/srv/irclogs.ubuntu.com/2010/08/13/#ubuntu-arm.txt

rsalvetiGrueMaster: still building, soon you'll get it02:12
GrueMasterNo rush.  I've started packing for my flight tomorrow.  I'll test first thing Monday once I'm setup.02:13
rsalvetiGrueMaster: oh, are you leaving tomorrow?02:13
GrueMasterI'm feeling rather bushed and my son is down for the weekend to celebrate his 21st birthday with his friends.02:13
GrueMasterI fly out tomorrow morning.02:13
rsalvetiGrueMaster: quite soon :-)02:16
rsalvetiwell, it's a long flight for you02:17
GrueMaster13 hours or something like that.02:17
rsalvetihm, about the same for me if I'd go there02:17
rsalvetitalking about trips I need to book my uds trip02:18
rsavoyewhere's UDS going to be this time ?02:19
rsalvetiorlando02:19
rsavoyeyeuch02:20
rsavoyedates ?02:20
rsalveti 25th - 29th October02:20
rsavoyethanks. I know some nice place in Colorado we could have it...02:20
rsavoyeI thought I saw a new kernel get uploaded today ?02:33
rsavoyestill trying to get my XM board to work stably...02:33
lagcooloney: Morning09:47
mopdenackerogra: Hi! I started testing your10:26
mopdenackerhttp://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/20100812/maverick-preinstalled-netbook-armel+omap4.img.gz image10:26
mopdenackeron my PandaBoard ES 1.0.10:28
mopdenackerI get the below error message after the reboot:10:28
mopdenackerhttp://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/20100812/maverick-preinstalled-netbook-armel+omap4.img.gz10:28
mopdenackerOops, here's the error: http://pastebin.com/iKWD7Umg10:28
cooloneylag: morning man10:34
lagcooloney: Hey Bryan10:35
lagcooloney: How are you getting on with your packaging work?10:35
cooloneylag: yeah, just rebased to our 2.6.35 master tree10:36
cooloneyand sent out an email to discuss the maintanenc method10:37
cooloneylag: how about your review?10:37
cooloneyi think ndec is still working on the integration tree, since there L24.9 branch is increasing10:37
mopdenackerogra: I wonder what's responsible for the "Exec format error" message.10:39
=== fta_ is now known as fta
lagcooloney: Yes, I saw your email10:43
lagcooloney: I was chatting with apw and tgardner yesterday about the maintenance issues10:44
lagcooloney: It's quite a complex one10:44
lagcooloney: It would be easier if TI's tree was based solely on Linus' 2.6.35 tree10:45
cooloneylag: right, i also asked that question in the email to ndec10:46
cooloneylag: but that's TI's internal process. i hope we can get such ti patch stack on Linux 2.6.35 release10:47
cooloneylag: currently, it is not easy for them to change.10:47
lagcooloney: How did you do the last one?11:09
cooloneylag: sorry, what's the last one?11:12
lagThe last release11:12
cooloneyoh, the last release is simpler, since we just rebased TI's 230+ patches on top of our tip tree.11:13
=== amitk is now known as amitk-afk
lagSo what makes this any different?11:18
cooloneylag: actually no big difference to me, just 1300+ patches vs 230+ patches.11:21
cooloneyand it is 2.5.35 based release11:21
cooloneywe cannot simple rebase on our -ti-omap4 branch, but need to rebase on our master11:22
cooloneysince our master is also 2.6.35 based11:22
lagWhen I do a git merge onto master, I only receive one (easily correctable) issue11:22
cooloneyyeah, that's it.11:22
cooloneyi got 2,11:22
cooloneytoday.11:22
cooloneybut 2 of them are omap3 related11:23
cooloneyso it won't be much difficult.11:23
cooloneyapw's method is not necessary this time11:24
lagYou have to revert ce8bd5fb89ad58d345a5eaef52e7fa6f5683f5a911:27
lagThen you only get one11:27
lagcooloney: -^11:28
cooloneylag: thx, man, you fixed 50% rebasing issues11:29
cooloneylag: and as ndec said in his latest email, we still need to wait for their integration release11:30
cooloneythat's our upstream, L24.9 is just their monthly release tree.11:30
lagWell, why don't we wait then?11:35
lagI'm guessing sebjan will ensure their tree _fits_ on ours11:35
lagcooloney: The other conflict should be easy too11:36
