**** BEGIN LOGGING AT Wed Jul 07 02:59:56 2010 Jul 07 09:01:36 sebjan: ping Jul 07 09:01:45 lag: pong Jul 07 09:02:24 Is all the Syslink code going upstream eventually? Or is it going to reside solely in the Ubuntu kernel? Jul 07 09:03:48 this version of the syslink code will not go upstream. Next version is more suitable to go upstream. Jul 07 09:04:04 Okay Jul 07 09:04:23 I sending my patch to the kernel mailing list now Jul 07 09:05:19 lag: great, thanks! Jul 07 09:08:30 Gone Jul 07 09:08:34 And now ... we wait Jul 07 09:08:37 :) Jul 07 09:15:56 any news on solution to the git clone issue on omap4? Jul 07 09:21:01 XorA: it is fixed with this patch: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-maverick.git;a=commit;h=d96ac19bf235fb639bc29ad3f7db210174e429c2 Jul 07 09:25:36 sebjan: thanks Jul 07 09:27:03 ogra: I am downgrading my bb to 6b18 openjdk Jul 07 10:12:29 hrw, dyfet now has the task to help you working on that Jul 07 10:16:31 * ogra_panda waves Jul 07 10:18:03 * cooloney waves back to ogra_panda smiling like a panda Jul 07 10:18:36 ogra_panda: we can get a working panda image for downloading now? Jul 07 10:19:57 cooloney, not yet, seb128 uploaded a new glib, once thats in the archive i can respin, while we have images, they still have the segfaulting kernel Jul 07 10:22:22 grmbl Jul 07 10:22:25 crashed again Jul 07 10:22:55 cooloney: What's the current state of bug 588243 Jul 07 10:23:09 Launchpad bug 588243 in linux-ti-omap (Ubuntu) (and 1 other project) "kernel BUG at /build/buildd/linux-ti-omap-2.6.33/drivers/video/omap2/dss/core.c:323! (affects: 1) (heat: 96)" [Undecided,New] https://launchpad.net/bugs/588243 Jul 07 10:23:25 lag: oh, sorry, no update. Jul 07 10:23:41 as long as i got the hw, i will start to debug that Jul 07 10:25:54 What hw do you have currently? Jul 07 10:26:23 lag: nothing at all -:< Jul 07 10:27:45 cooloney: That's rubbish Jul 07 10:30:06 i will got it tommorrow Jul 07 10:30:24 What will it be? Jul 07 10:30:26 C4? Jul 07 10:30:28 god, IO really sucks Jul 07 10:30:37 panda, i think Jul 07 10:31:30 cooloney: That bug I sent you affects Lucid - Beagleboard Jul 07 10:31:41 the MMC driver surely needs love Jul 07 10:32:17 ogra_panda: Which board? Going by your name, Panda? Jul 07 10:32:35 right Jul 07 10:32:43 What's wrong with it? Jul 07 10:32:59 tab completion can make the board stuck for 30sec Jul 07 10:33:17 * lag checks Jul 07 10:33:28 like hard hanging until the rest of the directory name is printed on the term Jul 07 10:33:48 but at least i have a working image now Jul 07 10:34:01 You do? What was wrong? Jul 07 10:34:27 segfaults over and over Jul 07 10:34:31 Are you running a full Desktop system? Jul 07 10:34:47 lag, i'm running what you should run as well, the ubuntu images := Jul 07 10:34:49 :) Jul 07 10:34:49 ogra_panda: what's the kernel version Jul 07 10:35:02 cooloney, now its the one uploaded yesterday Jul 07 10:35:07 i think we got some fixing patches in ti-omap4 Jul 07 10:35:07 901.4 i think Jul 07 10:35:17 * lag is a Kernel Engineer - GUIs are for weaners Jul 07 10:35:26 that one is a lot better but still sucks at SD I/O Jul 07 10:35:39 :) Jul 07 10:35:44 ogra: boot from sd but keep rootfs on