[00:00:22] <RobotGuy> I'm setting up a Beagle dev envronment now. So far, the beginner's instructions are very good.
[00:05:42] * _altered (n=Beagle7@sceapdsd43-72.989studios.com) Quit ()
[00:07:01] <jkridner|work> glad to hear that RobotGuy
[00:07:49] <jkridner|work> ds2: well, it was used for that demonstration.
[00:08:10] <jkridner|work> your demonstration is still visible!
[00:08:58] <jkridner|work> it isn't koen's fault. someone on the DLP team decided to give LinuxDevices a few days of exclusivity on the video. From the way they state in the article, I don't think they wanted exclusivity. :)
[00:10:31] <jkridner|work> From http://www.linuxdevices.com/news/NS2883080813.html: "The complete video is available as an exclusive download from LinuxDevices, here (Ogg Theora format -- It's worth noting that the projector's output looked crisper in the video's original MP4 format, but we had to transcode it to save on monthly bandwidth charges!)."
[00:10:57] <jkridner|work> so, the exclusive in that phrase made it where koen had to take his copy down for a few days.
[00:13:06] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[00:13:11] * guillaum1 (n=Guillaum@AMontsouris-153-1-22-195.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
[00:13:16] <Vegar> HDMI with I??C?
[00:13:41] * guillaum1 (n=gl@AMontsouris-153-1-22-195.w86-212.abo.wanadoo.fr) has joined #beagle
[00:13:42] <jkridner|work> HDMI normally has I2C for EDID.
[00:14:00] * _bernie (n=bernie@trinity.develer.com) Quit (Remote closed the connection)
[00:14:05] <jkridner|work> I think they use the I2C connection that is normally intended for EDID for some additional projector controls.
[00:14:27] <jkridner|work> goodnight all
[00:14:30] <Vegar> ah
[00:14:46] * jkridner|work (n=a0321898@nat/ti/x-e1e79f0764160f38) Quit ("Leaving.")
[00:14:47] <Vegar> night
[00:23:00] <RobotGuy> I am getting: ERROR: Please set the 'PERSISTENT_DIR' or 'CACHE' variable.
[00:23:09] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Connection timed out)
[00:23:09] <RobotGuy> What should these be set to in bitbake.conf?
[00:23:24] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
[00:23:48] <denix0> RobotGuy: you don't edit bitbake.conf, you need to create your local.conf
[00:24:41] <RobotGuy> Oh yeah, right, but what should those variables be set to in local.conf?
[00:25:33] <denix0> you don't need to set them. it means your configuration is not correct
[00:27:30] <RobotGuy> I followed the instructions on creating local.conf. I'll pastebin what I have.
[00:27:48] * flo_lap (n=fuchs@g228197174.adsl.alicedsl.de) Quit (Remote closed the connection)
[00:28:18] <RobotGuy> http://www.pastebin.ca/1260948
[00:30:02] <denix0> what's your BBPATH?
[00:30:41] <RobotGuy> BBPATH is /robotguy/ARM9/Beagle/openembedded
[00:30:53] <RobotGuy> Hmm, that is not right.
[00:31:48] <denix0> and where is your local.conf?
[00:32:57] <denix0> your BBPATH needs to point to openembedded and to conf/local.conf, if it's not in openembedded
[00:33:03] <RobotGuy> BBPATH is: /stuff/robotguy/ARM9/Beagle/beagleboard/:/stuff/robotguy/ARM9/Beagle/beagleboard/beagleboard:/stuff/robotguy/ARM9/Beagle/openembedded
[00:34:18] <RobotGuy> local.conf is where the instructions say it should be: $OE_HOME/beagleboard/beagleboard/conf
[00:35:27] <denix0> I guess OE_HOME is /stuff/robotguy/ARM9/Beagle
[00:35:33] <RobotGuy> Yes
[00:36:00] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[00:36:44] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 104 (Connection reset by peer))
[00:36:56] <RobotGuy> Do I need to have $OE_HOME/beagleboard/beagleboard/conf in BBPATH?
[00:39:08] <denix0> nope
[00:39:41] <ds2> wonder how slow would a point cloud to triangle mesh conversion run on the beagle ;)
[00:40:21] <RobotGuy> I still get "ERROR: Please set the 'PERSISTENT_DIR' or 'CACHE' variable."
[00:42:00] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[00:42:00] <denix0> btw, you can't have $OE_HOME in local.conf, you need to use ${OE_HOME} instead
[00:42:16] * JoeyBorn (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) has joined #beagle
[00:43:10] <RobotGuy> Fixed, but that does not show in the instructions.
[00:43:35] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 104 (Connection reset by peer))
[00:43:46] <denix0> read openembedded/conf/local.conf.sample
[00:44:18] <denix0> so, do you still have issues with 'PERSISTENT_DIR' and 'CACHE'?
[00:45:37] <RobotGuy> Yes
[00:46:14] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[00:46:21] <denix0> what's your bitbake version? where did you get it from?
[00:46:32] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 104 (Connection reset by peer))
[00:47:38] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit ()
[00:48:19] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[00:48:48] <RobotGuy> I can see I am going to have the same kinds of problems with Beagle that I've had with Hammer.
[00:48:53] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 104 (Connection reset by peer))
[00:49:17] <RobotGuy> denix0: I followed the instructions at http://elinux.org/BeagleBoardAndOpenEmbeddedGit
[00:49:25] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[00:49:48] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 104 (Connection reset by peer))
[00:52:29] <ds2> RobotGuy: OE is a super hard version of buildroot }:-)
[00:54:50] * JuanG (n=Juan@nat/ti/x-ca60a9a691cd0430) has joined #beagle
[00:54:56] * JuanG (n=Juan@nat/ti/x-ca60a9a691cd0430) has left #beagle
[00:57:43] <denix0> RobotGuy: something is still wrong with your BBPATH
[01:00:53] * harry-plenware (n=harry@221.10.25.6) has joined #beagle
[01:01:45] <RobotGuy> If I put exactly what is in the instructions into my profile.sh, it is even worse.
[01:02:48] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[01:02:56] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 104 (Connection reset by peer))
[01:09:44] <denix0> RobotGuy: what do you mean worse?
[01:10:54] <RobotGuy> denix0: Those "\" mess things up majorly, at least in most places.
[01:13:25] <Crofton|irssi> ok, demo-iamge built for me again
[01:15:13] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[01:16:47] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 104 (Connection reset by peer))
[01:19:21] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
[01:21:30] <RobotGuy> What is BBPATH supposed to be??
[01:27:18] * harry-plenware (n=harry@221.10.25.6) Quit (Remote closed the connection)
[01:28:39] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit (Read error: 60 (Operation timed out))
[01:28:44] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) has left #Beagle
[01:28:54] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
[01:30:53] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
[01:43:26] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Connection timed out)
[01:43:41] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
[01:49:18] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) has joined #Beagle
[01:50:24] * hli (i=chaton@vig91-2-82-232-97-149.fbx.proxad.net) Quit (Read error: 54 (Connection reset by peer))
[01:53:31] <RobotGuy> I need some help to fix whatever is stopping things working for me on this Angstrom build for Beagle.
[01:53:49] <RobotGuy> I don't know what is wrong with my BBPATH.
[02:00:14] * harry-plenware (n=harry@221.10.25.6) has joined #beagle
[02:05:55] <denix0> RobotGuy: paste your local.conf and whatever profile.sh you have
[02:07:48] <RobotGuy> http://www.pastebin.ca/1261036
[02:10:24] <denix0> profile?
[02:10:35] <RobotGuy> Oh wait, that was my local.conf, sorry. Hang on.
[02:12:01] <RobotGuy> profile.sh: http://www.pastebin.ca/1261042
[02:17:03] <denix0> also, please paste: find $OE_HOME -maxdepth 3
[02:20:25] * RyoKimball (n=Ryo_Kimb@75-120-202-251.dyn.centurytel.net) has joined #beagle
[02:23:52] <RobotGuy> http://www.pastebin.ca/1261051
[02:24:13] <tomasg> Question on USB OTG... I know this has been asked a few times. I have the stock linux-omap2_git 2.6.26-r64 kernel built up which has OTG support. My musb/mode is b_idle, unless I boot with it connected to a host in which case it goes to b_peripheral. I cannot switch it to host or otg mode.
[02:24:31] <denix0> RobotGuy: your OE_HOME is wrong
[02:25:14] <RobotGuy> How so?
[02:25:55] <denix0> your OE_HOME should be /stuff/robotguy/ARM9/Beagle
[02:26:25] <tomasg> I didn't pick up an "OTG" cable per se, which seems like this is the trouble. I do have a connector from Frys that goes from a 5-pin miniB to USB A female. Will this work w/ a cable when plugged into a powered hub? Do I need to mod my own cable/connector?
[02:27:01] <denix0> RobotGuy: echo $OE_HOME
[02:27:07] <RobotGuy> denix0: That is what my OE_HOME is.
[02:27:41] <denix0> then why is it showing listing of bitbake directory?
[02:28:01] <RobotGuy> I don't know what you mean.
[02:28:14] <denix0> <denix0> RobotGuy: echo $OE_HOME
[02:28:36] <RobotGuy> /stuff/robotguy/ARM9/Beagle
[02:28:44] <denix0> also, it's recommended to execute bitbake from your OE_HOME
[02:28:55] <RobotGuy> I know that and that is what I am doing.
[02:29:02] <denix0> then your pastebin is wrong - http://www.pastebin.ca/1261051
[02:29:28] <denix0> did you do - find $OE_HOME -maxdepth 3?
[02:29:48] <RobotGuy> That is what I pasted.
[02:31:02] <denix0> it should have beagleboard and openembedded directories, as well as opt/bitbake, but it only shows content of opt/bitbake
[02:31:06] * n2o_2k (n=n2o_2k@204.153.78.227) has joined #beagle
[02:31:31] <RobotGuy> You are right - my pastebin was wrong.
[02:32:05] <RobotGuy> I don't know how to paste all the info that find $OE_HOME -maxdepth 3 produces.
[02:33:10] <Crofton> OE_HOME, you'll need to add that to the list of env cars bitbake passes
[02:33:34] <RobotGuy> Env cars?
[02:33:37] <denix0> try export BB_PRESERVE_ENV=1
[02:33:45] <denix0> s/cars/vars
[02:34:25] <denix0> Crofton: then it would have complained about missing ${OE_HOME}...
[02:35:45] <RobotGuy> Something is happening now. :) NOTE: Handling BitBake files:
[02:36:31] <denix0> hmm, that'
[02:37:28] <denix0> Crofton: that's strange - if I don't preserve vars, I get absolutely different error message... :(
[02:37:40] <Crofton> sorry, I'm in a hotel and catchign up on a bunch of crap
[02:40:33] * harry-plenware (n=harry@221.10.25.6) Quit (Remote closed the connection)
[02:41:29] <RobotGuy> denix0: It's doing a whole bunch of stuff now, so this is a good sign. :)
[02:44:38] <Crofton> http://wiki.openembedded.net/index.php/AdvancedConfiguration
[02:44:48] <Crofton> some more info at the bottom of this page.
[02:45:18] <denix0> this env filter makes more problems than it solves...
[02:45:34] * emeb_mac (n=ericb@ip72-223-95-46.ph.ph.cox.net) has joined #beagle
[02:45:45] <Crofton> heh
[02:46:08] <Crofton> apparently there are one or two packages that fail mysteriously if you have certain vars set
[02:46:21] <Crofton> the goal is to have deterministic builds
[02:46:30] <Crofton> this is a lot harder than it would seem
[02:47:18] * emeb_mac (n=ericb@ip72-223-95-46.ph.ph.cox.net) Quit (Client Quit)
[02:47:57] <denix0> $ unset BB_PRESERVE_ENV
[02:48:05] <denix0> $ export OEBASE=$HOME/oe
[02:48:11] <denix0> $ bitbake nano
[02:48:30] <denix0> ERROR: opening ${OEBASE}/openembedded/packages/*/*.bb: file not found
[02:50:01] * RyoKimball1 (n=Ryo_Kimb@75-120-202-251.dyn.centurytel.net) has joined #beagle
[02:50:31] <denix0> at least I'm getting better error message :) instead of ERROR: Please set the 'PERSISTENT_DIR' or 'CACHE' variable.
