**** BEGIN LOGGING AT Mon Jul 26 02:59:57 2010 **** BEGIN LOGGING AT Mon Jul 26 05:00:15 2010 Jul 26 05:51:09 http://www.geeksphone.com/en/ Jul 26 06:03:40 TAsn: hmm... they say to ship limited units... that's neww Jul 26 06:13:41 Ainulindale: can't obey to rule #10 :P Jul 26 08:53:51 moin Jul 26 09:38:34 does android support dvd video playback Jul 26 10:16:49 Has anyone besides me problems with GSM detaching from the network in latest SHR-U? Jul 26 11:15:02 beniwtv: what revision of fsogsmd has latest SHR-U? Jul 26 11:15:07 * mrmoku checks Jul 26 11:17:30 beniwtv: have to build updated cornucopia... can't do that before tonight though Jul 26 11:17:36 but then that problem is fixed Jul 26 11:17:53 * Q-Master is tired of WSs Jul 26 11:18:05 beniwtv: if you need an urgent fix put 'pin = 1234' into your /etc/phonefsod.conf in the [gsm] section Jul 26 11:18:16 Q-Master: WS is gone for me Jul 26 11:18:31 mrmoku: how? Jul 26 11:18:42 kernel update Jul 26 11:19:05 no shr-u update yet Jul 26 11:19:17 it was one of the last kernel updates Jul 26 11:19:54 the kernel was not updated for a very long time already by opkg update/upgrade in SHR-u Jul 26 11:20:23 root@om-gta02 ~ # opkg list_installed kernel Jul 26 11:20:24 kernel - 2.6.32.13-oe3.3+gitr6+a9254be10ac2294ea20165a87c09ea6afcf66d94-r0.5 Jul 26 11:20:45 mrmoku: yep. this one is at mine neo me also Jul 26 11:20:57 Q-Master: hmm... for me WS is gone with that Jul 26 11:21:41 got this morning suspend problem. after an alarm wakeup no suspend possible. even autosuspend resumed after a short time. Jul 26 11:21:59 and got a WS just 2 min's ago when touched the display. Jul 26 11:22:27 Q-Master: what was the resume reason? Jul 26 11:23:28 lindi-: AFAIK as always on new kernel: unknown. no such file or directory. Jul 26 11:23:48 mrmoku|lunch: and LEDs are now not working also. 8( Jul 26 11:24:07 Q-Master: well can you find the resume_reason files manually? Jul 26 11:24:39 lindi-: the last time i've tried - found none. let me resume and check. 1 sec Jul 26 11:26:15 lindi-: 2010-07-26T11:23:51.310200Z [WARN] fsousaged : Can't read-open /sys/class/i2c-adapter/i2c-0/0-0073/neo1973-resume.0/resume_reason: No such file or directory Jul 26 11:26:15 2010-07-26T11:23:51.310927Z [INFO] LowLevelOpenmoko <>: No resume reason marked in /sys/class/i2c-adapter/i2c-0/0-0073/neo1973-resume.0/resume_reason Jul 26 11:26:15 2010-07-26T11:23:51.311698Z [INFO] UsageController <8 R>: Resume reason seems to be FSO_USAGE_RESUME_REASON_Unknown Jul 26 11:26:27 Q-Master: find /sys | grep resume Jul 26 11:26:37 1 sex Jul 26 11:26:44 s/sex/sec Jul 26 11:27:38 lindi-: /sys/devices/platform/s3c2440-i2c/i2c-0/0-0073/resume_reason Jul 26 11:28:03 the path is wrong Jul 26 11:28:56 Q-Master: there should be _two_ resume_reason files Jul 26 11:29:13 lindi-: nop. only 1 here Jul 26 11:29:21 Q-Master: 2.6.32 bug I guess Jul 26 11:29:29 might be Jul 26 11:29:40 but that was from beginning. Jul 26 11:29:50 there are too much bugs already. 8( Jul 26 11:29:53 yes I listed it myself too Jul 26 11:30:23 no report though since I thought it would just be non-implemented feature Jul 26 12:07:32 Q-Master: LED non working? Jul 26 12:07:45 Q-Master: are you sure you have up-to-date rules.yaml for oeventsd? Jul 26 12:08:04 i have already fixed LEDs some time ago Jul 26 12:09:00 dos1: 1 sec. Jul 26 12:11:31 dos1: I have this one: http://pastebin.org/420208 Jul 26 12:12:17 Q-Master: it's old Jul 26 12:12:34 Q-Master: you don't update conf files after opkg upgrade??? Jul 26 12:13:43 dos1: I have only this one. no -opkg files... Jul 26 12:13:57 dos1: what package contains this? Jul 26 12:14:04 strange Jul 26 12:14:11 frameworkd-config-shr Jul 26 12:14:51 hi guys Jul 26 12:15:01 there's a typo in the ncurses recipe Jul 26 12:15:09 -fforward-progagate Jul 26 12:36:29 dos1: nothing changed. Jul 26 12:37:05 dos1: /etc/freesmartphone/oevents/rules.yaml haven't changed a bit. Jul 26 12:37:10 Q-Master: http://git.shr-project.org/git/?p=shr-themes.git;a=blob;f=frameworkd/frameworkd-config-shr/om-gta02/rules.yaml;h=8f7e2a7a9fc849d816f2ad03b3f32d610bb5fc40;hb=HEAD Jul 26 13:00:16 xrmx: hi Jul 26 13:00:33 I built the QtWebKit, but it segfaults on JavaScript Jul 26 13:00:52 Gabrys, with or without the patch? Jul 26 13:00:57 with JIT Jul 26 13:01:05 then I reverted the patch, compiled again Jul 26 13:01:09 and it crashes as well Jul 26 13:01:30 I'm not sure if I cleaned the place in the proper way Jul 26 13:03:05 yes, it's badly cleaned Jul 26 13:03:08 Gabrys, this suggests that qt use .pro files and not autotools http://trac.webkit.org/wiki/BuildingQtOnLinux Jul 26 13:03:12 the same md5sum of libQtWebKit.4.9.0 Jul 26 13:03:28 Gabrys, so my patch does not do any difference to your build Jul 26 13:04:18 xrmx: ah ok Jul 26 13:04:28 xrmx: I'm going to find out which revision works anyway Jul 26 13:04:41 Gabrys, have fun :P Jul 26 13:05:27 