• [01:22:26] * bazbell (n=a0192809@nat/ti/x-000a7c94867ebcc7) has joined #beagle
  • [01:30:12] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [01:42:45] <mru> been running entirely off mmc card for a while now, no crashes
  • [02:12:24] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) Quit ("Aloha")
  • [02:19:50] * JoeBorn_ (n=jborn@adsl-75-3-15-70.dsl.chcgil.sbcglobal.net) has joined #beagle
  • [02:58:50] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [02:59:09] * Olipro_ (n=Olipro@unaffiliated/olipro) Quit (Read error: 104 (Connection reset by peer))
  • [02:59:13] * Olipro (n=Olipro@unaffiliated/olipro) Quit ("Leaving")
  • [02:59:27] * Olipro (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [03:01:26] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Client Quit)
  • [03:01:35] * Olipro (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [03:14:08] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [03:21:23] * lgentili (n=lgentili@host186.190-224-254.telecom.net.ar) has joined #beagle
  • [03:22:09] * lgentili (n=lgentili@host186.190-224-254.telecom.net.ar) Quit (Remote closed the connection)
  • [03:23:08] * banderson (n=irc@69.71.183.7) Quit (Read error: 110 (Connection timed out))
  • [03:23:11] * JoeBorn (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) Quit (Nick collision from services.)
  • [03:23:17] * JoeBorn_ is now known as JoeBorn
  • [03:23:35] * ScumBagBob (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) has joined #beagle
  • [04:09:00] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) Quit ("Trillian (http://www.ceruleanstudios.com")
  • [05:08:05] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [06:59:55] * trickie (n=trickie@basesoft.xs4all.nl) has joined #beagle
  • [07:02:34] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Read error: 60 (Operation timed out))
  • [07:19:47] <khasim> I see latest OMAP GIT tree hanging for Beagle.
  • [07:20:15] <khasim> I dont see any crash, but after a while the typing anything doesnt work
  • [08:04:16] <mru> I've had the rs232 console stop responding a few times
  • [08:04:19] <mru> ssh still works
  • [08:12:43] <koen> yeah, serial stops after a while
  • [08:13:20] <koen> and after a while sd stops as well, so you can only do stuff that's in ram, any IO will block
  • [08:13:58] <koen> it can take >1 day to get there, but mostly it's ~30 minutes
  • [08:14:25] <mru> I've never had that happen
  • [08:44:18] <khasim> Koen: did we discuss this hang on linux omap git mailing list?
  • [08:44:39] <koen> iirc not
  • [08:45:34] <khasim> I found OMAP3430 SDP working, but beagle doesnt
  • [08:45:45] <khasim> need to check any thing specific to our board.
  • [08:46:51] * dirk2 (n=dirk@F304d.f.strato-dslnet.de) has joined #beagle
  • [08:55:24] * dirk2 (n=dirk@F304d.f.strato-dslnet.de) has left #beagle
  • [09:52:46] * RogerMonk (n=a0740758@nat/ti/x-31b455e4ff66547b) has joined #beagle
  • [10:30:33] <khasim> Koen: I think disabling RTC will remove the hang
  • [10:30:45] <khasim> Do we need RTC for any of our demos ?
  • [10:36:41] * docelic (n=docelic@78.134.206.172) has joined #beagle
  • [10:44:18] * roxfan2 (n=dunno@126.198-201-80.adsl-dyn.isp.belgacom.be) has joined #beagle
  • [10:55:41] * roxfan (n=dunno@251.215-200-80.adsl-dyn.isp.belgacom.be) Quit (Read error: 110 (Connection timed out))
  • [10:55:44] <koen> khasim: I wonder why the rtc would cause a hang
  • [11:37:22] <khasim> I think the IRQ handling was changed for twl RTC, Battery and USB. This has caused the issue
  • [11:37:56] <khasim> I didnt look into the details of the change. Am working on porting all drivers to latest GIT, will take this up at the end.
  • [11:52:58] <Crofton|work> khasim, having all the drivers in git will be awesome
  • [11:54:19] <koen> iirc WTBU doesn't support RTC on beagle
  • [11:56:16] * bazbell (n=a0192809@nat/ti/x-000a7c94867ebcc7) Quit (Remote closed the connection)
  • [12:16:31] * rsalveti_ (n=salveti@200.184.118.132) has joined #beagle
  • [12:16:44] * rsalveti (n=salveti@189.70.97.120) Quit (Read error: 110 (Connection timed out))
  • [12:18:08] * rsalveti_ (n=salveti@200.184.118.132) Quit (Remote closed the connection)
  • [12:18:17] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
  • [12:23:12] <RogerMonk> Koen, Crofton, you there?
  • [12:25:54] <koen> I'm here
  • [12:26:39] <RogerMonk> Hey sir - can you point me at latest Angstrom stuff please? Trying to get myself up on the same version as you guys. Thanks, ~rog
  • [12:26:51] <koen> binaries?
  • [12:27:00] <koen> http://www.angstrom-distribution.org/unstable/autobuild/beagleboard/
  • [12:27:00] <RogerMonk> both please, binaries first
  • [12:27:21] <koen> and everything is in OE, so now patches needed there
  • [12:27:54] * docelic_ (n=docelic@78.134.193.152) has joined #beagle
  • [12:28:03] <RogerMonk> gr8
  • [12:31:19] * bazbell (n=a0192809@nat/ti/x-2ae0c95f65c0b289) has joined #beagle
  • [12:34:42] <Crofton> beagle or EVM?
  • [12:37:55] <RogerMonk> both actually if possible?
  • [12:38:49] <Crofton> let me run set for the evm
  • [12:39:02] <RogerMonk> thanks!
  • [12:41:15] * docelic (n=docelic@78.134.206.172) Quit (Nick collision from services.)
  • [12:41:16] * docelic_ is now known as docelic
  • [12:45:06] <koen> RogerMonk: http://www.sakoman.net/omap3/evm/
  • [12:45:13] * Olipro (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [12:45:21] <koen> oh wait, that one is a bit old
  • [12:49:43] <sakoman> http://www.sakoman.net/feeds/omap3/glibc/images/
  • [12:59:32] <sakoman> koen: still trying to track down my video issue
  • [13:01:12] <sakoman> I've finally found a combination of cable/monitor that will display the u-boot beagel splash: hdmi cable straight to the hdmi input on my 55" plasma monitor
  • [13:01:49] <koen> does fbdev work on that?
