• [00:00:10] * maelcum (n=quassel@e178167126.adsl.alicedsl.de) has joined #beagle
  • [00:08:38] * Adventus_ (n=lachlan_@ESS-p-144-138-159-144.mega.tmns.net.au) has joined #beagle
  • [00:11:50] * TAK2004 (n=Administ@dslb-088-074-043-005.pools.arcor-ip.net) Quit (Read error: 104 (Connection reset by peer))
  • [00:12:09] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [00:13:09] <Adventus_> hiya
  • [00:13:29] <Adventus_> has anyone manged to to run nbench with the Neon unit enabled?
  • [00:40:37] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has left #beagle
  • [00:43:10] * florian (n=fuchs@g227200042.adsl.alicedsl.de) Quit ("Verlassend")
  • [00:45:06] <Adventus_> I ask because, on the openpandora forums, we seem to be getting a low FP score.
  • [01:07:12] <ssvb> Adventus_: are you benchmarking double precision FP?
  • [01:12:19] * geckosenator (n=sean@c-24-8-193-190.hsd1.co.comcast.net) has joined #beagle
  • [01:13:07] <geckosenator> hi
  • [01:19:56] * ClaudeQC (n=claude@bas1-quebec03-1279636829.dsl.bell.ca) has left #beagle
  • [01:21:19] <Adventus_> ssvb: nope should be single
  • [01:23:03] <Adventus_> they're using these compiler flags: "-O3 -fomit-frame-pointer -mfloat-abi=softfp -mfpu=neon -mcpu=cortex-a8 -ftree-vectorize -ffast-math"
  • [01:23:04] <ssvb> are you using options like -march=armv7a -mcpu=cortex-a8 -mfpu=neon -mfloat-abi=softfp ?
  • [01:23:43] <ssvb> then it should be probably enabled
  • [01:24:33] <Adventus_> Hmm thats worrying then. We only get an FP index around 0.8.
  • [01:25:17] <ds2> what's something like a 2430 or 2420 scoring?
  • [01:27:04] <Adventus_> An omap 2420 gets around 0.908 according to http://wiki.hbmobile.org/index.php?title=Application_Processor_Benchmarks
  • [01:27:44] <Adventus_> The rest of our number are good, in the 5 range.
  • [01:28:03] <ds2> Hmmmm 0.8 is not too far out of line then... 2420 only has VFP
  • [01:29:26] * fulgas is now known as FuL|OUT
  • [01:30:06] <Adventus_> Yea but we're at 2x the clock and apparantly with a Neon.
  • [01:30:36] <ds2> 2x clock? how fast is the 2420 being run at?
  • [01:30:49] * JoeyBorn (n=jborn@adsl-75-3-7-160.dsl.chcgil.sbcglobal.net) has joined #beagle
  • [01:30:50] <Adventus_> at 330mhz.
  • [01:31:20] <ds2> I see, thought it was at 400MHz... but it isn't 2x...think the beagle w/o the other mods runs at 500 or 550
  • [01:32:10] <Adventus_> This aint a beagleboard, its an OMAP3 based handheld. Its been clocked stable at 900mhz.
  • [01:32:17] <ds2> ah
  • [01:32:45] <ds2> try using VFP instead of Neon?
  • [01:32:48] <ds2> j #reprap
  • [01:32:54] <ds2> bah
  • [01:33:44] <Adventus_> Yea i guess they should try using just the VFP-lite and see if its the same.
  • [01:34:36] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [01:34:54] <Adventus_> PS: I dont actually have access to the board, I'm just curious since my software is likely to be FP bottlenecked.
  • [01:37:50] <mru> you need to coerce the chip into what ARM calls "runfast" mode
  • [01:38:04] <mru> it lets more single-precision FP instructions run in the neon pipeline
  • [01:38:08] <mru> instead of the vfp
  • [01:38:15] <Adventus_> So those flags wont enable runfast mode?
  • [01:38:25] <mru> I don't think so
  • [01:39:02] <mru> runfast mode is a combination of fpu control register settings
  • [01:39:13] <Adventus_> Ahhhh, thats promising.
  • [01:39:14] <mru> things like rounding towards zero
  • [01:39:31] <mru> rather than to nearest
  • [01:40:06] <mru> also things to do with inexact exceptions and such
  • [01:40:12] <Adventus_> Will the enable runfast asm at http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dai0133c/index.html work?
  • [01:40:53] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit (Read error: 104 (Connection reset by peer))
  • [01:40:58] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [01:43:06] <mru> looks like it should work
  • [01:44:38] <Adventus_> I'll suggest they use that, cheers.
  • [01:55:58] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Connection timed out)
  • [01:56:13] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [02:08:39] * n6pfk (n=mike@c-76-104-40-104.hsd1.va.comcast.net) has joined #beagle
  • [02:09:29] <n6pfk> What is the difference in the two HDMI video cables offered by digi-key?
  • [02:11:52] <geckosenator> koen: Is your mode setting patch applied to any tree?
  • [02:21:02] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit (Read error: 131 (Connection reset by peer))
  • [02:21:13] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [02:33:13] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [03:07:35] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [03:15:45] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) has joined #beagle
  • [03:24:14] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Read error: 110 (Connection timed out))
  • [03:24:30] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [03:25:45] * JoeyBorn (n=jborn@adsl-75-3-7-160.dsl.chcgil.sbcglobal.net) Quit ("time for the bike")
  • [03:29:33] * tulpe (n=ask@p54AF659C.dip.t-dialin.net) Quit (Read error: 110 (Connection timed out))
  • [03:29:50] * tulpe (n=ask@p54AF7B3E.dip.t-dialin.net) has joined #beagle
  • [03:34:17] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [04:00:33] * Strontium (n=Strontiu@118.101.106.2) has joined #beagle
  • [04:03:22] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [04:07:00] * soman (n=somnath@61.16.248.242) has joined #beagle
  • [04:23:24] * Adventus (n=lachlan_@CWPP-p-203-54-191-235.prem.tmns.net.au) has joined #beagle
  • [04:25:30] * Adventus_ (n=lachlan_@ESS-p-144-138-159-144.mega.tmns.net.au) Quit (Read error: 60 (Operation timed out))
  • [04:27:54] * dcramer_ (n=davec@dcdsl.ebox.com) Quit (Read error: 104 (Connection reset by peer))
  • [04:28:34] * dcramer (n=davec@dcdsl.ebox.com) has joined #beagle
  • [04:50:10] * jsync (n=jess@59.160.172.220) has joined #beagle
  • [05:14:21] * Xerion_ (i=xerion@82-170-197-160.ip.telfort.nl) has joined #beagle
  • [05:30:23] * Xerion (i=xerion@82-170-197-160.ip.telfort.nl) Quit (Connection timed out)
  • [05:30:23] * Xerion_ is now known as Xerion
  • [05:50:24] <n6pfk> What is the difference in the two HDMI video cables offered by digi-key?
  • [05:55:28] * gregoiregentil (n=zonbu@adsl-71-135-114-242.dsl.pltn13.pacbell.net) has left #beagle
  • [05:59:01] * supo (n=supo@s-inf-pc117.oulu.fi) has joined #beagle
  • [06:41:54] * macneib (n=macneib@70.29.45.5) has joined #beagle
  • [06:42:31] <macneib> Hello, i was wondering if anyone has more precise instructions for setting up zd1211 based usb wifi dongles on the BB?
  • [06:42:49] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [06:44:41] <macneib> 'magically' it started behaving...