lagcooloney: The one in HEAD looks far more correct than the other11:36
lagJust delete11:37
lag=======11:37
lagdss_clk_enable();11:37
lagr = dsi_pll_init(dssdev, 0, 1);11:37
lag>>>>>>> L24.911:37
lagand obviously <<<<<<< HEAD11:37
lagThe result of the successful merge can be found here: http://kernel.ubuntu.com/git?p=lag/ubuntu-maverick.git;a=shortlog;h=refs/heads/ti-omap4-3511:37
lagNow I just need your packaging work to test11:38
cooloneylag: how did you do that? git merge?11:50
lagcooloney: Yeah11:50
cooloneylag: i found you missed some patches from L24.9 branch11:50
lagOh?11:50
lagWhen were they released?11:51
lagGive me the commit ids11:51
cooloneylag http://dev.omapzoom.org/?p=integration/kernel-omap4.git;a=shortlog;h=refs/heads/L24.911:51
=== fta_ is now known as fta
cooloneythey updated their tree yesterday, i think11:52
cooloneythe release for us is supposed to be next week.11:52
cooloneyndec and sebjan will work on it.11:52
cooloneyplease find my rebase result here11:52
cooloneyhttp://kernel.ubuntu.com/git?p=roc/ubuntu-maverick.git;a=shortlog;h=refs/heads/ti-release-rebasing-testing11:52
lagcooloney: I didn't miss anything - they weren't in the tree then11:53
lagI thought it was for Friday?11:53
lagcooloney: It's pointless merging now, they are going to be updating it daily11:54
lagWe should wait until our proper release date11:54
cooloneylag: right, we have to wait11:54
lagndec: Are you there?11:54
cooloneyso i just test rebasing is simple for this release11:54
lagcooloney: Me too11:54
ndeclag: hi!@11:54
lagndec: Hello :)11:55
lagWe've been discussing how best to handle your tree11:55
ndeclag: oh i see... i saw the emails but was not watching irc...11:55
lagWhen it our release date?11:56
lagis*11:56
ndeclag: cooloney: the current l24.9 branch is quite broken, especially on panda, we are trying to fix this.11:56
lagndec: Do you have people working on the checkpatch errors too?11:56
ndeclag: unfortunately, I will be on vacation next ;-)... but sebjan will come back mid of next week.11:57
ndeclag: for checkpatch, this is a bit more complicated...11:57
ndeclag: our dev process implies that our dev teams are moving to .36, so very likely they will not addess them in a .35 release.11:58
cooloneyndec: after a quick review of the checkpatch result file, i think just several patches cause most of errors11:59
ndeclag: i will need to see what we can do with sebjan for the most critical one11:59
cooloneyndec: if we pointed them out, they can help to fix that in .3611:59
ndeccooloney: well .36 won't help since we will stay in .35. but since you have the list already, you can send that to me and sebjan and we can address them12:00
cooloneyndec: yeah, i know that. but i don't think we can fix those 2800++ coding style errors in this release.12:00
lagndec: Would it be possible to insure your developers run checkpatch before applying their work in future12:00
cooloneyjust wanna let them know this.12:00
ogra_cmpcmopdenacker, i wish too i would know that, we're hunting this bug since monday12:01
ndeccooloney: lag: we are working on this...12:01
ndeccooloney: lag: they know.12:01
lagndec: If you every want to upstream your code, it will have to become a common operating procedure12:01
ogra_cmpcmopdenacker, if you want a working install, use the alpha3 milestone image12:01
mopdenackerogra_cmpc: thanks. That's good to know that I'm not the only one...12:02
cooloneyndec: actually, i think for TI git server we can add some git post-commit script which will call checkpatch.pl to check every commit12:02
mopdenackerogra_cmpc: OK, I will do that. Thanks!12:02
ndeclag: it's part of the procedure. the main problem is time pressure and h/w problems that we need to work with12:02
lagndec: Don't we all :)12:02
cooloneyndec: technically it will drop those commit having coding issues12:03
ndeccooloney: we know all of this...12:03
lagndec: :D12:03
ndeclag: yes we do. then it's a matter of prioritization ;-)12:03
cooloneyndec: right, i understood12:03