  • [13:01:51] <sakoman> but as soon as beagle boots it goes black
  • [13:02:05] <sakoman> er, as soon as linux boots
  • [13:02:17] <koen> right
  • [13:02:32] <koen> it will display psplash as soon as userspace boots
  • [13:02:35] <sakoman> never get anything on the screen after that
  • [13:03:18] <sakoman> can't really tell whether it doesn't like timing or whether there is some other issue
  • [13:04:11] <koen> you could try adjusting the lcd-beagle file to match uboot
  • [13:04:16] <sakoman> with my Apple and Viewsonic monitors I can't even get the u-boot splash
  • [13:05:19] <sakoman> koen: yeah that's my next step. if you looked at the u-boot code you will know it will not be fun :-)
  • [13:08:29] <koen> I'm still wondering why linux only uses 48MHz as pixclock
  • [13:13:59] <Crofton|work> sakoman, I have a DELL FP
  • [13:15:15] <sakoman> koen: I think I understand that part -- I'll look at fixing that while trying to resolve my video problem
  • [13:16:03] <sakoman> Crofton: I'm not quite ready to solve the problem by buying a new monitor ;-)
  • [13:16:17] <sakoman> Hopefully it's just a timing thing!
  • [13:16:25] <sakoman> well, off for coffee
  • [13:16:25] <Crofton|work> yeah
  • [13:16:29] <Crofton|work> l8r
  • [13:16:41] <Crofton|work> at some point I need to buy a monitor for the beagle
  • [13:18:43] <koen> at some point I need to clean my desk so I can sit behind it
  • [13:19:00] <Crofton|work> heh
  • [13:19:06] <Crofton|work> no wonder you work from the couch
  • [13:28:28] <khasim> Porting drivers is no more easier, there are three versions of every damn thing on OMAP, one for EVM, one for Zoom, one for SDP, +1 already on OMAP GIT
  • [13:33:59] <koen> the goal is to have only one in git :)
  • [13:34:09] <koen> mplayer on beagleboard: http://scap.linuxtogo.org/files/fc987d4acb2c745fb7e19cf4dca8de70.png
  • [13:35:33] * BThompson (n=BThompso@nat/ti/x-778f45b85669c1ca) has joined #beagle
  • [13:37:06] <RogerMonk> Koen - that's awesome!!
  • [13:37:21] <RogerMonk> do you have a pre-built image you could make available?
  • [13:37:57] <RogerMonk> Now we just need some of mru's Neon acceleration added :)
  • [13:38:28] <RogerMonk> Which window are you writing to with mplayer?
  • [13:39:59] <koen> RogerMonk: plain x11, there is no accel yet
  • [13:40:13] <koen> RogerMonk: grab the demo-image from http://www.angstrom-distribution.org/unstable/autobuild/beagleboard/
  • [13:40:24] <RogerMonk> it's coming down...
  • [13:40:47] <koen> I sadly have no clue on how to get mru's ffmpeg work into an mplayer build
  • [13:41:19] <koen> Crofton|work: RP just received his beagleboard
  • [13:41:21] <RogerMonk> ah-ha - should be ok, I think - mru has it in his FFMPEG git tree - should be able to replace the libavcodec libraries with his new ones
  • [13:41:40] <Crofton|work> awesome
  • [13:41:41] <RogerMonk> mru - you there?
  • [13:41:46] <koen> the code is in http://git.mansr.com/?p=ffmpeg.mru;a=shortlog;h=refs/heads/arm-neon
  • [13:41:55] <koen> and I can already build ffmpeg from there
  • [13:42:03] <RogerMonk> *saweet!*
  • [13:42:06] <koen> but mplayer wants to include it and statically link to it
  • [13:42:12] <Crofton|work> this is becoming some kind of TI plot to get all the OE devs working on OMAP3
  • [13:42:14] <RogerMonk> ah-ok
  • [13:42:26] <RogerMonk> :)
  • [13:42:31] <RogerMonk> it's where it's all happening?
  • [13:42:37] <RogerMonk> !
  • [13:42:54] <koen> well, 4 OE core-developers now have a beagle :)
  • [13:43:20] <RogerMonk> ... that's the same as the total number of boards in europe!
  • [13:43:23] <koen> I think the OE developer community now has 6 to 8 beagles in total :)
  • [13:55:09] * BThompson (n=BThompso@nat/ti/x-778f45b85669c1ca) Quit ("Trillian (http://www.ceruleanstudios.com")
  • [14:04:32] <koen> khasim: I disabled RTC and the serial console still froze :(
  • [14:06:23] <sakoman> koen: IIRC turning off APM ends the console freezes ;-)
  • [14:06:53] <koen> gah
  • [14:06:55] <koen> stupid apm
  • [14:08:19] <koen> sakoman: apm or apm emulation?
  • [14:08:50] <sakoman> I think it was apm emulation -- was a while back that I was experimenting with config changes
  • [14:09:31] * koen tries
  • [14:09:47] <sakoman> I hope that's right!
  • [14:10:48] <sakoman> Whatever it is is disabled in the EVM defconfig and I never get console hangs there
  • [14:30:27] <khasim> ON EVM Defconfig RTC is not enabled
  • [14:30:37] <khasim> on Beagle defconfig RTC is enabled
  • [14:30:45] <khasim> Koen: How did you disable RTC
  • [14:31:07] <khasim> I suggest do a clean build, or at least make sure the RTC files are not getting compiled
  • [14:31:29] <khasim> Class structure changes in latest kernel are killing me
  • [14:31:51] <koen> khasim: by changing the defconfig
  • [14:32:06] <koen> khasim: I disabled the twl4030 rtc driver
  • [14:32:25] <koen> CONFIG_RTC_DRV_TWL4030
  • [14:33:56] <koen> RogerMonk: it seems that just dropping in the libavcodec dir works
  • [14:37:08] <RogerMonk> cool!
  • [14:38:28] <RogerMonk> Koen, for your demo image - what does your USB hookup look like?
  • [14:39:27] <koen> RogerMonk: beagleboard -> hub -> {keyboard,mouse,network,lcd}
  • [14:39:52] <RogerMonk> are you able to run NFS ok?
  • [14:40:00] <RogerMonk> over USB network?
  • [14:40:13] <RogerMonk> (what's lcd ?)
  • [14:40:30] * BThompson (n=BThompso@nat/ti/x-e96d1c833bd882aa) has joined #beagle
  • [14:41:09] <koen> RogerMonk: lcd: http://www.youtube.com/watch?v=R33dzREZGEk
  • [14:42:55] <RogerMonk> nice!
  • [14:42:58] <RogerMonk> I need one of those!
  • [14:44:27] <koen> you can build one: http://www.harbaum.org/till/lcd2usb/index.shtml :)
  • [14:48:03] <RogerMonk> Koen - NFS ?