I'm tired of this Jul 26 13:05:34 but what else to do Jul 26 13:05:37 I need to have this working :-) Jul 26 13:06:06 Gabrys, then use a release instead of svn no? Jul 26 13:06:37 I need the improvements from QtWebKit2 Jul 26 13:07:07 Gabrys, ah! Jul 26 15:01:50 hey folks Jul 26 15:03:16 My SHR-testing dmesg gives me "JFFS2 notice: (50) check_node_data: wrong data CRC in data node at 0x0d263ec0: read 0x4e5f6846, calculated 0x20f14315" - should I start to worry? Jul 26 15:03:23 I also see "JFFS2 warning: (868) jffs2_sum_write_data: Not enough space for summary, padsize = -900" Jul 26 15:27:30 hi Jul 26 15:27:44 i want to write a small game for the freerunner Jul 26 15:27:50 using SDL Jul 26 15:28:07 i want to blit as much as possible. Jul 26 15:28:39 i read that the glamo can blit (via SDL). but only support 512*512 pixel Jul 26 15:29:06 can i use 640*480 and use hardware accelerated blitting of the glamo? Jul 26 15:29:25 or is the 512*512 the maximum texture size and the glamo can blit in larger areas? Jul 26 15:33:14 rohezal: 511x511 is the maximum size of a target buffer for 3D in Glamo Jul 26 15:33:30 you can blit as much as you like (actually up to 640x640 IIRC) Jul 26 15:33:38 cool thx guys :) Jul 26 15:33:54 you could use DRI/DRM for your access to the GPU Jul 26 15:34:10 (that's what it's for) Jul 26 15:36:17 im not so deep in gpu stuff Jul 26 15:36:30 when i use SDL_BlitSurface Jul 26 15:36:38 does it use the glamo? Jul 26 15:36:50 the main idea is the avoid sending the frame from the cpu to the gpu Jul 26 15:37:02 because the !"&/() 7 MB/s cpu<->gpu bus Jul 26 15:37:06 not sure actually - maybe it does, via X Jul 26 15:37:29 http://wiki.openmoko.org/wiki/Smedia_Glamo_3362 Glamo's 2D engine is currently used to accelerate solid fills and blits in both XGlamo (Kdrive) and X.org (xf86-video-glamo). Jul 26 15:37:49 hmmm ok Jul 26 15:37:53 yep, but whether or not your SDL commands make it through SDL to X to EXA in a useful way is a different matter Jul 26 15:37:53 can i test it^^ Jul 26 15:38:06 can i test it?^^ Jul 26 15:38:33 sure - do continuous fullscreen 640x480 blits and see how fast it is - if it's accelerated, it'll be >50fps (should be closer to 100fps if everything is working optimally) Jul 26 15:38:53 hmm ok Jul 26 15:38:56 thx weiss Jul 26 15:39:10 if it's not accelerated, it'll be down at ~11-12 fps Jul 26 15:48:19 k on my pc my code gets 486.5 FPS Jul 26 15:48:22 let try on freerunner Jul 26 15:48:36 it won't be that high, for sure :) Jul 26 16:02:48 do you know if qtopia (with QX) is using a normal x server with hardware acceleration for blitting? Jul 26 16:03:15 don't think so, but I don't know much about qtopia Jul 26 16:05:01 hmm gamerunner^^? Jul 26 16:05:14 i got the distro on a sd card Jul 26 16:15:48 rohezal: Driver "Xglamo" Jul 26 16:16:00 rohezal: in QX. Jul 26 16:16:47 cool thx @ gena2x Jul 26 16:30:10 fu*k cross compiling Jul 26 16:30:27 (just want to say it :P) Jul 26 16:30:49 tame it. Jul 26 16:31:18 hehe Jul 26 16:31:32 * Weiss agrees that cross-compiling is indeed an absolute pain Jul 26 16:31:55 (http://warmcat.com/_wp/2010/02/08/fosdem-and-the-linux-cross-niche/) Jul 26 16:33:12 * gena2x disagree. 'absolute pain' is subjective. in fact sdram memory diagram fall to this category. Jul 26 16:33:40 hehe, indeed Jul 26 16:34:38 working Jul 26 16:34:42 a bit make file hacking Jul 26 16:34:51 hmm ok lets get it to the freerunner Jul 26 16:36:48 compiling stuff natively on the FR takes the same amount of time, but you get to read a book instead of stressing out Jul 26 16:37:34 i call this 'interrupt driven development'. Jul 26 16:39:13 you start thing 1, wait for compilation, start thing 2, then interrupt thing 2, to see why compilation fail on 1. then restart 1, and resume 2. you end up in both 1 and 2 not finished, as you can't concentrate neither. Jul 26 16:39:42 s/neither/ on any of 1 or 2/ Jul 26 16:39:42 gena2x meant: you start thing 1, wait for compilation, start thing 2, then interrupt thing 2, to see why compilation fail on 1. then restart 1, and resume 2. you end up in both 1 and 2 not finished, as you can't concentrate on any of 1 or 2. Jul 26 16:40:26 either way takes a long time, however Jul 26 16:40:37 yes Jul 26 16:40:44 <- needs an arm cpu desktop Jul 26 16:40:55 hmmm snap dragon or omap4? ^^ Jul 26 16:44:11 hm, reminds me there was some ARM-based tablet/laptop... can't remember its name Jul 26 16:47:08 touchbook Jul 26 16:47:08 people from french enlightenment association had one at RMLL, they were our neighboor on the association village Jul 26 16:47:55 yes, that thing Jul 26 16:49:03 misc, did you get to play with it? Jul 26 16:49:55 mickey|, Jul 26 16:49:57 hi Jul 26 16:50:01 we must talk privately Jul 26 16:50:28 about the next target Jul 26 16:51:42 ThibG: i just took a look, but you should ask directly on the channel of the french association Jul 26 