usb (if it works) Jul 07 10:35:45 I'll download the latest Jul 07 10:35:47 lag, thats the reason why we always get broken kernels from the kernel team :P Jul 07 10:35:51 Do you have a quick link? Jul 07 10:35:59 you guys only test if serial works :P Jul 07 10:36:08 * lag sees ogra's lips moving, but doesn't hear the noise Jul 07 10:36:15 ogra: you need more? really? Jul 07 10:36:18 haha Jul 07 10:36:59 lag, http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/ wait for 20100707, i'll do a rebuild shortly Jul 07 10:37:05 with the latest kernel Jul 07 10:37:15 06 isnt really usable Jul 07 10:37:31 due to permament segfaults in 900.1 Jul 07 10:37:52 * ogra_panda ponders if he should be brave and try firefox Jul 07 10:38:01 lets see :) Jul 07 10:38:11 run firefox, come back tomorrow - maybe it will be loaded Jul 07 10:38:48 runs fine :) Jul 07 10:39:34 couls scroll a bit smoother Jul 07 10:39:39 *could Jul 07 10:40:05 but is definately the fastest i have seen on arm framebuffer yet Jul 07 10:40:27 still fbdev in use? Jul 07 10:40:46 sure, we dont have an X server for panda in the archive Jul 07 10:41:19 * lag 's monitor still doesn't want to work with the Panda board Jul 07 10:41:24 :( Jul 07 10:42:14 font rendering sucks Jul 07 10:42:20 i wonder why Jul 07 10:43:11 it runs at 1920x1080 Jul 07 10:43:19 the fonts should be smooth Jul 07 10:44:16 beyond the IO issues and the ugly fonts its as fast as an ATOM i'd say Jul 07 10:44:30 it even boots in 15sec after u-boot Jul 07 10:44:41 i.e. switching on screen to gdm Jul 07 10:45:35 htop shows 281MB used with gnome-terminal, FF and xchat open Jul 07 10:46:22 shot? Jul 07 10:46:31 lag, i think GrueMaster experimented with display settings yesterday and found several options to use DVI adapters Jul 07 10:47:11 lag: using hdmi or dvi porT? Jul 07 10:48:46 hrw, DVI port doesnt work Jul 07 10:49:43 hrw, http://people.canonical.com/~ogra/panda_shot.png Jul 07 10:49:50 23:13 < GrueMaster> IF I plug the switch into the DVI port and the monitor HDMI>DVI cable into the HDMI port, I get video. Jul 07 10:49:57 heh, the fonts look fine in the screenshot Jul 07 10:50:25 ogra: check cables? Jul 07 10:50:27 hrw, weird, afaik there is no signal on the DVI port in the current revision Jul 07 10:50:34 the cable is fine Jul 07 10:50:40 i think the frequency isnt Jul 07 10:50:57 it runs at 60Hz while it should be running at 50 Jul 07 10:51:11 sadly you cant influence the resolution at all Jul 07 10:51:24 kernel reads EDID and automatically picks the highest res Jul 07 10:51:26 it does what monitor says Jul 07 10:51:30 ogra: you can Jul 07 10:51:44 you cant Jul 07 10:51:51 hrw: I can use either Jul 07 10:51:52 18:32 <@prpplague> Magdalena: for the panda you can just pass the bootargs omapdss.hdmimode=0 omapdss.hdmicode=35 Jul 07 10:51:53 omap4 doesnt accept resolution options Jul 07 10:52:00 18:32 <@prpplague> Magdalena: that will get you the basic 1024x768 mode that is common to most displays that support vesa/dvi Jul 07 10:52:03 18:33 <@prpplague> GrueMaster: if you check the hdmi code you can see a number of display modes, using the HDMI to run a DVI display you need to use the VESA modes Jul 07 10:52:36 well, if i enforce