  • [14:49:27] <koen> I don't use NFS :)
  • [14:49:39] <koen> sd cards are much faster and easy to use
  • [14:49:52] <RogerMonk> ok
  • [15:00:16] <keesj> I have RSI from sdcard based development.
  • [15:07:48] <khasim> The latest kernel has changed SYSFS file creation methods http://lkml.org/lkml/2008/4/30/591
  • [15:08:44] <khasim> Koen: I have disabled entire RTC from the device drivers list
  • [15:15:38] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [15:50:36] * banderson (n=irc@69.71.183.7) has joined #beagle
  • [15:58:25] <denix> koen: what kind of fancy top you have on the back of your screenshot?
  • [16:00:49] * hagisbasheruk (n=hagisbas@78.148.131.1) has joined #beagle
  • [16:24:42] <RogerMonk> yeah - I was wondering the same - looks pretty nice
  • [16:49:14] * dirk2 (n=dirk@F30fd.f.strato-dslnet.de) has joined #beagle
  • [16:54:03] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [16:56:04] <dirk2> khasim: ping
  • [16:57:01] <khasim> hi dirk
  • [16:57:47] <dirk2> just read through the logs: What do you think is wrong with RTC (interrupts)?
  • [16:58:08] <dirk2> "<khasim> I think the IRQ handling was changed for twl RTC, Battery and USB. This has caused the issue"
  • [16:59:43] <khasim> When booted the GIT kernel I found that after some time there are no timer or PRCM interrupts and the control resides in cpu_idle
  • [17:00:34] <khasim> Then I started comparing my previous versions and 3430sdp defconfig, found that RTC was not enabled. I just disabled it for Beagle and it stopped hanging
  • [17:00:48] <khasim> I dont know where exactly the problem in RTC is
  • [17:01:40] <khasim> While browsing the history of files related to RTC I found that some changes were done to the files to handle T2 interrupts in a centralized way for USB, Battery and RTC
  • [17:01:49] <khasim> Even Battery broke with this patch set
  • [17:03:26] <khasim> http://source.mvista.com/git/gitweb.cgi?p=linux-omap-2.6.git;a=commitdiff;h=68d7477caca19c0b52b5d4e85700cd3e6115577f
  • [17:03:40] <khasim> May be this is the change that is affecting RTC.
  • [17:04:23] <khasim> I still cant understand why it is affecting GP timer and PRCM interrupts.
  • [17:04:54] <khasim> I also have code snippet to handle IRQ -33 interrupts.
  • [17:05:14] <khasim> I lost some time today to debug the RTC and SYSFS issues
  • [17:05:44] <khasim> I am currently working on my long time pending task to give complete driver set on latest GIT kernel in the form of Quilt
  • [17:05:58] <dirk2> Thanks! No, I don't think you loose time while debugging issues ;) While it is nice to port all drivers to git, wouldn't it be better to fix issues found in existing code, first, before adding new code? Or, if there is no time for this, do you like to drop a mail with these findings to omap list? Maybe the experts there which did the changes have an idea if we give them a note.
  • [17:07:37] <khasim> I heard some TI folks working on Battery driver, I didnt find any one working on RTC though. Will update the list with my findings
  • [17:08:03] <khasim> I understand your suggestion for fixing the issues
  • [17:08:55] <khasim> But we are in desparate need for a complete source code for demos and Beagle release
  • [17:09:44] * decriptor (n=decripto@166-70-62-135.ip.xmission.com) has joined #beagle
  • [17:09:53] * decriptor (n=decripto@166-70-62-135.ip.xmission.com) has left #beagle
  • [17:10:33] <dirk2> If "complete source code == move all drivers to git" then it wouldn't help you if source code is complete but has issues here and there, i.e. isn't stable
  • [17:11:14] <khasim> I want it to be stable in terms of functionality - meaning it shouldnt be broken or crashing or hanging
  • [17:11:36] <khasim> I was thinking of putting patches to OE as a recipe
  • [17:11:44] <khasim> then start cleaning up for upstream
  • [17:11:57] <khasim> The major issue we have is for DSS
  • [17:12:18] <khasim> this has multiple dependency as you know
  • [17:13:13] <dirk2> "in terms of functionality": In which other terms can it be stable?
  • [17:13:26] <khasim> :)
  • [17:14:10] <khasim> unstable in terms of code
  • [17:14:19] <dirk2> Ah ;)
  • [17:14:21] <khasim> meaning unable to submit upstream
  • [17:14:54] <khasim> but I am also aware that we are not encouraging community to work on latest GIT
  • [17:15:24] <khasim> but as I mentioned I will continue working on making patches very minimum
  • [17:15:37] <khasim> I think we should only have DSS as the delta
  • [17:15:44] <khasim> and take Steve's work for ASOC
  • [17:15:59] <dirk2> Sorry, why are we "not encouraging community to work on latest GIT"?
  • [17:16:52] <khasim> meaning if we put a complete tar ball out - community should not take that alone, but also help us in fixing the issues pending to make it minimal
  • [17:18:17] <dirk2> So you plan to put a (kernel git) tar ball out including fixes for demos and beagle releases?
  • [17:21:41] * hagisbasheruk (n=hagisbas@78.148.131.1) Quit ("Ex-Chat")
  • [17:22:49] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [17:29:00] * jkridner|work (n=a0321898@nat/ti/x-78351b1494d27875) has joined #beagle
  • [17:29:47] <jkridner|work> good morning
  • [17:30:31] <jkridner|work> mru: you planning to get DVI working?
  • [17:31:48] <denix> good ugt morning
  • [17:34:38] <jkridner|work> indeed. I have demonstrated my understanding of ugt.
  • [17:35:11] * khasim (n=a0393720@117.192.98.244) has joined #beagle
  • [17:36:54] <khasim> sorry
  • [17:37:02] <khasim> my network was gone
  • [17:37:12] <khasim> I missed some important conversation
  • [17:37:26] <khasim> dirk r u there
  • [17:38:06] <khasim> Dirk: So you plan to put a (kernel git) tar ball out including fixes for demos and beagle releases?
  • [17:38:19] <khasim> Ans:
  • [17:38:58] <khasim> No, I am planning to get complete functionality on latest GIT, maintain them as patches and have them added to OE distribution
  • [17:39:22] <khasim> I would like to have single site for all sources, maintained, managed
  • [17:40:28] <khasim> Can some one type some thing please? Just to confirm if my messages are going through as I recycled my network hardware here :(
  • [17:40:42] <jkridner|work> hi
  • [17:40:44] <jkridner|work> :)
  • [17:40:52] <khasim> thanks Jason :)
  • [17:41:28] <khasim> For u-boot, we can continue with Steve's tree + a mirror at beagle board.org
  • [17:41:58] <ds2> you could always do a server request, i.e. "/date" to verify server connection and if you wait about 2 minutes, the automatic server PING checking will kick you off if there is no connection :)
  • [17:42:45] <khasim> For kernel OE, I need to find a server and mirror site - Having it on Steve's machine might be difficult to handle traffic and maintainence
  • [17:43:37] <khasim> Binary images, rootfs for demos and docs can be maintained at code.google
  • [17:43:41] <sakoman> khasim: I'm looking at hosted git
  • [17:44:00] <khasim> steve: you mean for kernel?