[02:53:56] * harry-plenware (n=harry@221.10.25.6) has joined #beagle
[02:55:01] * n2o_2k (n=n2o_2k@204.153.78.227) Quit ("leaving")
[02:55:49] <jkridner> has this already floated around: http://lizards.opensuse.org/2008/11/18/arm-support-for-opensuse-buildservice-and-opensuse/ ?
[02:58:17] * robtow (n=rtow@nat/montavista/x-29cff3d44e6e8d2e) Quit ("Leaving.")
[03:05:51] * RyoKimball (n=Ryo_Kimb@75-120-202-251.dyn.centurytel.net) Quit (Read error: 110 (Connection timed out))
[03:05:59] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) Quit (Read error: 104 (Connection reset by peer))
[03:07:31] * JAF (n=chatzill@129.108.41.175) has joined #beagle
[03:08:05] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
[03:11:06] <RobotGuy> mpfr.org seems to be offline, and has been since at least yesterday.
[03:17:25] <RobotGuy> Is there a way to tell OE to look in an alternative place for stuff it can not find?
[03:17:47] <denix0> it does
[03:19:43] <RobotGuy> It keeps trying mpfr.org. which does not seem to have what it wants (mpfr-2.3.1).
[03:20:51] <denix0> did it timeout?
[03:21:07] <RobotGuy> Yes, several times.
[03:21:57] <RobotGuy> I'm seeing: HTTP request sent, awaiting response... Read error (Connection reset by peer) in headers.
[03:22:04] * tulpe (n=ask@p54AF78A9.dip.t-dialin.net) Quit (Read error: 145 (Connection timed out))
[03:22:24] <RobotGuy> It's on try #4
[03:22:29] * tulpe (n=ask@p54AF7D96.dip.t-dialin.net) has joined #beagle
[03:23:18] <Crofton> it may grab from the angstrom source mirror after a few failuers, try letting it succeed or error
[03:25:03] <RobotGuy> It's just erroring so far, and I would prefer it did not just keep trying like this.
[03:25:12] <RobotGuy> It increases build times.
[03:25:43] <denix0> once it downloads in your DL_DIR, it won't try it again next time
[03:26:21] <RobotGuy> denix0: I understand that, but it has to successfully download first.
[03:26:35] <denix0> it will try mirrors eventually
[03:26:55] * mturquette (n=mturquet@CC0F872D.ptr.provps.com) has joined #beagle
[03:26:56] <RobotGuy> denix0: How many unsuccessful tries does it take, and can I control this?
[03:27:36] <RobotGuy> 3 tries max should be more than enough to make it move on to mirrors.
[03:28:10] <denix0> if it can't find a file (error 404), it tries mirrors next
[03:28:33] <denix0> not sure how many tries it does when it fails to connect
[03:29:30] <denix0> aha, wget -t 5
[03:29:40] <RobotGuy> denix0: Well, more than three tries is more than I want to wait.
[03:30:13] <denix0> FETCHCMD_wget = "/usr/bin/env wget -t 5"
[03:31:20] <denix0> that's in bitbake.conf
[03:32:11] <denix0> and FETCHCOMMAND_wget = "/usr/bin/env wget -t 5 --passive-ftp -P ${DL_DIR} ${URI}"
[03:32:43] <denix0> change -t 5 to set number of tries
[03:33:00] <RobotGuy> So I put it in my local.conf then, right?
[03:33:07] <denix0> you can add those 2 into your local.conf
[03:33:08] <denix0> right
[03:33:56] <denix0> and for completeness, RESUMECOMMAND_wget = "/usr/bin/env wget -c -t 5 --passive-ftp -P ${DL_DIR} ${URI}"
[03:34:26] <RobotGuy> denix0: OK, got it, thanks. I appreciate your help! :)
[03:34:57] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit ()
[03:35:47] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
[03:36:38] <RobotGuy> Maybe I can come up to speed with Beagle faster than I was able to do with Hammer.
[03:36:40] * emeb (n=ericb@ip72-223-95-46.ph.ph.cox.net) has left #beagle
[03:38:57] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has left #beagle
[03:51:41] <jkridner> I just found http://elinux.org/U-boot_environment_variables_in_linux, but it isn't working for me with the latest u-boot on the denx.de u-boot-arm/omap3 tree.
[03:51:51] <jkridner> I was getting header problems, so I tried to copy over the latest Linux headers, but that created problems.
[03:52:02] <jkridner> I then tried setting MTD_VERSION=old and that has problems too.
[03:52:34] <jkridner> I wish these wiki page entries would do a bit more to preserve the version info.
[03:52:56] <jkridner> u-boot-git-r18 might be enough info, but I don't know.
[03:57:41] <sakoman_> jkridner: it isn't enough info
[03:57:51] <jkridner> k.
[03:58:00] <jkridner> have you tried 'make env' in u-boot?
[03:58:02] <sakoman_> I tried to build those tools last week and ran into the same missing header issue
[03:58:15] <jkridner> k. putting it in from Linux doesn't help.
[03:58:16] <sakoman_> What are you trying to accomplish?
[03:58:52] <jkridner> I'm working on a clean OE build of a demo system that enables me to setup Beagle Boards for demos.
[03:59:08] <sakoman_> What sort of setup do you need?
[03:59:48] <jkridner> well, I'm looking for something very flexible. I'm producing a RAMDISK that can be booted directly from SD, without any reading of environment variables on the NAND flahs.
[04:00:22] <jkridner> I've done that already at http://beagleboard.org/demo/esc...
[04:00:31] <jkridner> but, I'm looking to do it more cleanly.
[04:01:25] <sakoman_> You might want to do something like I've done for the default overo environment
[04:01:41] <jkridner> I want to produce an SD card image that people can recreate easily in Windows and will not depend on the contents of the flash.
[04:01:44] <jkridner> dd images?
[04:02:20] <sakoman_> With that setup uboot first looks on microSD for a u-boot script called boot.script. If it is there it runs it
[04:02:46] <jkridner> I'm considering also making the default bootargs read a kernel from the ext3 partition to avoid people using mismatched kernels (as you can see happens on the mailing list).
[04:02:50] <sakoman_> If it isn't there, then it looks for uImage on the microSD, if present it loads it and boots
[04:03:01] <sakoman_> if neither is there then it does a nand boot
[04:03:36] <jkridner> sounds good. my u-boot hack looks for u-boot.img for a script on the SD card, but it doesn't fail gracefully...
[04:03:41] <sakoman_> This approach gives you almost infinite flexibility since you can put any uboot commands you want in the boot.scrip
[04:03:58] <jkridner> u-boot normally hangs on 'fatload' if there is no card in the socket.
[04:04:31] <jkridner> I've already got that capability (u-boot scripts), but I'm using the autoscr for it and it requires the script to be pre-processed.
[04:04:34] <sakoman_> I had to do some u-boot patches to fix mmc detection and improve scriptability
[04:04:58] <jkridner> I'd really like to have some basic expression and branching as well.
[04:05:05] <sakoman_> Yes, that is still required for the boot.script, but is pretty simple to do
[04:05:08] <jkridner> I'm discovering that u-boot is pretty weak.
[04:05:21] <jkridner> are you using mkimage?
[04:05:32] <sakoman_> I have a pre-built version of u-boot for beagle with that setup if you want to try it
[04:05:47] <sakoman_> Yes, I use mkimage for boot.script
[04:05:55] <jkridner> I'd like to look at the patches first.
[04:06:06] <sakoman_> It is used by the gumstix production folks for factory flashing and test
[04:06:16] <jkridner> also, I like to avoid any non-8.3 file names for FAT formatted cards, just to avoid any funniness.
[04:06:55] <sakoman_> pre-built images are here: http://www.sakoman.net/feeds/omap3/glibc/images/
[04:07:37] <jkridner> I am wondering if I got the default to NAND flash wrong (button for SD) in looking for predictability in boot-loading.
[04:08:01] <sakoman_> the u-boot branch my changes are in is: http://www.sakoman.net/cgi-bin/gitweb.cgi?p=u-boot-omap3.git;a=shortlog;h=refs/heads/omap3
[04:08:43] <sakoman_> It's a working repo, so you see all the steps along the way, not the clean end patches
[04:09:48] <sakoman_> jkridner: yeah, I made Gordon default to microSD
[04:09:55] <jkridner> yeah, it doesn't look like you have rebased on u-boot-arm/omap3.
[04:10:21] <jkridner> what experience told you that was the right solution?
[04:10:59] <sakoman_> I will -- when I was doing this work last week u-boot-arm/omap3 wouldn't boot
[04:12:01] <sakoman_> jkridner: well, mostly gut feel :-)
[04:12:22] <jkridner> k. I might have overanalyzed it.
[04:12:43] <sakoman_> It just seemed that if a person stuck a bootable microSD in the slot, it ought to boot from it
[04:12:56] <sakoman_> It seemed to be the rational deault behaviour
[04:12:59] <sakoman_> default
[04:13:52] <jkridner> I was worried that depending on the card format would mean that you could get surprises.
[04:14:36] <sakoman_> That's why I make the dd images available, sort of a fail safe for folks who can't get the formatting right
[04:14:45] <jkridner> I wanted the behavior such that if you had programmed the NAND flash, it would always boot from the NAND flash unless you did something special.
[04:15:05] <sakoman_> Most gumstix customers are linux folks though, so there isn't the big Windows issue
[04:15:10] <jkridner> but 'dd' is still a real pain. I think we need another executable.
[04:15:21] <jkridner> ah, well, if you already know dd, then it is easy.
[04:15:54] <jkridner> one problem I've been looking at is creating a 'dd' image from within OE.
[04:16:06] <jkridner> I can partition a file, but I don't know how to format the partitions within the file.
[04:16:10] <sakoman_> The audience for Gumstix is embedded developers, most of whom are completely comfortable with dd
[04:16:44] <jkridner> k. I've always thought of Beagle supporting more than just Linux.
[04:16:49] <RobotGuy> Wouldn't that be the same audience for Beagle too?
[04:16:50] <ds2> sakoman_: is the MMC detect on the TWL4030 on the Overo?
[04:17:11] <RobotGuy> But Beagle may have an expanded audience too.
[04:17:19] <sakoman_> ds2: yes -- the standard connection just like beagle/evm
[04:17:29] <jkridner> RobotGuy: to some extent, desktop developers are being targeted by Beagle as a cross-over to embedded.
[04:17:43] <RobotGuy> Desktop as i Windows?
[04:17:44] <ds2> sakoman_: how do you talk to the twl4030 in u-boot then? you wrote a polling I2C driver?
[04:18:03] <jkridner> RobotGuy: yes, and Linux desktop people as well.
[04:18:25] <sakoman_> ds2: I enabled i2c in u-boot
[04:18:34] <denix0> jkridner: Linux desktop people would know how to use dd
[04:18:41] <sakoman_> but I just rely on the error return from mmcinit
[04:18:41] <ds2> u-boot has a working polling driver for OMAP3?
[04:18:41] * RyoKimball1 (n=Ryo_Kimb@75-120-202-251.dyn.centurytel.net) Quit (Read error: 110 (Connection timed out))
[04:18:47] <RobotGuy> I've been doing embedded, but not Desktop development.
[04:18:56] <jkridner> denix0: I did answer yes to the Windows question though.
[04:19:02] <denix0> :)
[04:19:07] <sakoman_> ds2: yes, there are even commands to access i2c devices from the u-boot command line
[04:19:10] <sakoman_> try it!
