| 23:59 | unununium_ | Ah, thank you very much! |
| 00:01 | lioka | 'author' there shold be 'committer' at most :] |
| 00:05 | unununium_ | Now I begin feeling stupid... but I will find the place where the patch will help me... |
| 00:09 | unununium_ | Oh. I begin to learn. |
| 00:10 | unununium_ | (Maybe I just begin to think... ;) ) |
| 00:14 | unununium_ | Hmm. Well. I *think* I applied the patch. |
| 00:15 | unununium_ | ("error: 'FOURCC_RGBA32' undeclared (first use in this function)") But... google might help me. :) |
| 00:16 | lioka | try again with -DMRVL_XV_SUPPORT_RGB_FORMAT |
| 00:18 | lioka | actually, there is a lot of defines, look at debian/rules |
| 00:20 | unununium_ | Ah, yes, that seems to have solved the problem, but not all. I think having a look there might be helpful. |
| 00:21 | unununium_ | Holy moly... compiling seems to be successful... |
| 00:22 | cbxbiker61 | Holy moly? you must be in US |
| 00:22 | unununium_ | No, just watched Simpsons to often... |
| 00:23 | _rmk_ | or Die Hard |
| 00:23 | unununium_ | Haven't seen that, sorry. ;) |
| 00:26 | unununium_ | Well. It is compiled. It is installed. And it doesn't tell me, that it was compiled for the wrong version. So far it is a success! :) |
| 00:26 | unununium_ | But... in fact... "Segmentation fault at address 0x18" when starting Xorg is not exactly what I expected. ;) |
| 00:45 | unununium_ | Well... I think I'll just give up for today. |
| 00:46 | unununium_ | But thank you very much for your help, lioka! |
| 00:47 | unununium_ | Welcome back, cbxbiker61. |
| 00:50 | cbxbiker61 | my synaptics touchpad goes wonky periodically |
| 00:51 | cbxbiker61 | i don't know what triggers it |
| 00:55 | unununium_ | Oh... that's not exactly what you want... |
| 00:56 | unununium_ | I... must... stop... looking... for... my... problem! Have... to... get... up... early! |
| 02:04 | unununium_ | Finally I'll go to bed right now. :) cbxbiker61, lioka: Thank you very, very much for your help! :) And have a good night! |
| 02:04 | unununium_ | (Oh, to answer your question: No, I was not successful yet.) |
| 08:36 | MikeSeth | rabeeh: oh, I wanna learn the arm architecture, perhaps write a mini-OS |
| 09:07 | unununium_ | Good morning... |
| 09:07 | unununium_ | lioka: I got the module working. Having a look at debian/rules was a good idea! *thumbs up!* |
| 09:08 | unununium_ | (But still it is not accelerated. But that is something I'll take care about later.) |
| 09:18 | jnettlet | unununium_, saw your messages last night. I am almost done with a new driver. Maybe today or tomorrow. I started with the IMX support and am now bringing in armada. |
| 11:05 | rabeeh | linxbmc will support cubox-i too |
| 11:06 | rabeeh | MikeSeth: if you want to learn ARM hardcore; try developing a boot loader :) |
| 11:06 | MikeSeth | I do :P |
| 11:06 | rabeeh | like porting this - https://github.com/alexandrebelloni/whoosh |
| 11:07 | rabeeh | they were able to boot a GL application in 600mSec |
| 11:07 | MikeSeth | rabeeh: I'm thinking this toy: http://iqmedia.co.il/2011-03-02-08-35-22.html?page=shop.product_details&flypage=flypage.tpl&product_id=277&category_id=47 |
| 11:08 | rabeeh | i played a bit with MCUs |
| 11:08 | rabeeh | i enjoy X rather than counting image size in bytes |
| 11:12 | MikeSeth | well this is a toy, i want to write a forth for it |
| 11:24 | jnettlet | rabeeh, I was looking at whoosh last week. looks pretty cool |
| 11:24 | rabeeh | jnettlet: xbmc in 2 seconds :) |
| 11:24 | rabeeh | thats the target |
| 11:24 | jnettlet | could be very cool |
| 11:25 | rabeeh | they have a long presentation about that; mainly they shrunk boot loader + uImage + rootfs for ~5MB |
| 11:25 | rabeeh | so they don't need to load too much binaries. |