  • [17:44:12] <jkridner|work> ds2: It seems that Pidgin doesn't support /date
  • [17:44:47] <jkridner|work> github.com git.or.cz
  • [17:46:33] <jkridner|work> khasim: fyi, it helps to use the actual nick (sakoman) due to assumptions by chat clients.
  • [17:46:52] <khasim> yeah,
  • [17:47:00] <khasim> even I thought so
  • [17:47:12] * jkridner|work reboots machine due to bad behavior.
  • [17:47:50] <khasim> I will first get the patch set working, then will back to you all for maintaining...
  • [17:48:00] <khasim> I think this itself will take me a day or two :(
  • [17:48:55] <dirk2> khasim: why not send patch by patch, why one patch set?
  • [17:49:58] * jkridner|work (n=a0321898@nat/ti/x-78351b1494d27875) Quit (Remote closed the connection)
  • [17:50:09] <dirk2> khasim: If you have patch A ready (if I understand correctly ready == functionality & stability but ready != upstream ready) just send it.
  • [17:50:32] <dirk2> khasim: release early, release often ;)
  • [17:50:46] <sakoman> khasim: for u-boot, kernel, and mt oe branch
  • [17:51:06] <sakoman> s/mt/my
  • [17:51:54] <ds2> jkridner: what about /luser or "/quote date"?
  • [17:52:03] <ds2> doh
  • [17:53:20] <koen> denix,RogerMonk: the fancy top is htop
  • [17:54:03] <denix> koen: thanks
  • [17:54:40] <denix> koen: and thanks for fixing those patches!
  • [17:55:54] <dirk2> What I don't understand: Why do we always talk about "we need git repositories"? What's wrong with http://www.beagleboard.org/gitweb/ (except uboot and that they are more or less unused)?
  • [17:56:41] * jkridner|work (n=a0321898@nat/ti/x-0b385f5baa292298) has joined #beagle
  • [17:57:25] <denix> dirk2: ask jkridner|work
  • [17:57:35] <jkridner|work> ask me what?
  • [17:57:54] <denix> <dirk2> What I don't understand: Why do we always talk about "we need git repositories"? What's wrong with http://www.beagleboard.org/gitweb/ (except uboot and that they are more or less unused)?
  • [17:59:01] <jkridner|work> there are some headaches with submitting in that the system isn't setup for easy creation of user accounts.
  • [17:59:52] <jkridner|work> I can handle a small number, but my preference is for people to use services with better known reliability, like github.com or git.or.cz, and then to register their projects (working on a project registration page now).
  • [18:00:27] <koen> or http://gitorious.org/
  • [18:00:30] <jkridner|work> still, for top-priority stuff, I really want to mirror the git repositories on BeagleBoard.org.
  • [18:00:44] <jkridner|work> ah, didn't know about that one.
  • [18:00:59] * Crofton doesn't care where it is, just that it is maintained and good community works on it :)
  • [18:01:14] <Crofton> ask MV to host it with linux-davinci and linux-omap?
  • [18:01:17] <jkridner|work> I believe I finally have a mirror of sakoman's git repository for u-boot setup properly, but it still needs to be pounded a bit.
  • [18:01:36] <jkridner|work> only beagleboard.org.git and u-boot.git have anything useful in them right now.
  • [18:01:38] <ds2> why not make it a branch of linux-omap?
  • [18:02:03] <jkridner|work> I should delete/replace the others with mirroring of the appropriate repositories.
  • [18:02:17] <jkridner|work> ds2: which one?
  • [18:02:23] <jkridner|work> is that to khasim?
  • [18:02:45] <ds2> jkridner: sorry... I mean as the repo instead of having yet another repo for beagle code
  • [18:03:28] <jkridner|work> would we want to have longer-lived patches managed in git or as OE patches?
  • [18:03:40] <jkridner|work> (which could be in git/mtn themselves)
  • [18:03:56] * jkridner|work needs to visit the log to catch up on this conversation. :)
  • [18:04:09] <ds2> heh
  • [18:04:27] <jkridner|work> still, whatever is needed for hosting, I can help with some bandwidth and disk space.
  • [18:04:30] <ds2> OE patches is a bad idea since that would make it hard for non OE users
  • [18:05:02] <RogerMonk> Koen, ran your latest beagle OE kernel and filesys on OMAP3EVM - worked pretty well - able to login in to desktop and bring up console, etc
  • [18:05:57] <RogerMonk> then tried to use sakoman's kernel and filesys - doesn't seem to like mounting nfs. do you have a nfs network enabled kernel to go with this? - fails at root mount?
  • [18:06:09] <jkridner|work> I don't see how a branch is any easier than another repository.
  • [18:08:06] <khasim> a) We can have complete Kernel + uboot on respective GITs b) Tar images c) OE distribution
  • [18:08:48] <khasim> combination of these is what we have today and this will get difficult to maintain when we see more Beagle owners in coming months
  • [18:08:51] <ds2> jkridner: branch makes it easier to keep up with the linux-omap-26 tree
  • [18:09:01] * NishanthMenon (n=nmenon@nat/ti/x-d17da9f9ab031f66) has joined #beagle
  • [18:10:06] <ds2> seperate repo would mean the users will have to setup remote tracking branches to do diffs against where as a single repo would let diff's be generated much easier
  • [18:10:42] <ds2> unless you are expecting a significant divergence from linux-omap-26
  • [18:11:08] <khasim> meaning: we should be closest to repective GIT trees - this is definitely true
  • [18:12:02] <khasim> I am still unclear on handling OE
  • [18:12:30] <khasim> when it comes to Filesystem, apps, etc, we need a distribution than just GITs
  • [18:13:46] <khasim> We are also looking at integrating DSP side stuff to OE - this is still in preliminary discussion - but eventually some thing is required to handle this
  • [18:14:01] <khasim> any thoughts?