[04:19:24] <ds2> didn't think it worked on omap
[04:19:29] <sakoman_> yup
[04:19:36] <ds2> good to know
[04:20:43] <sakoman_> jkridner: I think people trying to develop for beagle or overo on Windows will have more than their fair share of grief
[04:21:01] <jkridner> sakoman_: I've looked at your patches and they aren't exactly what I'd do, but they are much better than what I've done so far. :)
[04:21:45] <jkridner> looks like I can build off of what you've done. I still think I need 'make env' to work to have a good image that people can use interactively to configure their systems.
[04:21:54] <sakoman_> What would you change? I'm always up for suggestions for a better way
[04:22:33] <jkridner> I'll also likely add some sort of ncurses UI once running Linux to select a lot of standard operations.
[04:22:53] * jkridner wishes Beagle had two SD card slots now.
[04:23:12] <vlad_> USB! :)
[04:23:39] * sakoman_ is looking forward to an overo buddy board with a second mmc/microSD slot
[04:23:59] * emeb_mac (n=ericb@ip72-223-95-46.ph.ph.cox.net) has joined #beagle
[04:24:25] <ds2> but it does
[04:25:37] <ds2> just a pcb fab + some soldering
[04:25:56] <jkridner> sakoman_: just silly things like not using a non-8.3 file name, giving some video mode options for non-mmc boot, ...
[04:25:59] <sakoman_> ds2: true
[04:26:12] <jkridner> the more I look at it, however, it seems to do everything I need.
[04:26:47] <jkridner> I was going to suggest reverting the 600MHz default, but you already did that. :)
[04:27:03] <sakoman_> it was just an experiment
[04:27:07] <jkridner> better to let Linux power management raise the clock speed.
[04:27:24] <jkridner> well, I must call it a night.
[04:27:27] <sakoman_> wanted to do some performance tests to see what difference it mafe
[04:27:29] <sakoman_> made
[04:27:31] <jkridner> you saved me a lot of time!
[04:27:39] <ds2> is enuff of pm in there on the main branch to do it?
[04:27:40] <sakoman_> no problem!
[04:27:47] <JAF> request time?
[04:27:49] <jkridner> wish I saw this before.
[04:27:50] <JAF> I would like access to the camera port :/
[04:28:07] <jkridner> let me know if anyone gets 'make env' to work.
[04:28:15] <jkridner> JAF: it is on the Overo.
[04:28:28] <sakoman_> jkridner: I'll look at it again when I get a moment
[04:28:39] <jkridner> JAF: I think that is a perfectly good request for Rev. D and one I'm thinking about more.
[04:29:03] <jkridner> JAF: It will be easier to escape the camera signals without adding layers on the 0.65mm package.
[04:29:13] <jkridner> good night all.
[04:29:23] <sakoman_> JAF: but it is available on the overo now if you really need it
[04:29:24] <ds2> JAF: where would you get the cameras?
[04:29:30] <JAF> thanks :)
[04:29:40] <JAF> I am using a CMOS chip from micron
[04:30:07] <JAF> but I bought it before they sold that division
[04:30:17] <denix0> gn
[04:30:20] <JAF> so I wouldn't know where to get them now
[04:31:29] <JAF> The overo has the camera pins on those really small connectors right?
[04:31:49] <ds2> hirose, IIRC
[04:31:56] <JAF> I am working on a very limited budget, student budget actually
[04:32:40] * JoeyBorn (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) Quit (Read error: 110 (Connection timed out))
[04:32:54] * harry-plenware (n=harry@221.10.25.6) Quit (Remote closed the connection)
[04:33:21] <sakoman_> JAF: camera signals are on a ribbon cable connector
[04:34:10] <sakoman_> so you would need a ribon cable and a matching ribbon cable connector on your camera pcb
[04:37:50] <JAF> hm I see
[04:38:18] * harry-plenware (n=harry@221.10.25.6) has joined #beagle
[04:38:40] <JAF> thanks, I'll try to get one
[04:39:38] * JAF (n=chatzill@129.108.41.175) has left #beagle
[04:39:53] * likewise (n=Leon_Woe@atwork-193.r-212.178.107.atwork.nl) has joined #beagle
[04:53:18] * khasim (n=a0393720@192.163.20.231) has joined #beagle
[05:02:00] * emeb_mac (n=ericb@ip72-223-95-46.ph.ph.cox.net) Quit ()
[05:06:46] * soman (n=somnath@61.16.248.242) has joined #beagle
[05:07:18] * jsync (n=jess@59.160.172.220) has joined #beagle
[05:25:12] * DaQatz (n=db@c-66-30-48-54.hsd1.nh.comcast.net) Quit ("Gone")
[05:33:36] * aanishn (n=anish@59.160.172.220) has joined #beagle
[05:37:34] * emeb_mac (n=ericb@72.223.95.46) has joined #beagle
[05:38:24] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
[05:40:28] * emeb_mac (n=ericb@72.223.95.46) Quit (Client Quit)
[05:57:30] * emeb_mac (n=ericb@ip72-223-95-46.ph.ph.cox.net) has joined #beagle
[06:02:15] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[06:04:23] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
[06:13:37] * JoeyBorn (n=jborn@adsl-75-2-240-22.dsl.chcgil.sbcglobal.net) has joined #beagle
[06:14:21] * emeb_mac (n=ericb@ip72-223-95-46.ph.ph.cox.net) Quit ()
[06:20:58] <ds2> the math for a 3D scanner is easy but interpreting a point cloud by inspection of X,Y,Z is a major pain in the @#@%#######
[06:23:14] * harry-plenware (n=harry@221.10.25.6) Quit (Read error: 104 (Connection reset by peer))
[06:27:29] * harry-plenware (n=harry@221.10.25.6) has joined #beagle
[06:42:02] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
[07:06:24] * shriram (i=3d5fc3cd@gateway/web/ajax/mibbit.com/x-fde64ae9d69e78e7) Quit ("http://www.mibbit.com ajax IRC Client")
[07:08:43] * rupeshgujare (n=Administ@59.160.172.220) has joined #beagle
[07:12:10] * oyvindje_ is now known as oyvindje
[07:12:53] * gduncan (n=spec@207-255-14-177-dhcp.jst.pa.atlanticbb.net) has joined #beagle
[07:26:12] * codepope (n=codepope@ip-217.146.111.161.merula.net) has joined #beagle
[07:35:43] * abitos (n=nixgibts@dslb-084-057-154-032.pools.arcor-ip.net) has joined #beagle
[07:55:59] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) Quit (Remote closed the connection)
[07:58:33] * pmeerw (n=pmeerw@hippogriff.cosy.sbg.ac.at) Quit (Read error: 104 (Connection reset by peer))
[07:59:45] * harry-plenware (n=harry@221.10.25.6) Quit (Remote closed the connection)
[08:00:02] * Beagle6 (n=Beagle6@208-100-140-231.bendbroadband.com) has joined #beagle
[08:00:05] * Beagle6 is now known as _altered
[08:14:54] * abitos (n=nixgibts@dslb-084-057-154-032.pools.arcor-ip.net) Quit (Read error: 113 (No route to host))
[08:19:52] * geckosenator (n=sean@c-24-8-193-190.hsd1.co.comcast.net) Quit (Read error: 110 (Connection timed out))
[08:27:04] <AV500> the beagles have landed!
[08:27:25] <AV500> just got 2 of them, ordered them 2 days ago, nice :-)
[08:29:52] <geist> congrats
[08:31:00] <AV500> pardon my ignorance, but where would I just download a "nice" SD-image to test?
[08:35:05] <AV500> the 10pin serial, it is "logically" or "physically" wired?
[08:37:12] <geist> it's documented on the wiki somewhere
[08:37:30] <geist> it turns out to be a semi standard for those 10 pin connectors on a motherboard to a db9 or db25
[08:37:53] <geist> i found an ancient one in my pile of cables for an ancient PC era 10pin->db25 case
[08:38:19] <AV500> yes, but even there are two "standard", one is to wire pin1 to pin1, 2 to 2 etc..
[08:38:27] <geist> it's documented on the wiki
[08:38:29] <geist> and no, it isn't
[08:38:52] <AV500> the other is to layout the pins so that a 9/10pin crimp header can fit a flat cable on
[08:39:31] <AV500> I guess it will be this one...
[08:39:35] <geist> it's documented on the wiki
[08:39:44] <AV500> ok ok, I get you :-)
[08:44:15] <AV500> according to BB.ORG it is wired "logically" 2-2, 3-3 5-5, like the OMAP3 ZOOM actually
[08:44:23] <geist> oh cool
[08:46:55] <AV500> ah, scratch that, the ZOOM is different.
[08:46:57] <ds2> itis documented in the SRM
[08:47:06] <ds2> Zoom/SDP cable will NOT work
[08:47:11] <AV500> yes, I am looking at that right now.
[08:47:23] * codepope (n=codepope@ip-217.146.111.161.merula.net) Quit ()
[08:47:39] <ds2> AT/Everex is the pinout used on the BB
[08:48:40] <AV500> and I'm fresh out of 10pin crimp headers... :-(
[08:48:59] <ds2> raid a PC
[08:49:12] <ds2> or WW 3 wires to a DB-9
[08:49:25] * and-ri (n=ubuntu@pd95b4498.dip0.t-ipconnect.de) has joined #beagle
[08:50:08] <AV500> yes, yes
[08:52:24] * codepope (n=codepope@ip-217.146.111.161.merula.net) has joined #beagle
[08:52:40] <AV500> or just use this cable here right before me
[08:53:00] <ds2> =)
[08:53:20] <AV500> OMAP3 beagleboard.org #
[08:53:23] <AV500> :-)
[08:55:22] <koen> now, get yourself a non broken u-boot :)
[08:55:44] <koen> I'm using http://www.angstrom-distribution.org/demo/beagleboard/u-boot.bin
[08:56:03] <koen> that fixes the l1neon errate in cortex r1p0-3
[08:58:17] <AV500> the good this is, I left my HDMI->DVI at home this morning... :-(
[09:14:53] * BBdev (n=Beagle5@bzq-219-125-195.static.bezeqint.net) Quit ()
[09:20:14] * methril|gone is now known as methril
[09:28:13] <mru> koen: is the l1neon patch in OE u-boot builds?
[09:28:27] <koen> it's in sakomans tree where OE builds from
[09:28:47] <mru> ok, it's also in mainline
[09:28:51] <mru> omap3 branch
[09:33:26] <koen> great
[09:34:35] <mru> the curious thing is that setting fixes the ffmpeg crashes on ES2.1 even without the dmb hack
[09:38:15] <koen> heh
[09:39:17] <mru> the other bug is probably much more rare than it seemed
[09:40:19] * Beagle3 (n=Beagle3@s-inf-pc117.oulu.fi) has joined #beagle
[09:40:20] * Beagle3 (n=Beagle3@s-inf-pc117.oulu.fi) Quit (Client Quit)
[09:40:24] * Beagle4 (n=Beagle4@s-inf-pc117.oulu.fi) has joined #beagle
[09:58:27] <ldesnogu> mru: or it's perhaps a real r1p2 Cortex-A8 in there...
[10:14:06] * florian (n=fuchs@217.146.132.69) has joined #beagle
[10:22:33] <koen> tomba: is there an ioctl or something similar to enable/disable planes?
[10:22:57] <koen> tomba: I've been thinking on how the XV driver can be changed to work with dss2
[10:23:01] * Beagle4 (n=Beagle4@s-inf-pc117.oulu.fi) Quit ()
[10:23:04] * Beagle4 (n=Beagle4@s-inf-pc117.oulu.fi) has joined #beagle
[10:36:14] * Beagle5 (n=Beagle5@213.139.154.7) Quit ()
[10:48:24] <koen> likewise: you probably have a stray " somewhere
[10:48:51] <koen> likewise: or a TMP_DIR with a trailing /
[10:56:09] * guillaum1 (n=gl@AMontsouris-153-1-22-195.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
[11:18:54] <likewise> koen: that last suggestion might be truth related. But it should be fully safe against 1 or more ///. Where would this fail?
