=== Ursinha` is now known as Ursinha === Ursinha is now known as Guest70417 === Guest70417 is now known as Ursinha === Ursinha is now known as Guest78022 [08:22] qgis porting to beagleboard help?? anyone here? [08:37] qgis porting to beagleboard help?? anyone here? [08:38] beyond_: smart questions, anyone read? [08:39] beyond_: http://www.catb.org/esr/faqs/smart-questions.html [08:44] qgis porting to beagleboard ubuntu error.. Help please [08:45] * LetoThe2nd sighs and heads back to work. [09:55] LetoThe2nd, I like it [09:57] lilstevie: what specifically [09:58] the smart-questions faq [09:58] lilstevie: its awesome. too bad nobody ever reads it. [09:59] * lilstevie is reading it right now [10:37] rsalveti, bug 1045491 for you [10:37] Launchpad bug 1045491 in pvr-omap4 "Moving mouse messes up the desktop" [High,Confirmed] https://launchpad.net/bugs/1045491 [11:10] rsalveti, hmm, i struggel to get any monitor output on the server image ... is there any trick to make that work if no xserver and no pvr driver are installed ? [11:17] yippie, my new ac100 screen arrived [11:20] marvin24, do you happen to know if i still need the kernel hacks if i want to use a 1280x720 panel with the 3.1 kernel ? [11:21] ogra_: yes, the resolution is hard coded [11:21] k, sad [11:21] maybe we could read out the edid [11:21] but I have no idea how [11:21] well, i guess it would be fixed if i used a kms enabled kernel [11:22] not really [11:22] no edid, no autoresolution [11:22] oh [11:22] the schematic says something about a ddc connection, though [11:22] i thought kms would rely on EDID data [11:23] anyway, first i need to replace the HW :) [11:23] * ogra_ is a bit scared to remove the bezel [11:24] * LetoThe2nd hands ogra_ an "r" [11:24] lol [11:24] bavarians ... [11:24] \o/ [11:25] a bavarian would never call it "brezel" though, but "brezn" [11:25] oh, indeed [11:28] ogra_: yes, kms needs it [11:28] but there's a hard code hack ... [11:28] lovely [11:28] seems the ddc data path is only connected to the hdmi port [11:28] well, i'm willing to live with a hack if needed, after all 720p panels arent std on the ac100 [11:29] (and i wouldnt have bought it if one of my ac100s wouldnt have a cracked screen) [11:44] ogra_: what is the issue with the server image? in theory you should be able to get the boot log or the console login interface just fine [11:44] unless your panda is not recognizing your monitor from the kernel side, at least not during boot [11:45] and the modeseting from pvr-omap4 is the one enabling and starting your monitor properly [11:45] rsalveti, it works just fine when doin a serial install, but without console= set no screen comes up on the monitor [11:45] there is no pvr on server installs [11:45] ogra_: and thanks for the buglink, got the kernel to build here, will give it a try now [11:45] we cant ship X there [11:45] that's fine, the kernel should still be able to set up the correct resolution and get everything in place for you [11:46] even if you're not using X [11:46] you mean if i dont have pvr installed ? [11:46] let me boot with text and see if I can reproduce your behavior [11:46] yup [11:46] i dont even get a console after install :/ [11:48] [ 0.288848] omap_sr_disable: omap_sr struct for sr_iva not found [11:48] [ 0.291564] omap_hwmod: sl2if: _wait_target_ready error: -16 [11:49] thats the only error i find in dmesg [11:49] well, a few more omap_sr errors [11:49] adding drm.debug=7 omapdss.debug=1 might help us understanding what is happening [11:49] * ogra_ does [11:53] http://paste.ubuntu.com/1185592/ [11:55] ogra_: here it seems to work fine, after booting with 'text' [11:55] rsalveti, what is bothering is that it works just fine on desktop ... [11:55] the monitor turned itself on while opening the tty1 console [11:55] on a server image ? [11:55] note that desktop has no issues here either [11:55] no, desktop, but with 'text' the X11 it not even started [11:56] and it didnt with the alpha release (where we had no pvr) [11:56] let me check the logs [11:57] ogra_: when booting your desktop image, do you always see the splash