| 11:25 | rabeeh | probably they used also other library optimizers to shrink it down to really the functions that are using in the rootfs |
| 11:26 | rabeeh | but 600msec is very impressive |
| 11:27 | jnettlet | and their bootloader is just doing the bare minimum. Init dram and then bootstrap the kernel from sdhc |
| 11:27 | jnettlet | don't double probe things |
| 11:30 | rabeeh | elinux.org/images/d/d1/Alexandre_Belloni_boottime_optimizations.pdf? |
| 11:34 | jnettlet | yeah the newer hotplug smp and memory functionality will be great for fast booting. |
| 11:34 | unununium_ | jnettlet: Thanks for your reply! :) That is very good to hear! |
| 11:34 | jnettlet | unununium_, this is for the original Cubox. Running what distro? |
| 11:35 | unununium_ | rabeeh: Will Debian work (including 3D-Support) with Cubox-i? (I ordered mine the day before yesterday) |
| 11:35 | unununium_ | jnettlet: Debian jessie. |
| 11:35 | jnettlet | okya |
| 11:35 | rabeeh | unununium_: should |
| 11:35 | unununium_ | rabeeh: Great! :D Hope it will arrive within this year. :) |
| 11:36 | unununium_ | (I live in germany...) |
| 11:36 | jnettlet | wow a lot of germans in the channel recently |
| 11:36 | rabeeh | the basic stuff is for sure (i.e. use armhf binaries); the 3D stuff is more trickier since it requires a smart developer that knows how to twist X and force it to use xegl |
| 11:36 | unununium_ | How are taxes handled? Will I be informed? |
| 11:36 | rabeeh | unununium_: which one did you order? |
| 11:37 | jnettlet | I guess OctoberFest is over so it is time to sober up and hack on some hardware |
| 11:37 | unununium_ | rabeeh: The i4-pro |
| 11:38 | unununium_ | jnettlet: I've never been there... in fact I'm not much in drinking beer... |
| 11:38 | rabeeh | regular mail or dhl? |
| 11:38 | unununium_ | Regular, I think. |
| 11:38 | unununium_ | I did not choose any advanced options. |
| 11:38 | rabeeh | i can ask the shipping guys; but typically it takes few weeks to arrive to Germany |
| 11:39 | rabeeh | sometimes they ask for you to pay customs; but if it's under 100eu typically they don't |
| 11:39 | unununium_ | Yes, that was already mentioned somewhere... it was written that it could take up to eight weeks. |
| 11:39 | rabeeh | we will probably have lots of inputs end of Nov about shipping and people's experience how it's delivered |
| 11:39 | unununium_ | At all I was at about 110,- including shipping. |
| 11:39 | rabeeh | i don't know why Germany takes so long |
| 11:39 | unununium_ | Maybe I'll one of the people reporting. :) |
| 11:40 | rabeeh | we always had issues there |
| 11:40 | unununium_ | I didn't want to blame you for that... |
| 11:40 | rabeeh | i think dhl took over the German post :) |
| 11:40 | unununium_ | I think they are somehow one company... but I'm not sure about that... |
| 11:40 | rabeeh | unununium_: if it will be late then please accept my apologies from now :) |
| 11:40 | rabeeh | it's something that we can't really control |
| 11:40 | unununium_ | rabeeh: I do, of course. ;) It's not your fault. |
| 11:41 | unununium_ | Ah, but I have another problem... wanted to register at the forums, but I don't get any activation mail... |
| 11:41 | rabeeh | oh; if it's the same company then dhl employees work for the dhl shipments and at the end of the day; if they have some time then they will deliver the regular mail :) |
| 11:41 | rabeeh | where? |
| 11:41 | rabeeh | imx.solid-run.com/forums? |
| 11:42 | unununium_ | Exactly. |
| 11:42 | unununium_ | No. Not exactly. |
| 11:42 | unununium_ | Wait. ;) |
| 11:42 | rabeeh | which nickname did you choose? |
| 11:42 | unununium_ | Give me one second, I was wrong about that... |
