• [00:02:15] * bazbell (n=a0192809@nat/ti/x-9dc072b03c588db8) Quit ("Leaving.")
  • [00:03:57] * amit (i=816ef12f@gateway/web/ajax/mibbit.com/x-bff8aafa60a019be) has joined #beagle
  • [00:04:13] <amit> hi
  • [00:04:25] * amit is now known as Guest4565
  • [00:05:15] <Guest4565> i am new to beagle board and trying to get angstom demo working and has a small problem .
  • [00:06:02] <Guest4565> i am trying to download MLO file and having trouble .... can anyone say m e how to download MLO file
  • [00:06:45] <Guest4565> hello is anyone online
  • [00:07:01] <mru> yes, there are people here
  • [00:07:09] <mru> maybe you could be a bit more specific about your probelm
  • [00:07:46] <Guest4565> ok when i click on MLO file name ... instead of asking me to say the file .... file directly opens
  • [00:08:05] <mru> right-click and choose "save link as"
  • [00:08:16] <mru> or something similar depending on your browser
  • [00:08:24] <Guest4565> i tried that ... but it saves the file in html format
  • [00:08:31] <mru> what browser?
  • [00:08:42] <mru> and which page are you downloading it from?
  • [00:09:06] <Guest4565> http://www.angstrom-distribution.org/demo/beagleboard/
  • [00:09:12] <Guest4565> i am using this link
  • [00:09:32] <Guest4565> working on Mozilla Firefox on ubuntu
  • [00:09:48] <mru> works fine here
  • [00:10:29] <mru> if firefox isn't playing nice, try wget or something
  • [00:11:01] <Guest4565> ok will try that
  • [00:11:13] <Guest4565> thanks for ur help
  • [00:12:46] * ScriptRipper (n=martin@opensuse/member/MartinMohring) Quit (Remote closed the connection)
  • [00:15:50] <Guest4565> hi
  • [00:19:52] * flameman (n=toor@net-93-65-1-167.cust.dsl.vodafone.it) Quit (Read error: 113 (No route to host))
  • [00:25:25] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-49cce26f5e32c713) Quit ("http://www.mibbit.com ajax IRC Client")
  • [00:28:15] * _fulgas (n=fn@85.138.15.70) has joined #beagle
  • [00:38:33] * Guest4565 (i=816ef12f@gateway/web/ajax/mibbit.com/x-bff8aafa60a019be) Quit ("http://www.mibbit.com ajax IRC Client")
  • [00:42:28] * fulgas (n=fn@85.138.15.70) Quit (Read error: 110 (Connection timed out))
  • [00:44:18] * abitos (n=nixgibts@dslb-084-057-146-023.pools.arcor-ip.net) Quit (Read error: 60 (Operation timed out))
  • [00:47:41] * brolin (n=brolin@190.157.13.75) Quit (Client Quit)
  • [00:50:16] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [00:54:24] * raster (n=raster@enlightenment/developer/raster) Quit ("Gettin' stinky!")
  • [00:59:16] <vlad_> mru: i'm slow, I just made the connection between http://hardwarebug.org/category/arm/ and you :p
  • [01:00:22] <vlad_> mru: do you know offhand how to get the assembler to not complain that the processor doesn't support movt? -mcpu=cortex-a8 doesn't do the trick
  • [01:08:31] <jkridner> anyone get eFfeM's entry running?
  • [01:08:43] <jkridner> I get hung at Downloading http://www.angstrom-distribution.org/feeds/2008/ipk/glibc/armv7a/base/libstdc++6_4.3.1-r17.1_armv7a.ipk
  • [01:10:49] * mib_uut961 (i=4dfc12d9@gateway/web/ajax/mibbit.com/x-8b177cd850c57133) has joined #beagle
  • [01:16:34] * DaQatz (n=db@c-66-30-48-150.hsd1.nh.comcast.net) Quit (Remote closed the connection)
  • [01:17:33] * DaQatz (n=db@c-66-30-48-150.hsd1.nh.comcast.net) has joined #beagle
  • [01:19:38] * _fulgas (n=fn@85.138.15.70) Quit (Read error: 104 (Connection reset by peer))
  • [01:20:04] * fulgas (n=fn@85.138.15.70) has joined #beagle
  • [01:24:28] * mib_uut961 (i=4dfc12d9@gateway/web/ajax/mibbit.com/x-8b177cd850c57133) Quit ("http://www.mibbit.com ajax IRC Client")
  • [01:27:33] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [01:31:18] * rsalveti (n=salveti@dasasob.nokia.com) Quit (Remote closed the connection)
  • [01:32:15] * ant_home (n=andrea@host37-249-dynamic.4-87-r.retail.telecomitalia.it) Quit ("Leaving")
  • [01:34:04] * Vegar (n=vegar@unaffiliated/vegar) Quit (Remote closed the connection)
  • [01:34:28] * Vegar (n=vegar@209-20-88-100.slicehost.net) has joined #beagle
  • [01:34:52] * Vegar is now known as Guest71345
  • [01:38:13] * mpr (n=mpr@aggr.com) Quit (Remote closed the connection)
  • [01:38:16] * mpr (n=mpr@aggr.com) has joined #beagle
  • [01:47:06] * fulgas is now known as FuL|OUT
  • [01:55:22] * robclark (n=robclark@nat/ti/x-dc113761456b5f24) Quit (Remote closed the connection)
  • [01:56:10] * Wowbagger_ (n=wowbagge@d154-20-187-250.bchsia.telus.net) has joined #beagle
  • [02:18:01] * Septalicia (n=bakljg@c-98-240-226-129.hsd1.mn.comcast.net) has joined #beagle
  • [02:18:33] * Xerion (i=xerion@82-170-197-160.ip.telfort.nl) Quit (Read error: 104 (Connection reset by peer))
  • [02:20:58] * Sept (n=bakljg@c-98-240-226-129.hsd1.mn.comcast.net) Quit (Read error: 60 (Operation timed out))
  • [02:27:11] * Cru_N_cher (n=luls_lol@dslb-084-058-098-215.pools.arcor-ip.net) Quit (Read error: 60 (Operation timed out))
  • [02:28:52] * lcuk (n=lcuk@cpc3-oldh7-0-0-cust590.manc.cable.ntl.com) Quit ("Leaving")
  • [02:29:03] * CruNcher (n=luls_lol@dslb-084-058-152-081.pools.arcor-ip.net) has joined #beagle
  • [02:29:31] * rsalveti (n=salveti@200.222.2.234) has joined #beagle
  • [02:33:25] * robclark (n=robclark@166.128.148.236) has joined #beagle
  • [02:50:36] * robclark (n=robclark@166.128.148.236) Quit (Connection timed out)
  • [03:30:56] * TAK2004 (n=thomas@dslb-088-074-060-212.pools.arcor-ip.net) Quit (Read error: 104 (Connection reset by peer))
  • [03:40:28] * rsalveti (n=salveti@200.222.2.234) Quit (Remote closed the connection)
  • [03:40:55] * rsalveti (n=salveti@200.222.2.234) has joined #beagle
  • [03:45:36] * rsalveti (n=salveti@200.222.2.234) Quit (Remote closed the connection)
  • [03:46:02] * rsalveti (n=salveti@200.222.2.234) has joined #beagle
  • [03:46:05] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [03:47:05] * arcclaro (n=arcclaro@78.130.65.13.rev.optimus.pt) has joined #beagle
  • [03:50:20] * rsalveti (n=salveti@200.222.2.234) Quit (Remote closed the connection)
  • [03:50:41] * rsalveti (n=salveti@200.222.2.234) has joined #beagle
  • [04:06:01] * ssvb (n=ssvb___@a88-114-221-132.elisa-laajakaista.fi) Quit ("Leaving")
  • [04:14:52] * arcclaro (n=arcclaro@78.130.65.13.rev.optimus.pt) Quit (Read error: 110 (Connection timed out))
  • [04:17:20] * Pyrodogg (i=ad116564@gateway/web/ajax/mibbit.com/x-2007be0ed578497f) has joined #beagle
  • [04:32:37] * ClaudeQC (n=ClaudeQC@bas1-quebec03-1279636431.dsl.bell.ca) has left #beagle
  • [04:36:40] <ZeZu> does movw affect the high halfword ?
  • [04:37:33] <ZeZu> hmm looks like it zero extends
  • [04:37:49] * mib_b3ogfr (i=4428515f@gateway/web/ajax/mibbit.com/x-cf75138e0089cf15) has joined #beagle
  • [04:41:24] * mib_b3ogfr (i=4428515f@gateway/web/ajax/mibbit.com/x-cf75138e0089cf15) Quit (Client Quit)
  • [05:08:20] <vlad_> arg, I need to get my beagle updated with the 3d drivers and all
  • [05:33:04] * TehUni (n=cory@c-24-30-35-55.hsd1.ga.comcast.net) Quit (Remote closed the connection)
  • [05:34:32] * Xerion (i=xerion@82-170-197-160.ip.telfort.nl) has joined #beagle
  • [05:51:15] * harbaum (n=harbaum@i59F4B795.versanet.de) has joined #beagle
  • [05:57:32] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has left #beagle
  • [06:28:56] * TehUni (n=cory@c-24-30-35-55.hsd1.ga.comcast.net) has joined #beagle
  • [06:34:20] <vlad_> somehow I ended up at home without a single SD card reader
  • [06:34:25] <vlad_> except for the one built-in to my monitor, which doesn't like SDHC
  • [06:39:58] * Pyrodogg (i=ad116564@gateway/web/ajax/mibbit.com/x-2007be0ed578497f) Quit ("http://www.mibbit.com ajax IRC Client")
  • [06:43:11] * likewise (n=likewise@p5B39F4C7.dip.t-dialin.net) has joined #beagle
  • [06:50:21] * kozak (n=subbu@117.192.0.126) has joined #beagle
  • [06:52:14] * Batko_Marto (n=Batko_Ma@CPE001346f996d2-CM001e6b1a8d1e.cpe.net.cable.rogers.com) Quit (Remote closed the connection)
  • [07:03:39] * BThompsonD (n=bernie@cpe-72-190-75-99.tx.res.rr.com) Quit ("Leaving.")
  • [07:13:56] * harbaum (n=harbaum@i59F4B795.versanet.de) Quit (Read error: 113 (No route to host))
  • [07:20:47] * likewise (n=likewise@p5B39F4C7.dip.t-dialin.net) Quit ()
  • [07:31:47] * arne-unicap (n=arne@p548198F3.dip0.t-ipconnect.de) has joined #beagle
  • [07:32:28] * kozak (n=subbu@117.192.0.126) Quit (Read error: 60 (Operation timed out))
  • [07:34:49] <watarat> anyone care to take a guess at boot time for a Beagle with FS, USB and not much else - just bit banging IO and writes to disk, No Network, No UI, No Apps etc?
  • [07:44:22] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-d6248b40a7b27877) has joined #beagle
  • [07:46:17] * Wowbagger_ (n=wowbagge@d154-20-187-250.bchsia.telus.net) Quit ("Leaving")
  • [07:58:50] <recalcati> morning
  • [07:59:52] <ZeZu> vlad_: and beagle doesn't mind sdhc? or using it for another purpose ?
  • [07:59:54] <kulve> watarat: a guess: 7 seconds
  • [08:00:08] <ZeZu> recalcati: morning
  • [08:02:14] <recalcati> I made a custom image from beagleboard-demo-image and swfdec-mozilla starts but doesn'r display youtube video
  • [08:02:25] <vlad_> sdhc works fine on beagle
  • [08:05:44] <recalcati> it comes out a big play image but the video doesn't start
  • [08:07:50] <watarat> Cool - obviously not 1 minute, which I was worried about, - better get hold of one and start playing :)
  • [08:10:20] <watarat> Typical boot time "fully optioned" ??
  • [08:11:14] <kulve> I guess it's something >30secs (or even close to a minute) if you start up a full blown desktop environment
  • [08:11:37] <kulve> but that depends very much on what you start
  • [08:11:41] <recalcati> this thread was stopped http://groups.google.com/group/beagleboard/browse_thread/thread/eeff1103e74daf9f/cb1a710d95171393?lnk=gst&q=swfdec#cb1a710d95171393
  • [08:12:28] <watarat> OK - that sort of gives me a feel for it - thanks.
  • [08:31:09] * Watarat1 (n=Rich@202-161-24-77.dyn.iinet.net.au) has joined #beagle
  • [08:40:19] * ScriptRipper (n=martin@host-82-135-36-197.customer.m-online.net) has joined #beagle
  • [08:43:58] * watarat (i=caa1184d@gateway/web/ajax/mibbit.com/x-47fc1cf45da253c1) Quit (Read error: 104 (Connection reset by peer))
  • [08:43:58] * fop518 (i=d078f0bd@gateway/web/ajax/mibbit.com/x-4ef0749c05e5439f) Quit (Remote closed the connection)
  • [08:43:58] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-d6248b40a7b27877) Quit (Remote closed the connection)
  • [08:43:58] * mib_v6ycjl (i=86868b47@gateway/web/ajax/mibbit.com/x-8e87588eeff59802) Quit (Remote closed the connection)
  • [08:44:39] * watarat (i=caa1184d@gateway/web/ajax/mibbit.com/x-8448a8852aed2df4) has joined #beagle
  • [08:45:11] * mib_v6ycjl (i=86868b47@gateway/web/ajax/mibbit.com/x-255dc85eb73bed93) has joined #beagle
  • [08:45:24] * harbaum (n=harbaum@i59F4B795.versanet.de) has joined #beagle
  • [08:45:32] * fop518 (i=d078f0bd@gateway/web/ajax/mibbit.com/x-5141aa87d12e45a6) has joined #beagle
  • [08:45:47] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-2e06a608ba7c4217) has joined #beagle
  • [08:46:06] * mib_6fd02v (i=4dfc12d7@gateway/web/ajax/mibbit.com/x-f699178c0e3ec0e4) has joined #beagle
  • [08:50:31] * flameman (n=toor@net-93-65-1-167.cust.dsl.vodafone.it) has joined #beagle
  • [09:01:12] * mib_6fd02v (i=4dfc12d7@gateway/web/ajax/mibbit.com/x-f699178c0e3ec0e4) Quit ("http://www.mibbit.com ajax IRC Client")
  • [09:01:38] * rupeshgujare (n=rupesh@59.160.172.220) has joined #beagle
  • [09:06:12] * raster (n=raster@enlightenment/developer/raster) has joined #beagle
  • [09:16:37] * Watarat1 (n=Rich@202-161-24-77.dyn.iinet.net.au) has left #beagle
  • [09:17:51] * valhalla (n=valhalla@81-174-25-183.dynamic.ngi.it) has joined #beagle
  • [09:19:14] * ogra (n=ogra@ubuntu/member/ogra) Quit (Read error: 104 (Connection reset by peer))
  • [09:20:08] * Watarat1 (n=Rich@202-161-24-77.dyn.iinet.net.au) has joined #beagle
  • [09:23:41] * harbaum (n=harbaum@i59F4B795.versanet.de) Quit (Read error: 113 (No route to host))
  • [09:30:02] * dev_543|away (i=c2c841ef@gateway/web/ajax/mibbit.com/x-2cdddcc4d11a45fe) Quit (Remote closed the connection)
  • [09:34:04] * mrc3 (n=ddiaz@189.157.115.100) has joined #beagle
  • [09:34:11] * ogra (n=ogra@ubuntu/member/ogra) has joined #beagle
  • [09:34:46] * vijay (n=vijay@203.199.213.3) has joined #beagle
  • [09:53:07] <mru> morning
  • [09:54:14] <vlad_> morning
  • [09:55:02] <mru> vlad_: did you solve your problem?