  • [06:59:45] * macneib (n=macneib@70.29.45.5) has left #beagle
  • [07:01:53] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) Quit ("Leaving")
  • [07:08:22] * soman (n=somnath@61.16.248.242) Quit (Read error: 60 (Operation timed out))
  • [07:08:42] * Sept (n=bakljg@c-98-240-226-129.hsd1.mn.comcast.net) Quit (Read error: 131 (Connection reset by peer))
  • [07:08:47] * Sept (n=bakljg@c-98-240-226-129.hsd1.mn.comcast.net) has joined #beagle
  • [07:15:06] * soman (n=somnath@61.16.248.242) has joined #beagle
  • [07:25:37] * mykro (n=notranc@zima.linwin.com) Quit (Remote closed the connection)
  • [07:42:29] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
  • [07:42:35] * _bernie (n=bernie@trinity.develer.com) Quit (Remote closed the connection)
  • [07:43:34] * _bernie (n=bernie@trinity.develer.com) has joined #beagle
  • [07:46:12] * _bernie (n=bernie@trinity.develer.com) Quit (Client Quit)
  • [08:05:40] * macneib (n=macneib@70.29.45.5) has joined #beagle
  • [09:00:25] * fr3men (n=fr3men@ip-5-146.sn2.eutelia.it) has joined #beagle
  • [09:07:31] * Adventus (n=lachlan_@CWPP-p-203-54-191-235.prem.tmns.net.au) Quit (Read error: 110 (Connection timed out))
  • [09:39:50] * _AV500_ (n=AV500@IP-213157024130.static.heagmedianet.de) has joined #beagle
  • [09:55:40] * chakie_work (n=jeh@vpn-pub2.hard.ware.fi) Quit (Read error: 54 (Connection reset by peer))
  • [10:00:45] * ssvb (n=ssvb___@a88-114-221-132.elisa-laajakaista.fi) Quit (Remote closed the connection)
  • [10:09:55] * guillaum1 (n=gl@AMontsouris-153-1-91-1.w90-2.abo.wanadoo.fr) Quit (Read error: 145 (Connection timed out))
  • [10:14:31] * abitos (n=nixgibts@dslb-084-057-197-153.pools.arcor-ip.net) has joined #beagle
  • [10:20:28] * RogerMonk (n=a0740758@nat/ti/x-03a4edb2bc116938) has joined #beagle
  • [10:22:27] * guillaum1 (n=gl@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) has joined #beagle
  • [10:27:24] * thomasg (n=thomasg@85.131.189.115) Quit (SendQ exceeded)
  • [10:34:06] * guillaum1 (n=gl@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [10:34:37] * guillaum1 (n=GZBA4143@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) has joined #beagle
  • [10:50:08] * abitos (n=nixgibts@dslb-084-057-197-153.pools.arcor-ip.net) Quit (Nick collision from services.)
  • [10:50:23] * abitos (n=nixgibts@dslb-084-057-206-030.pools.arcor-ip.net) has joined #beagle
  • [10:50:39] * florian__ (n=fuchs@217.146.132.69) has joined #beagle
  • [10:51:05] * florian__ (n=fuchs@217.146.132.69) Quit (Read error: 104 (Connection reset by peer))
  • [10:51:07] * florian (n=fuchs@217.146.132.69) has joined #beagle
  • [11:00:50] * felipec (n=felipec@a91-153-251-222.elisa-laajakaista.fi) Quit (Read error: 110 (Connection timed out))
  • [11:01:08] * felipec (n=felipec@a91-153-251-222.elisa-laajakaista.fi) has joined #beagle
  • [11:02:26] * thomasg (n=thomasg@85.131.189.115) has joined #beagle
  • [11:10:01] * Leon_Nardella (n=Leon@200-161-14-111.dsl.telesp.net.br) has joined #beagle
  • [11:14:27] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [11:20:42] * _apk (n=kows@85-18-136-98.fastres.net) Quit (Read error: 60 (Operation timed out))
  • [11:47:46] <koen> wow, that pico DLP is *small*
  • [11:55:05] * zedstar (n=john@fsf/member/zedstar) has joined #beagle
  • [12:07:51] * felipec (n=felipec@a91-153-251-222.elisa-laajakaista.fi) Quit ("Leaving")
  • [12:08:50] * Openfree (n=df@222.65.143.118) has joined #beagle
  • [12:17:16] * florian still waiting for DHL
  • [12:19:40] <koen> http://www.flickr.com/photos/koenkooi/3109688167/
  • [12:20:30] <koen> florian: you'll like http://www.flickr.com/photos/koenkooi/3110520884/in/photostream/ :)
  • [12:21:00] <DJWills> Damm, that is tiny.
  • [12:21:13] <florian> koen: yes, that's cool indeed
  • [12:21:17] <koen> DJWills: you lost an 'i'!
  • [12:22:50] <florian> koen: The shipment from TI came to Europe via Amsterdam, so you must have been the first receiving it ;)
  • [12:23:18] <koen> florian: I suspect it went through TI utrecht first and then got reshipped
  • [12:23:49] <DJWills> koen: what is the res of that DLP?
  • [12:24:13] <koen> DJWills: 640x480 input, 480x320 output, ~9 lumens
  • [12:24:20] <florian> koen: exactly... the tracking number I got was for a shipment with a weight of ~220kg :)
  • [12:24:23] <koen> the DLP asic does the scaling
  • [12:24:36] <koen> florian: heh, 'my' shipment was 530kg :)
  • [12:24:43] <florian> koen: hehe
  • [12:25:19] * koen is kind of regretting that he painted the nearest wall limegreen
  • [12:27:16] <DJWills> koen: hmmm, interesting, am I right in thinking the DPP1500 in that has no colour wheel just LED's doing the job?
  • [12:27:29] <_AV500_> koen: you can compensate be making the image more red!
  • [12:27:58] <koen> DJWills: I have no idea :)
  • [12:28:48] <DJWills> koen: either way, I want one. So many uses for a mini DLP.
  • [12:33:09] * and-r1 (n=ubuntu@pd95b4498.dip0.t-ipconnect.de) has joined #beagle
  • [12:36:21] * feig (n=ejf3@c-66-30-116-185.hsd1.ma.comcast.net) Quit ("Konversation terminated!")
  • [12:36:40] <and-r1> koen: i read that you had a similar problem in the past which i have now...."omapfb omapfb: controller initialization failed (-12)" ....the problem with the framebuffer....but i didn't understand what you have done to fix it......do you remember......you have written somthing about increasing dma to 4mb, which is default in my kernel, and booting very often....but this can't be the solution?
  • [12:37:26] <and-r1> http://www.beagleboard.org/irclogs/index.php?date=2008-05-09
  • [12:39:44] <koen> and-r1: use the defconfig of the OE kernel (zcat /proc/config.gz), that has the right DMA settings
  • [12:44:57] <and-r1> hmm...the problem is that this didn't work....i tried to use a config from a kernel which already runs with the omapfb...but i get another error when i use this config
  • [12:47:23] <and-r1> i thought you have a solution for this.....but thanks for the help
  • [12:50:24] * rsalveti (n=salveti@189.70.38.175) Quit (Remote closed the connection)
  • [12:50:41] <Crofton|work> koen, paint it white
  • [12:55:15] * soman (n=somnath@61.16.248.242) Quit (Remote closed the connection)
  • [12:56:36] * TehUni (n=cory@c-24-30-35-55.hsd1.ga.comcast.net) has joined #beagle
  • [12:56:36] * TheUni (n=cory@c-24-30-35-55.hsd1.ga.comcast.net) Quit (Read error: 104 (Connection reset by peer))
  • [13:03:03] * alecrim (n=alecrim@189.2.128.130) has joined #beagle
  • [13:05:10] * FuL|OUT is now known as fulgas
  • [13:06:15] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [13:11:07] * jsync (n=jess@59.160.172.220) Quit (Remote closed the connection)
  • [13:32:15] * alecrim (n=alecrim@189.2.128.130) Quit ("Leaving")
  • [13:35:12] * alecrim (n=alecrim@189.2.128.130) has joined #beagle
  • [13:38:09] * alecrim (n=alecrim@189.2.128.130) Quit (Client Quit)
  • [13:38:20] * alecrim (n=alecrim@189.2.128.130) has joined #beagle
  • [13:40:03] * n6pfk (n=mike@c-76-104-40-104.hsd1.va.comcast.net) Quit ("Leaving")
  • [13:44:00] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [13:46:51] * favor (n=favor@118.113.207.154) has joined #beagle
  • [13:47:49] <favor> hi. Can beagle board support 24 bit color display?
  • [13:48:14] <koen> yes, but your kernel needs to support it
  • [13:48:24] <koen> that's why I keep asking which kernel you use
  • [13:49:11] <favor> thank you....
  • [13:49:33] <favor> good friend
  • [13:50:16] <favor> my kernel is 2.6.26-omap1 , can it support 24 bit color?
  • [13:52:48] <favor> I tried today, I build qt-embedded with "-depths 24", and get the same error. so , i think this kernel don't support 24 bit. can you suggest me which newer version kernel to try?