[11:23:07] <koen> it fails somewhere in packaged-staging
[11:23:26] <koen> RP knows the details, he suggested I removed trailing / from TMPDIR
[11:24:49] * koen struggles with uboot
[11:25:27] <koen> I wish I could make the uboot recipes in OE a bit cleaner
[11:25:40] <koen> but uboot upstream is such a mess
[11:34:44] * Robot_kolej (n=robot@robot.kolej.mff.cuni.cz) has joined #beagle
[11:34:47] <Robot_kolej> hello
[11:35:22] <Robot_kolej> did someone try to build the gcc 2007q3 package in OE ? (without a fail)
[11:39:50] <likewise> koen: hmm, no trailing /, looking for strays "
[11:43:25] <koen> Robot_kolej: I even succeeded
[11:45:26] <Robot_kolej> koen: what are the prerequisites? , i have tried it several times with no luck
[11:46:16] <koen> a working OE installation are the only prereqs I know off
[11:46:28] <koen> a beagleboard build will build 2007q3 by default for the kernel
[11:46:58] <Robot_kolej> and how do i correctly tell bitbake to build the exact version of gcc ?
[11:49:36] * shenki (n=joel@202.174.42.5) has joined #beagle
[11:50:05] <Robot_kolej> i know that 2007q3 is built for kernel, but only c compiler, i need also the c++ compiler
[11:51:53] <koen> 'need'?
[11:51:57] <koen> 2007q3 is POS compiler
[11:52:09] <koen> is it a real need or did you buy into the marketing BS?
[11:54:52] <Robot_kolej> i thought that the 2007q3 does better optimalization for the cpu ...
[11:55:35] <koen> I haven't seen any benchmarks so far
[11:55:48] <koen> just handwaving or 5 line c programs
[12:00:50] <Robot_kolej> so you suggest to use the "normal" gcc, instead of the csl ?
[12:00:57] <jkridner> good morning all
[12:01:22] <Robot_kolej> and what about the neon instruction set, does the "normal" gcc support these ?
[12:04:26] <likewise> gm
[12:05:20] <koen> Robot_kolej: I suggest using the default OE picks unless you are sure you know better
[12:05:34] <koen> Robot_kolej: and yes, vanilla gcc supports neon
[12:05:36] <koen> hey likewise
[12:06:50] <Robot_kolej> koen: ok , thanks for the info, i will try both, and see what the difference is
[12:32:57] <tomba> koen: yes, there are ioctl's to change the planes. it's the same one as in the current DSS. (SETUP_PLANE or something)
[12:33:51] <tomba> koen: but if the XV driver works with current DSS, it should work with DSS2 also.
[12:36:50] <koen> with the new dss it doesn't seem to clear (or disable, not sure which) after the movie has finished
[12:37:12] <koen> the last frame stays visible, even when its parent window has disappeared
[12:37:38] <koen> and I suspect that the XV driver needs to set fb1 to nonstd8 when initializing the plane
[12:38:50] <tomba> it should do that also with the current DSS... I think the current DSS initializes all planes to rgb 565 by default. DSS2 uses rgb565 or rgb888 by default
[12:39:14] <tomba> do you have the giturl to that xv project again? I could check it out
[12:39:40] <tomba> and I have to say that I haven't tested the plane setup ioctls almost at all
[12:40:41] * shrirama (i=3d5fc3cd@gateway/web/ajax/mibbit.com/x-cc816c07e9761c6d) has joined #beagle
[12:47:05] * cbrake_away is now known as cbrake
[12:50:52] <tomba> koen: looking at http://gitweb.pingu.fi/?p=xf86-video-omapfb.git;a=blob;f=src/omapfb-driver.c;h=aba5b19a141a1c39701177ecbabd73aea1580a16;hb=e33601fe4c4c5e0d093321f86a633047056c7445 I can only see the code turning the overlay on, never off
[12:54:31] * soman (n=somnath@61.16.248.242) Quit (Remote closed the connection)
[13:03:09] * Crofton (n=balister@12.234.232.66) Quit (Read error: 110 (Connection timed out))
[13:06:55] * uwe__ is now known as uwe_
[13:09:47] * robclark (n=robclark@nat/ti/x-6d2238ac7eed23be) has joined #beagle
[13:11:01] * rsalveti (n=salveti@189.70.53.187) Quit (Remote closed the connection)
[13:11:16] <koen> tomba: the 'window' does dissappear with dss1, but you'd have to ask zuh of kulve about the details
[13:12:42] <kulve> umm
[13:12:58] <kulve> at least the video disappears after the playback is finished :)
[13:13:03] <kulve> maybe zuh can say something more :)
[13:13:05] * _altered (n=Beagle6@208-100-140-231.bendbroadband.com) Quit ()
[13:14:25] <koen> kulve: with dss2 scaling and fullscreen work, but the last frame stays on top
[13:14:43] <kulve> sounds like it's not turned of then..
[13:30:46] <koen> right
[13:31:20] <koen> so I wonder why it works with dss1
[13:32:36] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
[13:34:36] * Crofton (n=balister@12.234.232.66) has joined #beagle
[13:34:54] * koen tests a dsplink that hooks into Kbuild
[13:37:24] <koen> http://gitweb.pingu.fi/?p=xf86-video-omapfb.git;a=blob;f=src/omapfb-xv.c;h=cf598edefb8b7962ecdc6aceb148af06a326adaa;hb=e33601fe4c4c5e0d093321f86a633047056c7445#l209
[13:38:02] <koen> that's the setup_plane
[13:38:04] <koen> and this:
[13:38:05] <koen> http://gitweb.pingu.fi/?p=xf86-video-omapfb.git;a=blob;f=src/omapfb-xv.c;h=cf598edefb8b7962ecdc6aceb148af06a326adaa;hb=e33601fe4c4c5e0d093321f86a633047056c7445#l330
[13:38:18] <koen> I think that's where it should turn off the plane
[13:39:02] <koen> I suspect that with dss1 "close fb" means "turn plane off" and with dss2 it does nothing
[13:39:10] * koen aint a coder
[13:39:14] <Crofton> stop hotel wireless
[13:39:23] <Crofton> s/stupid/stop/
[13:44:03] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
[13:44:37] <koen> Crofton: ccache works for me
[13:44:37] * Beagle4 (n=Beagle4@s-inf-pc117.oulu.fi) Quit ()
[13:46:27] * fagius (n=fagius@static-71-111-254-119.rlghnc.dsl-w.verizon.net) has joined #beagle
[13:50:17] * gduncan (n=spec@207-255-14-177-dhcp.jst.pa.atlanticbb.net) Quit ()
[13:53:55] <koen> so dsplink and cmem hook into kbuild
[13:54:00] <koen> that only leaves lpm
[13:54:25] <Crofton> it has wrked for me, until I tried it on rhel4
[13:54:37] <Crofton> running the offending command by hand also worked
[13:54:50] <Crofton> I'll add a note to the mockumentation
[13:55:06] * Robot_kolej (n=robot@robot.kolej.mff.cuni.cz) has left #beagle
[14:00:09] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
[14:01:12] * JuanG (n=Juan@nat/ti/x-bffeec106ec435a8) has joined #beagle
[14:01:21] * JuanG (n=Juan@nat/ti/x-bffeec106ec435a8) has left #beagle
[14:19:47] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
[14:20:12] * prpplague^2 (n=dave@mail.americanmicrosystems.com) has joined #beagle
[14:20:47] * prpplague^2 is now known as prpplague
[14:21:14] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
[14:22:22] * guillaum1 (n=gzba4143@86.212.21.195) has joined #beagle
[14:23:22] * Openfree (n=df@222.65.143.118) has joined #beagle
[14:28:05] * tth (i=tth@spider.lemtek.fi) Quit ("leaving")
[14:32:17] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
[14:36:22] <sakoman_> good morning
[14:36:47] * codepope (n=codepope@ip-217.146.111.161.merula.net) Quit ()
[14:37:26] <prpplague> sakoman_: greetings
[14:37:53] <koen> hey sakoman_
[14:38:31] * _bernie (n=bernie@trinity.develer.com) Quit (Remote closed the connection)
[14:39:36] <sakoman_> koen: I fixed the gimp build on my machines last night. not sure if my methods are worthy of upstream merge, but here you go: http://www.sakoman.net/cgi-bin/gitweb.cgi?p=openembedded.git;a=commit;h=19dfc71bffe473d2c1142ea64ea36a791305c2a3
[14:40:20] <sakoman_> previously gimp would only build if your build machine had compatible headers in /usr/include
[14:41:54] * BThompson (n=BThompso@nat/ti/x-8393ff4bca7bf2cc) has joined #beagle
[14:43:35] <sakoman_> koen: BTW, thanks for moving UBOOT_MACHINE to the proper place in the machine.conf files. been meaning to talk to you about that!
[14:45:43] * bazbell (n=a0192809@nat/ti/x-38c5d41a9bf8d24e) has joined #beagle
[14:46:19] * surki (n=surki@gek7.kyla.fi) has joined #beagle
[14:50:39] * john280z (n=jmitchel@static-71-111-254-119.rlghnc.dsl-w.verizon.net) has joined #beagle
[14:51:43] * emeb (n=ericb@ip72-223-95-46.ph.ph.cox.net) has joined #beagle
[14:58:55] <zuh> koen, tomba: The plane enable/disable happens in the PutImage/StopVideo callbacks, for the generic case here: http://gitweb.pingu.fi/?p=xf86-video-omapfb.git;a=blob;f=src/omapfb-xv-generic.c;hb=HEAD#l186 and here: http://gitweb.pingu.fi/?p=xf86-video-omapfb.git;a=blob;f=src/omapfb-xv-generic.c;hb=HEAD#l295
[15:00:36] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
[15:07:21] * rupeshgujare (n=Administ@59.160.172.220) has left #beagle
[15:08:34] * JoeyBorn (n=jborn@adsl-75-2-240-22.dsl.chcgil.sbcglobal.net) Quit ("open.neurostechnology.com")
[15:12:05] * JoeyBorn (n=jborn@adsl-75-2-240-22.dsl.chcgil.sbcglobal.net) has joined #beagle
[15:13:57] * JoeyBorn (n=jborn@adsl-75-2-240-22.dsl.chcgil.sbcglobal.net) Quit (Client Quit)
[15:14:17] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[15:14:30] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 54 (Connection reset by peer))
[15:15:14] * and-ri (n=ubuntu@pd95b4498.dip0.t-ipconnect.de) Quit (Read error: 110 (Connection timed out))
[15:22:02] <koen> sakoman_: ah, it seems gimp change its makefiles
[15:22:10] * shrirama (i=3d5fc3cd@gateway/web/ajax/mibbit.com/x-cc816c07e9761c6d) Quit ("http://www.mibbit.com ajax IRC Client")
[15:22:45] <koen> sakoman_: as you can see from the sedding it filtered out ${includedir} already, but it seems they hardcoded .usr/include now :(
[15:23:00] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
[15:23:08] * _bernie (n=bernie@trinity.develer.com) Quit (Read error: 104 (Connection reset by peer))
[15:24:03] * aanishn (n=anish@59.160.172.220) has left #beagle
[15:27:18] <sakoman_> koen: it's a conspiracy!
[15:28:09] <sakoman_> all packages seem to keep trying to slip in /usr/include dependencies
[15:29:10] <koen> yeah
[15:38:25] <koen> it seems to be the silver bullet for idiots writing makefiles
[15:39:25] <koen> I had one of my students insisting that -L/usr/lib -lbluetooth was the proper way to do things
[15:40:27] <sakoman_> koen: I see you've been playing with nautilus. How does it look on Beagle?