  • [09:55:05] <mru> with movt
  • [09:59:12] <vlad_> yeah, I used .word :p
  • [09:59:47] <mru> that's not pretty
  • [09:59:50] <mru> it should work
  • [09:59:53] * likewise (n=likewise@p5B39CEA0.dip.t-dialin.net) has joined #beagle
  • [09:59:55] <vlad_> it's not, but guaranteed to work :)
  • [09:59:59] <mru> what version of binutils are you using?
  • [09:59:59] <vlad_> yeah
  • [10:00:08] <vlad_> um, whatever was with 2007q3
  • [10:00:15] <mru> it's broken
  • [10:00:18] <vlad_> fun
  • [10:00:21] <mru> use fsf 2.18 or later
  • [10:01:46] * likewise (n=likewise@p5B39CEA0.dip.t-dialin.net) Quit (Client Quit)
  • [10:06:15] * flameman (n=toor@net-93-65-1-167.cust.dsl.vodafone.it) Quit (Read error: 104 (Connection reset by peer))
  • [10:06:56] * flameman (n=toor@net-93-65-1-167.cust.dsl.vodafone.it) has joined #beagle
  • [10:07:15] * lcuk (i=lcuk@cpc3-oldh7-0-0-cust590.manc.cable.ntl.com) has joined #beagle
  • [10:12:44] * Robot_ (n=robot@rb5bx242.net.upc.cz) has joined #beagle
  • [10:23:43] * erbo (n=erik@c-7b7de455.115-16-64736c14.cust.bredbandsbolaget.se) has joined #beagle
  • [10:35:07] * abitos (n=nixgibts@dslb-084-057-146-023.pools.arcor-ip.net) has joined #beagle
  • [10:39:26] * Xenion (n=robert@p579FCF1B.dip.t-dialin.net) Quit (Read error: 101 (Network is unreachable))
  • [10:39:49] * Xenion (n=robert@p579FCE76.dip.t-dialin.net) has joined #beagle
  • [10:43:05] * fop518 (i=d078f0bd@gateway/web/ajax/mibbit.com/x-5141aa87d12e45a6) Quit ("http://www.mibbit.com ajax IRC Client")
  • [10:53:31] * ant_home (n=andrea@host37-249-dynamic.4-87-r.retail.telecomitalia.it) has joined #beagle
  • [10:54:44] * ant_home (n=andrea@host37-249-dynamic.4-87-r.retail.telecomitalia.it) Quit (Client Quit)
  • [10:55:21] * eFfeM (n=Frans@195-241-226-180.ip.telfort.nl) has joined #beagle
  • [10:58:14] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-2e06a608ba7c4217) Quit ("http://www.mibbit.com ajax IRC Client")
  • [11:03:37] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-2fd8317f8907d64d) has joined #beagle
  • [11:04:04] <Robot_> I have compiled u-boot in OE and it does not understand "nand ecc sw"
  • [11:04:30] <Robot_> is that a big problem ? or is the ecc set by default ?
  • [11:09:22] <mru> try "nandecc sw"
  • [11:10:45] <Robot_> thanks ;)
  • [11:11:02] <Robot_> i just tried googling it ... and its a bit faster :)
  • [11:22:27] * eFfeM (n=Frans@195-241-226-180.ip.telfort.nl) Quit (Read error: 104 (Connection reset by peer))
  • [11:24:42] * eFfeM (n=frans@195-241-226-180.ip.telfort.nl) has joined #beagle
  • [11:24:43] * pH5 (n=ph5@e178242205.adsl.alicedsl.de) has joined #beagle
  • [11:24:53] * Watarat1 (n=Rich@202-161-24-77.dyn.iinet.net.au) has left #beagle
  • [11:25:18] * rob_k (i=5a9c259a@gateway/web/ajax/mibbit.com/x-56222e2653d7748d) has joined #beagle
  • [11:26:46] * mib_hc9uw5 (i=caa1184d@gateway/web/ajax/mibbit.com/x-abd4c08e758cfb66) has joined #beagle
  • [11:27:31] * mib_hc9uw5 (i=caa1184d@gateway/web/ajax/mibbit.com/x-abd4c08e758cfb66) Quit (Client Quit)
  • [11:29:18] * harbaum (n=harbaum@i59F4B795.versanet.de) has joined #beagle
  • [11:29:24] <rob_k> Hi all, how can I add some more packages to my beagleboard build? I'm using console-image conf. I need opencv and wifi pkg like kismet etc.
  • [11:35:11] <eFfeM> rob_k: opkg install packagename
  • [11:35:24] <eFfeM> e.g. opkg install kismet
  • [11:36:30] <rob_k> ok but where can I find packages for arm platform?
  • [11:36:51] <rob_k> Is it some specific way of searching?
  • [11:37:44] <rob_k> I thought that best solution is creating whole build with included packages.
  • [11:42:11] <eFfeM> opkg list | more
  • [11:42:22] <eFfeM> then you download from the angstrom repo
  • [11:42:37] * Xenion (n=robert@p579FCE76.dip.t-dialin.net) Quit ("Verlassend")
  • [11:42:37] <eFfeM> you can build yourself if you want to, look intot the packages dir
  • [11:43:33] <recalcati> eFfeM: if I build all in my pc and I set /etc/opkg/beagleboard-feed.conf to my PC woth apache it is good, isn't it ?
  • [11:45:16] * rupeshgujare (n=rupesh@59.160.172.220) has left #beagle
  • [11:45:17] <eFfeM> recalcati: I guess it does, but I never tried it
  • [11:46:04] <recalcati> the reason is the I'd like not to recompile every time
  • [11:46:21] <eFfeM> recalcati: you gave me the 720p string yesterday, but what do I tell to Xorg ? I get a 720p display with Xorg using 640x480
  • [11:46:41] <eFfeM> don't understand the recompile part
  • [11:46:57] <recalcati> eFfeM: It is an awfull hack
  • [11:47:11] <eFfeM> afk
  • [11:50:32] <recalcati> for 720p: see first of all in linux-omap-2.6.28/Documentation/arm/OMAP/DSS some explanation. then1 remember that not all timings are possible due http://groups.google.com/group/beagleboard/browse_thread/thread/89654a5ac60b82a9/8423f1fadc6f5397#8423f1fadc6f5397 to
  • [11:52:36] <eFfeM> recalcati: i know about the kernel changes and that thread; my monitor is happy with the current settings (my tv is not); but for some reason u-boot does not want to save me env settings any more (guess I need to redo the fat partition)
  • [11:52:58] <eFfeM> i'll try a kernel patch so it works on tv later this week
  • [11:53:09] <mru> the uboot env is saved in nand
  • [11:53:32] <eFfeM> btw wrt packages:maybe opkg install http:.... does also work (no idea)
  • [11:53:43] <mru> maybe if you clear the env area it will start working again
  • [11:53:44] <eFfeM> mru: even if booting from mmc ?
  • [11:53:54] <mru> yes
  • [11:54:09] <eFfeM> mru I already tried a nand erase
  • [11:54:33] <eFfeM> i thought when using mmc it was saved in u-boot.bin
  • [11:54:57] <mru> uboot doesn't know where it was loaded from
  • [11:55:37] <eFfeM> true
  • [11:57:10] <recalcati> normally I get u-boot from mmc
  • [11:57:31] <eFfeM> me too
  • [11:58:29] * guillaum1 (n=gl@AMontsouris-153-1-89-247.w90-2.abo.wanadoo.fr) Quit (Read error: 110 (Connection timed out))
  • [11:59:07] <mru> I would too, except x-loader tells me Error: reading boot sector
  • [11:59:15] <mru> I've no idea what the nature of that error is
  • [11:59:37] <mru> uboot has no trouble loading the kernel from the same card
  • [12:00:20] * guillaum1 (n=gl@AMontsouris-153-1-22-157.w86-212.abo.wanadoo.fr) has joined #beagle
  • [12:00:27] <eFfeM> odd, have you tried another card ?
  • [12:01:21] <eFfeM> i have problems booting the .28 kernel from a 4GB sdhc card. Odd thing is that it worked fine with a .27 kernel
  • [12:01:25] <eFfeM> away for lunch
  • [12:01:42] <recalcati> After having lost a lot of time with fdisk now I used the Koen video to make a good partition
  • [12:01:44] <mru> I'm using 8GB cards
  • [12:02:01] <mru> and the kernel loads without issue
  • [12:02:14] <mru> I'm fine keeping uboot in nand
  • [12:03:24] <recalcati> I'm trying to add tmp/deploy/glibc/ipk/armv7a to my apache
  • [12:03:50] * pH5 (n=ph5@e178242205.adsl.alicedsl.de) Quit (Read error: 110 (Connection timed out))
  • [12:04:24] * pH5 (n=ph5@e178219220.adsl.alicedsl.de) has joined #beagle
  • [12:06:44] * FuL|OUT is now known as fulgas
  • [12:09:53] * TAK2004 (n=thomas@dslb-088-072-212-109.pools.arcor-ip.net) has joined #beagle
  • [12:10:36] * Guest71345 is now known as Vegar
  • [12:17:17] <eFfeM> back
  • [12:17:55] <eFfeM> mru for me everything loaded fine, but when booting the kernel I got errors
  • [12:20:44] * TAK2004 (n=thomas@dslb-088-072-212-109.pools.arcor-ip.net) Quit (Remote closed the connection)
  • [12:25:40] * Xenion (n=robert@p579FCE76.dip.t-dialin.net) has joined #beagle
  • [12:27:04] * arcclaro (n=arcclaro@62.169.121.6.rev.optimus.pt) has joined #beagle
  • [12:27:54] * harbaum (n=harbaum@i59F4B795.versanet.de) Quit (Read error: 113 (No route to host))
  • [12:28:45] * likewise (n=likewise@p5B39EE40.dip.t-dialin.net) has joined #beagle
  • [12:29:54] <eFfeM> koen, ping
  • [12:33:26] * arne-unicap (n=arne@p548198F3.dip0.t-ipconnect.de) Quit (Read error: 113 (No route to host))
  • [12:33:56] <recalcati> done, now I see ipk directory compile by myself, obviously I have some errors: opkg -V 2 -f /etc/opkg/beagleboard-local.conf update ... Downloading http://192.168.1.101/IPK/Packages.gz Inflating http://192.168.1.101/IPK/Packages.gz Updated list of available packages in /usr/lib/opkg/lists/beagleboard Collected errors: * ERROR: duplicate src declaration. Skipping: src beagleboard http://192.168.1.101/IPK
  • [12:34:22] * fulgas (n=fn@85.138.15.70) Quit (Read error: 104 (Connection reset by peer))
  • [12:34:30] <recalcati> probably I have to configure packages list starting from the compilation
  • [12:34:33] <mru> eFfeM: I've had errors booting the kernel many times
  • [12:34:36] * pH5_ (n=ph5@e178236226.adsl.alicedsl.de) has joined #beagle
  • [12:35:00] * fulgas (n=fn@85.138.15.70) has joined #beagle
  • [12:35:26] <eFfeM> mru for me i guess it is a driver issue, as the same image works like a charm with an SD card
  • [12:35:53] <recalcati> mru:
  • [12:35:59] <mru> eFfeM: and when it failed?
  • [12:36:00] <recalcati> ops sorry
  • [12:36:42] <eFfeM> mru the kernel booted but complained about the root fs; it said it was no good, replayed the log and remounted ro
  • [12:36:57] <eFfeM> even though i fsck-ed the card on my desktop before using it
  • [12:36:58] <mru> maybe the fs was damaged
  • [12:37:03] <mru> hmm
  • [12:37:16] <eFfeM> could be timing related
  • [12:37:36] <eFfeM> actually i should try with anoter one, i have a 2nd card same type & brand
  • [12:37:40] <mru> I've never had any problems reading sd cards
  • [12:37:42] <mru> of any size
  • [12:37:45] <recalcati> eFfeM: every time I had that kind of problem I made a gift to one of my friend
  • [12:38:18] <mru> in the kernel, that is
  • [12:38:23] <eFfeM> recalcati: lol, do you have any friends left after giving them stuff that might be broken "-)
  • [12:38:30] <mru> x-loader, as I said, seems to be a bit funny
  • [12:39:02] <mru> try s/gift/sacrifice/ s/friend/god/
  • [12:39:17] <eFfeM> lol
  • [12:40:24] <eFfeM> wtf, mediatomb tries to load /usr/lib/libjs instead of /usr/lib/libjs.so but it worked earlier this week
  • [12:40:38] <recalcati> I told them. I really didn't understand the reason, they sometimes told me that sdcard worked
  • [12:41:42] <eFfeM> sdhc seems a more picky than sd; i know that with the original eeepc some sdhc cards worked and some did not
  • [12:41:51] <eFfeM> could also be the interface
  • [12:51:21] * pH5 (n=ph5@e178219220.adsl.alicedsl.de) Quit (Read error: 110 (Connection timed out))
  • [12:53:18] * arcclaro_ (n=arcclaro@88.210.78.126.rev.optimus.pt) has joined #beagle
  • [12:57:36] <recalcati> going to lunch
  • [12:58:45] <eFfeM> enjoy
  • [13:08:19] * fulgas (n=fn@85.138.15.70) Quit ("changing servers")
  • [13:08:40] * ant_home (n=andrea@host37-249-dynamic.4-87-r.retail.telecomitalia.it) has joined #beagle
  • [13:08:48] * fulgas (n=fn@85.138.15.70) has joined #beagle
  • [13:11:02] * arcclaro_ (n=arcclaro@88.210.78.126.rev.optimus.pt) has left #beagle
  • [13:18:00] * fulgas (n=fn@85.138.15.70) Quit ("changing servers")
  • [13:18:33] * fulgas (n=fn@85.138.15.70) has joined #beagle
  • [13:20:32] * arcclaro (n=arcclaro@62.169.121.6.rev.optimus.pt) Quit (Read error: 110 (Connection timed out))
  • [13:21:11] * kozak (n=subbu@117.192.20.35) has joined #beagle
  • [13:22:56] <recalcati> I learn everytime a piece of OE.