16:52:18 ThibG: not sure if this #e.fr Jul 26 16:52:24 ok Jul 26 16:52:56 ( ask puppet_master about it ) Jul 26 16:56:17 hmmm Jul 26 16:56:22 damn seg fault on free runner Jul 26 16:57:00 someone want to look @ the makifle? http://pastebin.de/8556 Jul 26 16:57:21 misc, the device looks interesting, and not that expensive :) Jul 26 16:57:50 working Jul 26 16:57:53 ok damn Jul 26 16:58:03 12 FPS via blitting in SDL with QT Jul 26 16:58:18 someone knows if there is a distro Jul 26 16:58:24 which supports blitting via SDL? Jul 26 17:01:20 which supports *accelerated* blitting via SDL, I guess you mean..? Jul 26 17:01:30 yes^^ Jul 26 17:01:32 of course Jul 26 17:01:47 maybe gamerunner or shr? Jul 26 17:02:27 it depends more on SDL I think, and that will most likely be the same for all Jul 26 17:02:48 maybe SDL can be configured to use X commands (which X can accelerate) - maybe it's doing something else right now Jul 26 17:04:58 GNUtoo|laptop: ok, not now, but i'll ping you later Jul 26 17:11:31 GNUtoo|laptop: hmm, it seems snapdragon is also supported by upstream now Jul 26 17:11:42 nice Jul 26 17:12:19 1GHz should be enough Jul 26 17:12:37 even for ooffice and abiword Jul 26 17:12:44 running fluently Jul 26 17:13:16 Weiss you know how to force SDL using X commands? Jul 26 17:13:40 or an other to blit easily for a game without using the X Server directly? Jul 26 17:17:33 toshiba seems to have produced a nice snapdragon thing Jul 26 17:17:35 hmm Jul 26 17:17:41 rohezal: to the first: not at all. to the second: I can show you how to send hardware command sequences for a blit. it's tricky, but it'll give you the maximum possible performance (and also render your code specific to FR unless you write some kind of intermediate library) Jul 26 17:18:29 Weiss: this would be nice. i read the X11 cant use blitting with hardware acceleration at all. DAMN Jul 26 17:19:32 wherever did you read that? Jul 26 17:19:41 or do you mean SDL+X11 can't use accel? Jul 26 17:20:33 http://kevinlocke.name/inquiry/sdlblitspeed/sdlblitspeed.php Jul 26 17:20:36 SDL supports hardware accelerated blitting of ``normal'' surfaces to the Jul 26 17:20:36 screen surface. However, this mode is not supported for X11 (except for Jul 26 17:20:36 some cards when using DGA). In fact, the only backend hardware Jul 26 17:20:36 accelerated memory-to-screen blitting is well supported seems to be the Jul 26 17:20:36 DirectX5 backend for Windows. Jul 26 17:20:42 http://forums.libsdl.org/viewtopic.php?p=4943&sid=2414d7b2c07c79f5f9c8e41a12e55d8a Jul 26 17:21:01 maybe im misunderstand some things Jul 26 17:21:28 GNUtoo|laptop: http://htc-linux.org/wiki/index.php?title=Tg01 Jul 26 17:21:48 GNUtoo|laptop: http://en.wikipedia.org/wiki/Toshiba_TG01 Jul 26 17:21:55 GNUtoo|laptop: snapdragon Jul 26 17:21:57 :-D Jul 26 17:21:59 ok Jul 26 17:22:02 rohezal: ok, so that's just SDL. X11 can accelerate things for sure.. Jul 26 17:22:08 but first we make htcdream run properly Jul 26 17:22:23 Weiss: sachmal du sprichst deutsch oder? so vom namen her?^^ Jul 26 17:22:40 leviathan, yes and? Jul 26 17:22:46 leviathan, you plan to buy one? Jul 26 17:23:00 Weiss: yes i know. but it looks like sdl cant use accelerated blitting with x Jul 26 17:23:02 rohezal: openmoko-de Jul 26 17:23:08 for the dream I have the ability to recruit a new developer Jul 26 17:23:09 rohezal: ;-) Jul 26 17:23:25 GNUtoo|laptop: wonderfull! Jul 26 17:23:26 let's see how it goes Jul 26 17:23:38 leviathan, the only problem is that he lacks the hardware Jul 26 17:23:39 GNUtoo|laptop: we can always use new devs ^_^ Jul 26 17:23:45 leviathan: yes but weiss is in the chanel? why change to de? Jul 26 17:23:45 leviathan, but I can solve that Jul 26 17:24:11 leviathan, I know his capabilities and he's trusted Jul 26 17:24:20 so maybe Jul 26 17:24:23 I give him my dream Jul 26 17:24:26 rohezal: here is english given as language, daher für deutsch konversation bitte den deutschen channel ;-) Jul 26 17:24:30 and get mickey's dream Jul 26 17:24:39 or something like that Jul 26 17:24:51 hehe Jul 26 17:25:01 anyway I must finish dream soon Jul 26 17:25:06 before starting a new target Jul 26 17:25:10 GNUtoo|laptop: yes Jul 26 17:25:18 leviathan: wait, i though writing the first name with double dots is only visible to the user with that name... damn im an irc noob ... sory Jul 26 17:25:20 GNUtoo|laptop: I dunno how the meeting will go Jul 26 17:25:37 rohezal: np Jul 26 17:25:49 ok Jul 26 17:26:07 rohezal: so, join the german channel Jul 26 17:26:17 rohezal: you will see me there too ^_^ Jul 26 17:27:19 GNUtoo|laptop: I hope they will buy it... The idea of getting the development done by some people for free should be a hammer argument Jul 26 17:28:04 GNUtoo|laptop: they give us a platform, we can develop on, with theire cooperation Jul 26 17:28:05 ok Jul 26 17:28:18 leviathan, can I PM you? Jul 26 17:28:25 GNUtoo|laptop: of