[15:40:37] <koen> well
[15:40:45] <koen> you saw the commits that came after it
[15:40:57] <koen> haven't tried it yet, I tried to get it to install without errors
[15:41:47] <sakoman_> I need to update my cron job to sync more frequently with oe.dev
[15:42:33] <sakoman_> Right now it doesn't sync during the overnight hours
[15:43:15] <sakoman_> I hate all the "Merge branch" noise in gitweb
[15:46:38] <koen> git pull --rebase should get rid of that noise
[15:46:44] <koen> with a price of course
[15:47:02] <koen> e-wm doesn't like nautilus
[15:47:21] <sakoman_> heh, had a feeling that might be the case
[15:47:45] <koen> sakoman_: http://scap.linuxtogo.org/files/6e9f9a00fea7b92eaf8e2138ea06324b.png
[15:48:26] <sakoman_> Cool!
[15:48:40] <koen> the mime problem is sadly still present
[15:49:01] <sakoman_> what was the e-wm conflict?
[15:49:28] <koen> e-wm doesn't like that nautilus wants to claim the 'desktop', so it puts it in a window
[15:49:48] <koen> and nautilus is missing mime-info
[15:49:53] <koen> well 'missing'
[15:50:06] <koen> it clashes with the content from shared-mime-info
[15:50:26] <koen> someone that understands /usr/share/mime should have a look at it
[15:51:40] <sakoman_> koen: can nautilus replace e-wm?
[15:52:25] <thomasg> how should a filemanager replace a windowmanager?
[15:52:54] <koen> sakoman_: I suspect one can configure e-wm to give nautilus the WM_WORKAREA
[15:53:05] <koen> sakoman_: or add metacity to OE :)
[15:53:24] <sakoman_> thomasg: you assume I knew that nautilus was just a file manager :-)
[15:53:26] <thomasg> or just don't use that gnome-stuff :)
[15:53:57] <koen> people are getting hot and bothered about ubuntu/arm, which is just 'gnome-stuff' :)
[15:54:00] <thomasg> sakoman_: indeed, I did :)
[15:54:14] <sakoman_> thomasg: my ignorance is boundless!
[15:55:08] <thomasg> I don't blame you for this. once you've seen nautlius you will regret to have ever tried it :)
[15:56:09] <thomasg> koen: do you know why this problem exists? guess nautilus handles the desktop-/hal-icons, right?
[15:56:30] * vijay (n=vijay@203.199.213.3) has joined #beagle
[15:56:53] <koen> e-wm tries to be a desktop-shell and claims the desktop (I forget the proper NETWM term for it)
[15:57:19] <koen> e17 has its own filemanager
[15:57:22] <thomasg> yeah
[15:57:29] * methril is now known as methril|gone
[15:57:43] <thomasg> I recently did some patches for it, I know that :)
[15:57:44] <sakoman_> koen: said filemanager doesn't seem to work at all
[15:58:09] <thomasg> what problems do you have with it?
[15:58:09] <koen> sakoman_: it does after the recent updates
[15:58:18] <tomasg> Question on USB OTG... I know this has been asked a few times. I have the stock linux-omap2_git 2.6.26-r64 kernel built up which has OTG support. My musb/mode is b_idle, unless I boot with it connected to a host in which case it goes to b_peripheral. I cannot switch it to host or otg mode.
[15:58:23] <tomasg> I didn't pick up an "OTG" cable per se, which seems like this is the trouble. I do have a connector from Frys that goes from a 5-pin miniB to USB A female. Will this work w/ a cable when plugged into a powered hub? Do I need to mod my own cable/connector?
[15:58:45] <koen> sakoman_: http://scap.linuxtogo.org/files/3bd73f18962c09c9c6191591d89f31c2.png
[15:58:55] <thomasg> I find it quite usable right now - though - I still use terminal only because filemanaging without a proper shell sucks
[15:59:03] <koen> sakoman_: the icon topleft is the fat partition of the sd card
[15:59:26] <koen> sakoman_: evm with 2.6.28rc4 + DSS2
[16:00:02] <sakoman_> koen: just trying a fresh image right now, so I ought to see the same (fingers crossed)
[16:00:46] <sakoman_> been mucking about in low level stuff and factory tools for a few weeks, so I haven't done upstream merges in a while
[16:00:51] * NishanthMenon (n=nMenon@nat/ti/x-daa192b7e3ddf4aa) has joined #beagle
[16:00:55] * NishanthMenon (n=nMenon@nat/ti/x-daa192b7e3ddf4aa) Quit (Client Quit)
[16:01:07] <koen> I uploaded http://www.angstrom-distribution.org/demo/beagleboard/Angstrom-Beagleboard-demo-image-glibc-ipk-2008.1-test-20081119-beagleboard.rootfs.tar.bz2 today
[16:01:12] * koen hopes it works
[16:01:50] * jsync (n=jess@59.160.172.220) has left #beagle
[16:06:15] * Openfree (n=df@222.65.143.118) Quit (Remote closed the connection)
[16:12:56] <sakoman_> tomasg: just noticed your question. you definitely need an OTG cable. a mini-B to A will not work
[16:16:24] <Crofton|irssi> demo-image built again
[16:19:52] * koen turns his attention to making jambalaya
[16:27:15] * cringthis (n=cringthi@nat/ti/x-a080d61000a44504) has joined #beagle
[16:27:52] <zuh> It's funny how some words instantly start a mental playback of certain musical pieces...
[16:29:04] <Crofton|irssi> koen: I think bevan first attept was as root
[16:29:22] <Crofton|irssi> then he ran as a normal user and got permissions on /dev/null
[16:30:40] * robtow (n=rtow@64.62.142.114) has joined #beagle
[16:30:49] <Crofton|irssi> running bitbake as root changed perms on /dev/null?
[16:32:26] * robtow (n=rtow@64.62.142.114) Quit (Client Quit)
[16:32:43] * robtow (n=rtow@64.62.142.114) has joined #beagle
[16:40:36] <tomasg> sakoman_: Thanks! Does anyone have a breakdown on how to mod a standard cable/connector?
[16:42:34] <tomasg> http://www.internettablettalk.com/forums/showthread.php?t=20288 look good
[16:45:36] * jkridner|work (n=a0321898@nat/ti/x-8960058d9f1130c5) has joined #beagle
[16:46:28] * Stasi (n=stancho@78.90.99.168) has joined #beagle
[16:56:51] * _bernie (n=bernie@204.9.220.36) has joined #beagle
[17:02:12] * [X]Spot (n=stancho@78.90.99.168) Quit (Connection timed out)
[17:15:17] * _bernie (n=bernie@204.9.220.36) Quit (Read error: 110 (Connection timed out))
[17:26:25] * DaQatz (n=db@c-66-30-48-54.hsd1.nh.comcast.net) has joined #beagle
[17:29:06] <sakoman_> koen: hmmm . . . the demo image now boots into a screen asking me to select a language from an empty list and a profile from an empty list
[17:29:43] * zalassi (n=zalassi@port2.cjas.org) Quit (Remote closed the connection)
[17:29:43] <sakoman_> and then hangs at the profile screen because there is no valid response
[17:29:55] * zalassi (n=zalassi@port2.cjas.org) has joined #beagle
[17:29:55] <sakoman_> have you seen this?
[17:32:12] * robtow (n=rtow@64.62.142.114) Quit ("Leaving.")
[17:34:18] * dannyBlue|DDd (n=dannyBlu@213.63.151.69) has joined #beagle
[17:34:55] * dannyBlue|DDd (n=dannyBlu@213.63.151.69) Quit (Client Quit)
[17:35:27] * dirk2 (n=dirk@92.117.73.130) has joined #beagle
[17:35:51] * guillaum1 (n=gzba4143@86.212.21.195) Quit ("Leaving.")
[17:36:43] * guillaum1 (n=Guillaum@AMontsouris-153-1-22-195.w86-212.abo.wanadoo.fr) has joined #beagle
[17:39:48] <zuh> koen: I hid the mode setting in xf86-video-omapfb behind a ifdef, so you shouldn't have troubles from it anymore. (I naturally tested it by compiling so ymmv ;)
[17:46:39] * _bernie (n=bernie@24.147.221.165) has joined #beagle
[17:47:37] <koen> zuh: :)
[17:47:47] <koen> sakoman_: ah, that must be the new enlightenment stuff :(
[17:48:13] <koen> sakoman_: do you have IMAGE_LINGUAS = "" in your image file?
[17:51:56] <koen> sakoman_: could you post to oe-devel about the e profile thing?
[17:52:11] <koen> sakoman_: I suspect it's a leftover from the e-wm overhaul by the openmoko folks
[17:53:51] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) has joined #Beagle
[17:54:07] <sakoman_> koen: no IMAGE_LINGUAS in my image file. I'll try it
[17:54:29] <koen> IMAGE_LINGUAS = "" would remove every linguaE
[17:54:30] <RobotGuy> My build finished, but I didn't get anything in $OE_HOME/tmp/deploy/glibc/images/beagleboard
[17:54:35] <koen> which is probably not what you want
[17:54:51] <RobotGuy> In fact, there is no such path.
[17:56:56] <sakoman_> koen: looks like image.bbclass does a IMAGE_LINGUAS ?= "de-de fr-fr en-gb"
[17:57:10] <RobotGuy> I got this error at the start, which I didn't see at first: ERROR: '[]' RDEPENDS/RRECOMMENDS or otherwise requires the runtime entity 'openzaurus-sa' but it wasn't found in any PACKAGE or RPROVIDES variables
[17:57:22] * cringthis (n=cringthi@nat/ti/x-a080d61000a44504) has left #beagle
[17:57:41] * valhalla (n=valhalla@81-174-21-31.dynamic.ngi.it) has joined #beagle
[17:57:53] <koen> sakoman_: right, it tries to please OE developers
[17:57:54] <denix> RobotGuy: that message is harmless
[17:58:05] <koen> sakoman_: who are in .de, .fr and .uk :)
[17:58:08] <sakoman_> koen: ah, but then angstrom does IMAGE_LINGUAS = '${@base_less_or_equal("ROOT_FLASH_SIZE", "16", "", "en-gb", d)}'
[17:58:18] <koen> right
[17:58:35] <RobotGuy> OK
[17:58:35] <koen> if root flash size isn't set you get all 3
[17:58:57] <koen> or do I read that wrong?
[17:59:01] <sakoman_> koen: hmm . . . not none?
[17:59:13] <sakoman_> I think it would select ""
[17:59:18] <koen> hmmm
[17:59:25] <koen> someone screwed up I guess
[17:59:26] <sakoman_> I'm confused
[17:59:52] <sakoman_> The symptoms would indicate ""
[18:00:04] <RobotGuy> Wat would cause me not to get any build output?
[18:00:41] <ds2> n
[18:01:00] <ZeZu> what is the real need for a fat partition on the SD card ? limitation of uboot ?
[18:01:17] <RobotGuy> It could not fetch the codesourcery cross toolchain.
[18:01:25] <sakoman_> ZeZu: limitation of omap3 rom bootloader
[18:01:42] <RobotGuy> How can I fix this?
[18:01:53] <ZeZu> sakoman_ that is x-loader ?
[18:02:11] <sakoman_> no, it is code in rom on the chip that loads x-load
[18:02:22] <koen> sakoman_: I have 'glibc-localedata-en-gb_2.6.1-r15.1_armv7a.ipk' in my beagle demo image
[18:02:25] <sakoman_> it only knows fat
[18:02:42] <Crofton|irssi> RobotGuy: can you pastebin the output of bitbake?
[18:03:05] <RobotGuy> Yes, hang on.
[18:03:29] <ZeZu> sakoman_, so if it loads x-loader/uboot from nand this can be bypassed ?
[18:03:44] <sakoman_> ZeZu: yes
[18:04:42] <sakoman_> koen: http://pastebin.com/m558470df
[18:05:10] <sakoman_> looks like I have en-gb and en-us
[18:05:20] <sakoman_> yet still no language list
[18:05:32] <sakoman_> so it must be some e-wm thing
[18:05:34] <koen> sounds indeed like a bug in e-wm
[18:05:38] * koen curses e-wm
[18:05:53] <sakoman_> two steps forward, one back
[18:05:58] <RobotGuy> Crofton|irssi: http://www.pastebin.ca/1261535
[18:06:20] * FlipFlop (n=flipflop@146.244.179.229) has joined #beagle
[18:06:47] * dirk2 (n=dirk@92.117.73.130) has left #beagle
[18:08:19] <Crofton|irssi> The URL works with wget from the command line
[18:08:28] * Fl1pFl0p (n=flipflop@ip68-8-230-62.sd.sd.cox.net) Quit (Nick collision from services.)