  • [13:27:29] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [13:28:55] * dirk2 (n=dirk@p5B04081F.dip0.t-ipconnect.de) has joined #beagle
  • [13:29:31] * koen_ (n=koen@ip5451781a.direct-adsl.nl) has joined #beagle
  • [13:30:02] <koen_> Good morning all
  • [13:30:30] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-2fd8317f8907d64d) Quit ("http://www.mibbit.com ajax IRC Client")
  • [13:31:45] <raster> koen_: you're tail's sticking out
  • [13:31:53] <ant_home> hi koen
  • [13:33:11] <koen_> raster: That's the aerial of my phone ;)
  • [13:33:22] <raster> wow
  • [13:33:25] <raster> old-school
  • [13:33:31] <raster> it has an aerial :)
  • [13:33:33] <koen_> Hey ant_home
  • [13:33:48] <ant_home> koen: my build from scratch for c7x0 fails because two files are not found (updater.sh and sharp-flash-header.bin). Both are in EXTRA_IMAGEDEPEND...did it got borked?
  • [13:34:08] <koen_> *cough* iPhone *cough*
  • [13:34:49] <koen_> Zaurus in oe is 'different'
  • [13:35:30] <raster> aaah zaurusu build
  • [13:35:34] <raster> that used to work well
  • [13:35:36] <ant_home> iirc this implies the files are just staged
  • [13:35:38] <raster> then somewhere along the way.. broke
  • [13:35:41] <ant_home> not installed
  • [13:36:25] <ant_home> btw you know since long time I have a floating patch to kill that headers ;)
  • [13:36:37] <ant_home> and you know my opinion about zaurus-updater...
  • [13:36:53] <ant_home> but I don't dare to commit what I locally have ;-)
  • [13:37:47] <raster> it installed for me
  • [13:37:51] <koen_> I wouldn't be surprised if zaurus machines would install opie in every image
  • [13:37:53] <ant_home> btw now I havew 8gb ram and moved /tmp/angstrom to tmpfs
  • [13:37:53] <raster> but usbnet broke
  • [13:38:00] <raster> and so i lost any console :(
  • [13:38:01] <ant_home> well, it flies
  • [13:38:09] <raster> that made mr happy.. very sad.
  • [13:38:10] <ant_home> if we just split pstage and all
  • [13:38:17] <ant_home> out of tmp...
  • [13:38:26] <ant_home> would be perfect!
  • [13:38:30] <raster> koen_: nah - its actually a pretty lean image :)
  • [13:39:16] <koen_> I know how to fix the zarus stuff
  • [13:39:49] * ant_home just build zaurus-updater and sharp-flash-headers before
  • [13:39:52] <koen_> But i have better things to do
  • [13:39:59] <ant_home> np
  • [13:40:37] <raster> koen_: not surprising
  • [13:40:44] <ant_home> koen: indeed !
  • [13:40:53] <raster> frankly it'd be a fulltime job for like 5 people to keep oe working on all the platforms it supports
  • [13:41:05] <raster> and that would assume u can still GET hw for all of them somehow
  • [13:41:51] <koen_> I spend about 3 hours a day to keep beagle working
  • [13:42:27] * DaQatz (n=db@c-66-30-48-150.hsd1.nh.comcast.net) Quit (Read error: 110 (Connection timed out))
  • [13:42:29] <ant_home> and repairing breakages...
  • [13:42:38] <raster> hahaha
  • [13:43:02] <ant_home> well, not much time for development
  • [13:43:28] <ant_home> but your role is even more important
  • [13:43:34] <koen_> 'keep working' == repAiring breakages
  • [13:43:43] <ant_home> 'role' = 'missionary task
  • [13:44:22] <ant_home> well, </rant>
  • [13:45:03] <koen_> I like doing new stuff
  • [13:45:20] * koen_ gets bored quickly
  • [13:46:27] <raster> koen_: fix my gles stuff!
  • [13:46:29] <raster> :)
  • [13:46:32] <raster> thats new!
  • [13:46:33] <raster> :)
  • [13:46:36] <ant_home> I lost interest in the brand-new-hardware years ago
  • [13:46:42] * Crofton|work is happy demo-image built yesterday
  • [13:46:43] <koen_> If usps hurries up i can put some pictures online of my secret project
  • [13:47:14] <Crofton|work> although I still think there are some packages that break on || make
  • [13:47:24] <Crofton|work> net-tools maybe
  • [13:47:35] <koen_> HFS
  • [13:47:40] <koen_> Gah
  • [13:47:52] <koen_> Nfsutils is broken
  • [13:48:08] <Crofton|work> how?
  • [13:48:29] <koen_> Dorsn't vuile
  • [13:48:38] <koen_> Build
  • [13:48:40] <mru> koen_: are you drunk?
  • [13:48:51] <Crofton|work> heh
  • [13:49:01] <koen_> No, wordcompletion
  • [13:49:11] <koen_> With a dutch dict
  • [13:49:23] <Crofton|work> :)
  • [13:49:25] <mru> ah, drunken work completion
  • [13:49:33] <Crofton|work> same difference
  • [13:56:11] <ant_home> well, maximum occupation using tmpfs (and rm_work) none 4.0G 2.5G 1.6G 61% /oe/build/tmp/angstrom
  • [13:56:37] <ant_home> basically 2,5GB ram are enough, in case it should even swap
  • [13:58:50] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [14:00:30] <eFfeM> koen_ can you have a look at coherence/gstreamer? I brought coherence to 0.6.2 and it does not build. there is no dependence on gstreamer, yet it wants to build gstreamer which fails in configre on an xml2 test, i've added libxml2-native to the gstreamer recipe, but that did not help either
  • [14:01:01] <eFfeM> mediatomb seems to have the same problem (and wants /usr/lib//libjs without an extension :-( )
  • [14:01:40] <eFfeM> and i could not save my uboot env, so no pics of james on the monitor
  • [14:02:10] <eFfeM> puzzled, probably clean out tmp and rebuild tonight
  • [14:02:38] * watarat (i=caa1184d@gateway/web/ajax/mibbit.com/x-8448a8852aed2df4) Quit ("http://www.mibbit.com ajax IRC Client")
  • [14:02:42] <koen_> Ping me when i get home
  • [14:03:07] <eFfeM> koen_ will do, gone as well in a few minutes
  • [14:03:24] <eFfeM> if anyone else wants to see if they can build gstreamer: greatly appreciated
  • [14:12:42] * Robot_ (n=robot@rb5bx242.net.upc.cz) Quit (Remote closed the connection)
  • [14:18:23] <likewise> ant_home: how fast does the oe 'fly' when in tmpfs?
  • [14:18:31] <likewise> oe=oe build
  • [14:19:00] <likewise> oh, and hi all :-)
  • [14:23:04] * CruNcher (n=luls_lol@dslb-084-058-152-081.pools.arcor-ip.net) Quit (Read error: 60 (Operation timed out))
  • [14:23:11] * koen_ heads to the station
  • [14:23:15] * koen_ (n=koen@ip5451781a.direct-adsl.nl) Quit ("Get Colloquy for iPhone! http://mobile.colloquy.info/")
  • [14:24:04] * erbo (n=erik@c-7b7de455.115-16-64736c14.cust.bredbandsbolaget.se) Quit (Read error: 60 (Operation timed out))
  • [14:24:30] * CruNcher (n=luls_lol@dslb-084-058-113-049.pools.arcor-ip.net) has joined #beagle
  • [14:26:11] * erbo (n=erik@c-7b7de455.115-16-64736c14.cust.bredbandsbolaget.se) has joined #beagle
  • [14:40:07] <ant_home> likewise: I have yet to time it, but at least 50%
  • [14:40:21] <ant_home> (I have a SATA AHCI 320GB barracuda)
  • [14:40:44] <likewise> ant_home: and full tmpfs is not swapped out, i.e. you have enough ram?
  • [14:42:53] <jkridner> good morning all.
  • [14:43:50] <jkridner> eFfeM: libstdc++ never finishes downloading for me :(
  • [14:48:01] * mib_mfmaql (i=18289441@gateway/web/ajax/mibbit.com/x-95793c8482984cc9) has joined #beagle
  • [14:48:12] * mib_mfmaql (i=18289441@gateway/web/ajax/mibbit.com/x-95793c8482984cc9) Quit (Client Quit)
  • [14:48:35] * john280z (n=johnm@user-0c2h521.cable.mindspring.com) has joined #beagle
  • [14:50:24] <likewise> gm jkridner
  • [14:51:10] <jkridner> hi likewise.
  • [14:51:27] <jkridner> would have liked to have your entry for this contest instance.
  • [14:51:42] <jkridner> have you tried running James?
  • [14:52:19] <likewise> james?
  • [14:53:00] <likewise> ah, effem's project
  • [14:53:57] <jkridner> right.
  • [14:57:46] <likewise> jkridner: no, I barely get time to play with the beagle/omap3. Mostly because we will not do a design-in on it. When 1080p support is available, I'm there to step in again :-)
  • [14:58:36] * Prathap (i=ca9c0be2@gateway/web/ajax/mibbit.com/x-1047072de991994d) has joined #beagle
  • [15:00:59] * Prathap (i=ca9c0be2@gateway/web/ajax/mibbit.com/x-1047072de991994d) Quit (Client Quit)
  • [15:03:48] * john280z (n=johnm@user-0c2h521.cable.mindspring.com) Quit ("Leaving")
  • [15:05:11] * pmeerw_ (n=pmeerw@195-70-114-97.stat.salzburg-online.at) Quit (Remote closed the connection)
  • [15:06:41] * stoupha (i=29e1c597@gateway/web/ajax/mibbit.com/x-a6f5cffc7d06f3b1) has joined #beagle
  • [15:06:50] <stoupha> hi
  • [15:07:10] <stoupha> any one can help me ?
  • [15:08:56] * stoupha (i=29e1c597@gateway/web/ajax/mibbit.com/x-a6f5cffc7d06f3b1) Quit (Client Quit)
  • [15:11:04] * koen_ (n=koen@94.157.56.165) has joined #beagle
  • [15:11:32] * koen_ in a train now
  • [15:13:19] * koen_ (n=koen@94.157.56.165) Quit (Read error: 104 (Connection reset by peer))
  • [15:13:53] * koen_ (n=koen@94.157.56.165) has joined #beagle
  • [15:18:48] <jkridner> likewise: so, 1080p is a requirement for you today?
  • [15:19:31] <jkridner> does someone need to do an omap3 board with 1080p output (at higher than 25fps)?
  • [15:20:02] * koen_ wants 1
  • [15:20:21] <likewise> jkridner: not today, we started shipping 1080p products last year
  • [15:20:58] <jkridner> full-frame video, or just rendering at that resolution?
  • [15:21:33] <likewise> jkridner: 1080p50 and p60. but rendering would be ok at lower frame rates, might even output p25 or p30.
  • [15:22:44] <jkridner> k. Beagle can output 1080p25 for refreshing the screen, at a reduced blanking interval, but won't likely do full-frame video at that resolution/fps.
  • [15:23:50] <jkridner> as other devices come out, I'll try to think of your need. adding hardware to do the 1080p60 screen refreshes is relatively simple.
  • [15:24:34] <koen_> Hdmi audio would be nice as well
  • [15:25:22] * koen__ (n=koen@94.157.56.165) has joined #beagle
  • [15:25:23] * koen_ (n=koen@94.157.56.165) Quit (Read error: 104 (Connection reset by peer))
  • [15:25:45] <jkridner> I believe the Archos DVR station does both of those.
  • [15:26:06] <jkridner> http://www.archos.com/products/imt/archos_7/accessories.html?country=ru&lang=en
  • [15:26:11] <koen__> Dunno how that works out with licensing
  • [15:26:13] * jkridner is looking for the output resolution.
  • [15:26:46] <koen__> I'd like something hackable :)
  • [15:27:18] <likewise> jkridner: ok, so with reduced blanking I can get the beagle to output 1080p25?
  • [15:27:20] * Batko_Marto (n=Batko_Ma@CPE001346f996d2-CM001e6b1a8d1e.cpe.net.cable.rogers.com) has joined #beagle
  • [15:27:30] <jkridner> likewise: yes.
  • [15:27:35] <likewise> jkridner: are these modes supported in the current drivers?
  • [15:27:49] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [15:27:58] <jkridner> see http://elinux.org/BeagleBoardFAQ#linux-omap_display_driver
  • [15:28:06] <koen__> Sysfs allows any hardware mode
  • [15:28:08] <jkridner> likewise: yes.
  • [15:28:44] <koen__> There was a mail aboit it this weekend
  • [15:29:02] <likewise> I am pushing my luck asking for 1920x1200? :-)
  • [15:29:23] <likewise> I'm officially enjoying a holiday weekend, am not allowed to read mail :-)
  • [15:30:01] <koen__> Bbl
  • [15:30:01] * koen__ (n=koen@94.157.56.165) Quit (Client Quit)
  • [15:30:23] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-bd70e3e384bd41a6) has joined #beagle
  • [15:30:44] <likewise> jkridner: my use case is this: ftp://ftp.axon.tv/WhitePapers/Application%20note%20HGQW200PD%20in%20multiview%20applications.pdf
  • [15:31:05] <likewise> jkridner: which is currently XScale+FPGA based -> slow graphics.
  • [15:31:11] * robclark (n=robclark@32.176.5.152) has joined #beagle
  • [15:31:32] <likewise> jkridner: the screenshots look ugly in that pdf, we beefed it up quite nicely using cairo/pango/freetype graphics.
  • [15:31:52] <jkridner> if the blanking interval and fps aren't critical, then the BeagleBoard resolution limitation is 2048x2048.
  • [15:33:28] <likewise> jkridner: hmm, we might be able to have the FPGA handle anything, so this might be just working.
  • [15:33:48] <jkridner> obviously BeagleBoards don't have all those inputs. :)
  • [15:34:26] <jkridner> I think you are going to get with an OMAP3530 a lot more video performance than you can get out of an X-Scale device.
  • [15:34:27] <likewise> jkridner: it would make a perfect graphics overlay engine (for real-time audio metering, video vector scopes etc).
