• [00:18:31] * RogerMon1 (n=a0740758@nat/ti/x-a343ff9921460451) Quit (Remote closed the connection)
  • [00:21:58] * codyl (n=cody@adsl-71-142-83-140.dsl.pltn13.pacbell.net) has joined #beagle
  • [00:23:20] <codyl> koen, ping
  • [00:33:50] <codyl> What license is the design under?
  • [00:43:47] <ali_as> He went to sleep a while ago.
  • [00:44:15] <ali_as> About 4 hours.
  • [01:55:01] * codyl (n=cody@adsl-71-142-83-140.dsl.pltn13.pacbell.net) Quit (Read error: 110 (Connection timed out))
  • [02:10:02] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [02:28:30] * darkerror05 (n=darkerro@cpe-74-69-7-154.rochester.res.rr.com) has joined #beagle
  • [02:29:12] <darkerror05> where is the best place to order a board?
  • [02:43:50] <BThompson> beagleboard.org when they are available
  • [02:46:25] <darkerror05> Estimated timeline?
  • [02:53:12] <BThompson> I am not sure when the estimate is for availability on the site, jkridner could probably respond best to that, there are some early release boards around, but there are some usb issues being worked out before they are widely available
  • [03:07:48] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) Quit ("Trillian (http://www.ceruleanstudios.com")
  • [03:17:08] * bazbell (n=a0192809@nat/ti/x-8c63acd87ae003e2) Quit ("Leaving.")
  • [03:26:32] * jkridner|work1 (n=a0321898@nat/ti/x-39f63a877ab72bfa) has joined #beagle
  • [03:30:42] * jkridner|work (n=a0321898@nat/ti/x-0e843c8c296f0a9a) Quit (Remote closed the connection)
  • [03:43:12] * DaQatz (i=db@c-66-30-48-54.hsd1.nh.comcast.net) has joined #beagle
  • [04:03:47] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [04:59:02] * j24 (n=j24@ns32433.ovh.net) has joined #beagle
  • [04:59:04] <j24> hi
  • [04:59:59] <j24> is there any one from beagle support here?
  • [05:13:49] * DaQatz (i=db@c-66-30-48-54.hsd1.nh.comcast.net) Quit (Read error: 110 (Connection timed out))
  • [05:22:07] * jkridner|work (n=a0321898@nat/ti/x-ea3cbd92d7929c54) has joined #beagle
  • [05:26:13] * jkridner|work1 (n=a0321898@nat/ti/x-39f63a877ab72bfa) Quit (Remote closed the connection)
  • [05:32:48] <jkridner> j24, yes.
  • [05:33:18] * jkridner changes topic to 'Welcome to #Beagle | Discussion about the OMAP3 Beagle Board - http://beagleboard.org | Beagle search tools are on #dashboard at irc.gimp.org, NOT here ;) | Log is at http://www.beagleboard.org/irclogs'
  • [05:33:34] <sakoman> jkridner: you are up late!
  • [05:33:43] <jkridner> just got back from a show.
  • [05:33:56] <jkridner> The Cure.
  • [05:34:09] <sakoman> nice!
  • [05:34:13] <jkridner> won't be up long. :)
  • [05:34:36] <sakoman> my cat is telling me it is time to go to sleep
  • [05:34:53] <sakoman> she sits next to my keyboard and stares at me
  • [05:35:31] <jkridner> :)
  • [05:35:47] <sakoman> doing a little u-boot housekeeping
  • [05:36:10] <jkridner> an ARM u-boot maintainer contacted me about a board...
  • [05:36:16] <jkridner> guess they noticed us.
  • [05:36:35] <jkridner> listen to your cat. :) good night.
  • [05:37:16] <sakoman> good night1
  • [05:38:50] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [05:41:12] <sakoman> good morning khasim!
  • [05:58:01] <khasim> hello sakoman
  • [05:59:29] <khasim> sakoman, I think you will become full time maintainer for u-boot Beagle... I hope you don't mind ...
  • [06:09:12] <khasim> sakoman: are you there?
  • [06:20:58] <ds2> who is the full time maintainer for OMAP3 Uboot?
  • [06:24:29] <khasim> OMAP had no u-boot mainitainer
  • [06:25:26] <khasim> sakoman: I am not getting which branch of yours is targetted for upstream submission
  • [06:25:33] <khasim> is it the origin/test ?
  • [06:26:08] <khasim> I see lot of differences when I take a diff between test and the latest u-boot.
  • [06:41:28] <khasim> my problem, the uboot git I had pulled was not updated properly, but it used to show "Already updated"
  • [06:41:34] <khasim> pulling the u-boot GIT again
  • [07:27:51] * hagisbasheruk (n=hagisbas@78.148.135.75) has joined #beagle
  • [07:34:06] * likewise (n=Leon_Woe@atwork-186.r-212.178.104.atwork.nl) has joined #beagle
  • [07:58:45] <khasim> Use of volatile is usually wrong: see Documentation/volatile-considered-harmful.txt
  • [07:58:45] <khasim> #1499: FILE: board/omap3530beagle/omap3530beagle.c:66:
  • [07:58:45] <khasim> + __asm__ volatile ("1:\n" "subs %0, %1, #1\n"
  • [07:59:13] <khasim> Running checkpatch on beagle u-boot patches gives lot of these errors.
  • [07:59:27] <khasim> dont know if this gets accepted as is on u-boot git
  • [09:31:38] * RogerMonk (n=a0740758@nat/ti/x-d021757f07a9782c) has joined #beagle
  • [09:36:05] <ldesnogu> don't know if anyone noticed: http://linuxdevices.com/news/NS5852740920.html
  • [09:42:21] * jkridner|work (n=a0321898@nat/ti/x-ea3cbd92d7929c54) Quit (Remote closed the connection)
  • [09:42:54] * jkridner|work (n=a0321898@nat/ti/x-2045e20a841c60c2) has joined #beagle
  • [10:42:36] <keesj> hurry up with those boards! are you really wokring on a 4 layer design for rev B?
  • [10:47:34] * hagisbasheruk (n=hagisbas@78.148.135.75) Quit (Remote closed the connection)
  • [10:57:23] * ferret_0567 (n=ferret_0@unaffiliated/ferret0567/x-681468) Quit (Remote closed the connection)
  • [10:57:27] * ferret_0567 (n=ferret_0@unaffiliated/ferret0567/x-681468) has joined #beagle
  • [11:44:45] * jkridner|work (n=a0321898@nat/ti/x-2045e20a841c60c2) Quit (Remote closed the connection)
  • [11:44:48] * jkridner|work1 (n=a0321898@nat/ti/x-03f8dce41ebe30f7) has joined #beagle
  • [11:58:24] <jkridner> Rev. B is a 6 layer design.
  • [11:58:56] <jkridner> hmmm.... ldesnogu, thanks for the headsup.
  • [12:03:59] <jkridner> I wasn't quite ready for a LinuxDevices blast.
  • [12:04:22] <jkridner> didn't want that to happen until our distributor was stocked.
  • [12:04:36] <jkridner> I guess this is someone's idea to turn up the heat.
  • [12:05:18] <Crofton> I didn't recognize the name at the bottom
  • [12:06:14] <jkridner> me either.
  • [12:18:27] <ldesnogu> don't know him either
  • [12:19:25] <Crofton|work> well, I guess you need to get ready for the rush
  • [12:19:53] <ldesnogu> hasn't the rush already started? :)
  • [12:20:27] <Crofton|work> well, this is just the word of mouth rush
  • [12:43:21] <ali_as> Ok, so someones clearly not aware how unready this is for mass consumption!
  • [12:45:04] <ali_as> Jason, given the circumstances would you mind if I communicated the existance of the beagle in a more balenced way to the open risc os people?
  • [12:45:12] <ali_as> http://www.riscosopen.co.uk/
  • [12:45:48] <sakoman> khasim: sorry I missed you last night
  • [12:46:20] <jkridner> ali_as: well, it isn't a secret.