  • [13:54:31] * rsalveti (n=salveti@200.184.118.130) has joined #beagle
  • [14:00:28] * Leon_Nardella (n=Leon@200-161-14-111.dsl.telesp.net.br) Quit ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111318]")
  • [14:05:10] * supo (n=supo@s-inf-pc117.oulu.fi) Quit ("Ex-Chat")
  • [14:07:49] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit ()
  • [14:09:28] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [14:12:28] * MostAwesomeDude (n=simpson@c-24-21-147-156.hsd1.or.comcast.net) has joined #beagle
  • [14:34:04] * kevinsc (n=kevinsc@nat/ti/x-ba4a61e69ad2642f) has joined #beagle
  • [14:34:35] * kevinsc (n=kevinsc@nat/ti/x-ba4a61e69ad2642f) has left #beagle
  • [14:35:38] <geckosenator> hi
  • [14:38:31] * JuanG (n=Juan@nat/ti/x-fdf99d8a8e7a5dd6) has joined #beagle
  • [14:38:38] * JuanG (n=Juan@nat/ti/x-fdf99d8a8e7a5dd6) has left #beagle
  • [14:38:57] * rsalveti (n=salveti@200.184.118.130) Quit (kubrick.freenode.net irc.freenode.net)
  • [14:38:57] * GimpMaster (n=Beagle3@cpe-70-112-199-85.austin.res.rr.com) Quit (kubrick.freenode.net irc.freenode.net)
  • [14:38:57] * vlad (n=asdf@69-12-240-128.dsl.static.sonic.net) Quit (kubrick.freenode.net irc.freenode.net)
  • [14:38:57] * cbrake_away (n=cbrake@oh-69-34-21-229.sta.embarqhsd.net) Quit (kubrick.freenode.net irc.freenode.net)
  • [14:38:57] * Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) Quit (kubrick.freenode.net irc.freenode.net)
  • [14:39:02] * prpplague (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [14:39:07] <geckosenator> koen: your patch for various resolutions as a kernel parameter, is it applied to a tree?
  • [14:39:08] * vlad (n=asdf@69-12-240-128.dsl.static.sonic.net) has joined #beagle
  • [14:39:21] <geckosenator> koen: I am working with linux-omap, but I'm wondering if there is a better one
  • [14:39:25] * rsalveti (n=salveti@200.184.118.130) has joined #beagle
  • [14:39:42] <geckosenator> I'm hoping to support mode setting from fbset, I don't think it would be too hard
  • [14:39:47] * cbrake_away (n=cbrake@69.34.21.229) has joined #beagle
  • [14:40:27] * Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) has joined #beagle
  • [14:42:01] <MostAwesomeDude> geckosenator: Sorry, not familiar with your work. What are you working on?
  • [14:46:50] <khasim> I am on Zoom kernel, I see following lines with FBOUT :(
  • [14:46:52] <khasim> <6>omap24xxfb: Options "<NULL>"
  • [14:46:52] <khasim> <4>coherent allocation too big (requested 0x260000 mask 0xffffffff)
  • [14:46:52] <khasim> <3>omap24xxfb: cannot allocate framebuffer
  • [14:49:23] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [14:51:52] <geckosenator> MostAwesomeDude: I'm looking to add support for mode setting from ioctls
  • [14:52:28] <geckosenator> MostAwesomeDude: it's supported with most framebuffers, but not omap
  • [14:52:50] * JuanG (n=Juan@nat/ti/x-3cdfcf02e65dafe7) has joined #beagle
  • [14:52:53] * JuanG (n=Juan@nat/ti/x-3cdfcf02e65dafe7) has left #beagle
  • [14:55:29] <MostAwesomeDude> geckosenator: We may have things to discuss. I'm an Xorg developer, and I'm considering REing the driver to bring some acceleration to omap. :3
  • [14:56:21] <MostAwesomeDude> I haven't had a chance to actually dissect omapfb in the kernel yet. I assume it's pretty self-documented, yeah?
  • [14:56:21] * cbrake_away is now known as cbrake
  • [14:58:16] * geckosen1tor (n=sean@c-24-8-193-190.hsd1.co.comcast.net) has joined #beagle
  • [14:58:31] <geckosen1tor> MostAwesomeDude: sorry my internet is flaky
  • [14:58:44] <MostAwesomeDude> No worries.
  • [14:59:02] * fr3men (n=fr3men@ip-5-146.sn2.eutelia.it) has left #beagle
  • [15:00:48] <geckosen1tor> anyway, I'm looking to support mode setting without rebooting
  • [15:01:29] <geckosen1tor> and maybe also svideo support, I feel like it should be a separate framebuffer (/dev/fb2) and work at the same time as the normal output
  • [15:01:53] <MostAwesomeDude> Hm. That's one way, although...
  • [15:02:12] <MostAwesomeDude> Normally, dual outputs are done through a multiple FB map of some sort.
  • [15:02:24] <geckosen1tor> yes
  • [15:02:32] <geckosen1tor> so you can have consoles on all of them
  • [15:02:54] <geckosen1tor> there is an ioctl for that
  • [15:02:57] <geckosen1tor> CON2F
  • [15:03:02] <geckosen1tor> er, CON2FB
  • [15:03:17] <MostAwesomeDude> Have you had a chance to read up on KMS and GEM yet?
  • [15:03:24] <geckosen1tor> can set which console is on which fb, or figure out which fb is on which consol
  • [15:03:32] <geckosen1tor> what is KMS and GEM?
  • [15:04:29] <MostAwesomeDude> Interrupt me if I stop making sense.
  • [15:04:59] <MostAwesomeDude> KMS is Kernel Mode-Setting. It's an effort to move mode-setting code into the kernel, and provide decently generic controls for messing with the modes and card setup.
  • [15:05:01] * rsalveti (n=salveti@200.184.118.130) Quit (Client Quit)
  • [15:05:17] * rsalveti (n=salveti@200.184.118.130) has joined #beagle
  • [15:05:28] <MostAwesomeDude> KMS gives us a free fbcon. :3
  • [15:05:32] <geckosen1tor> I think it always was in the kernel?
  • [15:05:42] <geckosen1tor> what do you mean by free fbcon?
  • [15:05:49] <MostAwesomeDude> For omapfb, yes. Not for intel, radeon, etc.
  • [15:06:01] <MostAwesomeDude> And omapfb doesn't really have a super-special DDX driver, anyway.
  • [15:06:15] <geckosen1tor> well, hold on now..
  • [15:06:17] <MostAwesomeDude> fbcon, framebuffer console, gets full awesomeness if KMS is on.
  • [15:06:32] <geckosen1tor> to set the mode with any type of framebuffer, you set the varinfo with an ioctl
  • [15:06:51] <geckosen1tor> how is that not in kernel mode?
  • [15:07:25] * JimDonovan (n=jimd@c-66-30-221-249.hsd1.ma.comcast.net) has joined #beagle
  • [15:07:54] <geckosen1tor> that's how it's always been, and it works for radeon, vesa, matrox, and nvidia
  • [15:07:58] <MostAwesomeDude> Well, it has to do with how the driver's set up. KMS is a generic setup.
  • [15:08:02] <geckosen1tor> I'm pretty sure it works for others too
  • [15:08:17] <geckosen1tor> they also support setting it as a kernel parameter
  • [15:08:42] <MostAwesomeDude> More importantly, this lets radeon, matrox, nv/nouveau move mode-setting code for X *out of* the DDX drivers.
  • [15:08:55] <MostAwesomeDude> It's architectural.
  • [15:09:07] <MostAwesomeDude> The only reason I really bring it up, is because it's related to GEM.
  • [15:09:22] <geckosen1tor> I'm not talking about X at all
  • [15:09:36] <MostAwesomeDude> GEM is Intel's Graphics Execution Manager. It's a generic memory manager for DRMed cards.
  • [15:09:40] <koen> geckosen1tor: yes, those are applied to the 2.6.27 kernel
  • [15:09:53] <geckosen1tor> koen: the omap-linux kernel?
  • [15:10:05] <MostAwesomeDude> koen: Do you know about what I'm talking about, kinda? Maybe?
  • [15:10:10] <MostAwesomeDude> I don't wanna be rambling. :c
  • [15:11:03] <MostAwesomeDude> At any rate, the idea is that GEM handles allocating memory (framebufs, etc.) for the chipset, and KMS handles the mode-setting for it.