  • [18:14:10] <koen> RogerMonk: iirc the trouble is that ifupdown reinitializes eth0 during boot
  • [18:15:26] <ds2> OE is a possible distro but things that are naturally independant shouldn't be made OE specific.. so if a package needs patches to support beagle, I would suggest putting out patches for the package and OE can then pick that up
  • [18:15:33] <RogerMonk> Koen, it's not getting that far
  • [18:15:45] <RogerMonk> it won't mount the rootfs
  • [18:15:50] <ds2> that way nonOE users aren't burdened with deOE-ifying it
  • [18:16:04] <RogerMonk> how are you guys working with evm - nfs?
  • [18:16:12] <Crofton> ds2, all we want is a sanely packaged dsplink :)
  • [18:16:17] <ds2> nfs and jffs2
  • [18:16:21] <Crofton> that should improve ease of build for all
  • [18:16:29] <koen> ds2: OE isn't a distro
  • [18:16:37] <mru> hi guys
  • [18:16:42] <koen> hey mru
  • [18:16:43] <ds2> Crofton: yes, but I don't want patches for that in refer to the OE structure, etc
  • [18:16:51] <ds2> koen: it is more of a build system
  • [18:16:57] <koen> right
  • [18:17:13] <Crofton> ideally, the kernel modules would go into the kernel tree
  • [18:17:23] <Crofton> the arm user space would build via configure
  • [18:17:33] <Crofton> and I am not how the dsp side would be packaged
  • [18:17:57] <khasim> To summarize, we need to have kernel on GIT + patches maintained separately
  • [18:18:05] <khasim> same thing with u-boot
  • [18:18:07] <Crofton> this gets complicated because (I think) they are building for what we would think of as "odd systems"
  • [18:18:17] <khasim> both should be available independently
  • [18:18:22] <ds2> ideally, one day there will be no patches...it'll just be the OMAP kernel
  • [18:18:24] <koen> mru: what woould be the best format for the demo movie, mpeg2, h264, something else?
  • [18:18:40] <mru> h264 is very resource hungry
  • [18:18:44] <ds2> theora
  • [18:18:54] <koen> FWIW, the neuros people have a repackaged codec engine buildsystem
  • [18:18:57] <khasim> OE should be configured to collate all the patches and respective kernels/u-boot from respective GIT tree
  • [18:19:00] <koen> haven't tried it yet, though
  • [18:19:03] <mru> koen: h264 is very resource hungry
  • [18:19:03] <khasim> am I correct ?
  • [18:19:15] <mru> and I've done no neon optimisations for it yet
  • [18:19:32] <dirk2> khasim: GIT + patches maintained separately (uboot & kernel): Why? Do you know sakoman uboot git? this is fine, no?
  • [18:19:41] <mru> theora is imho a joke
  • [18:19:49] <ds2> khasim: I would agree with that. So first time users can just grab OE but more advance users can just goto the patches themselves if they don't need OE
  • [18:20:12] <khasim> I mean GIT for sure. yes it can be sakoman's tree itself - I just understanding the flow
  • [18:20:15] <RogerMonk> mru, koen : we have h264 on the dsp, so better to focus on codecs that aren't optimised for the dsp first
  • [18:21:02] <dirk2> khasim: No, sorry. I mean why "additional patches"? Why not have everything in git like with sakoman's uboot git
  • [18:21:04] <ds2> mru: why?
  • [18:21:13] <koen> theora is broken by design
  • [18:21:29] <mru> yes, it pales in comparison to mpeg2 even
  • [18:22:09] <ds2> is there specifics or are you going by the assumption that it lost the MPEG battle for being the standard?
  • [18:22:12] <koen> see http://web.mit.edu/xiphmont/Public/theora/demo.html
  • [18:22:20] <khasim> Dirk: that is fine as well, but we need to know and show the delta clearly to all, if everything is integrated in GIT, others will not know the effort involved in getting them upstream
  • [18:22:56] <khasim> If it is GIT + patches in form of quilt atleast, then others can help in pushing the patches upstream
  • [18:23:08] <khasim> Dirk: what do you think about this?
  • [18:23:49] <koen> mru: which codecs benefit from your ffmpeg work when being played with mplayer?
  • [18:24:13] <dirk2> khasim: no patches please. git please. example is sakoman's git tree (two branches). You can then use git to create the patches if you like (you == user).
  • [18:24:41] <khasim> :)
  • [18:25:03] <mru> koen: anything that uses 8x8 dct and full- or half-pel motion compensation
  • [18:25:18] <mru> mpeg2 and mpeg4 both see good speedup
  • [18:25:32] <mru> h264 uses other transforms heavily
  • [18:25:34] <khasim> ok, let me start working towards porting DSS and other important stuff to latest GIT... will keep updating the patches
  • [18:26:08] <koen> mru: and mplayer will select the neon dct automagically?
  • [18:26:18] <mru> it should
  • [18:26:27] <suihkulokki> what's DSS ?
  • [18:26:27] <khasim> will drop off, thanks to all.... :)
  • [18:26:46] <dirk2> khasim: Read log if you are back ;)
  • [18:26:51] <dirk2> khasim: Bye
  • [18:26:54] * docelic (n=docelic@78.134.193.152) Quit ("http://www.spinlocksolutions.com/")
  • [18:27:06] <khasim> Display Sub System : FBDEV, V4L2 based Video driver, TV OUT, DVI, LCD, etc
  • [18:27:20] <khasim> dirk: sure...
  • [18:27:22] <suihkulokki> ah ofcourse
  • [18:28:19] <koen> mru: and ' -lavdopts idct=21' would force it?
  • [18:28:50] <mru> koen: yes
  • [18:28:56] <ds2> koen: thanks. that was an interesting read on theora
  • [18:29:59] <koen> mru: strange, idct=17 gives me faster results on the matrixbench avi (57 seconds vs 73)
  • [18:30:25] <koen> I could be doing something hooribly wrong, though
  • [18:30:26] <mru> koen: which file?
  • [18:30:35] <koen> http://samples.mplayerhq.hu/benchmark/testsuite1/matrixbench_normdivx_vbrmp3.avi
  • [18:32:05] <koen> I'm downloading the big buck bunny mp4 now
  • [18:32:14] <koen> http://download.blender.org/peach/bigbuckbunny_movies/big_buck_bunny_480p_surround-fix.avi
  • [18:32:30] <banderson> saw trialier for it..looked good
  • [18:32:41] <mru> it's kind of funny
  • [18:34:00] <mru> koen: care to do some oprofiling?
  • [18:34:38] <mru> that file decodes much quicker here with neon idct
  • [18:34:46] <dirk2> khasim: to summarize: what I like: Do small patches, often. No big tarballs. No big patches (see. e.g. Tony's coment about "Add Dspbridge driver" patch). Fix issues before adding new code (if possible). Use git (with branches). Example is how we work(ed) with sakoman's uboot git. Git is easy updatable against mainline. If user want patches, they can be extracted easily.