  • [12:46:27] <sakoman> there is not yet a branch targeted for upstream submission since we still have a lot of cleanup to do
  • [12:46:35] <jkridner> speak nicely. :)
  • [12:47:10] <ldesnogu> ali_as: I didn't know RISC OS had been opened!
  • [12:47:14] <sakoman> Dirk and I have been working in the origin/test branch
  • [12:47:25] <ali_as> Difference between a secret though and not painting a picture of it on the moon ;)
  • [12:49:12] <ldesnogu> I someone ports RISC OS to BeagleBoard, I'm sure some old timers here will *need* a Beagle :)
  • [12:50:16] <ali_as> It's the most promising hardware for a port for some time, and you can't argue with the price/performance.
  • [12:52:11] <ldesnogu> is RISC OS using the EABI?
  • [12:52:21] <jkridner> well, I'd certainly like to stop short of putting a picture on the moon right now.
  • [12:52:56] <jkridner> once there are 2 things: stock at the distributor and an easy-to-user boot/distro solution...
  • [12:53:04] <jkridner> then, I'll want everyone in the world to know. :)
  • [12:53:08] * BThompson (n=BThompso@nat/ti/x-aab94aa90eae3039) has joined #beagle
  • [12:53:52] <kulve> there seems to be many people interested in a hw that you can build stuff on. They don't need a complete distro for that
  • [12:54:17] <jkridner> sure, but I don't want those people that need a distro to run for the hills...
  • [12:54:33] <jkridner> if I could find a way to only talk to those who are looking to build, then great.
  • [12:54:40] <jkridner> those are the people I want to talk to today.
  • [12:54:51] <jkridner> but, I don't expect Beagle to stop there...
  • [12:55:17] <kulve> when advertise, just emphasize that there's partly working kernel and no distro..
  • [12:55:21] <jkridner> it should be very easy to manage ease-of-use with Beagle such that non-techie types can even do code development.
  • [12:55:33] <jkridner> for example, have you ever seen Alice, the 3D programming language?
  • [12:55:49] <kulve> nope..
  • [12:55:50] <jkridner> well, there is a distro (Angstrom)...
  • [12:56:07] <ali_as> I don't think the riscos people are the moon, I think that's been done allready with the linuxdevices page, but I know there are a few people champing at the bit for A8 hardware and I'd hate to see the project eclipsed by the morouding hoards from the linux devices rush.
  • [12:56:08] <jkridner> and Ubuntu for ARM and Fedora for ARM should run readily....
  • [12:56:16] <Crofton|work> what we need to do is set up a branch in the meta-data so you can build packages
  • [12:56:32] <Crofton|work> sort of like what sakoman did for gumstix
  • [12:56:34] <jkridner> but, someone needs to make it a really easy process to run these distros.
  • [12:56:51] <kulve> jkridner: so, there's already 1 of the two things you wanted to see before telling the world?
  • [12:56:53] <Crofton|work> so you have a binary package repo
  • [12:57:17] <jkridner> I'm looking for even easier.
  • [12:57:17] <Crofton|work> I am hopeful that git works as advertised, and this is easier to do
  • [12:57:36] * cavok (n=cavok@host65-140-static.28-87-b.business.telecomitalia.it) has joined #beagle
  • [12:57:40] <jkridner> I'm looking for a PC app that can program the SD Card and/or boot directly over USB.
  • [12:57:47] <Crofton|work> heh
  • [12:58:02] <Crofton|work> be careful not to get hung up over being perfect :)
  • [12:58:05] <jkridner> something that doesn't require a user to even know Linux before they get started.
  • [12:58:25] <jkridner> well, I won't let not being perfect hold everything up.
  • [12:58:39] <Crofton|work> jkridner, I think you need a larger community for that to happen
  • [12:58:47] <Crofton|work> unless you plan to do everythign yourself
  • [12:59:40] <kulve> jkridner: I don't see any problems telling the world about the beagle even if it didn't have a easy to boot distro as many can boot also through the harder ways. But one just must advertise what it is now and make clear statement that it's not ready for non-techie end users yet
  • [12:59:46] <Crofton|work> it looks like you could ship a fairly full features linux desktop today
  • [12:59:47] <jkridner> yes. I want a larger community before plastering Beagle on the moon.
  • [13:00:05] <Crofton|work> so you need to be able to sell boards :)
  • [13:00:07] <jkridner> I really *need* to talk to low-level developers now, which is why I'm not keeping much secret about Beagle.
  • [13:00:29] <jkridner> hey, no mistaking that priority.
  • [13:00:37] <jkridner> building boards is priority #1.
  • [13:00:52] <kulve> working kernel is #2?
  • [13:01:12] <jkridner> kulve: I'll be counting a lot on people here to help keep it straight what Beagle can and cannot do in our communication to the world.
  • [13:02:02] <jkridner> kulve: working kernel is already pretty far along. still a few snags, but it is getting there. great power management will lag a big behind.
  • [13:02:17] * jkridner needs to head into the office for a fun-filled day of meetings.
  • [13:02:21] <kulve> power management seems to be always laggind behind..
  • [13:09:21] * jkridner|work1 (n=a0321898@nat/ti/x-03f8dce41ebe30f7) Quit ("Leaving.")
  • [13:19:50] <kulve> mru: what kind of hw were you planning to use with the X.Org driver?
  • [13:22:28] * prpplague (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [13:34:11] * bazbell (n=a0192809@nat/ti/x-f6a69bbd9e7dad7c) has joined #beagle
  • [13:51:33] <khasim> Ah,,, I missed good discussion between crofton and Jason....
  • [13:56:33] <sakoman> khasim: did you see my earlier comments about which u-boot branch to use?
  • [13:57:14] <sakoman> dirk has been doing good work on cleaning things up for upstream submission
  • [13:57:47] <sakoman> he's got experience in u-boot submissions, so it is great to have his help
  • [14:00:46] * NishanthMenon (n=nmenon@nat/ti/x-6c2d67c88292b287) has joined #beagle
  • [14:01:17] * jkridner|work (n=a0321898@nat/ti/x-0fbdd4448269bdbb) has joined #beagle
  • [14:06:05] * robclark (n=robclark@nat/ti/x-88cb553c606a42dd) has joined #beagle
  • [14:10:10] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [14:10:51] * cavok (n=cavok@host65-140-static.28-87-b.business.telecomitalia.it) has left #beagle
  • [14:22:20] * jkridner|work (n=a0321898@nat/ti/x-0fbdd4448269bdbb) Quit (Remote closed the connection)
  • [14:26:10] * gkarabin (n=gkarabin@64.212.175.249) has joined #beagle
  • [14:30:14] * dirk2 (n=dirk@F3089.f.strato-dslnet.de) has joined #beagle
  • [14:44:12] * gkarabi1 (n=gkarabin@64.212.175.249) Quit (Read error: 110 (Connection timed out))
  • [14:53:21] * jkridner|work (n=a0321898@nat/ti/x-50659f63475abf89) has joined #beagle
  • [15:03:17] * oelewapperke (i=d64aae46@85-158-215-1.powered-by.benesol.be) Quit (kubrick.freenode.net irc.freenode.net)
  • [15:03:17] * Crofton|work (n=balister@66-207-66-26.black.dmt.ntelos.net) Quit (kubrick.freenode.net irc.freenode.net)
  • [15:05:28] <prpplague> koen: ping
  • [15:05:28] * oelewapperke (i=d64aae46@85-158-215-1.powered-by.benesol.be) has joined #beagle
  • [15:05:28] * Crofton|work (n=balister@66-207-66-26.black.dmt.ntelos.net) has joined #beagle
  • [15:11:37] * codyl (n=cody@adsl-71-142-83-140.dsl.pltn13.pacbell.net) has joined #beagle
  • [15:12:52] * ali_as (n=as@ambix.plus.com) Quit ("Okaay.")