[18:08:38] * FlipFlop is now known as Fl1pFl0p
[18:08:43] * FlipFl0p (n=flipflop@ip68-8-230-62.sd.sd.cox.net) has joined #beagle
[18:09:01] <RobotGuy> Hmmmmm.
[18:09:44] <RobotGuy> Yes, it sure does. Why did it not work within the build?
[18:09:59] <koen> sh: -c: line 0: unexpected EOF while looking for matching `"'
[18:10:06] <koen> sounds like you have a typo in local.conf
[18:10:21] <Crofton|irssi> koen: good point
[18:10:37] <Crofton|irssi> RobotGuy: go ahead and pastebin local.conf
[18:11:59] <RobotGuy> http://www.pastebin.ca/1261545
[18:12:39] <koen> line 7
[18:13:38] <RobotGuy> I don't see the problem.
[18:13:54] <Crofton|irssi> me neither
[18:14:08] <Crofton|irssi> what does RESUME and FETCH COMMANND do?
[18:14:22] <sakoman_> closing quote
[18:14:25] <RobotGuy> Should I remove the "-c" ?
[18:14:29] <Crofton|irssi> some weird quoting thing?
[18:14:45] <koen> seriously
[18:14:47] <koen> check line 7
[18:14:56] <sakoman_> closing quote is missing
[18:15:13] <Crofton|irssi> crap
[18:15:33] <RobotGuy> Got it and fixed, thanks.
[18:15:50] * geckosenator (n=sean@c-24-8-193-190.hsd1.co.comcast.net) has joined #beagle
[18:16:54] <RobotGuy> I must have missed the closing quote when I copied the line.
[18:17:49] * florian (n=fuchs@217.146.132.69) Quit (Remote closed the connection)
[18:21:00] <RobotGuy> Fetching the cross toolchain still fails.
[18:21:41] <Crofton|irssi> I can't count
[18:21:42] <Crofton|irssi>
[18:21:51] <Crofton|irssi> doh, back
[18:22:01] <Crofton|irssi> RobotGuy: pastebin the message
[18:22:20] * FlipFlop (n=flipflop@146.244.179.229) has joined #beagle
[18:22:25] <RobotGuy> It's the same as the first paste showing my bitbake outout.
[18:22:42] <RobotGuy> I can repost though if you want.
[18:23:52] * _bernie (n=bernie@24.147.221.165) Quit (Read error: 60 (Operation timed out))
[18:26:15] <denix> Crofton|irssi: FETCH and RESUME commands override the standard ones, as he wants to reduce the number of tries from 5 to 2
[18:27:02] <denix> RobotGuy: actually on line 7 you got the command truncated
[18:27:20] <denix> RobotGuy: you are missing ${URI} at the end
[18:27:56] * Crofton|irssi (n=balister@66-207-66-26.black.dmt.ntelos.net) Quit (Remote closed the connection)
[18:28:18] <denix> and don't remove "-c" from RESUME - that's exactly to tell wget to resume, i.e. continue aborted download
[18:28:27] <ZeZu> when i do: sudo tar xvf ./path/to/console-image-beagleboard.tar.bz2 -C /mnt << i get a bunch of errors in making links to busybox for various console apps
[18:28:48] <ZeZu> anyone know why this might happen?
[18:29:07] <ZeZu> they are Operation not permitted errors
[18:29:12] <denix> ZeZu: what's the filesystem type?
[18:29:15] <ZeZu> haha
[18:29:25] <ZeZu> that is the problem i think i need to wakeup
[18:29:33] <ZeZu> trying to copy them over to the fat fs
[18:30:19] * felipec (n=felipec@a91-153-251-222.elisa-laajakaista.fi) has joined #beagle
[18:30:37] <koen> I think you are missing a j in the extract command as well
[18:31:03] <ZeZu> true
[18:31:36] * Fl1pFl0p (n=flipflop@146.244.179.229) Quit (Nick collision from services.)
[18:31:39] * FlipFlop is now known as Fl1pFl0p
[18:31:41] <ZeZu> thankfully tar is pretty good about it normally, but good practice, was partially pasting from a wiki page
[18:31:59] * guillaum11 (n=Guillaum@AMontsouris-153-1-67-44.w90-2.abo.wanadoo.fr) has joined #beagle
[18:32:26] <denix> RobotGuy: did it work?
[18:33:50] * Crofton|irssi (n=balister@66-207-66-26.black.dmt.ntelos.net) has joined #beagle
[18:34:31] <ZeZu> has anyone done any benchmarks against the arm compilers ? (vs code sourcery)
[18:35:03] * mru has done some
[18:35:08] <mru> arm wins
[18:35:19] <koen> stuff crashes when compiled with csl compilers
[18:35:25] <koen> unless you patch them
[18:35:37] <ZeZu> even the 2007q3 ones ?
[18:35:41] <mru> csl wins over vanilla gnu
[18:35:48] <ZeZu> mru, i figured as much, how badly does it win (arm) ?
[18:36:10] * guillaum1 (n=Guillaum@AMontsouris-153-1-22-195.w86-212.abo.wanadoo.fr) Quit (Read error: 60 (Operation timed out))
[18:36:10] <mru> I haven't had anything crash with 2007q3 without -Os or with the updated 2008q3
[18:36:19] <ZeZu> If its minor i wont pay for it
[18:36:33] <mru> no need to pay
[18:36:40] <ZeZu> does the 2008q3 still have vectorization issues ? ( i think that was problem ? )
[18:36:46] <ZeZu> arm compilers free now ?
[18:37:06] <mru> the gcc bit is always free, you just can't download it from csl
[18:37:11] <mru> want a copy of the latest?
[18:37:27] <ZeZu> sorry i mean the arm sdt/e
[18:37:31] <ds2> koen: is there a specific way of triggering a crash due to a bad compiler?
[18:37:54] <koen> http://gitweb.openembedded.net/?p=openembedded.git;a=blob_plain;f=packages/gcc/gcc-csl-arm-2008q3/gcc-csl-2008q3-hotfix.diff;hb=051f59ecc15130fb88187c4c6b602801cbb40c79
[18:38:07] <koen> ds2: compile a complete rootfs, boot it
[18:38:13] <RobotGuy> denix: It is working now. :)
[18:38:21] <ds2> koen: that's a userland crash then?
[18:38:22] * Beagle7 (n=Beagle7@smtp.alconsystems.com) has joined #beagle
[18:38:22] * Beagle7 is now known as BeagleDeveloper
[18:38:26] <mru> koen: and what crashes?
[18:38:38] <mru> I build everything with 2007q3
[18:38:58] <ds2> I have built kernels with bad CS compilers and they worked fine with a known good userland
[18:39:40] * dirk2 (n=dirk@92.117.212.213) has joined #beagle
[18:40:15] * Crofton|irssi (n=balister@66-207-66-26.black.dmt.ntelos.net) Quit (Remote closed the connection)
[18:41:05] <BeagleDeveloper> Guys, I can't find "rmmod" in the google rootfs, any ideas how can I remove module?
[18:41:28] <mru> hex editor on /dev/mem ;-)
[18:41:51] <BeagleDeveloper> :-)
[18:42:07] <BeagleDeveloper> the only way is to recompile the kernel?
[18:42:33] <dirk2> sakoman_, mru, koen: U-Boot v1 upstream: j24 applied the last remaining bugfix patch ~60min ago to u-boot-arm/omap3
[18:43:02] <dirk2> just tried it, it seems that u-boot-arm/omap3 is working now on Beagle without any further patches. If you like, please test
[18:43:07] <koen> nice
[18:43:38] <dirk2> the remaining ~10/11 patches are clean up only (to get mainline acceptance)
[18:44:12] <koen> so it isn't in yet?
[18:44:36] <dirk2> mainline? No, and I'm not sure that it will ever...
[18:45:08] * guillaum11 is now known as guillaum1
[18:45:53] <BeagleDeveloper> mru, really, any way to remove modules?
[18:45:55] <dirk2> the patch apply rate for u-boot-arm seems to be <= 1-2 patches/week
[18:47:24] <dirk2> and once we think we fixed all (mainly style) issues, we have to extract patches from u-boot-arm/omap3 and send them again to u-boot list for (final??) review...
[18:47:29] <ZeZu> mru, you were speaking of the realview arm tools right? did you build a whole system with them ?
[18:47:51] <BeagleDeveloper> mru, forget it, my mistake...
[18:51:05] * koen stabs aclocal.m4 users
[18:51:20] <mru> ZeZu: no, only ffmpeg
[18:51:55] <ZeZu> ffmpeg w/o neon or w/ neon ? and how much of a perf. difference was there ?
[18:51:56] <mru> dirk2: if you think that's slow, come visit my dayjob
[18:52:52] * vijay (n=vijay@203.199.213.3) Quit (Read error: 110 (Connection timed out))
[18:53:16] <mru> anything non-trivial generally takes a year or more to get accepted
[18:54:33] <dirk2> mru: It's a difference if you are paid for ;)
[18:55:54] <BeagleDeveloper> guys, another linux question (sorry for that), I want to dismount the mmc and nand drivers. I actually don't want them to access the hardware. is there a linux command that can do that? I tried lsmod, but it returns nothing...
[18:56:11] <dirk2> mru: and regarding U-Boot, I think fixing all multiline comment styles is really trivial, no? ;)
[18:56:15] <BeagleDeveloper> any ideas?
[18:56:20] <geist> if it's built into the kernel you should remove it from the kernel config
[18:56:49] <BeagleDeveloper> geist, no way to do it in run time?
[18:56:49] <mru> dirk2: at work cosmetic changes are strictly forbidden
[18:56:56] <geist> i dont think so, no
[18:57:08] <mru> the code is an ever-increasing mess
[18:57:21] <geist> though unmounting the file systems should be pretty safe
[18:57:39] <sakoman_> dirk2: I will give it a try
[18:58:16] * ldesnogu_ (n=ldesnogu@ven06-2-82-247-86-183.fbx.proxad.net) has joined #beagle
[18:58:21] <BeagleDeveloper> geist, I don't have them mounted in any way (I'm using the ram-drive), but still I can see the hardware being accessed by the kernel.
[18:58:33] <geist> then you have to remove the drivers from the kernel
[18:58:53] <dirk2> sakoman_: Let me know the result (if I'm offline by mail). If there are any issues, I will try to fix them. Getting it applied then, well, ...
[18:59:08] <BeagleDeveloper> If i'll remove them, will I be able to boot from SD/MMC or NAND?
[18:59:25] <geist> if uboot does the loading, sure
[18:59:54] <BeagleDeveloper> hmm, good point!
[19:01:32] <koen> ds2: ' [..] we are trying to move all the omap driver
[19:01:32] <koen> development to the appropriate driver lists to get in sync with the
[19:01:33] <koen> mainline tree.'
[19:01:54] <koen> ds2: that's that tony said a few minutes ago
[19:02:26] * Crofton|irssi (n=balister@66-207-66-26.black.dmt.ntelos.net) has joined #beagle
[19:05:25] * Keithamus (n=keith@cpc1-port6-0-0-cust824.cos2.cable.ntl.com) has joined #beagle
[19:05:25] * Keithamus (n=keith@cpc1-port6-0-0-cust824.cos2.cable.ntl.com) Quit (Remote closed the connection)
[19:05:33] <BeagleDeveloper> geist, any idea what config file should i change?
[19:06:12] <geist> the kernel one. how are you building the kernel?