course Jul 26 17:34:23 ok weiss you said there is a way to use hw accelerated blitting. do you know some good links? is it possible to store the images at the gpu ram so it doesnt need to get the images via the cpu for every frame? Jul 26 17:37:12 rohezal: sure, here are the demo programs: http://git.bitwiz.org.uk/?p=glamo-dri-tests.git;a=tree Jul 26 17:37:12 GNUtoo|laptop: ok, i'm here for a bit now Jul 26 17:37:27 mickeyl, ok Jul 26 17:37:30 this is the best example: http://git.bitwiz.org.uk/?p=glamo-dri-tests.git;a=blob;f=gdrm-waitq.c;h=159db7dbb9922b039aa3803495a2f5aed67dde45;hb=HEAD Jul 26 17:37:57 (although, that doesn't use X) Jul 26 17:39:36 ok you are using some headers Jul 26 17:39:38 66 #include "drm.h" Jul 26 17:39:38 67 #include "glamo_drm.h" Jul 26 17:39:38 68 #include "drmtest.h" Jul 26 17:39:38 69 #include "glamo-regs.h Jul 26 17:39:53 where can i get them? are you linking against a special lib? Jul 26 17:41:43 rohezal: yep, libdrm.. Jul 26 17:41:57 rohezal: but hold on.. are you trying to run within X, or QtE/Qtopia? Jul 26 17:44:27 X Jul 26 17:44:37 qtopia can run X Applications Jul 26 17:44:40 via QX Jul 26 17:44:47 its seems to a normal X Server Jul 26 17:44:49 i hope^^ Jul 26 17:45:11 I don't think it is.. I think it's more of an X protocol->Qt stuff translator Jul 26 17:45:28 do you know gamerunner? Jul 26 17:45:33 the distro Jul 26 17:45:38 i got it on a sd card Jul 26 17:45:45 it uses X Jul 26 17:45:52 maybe its worth a try Jul 26 17:45:54 Xglamo? Jul 26 17:45:58 hmmm Jul 26 17:45:59 damn Jul 26 17:46:00 that doesn't have any of this acceleration stuff Jul 26 17:46:01 dont know Jul 26 17:46:11 what distro would you recommend Jul 26 17:46:12 ? Jul 26 17:46:16 SHR :) Jul 26 17:46:23 to slow for daily use :( Jul 26 17:46:48 hope a stable release with a none debug kernel will released in the near future^^ Jul 26 17:47:27 lol i googlet for QX... second hit: QX Gay UK Scene Magazine London England - Home-Index Jul 26 17:47:27 - [ Diese Seite übersetzen ] Jul 26 17:47:50 there are only two ways to get acceleration from a user app: [1] program Glamo yourself: old-style, likely to break, difficult, only one thing can use the hardware at once (and if your X server/other display system is already using it and doesn't want to share, you're stuffed). [2] use DRI - which is designed to allow apps to get access to the hardware like this while sharing with other apps and the display system Jul 26 17:49:12 ok so DRI looks good. will it work with QX? Jul 26 17:49:55 where can I learn to programm with DRI? I know a bit about OpenGL and SDL. And a lot about raytracing (witch is quite useless on a freerunner :P) Jul 26 17:49:57 wait, wait. why indirect rendering can't be accelerated??? Jul 26 17:51:36 @ gena2x: indirect rendering means you must render the whole image on the cpu and put it over the slow 7MB/s bus the gpu or? this is what i want to avoid with blitting Jul 26 17:52:07 no. Jul 26 17:52:20 and 1st - where is no 7mb/s bus. Jul 26 17:52:23 * misc watch as gena2x is going to explain the bus is faster Jul 26 17:52:41 ohhh is it faster? Jul 26 17:52:54 rohezal: about bus: http://lists.openmoko.org/pipermail/community/2010-July/062495.html Jul 26 17:53:08 rohezal: and http://lists.openmoko.org/pipermail/community/2010-July/062504.html Jul 26 17:54:12 cool thx 4 links Jul 26 17:56:05 and about direct/indirect rendering: i thought that 'direct' is then application controls hw 'directly', and indirect rendering then hw controlled by X. both can be accelerated. Jul 26 17:56:19 so both are fast. Jul 26 17:56:33 of course, only weiss know how this can be applied to glamo :) Jul 26 17:57:22 but aiglx is exaple of accelerated indirect rendering... Jul 26 17:57:50 brr. Jul 26 17:58:14 i don't really understand this very good. Jul 26 17:58:19 hmm ok Jul 26 17:58:22 looks cool Jul 26 17:58:27 14 or even 23 mb Jul 26 17:58:30 bus speed Jul 26 17:58:41 there are some changed settings mentioned Jul 26 17:58:44 what settings? Jul 26 17:59:00 Wait cycles. Jul 26 17:59:28 do you wish details? Jul 26 17:59:42 see patch (in same dir with u-boot) Jul 26 17:59:51 and s3c manual. Jul 26 18:00:17 ask if in doubt after that. Jul 26 18:00:22 on a normal pc you can change it via bios Jul 26 18:00:30 its row latency and col latency or? Jul 26 18:00:43 glamo is not 'sdram'. Jul 26 18:00:53 it connected via 'sram' interface. Jul 26 18:01:09 diffirent (and simplier) timings used in sram. Jul 26 18:01:18 wait wait waitr Jul 26 18:01:21 s/diffirent/different/ Jul 26 18:01:22 gena2x meant: different (and simplier) timings used in sram. Jul 26 18:01:41 the GPU is connect via a sdram interface? why? Jul 26 18:01:51 why not :) Jul 26 18:02:09 it's not 'sdram interface' its 'memory bus'. Jul 26 18:02:15 or something like this. Jul 26 18:02:19 ok Jul 26 18:02:24 and we can change the timings Jul 26 18:02:29 so more data can be passed