a mode i'm not sure my monitor will be able to use it on HDMI Jul 07 10:52:45 try? Jul 07 10:52:53 i think it only can 720p and 1080p on the HDMI port Jul 07 10:53:03 its more flexible on DVI Jul 07 10:53:11 use hdmi->dvi? Jul 07 10:53:26 well, first i have a lot of other stuff :) Jul 07 10:53:34 display is my least sorrow Jul 07 10:53:55 Display is my largest-bugbear Jul 07 10:53:58 Trying ... Jul 07 10:54:00 * ogra_panda tries some high load websites Jul 07 10:55:41 hrw: Nope Jul 07 10:56:10 * ogra_panda wonders why avahi cant resolve Jul 07 10:56:47 i can only ping myself on a .loacl address Jul 07 10:57:36 *local Jul 07 10:57:41 tell me when you will resolve that Jul 07 10:57:56 ssh hrw@192.168.1.137 is less fun then ssh hrw@beagle.local Jul 07 10:58:06 yeah Jul 07 10:59:19 ogra@osiris:~$ ping panda.local Jul 07 10:59:19 PING panda.local (192.168.2.114) 56(84) bytes of data. Jul 07 10:59:19 64 bytes from panda.local (192.168.2.114): icmp_seq=1 ttl=64 time=35.6 ms Jul 07 10:59:26 hmm, the other direction works Jul 07 10:59:49 but i cant ssh into my laptop via the local address Jul 07 11:00:06 hrw: When we use HDMI->DVI, which port do I use on the PB? Jul 07 11:01:48 lag, try both ? Jul 07 11:02:05 Am doing Jul 07 11:03:55 lag: from how I understood GrueMaster config he has DVI occupied but not used and hdmi output used for display Jul 07 11:04:05 HDMI->HDMI[X] HDMI->HDMI&DVI-D->DVI[X] HDMI->DVI[X] Jul 07 11:04:07 :( Jul 07 11:04:47 And DVI-D->HDMI[X] Jul 07 11:04:51 :'( Jul 07 11:05:02 lag: connect dvi and hdmi at same time Jul 07 11:05:38 no sound :( Jul 07 11:07:33 I tried that Jul 07 11:07:46 HDMI->HDMI&DVI-D->DVI[X] Jul 07 11:09:51 setenv bootargs omapdss.hdmimode=0 omapdss.hdmicode=35 vram=32M, omapfb.vram=0:8M console=tty2 console=ttyO2,115200n8 noinitrd mem=512M root=/dev/mmcblk0p2 rootdelay=1 ip=none; mmcinit 0; fatload mmc 0 0x80200000 uImage; bootm 0x80200000 Jul 07 11:10:04 What's our cmdline ogra? Jul 07 11:11:39 only vram=32M, no further video related options Jul 07 11:11:54 mem=512M is definately wrong Jul 07 11:11:54 I've tried that oo Jul 07 11:11:56 too* Jul 07 11:12:04 What should it be? Jul 07 11:12:13 that will try to write in coprocessor occupied ram regions Jul 07 11:12:15 463 Jul 07 11:12:43 (as i understood mem=463M is mandatory, thats why we use it on the images) Jul 07 11:13:02 why the heck are you using rootdelay ? Jul 07 11:13:14 i thought kernel devs know thats deprecated :P Jul 07 11:13:19 lag: use rootwait Jul 07 11:14:07 ogra@panda:~$ cat /proc/cmdline Jul 07 11:14:07 quiet splash vram=32M mem=463M root=UUID=bdee5e48-a61b-492f-a060-6b8d1afe4364 fixrtc Jul 07 11:14:10 ... Jul 07 11:14:35 hmm, there should be a ro too Jul 07 11:14:38 weird Jul 07 11:14:43 Still no console on the monitor Jul 07 11:14:48 oh, the image used an older jasper Jul 07 11:16:09 SIGH !!! Jul 07 11:16:23 now glib is in the archive and the image builds fail on empathy Jul 07 11:16:57 * ogra wishes the desktop team would stop uploading .... our desktop is fine as it is, damned ! Jul 07 11:17:05 j/k Jul 07 11:23:01 ogra: I would say "rebuild it on your fast arm" but it suck on i/o... Jul 07 11:23:38 i guess USB would be fast Jul 07 11:23:47 but