  • [15:11:10] <koen> geckosen1tor: the linux-omap_2.6.27.bb recipe in OE has those patches
  • [15:11:19] * geckosenator (n=sean@c-24-8-193-190.hsd1.co.comcast.net) Quit (Read error: 110 (Connection timed out))
  • [15:11:31] <MostAwesomeDude> fbcon can use KMS to get a console, and X can use it to get a graphical context, or whatever Intel's calling it.
  • [15:11:36] <geckosen1tor> koen: oh, so the recipe applies the patch, it isn't anything I can pull from git?
  • [15:11:53] <koen> right
  • [15:12:00] <geckosen1tor> koen: thanks, that clears it up
  • [15:12:04] <geckosen1tor> MostAwesomeDude: have you heard of egl?
  • [15:12:05] <MostAwesomeDude> koen: So is your patch just for the omapfb ioctls?
  • [15:12:17] <MostAwesomeDude> geckosen1tor: EGL? The OGL subset?
  • [15:12:51] <geckosen1tor> MostAwesomeDude: it allows you to have hardware accelerated opengl without using glx
  • [15:12:52] * fulgas is now known as FuL|OUT
  • [15:12:53] <koen> MostAwesomeDude: dunno, I only apply patches, I don't write them :)
  • [15:13:18] <MostAwesomeDude> geckosen1tor: Right, I'm aware of its existence.
  • [15:13:34] <MostAwesomeDude> Still need Mesa.
  • [15:13:37] <geckosen1tor> MostAwesomeDude: the idea is that modesetting is done through framebuffer ioctls
  • [15:13:52] <geckosen1tor> I"m not sure how x changes the mode normally
  • [15:14:22] <MostAwesomeDude> X normally sets the mode by asking the DDX to change it.
  • [15:14:32] <MostAwesomeDude> And then the DDX does mumbo-jumbo to change the mode.
  • [15:14:56] <geckosen1tor> ok
  • [15:15:03] <MostAwesomeDude> Only a couple DDX drivers actually use the framebuffer directly. Most drivers have some sort of command submission or management.
  • [15:15:20] <geckosen1tor> it's kind of ugly how it works
  • [15:15:24] <MostAwesomeDude> omapfb just happens to be one of those drivers where the FB is the only useful thing. :3
  • [15:15:43] <MostAwesomeDude> And yeah, we know it's ugly. KMS is undoing the mistakes of the XF86 years.
  • [15:15:43] * dirk2 (n=dirk@p5B040B65.dip0.t-ipconnect.de) has joined #beagle
  • [15:16:08] <geckosen1tor> well it seems to me like the modesetting should be done with the framebuffer
  • [15:16:15] <geckosen1tor> since this is already supported and it's in the kernel
  • [15:16:35] <MostAwesomeDude> And we arrive at the big problem. Both X and the FB driver wanna have modesetting control.
  • [15:16:56] <geckosen1tor> well X uses the fb driver
  • [15:17:01] <geckosen1tor> so it should be ok right?
  • [15:17:03] <MostAwesomeDude> So, rather than write a lot of BS code that does that in X, I'd rather do that in a KMS way.
  • [15:17:21] <geckosen1tor> yeah, just call the ioctl on the fb device
  • [15:17:22] <MostAwesomeDude> And yes, right now, X uses the FB driver. That will change the moment I figure out how to do some sort of accel.
  • [15:17:43] <geckosen1tor> are you working on direct rendering?
  • [15:18:06] <MostAwesomeDude> I will be, as soon as I get the black boxes I'm supposed to RE.
  • [15:18:21] <MostAwesomeDude> First, though, I think it'd be nice to have some basic accel, some basic infrastructure.
  • [15:18:37] <geckosen1tor> are you using the dsp for that?
  • [15:18:42] <MostAwesomeDude> Even cards that we have no docs for, like certain newer Radeons, can still have GEM+KMS, even if it doesn't do anything.
  • [15:19:06] <geckosen1tor> I would prefer you leave x to use the fb exactly how it does, and have the dsp render to the framebuffer through commands sent to it
  • [15:19:06] <MostAwesomeDude> Remind me of the Beagle's DSP. I'm still new to the HW. :3
  • [15:19:35] <geckosen1tor> I don't know what you were planning
  • [15:19:53] <MostAwesomeDude> That's ugly and slow. Doesn't the SGX GPU have the ability to accelerate OGL 2.1?
  • [15:20:41] <geckosen1tor> isn't video memory still ram?
  • [15:20:47] <geckosen1tor> I don't see how it would be any slower
  • [15:22:20] <geckosen1tor> MostAwesomeDude: what docs are you reading about the gpu support? I would like to take a look
  • [15:22:42] <MostAwesomeDude> geckosen1tor: I have no docs. This is an RE job.
  • [15:22:54] <geckosen1tor> you have a lot of acronyms I never heard of
  • [15:22:56] <MostAwesomeDude> The DSP looks like it can be used to accelerate video decoding.
  • [15:22:57] <geckosen1tor> what is an RE job?
  • [15:23:05] <MostAwesomeDude> Reverse-engineered. :3
  • [15:23:16] <geckosen1tor> the DSP should be able to do fragment shaders shouldn't it?
  • [15:24:05] <MostAwesomeDude> Probably, yes, but the SGX core is also used to make Intel GMA GPUs, or so I've gathered from a few buddies.
  • [15:24:21] <MostAwesomeDude> And I have a self-documented OGL 2.0 driver for *that* hardware.
  • [15:25:00] <geckosen1tor> I'm happy with software rendering for now :-P
  • [15:25:24] <MostAwesomeDude> I know. This is not something essential. This is something fun, and something we can use to get/keep students interested.
  • [15:25:51] <MostAwesomeDude> It's also kinda sad to know that a 10-million+ polygon engine is sitting idle.
  • [15:26:13] <geckosen1tor> aren't there docs on how to use it though?
  • [15:26:25] <geckosen1tor> are you saying the intel one is the same?
  • [15:27:13] <MostAwesomeDude> Well, Intel paid Tungsten Graphics to write those drivers, and then asked them to not open the docs.
  • [15:27:32] <MostAwesomeDude> But the Tungsten guys, being the cool cats they are, wrote the important docs into the driver.
  • [15:28:05] <MostAwesomeDude> And the cores are similar (if not identical), but the registers and such might be subtly different. That's where mmiotrace comes in. It lets me see what's writing, and to where.
  • [15:28:16] * Openfree (n=df@222.65.143.118) Quit (Read error: 110 (Connection timed out))
  • [15:29:22] <MostAwesomeDude> I'm not saying this is the kind of thing where four lines of glue code will turn it on magically, but it's not like I'm completely in the dark, like the nouveau guys.
  • [15:29:44] <geckosen1tor> I didn't realize there weren't docs
  • [15:30:36] <MostAwesomeDude> Well, Intel's docs are under NDA, as are some Radeon and VIA docs.
  • [15:31:02] <geckosen1tor> I thought intel's graphics were open
  • [15:31:03] <MostAwesomeDude> I'm not really quite sure who figured out the omapfb stuff for Xv, although he did a good job, whoever it is.
  • [15:31:10] <MostAwesomeDude> Hmm, it *is* open.,
  • [15:31:17] <MostAwesomeDude> Open source, not open hardware.
  • [15:31:47] <geckosen1tor> that doesn't make sense to me why they do that
  • [15:32:01] <geckosen1tor> I would rather just implement hardware acceleration in the dsp :-P
  • [15:32:09] <MostAwesomeDude> Go for it.
  • [15:32:21] <geckosen1tor> heh, mode setting and svideo first
  • [15:32:25] <MostAwesomeDude> I mean, I'm sure it's possible to move most of the OGL pipeline to the DSP.
  • [15:32:52] <geckosen1tor> can't gcc compile to the dsp?
  • [15:32:56] <MostAwesomeDude> But I'd rather use the actual 3D engine for that fun stuff, and reserve the DSP for video and whatever else people have in mind.
  • [15:33:33] <geckosen1tor> whatever the 3d hardware doesn't support
  • [15:33:38] <geckosen1tor> like fragment shaders
  • [15:34:12] <_AV500_> the 3d engine support 2 kind of shaders, vertex and the other one
  • [15:34:43] <geckosen1tor> vertex and fragment
  • [15:35:15] <MostAwesomeDude> If the 3D HW is what I've heard it is, it supports a full OGL 2.1/Dx 9.0 pipeline.