Jul 26 18:02:30 sure. Jul 26 18:02:39 like normal sd ram rimings Jul 26 18:02:41 timings Jul 26 18:02:48 my letter is about 2 things. Jul 26 18:02:57 first - it's already not 7Mb/s. Jul 26 18:03:04 ok Jul 26 18:03:06 understand Jul 26 18:03:19 second - it can be even more if timing set more properly. Jul 26 18:04:03 new settings were set not after guess, but after checking glamo docs by weiss (and my initial testing). Jul 26 18:04:37 one problem still exist - according to docs it should be 1-4-2, but this simply not working. Jul 26 18:05:04 and i've tired to test it - one wrong setting and i have to restore sd from backup :) Jul 26 18:05:38 ahh ok Jul 26 18:05:49 good but i still want to learn how to blit^^ Jul 26 18:05:56 with the glamo :) Jul 26 18:06:07 i need libdrm for it Jul 26 18:06:09 and the headers Jul 26 18:06:16 i don't understand, is uses x blit operation is not accelerated? Jul 26 18:06:24 s/uses/usual/ Jul 26 18:06:24 gena2x meant: i don't understand, is usual x blit operation is not accelerated? Jul 26 18:06:28 SDL cant use the X blit operation Jul 26 18:06:30 i think Jul 26 18:06:35 Weiss: ? ^ Jul 26 18:06:59 i just put a baseball via blitting over a 640*480 field Jul 26 18:07:14 which is blitted by one 32*32 plant image Jul 26 18:07:23 around 400-500 images Jul 26 18:07:29 on my desktop Jul 26 18:07:33 12 on freerunner Jul 26 18:07:42 rohezal: QX is just little program that launches regular X server and bypasses all Qtopia redrawing - so that they dont clash Jul 26 18:07:42 with HW blitting it should be around 50-100 fos Jul 26 18:07:43 fps Jul 26 18:07:59 radekp: what is clash? Jul 26 18:08:18 rohezal: so that qtopia and xserver dont draw at the same time Jul 26 18:08:19 also keep in mind that glamo internally is not instant too :) Jul 26 18:08:46 radekp: lol clash... sorry im not a native english speaker Jul 26 18:08:55 rohezal: np, me too :) Jul 26 18:09:24 well im running the programm since 15 minutes or something Jul 26 18:09:27 12 fps :( Jul 26 18:09:57 you can compare with x via framebuffer. Jul 26 18:10:06 which is not accelerated for sure :) Jul 26 18:10:14 *g* Jul 26 18:10:31 weiss said above i can use accelerated blitting via DRM / DRI with libdrm Jul 26 18:10:42 do you know how to get it and the header files for it? Jul 26 18:11:14 only weiss knows such things... Jul 26 18:14:28 maybe after upgrade to debian sid you can install libdrm dev packages, but i have never tried Jul 26 18:14:30 hmm ok^^ Jul 26 18:15:11 or you can try build from sources Jul 26 18:16:15 I've bumped my head into this error a few times over the last couple of days, and make update && make build doesn't make it go away: http://pastebin.com/gAbiP9Yh Jul 26 18:16:28 (I'm trying to build SHR-U) Jul 26 18:17:42 ohh its on the freerunner... maybe i can just copy it to the pc for cross compiling Jul 26 18:17:49 but i still need the headers Jul 26 18:19:41 * gena2x always thought that all this 'dri' 'drm' is for gl... and normal blit acceleration is EXA or XAA... Jul 26 18:21:54 i dont know Jul 26 18:25:02 hmm ok Jul 26 18:25:13 opkg-target install libdrm-dev Jul 26 18:26:38 but no header files Jul 26 18:26:49 dont tell me i have to sign this dna to get it Jul 26 18:27:44 rohezal: so... direct and indirect rendering are distinct from accelerated/non-accelerated Jul 26 18:28:34 you can have accelerated direct rendering (that's DRI), accelerated indirect (that's AIGLX), but you can also have unaccelerated direct (that's "swrast_dri.so"), and of course non-accelerated indirect (version 0.0.1 GLX) Jul 26 18:29:22 but hm... all this is gl-related. Jul 26 18:29:28 ok... im ok with both unaccelerated (when AIGLX is supporting what is needed) Jul 26 18:29:50 but your sample is using some headers files. how can i get them? Jul 26 18:30:00 #include "drm.h" Jul 26 18:30:00 #include "glamo_drm.h" Jul 26 18:30:00 #include "drmtest.h" Jul 26 18:30:00 #include "glamo-regs.h" Jul 26 18:30:56 ./local/openmoko/arm/arm-angstrom-linux-gnueabi/usr/include/drm/drm.h Jul 26 18:30:58 is there Jul 26 18:31:01 but the other files not Jul 26 18:31:20 gena2x: yep, it is. but "direct/indirect" has little to no meaning outside GLX Jul 26 18:31:53 so, to have accelereated blit, we have to have XAA. Jul 26 18:33:10 Weiss: but you mentioned this code: http://www.mail-archive.com/devel@lists.openmoko.org/msg06145.html Jul 26 18:33:15 for blitting Jul 26 18:34:09 gena2x: or EXA, or KAA, but all of those are just ways for the X server to tell the X driver what it wants to happen, not ways for the app to tell X what it wants Jul 26 18:36:16 Weiss: its for DRI or? is there a tutorial how to do it? There are some header files i cant find on my system. How can I get them? Jul 26 18:41:55 Weiss: got it. so some other way other than Xlib exists to control X server (for 2d)? Jul 26 18:53:58 ... Jul 26 18:54:21 guys need i get the source with GIT from the