  • [18:35:48] <sakoman> dirk2: I agree 100%
  • [18:35:55] <koen> mru: building oprofile now
  • [18:36:45] <mru> maybe you didn't build neon support properly
  • [18:36:59] <sakoman> dirk2: I write my OE recipes to build against upstream git, with patches generated by git diff between test and master branches
  • [18:37:06] <koen> mru: that would be my guess
  • [18:37:25] <sakoman> makes it very simple to deal with upstream submission/acceptance process
  • [18:37:44] <dirk2> khasim: For git it doesn't matter where it is hosted. Preferred are already known and used repositories. E.g. uboot and omap git. Sometimes this is not possible because (a) slow upstream or (b) bad code quality
  • [18:38:14] <dirk2> khasim: (a) and (b) apply to uboot, therefore sakoman created a new "uboot beagle" repository
  • [18:38:27] <banderson> sakoman: So more news from the land of evm weirdness...So I compiled a uboot version to use uart3 and now I can run you uImage without issue.
  • [18:38:35] <dirk2> khasim: But: for kernel I'm not sure.
  • [18:39:10] <dirk2> khasim: tony is relatively fast (a) and not so strict (b)
  • [18:39:23] <sakoman> dirk2: I've set up a linux-omap git that I'm using for the asoc work
  • [18:39:36] <sakoman> also the pending nand stuff
  • [18:39:52] <dirk2> khasim: sakoman: And here is our kernel git :)
  • [18:40:05] <sakoman> tony isn't so fast for stuff that crosses into other specialties like asoc, nand, usb
  • [18:40:09] <jkridner|work> sakoman: when did you last update your u-boot? I want to check if my crontab for 'git pull' is working properly.
  • [18:40:20] <dirk2> As already discussed: I think everything we need is already there :)
  • [18:41:03] <sakoman> jkridner: last u-boot update was 6 days ago
  • [18:41:44] <sakoman> time for lunch -- back shortly. leave a note if you run into any issues
  • [18:42:20] <RogerMonk> me too - l8r guys
  • [18:46:10] <dirk2> khasim: proposal: send some small kernel patches (do you have some pending fixes unkown outside from TI?) to beagle list and let us see if we get them integrated somewhere. If then can see if we get them integrated in Tony's git or we have to maintain them some time beagle locally, e.g. in sakoman's kernel git.
  • [18:46:53] * khasim (n=a0393720@117.192.98.244) Quit (Read error: 110 (Connection timed out))
  • [18:47:53] <jkridner|work> sakoman has a kernel git now?
  • [18:48:26] <dirk2> jkridner|work: See above: "dirk2: I've set up a linux-omap git that I'm using for the asoc work"
  • [18:49:03] <dirk2> http://www.sakoman.net/cgi-bin/gitweb.cgi
  • [18:51:04] <jkridner|work> I was just in the process of cloning linux-omap-2.6.git from source.mvista.com. I'd like to make sure I have clones of x-load, u-boot, kernel, and website, at least. We'll have to hear from khasim where he will maintain his patches.
  • [18:52:32] <jkridner|work> I like having git that clone Tony's git and have branches, as long as khasim can get his patches in, even the ones that will need a lot of rework before submitting to Tony.
  • [18:52:52] <dirk2> jkridner|work: For what do you think the clones are needed?
  • [18:53:19] <dirk2> You know created the 3rd uboot and x-load repositoryclone?
  • [18:55:08] <dirk2> s/know/now/ s/repositoryclone/repository clone/
  • [18:56:41] * RogerMonk (n=a0740758@nat/ti/x-31b455e4ff66547b) Quit (Remote closed the connection)
  • [18:58:06] * jconnolly (n=jconnoll@ip-64-32-229-194.dsl.nyc.megapath.net) Quit (Read error: 104 (Connection reset by peer))
  • [19:01:36] <koen> mru: do I only need to copy over libavcodec and edit config.mak for a neon mplayer build ?
  • [19:02:25] * jconnolly (n=jconnoll@ip-64-32-229-194.dsl.nyc.megapath.net) has joined #beagle
  • [19:07:46] <koen> D'oh
  • [19:07:52] <koen> and config.h of course
  • [19:19:52] <sakoman> banderson: is that weird or what! I changed my standard builds to turn off the LL_DEBUG stuff to uart3.
  • [19:20:22] <sakoman> The guys at Mistral now say it works for them with no issues. The used to have the same problem you did.
  • [19:20:46] <banderson> sakoman: yes, very strange. So what did they do to fix it?
  • [19:20:55] <sakoman> They were using the same u-boot you were too :-) So I suspect something wonky in u-boot
  • [19:21:09] <banderson> yes, Looking at that now...
  • [19:21:33] <sakoman> They just switched to my new build
  • [19:21:39] <sakoman> no other changes
  • [19:21:55] <banderson> ahh..the one without LL_DEBUG
  • [19:22:01] <sakoman> It seems like a reasonable change with no downsides
  • [19:22:15] <banderson> ya...really curious as to the why
  • [19:22:43] <sakoman> same here!
  • [19:23:08] <sakoman> But right now I'm even more curious why video out on my beagle doesn't work ;-)
  • [19:29:53] * dirk2 (n=dirk@F30fd.f.strato-dslnet.de) has left #beagle
  • [19:34:06] * RogerMonk (n=a0740758@nat/ti/x-4bc3aa0ebd43ddd4) has joined #beagle
  • [19:53:15] * jkridner (n=jason@c-76-31-18-64.hsd1.tx.comcast.net) Quit (Read error: 110 (Connection timed out))
  • [20:03:42] <RogerMonk> koen - ping
  • [20:04:31] <koen> pong
  • [20:04:49] <RogerMonk> hi sir - have you ever tried building gnash with OE?
  • [20:05:11] <Crofton> heh
  • [20:05:12] <koen> I tried yesterday, for avr32
  • [20:05:16] <Crofton> I think there is a recipe :)
  • [20:05:23] <koen> gnash is horrendous
  • [20:05:27] * koen is in the swfdec camp
  • [20:05:35] <RogerMonk> a good test for OE :)
  • [20:05:43] <Crofton> heh
  • [20:06:02] <Crofton> we really should talk him into fixing up the autofoo
  • [20:06:12] <Crofton> We talked to the gnash guy at FOSDEM
  • [20:07:06] <RogerMonk> how does swfdec compare to gnash?
  • [20:07:22] <koen> better design, less hype
  • [20:07:34] <RogerMonk> does it claim support for youtube yet?