| 11:42 | unununium_ | http://www.solid-run.com/phpbb/ |
| 11:43 | rabeeh | oh; this is the forums for the older cubox |
| 11:43 | rabeeh | maybe we have issue there; but for now please try registering at imx.solid-run.com/forums |
| 11:43 | unununium_ | I know, but I'm playing around with that at the moment. |
| 11:46 | unununium_ | A friend of mine gave me his CuBox because he bricked it... that way I got in contact with this device. :) |
| 11:51 | unununium_ | And I think I'll need a mircoSD with more than 4GB. ;) |
| 11:51 | unununium_ | Maybe I'll install jessie onto the 8GB... |
| 11:55 | unununium_ | Well... FPS improved from about 17 to 25. But the reason might be E17 -> Fluxbox, as E17 no longer wants to start. |
| 12:03 | Bluerise | jnettlet: I haven't really been at the Oktoberfest... Been hacking the whole time. ;) |
| 12:04 | jnettlet | What kind of Germans do we have in this channel? |
| 12:04 | jnettlet | Next thing you know I will find out they are all vegetarians |
| 12:04 | Bluerise | One Bavarian, who isn't doing what he's supposed to do. (Drinking beer all the time.) |
| 12:05 | Bluerise | Heh. |
| 12:06 | Bluerise | Well, I like Formula 1 and Golf. Usually one should be interested in Football instead. (Soccer, for the Americans.) |
| 12:06 | jnettlet | I have adopted European Football, I think it is a fair compromise |
| 12:07 | _rmk | 12:07 * _rmk_ prefers to think of soccer as a game of socks, particularly when you find that you have one less than you're supposed to. :) |
| 12:08 | liok | 12:08 * lioka knows that game |
| 12:08 | jnettle | 12:08 * jnettlet thinks his power supply might be dying. My displayport on my nvidia card keeps blinking out on me. |
| 12:14 | unununium_ | Bluerise: Me isn't interested in soccer neither... |
| 12:17 | Bluerise | But, all Germans are! (Apart from us apparently.) |
| 12:18 | _rmk | 12:18 * _rmk_ is also one brit who has no interest in "the beautiful game" |
| 12:18 | unununium_ | jnettlet: To answer your question: On is coming from the ,,Ruhrgebiet". (Is there a translation for that? Well. It's North Rhine-Westphalia...) |
| 12:19 | unununium_ | Bluerise: Obviously. ;) |
| 12:19 | Bluerise | The dictionary translates Ruhrgebiet to Ruhr ;) |
| 12:19 | unununium_ | Hehe. ;) |
| 12:20 | unununium_ | Although that's not exactly the same... but... almost... |
| 12:20 | Bluerise | _rmk_: Any news on the Device Tree thingy? |
| 12:22 | _rmk_ | and DRM? Grant has said he'll write up what he sees as the outcome of the discussion in the next few days |
| 12:22 | Bluerise | and GPL ;) |
| 12:22 | _rmk_ | oh. |
| 12:22 | jnettle | 12:22 * jnettlet has yet to venture down into Germany. Should do that at some point. |
| 12:22 | _rmk_ | not heard back from Grant yet - I think he's still conferencing with linaro connect this week. |
| 12:23 | Bluerise | ok. :) |
| 12:23 | Bluerise | Will still start with Linux as reference. Just not push it somewhere. |
| 12:23 | Bluerise | (When I have time...) |
| 12:31 | jnettlet | rabeeh, do you have a build env for Android on the cubox-i accessible anywhere? |
| 12:31 | rabeeh | just finished one now |
| 12:32 | rabeeh | my core i7 fans were on full speed |
| 12:32 | rabeeh | want it? |
| 12:32 | rabeeh | or let me test first and then ask you if you want it :) |
| 12:32 | jnettlet | rabeeh, sounds good. |
| 12:32 | rabeeh | we are about to launch a boot logo contest for android |
| 12:33 | rabeeh | boot animation |
| 12:33 | jnettlet | any chance you could test u-boot for the other Microsom's today? |
| 12:33 | jnettlet | oh nice. probably won't have time to play with that. |
| 12:33 | rabeeh | http://forum.xda-developers.com/showthread.php?t=2334927 |
