 | hno | mnemoc, so bill them for 9am to midnight. |
| 23:27.11 | mnemoc | hno: can't... it's fixed, weekly... as long as they are pleased :| |
| 23:27.15 | Mehhh | hno, i envy your billing system on your website |
| 23:27.31 | mnemoc | hno: and there is zero chance to get a local job in this f* country |
| 23:27.36 | Mehhh | If I tried something like that here they would laugh |
| 23:29.49 | techn_ | mnemoc: In my company there is ppl at TZ +10 to -10.. so no one expects ppl to answer right away |
| 23:30.17 | *** join/#arm-netbook freakazoid0223 (~matt@pool-173-75-233-172.phlapa.fios.verizon.net) |
| 23:31.28 | mnemoc | techn_: in my case I'm the only "offshore" :| |
| 23:32.25 | hno | Mehhh, You mean my paypal form? Most payments are sent using bank transfer. |
| 23:32.38 | penguin42 | http://www.globalscaletechnologies.com/p-58-mirabox-development-kit.aspx is nice |
| 23:32.53 | Mehhh | Maybe, i could've sworn it was a CC one |
| 23:33.07 | Mehhh | My friend and i were like damn... what a baller |
| 23:33.43 | penguin42 | actually, that's very nice: http://www.globalscaletechnologies.com/p-58-mirabox-development-kit.aspx |
| 23:33.46 | hno | Mehhh, no, it's only a standard paypal form. All CC processing is done by paypal. |
| 23:33.50 | Mehhh | ah |
| 23:33.57 | hno | including entry of cc details. |
| 23:34.23 | techn_ | mnemoc: how about job from ie. norway.. annual salary over 100k.. If you are that good to apply from offshore? :) |
| 23:34.49 | Mehhh | ironically, Norway has an awesome movie named Headhunters |
| 23:35.55 | penguin42 | Mehhh: Is it gory and involve erm decapitation? |
| 23:36.07 | Mehhh | Shooting mostly |
| 23:36.18 | Mehhh | it's about a recruiter / head-hunter |
| 23:36.27 | mnemoc | techn_: it's too burned out atm.. need to finish the current project and try to recover my brain first |
| 23:36.29 | penguin42 | boring, I thought it might be a play on words |
| 23:37.13 | mnemoc | techn_: s/it's/i'm/ |
| 23:38.09 | mnemoc | techn_: I assume the northern "linux market" was overflooded by former nokia people |
| 23:38.17 | hno | not really. |
| 23:38.52 | techn_ | mnemoc: hno: yeah.. since most of the ppl was symbian ppl.. like I |
| 23:39.19 | techn_ | luckily I found job.. but over 1000 didn't |
| 23:39.40 | rz2k | once tried to write symbian apps, got lost in million of enum's |
| 23:40.07 | hno | there is big needs for Linux people at all levels in Stockholm. |
| 23:40.20 | Mehhh | ./moves |
| 23:40.21 | mnemoc | .oO |
| 23:41.22 | techn_ | mnemoc: but you are true.. at last summer there was kicked out over 1000 linux gurus, but I think most of them was hired quite fast |
| 23:41.37 | hno | 1000 is not very many. |
| 23:42.07 | techn_ | for Finland it is :p |
| 23:42.19 | techn_ | only 5 million ppl lives in finland |
| 23:42.32 | Mehhh | and 10 million guns |
| 23:43.16 | hno | And ~10M people in Sweden. |
| 23:43.51 | techn_ | Mehhh: that could be true.. moose and duck hunting is quite popular here :p |
| 23:44.06 | Mehhh | At least in Sweden the greeting sounds like "hey" |
| 23:44.18 | hno | Hej |
| 23:44.32 | techn_ | Terve/Moro/Hei :) |
| 23:47.39 | techn_ | Buy anyway.. currently atleast two cities has been declared as cities at emergency state due Nokias situation |
| 23:48.13 | techn_ | (same as with Saab at Sweden) |
| 23:48.23 | mnemoc | techn_: you can't bet 27% of general unemployment and >80% in the under-30 segment |
| 23:48.48 | mnemoc | spain is doomed |
| 23:49.16 | techn_ | mnemoc: yep.. that's weird with your country.. and other mediterrian countries |
| 23:49.39 | Mehhh | Doesn't matter. Visa estimates that only 3-6% of modern countries will "need" to work in 5-10 years |
| 23:49.42 | Mehhh | due to automation |
| 23:50.01 | mnemoc | spaniards are lazy as hell |
| 23:50.10 | Mehhh | tell me moor |
| 23:50.15 | techn_ | mnemoc: your politics are draining all the money? |
| 23:50.37 | mnemoc | currently banks are the biggest drain |
| 23:51.12 | mnemoc | but before coming to spain (fooled by my current remote-employer) I thought the laziness was related to the heat |
| 23:51.51 | mnemoc | but this region has a similar weather than uk or ie... and people is just as lazy as those in the hot parts |
| 23:51.54 | penguin42 | (Is there anywhere that isn't doomed?) |
| 23:52.22 | penguin42 | mnemoc: Hang on, there can't be somewhere in spain with as bad weather as us surely |
| 23:52.30 | Mehhh | penguin, south korea |
| 23:52.51 | penguin42 | ah |
| 23:53.35 | mnemoc | penguin42: in my city it rains all year. summer never passes 30C and winter never passes under 5C :) |
| 23:53.38 | mnemoc | penguin42: I love it |
| 23:53.49 | mnemoc | penguin42: but commerce still closes from 2pm to 5pm! |
| 23:53.52 | mnemoc | for a nat |
| 23:53.53 | mnemoc | nap |
| 23:53.58 | mnemoc | come on |
| 23:54.07 | penguin42 | mnemoc: Yeh doesn't sound much off the UK, except here we might go down to -5c on a bad night in the middle of winter |
| 23:54.09 | mnemoc | and nothing opens on saturdays or sundays |
| 23:54.33 | penguin42 | mnemoc: That's bad - in a big city in the UK there's something open most of the time |
| 23:54.53 | techn_ | here has been -20 for this week :) |
| 23:55.11 | mnemoc | penguin42: i only mean that in hot places I understand people been slow, and sleeping during the hotest hours |
| 23:55.12 | techn_ | first snow came 2 months ago.. but it melt :p |
| 23:55.22 | techn_ | and came again this week :) |
| 23:56.28 | penguin42 | mnemoc: Yeh I guess so, I've never been anywhere that is like that |
| 23:59.22 | mnemoc | try to go on summer to the south of italy or spain... it's beautiful, but the brain melts |
| 23:59.28 | techn_ | .. anyway euro seems to be doomed.. greece most likely wont pay their debt never.. they are just raising thei salaries.. after that spain and italy will fall |
| 23:59.36 | techn_ | then france |