before X11 starts? [11:57] hmm, i didnt check, i think i didnt see it after install, but before [11:58] i.e. booting the image showed it ... but it was gone after booting into the installed system [11:58] ogra_: also, can you check what happens when you use the dvi output instead of hdmi? [11:58] already tried, no change [11:58] i would blame my monitor (LG) ... if it wouldnt work just fine in desktop [11:59] with the same kernel [11:59] [ 508.846771] omapdss DISPC: channel 1 xres 1920 yres 1080 [11:59] [ 508.879577] omapdss HDMI: hdmi_runtime_get [11:59] [ 508.879608] omapdss HDMI: hdmi_runtime_put [11:59] [ 508.879608] omapdss HDMI: Enter hdmi_display_disable [11:59] seems thats printed over and over in dmesg now [12:02] ogra@panda:~$ who [12:02] ogra tty4 2012-09-04 14:01 [12:02] ogra pts/0 2012-09-04 13:51 (192.168.2.91) [12:02] after logging in blindly ... [12:02] so getty is there and i can log in ... just cant see a thing [12:03] there's a lot of sync_lost as well... [12:03] yes === doko_ is now known as doko [12:07] let me just give the new kernel a try (with in theory fixes the flickering issue) [12:07] this new kernel changes quite a few bits at the dss and display driver [12:07] and reverts a few changes for omap5 as well [12:07] ok [12:08] if you think it hleps i can test here too [12:08] sure [12:19] rsalveti: so you decided to go with the ~try kernel? [12:20] where do you host the code? it's not coming from tilt-3.4.. [12:20] ndec: not yet, still checking what would be the side effects [12:20] we're using robclark's tree [12:20] yeah, i know. [12:21] but it breaks some display interface such as DSI at least. not sure if some of your users will care about that. [12:21] yeah, that's one of the issues I noticed here as well [12:22] DSI? you mean DVI, no? [12:23] there is a cleaner solution that has been designed in omapdss, to fix the same issue. and that is being planned for upstreaming in the upcoming kernel. [12:24] what is the current state of that? [12:24] because we either go with the same solution from ~try or decide to backport the proper one that will land upstream later on [12:24] i am hoping that we will get omapdss changes in our upcoming 3.6 tree. [12:24] but there are a bunch of omap5 related changes at our tree as well [12:25] and i am also interested in a backport of the 'clean' solution. [12:25] omap5? [12:25] because we got the tree from ti-3.4 branch [12:25] and what do you call 'our' tree in fact? is it Andy's tree? [12:25] yup [12:26] yes, tilt-3.4 has some support for OMAP5, even though it's still missing a few bits. we have another tree we use for full releases on OMAP5. [12:27] we will get a tilt-3.6 eventually with similar 'feature' set as tilt-3.4. except that our primary 'dev' and test platform will become OMAP5, instead of OMAP4. [12:27] and that we will get the rework from omapdrm/omapdss. [12:27] for OMAP4 'products' we will continue to use tilt.3.4, well our TI tree based on tilt-3.4 [12:28] got it [12:37] rsalveti, DVI should be working on my branch [12:39] robclark: yup, but for some reason I'm not even getting my hdmi out to work [12:39] still trying to understand why [12:40] had to disable OMAP2_DSS_VENC and OMAP2_DSS_DSI for it to build [12:42] rsalveti, hang on, I can pastebin my .config for you to cross check against [12:43] rsalveti, http://hastebin.com/raw/gopiweniqi [12:43] robclark: thanks, will check here [13:04] rsalveti, btw, the weston shader optimizations were pushed to master last fri.. this should give a nice performance boost w/ pvr [13:04] robclark: cool, nice to know [13:04] ogra_: heya [13:05] ogra_: on https://bugs.launchpad.net/ubuntu/+source/linux-ti-omap4/+bug/1045788 - I was trying ubuntu-server on my panda this morning and I'm seeing something very different [13:05] Ubuntu bug 1045788 in linux-ti-omap4 "screen on quantal server images stays black" [High,New] [13:06] ogra_: with hdmi plugged in, I see the language selection screen come up, but my keyboard does not work [13:06] plars: hey :-) [13:07] hey rsalveti [13:07] rsalveti: your name sounds familiar...have we