lagndec: Indeed12:03
lagndec: When is your release date?12:04
ndeclag: at the end of the day we make h/w... and this is much less soft than software ;-)12:04
cooloneyndec: i agree the first thing we need is working stable 2.6.35 based release for both panda and blaze12:04
ndeclag: on 17th we should get an updated BSP12:04
lagSo we'll pull your tree then?12:05
ndeccooloney: they have been some updates this night in L24.9 branch. I still haven't had a chance to build, but they are supposed to have fix the panda boot problem.12:06
lagrcn-ee: ping (when you're up)12:06
lagndec: cooloney: We should wait until the release date12:06
cooloneyndec: got it, awesome. need any help, please let use know12:06
lagndec: cooloney: It's pointless merging until it's finished12:07
ndeclag: for this specific release where our dev team release is on .35, yes, you can pull from L24.9 branch. then we will create a stable branch on http://dev.omapzoom.org/?p=integration/kernel-ubuntu.git;a=summary for the updates12:07
cooloneylag: yeah,12:07
=== amitk-afk is now known as amitk
lagndec: And when's that due?12:07
cooloneyndec: indeed. that's way we worked for several monthes12:07
lagcooloney: When will your packaging be finished?12:07
ndeclag: agreed. it's pointless untill I or sebjan tells you that it's booting on panda and blaze, let us do the dirty stuff first ;-)12:08
cooloneylag: i have to wait for that release12:08
ndeclag: final L24.9 branch from dev team due on 17th12:08
lagndec: I have every intention :)12:08
* ogra_cmpc hands ndec some gloves12:08
lagndec: Noted12:08
cooloneylag: need to sync config files with the latest release.12:08
ndecogra_cmpc: for this specific item I will be on vacation ;-), so I am passing the gloves to sebjan12:09
lagogra_cmpc: Why? Do you have something stuck up there which you'd like removing?12:09
ogra_cmpcndec, haha12:09
ogra_cmpclag, no, for not getting his hands so dirty12:09
lagogra_cmpc: ;)12:09
cooloneyheh12:09
ogra_cmpclag, i'm well able to remove stuck things myself :P12:10
lagogra_cmpc: Comes with practise I suppose :)12:13
lagogra_cmpc: Where should the jasper.log file be?12:13
ogra_cmpclag, after the install run in /var/log, during initramfs in /dev/.initramfs/12:17
ogra_cmpc(jasper copies it over as the last step before the reboot)12:18
lagIt's not in /var/log12:18
ogra_cmpcthen check /dev/.initramfs/12:19
ogra_cmpc(if it still exists)12:19
lagogra_cmpc: http://paste.ubuntu.com/477403/12:20
lagDoes that look okay to you?12:20
lagogra_cmpc: My Panda comes to a crashing halt when the purple screen appears12:21
lagThe cursor and the background is visible but nothing else12:22
ogra_cmpchit enter after waiting a few mins12:22
lagAnd my USB mouse and keyboard no longer work12:22
ogra_cmpci suppose you are using a newer image than A3 ?12:22
lagI am using the latest daily build12:22
ogra_cmpcoem-config dies for no apparent reason since monday12:22
lagOkay12:23
lagWhat's the link to the most stable version?12:23
ogra_cmpctake A312:23
ogra_cmpchttp://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/maverick/12:24
lagI have it12:24
rcn-eelag pong. ;)12:33
lagGood morning Robert12:33
rcn-eeHi Lee, how's it going.12:34
lagAre you well?12:34
lagVery good thanks12:34
lagHave you heard anything regarding your ro patch?12:34
rcn-eei'm good, it's friday.. ;)  I need to quickly verify on my C2/C4 before i post version 3... (it'll be this afternoon as i get off at noon)12:35
lagAre you going to shift all the code to board-omap3beagle.c on v3?12:36
rcn-eeyeap i did..12:36
lagGood stuff12:36
lagHopefully they're leave you alone on this one12:36
lagWhat happens once it's been approved on linux-omap?12:36
lagDo you have to submit it upstream?12:36
rcn-eenah i'm not woried. ;) they are here for reference.. http://bazaar.launchpad.net/~beagleboard-kernel/%2Bjunk/2.6.35-devel/files/head:/patches/gpio/12:37
rcn-eeafter linux-omap, tony will probally push them for 2.6.37... (in for-next)  (the 2.6.36 merge can't last for maybe another day or two..)12:38