[19:09:07] <BeagleDeveloper> http://code.google.com/p/beagleboard/wiki/BeagleSoftCompile
[19:10:14] <geist> you'll probably want to hack the config file in arch/arm/configs/*beagle*
[19:10:37] <geist> I dont know what it's called, but the make omap3_beagle_defconfig copies it out to .config in the root of your kernel dir and builds with it
[19:10:50] <ds2> koen: saw that :/
[19:11:07] <geist> or you can just copy it yourself then run make menuconfig and edit out the mmc and nand bits
[19:11:13] <geist> and then make uImage
[19:11:22] <geist> that's probably what you should do
[19:12:02] <BeagleDeveloper> thanks, I'll try that...
[19:12:18] <geist> shouldn't be too hard to find where it is in the options
[19:12:43] <ds2> geist: Isn't there a sysfs thing he could use to detach the driver from the hardware?
[19:13:00] <geist> i dunno, he seems to be pretty intent on it never running
[19:13:05] <geist> but you might be right too
[19:13:43] * _bernie (n=bernie@fw-1-user-net-flrs.cictr.com) has joined #beagle
[19:14:09] * koen wonders what the point of running linux is when you aren't going to use any drivers
[19:14:51] * Crofton|irssi (n=balister@66-207-66-26.black.dmt.ntelos.net) Quit (Read error: 113 (No route to host))
[19:15:08] <BeagleDeveloper> ds2, don't know regarding sysfs, I tried lsmod an it shows nothing...
[19:15:11] <ZeZu> how does OE decide what to rebuild? if i've added an overlay for a new distro, but all of the target flags are the same, will it rebuild all of the packages ?
[19:15:31] <geist> lsmod only lists what modules are loaded
[19:15:46] <geist> if the driver is built into the kernel, there will be no module associated with it
[19:16:05] <BeagleDeveloper> o, I didn't know that
[19:16:06] <geist> and for most embedded things like this, there's no point building the drivers as a module
[19:16:10] <geist> since hte hardware is fixed
[19:16:15] <koen> ZeZu: what do you need a new distro for?
[19:16:28] * dirk2 (n=dirk@92.117.212.213) has left #beagle
[19:16:35] <BeagleDeveloper> geist, so what sysfs does?
[19:16:39] <ZeZu> koen, i'm just playing around to learn a bit more about OE really
[19:16:42] <geist> no idea, ds2 seems to know
[19:16:47] <BeagleDeveloper> lol
[19:16:55] <koen> ZeZu: starting a new distro isn't a good way to do that
[19:16:59] <ZeZu> In the end i'll probably just modify which packages angstrom uses
[19:17:02] <geist> i was not aware that you can shut a driver instance down via sysfs
[19:17:02] <ZeZu> koen, why ?
[19:17:15] <geist> i've just never needed to do something like that
[19:17:27] <ds2> there is suppose to be a path to ask the driver to detech or attach
[19:17:37] <ZeZu> koen, what if i wanted to use uclibc rather than glibc for instance ?
[19:17:44] <ds2> all the machines I have shells on at the moment are modular so I can't find the path :(
[19:18:02] <koen> ZeZu: ANSTROM_MODE = "uclibc" bitbake <foo>
[19:18:24] <BeagleDeveloper> I guess I'll stick to rebuilding the kernel
[19:18:59] * koen stabs gnutls
[19:19:04] <ds2> the unbind/bind stuff
[19:19:41] <geist> that's good to know
[19:20:46] * denix thinks koen is in the stabbing mood today...
[19:21:45] <ZeZu> koen, thanks! whats the easiest way to see which packages angstrom is using for console-image or the like and modify that ? what i'm after is uclibc/busybox/Xorg/XFCE or maybe even just TWM :)
[19:22:30] <denix> ZeZu: packages/images/console-image.bb?
[19:23:44] <ZeZu> ok thanks
[19:24:50] <koen> twm + xeyes!
[19:25:11] <ZeZu> hehe
[19:26:00] <ds2> xneko
[19:33:18] <koen> xneko isn't in OE, but oneko is :)
[19:35:03] <ZeZu> beagleboard-demo-image uses enlightenment ?
[19:35:13] * garren (n=garren@dsl-244-222-39.telkomadsl.co.za) has joined #beagle
[19:43:30] * uwe_ (n=uwe_@dslb-084-056-051-163.pools.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
[19:44:11] * hli (i=chaton@vig91-2-82-232-97-149.fbx.proxad.net) has joined #beagle
[19:45:42] * ZeZu (n=null@c-24-14-8-40.hsd1.in.comcast.net) Quit (Read error: 113 (No route to host))
[19:46:31] * ZeZu (n=null@c-24-14-8-40.hsd1.in.comcast.net) has joined #beagle
[19:51:59] * _zAlt (n=null@c-24-14-8-40.hsd1.in.comcast.net) has joined #beagle
[19:54:44] * ZeZu (n=null@c-24-14-8-40.hsd1.in.comcast.net) Quit (Read error: 60 (Operation timed out))
[19:56:03] * _zAlt is now known as ZeZu
[19:59:16] * _zAlt (n=null@c-24-14-8-40.hsd1.in.comcast.net) has joined #beagle
[20:01:26] * Crofton|irssi (n=balister@66-207-66-26.black.dmt.ntelos.net) has joined #beagle
[20:02:56] <denix> koen: what machine feeds Angstrom has for i686 arch?
[20:03:21] * ZeZu (n=null@c-24-14-8-40.hsd1.in.comcast.net) Quit (Read error: 60 (Operation timed out))
[20:04:50] * FlipFlop (n=flipflop@146.244.179.229) has joined #beagle
[20:06:06] * _zAlt is now known as ZeZu
[20:10:05] <ZeZu> OE / beagleboard target uses csl by default ?
[20:12:43] * Fl1pFl0p (n=flipflop@146.244.179.229) Quit (Read error: 60 (Operation timed out))
[20:14:14] * robtow (n=rtow@nat/montavista/x-09c40c6eed3fc5c2) has joined #beagle
[20:16:29] * valhalla (n=valhalla@81-174-21-31.dynamic.ngi.it) Quit ("Leaving")
[20:17:30] <Crofton|irssi> ZeZu: for the kernel
[20:17:49] <koen> denix: none for the 2008 feeds
[20:18:01] * uwe_ (n=uwe_@dslb-088-064-102-103.pools.arcor-ip.net) has joined #beagle
[20:18:03] <koen> denix: I don't think we have people testing i686
[20:19:06] <denix> koen: yeah, I couldn't find any. I'm doing some testing with i686-generic
[20:19:50] <koen> I can add it to the autobuilder if that would make your life easier
[20:19:54] * Fl1pFl0p (n=flipflop@146.244.179.229) has joined #beagle
[20:20:14] <koen> qemux86 is there, but that's only i586
[20:20:20] <denix> koen: no, don't bother. I'm building things myself anyway
[20:21:14] <denix> yeah, saw it. I tried qemux86 for i586 before, until glibc was fixed to support i686
[20:22:29] <koen> I got my old 12" powerbook back, so need need for an eee901 anymore
[20:23:11] <denix> heh, I like my 901
[20:23:18] <koen> I bet
[20:23:21] <denix> not a big apple fan though...
[20:25:46] <koen> denix: I integrated the new dsplink makefile the link team created this week
[20:26:14] <Crofton|irssi> koen: does it make life better?
[20:26:24] <denix> I saw your commits, but haven't tried it yet
[20:26:29] <koen> denix: it now hooks into kbuild so no more need to force module loading
[20:26:32] <denix> does it work with 2.6.27?
[20:26:50] <koen> denix: no, since the include paths in the .c files are still wrong
[20:27:04] <koen> denix: that can be solved with a trivial sedscript if needed
[20:28:12] * jkridner|work (n=a0321898@nat/ti/x-8960058d9f1130c5) Quit ("Leaving.")
[20:28:36] <koen> s:arch:mach:g or something similar
[20:29:11] <ZeZu> Crofton|irssi, so it then builds two toolchains ? reg. gcc cross for userspace and then csl for kernel ?
[20:29:17] <ZeZu> little wonder it took so long :|
[20:29:25] <denix> ZeZu: yes
[20:29:35] <ZeZu> and what opts does it use by default ?
[20:29:41] <Crofton|irssi> we like to take long coffee breaks
[20:29:52] <ZeZu> yea
[20:30:09] <ZeZu> on a c2d w/ 2gb of ram it must have taken 6hours just for console-image
[20:30:42] <Crofton|irssi> do you have || mke enabled
[20:30:43] <denix> build time is one of the reasons for us to use binary csl as an external toolchain...
[20:31:22] <denix> PARALLEL_MAKE = "-j 4" and BB_NUMBER_THREADS = "4"
[20:31:32] <denix> that's for my c2q
[20:31:34] <garren> has anyone got an ipod touch?
[20:31:59] * garren (n=garren@dsl-244-222-39.telkomadsl.co.za) has left #beagle
[20:32:02] * mru has touched an ipod
[20:32:33] <denix> has ipod touched mru?
[20:32:47] * mru is untouched by ipod
[20:33:00] <mru> untainted even
[20:33:32] <ZeZu> denix: specify those in local.conf ?
[20:33:41] <denix> ZeZu: yep
[20:34:33] <mru> koen: I'd like to remove those dmb instructions from ffmpeg
[20:34:50] <mru> but that will make beagles start crashing without a u-boot update
[20:34:51] <ZeZu> i've built a few systems by hand before that didn't take that long, but all the files were already downloaded and there wasn't as much going on behind the packaging
[20:41:37] * FlipFlop (n=flipflop@146.244.179.229) Quit (Success)
[20:45:35] <koen> mru: people getting a new ffmpeg should be smart enought to get a new uboot as well
[20:46:00] <koen> mru: I just hope TI will put a decent uboot on new beagles instead of that buggy TI uboot
[20:46:19] <mru> they will eventually
[20:46:25] <mru> jkridner said so
[20:48:14] * chelli (n=chelli@p54B86AC2.dip.t-dialin.net) has joined #beagle
[20:49:41] <mru> btw, qemu is much more broken for sh4 than for arm
[20:50:26] * mru noticed that ffmpeg has been broken on sh4 for 5 years
[20:51:12] <koen> I trust qemu as far as I can throw a piano
[20:51:46] <mru> so when they build that moon base, will you use qemu there?
[20:53:51] <koen> maybe it can be compiled with gcc 4.0 by then
[20:54:08] <mru> yeah, what's with the reliance on gcc 3?
[20:54:28] * FuL|OUT is now known as fulgas
[20:54:52] <koen> something to do with their jit depending on gcc internals
[20:55:02] <mru> scary thought
[20:55:17] <koen> AIUI they assume that the code is always going to be compiled in a certain way
[20:55:57] <koen> at least, that's how people explained it to me
[20:55:58] <mru> if you need specific instruction sequences, you write assembler
[20:56:37] <Crofton|irssi> mru: you are far to logical
[20:57:18] <suihkulokki> the svn version of qemu supposedly doesn't need gcc3 anymore
[20:57:20] <mru> koen: are my ears pointy?