  • [15:34:29] <likewise> bbiab
  • [15:35:40] <recalcati> swfdec-mozilla: anybody get it working in youtube ?
  • [15:37:06] <recalcati> no money to buy flash-lite
  • [15:38:58] <jkridner> have you tried gnash instead? http://wiki.gnashdev.org/YouTube
  • [15:39:01] * TAK2004 (n=thomas@dslb-088-072-106-205.pools.arcor-ip.net) has joined #beagle
  • [15:40:00] <Xenion> hello thomasg
  • [15:40:04] <Xenion> hello TAK2004
  • [15:40:05] <Xenion> :-)
  • [15:40:28] <Xenion> likewise, what FPGA are you going to use ? ( one from xilinx ? )
  • [15:41:27] * brolin (n=brolin@190.157.13.75) has joined #beagle
  • [15:41:52] * brolin (n=brolin@190.157.13.75) Quit (Client Quit)
  • [15:47:37] <likewise> Xenion: Altera Stratix 4
  • [15:47:48] <Xenion> likewise, i see
  • [15:49:38] * koen_ (n=koen@89.205.166.169) has joined #beagle
  • [15:49:45] <recalcati> jkridner: no, but I'll do if there is any chance
  • [15:49:54] <koen_> Grrrr
  • [15:50:14] <jkridner> koen_ ?
  • [15:50:15] * koen_ missed the connection
  • [15:50:45] <koen_> Trying to get home by train
  • [15:50:46] <jkridner> for recalcati? trying gnash for youtube.
  • [15:50:51] <recalcati> koen: you told that swfdec was better than gnash. as mozilla plugin I mean
  • [15:51:20] <jkridner> koen_: are you playing youtube vids on your beagle?
  • [15:51:38] * jkridner wonders what the #neuros folks use.
  • [15:51:47] <recalcati> swfdec-mozilla , there is gnash-mozilla ?
  • [15:52:15] <koen_> The gnash plugin doesn't build
  • [15:52:40] <recalcati> aaahh
  • [15:52:54] <recalcati> inside packages/gnash/gnash_0.8.4.bb ?
  • [15:53:09] <koen_> Gnash claims it's going to build it but dorsn't
  • [15:53:37] <recalcati> --with-npapi-plugindir=${libdir}/mozilla/plugins is present
  • [15:54:10] <koen_> See what i said above
  • [15:56:02] * koen_ ponders about dmaivideosink
  • [15:56:47] <jkridner> http://www.engadget.com/2007/06/27/neuros-osd-gets-beta-youtube-browser/ should be all open source on ARM.
  • [15:58:06] <recalcati> thx
  • [15:58:51] <koen_> That probably downloads the raw movie and hands it to a mpeg4 decoder
  • [15:59:09] <koen_> Like the iPhone does
  • [15:59:23] * erbo (n=erik@c-7b7de455.115-16-64736c14.cust.bredbandsbolaget.se) Quit (Read error: 104 (Connection reset by peer))
  • [16:00:41] <jkridner> maybe.... I'm poking around at http://wiki.neurostechnology.com/index.php/Summer_of_Code_2008/N-Tube_FLV_Browser
  • [16:01:37] <koen_> Did it actially work on arm?
  • [16:01:45] <recalcati> It seems an old post, but maybe. By now in PC gnash http://uk.youtube.com/active_sharing.swf works
  • [16:02:46] <recalcati> not a firefox plugin. but first of all get it work!
  • [16:03:55] <recalcati> they sell it: http://www.neurostechnology.com/
  • [16:04:36] * koen_ has an osd2
  • [16:04:38] <jkridner> it is an open source box.
  • [16:04:52] <Crofton|work> jkridner, can you help with the SoC stuff?
  • [16:04:59] <jkridner> yes.
  • [16:05:05] <Crofton|work> I am thinking about the application bits about what bb.org is
  • [16:05:05] <jkridner> Cathy has also agreed to help.
  • [16:05:14] <Crofton|work> Cathy ?
  • [16:05:25] <jkridner> University Programs @ TI.
  • [16:05:36] <Crofton|work> I think it will be helpful having a non-TI person listed as the admin
  • [16:05:38] <Crofton|work> cool
  • [16:05:39] <jkridner> should this be a TI think or a bb.org thing?
  • [16:05:42] <Crofton|work> that will be wesome
  • [16:05:46] <Crofton|work> bb.org
  • [16:06:14] <Crofton|work> can you ask her to review other peoples idea lists and web pages?
  • [16:06:24] <Crofton|work> there are some really strong groups out there
  • [16:06:25] <koen_> .org has more streetcred
  • [16:06:32] <Crofton|work> :)
  • [16:06:36] <Crofton|work> exactly
  • [16:06:52] <Crofton|work> I think we should shoot for about 10 good ideas
  • [16:07:03] * jkridner_ (n=jason@c-76-31-18-64.hsd1.tx.comcast.net) has joined #beagle
  • [16:07:06] <Crofton|work> and we need to make sure they are really doable
  • [16:07:22] <Crofton|work> not to broad and nebulous
  • [16:07:26] <koen_> And supportable
  • [16:07:46] <koen_> E.g. Licences
  • [16:08:25] <Crofton|work> awesome more FOSDEM video is up!@
  • [16:09:10] <jkridner_> there are GSoC requirements on responsiveness.
  • [16:09:46] <Crofton|work> heh
  • [16:09:50] <Crofton|work> for next year :)
  • [16:10:11] <Crofton|work> this year we need to get accepted as a mentoring organization, so we have no track record
  • [16:12:16] <Crofton|work> first step will be get accepted as a mentoring organization and have some slots assigned
  • [16:12:36] <jkridner_> I think we need to come in with some good project ideas.
  • [16:12:44] <Crofton|work> eaxctly
  • [16:13:22] <jkridner_> my main areas of care-about are browsers and DSP.
  • [16:13:38] <koen_> 3d, dsp, tweaks for things like xv, mplayer, myttv, etc
  • [16:13:43] <Crofton|work> we'll need a good mentor for DSP :)
  • [16:14:14] <jkridner_> I think we should be able to get a very good one.
  • [16:14:50] <Crofton|work> we need someone who knows exactly how it works, and is good working with really green people :)
  • [16:14:52] * ceyusa (n=ceyusa@cm40051.red83-165.mundo-r.com) has joined #beagle
  • [16:14:52] <koen_> Gat-
  • [16:15:27] <ceyusa> which kernel repository do you use with dspbridge?
  • [16:15:35] <koen_> Gst-ti should be ready for general consumption next month
  • [16:15:49] <jkridner_> ceyusa: probably the one from omapzoom.org.
  • [16:16:05] <ceyusa> jkridner, thanks, I'll look at it
  • [16:16:09] <jkridner_> not sure if they pulled BeagleBoard support into that kernel yet.
  • [16:16:52] <jkridner_> should be relatively simple to pull in BeagleBoard support there.
  • [16:17:14] <ceyusa> mmmhh. interesting...
  • [16:17:32] <jkridner_> I believe felipec's BeagleBoard-dspbridge howto was written using the linux-omap kernel.
  • [16:17:47] <jkridner_> ...but, support for dspbridge is best on omapzoom.org.
  • [16:18:06] <koen_> Dsplink would be better supported on beagle
  • [16:18:15] <jkridner_> ceyusa: have you seen http://elinux.org/BeagleBoard/DSP_Howto ?
  • [16:18:23] <likewise> sounds like a merger project :-)
  • [16:18:50] <ceyusa> jkridner, i'm looking at it at this moment
  • [16:19:00] <koen_> Omapzoom is a 'vendor' kernel
  • [16:19:39] <koen_> So noone is touching that with a bargepole
  • [16:20:07] * jkridner_ is not disagreeing with koen), but there is a push to merge dspbridge into linux-omap. for BB, however, we like dsplink.
  • [16:20:50] <kulve> the dspbridge folks have been (at least were in december) updating a dsp bridge patch against l-o
  • [16:20:59] <kulve> but it seems not to work anymore
  • [16:21:49] * ceyusa heard by first time the expression "Wouldn't touch it with a bargepole" :D
  • [16:22:21] <ceyusa> hi kulve
  • [16:22:26] <kulve> hi
  • [16:23:34] * jkridner (n=jason@c-76-31-18-64.hsd1.tx.comcast.net) Quit (Read error: 110 (Connection timed out))
  • [16:23:44] * jkridner_ is now known as jkridner
  • [16:25:09] * jkridner points to http://wiki.omap.com/index.php?title=GStreamer for dsplink info
  • [16:25:50] * as_leep is now known as ali_as
  • [16:25:50] <ceyusa> afaik dsplink will replace dspbridge, isn't it?
  • [16:26:06] <kulve> lot's of omap related domains by TI..
  • [16:26:56] <likewise> actually, that's my main problem; gathering all the info from too many sources. :-)
  • [16:28:17] <Xenion> likewise, ACK
  • [16:28:47] <Xenion> best anchor seems to be the elinux wiki ( but the informations are somehow not alway acurate )
  • [16:28:47] <jkridner> my belief is that they will be merged.
  • [16:28:57] <jkridner> *belief*
  • [16:29:20] <jkridner> they share a similar legacy.
  • [16:29:33] <jkridner> they just evolved a bit differently.
  • [16:29:49] <likewise> jkridner: yeah, evolution
  • [16:30:19] <recalcati> there should be a list of stable things: oe hash could be enough to resume all maybe
  • [16:30:49] <kulve> jkridner: the example gst pipeline on that page uses some TI specific video sink (which probably doesn't interact with X?). Does the decoder assume that or can I just use the defacto standard xvimgesink?
  • [16:30:51] <likewise> recalcati: for beagleboard builds only?
  • [16:30:52] <jkridner> the most examination by TI broad-market folks happens at wiki.omap.com.
  • [16:31:09] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has left #beagle
  • [16:31:11] <jkridner> most work for TI wireless folks happens at http://omapzoom.org.
  • [16:31:37] <likewise> recalcati: in fact, the hash + the bitbake revision go into my manual log book for most targets
  • [16:31:42] <jkridner> eLinux is a community wiki for embedded Linux and has a lot of interest in BeagleBoard.
  • [16:31:57] <ceyusa> the wiki seems more about the davinci family
  • [16:32:07] <ceyusa> meanwhile omapzoom goes more for... omap :)
  • [16:32:13] <jkridner> indeed.
  • [16:32:22] <koen_> kulve: Ti sinks can help to avoid extra memcpys
  • [16:32:24] <likewise> jkridner: there should be some meta page that links to all the different sources. the next problem then is that some sources are outdated and some info supersedes other.
  • [16:32:31] <jkridner> that is because the folks that brought us davinci are the ones that took omap3 to the broad market.
  • [16:32:40] <kulve> koen_: and needs some extra work from everybody using them
  • [16:32:44] <jkridner> whereas omapzoom.org is the wireless folks.
  • [16:32:47] <kulve> there are APIs for a reason..
  • [16:33:10] <jkridner> we are going to do some clean-up to make it more "one TI", but that isn't a top priority in my view.
  • [16:33:16] <koen_> kulve: Xvsink works, btw
  • [16:33:25] <kulve> ok
  • [16:33:36] <koen_> Cluttersink as well :D
  • [16:33:54] <recalcati> I'm really worried about a thing told me koen: if I 'git pull' today having ABCDE hash that is stable, then tomorrow I do it again and so on, I'm not sure to get the ABCDE stable version, due to binary packages not recoverable
  • [16:34:41] * koen__ (n=koen@89.205.166.169) has joined #beagle
  • [16:34:41] * koen_ (n=koen@89.205.166.169) Quit (Read error: 104 (Connection reset by peer))
  • [16:35:19] <likewise> recalcati: hmm, did koen mean the binary downloads/ or SCM managed binaries?
  • [16:35:50] <koen__> Using an old oe version with recent angstrom feeds is not recommended
  • [16:35:54] <recalcati> I suppose
  • [16:36:18] <likewise> ah, this was about feeds, not builds from scratch?
  • [16:36:21] <recalcati> angstrom feeds, exaclty what are?
  • [16:36:52] <jkridner> Crofton|work: did you answer my question about who the administrator should be?
  • [16:37:22] <Crofton|work> I think so, I volunteered :)
  • [16:37:40] <jkridner> ah, excellent.
  • [16:37:54] <Crofton|work> however, help is appreciated :)
  • [16:37:59] <jkridner> do you have the bandwidth? can you meet the 48 hour response-time and other requirements?
  • [16:38:03] <recalcati> I always build from scratch, but I'm afraid there is something not versioned.
  • [16:38:04] <Crofton|work> yeah
  • [16:38:26] <Crofton|work> if things get insane, I'm sure we can develop a backup plan
  • [16:38:30] <Crofton|work> which we should have
  • [16:38:33] <jkridner> you have Cathy and I to pull on. I'll send an intro to Cathy.
  • [16:38:48] <jkridner> you just need to pull us in quickly if there are times you cannot be responsive.
  • [16:38:54] <Crofton|work> yep
  • [16:38:59] <Crofton|work> thanks
  • [16:39:01] <jkridner> I like the idea of having it be someone non-TI.
  • [16:39:11] <Crofton|work> Having non-Ti as primary admin is good for appearances
  • [16:39:21] <jkridner> ...and for reality.
  • [16:39:24] <Crofton|work> heh
  • [16:39:34] <Xenion> Crofton|work, but not good for "first hand knowledge"
  • [16:39:45] <Crofton|work> Hopefully someone proposes an Android idea :)
  • [16:39:48] <jkridner> it should be about advancing the state of the art, not TI's agenda.
  • [16:39:53] <Crofton|work> admin is just that
  • [16:40:02] <Crofton|work> TI-mentors would be awesome
  • [16:40:08] <jkridner> Mentors need to have the first-hand knowledge.
  • [16:40:24] <Crofton|work> or be willing to learn :)
  • [16:40:31] <jkridner> I'm going to try to convince Gagan to be a mentor.
  • [16:40:46] <jkridner> he worked on the 'dsplib'.
  • [16:40:51] <Crofton|work> mru has said he is willing to help also, but I do not think he wants to be an official mentor
  • [16:40:56] <Crofton|work> awesome
  • [16:41:30] <Crofton|work> I think I can steer someone into the right place to insert the first NEON into fftw
  • [16:41:37] <jkridner> so, main thing is for us make a good "ideas" page now?
  • [16:41:41] <Crofton|work> yes
  • [16:42:19] <koen__> Btw, bluez has neon code for sbc
  • [16:42:21] <Crofton|work> then I may need some help
  • [16:43:05] <Crofton|work> we have four days when the application is open, I'm going to be skiing from March 4-9
  • [16:43:06] <jkridner> Crofton|work: do you know what the mentoring organization application requirements/layout and e-mail address?