  • [20:07:46] <koen> it can play youtube
  • [20:07:55] <koen> but it's swfdec policy to not make claims :)
  • [20:09:09] <denix> koen, sakoman: omap3evm machine does not need those 3 SmartReflex patches - they are already part of linux-omap2 git
  • [20:10:15] <koen> ehm
  • [20:10:22] <koen> your OE checkout is too old
  • [20:10:27] <sakoman> denix: they are long gone in my oe branch -- I don't think koen has synced up with me recently
  • [20:10:45] <koen> SRC_URI_append_omap3evm = " file://no-harry-potter.diff;patch=1 \ file://0001-ASoC-OMAP-Add-basic-support-for-OMAP34xx-in-McBSP.patch;patch=1 \ file://0001-omap3-cpuidle.patch;patch=1 \ file://0002-omap3-cpuidle.patch;patch=1 \
  • [20:10:46] <sakoman> ah, or perhaps *you* haven't synced up recently ;-)
  • [20:10:49] <koen> "
  • [20:11:22] <denix> I'm at org.openembedded.dev/master...
  • [20:12:04] <koen> http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=commitdiff;h=058646e4b3e92f736926cd1cf18c5066fb8e9a31
  • [20:19:29] * jconnolly (n=jconnoll@ip-64-32-229-194.dsl.nyc.megapath.net) Quit ("Ex-Chat")
  • [20:23:37] <denix> oops, sorry for making noise - my checkout was definitely out of sync and in some weird state. fixed it now. thanks
  • [20:26:22] <sakoman> denix: no problem!
  • [20:27:51] * jconnolly (n=jconnoll@ip-64-32-229-194.dsl.nyc.megapath.net) has joined #beagle
  • [20:28:47] * jconnolly (n=jconnoll@ip-64-32-229-194.dsl.nyc.megapath.net) Quit (Client Quit)
  • [20:30:31] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [20:30:39] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Nick collision from services.)
  • [20:30:41] * Olipro_ is now known as Olipro
  • [20:42:17] <RogerMonk> sakoman - how are you controlling the UI on your EVM demo image? do you have keyboard control somehow?
  • [20:43:10] <RogerMonk> koen - have you tried swfdec on beagle?
  • [20:43:32] <sakoman> I use the touchscreen w/ onscreen keyboard most of the time
  • [20:43:46] <RogerMonk> how do I pull up the on-screen keyboard?
  • [20:43:47] <sakoman> You can also connect USB devices through a hub
  • [20:43:57] <sakoman> it should just pop up at boot
  • [20:45:06] <RogerMonk> I get the calib screen, then goes to the desktop with a few folders at the top - any way to pull up the keyboard from there?
  • [20:45:34] <sakoman> do you see the keyboard icon in the bar at the bottom of the screen?
  • [20:45:39] <sakoman> tap that
  • [20:47:37] <koen> RogerMonk: not yet
  • [20:49:49] <RogerMonk> sakoman -- hmm ok - I look tomorrow when I reboot... - I'm not currently using your kernel, since I needed nfs root mount support, which your kernel didn't seem to like - have you tried that at your end?
  • [20:51:30] <sakoman> RogerMonk: haven't tried nfs root, I don't think it is enabled in my kernel. I've been using nand rootfs, occasionally mmc rootfs
  • [20:52:18] <sakoman> Just adding some fixes for touchscreen jitter
  • [20:52:28] <sakoman> it's been pretty bad up to now
  • [20:54:54] <RogerMonk> any chance you could build me a kernel with it enabled :)?
  • [20:55:00] <RogerMonk> (please)
  • [20:55:18] <RogerMonk> + yep, the ts jitter fix would be great!
  • [20:56:03] * BThompson (n=BThompso@nat/ti/x-e96d1c833bd882aa) Quit (Remote closed the connection)
  • [20:56:31] * rsalveti (n=salveti@200.184.118.132) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:56:31] * bernard_ (n=bernard@mersenne.largestprime.net) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:56:31] * denix (n=denys@nat/ti/x-bc75499cc96fe88e) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:56:31] * sakoman (n=sakoman@static-74-41-60-154.dsl1.pco.ca.frontiernet.net) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:56:31] * KeyserSoze (n=Miles@67.107.206.170.ptr.us.xo.net) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:56:31] * ldesnogu (n=ldesnogu@fw-tnat.cambridge.arm.com) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:57:07] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
  • [20:57:07] * bernard_ (n=bernard@mersenne.largestprime.net) has joined #beagle
  • [20:57:07] * sakoman (n=sakoman@static-74-41-60-154.dsl1.pco.ca.frontiernet.net) has joined #beagle
  • [20:57:07] * KeyserSoze (n=Miles@67.107.206.170.ptr.us.xo.net) has joined #beagle
  • [20:57:07] * ldesnogu (n=ldesnogu@fw-tnat.cambridge.arm.com) has joined #beagle
  • [20:57:21] * denix (n=denys@nat/ti/x-a7216ca9abac8a21) has joined #beagle
  • [21:50:24] * rsalveti (n=salveti@200.184.118.132) Quit (Read error: 113 (No route to host))
  • [21:54:26] <banderson> sakoman: do we have a way of knowing what is different between your uboot version and the one in 0.9.5 SDK?
  • [21:55:34] <denix> can you diff them?
  • [21:56:03] <banderson> Not sure if we have source..or can get source
  • [21:56:12] <denix> which one?
  • [21:56:25] <denix> you have sources in SDK though
  • [21:57:01] <banderson> sakoman has a uboot version on his board that is like 3 days newer then the provided u-boot in 0.9.5 SDK
  • [21:57:10] <banderson> err I meant older
  • [21:57:54] <denix> Mar 11?
  • [21:58:23] <denix> or Mar 07?
  • [21:58:31] <banderson> Would have to look back through irc log to be sure but though his was the 7 or 9
  • [21:58:45] <banderson> could be
  • [22:00:20] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
  • [22:00:36] <denix> 0.9.5 - U-Boot 1.1.4 (Mar 11 2008 - 18:01:24)
  • [22:00:49] <denix> 0.9.4 - U-Boot 1.1.4 (Mar 11 2008 - 03:32:16)
  • [22:01:01] <denix> 0.9.3 - U-Boot 1.1.4 (Mar 7 2008 - 23:16:06)
  • [22:01:13] <denix> 0.9.1 - U-Boot 1.1.4 (Jan 24 2008 - 18:22:01)
  • [22:01:32] <banderson> It would be the Mar 7th one.
  • [22:01:33] <denix> make your pick :)
  • [22:01:59] <banderson> diff between 0.9.3 and 0.9.5
  • [22:02:02] <denix> 0.9.5 was the first public release, but I can run diffs against older ones for you
  • [22:02:20] <banderson> that would be great
  • [22:02:37] <sakoman> U-Boot 1.1.4 (Mar 7 2008 - 16:57:15)
  • [22:03:56] <sakoman> banderson: have you ever gotten to the point of running the omap3evm-demo-image?