  • [15:35:23] <_AV500_> at least I remember that from some TI presentation
  • [15:35:45] <MostAwesomeDude> Polys, shaderful TCL, multitex, shaderful fragments, fog, etc.
  • [15:36:26] <geckosen1tor> MostAwesomeDude: maybe you should ask TI for the docs
  • [15:36:47] <geckosen1tor> can it render to svideo?
  • [15:37:03] <MostAwesomeDude> geckosen1tor: They're going to get me the kernel drivers (open) and the DDX/Mesa drivers (closed).
  • [15:37:25] * Openfree (n=df@222.65.143.118) has joined #beagle
  • [15:37:28] <geckosen1tor> I'm at the point where I don't think I want it if it's closed
  • [15:37:41] <_AV500_> ok, found the doc, it says vertex AND fragment shader
  • [15:37:43] <MostAwesomeDude> Heh, neither do I. That's why I'm reversing it.
  • [15:37:53] <MostAwesomeDude> _AV500_: Is it just publicity fluff, or does it have meat?
  • [15:38:36] <_AV500_> it is under NDA
  • [15:39:09] <geckosen1tor> but would it be as fast as running the shaders in the dsp?
  • [15:39:53] <MostAwesomeDude> Faster.
  • [15:39:59] <_AV500_> Triangle Rate ** ~10 MPoly / Sec
  • [15:40:13] <_AV500_> Effective Fill Rate * 275 Mpix / sec
  • [15:40:20] <MostAwesomeDude> _AV500_: I may have to be extremely ticked at you for having docs.
  • [15:41:04] <_AV500_> MostAwesomeDude: it is easy, just buy a few 100k of chips
  • [15:41:45] <MostAwesomeDude> _AV500_: We're building our own. The Beagle just happens to have nearly the same chipsets.
  • [15:41:49] <MostAwesomeDude> But yeah.
  • [15:42:02] <MostAwesomeDude> I suppose you're not allowed to write code, then, either?
  • [15:42:09] <_AV500_> ?
  • [15:42:43] <MostAwesomeDude> The terms of the NDA.
  • [15:43:02] <_AV500_> no idea, it is for the company, does not name me :-)
  • [15:43:12] <MostAwesomeDude> XD
  • [15:43:21] <MostAwesomeDude> Well, I'm not going to ask you to do anything...silly.
  • [15:43:41] <MostAwesomeDude> Really, RE is kinda fun, in a masochistic kind of way.
  • [15:44:14] <MostAwesomeDude> Anyway, out of curiosity, who authored the docs?
  • [15:44:31] <MostAwesomeDude> My guess is TI, but it'd be nice to know who else may have been involved.
  • [15:44:35] <_AV500_> isnt the idea about OGL standard that there is nothing to RE? you just write using the OGL api and the driver/HW does the rest,no?
  • [15:44:56] <geckosen1tor> _AV500_: he is talking about writing the driver
  • [15:45:00] <MostAwesomeDude> _AV500_: Yes.
  • [15:45:04] <MostAwesomeDude> But we have no driver. :3
  • [15:45:19] <geckosen1tor> hey, we always have software mesa
  • [15:45:37] <MostAwesomeDude> The chain is, app->GLX->Mesa->card-specific driver->card.
  • [15:45:48] <MostAwesomeDude> Or, with new stuff, app->Gallium->Mesa...
  • [15:45:55] <_AV500_> yes, but driver will be there eventually, no?
  • [15:46:38] <MostAwesomeDude> Yes, as soon as I (or somebody else) reverse the closed-source driver and add it to Mesa.
  • [15:47:20] <geckosen1tor> yay, just built uboot on the beagle
  • [15:55:17] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [16:00:15] * favor (n=favor@118.113.207.154) Quit ("Leaving")
  • [16:00:25] <koen> jkridner: http://dominion.thruhere.net/koen/cms/pico-dlp-arrived
  • [16:05:41] * n6pfk (n=mike@c-76-104-40-104.hsd1.va.comcast.net) has joined #beagle
  • [16:08:01] * fr3men (n=fr3men@ip-5-146.sn2.eutelia.it) has joined #beagle
  • [16:12:05] <and-r1> what is wrong, if i get only some trash on the screen.....i mean: i try to use hdmi to lcd display but when the kernel starts i get only a lot of data-trash on the screen
  • [16:13:18] <simon42> and-r1: you use the hdmi input on your screen?
  • [16:14:26] <and-r1> i use hdmi from beagle to dvi on lcd
  • [16:15:28] <and-r1> it must be a kernel issue, cause other kernel works fine...but i need to work with this kernel 2.6.27-omap1
  • [16:15:54] <and-r1> but he only print trash on the screen
  • [16:16:41] * robtow (n=rtow@nat/montavista/x-05ad05f3d240db90) has joined #beagle
  • [16:19:22] <simon42> uboot/xloader up to date?
  • [16:20:15] <geckosen1tor> and the lcd works from other inputs besides the beagle?
  • [16:20:32] * philippe (n=fille@a83-245-252-47.elisa-laajakaista.fi) Quit (No route to host)
  • [16:22:08] * JimDonovan (n=jimd@c-66-30-221-249.hsd1.ma.comcast.net) Quit (Read error: 110 (Connection timed out))
  • [16:23:10] <and-r1> lcd works
  • [16:23:28] <and-r1> how can i find out the uboot version?
  • [16:24:14] <and-r1> oh i see, at the beginning he said uboot 1.3.3 ...hmm i think it is a little bit to old
  • [16:27:57] * TAK2004 (n=Administ@dslb-088-074-046-227.pools.arcor-ip.net) has joined #beagle
  • [16:31:09] * MostAwesomeDude is now known as MAD|adventures
  • [16:41:19] * florian (n=fuchs@217.146.132.69) Quit (Remote closed the connection)
  • [16:50:44] * atin_ (n=atin@c-71-232-7-71.hsd1.ma.comcast.net) has joined #beagle
  • [16:52:02] <atin_> jkridner, sakoman: I wrote up the mods for using a local overlay package with openembeded: http://elinux.org/BeagleBoardOpenEmbeddedDevelopment
  • [17:03:35] * RogerMonk (n=a0740758@nat/ti/x-03a4edb2bc116938) has left #beagle
  • [17:16:33] * sjam (n=sojan_ja@121.247.121.94) has joined #beagle
  • [17:18:20] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) has joined #beagle
  • [17:18:54] * sjam slaps sjam around a bit with a large trout
  • [17:19:00] * sjam slaps sjam around a bit with a large trout
  • [17:19:11] * mkhhh_ is now known as markus
  • [17:28:09] * fr3men (n=fr3men@ip-5-146.sn2.eutelia.it) has left #beagle
  • [17:31:19] * sjam (n=sojan_ja@121.247.121.94) Quit (Read error: 60 (Operation timed out))
  • [17:32:04] * uwe_ (n=uwe_@dslb-088-064-102-038.pools.arcor-ip.net) Quit (Read error: 60 (Operation timed out))
  • [17:32:17] * uwe__ (n=uwe@dslb-088-064-096-254.pools.arcor-ip.net) has joined #beagle
  • [17:34:30] * magnet_ (n=magnet@AMontpellier-259-1-21-168.w92-133.abo.wanadoo.fr) has joined #beagle
  • [17:36:07] * DVbeagle1 (n=dvescovi@cpe-72-184-136-6.tampabay.res.rr.com) has joined #beagle
  • [17:43:50] <DVbeagle1> I am hoping Mr. Sakoman can answer a question I have on the Overo ..are you on line?