  • [16:43:14] <Crofton|work> not sure how well connected we will have
  • [16:43:24] <Crofton|work> they will want my gmail account info
  • [16:43:50] <Crofton|work> the application is covered in the faq, but I do not think we can see the app until it opens
  • [16:44:01] <koen__> jkridner: Maybe pratheesh has some ideas for sgx stuff?
  • [16:44:09] <Crofton|work> There was a talk at FOSDEM but the GSoC woman
  • [16:44:27] <jkridner> koen__: probably.
  • [16:44:27] <Crofton|work> we will need to be careful with sgx due to licensing headaches
  • [16:44:43] <Crofton|work> no video yet, but when it appears, I'll let you know
  • [16:44:46] <jkridner> would be good to have Mentors in India if there are students there.
  • [16:45:19] <Crofton|work> anyone can be a mentor or student, unless you are in the "Axis of Evil"
  • [16:45:21] * koen__ (n=koen@89.205.166.169) Quit (Read error: 104 (Connection reset by peer))
  • [16:45:33] <Crofton|work> Google has figured out how to pay everyone
  • [16:45:42] <Crofton|work> also, we will get $500 per student
  • [16:45:53] <Crofton|work> but that is getting ahead of ourselves :)
  • [16:46:07] * jkridner wants Ubiquity with voice recognition. :)
  • [16:46:23] * Robot_ (n=robot@rb5bx242.net.upc.cz) has joined #beagle
  • [16:46:44] * Crofton|work wants about 10 solid ideas doable by college age people over the summer
  • [16:47:18] <Crofton|work> koen, feel free to submit an idea for a project and we can find a mentor
  • [16:47:19] <robclark> kulve: I have some interest to make TI video sink play nicely with X11
  • [16:47:32] <Crofton|work> I think we can work around your schedule for the evaluations
  • [16:47:57] * DaQatz (n=db@c-66-30-48-150.hsd1.nh.comcast.net) has joined #beagle
  • [16:49:02] <robclark> at least the gst-goo on omapzoom omx..
  • [16:51:06] <robclark> not exactly sure the best route yet, but I'm just starting to look into it
  • [16:51:20] <jkridner> robclark: do you know if the omapzoom builds a kernel for the BeagleBoard?
  • [16:51:52] <jkridner> should patches be submitted to linux-omap for that purpose?
  • [16:52:42] * geckosenator (n=sean@71.237.94.78) has joined #beagle
  • [16:53:43] * robclark is wondering if there is a way to get X11 to not draw certain pixels on the screen so we could have OMX render to a framebuffer beneath what X11 is using, and visible thru the "hole"
  • [16:54:43] <robclark> not sure..
  • [16:54:48] <kulve> robclark: overlay
  • [16:54:49] <robclark> I guess I'll find out when revC is avail, and I actually have a BB of my own ;-)
  • [16:55:21] <Crofton> jkridner, http://code.google.com/p/google-summer-of-code/wiki/AdviceforMentors
  • [16:55:36] <kulve> xf86-video-omapfb currently draws the video on the overlay, so it's always on top of the framebuffer X uses (including the pointer cursor atm)
  • [16:57:41] <robclark> well, I guess if BB support was in linux-omap kernel, it would eventually make it's way into omapzoom kernel (but I confess to not really knowing how and when things get merged between all these different trees)
  • [16:58:25] <kulve> xf86-video-omapfb works only with the l-o DSS1 and DSS2. It doesn't work with the v4l in TI's kernels
  • [16:58:25] <robclark> exactly
  • [16:59:58] * jkridner feels unable to deliver points for this month's software contest as I cannot see or reproduce any of the entries. :(
  • [16:59:59] <robclark> just need to make X11 draw some 100% transparent pixels where the video output shows through
  • [17:00:52] <jkridner> I can tell that all are very good efforts, but they have some work to do to declare completion.
  • [17:02:08] <kulve> robclark: there's no point in drawing pixels that are not shown ;) Waste of cpu and memory bandwidth
  • [17:03:12] <robclark> it should work as long as the user doesn't move the X11 window "containing" the video output.. if the user moves, I'm not really sure how to keep the window position and video output planes synchronized
  • [17:03:38] * jkridner is inclined to send them each a Rev C as "approved projects", instead of declaring them "contest winners".
  • [17:04:03] <kulve> robclark: XV does that. The whole point of the XV is that the video moves when the window moves. And the video uses hw scaling, when the X window is resized
  • [17:04:30] <kulve> robclark: http://cgit.pingu.fi/xf86-video-omapfb/tree/src
  • [17:04:45] <kulve> http://cgit.pingu.fi/xf86-video-omapfb/tree/src/omapfb-xv-generic.c
  • [17:05:30] * ssvb (n=ssvb___@a88-114-221-132.elisa-laajakaista.fi) has joined #beagle
  • [17:07:34] <robclark> kulve: yeah, I was looking at that... but my understanding is that there is an accelerated kdrive xserver from imagination which we'd like to use.. so something that depends on xf86-video-omapfb might not be an option.. at least for my "day job" part of this.. and anyways, I have to use v4l in omapzoom kernel
  • [17:08:28] * jkridner has voted: http://elinux.org/BeagleBoard/contest#Points_2
  • [17:08:41] <kulve> robclark: there's probably no benefit from using the sgx stuff for video playback.
  • [17:08:42] <robclark> but I've only just started looking into this a couple days ago, so who knows
  • [17:09:46] <kulve> I've started to add the v4l playback for TI kernels there but haven't really made any progress yet
  • [17:09:46] <robclark> well, the framebuffer should overlay the different video planes.. so that isn't CPU bandwidth..
  • [17:09:46] <robclark> hmm
  • [17:10:49] <kulve> robclark: atm the x-v-o copies the input data to the video plane. AIUI that might be avoided when no color conversion needs to be done. There's a neon version of the color conversion implemented already though
  • [17:10:50] <robclark> yeah, I'd been looking at that the other day
  • [17:13:56] <robclark> kulve: can I get the X11 server to allocate some physically contiguous memory that I can pass down to OMX to render output to? Otherwise I think there would still be an extra memcpy()
  • [17:14:47] * fischer1 (n=tfischer@fischerfamily.org) Quit (Remote closed the connection)
  • [17:15:34] <kulve> I'm not sure about that, it might be possible but needed something extra in the XV driver. zuh might remember (he's implemented the x-v-o driver mostly)
  • [17:15:34] <koen> robclark: I've tested Xsgx
  • [17:15:36] <robclark> well, I think the graphics core could even handle color conversion and scaling.. (but I'm still learning about this)
  • [17:15:56] <koen> robclark: http://dominion.thruhere.net/koen/OE/Xsgx-Xord.perfdiff.txt
  • [17:16:13] <koen> robclark: the non IMGtec xserver is up to 22 times faster
  • [17:16:17] <kulve> robclark: depending on the codec there's not neccessarily need for a color conversion
  • [17:16:45] * koen should redo those benchmarks no pixman has some neon
  • [17:16:45] <robclark> actually, I think right now the colorspace conversion is done on the DSP.. or at least it is at the OMX layer (which I've not looked too closely at yet.. but I think OMX just mmaps() /dev/video1 and simply copies the decoded frame on the ARM side
  • [17:16:46] <kulve> robclark: and the framebuffer hw does the scaling, no need to use extra hw for that
  • [17:17:25] <koen> robclark: with dsplink the dsp will give you yuv422 data, which can can feed to the resizer hardware or overlay directly
  • [17:17:25] <robclark> I'm just looking at OMAPFBXVPutImageGeneric()... what happens to 'buf' if the buffer is already in RGB?
  • [17:18:12] * fischer (n=tfischer@fischerfamily.org) has joined #beagle
  • [17:19:16] <robclark> kulve: ok, so you just hit the default: in that switch and nothing needs to be done
  • [17:19:49] * valhalla (n=valhalla@81-174-25-183.dynamic.ngi.it) Quit ("Leaving")
  • [17:19:51] <robclark> koen: hmm, interesting
  • [17:20:01] <kulve> robclark: no decoder gives out RGB, they are always some of those yuv formats
  • [17:20:02] <robclark> koen: well, if xf86-video-omapfb is faster, then it is an easier sell to use it
  • [17:20:35] <koen> robclark: I hope to get some 'numbers' on using different sinks and resize methods
  • [17:21:58] * koen only has 4 days left to get work done for his bossa presentation
  • [17:23:33] <koen> robclark: there's #gst_ti where gstreamer-on-dsplink stuff is being done
  • [17:23:36] <robclark> hmm, ok.. well I'll I've seen from the code so far is it is copied straight to mmap's /dev/video1.. but I wasn't sure what format it was
  • [17:23:46] <koen> robclark: also: gstreamer.ti.com :)
  • [17:23:47] <robclark> koen: ok, that would be interesting to know
  • [17:25:25] <kulve> robclark: keep me posted on your progress. And please ask if there's questions (although I might not be able to answer then :)
  • [17:26:29] <robclark> at any rate, once I wrap up a couple unrelated task (hopefully this next week), then I should have more time to dig in and learn a few things (rather than just in the time that I'm waiting for other stuff to compile ;-))
  • [17:28:14] * BThompsonD (n=bernie@cpe-72-190-75-99.tx.res.rr.com) has joined #beagle
  • [17:30:53] <eFfeM> jkridner: i once saw the problem with libc++, what I did is abort after a while, and restart and then it works.
  • [17:30:53] <robclark> koen: fwiw, I'm actually on the WTBU side.. so dspbridge and socketnodes.. I'm not sure how much that differs from the gstreamer.ti.com stuff (I'm still confused about all the different TI approaches to dsp stuff ;-))
  • [17:31:21] <eFfeM> not sure if this is related to opkg or the angstrom server
  • [17:31:21] <robclark> (disclaimer: I'm pretty new to WTBU/OMAP stuff)
  • [17:31:38] <eFfeM> and you can always use the image that I provided as an alternative
  • [17:31:41] * kozak (n=subbu@117.192.20.35) Quit (Read error: 60 (Operation timed out))
  • [17:31:41] <robclark> will do... I'm sure I'll have plenty of questions
  • [17:31:47] <eFfeM> koen, have you tried if gstreamer works for you?
  • [17:32:42] <eFfeM> rwhitby will also send pictures but have problems getting my u-boot env to save no idea what is wrong there, so can't boot into 720p
  • [17:33:24] <kulve> robclark: TI wireless have gst-goo on top of their openmax on top of dsp bridge. TI catalog group have their gst on top of dsk link
  • [17:33:49] <Crofton|work> http://ftp.heanet.ie/mirrors/fosdem-video/2009/maintracks/reverse.ogv
  • [17:33:54] <recalcati> I still a script to get 720p, but which is the right kernel? I see in openembedded that 2.6.28 and 2.6.29 are no more updated
  • [17:34:02] <recalcati> I still have
  • [17:34:13] <koen> hmmm
  • [17:34:19] <koen> why do you want it both ways?
  • [17:34:30] <koen> first you complain about OE being updated so much
  • [17:34:41] <koen> and now you complain about something not being updated so much
  • [17:34:46] <koen> make up your mind!
  • [17:34:47] <robclark> kulve: in theory, it should be the same OMX API, so the stuff below OMX shouldn't matter...
  • [17:34:58] <kulve> robclark: that would be nice
  • [17:35:11] <kulve> but my guess is that's only in theory..
  • [17:35:46] <recalcati> koen: I try to understand, not complain. I know yours is "the work" . mine are only cilly ideas
  • [17:36:20] <kulve> robclark: if you think OE updates too frequently, checkout http://linux.onarm.com. It has stable releases every 6 months. Although it doesn't have most of the stuff OE/?ngstr?m has (e.g. TI's openmax..)
  • [17:36:50] <kulve> I'm working with that one..
  • [17:38:20] <recalcati> kulve: is something similar to oe? it isn't oe enough?
  • [17:38:21] <robclark> yeah, it would be nice if theory == reality, but I'll believe it when I see it
  • [17:38:21] <robclark> ok
  • [17:39:20] * vijay (n=vijay@203.199.213.3) Quit ("Leaving")
  • [17:40:22] <robclark> I've actually been playing with this: http://github.com/mturquette/meta-texasinstruments
  • [17:41:54] <recalcati> so we have: oe,gentoo,http://linux.onarm.com,http://github.com/mturquette/meta-texasinstruments,android,ubuntu, .... really crazy
  • [17:41:55] <robclark> so in theory it should be possible to pull in some recipes from OE
  • [17:45:01] <robclark> but I might do something different on my own time when I get a beagle, since poky is a bit better suited to a small screen
  • [17:45:27] <kulve> recalcati: those are also quite different things. e.g. ALIP (from linux.onarm.com) isn't meant to be a ready end-user distro. But I like it a lot for developing
  • [17:45:38] * eFfeM is now known as eFfeM-away
  • [17:47:05] <recalcati> kulve: I'll see it better
  • [17:47:53] <recalcati> my daughter is hungry
  • [17:47:56] <robclark> recalcati: variety is the spice of life ;-)
  • [17:49:03] <recalcati> I'm afraid of the atom enemy
  • [17:49:09] <kulve> recalcati: it uses Scratchbox for cross compiling. That's one big difference compared to the rest of the mentioned
  • [17:50:13] <robclark> the meta-ti isn't really a distro, but an overlay on top of poky.. so more suitable for small screens (ie. phones)
  • [17:51:23] * Xerion (i=xerion@82-170-197-160.ip.telfort.nl) Quit (Read error: 104 (Connection reset by peer))
  • [17:52:28] * Xerion (i=xerion@82-170-197-160.ip.telfort.nl) has joined #beagle
  • [17:55:09] * arne-unicap (n=arne@p548198F3.dip0.t-ipconnect.de) has joined #beagle
  • [18:00:13] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [18:08:04] * rsalveti (n=salveti@200.222.2.234) Quit (Remote closed the connection)
  • [18:09:16] * rsalveti (n=salveti@200.222.2.234) has joined #beagle
  • [18:10:15] * hli (i=chaton@vig91-2-82-232-97-149.fbx.proxad.net) has joined #beagle
  • [18:11:17] <hli> hi, i'd like to have an info about linux memory mapping : where the user space resides ?
  • [18:13:16] <ali_as> Hmm, this can vary.
  • [18:14:25] <ali_as> I patched a program that was aborting because the allocated memory was in a different area and it was doing a signed check.