  • [15:15:30] * ali_as (n=ali_as@ambix.plus.com) has joined #beagle
  • [15:33:31] * BThompson (n=BThompso@nat/ti/x-aab94aa90eae3039) Quit ("Trillian (http://www.ceruleanstudios.com")
  • [15:36:43] <jkridner|work> http://www.us.design-reuse.com/news/18429/aac-mp3-mpeg-4-h-264-fft-openmax-cortex-a8-neon-arm11-processors.html
  • [15:36:53] <jkridner|work> Cortex-A8 libraries.
  • [15:37:03] <ldesnogu> This is the library I mentionned a few weeks ago
  • [15:37:13] <ldesnogu> the EULA looks not friendly :(
  • [15:37:30] <jkridner|work> ah. I'd seen AAC and H.264 before, MP3 and MPEG4 look new to me.
  • [15:37:46] <ldesnogu> the release date didn't change I think
  • [15:37:50] <jkridner|work> too bad about the EULA. shouldn't you do something about that? ;)
  • [15:38:00] <ldesnogu> I am supposed to yes...
  • [15:38:30] <ldesnogu> finding the right person to talk to is difficult, I guess you know that :)
  • [15:38:33] <jkridner|work> sorry to tease. you know I have an appreciation for the difficulty.
  • [15:38:41] <ldesnogu> hehe
  • [15:39:03] <codyl> jkridner|work, what license will the design be released under?
  • [15:39:17] <jkridner|work> did I mention the free C6000 compiler recently? ;)
  • [15:39:29] <Crofton> free in "quotes" :)
  • [15:39:38] <jkridner|work> the schematic and layout files will be completely free for use.
  • [15:39:53] <jkridner|work> Gerald is only holding onto the design because he's not comfortable with the quality yet.
  • [15:40:07] <codyl> jkridner|work, GPL, creative commons, public domain ?
  • [15:40:25] <jkridner|work> you can see PDF of the layout and schematic in the HW ref manual now.
  • [15:40:38] <codyl> yes, I already have (its missing the BOM btw)
  • [15:40:40] <jkridner|work> layout and schematic will be either creative commons or public domain.
  • [15:41:07] <codyl> jkridner|work, awesome- do you know when that might happen?
  • [15:42:54] <jkridner|work> for everybody, it'll probably be November. for individuals who understand the limitations with the current design and are willing to take them into consideration, we can provide now.
  • [15:44:06] <jkridner|work> 0.4mm ball pitch can be really tough to work with. I'd think most people are better off trying to use the LogicPD SOM, a Beagle, or use the 0.65mm ball-pitch package.
  • [15:46:07] * DaQatz (n=db@c-66-30-48-54.hsd1.nh.comcast.net) has joined #beagle
  • [15:58:17] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [16:03:55] <khasim> I owe some quick images for Gerald to get Beagle production going, I will be working on these tomorrow, I have to align on OMAP3 EVM patches to GIT internally
  • [16:04:15] <khasim> As soon as I am done with these, I will work on clock and mux cleanup
  • [16:04:45] <khasim> By end of this week we should have a good code base for submission.
  • [16:05:14] <khasim> We should also support Nishanth Menon for u-boot V2 work, he is getting alone I believe...
  • [16:06:33] * BThompson (n=BThompso@nat/ti/x-379e1e68fc9bc377) has joined #beagle
  • [16:10:34] * RogerMonk (n=a0740758@nat/ti/x-d021757f07a9782c) Quit (Remote closed the connection)
  • [16:12:43] <NishanthMenon> khasim, thanks.. lone ranger = me ;)
  • [16:15:42] <Crofton> I think he meant "getting along" ?
  • [16:16:11] <jkridner|work> or "going alone"
  • [16:17:33] <Crofton> the bleeding edge is always lonely
  • [16:17:39] <NishanthMenon> yeah i know :)
  • [16:17:57] <NishanthMenon> but it is fun :).. think about doing the i2c core driver
  • [16:17:59] <NishanthMenon> :D
  • [16:18:55] <NishanthMenon> and discovering that oh.. i dont have bus architecture anymore.. I need to write the i2c core for the new device architecture in u-boot v2, + hmmm.. i2c-dev does not actually fit normally.. oh yeah.. i need to port lm_sensor-i2c tools to u-boot v2 ;)
  • [16:19:33] <NishanthMenon> but the returns are slow :(...
  • [16:20:14] * NishanthMenon thinking how my wife would comment on me being alone ;)
  • [16:21:11] * cody_ (n=cody@adsl-76-200-146-114.dsl.pltn13.sbcglobal.net) has joined #beagle
  • [16:24:20] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [16:25:20] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
  • [16:32:22] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [16:36:18] * hagisbasheruk (n=hagisbas@78.148.135.75) has joined #beagle
  • [16:37:28] * codyl (n=cody@adsl-71-142-83-140.dsl.pltn13.pacbell.net) Quit (Read error: 110 (Connection timed out))
  • [16:38:49] * ldesnogu_ (n=ldesnogu@ven06-2-82-247-86-183.fbx.proxad.net) has joined #beagle
  • [16:40:50] * like2wise (n=chatzill@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [16:44:01] <hagisbasheruk> http://linuxdevices.com/news/NS5852740920.html nice :)
  • [16:49:15] <sakoman> NishanthMenon: ping!
  • [16:49:24] <NishanthMenon> sakoman, pong
  • [16:49:57] <sakoman> Would you be willing to take a look at a register dump of the twl 4030 audio section?
  • [16:50:19] <sakoman> I'm having trouble getting sound out of both channels :-)
  • [16:50:29] <NishanthMenon> sakoman, been a while since i looked at audio :(
  • [16:50:38] <sakoman> ah, OK
  • [16:51:01] <sakoman> Right channel is perfect, volume controls work
  • [16:51:16] <sakoman> Left channel is basically silent
  • [16:51:19] <NishanthMenon> did you check the gain regs?
  • [16:51:42] <NishanthMenon> there used to be a analog and digitial gain reg.. if i recollect
  • [16:51:43] <sakoman> yes, the very first thing I looked at :-)
  • [16:51:52] <sakoman> yup, one of each
  • [16:52:01] <sakoman> well actually two of each :-)
  • [16:52:15] <NishanthMenon> yeah... i used to goof out there
  • [16:52:17] <NishanthMenon> this is one place
  • [16:52:21] <NishanthMenon> second place could be i2s
  • [16:52:28] <NishanthMenon> you could be sending 0s on left
  • [16:52:30] <sakoman> better look again, you know how sometimes you can look at dsomething too often and not see the obvious stuff
  • [16:52:55] <NishanthMenon> hmmm..
  • [16:53:04] <sakoman> No, PCM data is perfect. Data on both channels, proper clock edges changes, etc
  • [16:53:11] <NishanthMenon> PCM?
  • [16:53:16] <NishanthMenon> not i2s?
  • [16:53:17] <sakoman> I2s
  • [16:53:26] <sakoman> sorry
  • [16:53:27] <NishanthMenon> ok.
  • [16:53:49] <NishanthMenon> just being lazy.. but any place i can pull the code? wondering if it compiles for SDP/beagle
  • [16:53:54] <NishanthMenon> the 2 boards i have..
  • [16:54:36] <sakoman> That's next on my list. Currently only have a EVM machine driver (cause it is easier to probe signals)
  • [16:54:58] <sakoman> Beagle will be trivial after I get this working
  • [16:55:18] <sakoman> I'll set up a git repo for it
  • [16:56:40] * NishanthMenon looking at omap2-audio-twl4030.c from TI kernel
  • [16:57:59] <sakoman> I've looked through that code quite a bit :-) Very complex!
  • [16:58:14] <NishanthMenon> sakoman, hoping you have setup the BIT_AVDAC_CTL_ADACL2_EN_M and BIT_AVDAC_CTL_ADACR2_EN_M (left and right DACs) also BIT_OPTION_ARXR2_EN_M and BIT_OPTION_ARXL2_EN_M
  • [16:58:35] <NishanthMenon> look at twl4030_enable_output
  • [16:58:45] <sakoman> think so, let me check again
  • [16:59:41] <NishanthMenon> also hoping that you have configured it for the right option .. drat.. i need to revisit trm :(...