http://git.bitwiz.org.uk/?p=libdrm.git;a=shortlog;h=refs/heads/glamo ? Jul 26 18:54:30 need i to build libdrm myself? Jul 26 18:54:40 are there the headers for the drm lib? Jul 26 18:55:14 what is the login for git to check the source out? Jul 26 18:55:27 ahh... ok, i got it "The main consumer of this extension will be a direct rendering OpenGL Jul 26 18:55:27 driver, but the DRI2 extension is not designed to be OpenGL specific. Jul 26 18:55:27 Direct rendering implementations of OpenVG, Xv, cairo and other Jul 26 18:55:27 graphics APIs should find the functionality exposed by this extension Jul 26 18:55:27 helpful and hopefully sufficient. Jul 26 18:55:28 " Jul 26 18:55:38 * gena2x will be silent forever on this topic. Jul 26 18:56:00 * gena2x thinks it's a bit overcomplicated. Jul 26 18:56:30 * gena2x gone to do something more useful :) Jul 26 18:56:38 hi gena2x Jul 26 18:56:47 do you know where to obtain the headers? Jul 26 18:56:49 last try Jul 26 18:56:54 ^^ Jul 26 18:58:19 ... Jul 26 18:58:20 http://git.bitwiz.org.uk/?p=libdrm.git;a=snapshot;h=HEAD;sf=tgz Jul 26 18:58:22 sry guys Jul 26 18:58:25 <- stupid :( Jul 26 19:07:35 checking for PTHREADSTUBS... configure: error: Package requirements (pthread-stubs) were not met: Jul 26 19:07:35 No package 'pthread-stubs' found Jul 26 19:07:35 Consider adjusting the PKG_CONFIG_PATH environment variable if you Jul 26 19:07:36 installed software in a non-standard prefix. Jul 26 19:07:36 Alternatively, you may set the environment variables PTHREADSTUBS_CFLAGS Jul 26 19:07:36 and PTHREADSTUBS_LIBS to avoid the need to call pkg-config. Jul 26 19:07:38 See the pkg-config man page for more details. Jul 26 19:07:40 FATAL: oe_runconf failed Jul 26 19:08:54 om-conf libdrm Jul 26 19:50:16 Weiss, beware of new Cowon firmwares... Jul 26 19:51:24 I don't know for your model, but latest S9 upgrade ignores vorbis comments... completly... it's not even the ARTISTSORT thing Jul 26 19:53:53 ThibG: ick :S Jul 26 19:54:04 ThibG: my X5 is quite old.. was thinking of upgrading Jul 26 19:54:23 fortunately, I can revert to the previous firmware, and it works great Jul 26 19:55:39 freesmartphone.org: 03morphis 07msmcomm * rf12688f7943e 10/msmcommd/src/lowlevelcontrol.vala: msmcommd: initialize string array before using it Jul 26 19:55:40 freesmartphone.org: 03morphis 07msmcomm * r21393bcecbc2 10/msmcommd/src/ (remoteclient.vala remoteclienthandler.vala): msmcommd: disconnect the client the right way Jul 26 19:55:41 freesmartphone.org: 03morphis 07msmcomm * r3b0d8780f69c 10/libmsmcomm/ (5 files in 3 dirs): libmsmcomm: implement ref id for test_alive msg/resp Jul 26 19:55:59 my concern is more about headphones, now... Jul 26 19:57:15 mines are dead, and the advertised life-time warranty is a real challenge here Jul 26 19:58:23 on the headphones? Jul 26 19:58:38 yes (not Cowon's) Jul 26 19:58:39 I never used the ones which came with it Jul 26 19:58:42 ah Jul 26 19:58:48 Koss :) Jul 26 20:00:22 the headphones that comes with Cowon's players are crap... Jul 26 20:00:42 that said, I've just seen they are selling some other headphones now Jul 26 20:16:31 <\marco> hi to all Jul 26 20:16:48 <\marco> I've a very strange problem with the FR and a thomson router Jul 26 20:17:34 <\marco> using iliwi it result successful connected Jul 26 20:18:06 <\marco> but I can't ping the router or the internet Jul 26 20:18:34 \marco: wlan? wpa? Jul 26 20:18:39 <\marco> inspecting with wireshark I've found that the FR don't answer to the router arp request Jul 26 20:18:58 <\marco> lindi-: wlan with wpa Jul 26 20:19:06 \marco: uname -r? maxperf? Jul 26 20:19:20 <\marco> I've tried maxperf, same result Jul 26 20:19:52 <\marco> the kernel is the last but the problem always exist from the first kernel I've used Jul 26 20:20:39 <\marco> If i ping another pc connected in wifi it ping! and FR correctly answer to arp request Jul 26 20:20:47 \marco: where did you run wireshark? Jul 26 20:20:54 <\marco> from another pc Jul 26 20:21:06 \marco: how can it see ARP traffic? Jul 26 20:21:24 ah, it's in the same LAN as the wlan AP? Jul 26 20:21:25 <\marco> well.. i mean the arp broadcast Jul 26 20:21:36 <\marco> yes Jul 26 20:22:31 <\marco> the router ask for the mac addres of the ip 192.168.1.67 ( the FR ) but there's no answer from the FR Jul 26 20:22:44 \marco: sniff the traffic on FR. does it see that request? Jul 26 20:23:08 <\marco> mhm I've to install wireshar on the FR Jul 26 20:23:14 <\marco> *wireshark Jul 26 20:23:14 \marco: just install tcpdump Jul 26 20:23:19 <\marco> ok Jul 26 20:23:33 \marco: you can load it to wireshark then Jul 26 20:23:41 <\marco> oh good Jul 26 20:24:18 <\marco> just a moment ( I'm not at home and I've to to take the networking from this pc via usb ) Jul 26 20:27:19 <\marco> ok installing tcpdump Jul 26 20:30:34 hi all! Jul 26 20:31:04 mrmoku: TAsn: seems vibro is br0ken again in latest shr-u