  • [18:15:51] <hli> well, i'd like to know if i can reserve the first 256 MiB (strating from 0) or the last 256 MiB of 4GiB
  • [18:16:17] <ali_as> With an MMU is that even a sane question?
  • [18:16:46] <hli> i need to have it on a specific address
  • [18:17:07] <hli> i'm working on an emulator
  • [18:18:08] <ali_as> Hmm, I see.
  • [18:18:40] * mib_4wibkx (i=614d31af@gateway/web/ajax/mibbit.com/x-dd5a074bbc253ceb) has joined #beagle
  • [18:18:55] <mib_4wibkx> hi
  • [18:19:06] <mru> hli: check what qemu does
  • [18:19:14] <ali_as> hli, what exactly are you emulating?
  • [18:20:41] <ali_as> Are you trying to make it so the memory of the emulated system matches resides in the host with 1:1 address matching?
  • [18:20:48] <hli> yes
  • [18:21:26] <hli> but with a AND or a OR to make the emulated address into the 256 MiB reserved space
  • [18:21:40] <ali_as> I see.
  • [18:22:09] <ali_as> Would adding a base address not be just as fast?
  • [18:22:13] <mru> you can make the emulator executable load at some unused address with linker flags/scripts
  • [18:22:40] <hli> MIPS to ARM
  • [18:22:47] <mru> then map the areas you want with mmap(MAP_FIXED) before anything else steals it
  • [18:23:12] <mru> watch out for shared libraries
  • [18:23:28] <hli> I need to AND or OR anyway
  • [18:23:46] <mib_4wibkx> I am new to beagle board .. .. my current status is i have installed Angstrom on SD card and get the demo working .... Now I would like to write a simple Hello World Program in C .... I will be using ARM cross-compiler from CodeSourcery. But now how to execute the generated binary on Beagle board
  • [18:23:54] <ali_as> Why do you need to AND/OR anyway?
  • [18:24:25] <ali_as> Are you trying to emulate a mips system on an arm system or the other way around?
  • [18:24:33] <hli> as 0x00000000 and 0x80000000 and 0xC0000000 is same physical address in emulated machine (MIPS
  • [18:24:57] * hli_ (i=chaton@vig91-2-82-232-97-149.fbx.proxad.net) has joined #beagle
  • [18:25:00] <recalcati> bye bye . this evening no pc.
  • [18:25:06] <ali_as> Ahh, I see, yes it is.
  • [18:25:38] <ali_as> Though you may have bigger problems.
  • [18:25:59] <ali_as> There is a 36bit addressing mode and a 64bit addressing mode.
  • [18:26:01] <mru> hli: isn't 0xa0000000 also an alias (uncached)?
  • [18:26:02] * recalcati (i=5d907169@gateway/web/ajax/mibbit.com/x-bd70e3e384bd41a6) Quit ("http://www.mibbit.com ajax IRC Client")
  • [18:26:27] <hli_> mips addr ==> (addr & (256MiB-1) + base)
  • [18:26:41] <hli_> mru : yes you're right
  • [18:26:56] <hli_> it is A0000000 not C0000000
  • [18:27:23] <ali_as> What is the mips system?
  • [18:27:28] <hli_> if i could use bas as offset
  • [18:27:32] <hli_> psp
  • [18:27:36] <ali_as> Ahh.
  • [18:28:06] <mib_4wibkx> I am new to beagle board .. .. my current status is i have installed Angstrom on SD card and get the demo working .... Now I would like to write a simple Hello World Program in C .... I will be using ARM cross-compiler from CodeSourcery. But now how to execute the generated binary on Beagle board
  • [18:28:52] <hli_> i work on jpcsp and pcsp and have my own psp emulator i plan to make it run on pandora
  • [18:29:01] <hli_> but i also have a beagleboard
  • [18:29:06] <geckosenator> mib_4wibkx: did you compile the binary yet?
  • [18:30:05] <hli_> so i'm looking at a very fast way to emulate read and write access of psp on cortex A8 taking into account of dual-issue
  • [18:30:37] <mib_4wibkx> geckosenator: no not yet
  • [18:30:44] <geckosenator> mib_4wibkx: try to do that first then
  • [18:31:18] <mib_4wibkx> I know how to do that ... i used the same for omapzoom
  • [18:32:28] <koen> mib_4wibkx: angstrom doesn;t use codesourcery, so you binary might not run
  • [18:33:22] <mru> koen: codesourcery compilers are binary compatible with any gcc4
  • [18:33:22] <hli_> if base can be 0xFF000000 i can make "LW %0, off(%1)" into : MOVW %t1, #off; AND %t2, %1, #FF000000; LDR %0, [%t1 + %t2] <- 1.5 cycles !
  • [18:33:44] <koen> mru: even C++?
  • [18:33:44] <mib_4wibkx> oh .. so how should i compile my C code such that it can run on Angstrom
  • [18:33:59] <koen> mib_4wibkx: use openembedded
  • [18:34:11] <mru> koen: should be
  • [18:34:25] * robclark (n=robclark@32.176.5.152) Quit (Read error: 110 (Connection timed out))
  • [18:34:38] * CmosReloaded (n=netdevil@cpc4-glfd1-0-0-cust691.glfd.cable.ntl.com) has joined #beagle
  • [18:34:43] <koen> mru: I don't trust c++ compat between gcc minor releases
  • [18:35:08] <mru> I don't trust c++
  • [18:35:17] <koen> true :)
  • [18:35:31] * dirk2 (n=dirk@p5B04081F.dip0.t-ipconnect.de) Quit ("Konversation terminated!")
  • [18:35:49] <mib_4wibkx> mru : so can i used binaries generated using codesourcery angstrom
  • [18:36:11] <mru> if you have compatible glibc versions, yes
  • [18:36:34] <mru> koen: what glibc version does angstrom use?
  • [18:36:41] <koen> mru: 2.7
  • [18:36:54] * arne-unicap (n=arne@p548198F3.dip0.t-ipconnect.de) Quit ("Verlassend")
  • [18:36:54] <koen> you need to match compile and linkerflags as well
  • [18:37:07] <mib_4wibkx> koen : how to do that
  • [18:37:14] <koen> mib_4wibkx: use openembedded
  • [18:37:17] <mru> koen: what do you mean?
  • [18:37:38] <mru> unless you do something crazy, like select fpa float, it should work fine
  • [18:37:42] <koen> mru: if you instruct the csl toolchain to make OABI binaries...
  • [18:37:49] <mru> that would be stupid
  • [18:37:53] <mru> it's not even the default
  • [18:37:56] <koen> right
  • [18:38:01] <koen> I've seen worse
  • [18:38:03] <ali_as> hli, I am very rusty, but your last arm instruction, is that valid?
  • [18:38:04] <hli_> ali_as : 36-bit and 64-bit addressing mode ? for what ?
  • [18:38:07] <geckosenator> mru: unless you are debugging it..
  • [18:38:23] <mru> geckosenator: why would you care about oabi at all?
  • [18:38:24] <koen> people insisting using a toolchain instead of OE are usually stubborn
  • [18:38:35] <koen> and 90% of the time not the good kind of stubborn
  • [18:38:36] <mru> koen: you mean people like me?
  • [18:38:36] <geckosenator> koen: the latest rootfs tarball on your site gives me errors when extracting
  • [18:38:44] <hli_> ali_as : LDR ? i guess so
  • [18:38:54] <koen> mru: you fall in the 10% of good type of stubborn :)
  • [18:38:56] <ali_as> LDR with 2 variables to add.
  • [18:39:03] <hli_> yes
  • [18:39:04] <koen> i.e. know what you're doing
  • [18:39:07] <mru> hli_: ldr only supports 32-bit addresses
  • [18:39:11] <ali_as> I thought you could add a fixed offset.
  • [18:39:14] <geckosenator> mru: it wouldn't be for the beagle.. but you could debug it on the beagle
  • [18:39:23] <mru> the mmu on some chips might support wider physical addresses
  • [18:39:25] <hli_> LDR address operand is very flexible
  • [18:39:51] <mru> geckosenator: or you could "debug" the other system by switching it to eabi
  • [18:39:59] <hli_> you can even have : LDR Rd, [Rn +/- Rm <SHIFT> #n]
  • [18:40:19] <mru> hli_: don't for get the base writeback
  • [18:40:29] <mru> forget
  • [18:40:43] <ali_as> hli, I don't know which addressing modes the psp chip supports. I did something a year or so ago with a 64bit chip for set top boxes. It had a 64bit addressing mode, a 36bit addressing mode, a 32bit addresing mode and something like a 26bit physical hardware address. MIPS can be nuts.
  • [18:40:51] * rob_k (i=5a9c259a@gateway/web/ajax/mibbit.com/x-56222e2653d7748d) Quit ("http://www.mibbit.com ajax IRC Client")
  • [18:41:02] <mru> ali_as: oh, you're talking about the mips
  • [18:41:09] <hli_> mru: base writeback ?
  • [18:41:14] <mru> mips64 does have 64-bit addresses
  • [18:41:26] * hli (i=chaton@vig91-2-82-232-97-149.fbx.proxad.net) Quit (Read error: 110 (Connection timed out))
  • [18:41:36] <mru> hli_: ldr Rd, [Rn, index]!
  • [18:41:39] <hli_> ali_as : psp is only 32-bit (register and address)
  • [18:42:01] <ali_as> That simplifies things hugely, I see why you want to fiddle with host address space.
  • [18:42:33] <hli_> mru : psp has no writeback so i don't need for arm generated ARM code unless i miss something
  • [18:43:02] <mru> no, mips doesn't have such addressing modes
  • [18:43:04] <ali_as> He's not implimenting a stack, he's converting between address spaces.
  • [18:43:18] <mru> that doesn't mean it might not be useful when emulating something
  • [18:43:56] <hli_> hummm, maybe you're right
  • [18:44:17] <hli_> but for what i need to emulate i don't see the use of writeback
  • [18:44:32] * koen notices he now has an ajay kumar and a jaya kumar in his beagle addressbook
  • [18:44:39] <mru> perhaps not, I just thought I'd mention it
  • [18:44:52] <ali_as> LDR was at least 2 clocks on the ARM9 I think. Hmm.
  • [18:45:20] <hli_> ali_as: immediate offset is only interesting if i could fit it (mips have 16-bit offset in load/store operations)
  • [18:45:31] <mru> ldr has two cycles result latency on cortex-a8
  • [18:45:35] <mru> one issue cycle
  • [18:45:49] <mru> so you can do something else while awaiting the result
  • [18:45:55] <ali_as> Excellent.
  • [18:46:47] <ali_as> hti, is this a recompiler of some sort btw?
  • [18:46:49] <hli_> and as A8 is dual issue i cannot make AND %t1, %1, 0xFF000000; LDR %0, [%t1, #off] into 1 cycle but into 2 cycles instead :/
  • [18:46:52] <hli_> yes
  • [18:47:40] <ali_as> I remeber the N64 HLE a few years ago. SPent some time in softice looking at what it did ;)
  • [18:48:06] <ali_as> Actually, damn, a lot of years ago.
  • [18:48:12] <hli_> having MOVW %t1, #off; AND %t2, %1, #0xFF000000; LDR %0, [%t1 + %t2] make it into 1.5 cycles which equivalent kinda
  • [18:48:55] <hli_> so it doesn't worth while considering an immediate offset in LDR
  • [18:50:00] <ali_as> Why is that not 1.5 cycles with immediate offset?
  • [18:50:10] <hli_> dual issue
  • [18:50:18] <hli_> register dependecy
  • [18:50:44] <hli_> AND and LDR have register dependency so they should not be in the same row
  • [18:50:44] <ali_as> But the second cycle you can add another instruction, no?
  • [18:51:17] <hli_> MOVW and AND has no register dependency so they can bow run in the same cycles
  • [18:51:44] <hli_> s/bow/both/g
  • [18:52:11] <mru> I don't see any dependency
  • [18:52:48] <hli_> AND %t1, ...; LDR ...,[%t1 + ...]
  • [18:53:02] <mru> ah, was reading the wrong thing...
  • [18:53:05] <hli_> they cannot be run in the same cycle because of %t1
  • [18:53:13] <ali_as> I see that as 1.5 cycles.
  • [18:53:16] <mru> but whatever comes after ldr can run in the same cycle
  • [18:53:27] <hli_> yes but this is not teh issue
  • [18:53:44] <geckosenator> does anyone get "tar: Error exit delayed from previous errors" when trying to extract Angstrom-Beagleboard-demo-image-glibc-ipk-2009.X-test-20090210-beagleboard.rootfs.tar.bz2?
  • [18:53:46] <hli_> as i need to run AND *BEFORE* LDR
  • [18:54:01] <mru> yes, but surely you'll have *something* after the ldr
  • [18:54:12] <mru> and whatever that is, can run in the same cycle as ldr
  • [18:54:31] <mru> if ldr dual-issues with the previous instruction, the following one will need a new cycle
  • [18:54:44] <mru> same number of cycles for all three instructions
  • [18:54:54] <hli_> LW %0, ...; ADD %0, %0, 1; SW %0, ... <-- not sure
  • [18:55:03] <mru> so do something else
  • [18:55:22] <mru> actually, you want to put two instructions between there
  • [18:55:29] <mru> otherwise the add will stall
  • [18:55:46] <hli_> that's MIPS
  • [18:56:43] <mru> well, read ahead a bit and see if you find something that doesn't depend on that sequence
  • [18:57:00] <ali_as> That might require making the code very complex.
  • [18:57:05] * ghaz`` (n=theghaz@pomm.es) Quit ("changing servers")
  • [18:57:20] <hli_> anyway, MOVW/AND/LDR is a viable option to cover any 16-bit offset. I could consider a better way to issue ARM instruction for specific offset but that's not a priority
  • [18:57:28] <mru> you should read up to the next branch, work out dependency chains, and then issue arm instructions in the optimal order
  • [18:57:54] <ali_as> That would make sense if he knew the code that was going to be produced.
  • [18:58:06] <ZeZu> hli_ = hlide ?
  • [18:58:08] <ali_as> But it's a recompiler.
  • [18:58:12] <mru> so?
  • [18:58:15] <hli_> mru: i'm considering the possibility to have some ARM instruction reordering
  • [18:58:17] <hli_> yes
  • [18:58:22] <hli_> hli_ = hlide
  • [18:58:23] <ZeZu> what emulator you working on ?
  • [18:58:24] <mru> does it matter to a compiler whether the source is C or mips?
  • [18:58:39] <hli_> dynamic recompiler
  • [18:58:45] <ZeZu> for what processor ?