| 12:33 | jnettle | 12:33 * jnettlet will probably port LiquidSmooth's rom over as that is what I run on most everything. |
| 12:33 | rabeeh | jnettlet: i'v tested on the quad; but it doesn't work |
| 12:34 | rabeeh | not that i like those boot animations; but those are examples |
| 12:34 | rabeeh | i'v tried creating one using linux pencil with our logo |
| 12:34 | rabeeh | complete disaster i must say :) |
| 12:35 | jnettlet | rabeeh, hmmm. will have to double check my work later. |
| 12:36 | Bluerise | http://i.minus.com/i1YhN9wisnays.gif that one is quite nice |
| 12:36 | rabeeh | jnettlet: i'm not sure it's your fault |
| 12:36 | rabeeh | all my quads are abused |
| 12:36 | Bluerise | The rest is too 3d-y ;) |
| 12:36 | jnettlet | yeah, but I should double check my work regardless. |
| 12:55 | unununium_ | Hmm... no it seems to use dovefb, but still no acceleration... |
| 12:57 | unununium_ | *now |
| 13:03 | rabeeh | jnettlet: ok; i have an Android image now |
| 13:03 | rabeeh | want it? |
| 13:03 | jnettlet | sure. |
| 13:05 | jnettlet | I can't wait to get my new capacitors in the mail. Nothing like a 42" test lcd for free. |
| 13:07 | rabeeh | what did you do? |
| 13:07 | rabeeh | the power supply? is it bricked because of the capacitors? |
| 13:08 | jnettlet | not me. It is a semi-old 720p lg tv. |
| 13:08 | jnettlet | somebody had thrown it out for recycling. I walked by and gave it a look. Looked nearly brand new. A little googling showed problems with the capacitors going. |
| 13:09 | unununium_ | Wow. :) |
| 13:09 | jnettlet | brought it home plugged it in, got power no video or sound. opened it up and yep two blown capacitors on the board. |
| 13:09 | rabeeh | jnettlet: hold on with android a bit |
| 13:10 | rabeeh | this build is not stable (power management issues) |
| 13:10 | jnettlet | ordered a new capacitor kit for $16USD |
| 13:10 | jnettlet | rabeeh, okay no rush |
| 13:10 | rabeeh | which one? |
| 13:10 | jnettlet | okay to the Android |
| 13:10 | rabeeh | notice their temperature rating |
| 13:11 | jnettlet | these are matched to the model of the tv. |
| 14:47 | _rmk_ | jnettlet: sometimes its just poor design: caps also have a habbit of failing if they're being overstressed (eg, too much ripple current) |
| 14:49 | jnettlet | very true. |
| 14:49 | jnettle | 14:49 * jnettlet will take a "garbage" tv and replace some caps any day. |
| 14:50 | jnettlet | This one even has a built in PVR system with a 160GB sata drive. |
| 14:50 | unununium_ | Not bad! |
| 14:51 | jnettlet | yeah should be good fun to play with. If the recording part is useless I will rip it out and mount my Carrier one board in it's place :-) |
| 14:52 | Bluerise | hmm |
| 14:52 | _rmk_ | I have someone's book reading system (he's got severe sight issues) which is basically a camera and a modified TV. It lost colour a couple of weeks ago due to a poor solder joint... but last week the tv part stopped working... turns out the frame coil drive has self-destructed. (40V across a 47ohm resistor due to two transistors going short isn't a good sign.) |
| 14:52 | Bluerise | Carrier-One TV? |
| 14:53 | _rmk_ | that resistor became... almost charcoal. |
| 14:53 | jnettlet | ouch |
| 14:53 | jnettlet | Here are my bad capacitors. https://lh5.googleusercontent.com/-GDGq7wwpEFE/Um5vFLOmqMI/AAAAAAAAFjs/glPr6SzyrlA/w1144-h858-no/IMG_20131028_150447.jpg |
| 14:57 | jnettlet | oh sure. I could just use the sata power connector to power the carrier-1 whenever I turn the TV on. That and the whoosh boot loader and that would be a nice little setup. |
| 14:57 | jnettlet | I have a few extra usb wifi dongles around. |
| 14:58 | _rmk_ | jnettlet: could be either heat or too much ripple |