  • [17:00:00] <NishanthMenon> i remember it had two options and different audio filters enabled for each
  • [17:00:34] <sakoman> In CODEC_MODE register?
  • [17:01:16] <NishanthMenon> yeah CODEC_OPTION_1
  • [17:01:46] <NishanthMenon> that mode kind of killed voice path if i recollect.. but pretty good on audio path
  • [17:02:52] <sakoman> I've tried both. Currently using option 2, I'll switch back
  • [17:04:06] <NishanthMenon> do you have the reg dump?
  • [17:04:30] <sakoman> hmm . . . I'm using dac1, not dac2 could that be the issue?
  • [17:04:36] <sakoman> didn't think it would matter
  • [17:04:50] <NishanthMenon> i think u need both
  • [17:04:53] <NishanthMenon> for left and right
  • [17:05:27] <NishanthMenon> i think.. my old brain is trying to catch up with you ;)
  • [17:06:02] <sakoman> http://www.sakoman.net/omap3/twl4030-audio-regs.txt
  • [17:06:34] <sakoman> There are 2 stereo dacs. I've set up the routing for dac1
  • [17:07:12] <sakoman> And yes, I did enable ADACL1_EN and ADACR1_EN
  • [17:08:20] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [17:08:53] * NishanthMenon looking at trm
  • [17:10:53] <sakoman> doh, I think I found it
  • [17:10:59] <NishanthMenon> ok -> you need option 1 if you want tx stereo audio path
  • [17:11:33] <sakoman> The left channel enable bit enables voice in Option 2, not the left channel!
  • [17:11:43] <sakoman> Thanks!
  • [17:12:30] <NishanthMenon> ok cool
  • [17:12:32] <NishanthMenon> :)
  • [17:12:43] <sakoman> recompiling now
  • [17:12:57] <sakoman> Too damn many registers!
  • [17:13:06] <sakoman> To damn many modes!
  • [17:13:09] <sakoman> :-)
  • [17:13:15] <NishanthMenon> yeah.. twl4030 was a nightmare when i wrote it first time after coming from tsc2101 :)
  • [17:13:20] <NishanthMenon> but darn flexible
  • [17:13:43] <sakoman> yeah, totally. If I was still building phones I would love it.
  • [17:14:06] <sakoman> But it is way overkill for beagle
  • [17:14:22] <sakoman> amazing feature set though
  • [17:14:24] <NishanthMenon> yeah, but you get a bunch of devices in a single package :)
  • [17:15:39] <sakoman> Trying to do a nice simple alsa soc driver for it
  • [17:16:06] <sakoman> One that support the beagle & evm feature sets and hides the complexity
  • [17:16:08] <NishanthMenon> it is easier without speaker and digitial mics :)
  • [17:18:22] <sakoman> Thanks again for helping me look at those registers again with a fresh perspective
  • [17:18:45] * cody_ (n=cody@adsl-76-200-146-114.dsl.pltn13.sbcglobal.net) Quit ("Leaving")
  • [17:18:50] <NishanthMenon> sakoman, thanks for letting me refresh myself a bit ;)
  • [17:19:39] * NishanthMenon off for chow
  • [17:19:46] <sakoman> later!
  • [17:34:53] <koen> good morning all
  • [17:35:30] <jkridner|work> good morning koen.
  • [17:35:56] <jkridner|work> sakoman: looks like you are close to a breakthrough.
  • [17:41:49] <DJWillis> Evening all
  • [17:43:19] <sakoman> jkridner: fingers crossed. first try made it worse. the right channel disappeared too :-)
  • [17:44:34] <koen> prpplague: pong
  • [17:44:55] <prpplague> koen: nm, answer my own question, thanks anway
  • [17:45:59] <koen> jkridner|work: omap3 cpuidle support got submitted to linuxomap today :)_
  • [17:46:38] <DJWillis> koen: I noticed that, was going to play with it later, could be very cool.
  • [17:48:04] <ldesnogu_> koen, I can't remember the power usage you measured (and I couldn't find it in the logs)
  • [17:49:27] <jkridner|work> koen: sweet! do you know if it goes into retention mode?
  • [17:52:16] <koen> ldesnogu_: 0.345A, and sakoman measured it :)
  • [17:52:27] <ldesnogu_> oops
  • [17:52:29] <koen> jkridner|work: no idea, didn't look at the patches themselves yet
  • [17:55:14] <ldesnogu_> sakoman, your power measures were for all of the board and @5V, right? also IIRC that was at the prompt so the cpu load was low (for information, I am trying to convince x86 fanboys how low some chips can go while being usefule :)
  • [17:55:50] <sakoman> yes 5v, whole board, at console prompt
  • [17:56:05] <koen> so including audio and dvi
  • [17:56:21] <ldesnogu_> when you have time could try to push it by running the CPU at full load?
  • [17:56:47] <sakoman> sure
  • [17:56:55] <ldesnogu_> great, thanks :)
  • [17:57:17] <ldesnogu_> jkridner|work, the idle patch first mail is here: http://marc.info/?l=linux-omap&m=121308176604994&w=2
  • [17:58:32] <koen> mru: did you try adding NEON support to libavcodec yet?
  • [17:58:58] <koen> sakoman: did I understand correctly that you made some progress on ASoC?
  • [17:59:15] <sakoman> no, false alarm
  • [18:00:37] <ldesnogu_> I am not sure what MPU and core are in the cpuidle mail; I would have assumed MPU = ARM, but then what is core?
  • [18:04:25] <koen> maybe mpu + dsp?
  • [18:04:55] <koen> I'll try to build a kernel with those patches in a moment
  • [18:06:47] * jjg (n=jjg@12.40.200.74) has joined #beagle
  • [18:06:58] <jjg> hi
  • [18:07:04] <mru> koen: no, I haven't really got going yet
  • [18:07:22] <mru> still setting up a good environment to work in
  • [18:08:03] <robclark> koen, ldesnogu_: I suspect "core" == "core power domain".. which probably means things like the interrupt controllers and some of the core peripherals?
  • [18:08:18] <mru> and I managed to hose something, causing tty related problems
  • [18:08:32] <ldesnogu_> probably, I found core referenced in the context of the l4 interconnect
  • [18:08:33] <mru> and as always with tty problems, there are very few clues printed...
  • [18:08:58] <jjg> any youtube vids of the board doing 3d?
  • [18:09:16] <jkridner|work> as robclark said, it is some peripherals that must always be powered and infrastructure stuff.
  • [18:09:21] <DaQatz> Convert it all to hex and store it in binary chunks instead! Muahahahhahaha!
  • [18:09:39] <DaQatz> wc
  • [18:10:15] <jkridner|work> jjg: I'm not aware of any. the general community doesn't have 3D libraries yet.
  • [18:10:27] <jjg> ah, ok, thanks
  • [18:10:30] <jkridner|work> you can see other OMAP3 boards running some 3D.
  • [18:10:33] <ldesnogu_> some videos were posted
  • [18:10:41] <jkridner|work> with 3D?
  • [18:10:43] <ldesnogu_> on EVM
  • [18:10:45] <ldesnogu_> yes
  • [18:10:57] <ldesnogu_> MWeston (pandora HW engineer posted them)
  • [18:11:08] <jkridner|work> not just http://youtube.com/watch?v=-UFUbqoNgs8&feature=related ?
  • [18:11:12] <ldesnogu_> gimme a few minutes and I will try to locate them
  • [18:11:29] <jjg> does the omap have davinci technology under the hood? I heard that the davinci line of chips was getting merged with omap and it was going to be referred to as a technology rather than a discrete product
  • [18:12:17] <jkridner|work> yes. OMAP3 includes a C64x+ DSP and the DaVinci Link and Codec Engine technologies can be used to drive the DSP.