ograndec, can you include asac in the invite for the friday call ?12:39
lagrcn-ee: Is Tony the linux-omap maintainer?12:41
amitklag: he is12:42
amitklag: want to meet him?12:42
lagSure, why not!12:42
amitklag: come to Helsinki on Tuesday, I'm having lunch with him12:43
lagamitk: Not that I have anything to say :)12:43
lagamitk: I'll just nip over :)12:43
rcn-eeYeap Tony is the linux-omap maintainer..12:43
amitkrcn-ee: how are your stocks of XMs? :)12:44
rcn-eerumor up the grape vine... possibly next week..  but i'll probally say that again next week.. ;)12:45
lagrcn-ee: You can have mine when amitk sends me the new version :)12:46
rcn-eei'll trade you my 256 xMA P7 for your 512. ;)12:47
rcn-eeeven  thou the 512's have memory problems, they stay up longer then my 256. ;)12:47
lagrcn-ee: I'll think about it12:48
lagrcn-ee: nah12:48
=== fta_ is now known as fta
ogralag, you are not seeing Bug 605488 because we dont use swap currently, create a swapfile and add it to fstab, then see if you get the errors13:04
ubot2Launchpad bug 605488 in linux-ti-omap4 (Ubuntu Maverick) (and 1 other project) "BUG: scheduling while atomic: mmcqd/46/0x00000002 (affects: 1) (heat: 132)" [High,In progress] https://launchpad.net/bugs/60548813:04
=== fta_ is now known as fta
lagrcn-ee: Not having much luck there are we? :)13:54
rsalvetircn-ee but nice you're working on this issue, things will be much easier to merge when we get your patches at least linux-omap upstream :-)14:07
rsalvetircn-ee: some other questions I have, related with xM14:08
rcn-eelag, getting closer. ;) one little tweak at a time.. ;)14:08
rsalvetimaybe lag and ogra can tell more about it14:09
rsalvetiI wanted to understand the known issues we got with the hardware14:09
rsalvetiI have a p8 board14:09
rsalvetilag: what rev is your xm?14:10
rcn-eethat has the bad 512Mb Micron, some work, some don't...  Production boards are using Neomonix.. well until Micron figures out why they corrupt randomly.14:10
rcn-eei have a P7, it's 256Mb Micron..14:10
rsalvetircn-ee: oh, got it, thanks for the info14:10
* ogra has micron too, but rev. A 14:10
rcn-eeit has other issues, like crashing into a ball of fire when your run "sudo aptitude update"14:10
ogramy former one was P7 as well14:10
ograbut with 512M14:11
rsalvetirandom issues doesn't seems good hehe14:11
ograi think the rev. A ones are supposed to be more stable but i might be wrong14:11
rcn-eethat's weird.. according to my cheat sheet from gerald, the P8 was the first 512, but considering those are marked with a  marker. ;)14:11
ograhmm, i might misremember and it might have been an 814:12
* lag has a Micron 14:12
rsalvetilag: but what rev?14:12
rcn-eei think all out are micron right now.. the actuall production ones will be Nuemoinx..14:12
rcn-eelet see "A" changed gpio related to mSecure, USB host1 (to a header), USB reset and Sysboot changes..14:13
=== fta_ is now known as fta
rsalvetircn-ee: cool, you have the logs :-)14:15
rcn-eeyeah the early one... hasn't been updated since May thou.. the current log is on the schematic of the "A2" on beagleboard.org (it doesn't have the old Px changes)14:16
* rcn-ee runs work back in a couple hours.14:16
lagI think the USB is broken on my Panda :(14:19
rsalvetilag: why?14:21
lagIt doesn't work14:24
lagusb 1-1: new high speed USB device using musb_hdrc and address 314:24
lagusb 1-1: device descriptor read/64, error -11014:24
lagusb 1-1: device descriptor read/64, error -11014:24
lagusb 1-1: new high speed USB device using musb_hdrc and address 414:24
lagusb 1-1: device not accepting address 4, error -11014:24
lagusb 1-1: new high speed USB device using musb_hdrc and address 514:24
lagusb 1-1: device not accepting address 5, error -11014:25
laghub 1-0:1.0: unable to enumerate USB device on port 114:25
prpplaguelag: es1.0 board:?14:28
lagHow do I tell?14:28
ograblack or not ?14:28
prpplaguelag: is the pcb green or black?14:29
ograes2.0 are painted black14:29