met? [13:07] :-b [13:07] plars: ;-) [13:08] plars: so I guess you're at least getting your screen to work properly? [13:08] the usb issue is a bit weird [13:08] I'd not be surprised if this screen issue is related with the monitor you're using [13:08] as I'm able to get it to work here just fine as well [13:09] and ogra_ seems to be getting a bunch of hdmi_display_disable messages after boot [13:10] rsalveti: yeah, likely [13:25] rsalveti, ogra_: also, even with serial, mine appears to get stuck at some point... just have a black screen now after going through a couple of steps [13:25] :-( [13:26] wow, not sure what that was [13:26] I got it back though [13:26] no ^c or ^z, I think ^d finally did it [13:26] it was after network detection, after it asked me for hostname [13:26] screen just went black [13:26] now after ^d it continues though [13:27] if you didn't see this, maybe I will try to reimage this and reproduce in a moment... let me get through it as is for now thoug [13:29] oh, you have to manually make a partition on it now? === Guest78022 is now known as Ursinha === Ursinha is now known as Guest61980 [14:20] plars, ?? [14:20] plars, you need to install to USB nowadays [14:20] or yes, jump through several hoops and pre-partition the SD (with / and swap at minimum) [14:21] apart from the screen issue both, omap and omap4 server images work fine for me [14:28] I have finally managed to get omap4 installed - happened to see an "omapdss HDMI error : failed to enable GPIO's" in a tty - is that a known thing or should I report it - if so what should I report it against - thanks for any help :) [14:29] or is it tied up with https://bugs.launchpad.net/ubuntu/+source/pvr-omap4/+bug/1045491 [14:29] Ubuntu bug 1045491 in pvr-omap4 "Moving mouse messes up the desktop" [High,Confirmed] [14:29] does your system behave as expected otherwise ? [14:29] I believe so ogra_ - this is the first time I've managed to get installed :) [14:30] i think it prints that every time you swithc the display context ... i.e. from X to a tty or back, i see that here too [14:30] aah ok - thanks [14:32] though it is very very slow to respond to mouse [14:32] hmm, i didnt find it slow [14:32] i found firefox extremely slow though [14:32] everything thats 3D rendered seemed fine [14:33] (i dont think it is slower than unity-2d was with the framebuffer driver in 12.04) [14:33] but thats indeed totally subjective :) [14:34] 7 seconds to report back from free -m in aterminal, 20 secs to give the me the shutdown menu, a long time for nautilus to show anything [14:34] oh, thats disk I/O ... i thought you referred to graphics [14:35] I was responding to " does your system behave as expected otherwise" [14:35] well, using USB for / simply limits you to the max 24MB/s you can pump through a USB bus [14:35] thats still a lot more than with SD, but its just USB after all [14:36] aah ok - so that is 'expected' - if I'd thought more I'd have not asked lol [14:36] also note that there are a lot of tasks running right after install [14:36] ok - so yea - as far as graphics - I just have that bug I linked then [14:36] update-xapian-index ... locate ... lots of processes that put a horrible load on the little panda [14:37] yep [14:37] try to let it sit fro 30min until all these bits are done [14:37] ok [14:37] and also do an additional reboot, that helps ureadahead to set up a better profile [14:37] (i think it uses three runs (each run at a fresh boot) to actually build the proper profile [14:37] ) [14:38] ok - will do that - and then see how it looks - thanks ogra_ [14:38] thanks for helping !!! [14:38] :) [14:38] welcome :) [14:39] * ogra_ is happy about every additional tester [14:39] lol [14:39] I blame balloons - so you are safe :) [14:39] haha [14:40] I test this and xubuntu - ubuntu has enough people :) [14:40] true [14:41] (though i bet plars would disagree ... ) [14:41] :) [14:41] I guess there's never enough people testing in truth [14:41] yeah [14:41] hah! [14:43] ohai balloons [14:47] * balloons waves at elfy [14:48] * elfy would wave back but the channel's logged ... :p [14:49] * ogra_ ponders if having the bootloader partition automounted after install is a bu or not [14:49] *bug [14:49] we used to hide it in the past [14:54] yay, my ac100 kernel with 1280x720 hack works [14:55] ogra_, nice, I kinda want an ac100 [14:55] seems like a bit of fun [14:56] buy one then :) [14:56] hah [14:56] far too expensive. [14:56] * lilstevie doesn't think he could convince the mrs that he needs it [14:57] * LetoThe2nd repeatedly thought hard about if he needs another arm tablet/notebook device. i couldn't find any use case, so the answer obviously is no. [14:57] LetoThe2nd, http://www.amazon.de/Toshiba-AC100-10V-Netbook-Android-schwarz/dp/B003YJ67T2 [14:57] 169€ [14:57] hmm... good price. [14:57] ogra_: do they ship to UK? [14:58] i think you can still get them used on amazon.co.uk somewhere :) [14:58] ogra_: jsut for _having_ but not _using_ 170€ is still too much. [14:58] xnox, you likely dont want a german kbd :) [14:58] ogra_: yes. I want US one [14:58] there are no US ac100's [14:58] ogra_: https://bugs.launchpad.net/ubuntu/+source/linux-ti-omap4/+bug/1045855 any suggestions what else I could try here? [14:58] Ubuntu bug 1045855 in linux-ti-omap4 "usb keyboard doesn't work during installation of ubuntu-server on panda" [Undecided,New] [14:59] uk, fr, russian, german iirc [14:59] ogra_: US never gets the cool toys :/ [14:59] plars, do you have a serial console enabled when it doesnt work ? [14:59] ogra_: according to amazon.co.uk 169 EUR is 287 GBP, instead of 133 GBP [15:00] i have seen d-i take the input from serial while displaying on framebuffer before [15:00] ogra_: I do, but when I switch over to the serial pre-env, it seems that nothing is shown on the graphical display, so there's nothing to interact with [15:00] xnox, heh, yeah, they are bad at math ... especially when it comes to GBP [15:01] ogra_: I did try pressing keys on the serial console, just to see if it would affect what was happing on my screen, no luck [15:01] plars, well, for laughs, try hitting the up/down keys in your serial console and watch the monitor [15:01] ogra_: i think they thing they are rather good with math, and customers aren't [15:01] ogra_: yeah, tried that [15:01] ah, you did [15:02] I'll play with it some more later... let me get through some more tests [15:02] yeah, as lon as seerial works i'm not too concerned [15:02] plain serial installs are fine for server ... [15:03] (not that we dont need to fix this, but its not a beta blocker imho) [15:03] ogra_: right [15:03] I know what you mean :) [15:03] it's possible to install for sure, so it's good enough for beta [15:03] yep [15:03] i'll have to write a little novel for the release notes though === Guest61980 is now known as Ursinha [16:16] Not sure if anyone has seen this yet, but it looks interesting. http://cubieboard.org [16:17] like a beaglebone :) [16:18] Except half price & with video + SATA. [16:19] that an all-winner based? [16:19] ahh right [16:19] yea A10 [16:20] the headphone jacks are really clever placed [16:29] indeed [16:34] someone needs start mainlining the allwinner stuff [16:34] hehe [16:54] another allwinner based board? [16:54] yep [16:54] very cheap, and mali GPU [16:54] suihkulokki: I think there are a few folks working on that already [16:55] problem is that people got it all wrong at the beginning, it needs rework before being able to send anything upstream [16:55] It looks a lot like what luke was trying to get done with PCMCIA-format A8, but only 1GB RAM :-( [16:58] wookey: yea it has some nice features and good size, but between lack of good mainline support and low amount of ram..... [17:07] robclark: argh, worked fine with your config, need to check what caused the issue with the original one I used === forestpiskie is now known as elfy [19:14] rsalveti: have a fun trip back home? === chrisccoulson_ is now known as chrisccoulson [19:38] well omap4+armhf is working better than it did except for some wierd flashing in unity. why are fonts corrupted in the terminal in openbox? [19:41] jimerickson: because they hate you [19:41] i thought as much!