| 15:00 | jnettlet | _rmk_, probably ripple as this model has reports of signal degradation on the second digital tuner when using the first for watching television. |
| 15:01 | jnettlet | can't supply enough juice for all the components. Probably didn't take into account the spinning 3.5" sata drive :-) |
| 15:02 | _rmk_ | hope your replacements have a higher ripple current rating then :) |
| 15:04 | unununium_ | They look pretty dead. |
| 15:04 | jnettlet | They are Samwha capacitors which don't have the best reputation from reading around the forums. The place I ordered the new ones from say they use Panasonic, Nichicon, or Rubycon |
| 15:07 | jnettle | 15:07 * jnettlet will have to decide if I want to just bite the bullet and replace all the caps, or just the failed ones |
| 15:08 | unununium_ | I'd suggest to replace them all |
| 15:10 | jnettlet | Of course the German would recommend that :-) |
| 15:10 | jnettle | 15:10 * jnettlet needs some more lazy Americans in this channel :-) |
| 15:11 | unununium_ | ;) |
| 15:11 | Bluerise | Nonono |
| 15:11 | Bluerise | I'd just replace the two. |
| 15:11 | unununium_ | Well, it is not too much of work, is it? |
| 15:11 | unununium_ | Wait. Are you really a german? |
| 15:11 | Bluerise | Then test if it works as expected, and then replace the rest. ;) |
| 15:12 | unununium_ | Okay, that's the best method. |
| 15:17 | unununium_ | You are the real german here! :-D |
| 15:18 | _rmk_ | jnettlet: allegedly, that 1000uF cap should stand 2.7A rms |
| 15:20 | _rmk_ | looks like samwha aren't producing that series anymore (wonder why!) |
| 15:20 | _rmk_ | so could only find the data on them here: http://www.evita.lt/getPdf.php?id=5845 |
| 15:22 | jnettlet | then I should probably replace them all. |
| 15:22 | jnettlet | I think I will just replace the blown ones and see if the thing works, and what the quality is like. If my wife wants this one downstairs then I will replace all of them. |
| 15:30 | unununium_ | Grr. It still does not work. But it won't give me any errors. |
| 15:30 | unununium_ | I want errors! |
| 15:31 | lioka | jnettlet: probably those things on metal plate are shorted too |
| 15:33 | _rmk_ | lioka: "those things on a metal plate" I assume you're not an electronic engineer using terminology like that :) |
| 15:33 | lioka | _rmk_: just not native speaker :] |
| 15:33 | _rmk_ | they're plastic TO220 packaged diodes - isolated from their heatsink. |
| 15:33 | lioka | yep |
| 15:34 | _rmk_ | if they're short it's likely the primary side has also gone bang |
| 15:35 | _rmk_ | looking through what's available from farnell, I see hardly any which meet the Samwha WB spec for ESR and ripple current... that tells you something :) |
| 15:35 | Bluerise | Being not an electronic engineer I gotta ask: why do you suspect it? |
| 15:36 | Bluerise | (That they are shorted, too) |
| 15:37 | _rmk_ | bluerise: if they've gone short, it would put the transformer output winding directly across the capacitor - and try to wiggle the capacitor at the switching frequency, which could result in a ripple current way in excess of their ratings... but I think that's actually very unlikely. |
| 15:37 | lioka | Bluerise: well, at least it is easy to check |
| 15:38 | _rmk_ | most SMPS will either blow up (quite literally go bang) and the primary side self destructs, or they will go into hiccup mode where the output current becomes very low. |
| 15:39 | _rmk_ | and... if they have gone short and caused those caps to do that, every bit of silicon on those supply lines will also have died |
| 15:40 | _rmk_ | right, time to see if I can get this CRT to scan its field again :) |
| 15:48 | unununium_ | Hmm, where can I have a look if I don't get the graphics accelerated, but Xorg itself doesn't give me any errors? And dmesg neither... |