Jul 26 20:32:45 <\marco> lindi-: mhm.. tcpdump on FR says the FR is answering to the arp request Jul 26 20:33:21 hi mickeyl Jul 26 20:34:17 <\marco> lindi-: another thing, ifconfig eth0 on the FR report broadcast 0.0.0.0 but changing it with an ifconfig eth0 .. (etc) doesn't solve the problem Jul 26 20:36:25 <\marco> any idea? Jul 26 20:37:36 mrmoku: hi Jul 26 20:40:48 <\marco> in the FR ARP reply, tcpdump say (oui unknown). Is it norma? Jul 26 20:41:07 <\marco> *normal Jul 26 20:42:07 \marco: if you can install "ip" command then output of iki.fi/lindi/print-net-settings would be interesting. ifconfig is really not very debug-friendly Jul 26 20:48:06 <\marco> eheh ok. opkg install iproute2 in progress Jul 26 20:50:23 \marco: however, if it answers the arp request correctly then I don't think there is much we can do Jul 26 20:50:52 <\marco> sigh :( Jul 26 20:50:58 leviathan: hey Jul 26 20:51:06 \marco: try without wpa Jul 26 20:51:12 <\marco> however the router seems to not receive the arp reply Jul 26 20:51:14 hi Jul 26 20:51:21 who was the person who bought a nexusone? Jul 26 20:51:38 <\marco> lindi-: ok! oh, I've installed iproute2 which "debug" you suggest? Jul 26 20:51:47 mickeyl, do you remember who got nexusone? Jul 26 20:51:59 GNUtoo|laptop: Deubeuliou has one Jul 26 20:52:10 \marco: iki.fi/lindi/print-net-settings should print all relevant info Jul 26 20:52:14 ok thanks Jul 26 20:52:16 hi Deubeuliou Jul 26 20:52:19 hi Jul 26 20:52:21 Kensan as well Jul 26 20:52:25 ok Jul 26 20:52:27 yeah, sorry, I didn't send a feedback :/ Jul 26 20:52:38 Deubeuliou, I've many possibilities Jul 26 20:53:00 roughly or get a free nexusone or buy a n900 which isn't cheap Jul 26 20:53:08 mickeyl: did you read my messages in backlog? Jul 26 20:53:15 (and since my vacation are beginning shortly, I won't be able to for a moment) Jul 26 20:53:19 mickeyl: and congrats for the new car :) Jul 26 20:53:22 <\marco> lindi-: mhm : 404 Not Found Jul 26 20:53:28 GNUtoo|laptop: why can't you get both ? Jul 26 20:53:40 Deubeuliou, that would be counter-productive Jul 26 20:53:45 if the nexusone is free, that shouldn't be a problem Jul 26 20:53:48 I see Jul 26 20:53:50 <\marco> oops Jul 26 20:53:54 <\marco> my fault sorry Jul 26 20:53:54 I have limited time Jul 26 20:54:04 ~lart limited time Jul 26 20:54:04 * apt brandishes Excalibur! "With this sword, I vanquish thee, limited time!" and lops off limited time's head Jul 26 20:54:05 <\marco> lindi-: it works sorry ^_^ Jul 26 20:54:05 if I don't use nexusone why getting it? Jul 26 20:54:31 <\marco> lindi-: let me see what the script says Jul 26 20:55:02 you could donate it to someone that wants to hack on it but can't afford one. (but that's maybe off-topic) Jul 26 20:55:10 mrmoku: thanks :). +CFUN=4 and +CFUN=1 does not correctly work w/ the states? Jul 26 20:55:31 ok Jul 26 20:55:43 Deubeuliou, to get it I'll have to do some minor replicant work Jul 26 20:55:52 Deubeuliou, so now to the point... Jul 26 20:55:54 GNUtoo|laptop: I heard the openess of the n900 was not so big ? Jul 26 20:55:59 what works,what doesn't Jul 26 20:56:01 ok Jul 26 20:56:06 Deubeuliou, n900 has battery decoded Jul 26 20:56:18 Deubeuliou, you can also talk to the bootloader with free software Jul 26 20:56:20 mickeyl: that was one... the other one that re-requesting works Jul 26 20:56:25 Deubeuliou, and....the kenrel is a good state Jul 26 20:56:28 ok Jul 26 20:56:41 mickeyl: and that there is open stuff in meego for gsm audio routing on n900 Jul 26 20:56:47 Deubeuliou, and....ophono has some sort of (incomplete) support for n900 Jul 26 20:56:50 the big problem with the nexus one, imho, is the kernel, actually Jul 26 20:56:59 mrmoku: ok. i'll have a look at the cfun issue Jul 26 20:57:06 mrmoku: ah, that's good news Jul 26 20:57:26 yup Jul 26 20:57:32 there are a lot of problem with module dependencies in the kernel at android.kernel.org Jul 26 20:57:54 mickeyl: it consists of a lib and a gstreamer module Jul 26 20:58:09 mrmoku: nice, and both is open? Jul 26 20:58:12 <\marco> lindi-: indeed, "ip" show the correct broadcast ip for eth0 Jul 26 20:58:13 yup Jul 26 20:58:15 mrmoku: ah, youre on Jul 26 20:58:15 excellent Jul 26 20:58:18 wonderful Jul 26 20:58:19 ^^ Jul 26 20:58:22 I don't know what is the state of the porting of the driver into a vanillia kernel Jul 26 20:58:33 Perhaps you've an idea, why the following thing happens Jul 26 20:58:39 I made a recipe for maxima Jul 26 20:58:57 basically it should produce arm code, when I inherit autotools Jul 26 20:59:05 mickeyl: http://meego.gitorious.org/meego-cellular/libcmtspeechdata Jul 26 20:59:14 mrmoku: but its always producing code for the buildhost Jul 26 20:59:14 <\marco> lindi-: I'll try without wpa :) thanks however ^^ Jul 26 20:59:38 Deubeuliou, forget about vanilla Jul 26 20:59:44 and http://meego.gitorious.org/maemo-multimedia/pulseaudio-modules-meego