  • [18:58:51] * ghaz`` (i=theghaz@pomm.es) has joined #beagle
  • [18:59:02] <hli_> although i'm pretty sure we can consider a static recompiler for psp emulation
  • [18:59:22] <ZeZu> ah emulating psp on omap3 ?
  • [18:59:32] <ali_as> mru, a human could read ahead and reorder very easily. Making a program that could do it and keep track of dependency chains :/
  • [18:59:56] <ZeZu> hli_ i've already got one going for nulldc
  • [18:59:57] <mru> ali_as: that's what any good compiler does
  • [19:00:27] <ali_as> Then I've never seen a good compiler.
  • [19:00:34] <mru> they're rare
  • [19:00:48] <mru> but most compilers definitely do *some* instruction scheduling
  • [19:00:50] <hli_> ZeZu : you're not Dr||Raziel ?
  • [19:01:11] <ZeZu> no
  • [19:01:15] <hli_> mru: i think gcc arm compiler is not trustworthy
  • [19:01:20] <ZeZu> i'm not drk||raziel
  • [19:01:34] <hli_> especiall when neon is not well generated from what i heard
  • [19:01:37] <mru> even gcc-arm does some optimisations
  • [19:01:45] <mru> it sucks bigtime at neon
  • [19:01:52] <ZeZu> sure does
  • [19:02:07] <hli_> psp has a vfpu which needs neon to compete with
  • [19:02:21] <mru> but I've often seen it interleave instructions belonging to different C statements
  • [19:02:30] <mru> which is essentially what I'm talking about
  • [19:02:58] <mru> mips floating point is *much* faster than the cortex-a8 vfp
  • [19:03:16] <mru> arm has never been very good at floating-point
  • [19:03:37] <hli_> mru : i'd like the idea to have a static recompiler for psp, generating C source to compile once for each PRX modules (ELF binary and dynamic so equivalent)
  • [19:04:13] <hli_> FPU and VFPU are two different things
  • [19:04:14] <ZeZu> why would you want it to generate C, why not just generate object files ??
  • [19:04:36] <hli_> and what generates object files ?
  • [19:04:44] <ZeZu> the static rec ?
  • [19:04:52] <hli_> and how ?
  • [19:05:32] <mru> make a mips frontend for gcc
  • [19:05:33] <ali_as> I have a great way to produce C source, dissasemble and put the code in __asm{.
  • [19:05:33] <ZeZu> by translating the code and generating an object file with the data / function naming matched to its functional details, like prxname and addr/stub
  • [19:05:43] <hli_> are you implying a gcc is embedded in pandora/beagleboard ?
  • [19:06:19] <ZeZu> well you can do as you like in it, and processing object files isn't terribly difficult
  • [19:06:31] <ZeZu> or make a dynamic library out of the prx's as they are meant to be
  • [19:06:41] <hli_> true
  • [19:06:50] <ZeZu> just suggesting, seems a bit odd to make C source out of them, but no doubt it would work
  • [19:07:14] <ZeZu> inline asm in gcc sucks balls btw, i'd generate a .S file if you go that route
  • [19:07:23] <mru> going via C would also require a C compiler
  • [19:07:44] <ZeZu> well they would use the one on the host pc and build the bin for the omap3
  • [19:07:47] <mru> so if you're anyway going to use a C compiler, presumably gcc, why not make it read mips code directly
  • [19:08:13] <ZeZu> make a modified objdump or such ? :)
  • [19:08:23] <mru> a gcc-mips frontend could run on a pc as well
  • [19:08:27] <hli_> ZeZu: i prefer to use inline gcc rather than .S as you can let gcc optimze the register usage or inline asm code in c functions
  • [19:08:40] <mru> gcc doesn't optimise anything with inline asm
  • [19:08:44] <ZeZu> gcc doesn't optimize iniline asm
  • [19:08:49] <mru> quite the opposite, in fact
  • [19:08:50] <ali_as> hli_, to think about your original question the only thing I can think about is if you use the MMU directly to map the memory. I have no idea if this is feasable and I don't have an A8 manual to look it and find out.
  • [19:08:56] <ZeZu> yes it screws up your C code opts
  • [19:09:17] <ZeZu> request it from ARM
  • [19:09:46] <hli_> well, i was able to have so good tuned asm code, not always but it means you know all the whistles and bells about it
  • [19:09:52] <ZeZu> you mean request the proper address from the mmu ?
  • [19:10:21] <hli_> 1:1 mapping between PSP and A8
  • [19:10:23] <ali_as> I mean map memory in such a way as it does the work hli wants.
  • [19:10:34] <ZeZu> i thought about doing that
  • [19:10:41] <ZeZu> first of all unless you modify the kernel
  • [19:10:45] <ZeZu> its just a request
  • [19:10:55] <ZeZu> but you can ask mmap to give you a certain address
  • [19:11:26] <ali_as> Doesn't make it empty though.
  • [19:11:34] <ZeZu> uh
  • [19:11:40] <ZeZu> yes ..i think that would be a requirement no ?
  • [19:12:02] <ali_as> That adress might have something critical in it.
  • [19:12:04] <ZeZu> well it may have some garbage in it, but its not going to let you destroy another processes memory or the kernels
  • [19:12:19] <ZeZu> it would be pretty poor code to let you do that, i think it would fail first
  • [19:12:38] <MostAwesomeDude> Okay, so I've generated a console-image.
  • [19:12:50] <ali_as> It can be used to access ports directly, so accidents can happen.
  • [19:12:51] <MostAwesomeDude> I'm going to test on my Beagle, but I don't know exactly what to copy over.
  • [19:13:05] <hli_> ali_as: mru proposal is very interesting, i need to tweak the linker file so i can reserve the first 256 MiB
  • [19:13:08] <ZeZu> well perhaps you know more about it than i do
  • [19:13:14] <MostAwesomeDude> I have a JFFS2 image, and then some other files, .bin and .ift. I also have an SD card that's been formatted with a split.
  • [19:13:29] <Qball> MostAwesomeDude: http://elinux.org/BeagleBoardAndOpenEmbeddedGit
  • [19:14:24] <ali_as> hli, I confess, I don't understand how that would work. :)
  • [19:14:29] <MostAwesomeDude> Qball: Rock. :3
  • [19:15:51] <hli_> ali_as : you tell to linux to load the emulator binary at another place after teh 256 MiB
  • [19:16:05] <ZeZu> well i didn't read what mru posted, but with a linker script you can do what you want with the binaries
  • [19:16:25] <ali_as> hli, How much memory does the pandora have?
  • [19:16:29] <hli_> ali_as : but you also need to quickly mmap the first 256 MiB to avoid anything else to allocate them before you
  • [19:16:36] <hli_> 256 MiB
  • [19:17:07] <ali_as> Okaay. So how can you load anything after the first 256M?
  • [19:17:54] <hli_> Ali_as : the 256MiB i reserved is just virtual, it doesn't mean i reserve all physical pages in 256 MiB
  • [19:18:42] <hli_> that's only a virtual space i want to reserve where I would map 64 MiB (main ram), 4 Mib (VRAM) and 16KiB (scratchpad)
  • [19:18:58] <ali_as> I don't see if you attempt to allocate that memory how linux/MMU is going to know what you are allocating for address space reasons and what areas need to have physical memory unless you tell it directly.
  • [19:19:07] <hli_> in such a way there is like a 1:1 mapping with PSP memory
  • [19:19:25] <ZeZu> I wouldn't shoot for that right away anyhow
  • [19:19:32] <ZeZu> i'd get it running first and save that until later
  • [19:20:00] <ZeZu> its going to be as much trouble as its worth
  • [19:20:19] <hli_> Ali_as : my pspe4all works this way in Windows
  • [19:20:45] <hli_> except i cannot choose where to place
  • [19:20:57] <ali_as> Interesting project though, I am looking at making a MIPS emulator are there any open source dynamic recompilers?
  • [19:21:01] <hli_> what i need in Linux is something similar
  • [19:21:03] * mib_4wibkx (i=614d31af@gateway/web/ajax/mibbit.com/x-dd5a074bbc253ceb) Quit ("http://www.mibbit.com ajax IRC Client")
  • [19:21:19] <ali_as> (For x86 host).
  • [19:21:20] <ZeZu> ali_as: mine will be when its complete
  • [19:21:24] <ZeZu> oh
  • [19:21:31] <ZeZu> pcsx2 but i wouldn't recommend using it
  • [19:21:44] <ZeZu> and i dont know their current status on "open source"
  • [19:21:50] <hli_> but with having a 256MiB reserved virtual space starting at 0 (it seems Linux load the binary at this place)
  • [19:22:02] <CmosReloaded> anyone know the depth (dimensions) of the latest board?
  • [19:22:21] <ali_as> Linux used to load binaries at this place, but modern ones I've tried in x86 don't.
  • [19:22:37] <hli_> ZeZu : it is open source as fas as i know
  • [19:22:57] <ZeZu> i dont know if they've changed anything
  • [19:23:07] <hli_> ali_as : well not exactly at 0 because it is use to trap null pointer
  • [19:23:14] <ZeZu> but they are a pain in the $%5 group and trying to get svn access ........
  • [19:23:18] <hli_> MIPS also traps at 0 so it is okay
  • [19:23:32] <ZeZu> and i dont think you can depend on linux loading a binary at a certain address
  • [19:23:54] <ZeZu> brb
  • [19:23:55] <hli_> ZeZu : mru told me it is possible
  • [19:24:19] <ali_as> The program I patched about 3 months ago was being loaded above 0x80000000 in debian and fedora and ubuntu.
  • [19:24:26] * mib_4wibkx (i=614d31af@gateway/web/ajax/mibbit.com/x-63b866dc599b3a04) has joined #beagle
  • [19:24:32] <ZeZu> ah well, anything that mru says you should listen to
  • [19:24:42] <suihkulokki> ali_as: what about qemu?
  • [19:25:12] <hli_> ali_as : :/
  • [19:25:47] <hli_> well if i could put the 256 MiB at 0xC0000000 it should be okay
  • [19:25:55] <ali_as> It was being loaded neer 0x0 in RH6 and Fedora 4.
  • [19:26:23] <hli_> that way i replace AND %t1, %1, #0xFF000000 with ORR %t1, %1, #0xC0000000
  • [19:26:35] <ali_as> But they are old old. I have no idea what happens on a beagle or pandora.
  • [19:26:51] <geckosenator> is there a kernel parameter I need to get usb otg mode?
  • [19:27:13] <Qball> with angstrom, you need the right cable
  • [19:27:55] <hli_> ali_as: i dunno as well
  • [19:28:11] <hli_> i still need to prepare my beagleboard
  • [19:28:13] <geckosenator> I think I have the right cable
  • [19:28:23] <ds2> just cable and compiled in support for OTG on MUSB
  • [19:28:30] <ali_as> hli, is your beagle revision B?
  • [19:28:42] <hli_> ali_as: B4 helas :/
  • [19:28:59] <ali_as> Ok so 128M then and may behave differently than pandora.
  • [19:29:12] <hli_> i should have waited for a C revision :/
  • [19:29:15] <ali_as> Are you developing on beagle becuase of the pandora shortages?
  • [19:29:36] <geckosenator> ds2: I'm just using koen's latest prebuilt uImage and rootfs
  • [19:29:43] <hli_> ali_as : i ordered it before Pandora first batch i think
  • [19:29:52] <hli_> i ordered a Pandora
  • [19:29:53] <ds2> Oh no idea then... don't use those
  • [19:29:57] <ali_as> Aha.
  • [19:29:58] <hli_> but still not available
  • [19:30:34] <geckosenator> ds2: heh.. so I need to compile the kernel?
  • [19:30:34] <ali_as> I am waiting for C2 beagle.
  • [19:30:49] * brolin (n=brolin@190.157.13.75) has joined #beagle
  • [19:31:08] <ds2> geckosenator: maybe... donno if he has it enabled and if there are patches needed in the version of the kernel
  • [19:31:13] * brolin (n=brolin@190.157.13.75) Quit (Client Quit)
  • [19:31:20] <hli_> if only bb B4 can simply load a linux image to flash through USB...
  • [19:31:27] <ds2> some versions are buggy in that in won't switch host to device and back
  • [19:31:53] <geckosenator> ds2: ok, I see
  • [19:32:01] <ali_as> suihkulokki, I was trying to work out what you meant, I think thats an interpreter isnt it?
  • [19:32:09] <ds2> geckosenator:does he have /proc/config.gz enabled?
  • [19:32:55] <geckosenator> yes
  • [19:33:20] <geckosenator> what should I look for?
  • [19:33:25] <ds2> see if CONFIG_USB_OTG is enabled first
  • [19:33:51] <ds2> then confirm withh CONFIG_USB_MUSB_OTG
  • [19:34:03] <geckosenator> all are =y
  • [19:34:38] <geckosenator> CONFIG_USB_OTG=y CONFIG_USB_MUSB_OTG=y CONFIG_USB_OTG_UTILS=y
  • [19:34:44] <ds2> that's the basics
  • [19:34:49] <ds2> what is it doing?
  • [19:34:55] <geckosenator> nothing
  • [19:35:10] <ds2> what is your sequence from power up? did you change cables?
  • [19:35:20] <geckosenator> I guess I can confirm with a voltmeter that my usb minia cord is shorting the right pins
  • [19:35:38] <geckosenator> I didn't even change cables..
  • [19:35:59] <ds2> Hmmm host mode should mostly just work unless you are pulling too much power
  • [19:36:05] <geckosenator> it works as device mode though
  • [19:36:16] * harbaum (n=harbaum@i59F4B795.versanet.de) has joined #beagle
  • [19:36:17] <ds2> Hmmm let me get the other option
  • [19:36:21] <geckosenator> yeah let me see if my cable is shorting the pins correctly
  • [19:36:45] <ds2> CONFIG_USB_MUSB_HDRC
  • [19:36:48] <ds2> is that enabled?
  • [19:36:50] <suihkulokki> ali_as: no, it does dynamic translation
  • [19:37:00] <ali_as> !
  • [19:37:19] <ds2> along with CONFIG_USB_HDRC_HCD?
  • [19:37:30] <geckosenator> ds2: yes
  • [19:37:33] <ali_as> I will look at it, I need something I can hack to bits so it may be more than I can handle.