  • [22:06:42] <denix> hmm, I see lots of temp files polluting 0.9.3 codebase, since it was not publicly released. do you want me to clean it and only diff the actual sources files?
  • [22:10:49] <sakoman> denix: yeah, just the important diffs
  • [22:11:25] <denix> ok, not too many changes, nothing important AFAICS - http://pastebin.org/48045
  • [22:12:20] <banderson> sakoman: I can run your uImage-2.6.25+2.6.26-rc8+r32+git050684c18f2ea0b08fdd5233a0cd3c7f96e00a0e-r32-omap3evm.bin
  • [22:12:47] <banderson> sakoman: If i move uboot to use uart3 and only do cold reboots
  • [22:13:11] <denix> eth port had a bug in 0.9.3, that was fixed in 0.9.5 - PWR_MGMT_PM_MODE_MSK was wrong
  • [22:13:38] <banderson> I am assuming your latest uimage has ll_debug for uart3 disabled
  • [22:14:11] <sakoman> banderson: one of the mistral guys can use my uimage unchanged. yes debug_ll is off
  • [22:14:12] <banderson> (which would work...at least my uImage compiles have worked that way)
  • [22:15:01] <banderson> do you know what verion of uboot is running on the one that can use uImage unchanged...
  • [22:15:06] <sakoman> he tried the omap3evm-demo-image.jffs2 and says it works, but that the spash screen has a red cast to it
  • [22:15:18] <sakoman> mine is grey as it should be
  • [22:15:34] <banderson> plus I am assuming we are talking older uImages (before removal of debug_ll for uart3)
  • [22:16:03] <banderson> Hey that sounds familiar!!!
  • [22:16:07] <banderson> I had that issue as well
  • [22:16:16] <RogerMonk> sakoman - your bz2 image worked fine for me - grey splash screen
  • [22:17:04] <sakoman> banderson: he is now using: U-Boot 1.1.4 (Jun 5 2008 - 17:53:37)
  • [22:17:21] <sakoman> don't know what that is! :-)
  • [22:17:59] <sakoman> RogerMonk: is your u-boot March 11?
  • [22:18:07] <banderson> another interesting twist to the red tint issue is that if I used orig ti uImage then restarted using the oe (yours or mine) it would work fine the first time....then if I restarted again it would be red ...until I ran ti uImage
  • [22:18:30] <RogerMonk> no June 5 - i'm using 0.9.8
  • [22:18:34] <RogerMonk> :)
  • [22:19:13] <banderson> strange...maybe uboot isn't issue and just ?
  • [22:19:26] <koen> sakoman: ever tried to run uboot git on the evm?
  • [22:19:58] <sakoman> koen: no, that would require me to do a port
  • [22:20:10] <sakoman> it is on the to do list
  • [22:20:50] <sakoman> nand vs onenand is probably the major change required
  • [22:21:09] <sakoman> but I am sure there are likely a few other little things
  • [22:21:27] <sakoman> like ethernet support (nice but not required)
  • [22:22:07] <denix> 0.9.8 - U-Boot 1.1.4 (Jun 5 2008 - 17:53:37)
  • [22:23:34] <sakoman> denix: any theories on the red tint issue?
  • [22:24:10] <sakoman> I suppose I should bite the bullet and upgrade my u-boot
  • [22:24:28] <sakoman> see if I begin experiencing any of these issues
  • [22:24:42] <denix> sakoman: 0.9.8 has micron bad block management, NAND subsystem is little bit different
  • [22:25:28] <sakoman> denix: my evm uses onenand
  • [22:25:56] <sakoman> does 0.9.8 have dual nand/onenand support?
  • [22:27:50] <denix> sakoman: yes
  • [22:28:00] <denix> 0.9.7 had onenand bad block management
  • [22:28:44] <sakoman> I thought onenand was supposed to hide bad blocks!
  • [22:28:53] <Crofton|work> RogerMonk, still working in the davinci images
  • [22:29:08] <Crofton|work> I messed up the dsplink bb file and lost the sample files
  • [22:29:18] <Crofton|work> I'll fix in the morning
  • [22:29:23] <Crofton|work> and finish images then
  • [22:29:53] <Crofton|work> I should start to free up some time from slumlording in a day or so
  • [22:33:38] <RogerMonk> Crofton - thanks - job for me tomorrow is to be able to build this stuff myself and stop burdening you with the updates! :)
  • [22:34:30] * bazbell (n=a0192809@nat/ti/x-2ae0c95f65c0b289) Quit (Remote closed the connection)
  • [22:36:41] <denix> sakoman: sorry for the delay, got interrupted here... onenand does not hide bad blocks
  • [22:39:35] <denix> sakoman: basically, bad block management in u-boot is not that smart - it just skips the block when writing/reading, but does not take into account partition boundaries
  • [22:45:15] <sakoman> denix: yes, I noticed that the bad black management in u-boot is fairly simplistic
  • [22:46:01] <sakoman> should be fine as long as we allow big enough partitions to handle sufficient bad blocks
  • [22:47:38] <denix> sakoman: except for the u-boot environment space, which is 2 blocks only and if both go bad, you are out of luck :)
  • [22:48:27] <sakoman> definitely -- therefore crucial to have a really good default u-boot environment :-)
  • [22:48:50] <denix> btw, besides those changes in 0.9.7 and 0.9.8 u-boot was synced up with wtbu...
  • [22:52:06] * rsalveti (n=salveti@200.184.118.132) Quit (Read error: 104 (Connection reset by peer))
  • [22:52:16] <sakoman> I really should just go ahead and do the 1.3.3 port for the evm
  • [22:53:46] <denix> I believe someone in TI is working on it...
  • [23:05:07] <denix> good night everyone
  • [23:05:54] * denix is away
  • [23:08:16] * NishanthMenon (n=nmenon@nat/ti/x-d17da9f9ab031f66) Quit ("Ex-Chat")
  • [23:18:48] * rsalveti (n=salveti@189.70.97.120) has joined #beagle
  • [23:20:24] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
  • [23:22:21] * bazbell (n=a0192809@nat/ti/x-77a2585946b18525) has joined #beagle
  • [23:23:27] <Crofton> RogerMonk, let us know if you need help
  • [23:23:46] <RogerMonk> thanks sir
  • [23:23:53] <Crofton> I need to get the update in to opackge the sample programs as it will fail atm
  • [23:27:39] <Crofton|work> I'd fix it now, but I need to go play darts