  • [18:12:20] <ldesnogu_> jjg, jkridner|work : http://youtube.com/user/MWeston2
  • [18:12:21] <Crofton|work> TI marketing should be shot :)
  • [18:12:33] <ldesnogu_> all marketing should be shot :)
  • [18:13:12] <jkridner|work> The SDK is not publicly available yet for Link and Codec Engine that is community compatible, but you can pull down Link v1.5 source from a TI web page.
  • [18:13:26] <jkridner|work> by "community compatible", I mean GPL in the kernel portions.
  • [18:13:39] <Crofton|work> but link does not work on OMAP3 yet?
  • [18:14:01] <suihkulokki> nvidia marketing certainly has balls: http://img.hexus.net/v2/news/nvidia/tegra/DSCF0121.jpg
  • [18:14:24] <koen> suihkulokki: right
  • [18:14:31] <jjg> ldesnogu, is this pandora platform built on omap?
  • [18:15:00] <ldesnogu_> yes
  • [18:15:01] <koen> suihkulokki: I wonder how the cortex perfoms with decoding SDTV resolutions
  • [18:15:01] <jkridner|work> heh, youtube vid of ubuntu on OMAP3: http://youtube.com/watch?v=5i9cWOK1spw
  • [18:15:23] * ldesnogu_ can't look at youtube on his x86_64 linux :(
  • [18:16:12] <jkridner|work> ldesnogu_: I don't see any 3D on the OMAP3 from MWeston2.
  • [18:16:35] <jjg> what about the opengl demo?
  • [18:16:41] <jjg> that on the omap?
  • [18:17:04] <suihkulokki> koen: mpeg2/divx probably well, h264 unlikely..
  • [18:17:47] <jkridner|work> no, that is iMX31. OMAP3 has better 3D performance than the iMX31.
  • [18:17:55] <DJWillis> ldesnogu: are you refering to the videos from MWeston posted some 7 months ago?
  • [18:18:26] <ldesnogu_> yes
  • [18:18:30] <ldesnogu_> http://youtube.com/watch?v=cpAj0jHglGw for instance
  • [18:18:51] <DJWillis> ldesnogu: as jkridner mentions that is all iMX31 from a long time ago (between MMSP2+ and OMAP3 ;-))
  • [18:18:52] <koen> does anyone have a sample mpeg2 clip at 720p and 1080i?
  • [18:19:00] <ldesnogu_> (can't check the content because as I told gnash stopped working for me)
  • [18:19:08] <ldesnogu_> oh OK my bad
  • [18:20:01] <jkridner|work> I've seen them do some 3D stuff w/o the 3D acceleration, which was fairly impressive in itself.
  • [18:20:15] <ldesnogu_> so let's say this gives a lower bound on what is achievable :)
  • [18:20:16] <jjg> lots of cycles to work with on this thing
  • [18:20:47] <ldesnogu_> koen, did you look on mplayer site?
  • [18:20:58] <ldesnogu_> koen, they have a benchmark archive IIRC
  • [18:23:04] <ldesnogu_> ftp://mplayer.hu/MPlayer/benchmark
  • [18:24:21] * jkridner_ (n=jason@c-76-31-18-64.hsd1.tx.comcast.net) has joined #beagle
  • [18:24:54] * jkridner (n=jason@c-76-31-18-64.hsd1.tx.comcast.net) Quit (Read error: 104 (Connection reset by peer))
  • [18:25:02] <koen> http://samples.mplayerhq.hu/benchmark/
  • [18:25:06] <ldesnogu_> jkridner|work, I checked that OpenMAX news on ARM site, all I can say is that the NEON version is the same I downloaded a month ago, that's strange
  • [18:27:57] * macneib (n=macneib@d221-71-252.commercial.cgocable.net) has joined #beagle
  • [18:34:13] <ds2> noticed Linuxdevices mentions a 3440 for low volume stuff? how's that diff from the 35xx's?
  • [18:34:19] * dirk2 (n=dirk@F3089.f.strato-dslnet.de) has left #beagle
  • [18:37:08] <ldesnogu_> IIRC 3440 is a 3430 with support for 720p HD
  • [18:37:35] <ldesnogu_> it also runs at 800 MHz...
  • [18:37:40] <ldesnogu_> http://focus.ti.com/general/docs/wtbu/wtbuproductcontent.tsp?templateId=6123&navigationId=12796&contentId=36505
  • [18:39:54] <ds2> hmmmm
  • [18:40:08] <ds2> thus leaving the 35xx in the dust?
  • [18:41:17] <ldesnogu_> it's not the same market
  • [18:41:36] <ldesnogu_> anyway 600 vs 800 is not that a big difference
  • [18:41:45] <koen> 33%
  • [18:41:53] <ldesnogu_> and you can overclock the 3530 if you feel the need
  • [18:42:12] <ds2> speed is speed
  • [18:42:20] <ldesnogu_> if you want speed buy a PC :)
  • [18:42:26] <koen> 209 second h264 480p clip: BENCHMARKs: VC: 409.427s VO: 0.042s A: 0.000s Sys: 1.092s = 410.561s
  • [18:42:38] <koen> that's with an unoptimized mplayer
  • [18:42:42] <ldesnogu_> how many frames?
  • [18:42:58] <koen> 209 times 60 I guess
  • [18:43:03] <ds2> the part that was confusing was the 3440 is targeted toward low volume...thought that's what the 35xx were?
  • [18:43:03] <ldesnogu_> hum 209 sec is the time
  • [18:43:04] <ldesnogu_> ok :)
  • [18:43:27] <ldesnogu_> koen, NEON or the c6x will make that much better :)
  • [18:43:30] <koen> (the "Get Smart" movie trailer from apple.com)
  • [18:43:42] <koen> ldesnogu_: vfp and even armv6 would make it better
  • [18:43:55] <ldesnogu_> mplayer supports v6
  • [18:43:59] <ldesnogu_> at least the idct
  • [18:44:27] <ldesnogu_> and the VFP would slow it down I am afraid
  • [18:44:33] <mru> vfp is useless for video
  • [18:44:42] <mru> video uses only integers
  • [18:44:46] <ldesnogu_> vfp is useless on Cortex-A8 :P
  • [18:44:58] <mru> yes, but the cortex has neon instead
  • [18:45:05] <ldesnogu_> yes :)
  • [18:45:09] <koen> ldesnogu_: 1.0rc2 doesn't have armv6 support iirc
  • [18:45:16] <mru> and studying the instruction set, neon looks very promising
  • [18:45:22] <ldesnogu_> oh, it's only on the trunk?
  • [18:45:28] <ldesnogu_> mru: it is
  • [18:45:37] <koen> mru: audio still uses floats :(
  • [18:45:40] <mru> ffmpeg/mplayer has had armv6 support since I got my nokia n800
  • [18:45:49] <mru> some audio uses floats, some not
  • [18:45:52] <ldesnogu_> mru: BTW I will spend some time playing with NEON and avcodec
  • [18:46:07] <ldesnogu_> that could be nice to cooperate
  • [18:50:00] <ds2> what's wrong with VFP on the cortex?
  • [18:50:16] <koen> ds2: it's slow compared to arm1136
  • [18:50:18] <ldesnogu_> it's not pipelined
  • [18:50:23] <ds2> ah I see
  • [18:50:38] <koen> mru: building mplayer trunk now
  • [18:50:52] <ds2> but the catch is neon only does integer right?
  • [18:51:03] <ldesnogu_> it does single precision FP
  • [18:51:06] <Crofton> no neon can do float
  • [18:51:19] <ds2> I see
  • [18:51:31] <ldesnogu_> and it does them in SIMD mode (2 operations at a time)
  • [18:51:44] <ds2> so for a cortex system, the fastest is to do what it can on Neon and fall back to VFP only if neon can't do it?
