[05:57] Think I can fully install Ubuntu on the Nook Color with this processor ARM Cortex A8-based Ti OMAP 3621? Instead of that stupid Ubuntu ARM over VNC on Android trick? === hrww is now known as hrw [09:56] can someone copy me bootcmd from normal ubuntu uboot for panda/beagle? [10:26] ~summon ogra [10:28] * ogra waves [10:28] http://pastebin.com/YXWdZTfe - lovely hang [10:29] ubuntu mlo + uboot + .35 kernel [10:29] same with linaro mlo (1.44) [10:29] worked with .37-linaro kernel [10:29] and my board shutdowns when ubuntu/xload + ubuntu/uboot + linaro.37 are used [10:30] weird [10:30] we dont have such probs [10:31] I will dig other psu to compare [10:31] http://cdimage.ubuntu.com/ubuntu-netbook/releases/natty/alpha-3/ [10:31] try with this image, it definitely worked on all our boards [10:31] .35 kernel, and the x-loader and u-boot from natty [10:32] afaik we are using all the same x-loader and u-boot linaro packages [10:32] will check [10:32] i have never seen such an x-loader hangs message that late in the boot [10:33] if image will fail to boot I will start to send more curses^Wcomplains in TI direction [10:33] only if the FAT wasnt ok but then usually before the kernel gets exectuted [10:33] fsck passed [10:34] doesnt matter [10:36] x-loader-omap4-panda and u-boot-linaro-omap4-panda are the binary packages we currently use in ubuntu === JamieBen1ett is now known as JamieBennett [10:43] outer_flush_range will flush L1 and L2 or only L2? =) === ericm_ is now known as ericm [10:49] hi, has anyone got s-video working correctly on the beagleboard xM? [10:49] it kinda works with the natty alpha-2 image [10:49] i think GrueMaster wanted to test it once [10:49] but you have to wait until he gets up (US Pacific TZ) [10:50] however the resolution is not correct (larger than 640x480) and the left part is off-screen [10:50] twaeaking the timings doesn't work. only the predefined "ntsc" and "pal" are accepted [10:51] everything else -> invalid argument [10:52] so far so good - booting [10:53] with what files ? [10:53] and what did you do to get there ? [10:53] orga: are u talking to me? [10:54] no to hrw [10:54] ok [10:54] i have no svideo around to even test so i cant tell much about it [10:55] ogra: alpha3 image [10:55] ah, k [10:55] intresting [10:55] ogra: booted to language selector and have to findout where my usb devices are [10:55] if you used the same packages it can only be your formatting [10:56] so linaro-image-tools [10:57] right, likely a wrong formatting of the vfat [10:57] time to report bugs against installer? [10:58] which installer ? [10:58] l-i-t ? [10:58] oem-setup or how ubuntu installer is named [10:58] hmm, why ? [11:00] keyboard selector lacks Polish chars in country/layout lists [11:00] "Po?udniowoafryka?ski" or "Pakista?ski" [11:01] font has "ł ń" characters cause they are used in other places [11:01] hmm, intresting [11:03] " S-video resolution is 720x574 (pal) or 720x482 (ntsc). The edges of that [11:03] area are usually not shown by the TVs. To get a framebuffer that is [11:03] fully visible, you need to resize and move the overlay." [11:03] ah! didn#t know that.. [11:03] http://42.pl/u/2z8N [11:03] ogra: http://42.pl/u/2z8N [11:05] intresting [11:06] they dont wear bras [11:06] :P [11:06] oO( erm... what? ) [11:06] the pic hrw linked to [11:07] gnah. that one again. [11:11] nice thing is that installation does not require mouse [11:11] desktop is unusable without it [11:12] yeah, there is a bug open for unity-2d accessibility [11:13] you can select classic gnome at the login manager though [11:13] that should be usable by kbd [11:13] cyanish framebuffer works [11:13] indeed [11:14] * ogra wonders whern someone will add fi fix for the color offset [11:14] ogra: there are rumours that 2.6.66 'ivil edition' will have it [11:15] should really not be to hard to manipulate the rgb defaults [11:16] your font stuff is intresting, seems to only affect the selection list, all other parts of the app have the right encoding [11:17] exactly [11:17] #ubuntu-installer exists? [11:17] yep [11:17] talk to ev [11:18] thx [11:36] installation works w/o mouse EXCEPT time zone setting [12:15] orga: ping [12:15] ogra: ping [12:16] ogra, meet ppisati ... [12:16] ogra, is going to be doing some arm work on the kernel team [12:34] looks like ogra is hiding from him [12:40] ppisati, apw, awesome [12:40] !! [12:40] janimo, whats that about