| 15:52 | lioka | unununium_: do you have libGAL.so in readelf -d dovefb_drv.so |
| 15:52 | lioka | .. output ? |
| 15:54 | jnettlet | unununium_, post your Xorg.log and we can tell what is up |
| 15:56 | jnettlet | actually post it and I will take a look when I get back. Need to walk the dogs while the sun is out |
| 15:59 | unununium_ | jnettlet: That's a good idea! (your dogs) And thanks in advance... |
| 16:00 | unununium_ | lioka: 0x00000001 (NEEDED) Shared library: [libGAL.so] |
| 16:03 | unununium_ | Xorg.0.log: http://pastebin.ca/2472500 |
| 16:08 | lioka | unununium_: log seems good for me |
| 16:08 | lioka | unununium_: do you have access to /dev/galcore, /dev/uio, /dev/bmm ? |
| 16:09 | unununium_ | Hmm, no. |
| 16:09 | unununium_ | /dev/uio is simply missing. |
| 16:10 | lioka | or /dev/uio0 -- can't remember |
| 16:10 | unununium_ | At least root has access. |
| 16:10 | lioka | unununium_: are you running X session as root ? really ? |
| 16:11 | unununium_ | No. |
| 16:12 | unununium_ | To me everything looks fine. But the result doesn't make me happy as everything is very slow within X. |
| 16:13 | unununium_ | glxinfo says Direct Rendereing would be enabled as well. |
| 16:13 | unununium_ | But uses Mesa... don't know, if that's the way it is supposed to be. |
| 16:14 | unununium_ | Files are accessible now and it did not change the result. |
| 16:17 | lioka | unununium_: well, what supposed to be accelerated: GLES/EGL based apps and video decoding via vmeta |
| 16:18 | lioka | as for regular 2d stuff -- it's quite slow, yes |
| 16:18 | unununium_ | Is it normal that even using firefox hangs? |
| 16:18 | unununium_ | Oh. Okay. |
| 16:18 | unununium_ | I mean... even moving Windows is not very... fast. |
| 16:18 | unununium_ | I will try a video, if I find one. :) |
| 16:20 | lioka | for video you probably should have gstreamer-based player and two gst-plugins with vmeta support |
| 16:20 | lioka | or patched xbmc |
| 16:20 | unununium_ | What applications are supposed to be "fast"? |
| 16:21 | lioka | old kaffeine (from kde3 or trinity) was good enough |
| 16:22 | lioka | by 'good' i mean easy enough to hook into vmeta stuff |
| 16:23 | lioka | but still 1080p was too laggy |
| 16:23 | unununium_ | Hmm, okay. |
| 16:24 | unununium_ | I hope CuBox-I will be faster. :) |
| 16:24 | lioka | lets hope, yes |
| 16:25 | unununium_ | We will see. ;) |
| 16:25 | dv_ | lioka: one big problem is the video output |
| 16:26 | dv_ | even if you decode with vmeta, if you do not have a way to output the frames without involving a frame copy done by the CPU, you cannot output 1080p smoothly |
| 16:27 | unununium_ | That sounds understandable to me... |
| 16:27 | lioka | dv_: that's why they invented bmmxv |
| 16:27 | dv_ | marvell added a hack to the dovefb driver to be able to pass a DMA buffer's physical address through Xv |
| 16:28 | dv_ | which is included in the bmmxv plugin, yes (bmmxv is just a hacked xvimagesink) |
| 16:32 | unununium_ | Well. It seems to be able to play the Simpsons. ;) |
| 16:32 | lioka | yay! |
| 16:33 | liok | 16:33 * lioka afk for a while |
| 16:33 | unununium_ | lioka: And again I thank you very much for your help! |
| 17:23 | jnettlet | unununium_, dovefb(0): EXA is disabled that is why 2d rendering is slow. |
| 17:29 | jnettlet | Has anyone attempted a Raspberry Pi camera module hookup to the C1 yet? |
| 18:04 | unununium_ | jnettlet: Thanks for your reply! Do you have an idea how to solve this problem? |
| 18:05 | unununium_ | And finally I have got my snow tyres mounted. :) |
| 18:23 | jnettlet | unununium_, great. I have to put my winter tires on soon |
| 18:24 | jnettlet | unununium_, are you using the xorg-dovefb that comes with the debian distro? |