[20:57:28] <mru> suihkulokki: it did yesterday
[20:57:41] <koen> mru: not really
[20:58:00] <mru> must be another explanation then
[21:00:20] * _zAlt (n=null@c-24-14-8-40.hsd1.in.comcast.net) has joined #beagle
[21:00:40] * ZeZu (n=null@c-24-14-8-40.hsd1.in.comcast.net) Quit (Read error: 54 (Connection reset by peer))
[21:00:52] * _zAlt is now known as ZeZu
[21:07:50] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Remote closed the connection)
[21:17:41] * JoeyBorn (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) has joined #beagle
[21:24:41] * Beagle7 (n=Beagle7@251.Red-213-97-34.staticIP.rima-tde.net) has joined #beagle
[21:29:58] * pk87 (n=Marlies@DSL01.83.171.180.196.ip-pool.NEFkom.net) has joined #beagle
[21:30:21] * pk87 (n=Marlies@DSL01.83.171.180.196.ip-pool.NEFkom.net) has left #beagle
[21:34:21] * fulgas (n=fn@a83-132-158-61.cpe.netcabo.pt) Quit (Remote closed the connection)
[21:35:05] * fulgas (n=fn@a83-132-158-61.cpe.netcabo.pt) has joined #beagle
[21:38:28] * Beagle7 (n=Beagle7@251.Red-213-97-34.staticIP.rima-tde.net) Quit (Connection reset by peer)
[21:43:14] * kevinsc (n=kevinsc@nat/ti/x-e172fec5e71abfdb) has joined #beagle
[21:45:14] * kevinsc (n=kevinsc@nat/ti/x-e172fec5e71abfdb) has left #beagle
[21:45:36] * kevinsc (n=kevinsc@nat/ti/x-7ba1af9cd964af90) has joined #beagle
[21:45:51] * john280z (n=jmitchel@static-71-111-254-119.rlghnc.dsl-w.verizon.net) Quit ("Leaving")
[21:48:05] <ZeZu> http://rafb.net/p/eLgnB739.html << anyone see whats going on here? I can't tell if its getting a bad file or ..
[21:50:09] <ZeZu> enchant-1.3.0.tar.gz is valid
[21:51:03] <mru> you're seeing the output of two or more commands interleaved
[21:53:00] <mru> it looks like it's trying to fetch ffmpeg from my git using an old revision id that doesn't exist
[21:53:41] * kevinsc (n=kevinsc@nat/ti/x-7ba1af9cd964af90) has left #beagle
[21:54:14] * Fl1pFl0p (n=flipflop@146.244.179.229) Quit ()
[21:57:26] * BThompson (n=BThompso@nat/ti/x-8393ff4bca7bf2cc) Quit ("Trillian (http://www.ceruleanstudios.com")
[21:57:32] <ZeZu> fun
[21:57:42] <ZeZu> let me update git
[21:57:49] <ZeZu> just did so not even an hour ago though :|
[22:00:55] <ZeZu> updated but doesn't fix ffmpeg, weird how can i update ffmpeg git version ?
[22:01:45] <mru> you'll have to ask one of the OE guys
[22:03:55] <geckosenator> can the beagle board act like a usb keyboard?
[22:04:17] <geckosenator> I mean, are there drivers for that
[22:06:07] <Crofton|irssi> http://gitweb.openembedded.net/?p=openembedded.git;a=blob;f=packages/ffmpeg/ffmpeg_git.bb;h=3f8785b6d7fa1ed2fdeb7fb46270ca51f2283f5d;hb=HEAD
[22:06:25] <Crofton|irssi> ZeZu: look at the SRCREV entries and adjust them :)
[22:25:59] <RobotGuy> My build is still running. :)
[22:26:11] * _fulgas (n=fn@a83-132-158-61.cpe.netcabo.pt) has joined #beagle
[22:26:11] <RobotGuy> I had to clear some disk space though.
[22:27:00] * fulgas (n=fn@a83-132-158-61.cpe.netcabo.pt) Quit (Read error: 104 (Connection reset by peer))
[22:32:19] <mru> ZeZu: I suggest revision 65acfde452afbd78eea7c369b297bb9bcb3544c4 for ffmpeg
[22:34:13] <mru> note: you will need the latest u-boot to use that revision
[22:45:33] * _bernie (n=bernie@fw-1-user-net-flrs.cictr.com) Quit (Read error: 110 (Connection timed out))
[22:47:33] <ldesnogu_> mru, qemu sh4 has been patched today; I don't know if that will make it work though
[22:47:47] <mru> I doubt it
[22:48:07] <ldesnogu_> also qemu does not need any special gcc anymore
[22:48:11] <ldesnogu_> except for ppc
[22:48:15] <mru> I encountered a missing instruction which hasn't been added
[22:48:35] <ldesnogu_> which one?
[22:48:56] <mru> fldi0
[22:49:51] <ldesnogu_> case 0xf08d: /* fldi0 FRn - FPSCR: R[PR] */
[22:50:01] <ldesnogu_> well that doesn't prove much I know :)
[22:50:41] <mru> and it the configure script still looks for gcc 3
[22:50:50] <ldesnogu_> you can force it
[22:51:12] <ldesnogu_> use --disable-gcc-check
[22:53:55] * flo_lap (n=fuchs@g228208103.adsl.alicedsl.de) has joined #beagle
[22:54:40] * fagius (n=fagius@static-71-111-254-119.rlghnc.dsl-w.verizon.net) Quit ("Leaving")
[22:54:59] <ldesnogu_> mru, if you find bugs, please send them to me (including statically linked tests if possible) and I'll post patches for inclusion
[22:55:00] * _fulgas is now known as FuL|OUT
[22:55:00] * FuL|OUT is now known as fulgas
[22:55:24] <mru> didn't know you cared about sh4
[22:55:38] <mru> not that I particularly do
[22:55:55] <ldesnogu_> I don't, but I care about qemu, and learning about a processor I don't know is always fun
[22:55:59] <mru> I just seem to be the only ffmpeg dev who knows anything about it
[22:56:18] <ldesnogu_> yeah I saw your mails and commits...
[22:56:37] <mru> I've dealt with it a bit at work
[22:56:51] <ldesnogu_> is it still used in some STB?
[22:56:58] <mru> had to disassemble parts of their libc
[22:57:11] <mru> it's in top-end, new STBs
[22:57:29] <mru> the STi71xx chips all use it
[22:58:03] * ldesnogu_ can't stop thinking MIPS is king in STB
[22:58:16] <mru> there's a lot of mips for sure
[22:58:21] <mru> quite a bit of arm too
[22:58:40] <mru> usually in conexant chips
[22:58:51] <ldesnogu_> yes I learned that during some internal marketing meeting, and was surprised
[22:59:03] <mru> it's even mentioned on the arm website
[22:59:14] <ZeZu> sh4 isn't bad
[22:59:29] <mru> no, it's not a bad architecture
[22:59:32] <ZeZu> emulated it for the dreamcast, the whole hw arch is nice, uses a pvr clx2 chipset as well
[22:59:57] <mru> the ST chips could be improved though
[22:59:58] <ldesnogu_> I thought SH was Hitachi
[23:00:02] <ZeZu> it is
[23:00:09] <ZeZu> Hitachi SuperSH series
[23:00:11] <mru> it is
[23:00:17] <mru> ST licences the core
[23:00:22] <ldesnogu_> they talk about SuperSH Inc
[23:00:35] <mru> renesas is involved too
[23:00:37] <ZeZu> which is now raneasas
[23:00:40] <ldesnogu_> so it's an hitachi subsidiary
[23:00:42] <ldesnogu_> ok
[23:01:45] <ldesnogu_> it seems to be very FP oriented, with not great integer performance
[23:01:47] <mru> qemu ppc is failing to build: dyngen: ret or jmp expected at the end of op_lswx_raw
[23:02:03] <mru> yeah, that's a bit disappointing
[23:02:06] <ldesnogu_> mru, I told you above ppc is the only target requiring gcc 3 :)
[23:02:15] <mru> so how can I disable it?
[23:02:18] <ldesnogu_> it's being converted
[23:02:37] <ldesnogu_> your need user emu?
[23:02:43] <ldesnogu_> or full system?
[23:02:50] <mru> I'm building only user
[23:03:12] <mru> that's all I need at the moment
[23:03:19] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
[23:03:25] <ldesnogu_> then either --target-list=sh4-linux-user or sh4eb-linux-user
[23:04:10] <mru> is this documented somewhere?
[23:04:11] * cbrake is now known as cbrake_away
[23:04:30] <ldesnogu_> hum good question :)
[23:04:46] * shoragan (n=shoragan@debian/developer/shoragan) Quit (Remote closed the connection)
[23:05:18] <ldesnogu_> qemu is not very well documented
[23:05:30] <ZeZu> Crofton|irssi, my SRVREV is the same as the one you posted
[23:05:33] <ldesnogu_> especially the svn version which is very different from last release
[23:05:35] <ZeZu> SRCREV *
[23:05:37] <mru> the ffmpeg dev won't comment on that...
[23:06:13] <ldesnogu_> :)
[23:06:15] <mru> ZeZu: and it doesn't exist
[23:06:30] <mru> the latest git version of that file has a good ffmpeg rev
[23:06:34] <ZeZu> i updated to the one you suggested
[23:06:49] <mru> that's the latest as of today
[23:06:53] <ZeZu> so git is broken here? i've updated the git repos.
[23:07:01] <mru> no, it's not broken
[23:07:42] <mru> koen apparently managed to set a revision that deleted moments later
[23:07:57] <mru> it probably lingered in my tree for some after that
[23:08:18] * shoragan (n=shoragan@debian/developer/shoragan) has joined #beagle
[23:09:27] <ZeZu> updating it did not help either ;|
[23:10:56] <ZeZu> wait i likely updated the wrong one trying again
[23:14:27] <mru> ldesnogu_: unknown instruction 0xf08d at pc 0x00793ee8
[23:14:32] * john280z (n=johnm@user-0c2h5av.cable.mindspring.com) has joined #beagle
[23:15:19] <ldesnogu_> hum
[23:15:55] <mru> ah, I see the problem
[23:16:02] <mru> and the code I'm trying to run is buggy too
[23:16:03] <ldesnogu_> fpscr is wrong?
[23:16:19] <mru> fpscr.pr is probably wrong
[23:16:25] <ldesnogu_> :)
[23:16:27] <mru> but in that case the instruction should be a noop
[23:16:39] <ldesnogu_> even in user mode?
[23:17:02] <mru> the manual says "possible exceptions: none"
[23:17:43] <ldesnogu_> ok, I will let the last committer know about that
[23:20:53] * abitos (n=nixgibts@dslb-084-057-137-013.pools.arcor-ip.net) has joined #beagle
[23:23:04] <mru> now to figure out what's wrong with the code
[23:35:02] * guillaum1 (n=Guillaum@AMontsouris-153-1-67-44.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
[23:35:19] * guillaum1 (n=gl@AMontsouris-153-1-67-44.w90-2.abo.wanadoo.fr) has joined #beagle
[23:49:18] * felipec (n=felipec@a91-153-251-222.elisa-laajakaista.fi) Quit ("Leaving")
[23:53:52] <ds2> general question - does anyone care about the details on how to setup the pinmux to use the expansion header?
[23:54:22] <jkridner> koen, ZeZu: we crossed 2^10 mailing list members now. I've started rejecting requests that don't have some text as I see them as potential spammers.
[23:56:04] <ZeZu> ? i haven't sent anything to the mailing list
[23:56:17] <robclark> who was the kilo-subscriber?
[23:56:50] <Crofton|irssi> ds3 yes
[23:56:53] <Crofton|irssi> er
[23:56:55] <Crofton|irssi> ds2: yes
[23:57:16] <ZeZu> weird when i start up angstrom demo image and put in username etc it gives me a blank list to choose a language, if i press next anyways it gives a blank profiles list and next btn is greyed out
[23:57:26] <ds2> ds2 and ds3 are both me ;)
[23:57:35] <Crofton|irssi> heh
[23:57:47] <ds2> have to load balance onto a difference client with some channels
[23:57:59] <Crofton|irssi> ZeZu: that sounds like a conversation sakoman_ was having a while back with some guys in #oe ....
[23:58:34] <ZeZu> ooh my net is down and firefox wont work but irc is still going
[23:58:36] <ZeZu> interesting
[23:58:53] <mru> maybe dns issue
[23:59:20] <ZeZu> it is, windows doesn't seem to like the subnet i setup for beagle usb/eth gadget
[23:59:20] <ds2> Crofton|irssi: is there any details you would like to see? I am thinking of writing up a page on setting up the pinmux instead of just answering stuff on the list
[23:59:27] <ZeZu> not sure why didn't look into it much
[23:59:40] * rsalveti (n=salveti@200.184.118.132) Quit (Remote closed the connection)
[23:59:57] <Crofton|irssi> ds2: I do not know enough to ask what I need to know yet