your NEON patch to Qt ? [12:45] ogra: Hi! i just got a mvl-dove board, do you know where i can find an image for it? [12:47] ogra, FTBFS fix, some files were not properly passed -mfpu=neon so gcc bailed [12:47] needed a change to gui.pro in the sources [12:47] I'll update the bugreport [12:47] janimo, i see that, but i would know about the implications [12:47] * janimo is annoyed by whatver changed lately in natty that fails to highlight xchat pings [12:48] ogra, we'll see the implications once it builds I guess [12:48] are there now *any* files in the archive that have NEON enabled in a hardcoded way ? [12:48] there are a few files built with neon options [12:48] the rest of Qt without [12:48] thats not acceptable [12:48] so hopefully only when those are enabled at runtime will run [12:49] I looked over the bugreport comments and am not 100% sure I understand everything [12:49] all NEON paths in code need to be runtime switched or off by default [12:49] however my patch so far is not assuming it fixes neon [12:49] just fixes the build [12:49] then we see if neon runtime detection works and fix it if it does not [12:50] well, -mfpu=neon will enable it hardcoded for these files [12:50] ogra, I believe that is what Qt does but have not checked. A few files which are built with neon should only be used when neon is detected [12:50] ogra, right, but those function will not be called all the time [12:50] k [12:50] lets test then [12:50] otherwise there is not neon at all if we do not build at least some functions with it [12:50] i guess i need to upgrade my ac100 to natty this week :( [12:51] ppisati, http://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/maverick/release/ was the last release we built for dove (not sure how well that image works though, you might be better off with the lucid one) [12:52] ogra I also took out gnome-panel from u-2d deps, hopefully it wil not be included on the image soon [12:52] as it is not used [12:52] it is used [12:52] by gdm [12:52] I don;t think it is [12:52] if so it will be pulled in by gdm then [12:52] gdm uses a cut down gnome standard session in the greeter [12:52] I have not seen anything resembling the gnome-panel in gdm login screens [12:53] just shutdown/reboot menu but that is different [12:53] the panel at the bottom is gnome-panel afaik [12:54] in any case we need gnome-panel for the classic session, if all deps are removed i need to seed it [12:54] ogra, we also have the classis session on the image?? [12:54] sure [12:54] I thiught unity and unity-2d is all [12:54] no [12:54] oh boy [12:54] we have a minimal classic session as -desktop does [12:55] the whole kitchen sink. So the chances of getting a mixed up nonfucntional environment like now compiz crashing+gnome classic creates on x86 [12:55] the only thing we additionally use is unity-2d, else our setup needs to match -desktop [12:55] doesnt carsh for me here [12:55] I have not yet run unity successfuly [12:55] wow [12:55] graphics card ? [12:56] and am not sure the bothed experience I get with classic now is due to unity bits remaning or gnome-calssic changed as well [12:56] intel [12:56] had workse on ati r300 [12:56] I mean it runs for a while then crashes [12:56] hmm, my 965 works fine here [12:56] with unity as well as witch classic [12:56] dual screen support looks weird - tooltips on a different screen than the buttons they belong to [12:57] I seea deluge of unity bugs and I do not have the time to fight it [12:57] ah, i only use an external screen on my laptop, but in single screen mode (laptop LCD off) [12:57] this is GMA4500 [12:57] ah, thats newer [12:58] maybe it is the dual setup, anyway I cannot image how this is going to be very stable in less than two months [12:58] heh [12:58] did the gnome-classic experience change too? [12:58] ask the DX team, they seem to be confident [12:58] (from what i gather in the release team meetings on fridays) [12:58] I see only one menu now in the corner (not app/places/system) [12:59] gnome-classic should look like it did before [12:59] i.e. in maverick [12:59] and the app menus are gone - I thought only unity moves them to top of screen [12:59] i have appmenus here [12:59] ok, so unity bits stay around even if compiz crashes. convenient [13:00] I guess I need to reboot more often and not just suspend/resume [13:00] i have to use classic because i just digitize my LP collection, audaciy (and i guess other wxwidget apps) doesnt work right with the appmenu in panel [13:01] runs as smooth as always [13:01] ok, that is reassuring [13:01] ogra, have you seen the debian-cd merge request? [13:02] hmm, no [13:09] janimo: I did [13:09] janimo: I flagged it as "I should really look at it" but didn't get the time [13:14] lool, ok thanks [14:00] gah [14:01] wh does update-manager install postfix on a maverick->natty upgrade === zul_ is now known as zul [14:09] morning [14:09] hey, isnt it a public holiday in brazil ? [14:10] or only if you dance ? [14:11] yup, carnival [14:12] but I'm more to rock and roll than the usual carnival dance ;-) [14:12] ogra: janimo: in theory the neon issue would be solved by compiling just the neon affected files with neon support === doko_ is now known as doko [14:12] and then at runtime it'd load the correct lib [14:12] that's what I think it was changed for qt 4.7.2 [14:13] rsalveti, yeah, the question is if thats really the case [14:13] we'll see :-) [14:13] but i'm just upgrading to natty here [14:13] so as soon as it built i can test [14:14] (if that upgrade ever finishes) [14:14] ogra, only one day left to go :) [14:14] yeah, possibly the same for my upgrade :P [14:14] it just finished buildin on a panda here, started lastd night [14:15] ppisati: welcome! [14:15] * ogra urgently needs to re-roll the ac100 kernel with all patches here ... the MMC driver doesnt know about fast cards and limits itself to 16M/s [14:16] ogra, do you have the sources to that kernel? can it be patched to 2.6.38? [14:16] it cant, but some people are working on moving it to .36 atm [14:16] once thats done the next level shouldnt be to hard [14:17] problem is that apparently the clocks arent set anywhere in the 2.6.29 code so they currently use random values and try to determine the right ones [14:37] rsalveti: yeah :) [14:39] ogra: you have devscripts installed? [14:39] hrw, only with --no-install-recommends [14:39] might be that u-m doesnt respect that though [14:40] which is annoying [14:42] ericm: ping [14:43] ppisati, pong [14:43] ppisati, did you get the package? [14:43] ericm: got the board, but i'm unable to get it running [14:43] ppisati, what's wrong? [14:43] ericm: wait let me tell you what i did [14:43] ppisati, 'k [14:44] ericm: i got a mvl image (http://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/maverick/release/ubuntu-netbook-10.10-netbook-armel+dove.img) [14:44] ericm: dd it on a sd card [14:44] ericm: put the sd card in sd 1 or 2 (tried both) [14:44] ppisati, ah - it doesn't boot from SD [14:44] ppisati, try USB [14:45] ericm: ah [14:45] ericm: so i need a usb stick? [14:45] ppisati, yep [14:45] ericm: k, let my try [14:49] ericm: does it make any difference in which usb port i plug the stick? [14:49] ppisati, nope [14:50] ppisati, but you can try [14:50] ppisati, if it doesn't boot on one or the other [14:51] ericm: the serial should work, right? [14:51] ppisati, I remember so - but last time I was unable to see anything from there [14:52] ppisati, cannot remember of if it's now using the USB serial besides that RS232 connector [14:52] ericm: but at least the vga output should spit something, right? [14:53] ppisati, yep [14:53] ppisati, or you may want to try a different monitor [14:53] ppisati, I have two monitors here, one work, the other doesn't [14:53] ppisati, :-) [14:56] ericm: no signs of life :( [14:56] ppisati, :-( [14:56] ericm: which .img did you use with that board? [14:56] ppisati, I used 10.04 [14:57] as i said above, maverick might not work [14:57] ogra: k [14:59] ogra, cannot find the 10.04 dove image though, strange [14:59] ericm: yep, it's not there [14:59] might be on releases.u.c [15:04] ogra: found it, let me try that [15:28] GrueMaster, hmm, when did you do your last natty upgrade test, cups seems to hang here upgrading the ac100 to natty [15:31] i wonder if thats a recent regression or something new [15:35] ogra: no luck either with 10.04 [15:37] weird [15:38] yep :( [15:38] ppisati, talk to GrueMaster and NCommander, they are the only ones still owning dove boards in the arm team [15:38] ogra: ok, thanks [15:38] (we havent touched it actively in the team since a year, NCommander did a special spinnoff for the maverick release) [15:39] NCommander: ping [15:39] GrueMaster: ping [15:40] what revision is your board ? [15:41] ogra: is it written on the pcb? [15:41] there should be