| 18:24 | unununium_ | I'm using testing and have http://download.solid-run.com/pub/solidrun/cubox/repo/debian in my sources list. But my dovefb comes from the packages file listed in the wiki |
| 18:25 | jnettlet | actually no. give me until tomorrow. I am finishing up the newest rev of the driver tonight with all the new fixen's I will use you as a guinea pig tomorrow or Thursday |
| 18:26 | unununium_ | That'd be great! Although I won't be here tomorrow because of work... but on Thursday you can do what ever you want with me. :) |
| 18:27 | jnettlet | sounds like a plan. A quick test would be to add these lines to the driver section of your xorg.conf. |
| 18:27 | jnettlet | Option "UseGPU" "true" |
| 18:27 | jnettlet | Option "XvAccel" "true" |
| 18:27 | jnettlet | Option "ExaAccel" "true" |
| 18:27 | unununium_ | Oh. |
| 18:28 | unununium_ | Wait. I have another idea. |
| 18:28 | unununium_ | Option "ExaAccel" "false" <- I have the idea, that this might be bad. But not sure... |
| 18:28 | jnettlet | Yep that will disable Exa acceleration |
| 18:29 | jnettlet | be prepared that if other things are misconfigured that could possibly crash your X server |
| 18:29 | jnettlet | just warning you |
| 18:29 | unununium_ | Well, a crashed X-Server would not be a problem... I can undo the changes. |
| 18:30 | unununium_ | Ah. Activating Exa results in "Segmentation fault at address 0x0". |
| 18:31 | jnettle | 18:31 * jnettlet warned you |
| 18:31 | jnettlet | Xorg log please |
| 18:31 | unununium_ | Yes, you did. :) |
| 18:31 | unununium_ | One second. |
| 18:33 | unununium_ | http://pastebin.ca/2472529 |
| 18:34 | jnettlet | Can you post your xorg.conf. I notice this line. Allocate all pixmaps from system memory, just for debug use |
| 18:34 | unununium_ | Sure I can, but this time I have to warn you: I just copied it from somewhere and made minimal changes. |
| 18:36 | unununium_ | http://pastebin.ca/2472533 |
| 18:37 | jnettlet | turn off shadowfb That causes a segfault with EXA enabled |
| 18:37 | jnettlet | and turn swsolid to true. The way the driver is now it is faster to do solid fills with the cpu |
| 18:37 | unununium_ | Yes. You are right. Now the Segmentation fault address moved to 0x24. Xorg.log coming... |
| 18:38 | unununium_ | And Segfault moved to 0x30 after enabling swsolid. |
| 18:40 | unununium_ | http://pastebin.ca/2472534 |
| 18:41 | jnettlet | hmmm something else isn't right Warning: Please check LCD offscreen memory space |
| 18:41 | unununium_ | Yes, locked at that line as well... |
| 18:41 | unununium_ | *looked |
| 18:42 | jnettlet | does dmesg give you any interesting output. Like galcore failing to allocate memory? |
| 18:43 | unununium_ | Used video mem: 0 KB contiguous: 0 KB virtual: 0 KB |
| 18:43 | jnettlet | yep there is your problem right there. |
| 18:43 | jnettlet | Lets pick this up on Thursday and I will have all sorts of fun goodies for you to play with |
| 18:43 | unununium_ | Does not look very normal to me. |
| 18:43 | unununium_ | Woohooo! :) |
| 18:44 | unununium_ | Thank you. :) |
| 19:07 | dv_ | jnettlet: btw I also had problems with EXA and the dovefb drive |
| 19:07 | dv_ | *driver |
| 19:08 | jnettlet | dv_, I know you aren't forgotten. I am juggling a lot right now. But have merged almost all of the Freescale/Marvell vivante 4.6.9 into a semi-coherent kernel driver |
| 19:08 | dv_ | :) |
| 19:08 | dv_ | I meant that I can relate to what unununium_ is seeing |
| 19:08 | dv_ | also, excellent news! |
| 19:09 | jnettlet | I am finishing userspace now. I have shifted gears a couple of times because I realizd Android needs a proper FB driver and can't use KMS |
| 19:09 | jnettlet | I am trying to cover all bases. |