rsalvetithe only issue I saw regarding the usb is the dma problem14:29
rsalvetimusb_hdrc.use_dma=0 could help, but not sure if you're getting the same issue14:29
ogralag, given that you didnt tell us about a new board you got this week i'd say es214:29
ograerr14:29
ograes114:29
rsalvetilag: es2 has the pandaboard logo14:29
prpplaguebecause of some hardware issues related to the the TWL6030 configuration on ES1.0 boards, not all the time when the board boots with something plugged into the MUSB does it enumerate properly14:30
ograand apart from tobin the es2.0's should all still be in delivery14:30
prpplaguetry rebooting with the MUSB unplugged and plug it in after booting14:30
* ogra wonders how much the pandas differ within the es1.0 line 14:31
ograi havent seen any issues regarding musb on my es114:32
prpplagueogra: it varies14:32
ograyeah, i was guessing so, like the XM14:32
lagGreen14:33
ogranow that was a prompt answer :P14:33
lagogra: shh14:33
lagogra: I'm doing more than one thing at once14:33
ograreally ?14:34
ograyou can ?14:34
rsalvetilol14:34
lagI think you misunderstand me - it used to work14:34
prpplaguelag: you mutli-task?14:34
* ogra always secretly suspected female genes in lag14:34
prpplaguelag: new kernel ?14:34
ograthere was a new meta today so likely a new kernel too14:34
lagprpplague: Yeah14:34
lagprpplague: I'll test the kernel separately - wait one14:35
prpplaguelag: np,14:35
ograhmm, the kernel is from the 5th though, i wonder why meta wasnt updated alongside14:35
rsalvetiI tested the newer kernel already, seems to be working fine14:35
prpplaguelag: try a simple test, just boot the boot without anything plugged into the MUSB and then plug it in after booting14:36
rsalvetiogra: cooloney forgot about it14:36
rsalvetiwhen he updated the main debs14:36
ograrsalveti, ah14:36
ograwould have been nice to have it in the alpha release though :P14:36
ograah well14:36
ogracant have everything14:37
* ogra glares at the maverick-changes ML14:37
ogracoinor-flopc++ ...14:38
ogralooks like someone just hit random keys to make up a package name14:38
* prpplague tries to decide what to do about panda audio14:39
ograprpplague, emmet (persia) will take care of it (at least he said so yesterday)14:39
prpplagueogra: ??14:39
prpplagueogra: you mean about the audio?14:40
ograthough i dont see him online today, seems he still has issues with his new appartment and internet14:40
ograprpplague, yep14:40
prpplagueogra: ahh, right, this is more for internal TI L24.x release14:40
ograthe ASoC module needs to expose the right stuff to make pulse aware of the different devices i think14:40
ograit doesnt do that atm14:40
prpplagueogra: i'm trying to decide if panda should share the sdp4430 audio file, or split it off for it's own14:40
ograah14:41
prpplagueogra: yea that is being done in the L24.9 release14:41
ogramain target needs to be to not need an asound.conf14:41
prpplaguehmm14:42
prpplagueogra: the thing is that the panda doesn't have a number of the channels that the sdp has14:42
prpplagueogra: none of the mic14:42
ograwell, then the driver needs fixing14:43
=== JamieBen1ett is now known as JamieBennett
ograasound.conf is a hack to override broken driver defaults, if the driver doesnt expose the right HW to userspace that needs fixing in the driver14:43
prpplagueogra: right, so the question is, do we share the sdp4430.c file and add a bunch of "machine_is_omap4_panda()" or do i create a seperate file14:43
ograheh, your choice i guess, whats working better for upstreamability later ?14:44
prpplagueyea that's what i'm trying to think through now14:44
ograprpplague, you could probably ask lrg as ASoC upstream buy ;)14:45
ogras/buy/guy/14:45
prpplagueogra: yea already have a ping out for him this morning14:46
ograah14:46
prpplagueogra: he's burried with the L24.9 release as well14:46
ograheh14:46
ograall of TI is it seems14:46
* ogra is only sitting with a bucket on his knees waiting to catch the stuff that comes downstream :)14:47
prpplaguehehe14:47
loolwhat a nice metaphor14:48