a sticker [15:41] i think we only supported X0 and upwards in the end [15:41] not sure though [15:41] ogra: actually ericm said he was using this [15:42] ppisati: morning. [15:42] hmm, k [15:42] (barely) [15:42] GrueMaster: morning [15:42] GrueMaster, ppisati is our new arm kernel victim :) [15:42] I'm sorry. [15:42] :) [15:42] lol [15:42] ppisati, it's A0 and lucid installation should be working [15:42] ppisati, I installed it days ago [15:43] ericm: usb stick and vga output, right? [15:43] ppisati, yes [15:43] janimo: saw a new qt4-x11 update today. this should be repeated after today's gcc-4.5 is in the archive [15:44] doko, you mean rebuilt :) [15:44] ppisati, is the hex LED increasing ever? [15:44] and the ugly hack dropped [15:45] ogra: actually it's completely off [15:45] ogra: while a couple of other green leds are on [15:45] ogra: whatever ... [15:46] ogra: did you use an atx power brick to power it? [15:46] doko, i meant to say ... its high up on our radar, dont worry :) [15:46] ppisati, yes and a sata disk ... but i only had a Y0 [15:47] thats loooong ago [15:47] ogra: k [15:47] ericm: and did you use an atx power brick too? [15:47] ppisati, yes [15:47] ericm: and do you remember if the HEX leds were on? [15:48] ericm: i mean, even when it didn't boot [15:48] ppisati, it's powered on as '7' [15:48] ppisati, or '0' [15:48] ericm: i'm trying to rule out a power problem [15:48] ppisati, and when kernel actually boots, it starts increasing from 0 to 7 [15:48] ppisati, it's a standard ATX, shouldn't be much a problem [15:48] ericm: uhmm... [15:50] ppisati, have to sleep now - let me know if you succeed with that crappy board :-) [15:50] ppisati, good luck === ericm is now known as ericm-Zzz [15:52] ppisati: I can work with you on getting your dove running, but I need my coffee first. Give me 15 minutes. [15:54] GrueMaster: k, thanks [16:20] ppisati: Ok, I'm semi-awake now. Have you gotten anywhere with your dove board yet? [16:23] GrueMaster: nope [16:23] Are you getting power? [16:23] GrueMaster: i tried with a Lucid and Maverick image on sd and usb [16:23] GrueMaster: well, i'm using an ATX power brick [16:23] GrueMaster: the HEX lcd is off, but a couple of green leds on the board turn on [16:24] GrueMaster: so i guess power is ok [16:24] GrueMaster: could you try to power on a board with no usb/sd inside and tell me if the LCD turns on? [16:25] No, that just means standby power is available. Press the power button. It is on the mainboard, next to the barrel power connector . [16:26] I am currently rebooting mine into Lucid. Works just fine. [16:27] GrueMaster: ... [16:27] GrueMaster: i fell really dumb now... [16:27] heh. [16:27] GrueMaster: thanks [16:28] No problem. And don't worry, I won't start your beer tab for at least a week. :P [16:30] ogra: What was the issue you had with Cups? [16:30] GrueMaster: :) [16:47] GrueMaster, service restart/stop from the preinst/postinst scripts seemed to hang [16:47] Hmmm. [16:47] This is in natty? [16:47] maverick to natty on the ac100 [16:48] Not that I have an ac100 to test with, but I'll try to reproduce it today on panda. May take a while (install, update, upgrade). [16:49] yeah [16:49] 3h up to now on the ac100 [16:49] i just used do-release-upgrade on cmdline though [16:49] Ok. I'll keep that in mind. [16:50] ogra: you finally upgrade? [16:50] yes [16:50] need to test neon bits [16:50] ;D [16:51] I am waiting for courier to replace alix.1c (amd geodelx x86) with d510mo (d510 64bit atom) [16:51] will see how good/bad ubuntu server will work on my router [16:52] its just x86 hw ... [16:52] it will work well [16:54] I had binutils-powerpc-linux-gnu on router... [17:09] ppisati: Did your system boot? [17:27] doko, I noticed the new gcc contains the fix, but I wanted to do an upload so I don't wait another day, and more importantly so that we have a 4.7.2 in a a hopefully working state with gcc 4.4 and have the compiler change as a separate step , in case somehtign goes wrong we don't have to scractj our heads againt whether it is a toolcahin on upstream change that caused the difference [17:27] but yes, we need to rebuild with gcc 4.5 ASAP [17:27] janimo, well, we also need to drop the patch anyway [17:28] i guess if we take care by thu. it will be ok [17:28] both packages will have built then [17:28] what is thu, another milestone? [17:28] yeah, the weekly thursday milestone ... comes right