| 19:13 | jnettlet | dv_, glad to see you are still lurking around. You seemed a little fed up with the current Freescale AV situation. |
| 19:13 | dv_ | fed up, no. |
| 19:13 | dv_ | I am just very busy at work. |
| 19:13 | dv_ | in fact, I am making progress with the gstreamer drivers |
| 19:14 | dv_ | err, plugins |
| 19:14 | jnettlet | oh great. I thought you had decided to hold off on them because Pengutronix were being aloof about what they were doing and when they would release it. |
| 19:14 | jnettlet | with the mem2mem v4l integration that is |
| 19:15 | dv_ | well, they dont seem to say or release anything |
| 19:15 | dv_ | and some people are already using my plugins |
| 19:16 | dv_ | so I wont wait |
| 19:16 | dv_ | (this is the perfect moment to remark that I find the pengutronix style of communication to be annoying) |
| 19:16 | jnettlet | great. |
| 19:16 | dv_ | why not develop in the open? |
| 19:18 | jnettlet | based on nothing but what is out in the community as far as code, I don't know how much they are actually developing at this point. |
| 19:18 | jnettlet | the definitely were very active for 6-8 months in 2012 but I haven't seen much of anything this year. |
| 19:19 | jnettlet | unless they have developed it and are keeping it "in house" till someone approaches them to finish it. |
| 19:20 | jnettlet | which is why I think FOSS should have some sort of kickstarter/indiegogo type site where developers can put up their projects to "release them to the community" |
| 19:20 | dv_ | which also sounds weird. they did all that work, invested time and money, and then hold back? |
| 19:21 | jnettlet | devs make some bank and the community gets the code. The more interesting the code/project the more money the devs can possibly collect. |
| 19:21 | jnettlet | maybe that is our calling. |
| 19:22 | dv_ | I mean, this is not closed source |
| 19:22 | dv_ | especially with regards to the kernel, the GPL is in effect |
| 19:22 | jnettlet | yes, but the GPL doesn't take effect until you sell or distribute the altered code. |
| 19:22 | dv_ | so either they get paid once to get access to the patches (which can then be copied freely afterwards because of the GPL), or they have a strange business strategy |
| 19:23 | jnettlet | and then they still don't have to release the source until someone that has purchased the product requests that they release it |
| 19:24 | dv_ | but on top of that, I thought freescale hired these guys to mainline the imx bits in the kernel |
| 19:25 | jnettlet | yeah you might have missed that. _rmk_ talked to these guys during the conference and they say that isn't the case. They only mainline if a customer requests it. |
| 19:26 | jnettlet | That is something that we all really need to sort out. |
| 19:26 | dv_ | but they are hired by freescale |
| 19:26 | dv_ | ? |
| 19:27 | jnettlet | can't answer that. rabeeh thought so. _rmk_ was pengutronix working for freescale? Or didn't you talk about that? |
| 19:27 | jnettle | 19:27 * jnettlet will be back in 5. I need to tend to my soup |
| 19:28 | dv_ | anyway, we are on our own |
| 19:28 | dv_ | so, business as usual |
| 19:39 | jnettlet | well at least we have a thriving community here. |
| 19:46 | _rmk_ | didn't talk about who their customers were |
| 20:10 | _rmk | 20:10 * _rmk_ tweaks his xorg armada drm driver a little |
| 20:22 | _rmk_ | ... and time to power up the carrier-1 and pull some of these changes across there |
| 22:07 | unununium_ | I think today I'll go to bed early... wish you a good night and thanks for the great help! :) cu on Thursday! |
| 22:07 | unununium_ | Well. |
| 22:07 | unununium_ | ls |
| 22:08 | unununium_ | A good night to everybody who is able to read it. ;) |