that wont get me the package into the archive faster Jul 07 11:23:55 its the publisher i'm waiting for Jul 07 12:12:42 morning Jul 07 12:20:57 ogra: nice you were able to boot panda :-) Jul 07 12:21:11 ogra: how is it now? still giving lots of seg faults? Jul 07 12:21:22 yup, its running fine, i only wish i could build images with the fixes now Jul 07 12:23:22 ogra: :-) Jul 07 12:23:37 new issues everyday haha Jul 07 12:23:49 heh, yeah Jul 07 12:23:58 people sadly upload packages ... Jul 07 12:28:46 hey robclark|panda :) Jul 07 12:29:06 gm ogra_panda Jul 07 12:29:07 * ogra_panda hands robclark|panda some bamboo Jul 07 12:29:17 yum yum Jul 07 12:29:23 :) Jul 07 12:30:05 * robclark|panda just got around to re-configuring pidgin on panda after nuking his home directory.. Jul 07 12:30:48 * ogra_panda is using xchat Jul 07 12:31:40 pidgin is at least relatively easy to get working behind proxy Jul 07 12:32:04 might be, i never tried xchat behind proxy :) Jul 07 12:32:27 i just find it painful to use IM clients for IRc Jul 07 12:32:57 yeah.. pidgin isnt really my favorite Jul 07 12:34:41 robclark|panda: irssi works well behind the proxy :) Jul 07 12:35:09 hmm.. ok.. something to try Jul 07 12:43:32 xchat works fine behind a proxy, that's how I always use it Jul 07 12:43:44 hence my lack of a hostmask :p Jul 07 12:59:46 ha ! Jul 07 12:59:57 finally an image build that doesnt die Jul 07 13:01:16 so in 2-3h we should have a panda image thats actually usable Jul 07 13:21:39 IT WORKS! Jul 07 13:22:12 lag: syslink? Jul 07 13:22:29 lool: That's old news - that worked yesterday Jul 07 13:22:30 :) Jul 07 13:22:40 My monitor - w/Panda Jul 07 13:22:45 erf Jul 07 13:23:25 Equine Rescue France? Jul 07 13:29:23 lag: what helped? Jul 07 13:30:30 I'm now running the release (CD) image Jul 07 13:30:41 I still don't know what the matter was with the other one? Jul 07 13:30:55 Perhaps the rootfs was incorrect? Jul 07 13:38:29 ogra_panda, ogra_cmpc, ogra: How long did your Panda image take to come up? Jul 07 13:39:30 Mine has been sat at the "will take approx 10mins" screen for about 15mins Jul 07 13:39:34 Then it just went off Jul 07 13:42:05 sebjan: Ping Jul 07 14:16:06 ogra? Jul 07 14:17:47 npitre, do you have any idea why Kirkwood keeps PCIBIOS_MIN_IO as 0x1000 and PCIBIOS_MIN_MEM as 0x01000000? Jul 07 14:19:49 ericm|ubuntu: no idea -- probably Lennert would know Jul 07 14:20:10 * GrueMaster yawns, stretches, and looks for coffee. Jul 07 14:20:22 npitre, ok - I'll drop him an email Jul 07 14:20:28 GrueMaster, wakeup from hibernation? Jul 07 14:20:42 Something likethat. Jul 07 14:20:56 * GrueMaster needs a new keyboard. Jul 07 14:21:47 ukleinek: would you mind removing that only patch of yours I object to from your Git tree, add the latest ACKs and send RMK a pull request? Jul 07 14:22:13 ukleinek: I think that stuff might be ready for mainline Jul 07 14:22:46 npitre: no, I can. My only concern is that the uImage target is broken then Jul 07 14:22:57 Ok, I have video on my panda working now using a HDMI>DVI cable to a pc monitor. It is in the HDMI port. I am also using the following kernel parameters: Jul 07 14:22:57 omapdss.hdmimode=0 omapdss.hdmicode=35 Jul 07 14:24:08 