Jul 26 20:59:48 Deubeuliou, that'll be done but it'll take too much time Jul 26 20:59:52 GNUtoo|laptop: I also read that the audio could cause problems Jul 26 21:00:01 Deubeuliou, you didn't try alsa? Jul 26 21:00:04 what did you try? Jul 26 21:00:07 mrmoku: http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=40abfa09ae5dc3310888576756b94e3e42e6b5ce Jul 26 21:00:08 I don't remember Jul 26 21:00:33 mrmoku: it ignores the configure and produces no arm code Jul 26 21:00:47 mrmoku: it builds for x86 no mater what I do Jul 26 21:00:55 mrmoku: heh, i guess we're going to have fun with PA again then :D Jul 26 21:01:07 not that it is bad, it just performed lousy... Jul 26 21:01:09 leviathan: do you have CPATH set in your env? Jul 26 21:01:11 mickeyl: hehe, yes Jul 26 21:01:19 mrmoku: CPATH? Jul 26 21:01:44 mrmoku: why? Jul 26 21:01:45 GNUtoo|laptop: ah, and the touchscreen is a little weird Jul 26 21:01:52 mrmoku: and where shall it point to? Jul 26 21:02:04 leviathan: or LIBRARY_PATH Jul 26 21:02:07 leviathan: must be unset Jul 26 21:02:10 Deubeuliou, that should be easy to change touchscreen Jul 26 21:02:13 ahh, ok Jul 26 21:02:22 Deubeuliou, so basically you didn't make a lot of progress? lack of time? Jul 26 21:02:28 leviathan: I had some problem like that due to having that in my env Jul 26 21:02:32 the bottom part of it actually are the four touch-buttons Jul 26 21:02:41 mrmoku: I don't have any of them set Jul 26 21:02:42 <\marco> I've to go.. bye! :) Jul 26 21:02:48 I don't know if I'm clear ... Jul 26 21:02:54 mickeyl: but omap has some kind of fp, no? Jul 26 21:02:59 that was the main problem I think Jul 26 21:03:29 Deubeuliou, ahh ok like palm pre? Jul 26 21:03:31 * mrmoku building from scratch again on buildhost :/ Jul 26 21:03:32 the four buttons at the bottom are part of the touchscreen Jul 26 21:03:44 I don't know the palm pre Jul 26 21:03:46 Deubeuliou, I understand...palm pre is a bit like that Jul 26 21:03:54 the touchscreen goes over the screen Jul 26 21:04:02 the "over the screen part" Jul 26 21:04:08 is a gesture area Jul 26 21:04:21 mrmoku: yes, but they also have some nice additional instructions which e.g. palm used to _massively_ speed up the pulseaudio resampler (which basically was the actual CPU hog on the FR) Jul 26 21:04:32 mrmoku: we can perhaps reuse those patches as well Jul 26 21:04:39 Deubeuliou, maybe you should report your findings to: Jul 26 21:04:59 http://wiki.freesmartphone.org/index.php/HardwareComparison Jul 26 21:06:44 apart from that ; I don't know how to make the compass work, neither the camera Jul 26 21:08:35 roughly same situation as the dream then Jul 26 21:09:02 I got to go. I the first days of August, I'll have some time to check more things Jul 26 21:09:12 ok Jul 26 21:09:17 me too I got to go Jul 26 21:09:22 mrmoku: hmm, it doesnt seem to be that, none of this environement variables is set Jul 26 21:09:23 hmm Jul 26 21:09:33 what takes me substantial amount of time is managing to get a kernel on which I can ssh to Jul 26 21:09:52 sometimes it works, sometimes it doesn't Jul 26 21:09:58 leviathan: hmm... I can't see anything in the recipe... probably something in the build system of the project? Jul 26 21:10:11 mrmoku: hmm Jul 26 21:10:16 perhaps Jul 26 21:10:19 its clisp Jul 26 21:10:24 I have seen that the USB connector could switch between several modes, that could be interesting Jul 26 21:10:27 bye Jul 26 21:10:59 * mrmoku got to go (to bed) too :) Jul 26 21:11:01 gnight all Jul 26 21:11:14 mrmoku|away: gn8 Jul 26 21:11:24 mickeyl: do you have a snapdragon device? Jul 26 21:11:46 mickeyl: http://htc-linux.org/wiki/index.php?title=Tg01 Jul 26 21:11:52 looks very nice ^_^ Jul 26 21:18:00 bye Jul 26 21:22:10 leviathan: no i don't have one. sounds all appealing, but given the nightmares we have with other devices, it won't become interesting for me until 100% of the peripherals work w/ standard linux interfaces... Jul 26 21:23:37 mickeyl: hmm, we have already rudimentary support in upstream Jul 26 21:23:54 I guess with 2.6.36 we will have official kernel working on dream Jul 26 21:24:13 which means, that we can get away from hackish leviathan-incoming Jul 26 21:24:22 and also use default kernel for dream Jul 26 21:24:47 I will also introduce alsa, as soon as we got the rest running properly Jul 26 21:25:10 interesting. I'll be pleasently surprised when that comes. GPS, Bluetooth, and the sensors will take some more time i'm afraid Jul 26 21:28:06 yes, I fear, that will take much longer Jul 26 21:28:44 good thing Jul 26 21:28:51 its leaded by codeaurora people Jul 26 21:29:00 they have the datasheets available Jul 26 21:29:12 because they have signed a NDA Jul 26 21:29:41 so, perhaps it won't take THAT long Jul 26 21:29:44 but still long Jul 26 21:29:45 -.- Jul 26 22:07:20 evening Jul 26 22:07:23 where is JaMa? **** ENDING LOGGING AT Tue Jul 27 02:59:56 2010