  • [18:52:11] <ldesnogu_> you don't need double precision that often anyway
  • [18:52:41] <ds2> hmm wonder if anyone has done that with a cortex specific libm
  • [18:53:22] <ldesnogu_> koen, do you know if gcc can use NEON for floating point without vectorization?
  • [18:53:42] <koen> ldesnogu_: it has an -mfpu=neon switch
  • [18:54:06] <ldesnogu_> did you give it a try?
  • [18:58:30] <koen> yes, but I didn't show disassembly to neon people yet :)
  • [19:03:01] <Crofton|work> did it generate any neon instructions
  • [19:03:07] <Crofton|work> and did it compile and run?
  • [19:03:24] <koen> see note about disassembly
  • [19:06:17] <ldesnogu_> I made a quick test a few week with NEON vectorization with a SP FP dot product and it generated NEON instructions
  • [19:06:29] <ldesnogu_> (and it did not crash :)
  • [19:06:38] <Crofton|work> what compiler?
  • [19:07:05] <ldesnogu_> 2008q1
  • [19:09:07] * RogerMonk (n=a0740758@nat/ti/x-e47deaa3999554de) has joined #beagle
  • [19:09:26] <kulve> mru: about the X.Org driver, what hw were you planning to use?
  • [19:09:44] <mru> eh, the beagle board of course
  • [19:09:51] <mru> I have a rev B board
  • [19:10:03] <koen> NOTE: Applying patch '0001-omap3-cpuidle.patch' (/OE/org.openembedded.dev/packages/linux/linux-omap2-git/beagleboard/0001-omap3-cpuidle.patch)
  • [19:10:06] <koen> there we go :)
  • [19:10:25] <kulve> mru: but what part of the cpu? Not SGX nor IVA iirc?
  • [19:10:41] <mru> whatever is best for the task
  • [19:10:44] <like2wise> koen: board isupgraded, but I forgot it to bring home. Will ship it on Friday.
  • [19:10:48] <koen> first dispc then iva, right?
  • [19:10:54] <like2wise> koen: cu, cinema time
  • [19:10:57] <koen> like2wise: cool, thanks a lot!
  • [19:10:57] * like2wise is now known as like|away
  • [19:11:03] <mru> iva should be useful for 2d operations
  • [19:11:10] <mru> and it's documented
  • [19:11:22] <kulve> oh, I though it wasn't..
  • [19:11:27] <like|away> koen: np 8 MB is NOT enough for anyone. :-)
  • [19:11:32] <kulve> mru do you happen to have the url?
  • [19:11:37] <mru> the display controller can do simple overlay and rotation only
  • [19:12:33] <mru> http://focus.ti.com/pdfs/wtbu/swpu114g.pdf
  • [19:12:39] <mru> it's somewhere in there
  • [19:13:15] <kulve> "The page you are looking for might have been removed, had its name changed, or is temporarily unavailable."
  • [19:13:23] <mru> oh
  • [19:14:20] <mru> it was there a week ago
  • [19:14:40] <kulve> happen to have a copy? :)
  • [19:16:29] <mru> of course I do
  • [19:17:10] <ldesnogu_> http://focus.ti.com/pdfs/wtbu/swpu114g.pdf
  • [19:17:20] <ldesnogu_> it works for me...
  • [19:18:32] <kulve> weird.. I tried we two browsers..
  • [19:18:36] <kulve> s,we,with,
  • [19:19:05] <ldesnogu_> koen, fpu=neon still generated vfp instructions with 2008q1...
  • [19:21:40] <kulve> I tried with 3 browsers and two computers and the above url doesn't work..
  • [19:23:21] * JoeBorn_ (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) has joined #beagle
  • [19:24:51] <kulve> others can't access it either, but I got a copy
  • [19:31:24] <koen> ldesnogu_: 2008q1 is broken, maybe 2007q3 works
  • [19:41:27] <prpplague> wheee beagle got a post on linuxdevices
  • [19:42:26] <jkridner|work> yup, a bit sooner than I'd hope (no stock at the distributor).
  • [19:43:18] <prpplague> jkridner|work: fun fun
  • [19:43:27] <prpplague> jkridner|work: still no jtag support right?
  • [20:02:52] <keesj> I plan to spend the following next few month on better jtag support for open source projects (openocd)
  • [20:03:57] <keesj> I reodedered me this one http://www.amontec.com/jtagkey-tiny.shtml
  • [20:04:18] <keesj> of course I will need to create some adapter so 16 pins :(
  • [20:04:49] <keesj> and spend to much time on arm11 support :p , and end-up....
  • [20:06:37] <koen> cpuidle seems to work on the beagle
  • [20:06:37] <hagisbasheruk> that JTAGkey-Tiny seems like a nice device to hav , bookmarked :)
  • [20:06:49] <koen> root@beagleboard:/sys/devices/system/cpu/cpu0/cpuidle# cat */usage
  • [20:06:49] <koen> 5579778
  • [20:06:49] <koen> 89
  • [20:06:49] <koen> 3551
  • [20:06:49] <koen> 3266
  • [20:06:56] <koen> still way too much time in C1, though
  • [20:09:20] <prpplague> hagisbasheruk: jtagkey-tiny is basically the same as the flyswatter, but atleast witht he flyswatter you get schematics
  • [20:09:40] <keesj> hagisbasheruk: the elinux.org beagle wiki suggests using a different one.
  • [20:10:30] <prpplague> keesj: you are going to add cortex-a8 support to openocd?
  • [20:12:06] <keesj> prpplague: it's a goal to be able to debug more recent arm cores
  • [20:12:46] <hagisbasheruk> jtagkey-tiny is cheaper :) i like cheaper :)
  • [20:13:00] <prpplague> keesj: well if you are planning to do it, and are serious about doing it, send me an email at dave123_aml@yahoo.com , i can send you a flyswatter
  • [20:13:12] <prpplague> hagisbasheruk: only by $5USD
  • [20:13:16] <keesj> hagisbasheruk: shipping is not that cheap with the jtag-tiny :p
  • [20:13:34] <prpplague> hagisbasheruk: and on top of that you get a second serial port with the flyswatter, hehe
  • [20:14:14] <hagisbasheruk> ohh.. i didn't notice the price for jtagkey-tiny was Euros ,
  • [20:15:31] <keesj> prpplague: are you involved in the flyswatter?
  • [20:17:09] <prpplague> keesj: yea, i did most of the schematic work for it
  • [20:17:38] <prpplague> keesj: which is to say, its mostly a reference design
  • [20:17:54] <prpplague> keesj: but made in a way to make it easy for people to hack/dev with it
  • [20:19:05] <keesj> oerding one item costs me
  • [20:19:05] <keesj> oerding one item costs me $149.09
  • [20:19:24] <prpplague> keesj: ??
  • [20:21:17] <keesj> the default mail thingy is a little to much , but I was able to select a cheaper on (going to order a flyswatter)
  • [20:21:39] <prpplague> keesj: yea, using the USPS service you can get pretty cheap
  • [20:25:53] <keesj> Total: $73.95 -> +- 50 euro haha!
  • [20:28:53] <prpplague> good deal
  • [20:36:17] <sakoman> NishanthMenon: ping
  • [20:41:11] <keesj> prpplague: yes indeed
  • [20:47:33] * koen tries mplayer trunk
  • [20:51:57] <NishanthMenon> sakoman, pong
  • [20:55:39] <sakoman> NishanthMenon: in your experience how important is it to do the automatic offset adjustment procedure?
  • [20:56:04] <koen> lastlog sakoman
  • [20:56:20] <NishanthMenon> sakoman, ?? twlcore?
  • [20:56:43] <sakoman> yes, bit 7 of ANAMICL
  • [20:56:54] <sakoman> not core, audio section
  • [20:57:04] <sakoman> koen: lastlog??