loolIt could have been worse14:48
prpplaguehehe14:49
ogralool, you mean it could be a cup ?14:52
loolit could be in another place than on your knees14:54
=== zyga is now known as zyga-lunch
=== JaMa is now known as JaMa|Away
=== zyga-lunch is now known as zyga
mopdenackerogra: I've tried the alpha3 installer and it works fine so far...15:48
ogragreat15:48
mopdenackerogra: The only bad thing is that the first boot phase is completely silent for a few minutes (while it extends the partition). I had to add "console=ttyO2,115200n8" to the bootargs to actually see something.15:49
mopdenackerThis could confuse users, thinking that their system is frozen. And they could reset their board and corrupt their rootfs which is currently being resized.15:50
ograwell, the image is designed in a way that you dont need to use serial15:50
ograyou should attach a HDMI monitor15:51
mopdenackerogra, I understand...15:51
ograhow big is your SD ?15:51
mopdenacker8 GB15:52
ograthe resizing shouldnt take more than a few mins (1-2) nowadays15:52
mopdenackerogra, you're right. It took approximately 2 minutes.15:52
laglag> It doesn't work15:52
lag<lag> usb 1-1: new high speed USB device using musb_hdrc and address 315:52
lag<lag> usb 1-1: device descriptor read/64, error -11015:52
lag<lag> usb 1-1: device descriptor read/64, error -11015:52
lag<lag> usb 1-1: new high speed USB device using musb_hdrc and address 415:52
lag<lag> usb 1-1: device not accepting address 4, error -11015:52
lag<lag> usb 1-1: new high speed USB device using musb_hdrc and address 515:52
lag<lag> usb 1-1: device not accepting address 5, error -11015:52
ograwe fixed that right before the alpha release (i didnt remove the thxt that tells you "10min per 4G yet)15:52
lag<lag> hub 1-0:1.0: unable to enumerate USB device on port 115:52
ogra*text15:53
mopdenackerI did attach an hdmi display, but nothing got displayed during the first boot phase (before the reboot to the installer)15:53
ograintresting, but it did work on second boot ?15:54
mopdenackerogra: Yes, it did.15:54
ograhmm, weird15:54
ograrobclark, ^^^ any idea why the first boot wouldnt work on HDMI but the second does ?15:55
ogra(we dont change any graphics related boot options or anything)15:55
robclarkogra: hmm.. is the first boot over too quickly before the monitor can sync?15:55
robclarkand is there any way you could get a bootlog w/ omapdss.debug=1?15:56
ograno, it takes a few mins15:56
lagSomeone's USB broken also?15:56
robclarkhmm.. and same kernel?15:56
ograand it does an auto-reboot so capturing the dmesg is a bit of an effort15:57
robclarkoh, ok..15:57
ograits the alpha3 release, works on my monitor for first and all subsequent boots15:57
robclarkwell, I can't think of any particular reason why it would not work on first boot, but would work on subsequent boots..   but I guess I don't know enough about what changes between first boot and second boot..15:58
robclarksame MLO/u-boot/uImage?15:58
ograyes, all the same15:58
robclarkany difference which could change timing and expose a potential race condition?15:58
ografirst boot is console only, second is using plymouth15:58
ograand i think GrueMaster saw some similar issues in the past15:59
lagndec: ping15:59
robclarkhmm.. but framebuffer should come up before you even get to userspace..15:59
ograyeah15:59
lagjayabharath: ping16:01
jayabharathlag: hi16:03
=== orbarron|OoO is now known as orbarron
jayabharathlag: did you try to plug in usb kybd after boot up..16:04
jayabharathlag: there was a problem that sometimes usb peripharl was not detected when you have it plugged in on cold boot16:04
lagjayabharath: I have16:06
prpplaguelag: can you send you me your kernel image, i'll test here16:07
prpplaguelag: and info on what kernel source you are using16:07
lagprpplague: Email address?16:07
prpplaguelag: x0132446@ti.com16:07
jayabharathprpplague: I remeber there was a TWL OTG VBUS issue.. was this a HW fix or kernel change16:09
prpplaguejayabharath: kernel fix16:09