  • [17:43:51] * uwe2 (n=uwe@dslb-088-064-102-038.pools.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
  • [17:44:05] <sakoman> yes
  • [17:45:19] <DVbeagle1> I am trying to debug the the OTG port ID detection .. on the Submmit it shows OTG ID going to GPIO 41
  • [17:46:02] * zedstar (n=john@fsf/member/zedstar) Quit ("Leaving")
  • [17:46:12] <DVbeagle1> But I question this ... I do not seem to detect a change of state when configured as input
  • [17:47:22] <DVbeagle1> do you have a way to verify if this is indeed wired up correctly on the motherboard? I know Beagle used the 4030 for host ID detection
  • [17:47:53] * and-r1 (n=ubuntu@pd95b4498.dip0.t-ipconnect.de) Quit (Read error: 110 (Connection timed out))
  • [17:49:30] <sakoman> DVbeagle1: I will check into this for you
  • [17:49:33] * magnet (n=magnet@AMontpellier-259-1-45-128.w92-133.abo.wanadoo.fr) Quit (Read error: 110 (Connection timed out))
  • [17:50:26] <sakoman> We should probably handle questions like this on #gumstix so as to not clog this channel with non-beagle chat
  • [17:51:21] <DVbeagle1> will do ..didn't know there was a gumstix channel THX
  • [17:52:05] * valhalla (n=valhalla@81-174-21-31.dynamic.ngi.it) has joined #beagle
  • [17:53:11] * ldesnogu_ (n=ldesnogu@ven06-2-82-247-86-183.fbx.proxad.net) has joined #beagle
  • [18:02:02] * Beagle8 (n=Beagle8@rrcs-76-79-16-226.west.biz.rr.com) has joined #beagle
  • [18:02:16] * Beagle8 is now known as mike18
  • [18:03:12] <mike18> I have Angstrom up and running on my BeagleBoard using the wiki HowTo page. I'm wondering if I can get ipkg packages without connecting my BB up to the internet
  • [18:03:29] <mike18> Either run an Angstrom VM from my desktop or download packages in windows from a website?
  • [18:04:11] <mike18> I suppose this would be a better questions for #Angstrom
  • [18:05:29] <Crofton|work> koen, ping
  • [18:06:18] * vlad (n=asdf@69-12-240-128.dsl.static.sonic.net) Quit (Nick collision from services.)
  • [18:06:18] * fraz__ (n=fraz@host143186050011.borland.com) has left #beagle
  • [18:06:21] * vlad (n=asdf@69-12-240-128.dsl.static.sonic.net) has joined #beagle
  • [18:06:30] * vlad (n=asdf@69-12-240-128.dsl.static.sonic.net) Quit (Nick collision from services.)
  • [18:07:59] <sakoman> DVbeagle1: I responded on #gumstix - I think I see you there :-)
  • [18:11:22] * vlad (n=asdf@69-12-240-128.dsl.static.sonic.net) has joined #beagle
  • [18:13:53] * guillaum1 (n=GZBA4143@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [18:15:04] * uwe_ (n=uwe_@dslb-088-064-096-254.pools.arcor-ip.net) has joined #beagle
  • [18:15:53] * guillaum1 (n=gl@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) has joined #beagle
  • [18:19:35] <koen> Crofton|work: pong
  • [18:19:53] <Crofton|work> I think I have worked it out
  • [18:20:06] <Crofton|work> sffsdr patch does not work against .27
  • [18:20:07] <koen> worked out what?
  • [18:20:16] <Crofton|work> sffsdr kernel build :)
  • [18:20:23] <Crofton|work> you warned me about it a coupel of weeks ago
  • [18:20:37] <Crofton|work> setting machine file to call for .25
  • [18:21:02] <Crofton|work> remove patch from versions it will never work for etc
  • [18:21:16] <Crofton|work> Hugo is working on an updated patch :)
  • [18:22:46] * MAD|adventures is now known as MostAwesomeDude
  • [18:23:05] * dioioib (n=dioioib@d121-173-32.home3.cgocable.net) has joined #beagle
  • [18:38:42] * gletelli__ (n=chatzill@fw-tnat.cambridge.arm.com) Quit ("ChatZilla 0.9.84 [Firefox 3.0.2/2008091620]")
  • [18:40:52] <DJWills> sakoman: did you sort that nautilus build issue you had some time ago?
  • [18:41:59] <sakoman> DJWills: No, I had to move on to working on some new prototype hw
  • [18:42:37] * chelli (n=chelli@debian/developer/tschmidt) has joined #beagle
  • [18:42:43] <sakoman> DJWills: are you having trouble too?
  • [18:49:51] * gletelli (n=chatzill@fw-tnat.cambridge.arm.com) has joined #beagle
  • [18:50:51] <DJWills> sakoman: exactly the same issue
  • [18:51:25] <sakoman> Bummer. It will likely be a few more days before I can get back to that
  • [18:54:11] * magnet_ (n=magnet@AMontpellier-259-1-21-168.w92-133.abo.wanadoo.fr) Quit ("Lost terminal")
  • [18:54:36] * magnet (n=magnet@AMontpellier-259-1-21-168.w92-133.abo.wanadoo.fr) has joined #beagle
  • [18:58:02] <DJWills> sakoman: no probs, just trying a new checkout of bitbake as I have a small known issue with the version I have and GIT but I suspect that is not the issue.
  • [19:03:57] * JustinLove (n=J_LOVE@64-135-210-33.FoxValley.net) has joined #beagle
  • [19:04:41] * florian (n=fuchs@g227179005.adsl.alicedsl.de) has joined #beagle
  • [19:05:23] * MostAwesomeDude is now known as MAD|munchies
  • [19:10:53] * Leon_Nardella (n=Leon@200-161-14-111.dsl.telesp.net.br) has joined #beagle
  • [19:17:36] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) has joined #Beagle
  • [19:26:57] * FuL|OUT is now known as fulgas
  • [19:29:01] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) Quit (Remote closed the connection)
  • [19:30:45] * zedstar (n=john@fsf/member/zedstar) has joined #beagle
  • [19:31:38] <mike18> can someone point me to which ipkg i'll want for a c++ compiler on the BeagleBoard?
  • [19:32:53] * fille_ (n=fille@a83-245-252-47.elisa-laajakaista.fi) has joined #beagle
  • [19:37:21] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) has joined #beagle
  • [19:38:13] * fille_ is now known as philippe
  • [19:39:20] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) Quit (Client Quit)
  • [19:41:45] <JustinLove> I've been getting SIGBUS accessing gpio register ever since trying a new kernel. I turned on FCLKEN, and ICLKEN; IDLEST is clear, but it still errors out trying to read a gpio register. What am I missing?
  • [19:42:26] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) has joined #beagle
  • [19:42:53] * dcordes (n=dcordes_@unaffiliated/dcordes) has joined #beagle
  • [19:50:19] <atin_> hci_recv_fragment: hci0 no memory for packet
  • [19:50:19] <atin_> DMA misaligned error with device 0
  • [19:50:27] <atin_> cwap/
  • [20:01:55] <dcordes> koen, but I can't see no armv6-novp packages in the feeds
  • [20:03:22] * _AV500_ (n=AV500@IP-213157024130.static.heagmedianet.de) Quit (Remote closed the connection)
  • [20:04:52] * _apk (n=kows@85-18-136-98.fastres.net) has joined #beagle
  • [20:06:50] * Leon_Nardella (n=Leon@200-161-14-111.dsl.telesp.net.br) Quit (Read error: 104 (Connection reset by peer))
  • [20:14:57] * dirk2 (n=dirk@p5B040B65.dip0.t-ipconnect.de) has left #beagle
  • [20:14:58] * valhalla (n=valhalla@81-174-21-31.dynamic.ngi.it) Quit ("Leaving")
  • [20:15:45] <JustinLove> Huh... in the older kernel the whole FCLKEN was enabled. Restoring all the bits I'm not using makes the modules I am using work....?
  • [20:18:52] <atin_> cwap - msub_giveback() crashing in the kernel when I plug in my bluetooth dongle.
  • [20:19:32] <atin_> I found something on the mailing list archives that had koen's kernel working...maybe I need r27.
  • [20:20:26] * guillaum1 (n=gl@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [20:20:47] * guillaum1 (n=Guillaum@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) has joined #beagle
  • [20:26:50] <koen> Crofton|work: ping
  • [20:26:59] <Crofton|work> pong
  • [20:37:09] <dcordes> koen, ping
  • [20:37:17] <koen> pong
  • [20:37:40] <dcordes> can't find armv6-novp package builds
  • [20:38:20] <dcordes> I guess nobody is in desperate need of them but you told me yesterday they are supposed to be present.
  • [20:38:51] <koen> angstrom@serenity:~/website/feeds/2008/ipk/glibc/unsorted$ find ../armv6-novfp/ | wc -l
  • [20:38:54] <koen> 5296
  • [20:39:06] <koen> over 5k of armv6-novfp packages
  • [20:39:55] <dcordes> oh
  • [20:41:46] <dcordes> koen, can I add some X image builds for htcraphael and htcdiamond machines when we sorted TS driver for use with tslib?
  • [20:41:55] <koen> sure
  • [20:47:41] <Crofton> koen, where do those come from and what are they?