GrueMaster: My board worked HDMI->HDMI once Jul 07 14:24:20 Then on reboot - nothing Jul 07 14:24:34 Then I re-flashed the SD card with the same image and rebooted Jul 07 14:24:39 Still nothing Jul 07 14:25:35 Try also adding console=ttyO2,115200 console=tty0. That way you can see what is happening in the kernel. Jul 07 14:26:18 It's not my cmdline - this is fresh from the CD archive Jul 07 14:26:28 lag, his a key, thats DPMS :) Jul 07 14:26:36 *hit even Jul 07 14:26:47 ? Jul 07 14:27:01 GrueMaster, wait for the new image that has a less buggy kernel that doesnt hang all the time Jul 07 14:27:11 Mine has been sat at the "will take approx 10mins" screen for about 15mins Jul 07 14:27:11 Then it just went off Jul 07 14:27:20 I tried that Jul 07 14:27:21 Some issues I ran into yesterday included not being able to use my HDMI switchbox with panda (sad face), sometimes losing keyboard input, but not mouse (I have a keyboard with built in touchpad), and occasional system hangs. Jul 07 14:27:22 But nothing Jul 07 14:27:29 oh, you mean after reboot ? Jul 07 14:27:41 I don't even know if my keyboard is compatible Jul 07 14:27:46 Yeah Jul 07 14:27:48 yeah, thats a known bug with the old images Jul 07 14:27:56 wait for 20100707 Jul 07 14:28:04 * GrueMaster goes to see if coffee is ready. brb Jul 07 14:28:06 I even flashed the card again, but nothing! Jul 07 14:28:06 that has a fix for it Jul 07 14:28:29 hmm, if you freshly flashed the card it should work indeed Jul 07 14:28:43 Correct Jul 07 14:28:46 though i wouldnt trus that kernel at all Jul 07 14:29:11 the new image should be up in about 1.5h Jul 07 14:29:58 Okay, I guess it's worth waiting for Jul 07 14:30:07 Is it a kernel bug? Jul 07 14:30:47 well, display should still work Jul 07 14:30:59 but the kernel throws random segfaults Jul 07 14:31:08 so it can die at any point Jul 07 14:31:29 i had to do five attempts until resizing worked yesterday Jul 07 14:31:43 because it locked up hard in the middle every time Jul 07 14:31:49 ukleinek: we'll need to not use the dynamic phys offset (which is off by default) until mkimage/u-boot is fixed, so that shouldn't prevent this stuff from being merged Jul 07 14:31:54 lag: pong Jul 07 14:32:27 ukleinek: could mkimage simply be used without any load addr? Jul 07 14:32:46 Hey sebjan Jul 07 14:32:55 Where you in a meeting with Nicolas? Jul 07 14:33:21 lag: yes, why? :) Jul 07 14:33:33 npitre, 0x0 should work Jul 07 14:33:38 Because you both came back at the same time :) Jul 07 14:33:57 I was wondering what the verdict was with regards to my patch? Jul 07 14:42:06 npitre: no, passing a loadaddr is obligatory Jul 07 14:44:49 lag: I would expect some experts reviewers to provide comments. Fyi, I will take your current patch for internal usage, and then update to whatever it ends-up after review. Jul 07 14:45:23 No problem Jul 07 14:45:57 npitre: pull request sent Jul 07 14:53:27 ukleinek: can this limitation be fixed? Jul 07 14:53:59 npitre: I didn't touch U-Boot for some time. I think they have a new image format, but I don't know the details Jul 07 14:54:05 ukleinek: something like a loadaddr of -1 meaning that it has to be provided at load time or the like Jul 07 14:54:23 