=== JaMa|Away is now known as JaMa
lagjayabharath: prpplague: I'm going to rebuild our latest kernel16:16
lagprpplague: When I've done so, I'll send you the .deb16:16
rsalvetiogra: thanks for targeting the bugs for beta16:29
rsalvetieasier to track now16:29
ograrsalveti, well, i have to else i'll get grilled in the release meeting :)16:29
rsalvetioh, true :-)16:30
mopdenackerogra: I got disconnected from the Internet, sorry. I may have missed replies from you.16:39
ogramopdenacker, i didnt reply any further to you, i just know that someone in here (i think GrueMaster) had a similar prob before16:41
GrueMasterSorry, what was the issue?16:42
mopdenackerogra: All right... we can continue to investigate. Thanks!16:42
ograGrueMaster, black screen on first boot on the panda but working monitor after the first reboot16:42
mopdenackerGrueMaster: having no graphical output with the preinstalled images....16:42
ogra(after resizing)16:42
GrueMasterI was seeing that too yesterday, even with my additional parameters I have had to set.  Not even seeing serial console output.16:43
mopdenackerThough in both cases, I had to add hdmi driver bootargs to get the display to work.16:43
mopdenackerRight, console serial output wouldn't hurt, even if a serial console is not a requirement.16:44
GrueMasterMy panda first boot cmdline:16:44
GrueMastersetenv bootargs vram=32M mem=463M   root=/dev/mmcblk0p2 fixrtc console=ttyO2,115200 console=tty0 omapdss.debug=1 omapdss.hdmimode=0 omapdss.hdmicode=3516:44
ograit breaks the splash screen16:44
mopdenackerogra: Ah, right.16:44
mopdenackerGrueMaster: Got the same bootargs as you do (without console=tty0, which is a good idea btw, and without omapdss.debug)16:46
mopdenackerogra: Are you supposed to detect the monitor type, and then load the display drivers from the initramfs with the corresponding settings?16:46
ogramopdenacker, the kernel is supposed to detect the EDID from the monitor and set the right stuff17:05
ogra(inside the driver)17:06
ograand its kind of a myth why it does that right on second boot but not on first17:06
mopdenackerogra: I understand, thanks.17:06
mopdenackerI can investigate  this within TI, reproduce it in a simple testcase and file a bug report. Thanks again!17:07
ogramopdenacker, robclark has worked out patches that we dont have on the images yet17:08
ogra*could* be related17:08
ograthough its still a myth why it works on second boot17:08
mopdenackerogra: I will try with our own images too. Thanks!17:09
prpplaguemopdenacker: what problem are you seeing?17:11
mopdenackerprpplague: no graphical output on the first boot phase (with pre-installed images), and then a correct display after rebooting. Same kernel, same bootargs (except a root= setting with UUID).17:13
prpplaguemopdenacker: which platform? panda?17:13
ograprpplague, panda with the ubuntu image17:14
prpplagueahh ok17:14
prpplagueand is it a HDMI or DVI display?17:14
mopdenackerogra: I had to hardcode the hdmi settings in the bootargs in both cases. So, the question in my case (and in GrueMaster's) is why we don't get anything on the display during the first boot phase.17:14
mopdenackerprpplague: with HDMI17:14
ograi dont think our kernel supports the DVI port yet at all17:15
prpplagueogra: it does, but only barely17:15
prpplaguemopdenacker: do it is a hdmi display connected to the HDMI port?17:16
prpplaguebecause GrueMaster 's boot args are for a DVI display17:16
mopdenackerprpplague: no, it's a DVI-D display.17:16
GrueMasterFor the record, I am using the hdmi port with an hdmi<>dvi cable to a dvi monitor.17:16
prpplagueright17:16
mopdenackerSame here. I was told by ndec that the DVI port didn't work yet.17:17
prpplaguerunning DVI-D via the HDMI doesn't read the EDID, which is why the args have to be passed17:17
rsalvetiI believe it's reading the edid with latest robclark's patches17:18
mopdenackerprpplague: Ah, right!17:18
rsalvetior at least something17:18
lagprpplague: I've send the email17:18
lagprpplague: 14MB of .deb coming your way :)17:19