  • [20:58:03] <geckosen1tor> does anyone know how to invoke mkimage on the Image file created when compiling the kernel to create an image uboot can boot?
  • [20:58:50] <dcordes> geckosen1tor, I had to install mkimage in debian
  • [20:59:07] <adj_> geckosen1tor: take a look at kernel's makefile
  • [20:59:12] <geckosen1tor> dcordes: I have compiled mkimage
  • [20:59:50] <dcordes> geckosen1tor, so you have it in PATH ?
  • [20:59:54] <dcordes> make should take care
  • [21:00:01] <dcordes> iirc
  • [21:00:08] <geckosen1tor> I was going to run mkimage directly
  • [21:00:16] * robtow1 (n=rtow@63.81.120.151) has joined #beagle
  • [21:00:19] <geckosen1tor> are you suggesting something like "make uImage" ?
  • [21:00:20] <dcordes> ok no clue sorry.
  • [21:00:23] <dcordes> yea
  • [21:00:27] <geckosen1tor> ok, let me try that
  • [21:01:16] <geckosen1tor> I'm doing it on the beagle so it takes a minute
  • [21:02:10] <geckosen1tor> am I the only one who gets a bunch of warnings like
  • [21:02:15] <geckosen1tor> <stdin>:1325:2: warning: #warning syscall ppoll not implemented
  • [21:02:21] * robtow (n=rtow@nat/montavista/x-05ad05f3d240db90) Quit (Read error: 60 (Operation timed out))
  • [21:02:57] <adj_> no, I get those too
  • [21:03:08] <geckosen1tor> ok, it says uImage ready
  • [21:03:12] <geckosen1tor> let me try it
  • [21:03:27] * robtow (n=rtow@63.81.120.151) has joined #beagle
  • [21:03:40] <koen> Crofton: what comes from where?
  • [21:04:39] <Crofton|work> the arm?-novfp
  • [21:05:04] <geckosen1tor> hmm, it is taking like 30 seconds to copy a 2.2meg kernel :-/
  • [21:05:07] * ssvb (n=ssvb___@a88-114-221-132.elisa-laajakaista.fi) has joined #beagle
  • [21:05:17] <koen> as with all angstrom packages, the autobuilder builds and uploads those
  • [21:06:39] <geckosen1tor> can someone try copying a file to the fat partition on the beagle and see how fast it is?
  • [21:06:59] <geckosen1tor> it's like 10kb/s for me
  • [21:07:46] <koen> mount /media/mmcblk0p1 -o remount,async
  • [21:08:17] <geckosen1tor> that made a very huge difference in speed
  • [21:08:33] * robtow (n=rtow@63.81.120.151) Quit ("Leaving.")
  • [21:08:35] <geckosen1tor> thanks.. is there a reason it isn't default?
  • [21:08:46] * robtow (n=rtow@63.81.120.151) has joined #beagle
  • [21:09:37] <koen> data integrity
  • [21:10:41] <koen> I think newer kernels have an option that gives better write performance, but I'm fuzzy on the details
  • [21:10:53] <geckosen1tor> even with sync?
  • [21:11:09] <geckosen1tor> hmm, maybe that is my problem now, it might not have flushed
  • [21:14:56] <geckosen1tor> no, still gets stuck at "Waiting for root device /dev/mmcblk0p2..."
  • [21:18:40] * robtow1 (n=rtow@63.81.120.151) Quit (Read error: 113 (No route to host))
  • [21:21:02] * dcordes (n=dcordes_@unaffiliated/dcordes) Quit (Remote closed the connection)
  • [21:21:11] * Beagle8 (n=Beagle8@cpe-70-112-199-85.austin.res.rr.com) has joined #beagle
  • [21:21:53] * Beagle8 is now known as GimpMaster
  • [21:22:09] * dcordes (n=dcordes_@unaffiliated/dcordes) has joined #beagle
  • [21:28:18] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) Quit (Remote closed the connection)
  • [21:39:34] <JustinLove> Well, my latest system build made mplayer much more reliable.
  • [21:39:47] <JustinLove> It crashes the system the second time, every time ;^)
  • [21:40:15] <mru> JustinLove: audio?
  • [21:40:32] <JustinLove> The error messages do seem to indicate the audio system.
  • [21:40:49] <JustinLove> And when I'm running it from my program, it keeps running, but there is no audio
  • [21:44:46] <JustinLove> After that, I can run it repeatedly without hanging the system, but no audio.
  • [21:44:54] <JustinLove> Running a file without audio works fine.
  • [21:47:04] * dcordes (n=dcordes_@unaffiliated/dcordes) Quit ("Verlassend")
  • [21:47:53] * lgentili (i=c82be332@gateway/web/ajax/mibbit.com/x-dd6886f00ac1c649) has joined #beagle
  • [21:48:22] <lgentili> another non-intel netbook!: http://news.cnet.com/8301-13924_3-10123149-64.html
  • [21:50:06] * DVbeagle1 (n=dvescovi@cpe-72-184-136-6.tampabay.res.rr.com) Quit ()
  • [21:50:20] <lgentili> use Qualcomm QSD8672 dual-core Snapdragon at 1.5Ghz (ARM Architecture!!).
  • [21:52:04] * alecrim (n=alecrim@189.2.128.130) Quit (Remote closed the connection)
  • [21:54:00] * lgentili (i=c82be332@gateway/web/ajax/mibbit.com/x-dd6886f00ac1c649) Quit ("http://www.mibbit.com ajax IRC Client")
  • [21:54:23] * mike18 (n=Beagle8@rrcs-76-79-16-226.west.biz.rr.com) Quit ()
  • [21:56:07] <JustinLove> I don't suppose anybody made progress on IRQ -33 on other fronts?
  • [22:05:20] <Crofton|work> I haven't seen that in ages ....
  • [22:09:31] <JustinLove> The strongly-ordered patch made things much better, but it will usually crash if left overnight. The latest kernel/mplayer has different problems.
  • [22:19:05] <JustinLove> While my kernel recompiles...
  • [22:19:20] <JustinLove> Anybody know an easy way to turn off the blinking cursor on the display?
  • [22:25:49] * philippe (n=fille@a83-245-252-47.elisa-laajakaista.fi) Quit (Read error: 60 (Operation timed out))
  • [22:30:17] * loki_val (n=pa@gentoo/developer/loki-val) has joined #beagle
  • [22:30:42] * loki_val (n=pa@gentoo/developer/loki-val) has left #beagle
  • [22:31:52] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) Quit ()
  • [22:34:08] * abitos (n=nixgibts@dslb-084-057-206-030.pools.arcor-ip.net) Quit (Read error: 113 (No route to host))
  • [22:45:51] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) Quit (Read error: 110 (Connection timed out))
  • [22:49:01] * dioioib (n=dioioib@d121-173-32.home3.cgocable.net) Quit ()
  • [23:04:32] * robtow1 (n=rtow@nat/montavista/x-13e0ef867e46894b) has joined #beagle
  • [23:06:08] <ds2> how many people are using ABS filament?
  • [23:07:07] * cbrake is now known as cbrake_away
  • [23:07:27] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has left #beagle
  • [23:12:52] * robtow (n=rtow@63.81.120.151) Quit (Read error: 148 (No route to host))
  • [23:16:21] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) has joined #beagle
  • [23:16:22] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
  • [23:17:14] * Beagle3 (n=Beagle3@68-114-156-188.dhcp.kgpt.tn.charter.com) has joined #beagle
  • [23:17:31] <Beagle3> i have a question
  • [23:17:58] <geckosen1tor> koen: none of the video modes work with my monitor
  • [23:18:03] <geckosen1tor> Beagle3: ...
  • [23:18:08] <Beagle3> can the beagle board be connect to this http://www.sparkfun.com/commerce/product_info.php?products_id=8511
  • [23:19:17] <geckosen1tor> Beagle3: you would have to do some work
  • [23:19:50] <Beagle3> for example?
  • [23:20:09] <geckosen1tor> well the easy way is to connect it to the arm9 one
  • [23:20:18] <geckosen1tor> then connect the arm9 to the beagle over usb ethernet
  • [23:20:34] * chelli (n=chelli@debian/developer/tschmidt) Quit ("IRC is just multiplayer notepad")
  • [23:20:41] <geckosen1tor> then run servers to play sound and mount the mini sd over nfs
  • [23:20:51] <geckosen1tor> that requires the least hardware work
  • [23:21:13] <geckosen1tor> also run ip tables with forwarding to make the ethernet work
  • [23:21:43] <ds2> easier to build your own then touse that SFE board
  • [23:22:22] <Beagle3> my idea is to use the pc cog and have it attached to a touch screen and that hole component attach to the beagle board to create a moble touch computer
  • [23:22:50] <geckosen1tor> but no arm9?