before the weekly friday one ;) [17:28] ogra, by tomorrow we'll see if neon runtime detection work adn whether another qt change is needed [17:28] :) [17:29] right [17:29] i dont think we are in a hurry with the rebuild ... though the kde people might disagree [17:29] not sure how much pressure has built up there [17:35] ogra, whay would they care that much? [17:35] because not all of kde is built yet ? [17:35] as long as it is agreed it is a temporary measure and has no drawbacks [17:35] * ogra hasnt checked the list [17:37] ogra, ah, I thought it was in reference to qt [17:37] nah, kde [17:38] indeed kde needs rebuilds, I think this is taken care of kubuntu devs unless it is arm specific FTBFS [17:38] I think it is just the usual deps not yet met thing [17:38] yep [17:42] xulrunner-1.9.2 (1.9.2.14+build3+nobinonly-0ubuntu2) wird eingerichtet ... [17:42] Segmentation fault (core dumped) [17:42] hrm [17:42] not good [17:47] Hey, I got Blaze running natty A3. replaced x-loader & u-boot with rsalveti's secret sauce and everything looks good, even unity-2d. [17:47] GrueMaster: \o/ [17:48] GrueMaster: which kernel are you using? [17:48] awesome [17:48] Default kernel that is in the image. 2.6.35-1101.4 I think. [17:48] GrueMaster: thx [17:54] ogra: rsalvetiWhat are the chances of having both panda & blaze versions of uboot & xloader in the deb package? Or are they heading towards a single binary? [18:04] Ugh. Fresh install of maverick, then dist-upgrade == 242 package updates. [18:08] GrueMaster: they will be the same binary [18:09] but we're still waiting correct sdp support from ti with x-loader upstream [18:09] Cool [18:09] oh, sorry, at least x-loader needs to be a different binary [18:09] because of the hardware id [18:10] It can't auto-detect? [18:10] it could be the same binary, but then it'd need a way to identify the hardware before actually initializing it [18:20] hrw: the x-loader you were using is a quite old one [18:20] that could be your problem [18:20] or even a simple fat problem with your first disk partition [18:49] rsalveti: linux-image-2.6.38-1204.5 working on my panda. Even have dvi output (shocking). Thought I might need a rework. [18:49] GrueMaster: awesome [18:49] GrueMaster: does it crash when you halt or reboot? [18:50] GrueMaster: please check if the sound and wireless is also working [18:50] if possible [18:50] Didn't appear to on reboot [18:50] (crash that is). [18:51] Had to reboot a couple of times. Once to add dvi settings to boot.script, and another because I forgot to add the kernel version to flash-kernel when updating boot.script (oops). [18:52] So I have booted back and forth a couple of times between .35 & .38. [18:52] Bluetooth appears to be enabled. Haven't tested yet. Wifi also appears to have been detected. [18:53] Does wifi need firmware? [19:05] GrueMaster: it needs a firmware [19:05] but probably already installed [19:05] Ok. Might need an antennae. Don't have one handy and my wifi AP is in another room. [19:07] could be even the lack of proper calibration values [19:07] sebjan may help here [19:10] GrueMaster: the wlan firmware coming with the linux-firmware package is the good one. [19:10] for calibration: http://wireless.kernel.org/en/users/Drivers/wl12xx#Calibration [19:10] (did not try that myself yet) [19:11] sebjan: ok, will check in a sec. Trying to get audio now. [19:11] GrueMaster: for WLAN, you may have a MAC address set to 0. This can hurt. For now, the best is to set a random MAC address manually (using ifconfig) [19:14] GrueMaster: fyi, here is the way I tested wlan and audio on this kernel: http://paste.ubuntu.com/577134/ [19:14] rsalveti: no audio at all. [19:15] GrueMaster: you need some mixers settings [19:15] hm, I think you need to set the mixers by hand [19:15] Yea, I was tweaking alsamixer manually. [19:16] While running speaker-test -c 2 -t wav in another terminal. [19:17] * sebjan has to go [19:17] GrueMaster: http://paste.ubuntu.com/577137/ [19:18] from sebjan's email [19:18] yep, you have this one plus the record mixers in my above http://paste.ubuntu.com/577134/ compilation ;-) [19:19] that should do it :-) [19:24] Nope, nothing. [19:24] This is on natty w/ 2.6.38-1204.5. [19:26] Hmmm. dmesg has "asoc: no valid backend routes for PCM: SDP4430 Media" repeated multiple times. [19:28] Could be pulseaudio related though. No hw in pulseaudio. No sound using alsa tools. [19:31]