IIRC U-Boot can boot zImages, too Jul 07 14:54:34 Barebox can, too Jul 07 14:55:21 ukleinek: last time I booted zImage with uboot, the ATAG block wasn't set up Jul 07 14:55:49 ukleinek: nor the r1 machine ID Jul 07 14:57:06 npitre: barebox does it right, don't know about U-Boot for sure Jul 07 15:10:28 ogra: I see uboot-envtools and uboot-mkimage packages in the archive. Where is the u-boot source? Jul 07 15:10:45 u-boot-omap3/4 Jul 07 15:11:01 i dont know why debian chose to use no dash Jul 07 15:11:09 upstream clearly does Jul 07 15:12:18 ogra: in lucid? Jul 07 15:12:30 only omap3 there Jul 07 15:12:54 ogra: don't see any u-boot-omap3 Jul 07 15:13:01 * amitk checks deb-src lines Jul 07 15:13:24 amitk, https://edge.launchpad.net/ubuntu/lucid/+source/u-boot-omap3 Jul 07 15:40:27 GrueMaster, lag, http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/20100707/ Jul 07 15:40:38 happy playing Jul 07 15:40:46 yea, I'm seeing them come in. Jul 07 15:41:03 i think we should disable zsync generation Jul 07 15:41:20 swap it for MD5sums. :P Jul 07 15:41:20 it gives a flase impression that users would gain anything by using it Jul 07 15:41:34 GrueMaster, assigned to NCommander Jul 07 15:41:42 dont blame me :) Jul 07 15:41:48 I know. I was kidding. Jul 07 15:42:02 ogra: GrueMaster: its sad that we can't use it, bu tI'll look at it once my headache subsides Jul 07 15:42:03 Will the omap one work for XM? Jul 07 15:42:28 lag, you could try it with a different u-boot binary Jul 07 15:42:40 lag, if you find one that works Jul 07 15:42:52 Well I have booted the XM Jul 07 15:42:53 beyond that it should work Jul 07 15:42:57 So I do have one Jul 07 15:43:04 When will it work out-of-the-box? Jul 07 15:43:08 oh, then *your* XM might even work with that image :) Jul 07 15:43:15 mine doesnt Jul 07 15:43:49 kernel surely is supposed to work and beyond that there should be no difference in userspace Jul 07 15:44:00 feel free to try it :) Jul 07 15:44:02 Are we doing a single image now or is the omap4 image still churning? Jul 07 15:45:09 i see both Jul 07 15:45:15 * ogra reloads Jul 07 15:45:21 still Jul 07 15:46:22 Can someone point me to a list of hdmi modes? Jul 07 15:46:31 35 is out of range on my monitor Jul 07 15:46:41 35 is 1280x1024 Jul 07 15:46:48 i think there are some in the omap4 dss driver code Jul 07 15:46:51 beyond that, I don't know. Jul 07 15:47:07 This is odd Jul 07 15:47:18 Booted with 35 "out of range" Jul 07 15:47:40 the panda ? Jul 07 15:47:58 Yeah Jul 07 15:48:12 oh, i forgot, it doesnt like your EDID Jul 07 15:48:22 But it's working Jul 07 15:48:25 With my own kernel Jul 07 15:48:29 But only with 35 Jul 07 15:48:35 11 doesn't even come up Jul 07 18:17:45 * zumbi sends armin76's dog to eat all the germans arround :) Jul 07 20:09:14 ogra: Hey there Jul 07 20:09:23 ogra: Do you know if the A2 kernels work on Beagleboard XMs? Jul 08 02:02:52 if i put the new kernel uImage-2.6.34 in the boot directory and extract the kernel modules to the root fs, will that successfully update the kernel? Jul 08 02:03:01 ..and try to reboot? Jul 08 02:04:49 or is there docs on how to upgrade the kernel on a sheevaplug? **** ENDING LOGGING AT Thu Jul 08 02:59:57 2010