  • [23:22:59] <geckosen1tor> since the cog already plugs right into it
  • [23:23:36] <geckosen1tor> it only consumes 150mW
  • [23:24:04] <ds2> why not just attach the touch screen to a TSC2046 and talk to it from the expansion header McSPI?
  • [23:25:04] <geckosen1tor> it's more software to write
  • [23:25:09] <ds2> HUH?
  • [23:25:26] <ds2> TSC2046 support is in there already... it is used by Zoom and SDP
  • [23:25:38] <ds2> only difference might be the McSPI to use
  • [23:25:43] <geckosen1tor> ok
  • [23:25:51] <geckosen1tor> but what about all the other stuff?
  • [23:26:06] <Beagle3> is the arm9 samething as the beagle?
  • [23:26:08] <ds2> most of it seems redundant unless really high quality audio is needed
  • [23:26:09] <geckosen1tor> like the ethernet and lcd
  • [23:26:17] <geckosen1tor> and audio
  • [23:26:19] <Crofton|work> does anyone know where I can find examples that use the gadgetfs?
  • [23:26:26] * n6pfk (n=mike@c-76-104-40-104.hsd1.va.comcast.net) Quit ("Leaving")
  • [23:26:48] <ds2> Crofton|work: please post if you find good examples... i think there is a page but I don't recall seeing code
  • [23:27:13] <ds2> geckosen1tor: SPI ethernet... LCD... well, isn't that what you are doing? =)
  • [23:27:34] * ldesnogu_ (n=ldesnogu@ven06-2-82-247-86-183.fbx.proxad.net) Quit ()
  • [23:27:35] <Crofton|work> will do
  • [23:27:38] <geckosen1tor> ds2: heh, true
  • [23:27:49] <Crofton|work> anyone using spi on the beagle either?
  • [23:27:53] <ds2> Crofton|work: what are you trying to emulate?
  • [23:28:01] <Crofton|work> USRP :)
  • [23:28:14] <ds2> Crofton|work: there is Kevin on the mailing list but he's having issues with the RX
  • [23:28:20] <ds2> Kevin U that is
  • [23:28:30] <ds2> why not do that in the kernel?
  • [23:28:46] <ds2> does the USRP use an ISOC EP or just a Bulk EP?
  • [23:28:55] <Crofton|work> bulk I think
  • [23:29:04] <Crofton|work> no iso-synchronous
  • [23:29:15] <ds2> In that case, I'd say, hack up g_zero or g_ether...
  • [23:29:40] <Crofton|work> where is the code for those?
  • [23:29:42] <ds2> g_zero can register a misc character device to talk to the userland
  • [23:29:57] <ds2> drivers/usb/gadget/{zero,ether}.c, IIRC
  • [23:30:24] <Crofton|work> hmm this is in kernel though
  • [23:30:25] <ds2> the zero.c cold is pretty straight forward... at least as of 2.6.24 O:-)
  • [23:30:28] <ds2> yep.
  • [23:30:31] <ds2> compile them as modules
  • [23:30:45] <koen> geckosen1tor: is your uboot up to date? older uboots misconfigured a pll so you couldn't get the rate you needed
  • [23:31:00] <geckosen1tor> koen: the pixel clock is 72mhz
  • [23:31:04] <ds2> in the init, add a misc_register() to create a /dev/foo device; then handle I/O from the read()/write() callbacks
  • [23:31:15] <geckosen1tor> well, up to 72mhz
  • [23:31:25] <geckosen1tor> koen: I have a pretty new uboot
  • [23:32:24] <geckosen1tor> I can verify the resolution with fbset, but for whatever reason the monitor doesn't like it
  • [23:36:37] * JustinLove (n=J_LOVE@64-135-210-33.FoxValley.net) has left #beagle
  • [23:37:42] <geckosen1tor> koen: do you use linux-omap-2.6.27 in openembedded?
  • [23:37:49] <geckosen1tor> or linux-omap2-git?
  • [23:39:04] <koen> I'm currently using 2.6.27 to test the SGX and DSP
  • [23:39:41] <geckosen1tor> it's hard to keep track of which branches are used for what
  • [23:39:54] * TAK2004 (n=Administ@dslb-088-074-046-227.pools.arcor-ip.net) Quit (Read error: 131 (Connection reset by peer))
  • [23:40:00] * koen uses no branches
  • [23:40:06] <ds2> koen: does your SGX version cooperate with omapfb?
  • [23:40:26] <koen> ds2: it seems to
  • [23:40:43] <koen> ds2: I had to apply a small patch to omapfb
  • [23:41:49] <ds2> koen: do you know the pedegree if that code?
  • [23:41:56] <koen> I'm using nokia's version of the sgx kernel drivers, which seems to be modified to work with mainline code instead of TI code
  • [23:42:34] <ds2> and I am wondering where did Nokia sourced that
  • [23:42:53] * koen stabs holiday season for making it impossible to get NDAs signed in a timely fashion
  • [23:42:57] <ds2> cuz there exists at least one other version that has been modified to run with mainline code
  • [23:44:27] * rsalveti (n=salveti@200.184.118.130) Quit (Remote closed the connection)
  • [23:47:49] <geckosen1tor> could someone please explain the difference between the current state of linux-omap-2.6.26 linux-omap-2.6.27 and linux-omap2-git in open embedded? thanks
  • [23:48:27] <koen> two of those are 2.6.26 the other one is 2.6.27
  • [23:48:54] <ds2> heheh
  • [23:49:16] <geckosen1tor> that explains a bit already
  • [23:49:19] <koen> and linux-omap_git is 2.6.26rc7
  • [23:49:25] <koen> ehm
  • [23:49:28] <koen> 28rc7
  • [23:51:21] <koen> zuh: I wonder if xf86-video-omapfb should allocate a 1280x720 overlay so mediaplayers like mplayer can output a 720p movie and have DSS downscale it to teh resolution of /dev/fb0
  • [23:52:14] <geckosen1tor> maybe you could allow mplayer to set the overlay resolution to whatever it wants
  • [23:52:15] <koen> zuh: case in point: the pico DLP needs 640x480 and most mediaplayers refuse to play source files wider than the output screen
  • [23:52:28] <koen> I don't want mplayer touching the overlay
  • [23:52:40] <koen> I want XV to do the heavy lifting
  • [23:52:59] <mru> something has to tell xv what to do
  • [23:53:32] <koen> right, but mplayer doesn't need to know that's it's using xv on beagle, only that it's using xv :)
  • [23:53:44] <mru> true
  • [23:54:31] <koen> it looks like omapfbplay just crops the image to the right dimensions
  • [23:55:12] <ds2> is there enough memory bandwidth to output in 720p and havethe DSS DMA it back out?
  • [23:55:40] * Beagle3 (n=Beagle3@68-114-156-188.dhcp.kgpt.tn.charter.com) Quit ()
  • [23:57:38] <ds2> what would be nicer for the picoDLP situation is if the FB can be setup report a fixed resolution and have the output resolution variable depending on the device; maybe configureable by sysfs
  • [23:58:06] <koen> I think DSS2 can do that
  • [23:58:23] * maelcum (n=quassel@e178167126.adsl.alicedsl.de) Quit (Read error: 110 (Connection timed out))
  • [23:58:38] <ds2> would be cooler even if it can do that scaling automatically for the S-Video port
  • [23:58:47] * maelcum (n=quassel@e178136103.adsl.alicedsl.de) has joined #beagle
  • [23:58:59] <koen> I know for sure DSS2 can scale for svideo
  • [23:59:15] <ds2> I would really like to do a side by side comparism between a 'toy quality' S-video input lcd based projector and the picoDLP
  • [23:59:27] <koen> that's how I did 1024x768 on my dvi screen and have dss scale it to svideo
  • [23:59:33] * robtow (n=rtow@nat/montavista/x-6ce4c4d0051e6c6f) has joined #beagle
  • [23:59:40] <ds2> it can vs there eixsts code to do it are 2 different things :(