prpplaguelag: dandy, i'll have a look between L24.9 testing17:19
lagprpplague: np17:19
prpplaguersalveti: i'll check with robclark17:19
robclarkprpplague: if you use HDMI->DVI adapter, then you get EDID..  there were some issues with DVI specific parts of the EDID not being parsed, but most of that is implemented now with latest patches17:19
lagprpplague: I have to go now - ping me an email when you have some news17:19
prpplaguerobclark: ahh ok, so L24.9 should have the patches included?17:20
rsalvetiprpplague: http://gitorious.org/~robclark/pandaboard/robclarks-kernel-omap4/commits/L24.8_panda-hdmi-patches17:20
rsalvetiI tried to test but got -vsync with my monitor17:20
robclarkprpplague: not all of them..  in particular, the ones about EDID parsing aren't yet ;-)17:20
prpplaguerobclark: ahh ok, i'll see about getting that in panda branches17:21
prpplaguemopdenacker: i'll merge those edid patches in about an hour17:22
mopdenackerprpplague: awesome, thanks!17:24
mopdenackerprpplague: on dev.opamzoom.org?17:24
mopdenackers/opamzoom/omapzoom17:25
mopdenackerprpplague: should be able to test next Tuesday. Thanks!17:28
rsalvetimopdenacker: I have a kernel deb file with these patches at http://rsalveti.net/pub/ubuntu/kernel/maverick/17:28
rsalvetiget 903.7 rsalveti217:28
rsalvetiif you want to test now17:29
mopdenackerrsalveti: thanks! Too late for today, really gotta go...17:29
rsalvetinp :)17:29
lagprpplague: Are you still there?17:44
lagprpplague: Here is the image:17:47
laghttp://people.canonical.com/~ljones/usb-stopped-panda/17:47
lagprpplague: Email didn't work17:47
=== JaMa is now known as JaMa|GoNe
prpplaguelag: pulling now19:23
* rsalveti needs coffee19:25
ndeclag: still around?19:26
prpplaguendec: hey, the panda board hdmi fixes have not been pull from display-next as of yet19:30
ndecprpplague: hi19:31
prpplaguendec: btw, this is Dave Anders19:31
ndecprpplague: so this is normal if it does not boot?19:31
ndecprpplague: i know you who are ;-)19:31
prpplaguendec: sorry just wanted to make sure19:31
prpplaguendec: yea19:31
ndecprpplague: which patch do I need to pull? i'd like to make a quick test before my vacations ;-)19:32
prpplaguendec: one sec19:32
prpplaguehttp://dev.omapzoom.org/?p=axelcx/kernel-display.git;a=commit;h=ff6125dde94754d36dcb12dbb4e1d8e35726bcbd19:32
prpplaguendec: plus you need to set the vram to atleast 1619:32
ndecprpplague: just this?19:32
=== fta_ is now known as fta
prpplaguendec: that should be all that is require to make it bootable19:32
prpplaguendec: you may have to turn off power management in the config, there is one patch (not panda specific) that needs to be pulled19:33
ndecprpplague: ok. have you enabled usb host?19:33
prpplaguendec: yea, the EHCI and MUSB have been tested19:33
prpplaguendec: i'll retest everything this afternoon19:34
prpplaguendec: the audio patches have been submitted and should be pulled soon as well19:34
ndecprpplague: thanks. it booted once... no display, no ethernet... i will try this again after my break... good luck;-)19:47
prpplaguendec: i'll keep testing19:48
prpplaguendec: np, there are still some things to be pulled from the other trees19:49
ndecprpplague: ok. thx19:50
=== fta_ is now known as fta
=== sbambrough is now known as sbambrough-afk
MartynArm L2 cache -- generally write-back, or write-through?20:33
=== fta_ is now known as fta
rsavoyeif anybody is also building icedtea, please try this patch. I'm also doing more testing of it21:27
rsavoyehttp://www.senecass.com/projects/OpenJDK-ARM/thumb2-081310.patch21:27
loolrsavoye: I met Tom Marble at Debconf, he was interested in your work21:30
rsavoyeyeah, I'm donating some old HPPA HW to Tom for OpenJDK testing21:32
rsavoyelool: it's been interesting getting my head back into assembler programming21:33
rsavoyeplus Ed doesn't need comments in his asm code :-)21:33
=== fta_ is now known as fta
=== fta_ is now known as fta
=== fta_ is now known as fta

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!