  • [20:57:29] <koen> sakoman: missing '/'
  • [20:57:41] <koen> I was trying to see if you responed to my alsa question
  • [20:57:55] <koen> (you did: "no, false alarm")
  • [20:58:07] <sakoman> yes, the answer was "no breakthrough"
  • [21:00:01] <NishanthMenon> sakoman, i dont think i recollect seeing this bit
  • [21:00:20] <sakoman> NishanthMenon: if I crank up the analog gain I can hear the left channel. looking at it on a scope it looks clipped, but only in one direction
  • [21:00:31] <sakoman> much like an offset error would look
  • [21:00:43] <NishanthMenon> sakoman, no.. that could be coz of saturation.
  • [21:00:53] <NishanthMenon> let me rephrase:
  • [21:00:54] <koen> mplayer trunk: BENCHMARKs: VC: 403.271s VO: 0.038s A: 0.000s Sys: 2.415s = 405.724s
  • [21:00:57] * BThompson (n=BThompso@nat/ti/x-379e1e68fc9bc377) Quit ("Trillian (http://www.ceruleanstudios.com")
  • [21:00:57] <sakoman> any ideas?
  • [21:01:16] * hagisbasheruk (n=hagisbas@78.148.135.75) Quit (Remote closed the connection)
  • [21:01:32] <NishanthMenon> a) saturation, b) mcbsp bit loss
  • [21:01:35] <koen> 5 seconds faster than 1.0rc2
  • [21:02:34] <NishanthMenon> what twl4030 version do u have?
  • [21:02:37] <NishanthMenon> 3.0 or 2.1?
  • [21:03:01] <sakoman> I tried playing same data on both channels and looked at i2s on my scope (in storage mode)
  • [21:03:27] <NishanthMenon> i2s will look good.. it is one of the versions of twl4030 that got it's sampling goofed up
  • [21:03:29] <sakoman> data, clocks, and framing looked like the data sheet
  • [21:03:44] <NishanthMenon> it should be in twl4030 errata (if you have your hands on it)
  • [21:04:27] <NishanthMenon> my recommendation: generate a sinewave data and give it downstream
  • [21:04:36] <NishanthMenon> if you probe it, the issue is evident immediately
  • [21:04:46] <sakoman> what will I see?
  • [21:04:56] <NishanthMenon> clipped sinewave is fine
  • [21:05:08] <NishanthMenon> it will look almost like square if you have saturation issues
  • [21:05:27] <NishanthMenon> if your data bit is being messed, you will get a weired choppy sinewave on analog output
  • [21:05:36] <sakoman> How can I tell what version I have?
  • [21:05:45] <NishanthMenon> there is some revid reg
  • [21:05:50] * NishanthMenon looking it up
  • [21:10:12] <NishanthMenon> physical reg 0x85, audio device address, total of 4 bytes. if you see 0x0009002F it is TWL5030 ES1, 0x1009002F is TWL4030 es2.0, 0x2009002F is 2.1, 0x4009002F is es3.0, 0x5009002F is es3.1
  • [21:11:55] <NishanthMenon> ok, the work around applies for ES3.0 and ES3.1 of TWL4030 -> clkxp=0 instead of clkxp=1 for all other devices.
  • [21:12:07] <NishanthMenon> this is in mcbsp configuration.
  • [21:12:45] <sakoman> OK, let me see what mcbsp in git is using
  • [21:13:14] <koen> does the kernel output the twl version?
  • [21:13:30] <NishanthMenon> koen, dunno :( not looked at kernel for long :(
  • [21:13:47] <sakoman> it uses clkxp=1
  • [21:14:10] <sakoman> I don't see that register in the audio section of the trm. is it "secret"?
  • [21:15:26] <NishanthMenon> no it is not part of the audio module
  • [21:15:38] <NishanthMenon> it is just on the same physical bus. search for IDCODE
  • [21:15:47] <sakoman> OK
  • [21:17:44] <sakoman> I'll modify the driver to print that out
  • [21:19:00] <sakoman> NishanthMenon: if that is the issue, shouldn't I be seeing the same problem on *both* channels?
  • [21:19:02] <NishanthMenon> sakoman, you might be better off debugging driver with sinewave though.
  • [21:19:26] <NishanthMenon> sakoman, probably yes. hmm..
  • [21:19:44] <sakoman> right channel looks and sounds quite perfect
  • [21:20:05] <NishanthMenon> hmm.. did you check the digital gain b/w left and right?
  • [21:20:19] <sakoman> yes, identical setting in the two registers
  • [21:20:26] <NishanthMenon> also, are u probing without the jack connected?
  • [21:20:31] <sakoman> identical setting for l & r analog gain too
  • [21:20:41] <sakoman> yes probing with jack connected
  • [21:21:00] <sakoman> will try with headphones disconnected
  • [21:21:01] <NishanthMenon> drat.. i thought probably variance in load
  • [21:21:27] <NishanthMenon> sakoman, without headset, it is going to spike up.. you may want to reduce gain down
  • [21:21:43] <sakoman> left signal looks like only peaks are getting through
  • [21:21:51] <sakoman> but only positive peaks
  • [21:21:58] <NishanthMenon> that is really weird
  • [21:22:05] <sakoman> no kidding!
  • [21:22:18] <NishanthMenon> ext mute?
  • [21:22:22] <sakoman> only when analog gain is cranked up to 10-12db
  • [21:22:51] <sakoman> when it is lower analog gain I don't see much of anything!
  • [21:23:06] <sakoman> not sure what you mean by ext mute
  • [21:23:14] <NishanthMenon> external mute
  • [21:23:26] <NishanthMenon> it has that stuff for preventing pop noise
  • [21:23:28] <sakoman> you mean on headset?
  • [21:23:34] <NishanthMenon> no
  • [21:23:38] <sakoman> oh, ok
  • [21:23:41] <NishanthMenon> i think there was some sort of gpio thingy
  • [21:23:48] <NishanthMenon> not sure about evm though..
  • [21:24:16] <sakoman> I looked at the schematic and didn't see any 4030 gpio's being used for external mute
  • [21:24:42] <sakoman> Didn't see any such thing on Beagel either
  • [21:24:55] <sakoman> But I did see it in the code. Found that confusing!
  • [21:25:46] <NishanthMenon> wow.. i seemed to have commented something of interesting to you
  • [21:26:40] <NishanthMenon> http://pastebin.com/d72cf65a9
  • [21:27:05] <NishanthMenon> "ensuring the existance of the negative phase of analog signal" dunno why made that statement
  • [21:28:23] <NishanthMenon> sakoman, brb meetings :(
  • [21:28:32] <sakoman> well there you go! thanks for the pointer!
  • [21:28:40] <sakoman> enjoy your meeting!
  • [21:28:43] <NishanthMenon> :(
  • [21:29:36] <ldesnogu_> gn all
  • [21:29:41] * ldesnogu_ (n=ldesnogu@ven06-2-82-247-86-183.fbx.proxad.net) Quit ("Leaving")
  • [21:34:32] <koen> " blinking leds are always considered the highhest priority
  • [21:34:33] <koen> patches ;)"
  • [21:34:42] <koen> led patch seems to be upstream now :)
  • [21:36:01] * hagisbasheruk (n=hagisbas@78.148.135.75) has joined #beagle
  • [21:38:56] * hagisbasheruk (n=hagisbas@78.148.135.75) Quit (Client Quit)
  • [21:39:38] <NishanthMenon> sakoman, look at sound/oss/omap2-audio-twl4030.c http://linux.omap.com/pub/kernel/3430zoom/linux-ldp-v1.0b.tar.gz that used the original code i had written couple of years back
  • [21:40:10] <sakoman> OK, will do
  • [21:40:20] <sakoman> hey you're supposed to be in a meeting!
  • [21:40:25] <NishanthMenon> it wont scale to the git kernel, but you should be able to get most of the i2c op sequence..
  • [21:40:36] <NishanthMenon> sakoman, just got back.. short talk with boss :(
  • [21:40:54] <sakoman> that's the way I like them!
  • [21:41:14] <sakoman> In a one person company my meetings are amazingly short :-)
  • [21:41:20] <NishanthMenon> lol :)
  • [21:42:00] <koen> yay
  • [21:42:03] <koen> gcc 4.3.1 released
  • [21:44:37] <sakoman> koen: I sure hope this means that I can use the same compiler for both u-boot and linux now!
  • [21:48:35] <koen> I hope so too!
  • [21:48:38] * JonHunter (n=a0741266@nat/ti/x-7b910f2683f67cc7) has joined #beagle
  • [21:48:52] <Crofton> My boss is the cat
  • [21:48:54] <koen> I'm sorting out patches atm
  • [21:52:54] <sakoman> hmm . . . Tony didn't like my nand patches. Strange. I always run checkpatch on them before submitting
  • [21:53:04] <sakoman> perhaps gmail screwed them up
  • [21:54:47] * bazbell (n=a0192809@nat/ti/x-f6a69bbd9e7dad7c) Quit (Remote closed the connection)
  • [21:55:22] <koen> sakoman: I'm using git-send-email nowadays
  • [21:55:49] <sakoman> NishanthMenon: I just disabled the anti-pop feature entirely and I now get both channels
  • [21:56:23] <sakoman> Need to throttle back the analog gain, but this is promising
  • [21:56:38] <NishanthMenon> Coooooool...
  • [21:56:58] <NishanthMenon> only you'd not want to wear a headphone...
  • [21:57:13] <sakoman> I think I'll do a first cut without any of the fancy features
  • [21:57:20] <NishanthMenon> darn pop almost blew my ears out the first time i tested.. with max volume :(
  • [21:57:22] <koen> the wtbu kernel is real bad pops
  • [21:57:29] <koen> s/is/has/
  • [21:58:07] <sakoman> No pops when playing mp3's
  • [21:58:12] <NishanthMenon> koen, i dont know if someone has spend time looking at the pop code.. i put a scope and could clearly see the bump.. i think i goofed up on the ioctl part..
  • [21:58:22] <sakoman> Just too loud, even with digital volume cranked way down
  • [21:58:24] <NishanthMenon> sakoman, that is good news :)
  • [21:58:32] <koen> NishanthMenon: you should ask the ASoC guys about pop&click reduction :)
  • [21:59:30] <sakoman> Need to run into town now. I'll look at audio in this evening. Hopefully it will "just work"
  • [21:59:31] <NishanthMenon> koen, that will be the day when ASIC guys listen to software guys.. :P
  • [21:59:43] <NishanthMenon> sakoman, happy hunting..
  • [21:59:45] * rsalveti (n=salveti@200.184.118.132) Quit (Read error: 113 (No route to host))
  • [22:00:06] <sakoman> Thanks for the help. That negative signal comment was the key to the whole mystery!
  • [22:00:18] <sakoman> Very bizarre symptom!
  • [22:00:37] <NishanthMenon> i must have picked the comment from some where in TRM.. dont recollect now though :(
  • [22:01:04] <sakoman> I don't remember seeing that. I'll have to reread that section when I turn anti-pop back on
  • [22:01:22] <NishanthMenon> probably the external mute line is pulling line up? wierd that it should do only to left channel...
  • [22:02:09] <NishanthMenon> there are some 4 wire and 5 wire thingy about connecting headset.. ext mute, if i recollect has to be configured in the right way.. so maybe, it might have to do with schematics..
  • [22:02:11] * gkarabin (n=gkarabin@64.212.175.249) has left #beagle
  • [22:02:17] * gkarabin (n=gkarabin@64.212.175.249) has joined #beagle
  • [22:03:20] * NishanthMenon wondering if there is some way of spell+grammar checking incode comments..
  • [22:04:17] <sakoman> I hate the EVM schematics. Who decided on using red and blue text & lines? Practically unreadable!
  • [22:04:53] <prpplague> sakoman: probably marketing
  • [22:05:08] <sakoman> :-) The colors are so pretty!
  • [22:05:53] <prpplague> sakoman: engineer "yes we are starting the dev for our next gen product, what features would you like to have?" marketing "uh, it needs to be blue"
  • [22:06:47] <sakoman> no, pink. hmm . . Ok, I have an idea. Let's make the lines blue and the text pink!
  • [22:07:14] <NishanthMenon> sakoman, you can ping khasim also on audio if i am not online in the morning- if i recollect, he'd dabbed a bit in it too..
  • [22:07:53] <sakoman> Thanks! Hopefully it will all go much more easily now.
  • [22:08:14] <NishanthMenon> yep.. touch wood
  • [22:08:33] <sakoman> Mixer controls work. I need to leave before I try audio in, otherwise I might not get away anytime soon :-)
  • [22:10:02] <jkridner|work> sakoman: so, audio out is working now?
  • [22:11:43] <NishanthMenon> jkridner|work, i think yes at least the initial problem seems gone.. though that POP reduction external mute line may need a clean look
  • [22:11:52] <Crofton|work> koen, should I nuke tmp and start over
  • [22:16:14] * JonHunter (n=a0741266@nat/ti/x-7b910f2683f67cc7) has left #beagle
  • [22:20:26] * rsalveti (n=salveti@189.70.96.19) has joined #beagle
  • [22:37:30] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
  • [22:39:01] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
  • [22:43:41] * bazbell (n=a0192809@nat/ti/x-2dccdf26a205e1f1) has joined #beagle
  • [23:14:41] <sakoman> jkridner: can you confirm whether beagleboard and evm have external mute circuitry
  • [23:15:39] <ds2> external? doesn't the triton have that?
  • [23:17:12] <sakoman> The evm code seems to diddle one of the 4030 gpio's for mute
  • [23:17:30] <sakoman> but I don't see any circuitry connected to that gpio :-)
  • [23:18:02] <ds2> oh I see
  • [23:18:22] <sakoman> could be that it is on some schematic page nowhere near the other audio stuff, but as far as I can tell it just goes to a test point
  • [23:18:28] <jkridner|work> sakoman: I didn't think there was any, but I'll check.
  • [23:18:38] <ds2> maybe it is for debugging?
  • [23:18:42] <sakoman> I don't want to carry that code along if it isn't actually doing anything!
  • [23:19:35] <ds2> or maybe it is to drive a muted LED
  • [23:20:04] <ds2> is this a beagle/EVM specific thing or does it do it for the SDP also?
  • [23:20:13] <sakoman> the test point connection is on the evm. can't tell what is happening with beagle since the schematics in the manual are fairly limited
  • [23:20:42] <sakoman> ds2: don't know, that's why I'm asking :-)
  • [23:20:50] <ds2> also which driver, SoC or the other one?
  • [23:21:23] <jkridner|work> I'm pretty confident nothing is connected for external mute.
  • [23:21:40] <sakoman> OK, I figured as much
  • [23:26:23] <NishanthMenon> jkridner|work, mebbe someone could look at "Headset Pop-Noise Attenuation Diagram" figure in TWL4030 trm?
  • [23:28:02] <jkridner|work> sakoman: do you have the TWL4030 TRM?
  • [23:28:11] <ds2> sakoman: what code are you starting with?
  • [23:28:34] <sakoman> jkridner: I have the catalog version trm
  • [23:28:57] <sakoman> ds2: it is my own alsa soc driver
  • [23:29:55] <ds2> i see
  • [23:30:12] <sakoman> I looked at the 2.6.22 alsa code, but decided to start from scratch since it is quite complex, not in the proper form for SOC, and supports lots of features that aren't present on beagle or the evm
  • [23:33:37] * jkridner|work is headed home and will be back online later.
  • [23:33:47] * jkridner|work (n=a0321898@nat/ti/x-50659f63475abf89) Quit ("Leaving.")
  • [23:41:52] * pdg37 (n=paul@wsip-24-234-78-58.lv.lv.cox.net) has joined #beagle
  • [23:50:55] * like|away (n=chatzill@82-171-51-231.ip.telfort.nl) Quit ("ChatZilla 0.9.82.1 [Firefox 3.0b5/2008050509]")