  • [19:37:38] <geckosenator> and also CONFIG_USB_GADGET_MUSB_HDRC=y
  • [19:38:04] <ds2> then it should work unless it was one of the buggy kernels
  • [19:38:18] <geckosenator> ok, let me test the cable
  • [19:38:40] <suihkulokki> ali_as: http://repo.or.cz/w/qemu.git?a=blob;f=tcg/README
  • [19:38:51] <ds2> there might be a sysfs entry somehwere to give you the status but I don't recall where it is; it used to be in /proc/driver
  • [19:39:45] <geckosenator> it looks like pins 4 and 5 weren't shorted
  • [19:39:57] <ds2> Oh that is a problem
  • [19:40:09] <ds2> ID needs to be grounded
  • [19:40:19] <geckosenator> I don't know why they make cables with mini-a and female A that don't ground ID
  • [19:40:25] <geckosenator> but apparently I got one
  • [19:40:32] <ds2> sure it is mini A?
  • [19:40:42] <geckosenator> I guess it's mini B
  • [19:40:51] <MostAwesomeDude> Qball: Thanks for the help, I achieved boot.
  • [19:40:52] <ds2> I got mini-B's with female A; apparently it is used by quite a few import gadgets
  • [19:41:03] <MostAwesomeDude> Although it refuses to actually load sysvinit.
  • [19:41:10] <ds2> also handy on the N800 when they only installed a MiniB socket; not an AB socket
  • [19:44:17] <koen> MostAwesomeDude: I wouldn't waste time with jffs2, try ubifs
  • [19:44:58] * koen knows ubifs works since he's actually using it on a beagle
  • [19:46:50] <MostAwesomeDude> koen: Actually, I went ahead and used the tar.bz2 instead.
  • [19:46:52] <MostAwesomeDude> ext2.
  • [19:47:21] <MostAwesomeDude> So init complains that it can't execute /etc/init.d/rc, but I just checked and it's set 755.
  • [19:48:48] * koen would recommend ext3 on SD, since you'd run out of power more often that you'd gain with saving writes to the journal
  • [19:49:01] * rsalveti (n=salveti@200.222.2.234) Quit (Remote closed the connection)
  • [19:49:20] * MostAwesomeDude will do tune2fs -j
  • [19:49:27] <jkridner> I finally got most of "James" to install, but the browser only shows "It works!".
  • [19:49:28] <ali_as> suihkulokki, I'm having a long read, thanks!
  • [19:49:56] <MostAwesomeDude> Anyway, is there any reason why init might fail that could be easily explained? I'm only using custom kernel right now.
  • [19:50:25] * rsalveti (n=salveti@200.222.2.234) has joined #beagle
  • [19:51:23] <koen> MostAwesomeDude: you unpacked as root?
  • [19:51:36] <koen> MostAwesomeDude: also, there are some sd cards that just hate ext2
  • [19:51:49] <koen> (and ext3)
  • [19:52:18] <MostAwesomeDude> koen: I unpacked as root; do I need to preserve users, or just permissions?
  • [19:52:36] <MostAwesomeDude> Also it is a Sandisk card, so not super-cheap but not super-awesome either.
  • [19:53:52] <MostAwesomeDude> Hm. Looks like mmc0 was found correctly, mounted ext2 FS...
  • [19:54:05] <jkridner> CmosReloaded: board dimensions haven't changed, but location of the HDMI connector has.
  • [19:54:11] <MostAwesomeDude> Yeah, not seeing any boot problems. I'll try unpacking again.
  • [19:56:50] * jkridner finds that 'james-install' doesn't think about what directory it is being run from properly.
  • [19:58:51] * jkridner finds that all of the "channels" in James are undefined.
  • [19:59:38] * jkridner is surprised to find content under the EPG, but doesn't know if is refreshable.
  • [20:00:42] <jkridner> ah, playback is disabled. :(
  • [20:03:26] * ZeZu should have quickly come up with another entry for that contest
  • [20:10:20] <MostAwesomeDude> Hm. INIT: Entering runlevel: 5/init.d/rcS"
  • [20:10:30] <MostAwesomeDude> That doesn't look right at all.
  • [20:15:32] * rsalveti (n=salveti@200.222.2.234) Quit (No route to host)
  • [20:21:27] <koen> jkridner: the plan is to send Jaya and Frans both a beagle for effort and have a longer time (and more marketing) for contest #3?
  • [20:22:57] <jkridner> I'll need to find how many boards Gerald can scrape up, but I'd add ZeZu to that as well. Voting can just help set the priority.
  • [20:23:14] <jkridner> so, yes is the answer.
  • [20:23:33] <koen> ok, I'm cool with that
  • [20:23:48] <jkridner> I felt like with contest #1, we had great results with very little time and we could accelerate on that.
  • [20:24:21] <jkridner> I believe, however, that we really need more time to have good entries--and that people are bottlenecked on having hardware.
  • [20:24:28] <koen> I can't really say why #2 was so different than #1
  • [20:24:51] <jkridner> I think the amount of time for Rev C was one difference.
  • [20:25:18] <jkridner> we couldn't really give out hardware at the beginning for "accepted projects" before, because we didn't want to give away Rev B's for that.
  • [20:25:59] <jkridner> now that Rev C's are being built, and will be available for everyone at the end of the month, I think the timing makes sense to switch to getting boards for "approved projects".
  • [20:26:17] <ZeZu> :)
  • [20:26:22] <jkridner> I think that should increase the project entries quite a bit... as for these two we were limited to only 4 total boards.
  • [20:26:42] <ZeZu> I didn't even know there was a contest until last weekend
  • [20:26:56] <jkridner> koen: are you just about ready to vote?
  • [20:26:58] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [20:27:29] <jkridner> maybe you can tell me a few more things I can try with "James".
  • [20:27:52] <jkridner> it is the only one I can personally try out, but is still not a clean setup and not a huge amount of out-of-the-box functionality.
  • [20:27:53] <koen> the James wiki pages lists the highlights
  • [20:28:11] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Client Quit)
  • [20:28:20] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [20:29:02] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [20:29:30] <koen> I think James could use some glue to tie things together, it's a bit of a 'mixed bag' (if I use the phrase correctly)
  • [20:29:53] <jkridner> is there a recommended USB CD drive?
  • [20:30:00] <koen> The e-ink think is nice, but I haven't figured out an affordable ay to test it
  • [20:30:27] <jkridner> It seems that people really want to build *hardware*.
  • [20:31:15] <jkridner> I was expecting some nice demos, ala: http://www.youtube.com/watch?v=evJ0Vr8LjYk
  • [20:32:02] * jkridner wonders how to get the attention of the 'demoscene'.
  • [20:33:04] <jkridner> so, we'll definitely be looking how to do more with hardware, since that seems to be a focus area.
  • [20:33:42] * jkridner will listen to where others want to take this, rather than to try to push my own vision of a software contest.
  • [20:33:49] <koen> I suspect 98% of the beagleboard users are hobbyist using it for fun
  • [20:34:07] <jkridner> sure, but what does that mean?
  • [20:34:25] <jkridner> I really like where most of the projects are headed.
  • [20:34:43] <jkridner> I like the idea of an open source PVR with web-based controls.
  • [20:34:48] <jkridner> I like gaming emulators.
  • [20:34:53] <jkridner> I like e-ink displays.
  • [20:34:57] <koen> it means that only the remaining 2% is going to enter in the contest
  • [20:35:10] <koen> and a large part of those are doing paid/credited work on it
  • [20:35:19] <koen> e.g. the pixhawk chopper
  • [20:35:44] <jkridner> who pays the pixhawk guys? do they just sell the finished chopper?
  • [20:35:57] * watarat (i=7cab4a53@gateway/web/ajax/mibbit.com/x-fc7ed35c0ce6c4c5) has joined #beagle
  • [20:35:59] <koen> eth.cz is a univeristy
  • [20:36:27] <koen> http://en.wikipedia.org/wiki/ETH_Zurich
  • [20:36:57] <jkridner> k. so it is a research project. any reason they could not enter our little contest and gain some more attention?
  • [20:37:46] <koen> I suspect they feel they aren't allow to enter
  • [20:38:02] <jkridner> obviously there is some attention paid to what goes on here, given that linuxdevices picked up the last contest and did a really nice write-up without any organizer contact.
  • [20:38:17] <koen> did LD really do a write up?
  • [20:38:35] * koen only know about LD publishing stuff that got sent in
  • [20:39:23] <jkridner> I'm just talking about http://www.linuxdevices.com/news/NS6362979448.html
  • [20:40:46] <koen> that still looks like a sent-in piece with an LD paragraph on top :)
  • [20:41:10] <koen> it is a real nice piece, I just have a hard time believing LD wrote it
  • [20:41:15] <jkridner> well, someone sent it in and they accepted it.
  • [20:41:23] <jkridner> and I didn't write it or ask anyone to write it.
  • [20:42:06] * florian (n=fuchs@f049013075.adsl.alicedsl.de) has joined #beagle
  • [20:42:08] <jkridner> they usually thank someone for sending something in.
  • [20:42:22] * rsalveti (n=salveti@189.64.114.99) has joined #beagle
  • [20:42:54] <jkridner> anyway, I think that the contest has clear opportunities for exposure, and that eth.cz would likely welcome that.
  • [20:43:04] <jkridner> what do you think we are doing to make them feel they cannot enter?
  • [20:43:36] <koen> most projects are of the one-guy-coding type
  • [20:43:56] <florian> hi all
  • [20:43:58] <koen> If I was part of a uni team I'd think it would be a bit unfair to enter
  • [20:44:01] <jkridner> hi florian.
  • [20:44:04] <koen> hey florian
  • [20:44:30] <koen> but then again, I would enter, since I'm a sucker for (free) gadgets
  • [20:44:46] <jkridner> it is my experience that one-guy-coding can often compete against a university team in terms of making something really interesting.
  • [20:44:50] <florian> koen: you are pretty good in this, right? ;)
  • [20:45:01] <koen> :)
  • [20:45:35] * koen turns his attention to the TV
  • [20:45:39] <jkridner> well, the prize is certainly more geared at individuals, so I think we need to revisit that.
  • [20:45:47] <florian> That reminds me that I still have to present somethgin useful with the pico
  • [20:49:59] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) Quit ("reboot")
  • [20:57:56] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) has joined #beagle
  • [21:03:10] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) Quit ()
  • [21:04:21] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit ("Ex-Chat")
  • [21:04:47] * bgamari (n=ben@c-24-61-223-47.hsd1.nh.comcast.net) Quit ("Terminated with extreme prejudice - dircproxy 1.0.5")
  • [21:07:31] * bgamari (n=ben@c-24-61-223-47.hsd1.nh.comcast.net) has joined #beagle
  • [21:18:26] <CmosReloaded> jkridner: thanks for the reply (about board dimensions). i'm currently looking at the release documentation and the "max height" (which i presume is the board + component height) is not listed. do you, or anyone know what the max height would be?
  • [21:19:07] * jkridner looks around for calipers.
  • [21:19:19] <jkridner> how accurate do you need this info?
  • [21:19:40] <jkridner> is that not in the System Reference Manual?
  • [21:20:28] <CmosReloaded> i need it to be accurate
  • [21:20:42] <CmosReloaded> the srm states that the max height is tbm (to be measured?)
  • [21:22:00] <jkridner> guess so. Best to ask on the mailing list and get Gerald's answer.
  • [21:23:14] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) has joined #beagle
  • [21:23:40] <CmosReloaded> ok, thanks for the advice
  • [21:23:56] * ZeZu_ (n=null@c-98-227-56-17.hsd1.in.comcast.net) has joined #beagle
  • [21:23:59] * ZeZu_ (n=null@c-98-227-56-17.hsd1.in.comcast.net) Quit (Remote closed the connection)
  • [21:24:01] <CmosReloaded> i definetly will be buying a board for my project, so i could always wait until then :P
  • [21:24:33] * mrc3 (n=ddiaz@189.157.115.100) Quit (Remote closed the connection)
  • [21:25:17] <jkridner> http://groups.google.com/group/beagleboard/browse_thread/thread/c83fbd8e81d80979
  • [21:25:39] <koen> there's a beagle in googles sketchup model warehouse
  • [21:26:13] <jkridner> guess that is a bit out of date.
  • [21:26:27] <jkridner> http://sketchup.google.com/3dwarehouse/cldetails?mid=decd9e6286323e92a3374bbb8941adff
  • [21:26:30] * mrc3 (n=ddiaz@189.157.115.100) has joined #beagle
  • [21:26:42] * harbaum (n=harbaum@i59F4B795.versanet.de) Quit (Read error: 113 (No route to host))
  • [21:39:47] * Xerion (i=xerion@82-170-197-160.ip.telfort.nl) Quit (" ")
  • [21:40:28] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) Quit (Read error: 104 (Connection reset by peer))
  • [21:42:36] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [21:52:35] * gletelli_ (n=chatzill@fw-tnat.cambridge.arm.com) Quit (Read error: 104 (Connection reset by peer))
  • [21:53:13] * Xerion (i=xerion@82-170-197-160.ip.telfort.nl) has joined #beagle
  • [21:53:40] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [21:58:00] <eFfeM-away> jkridner: i thought i made james-install such that it is dir independent (for opkg install it does not matter and on all other places there is a hard coded path or a cd command before it; the missing channels is probably an oversight in the install script; the content for EPG is preloaded and in /usr/share/apache2/cgi-bin/tvguide.xml; it is the output from xmltv (http://wiki.xmltv.org/) and as such in principle refreshable, but I could not
  • [21:58:10] <jkridner> koen: don't forget to vote!
  • [21:58:30] * gletelli_ (n=chatzill@fw-tnat.cambridge.arm.com) has joined #beagle
  • [21:58:35] <jkridner> mru, sakoman, ds2: ditto.
  • [21:58:46] * guillaum1 (n=gl@AMontsouris-153-1-22-157.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [21:59:00] <ds2> 'k
  • [21:59:01] <jkridner> I found "<br>" in "james-install" and had to remove those.
  • [21:59:12] <ds2> need to layout a PCB first
  • [21:59:15] * guillaum1 (n=Guillaum@AMontsouris-153-1-22-157.w86-212.abo.wanadoo.fr) has joined #beagle
  • [21:59:52] <eFfeM-away> jkridner: still reading the log, I would say any usb cd drive would work; recording does not; there is an issue with cdrkit on oe
  • [22:00:23] <jkridner> k. you need to be in / to extract pvr.tar.gz (or perhaps you can add -C /)
  • [22:00:33] * rsalveti (n=salveti@189.64.114.99) Quit (Read error: 104 (Connection reset by peer))
  • [22:00:34] <eFfeM-away> koen, i still need to add some glue and also a more friendly install script and configuration software, but this is a 1 person project which I can work on only part-time
  • [22:01:15] <jkridner> bye for now
  • [22:01:17] <eFfeM-away> jkridner: and if you have problems running things, ping me tomorrow or so (almost bedtime now) and I gladly give help
  • [22:02:03] <Crofton> https://wiki.har2009.org/page/Main_Page
  • [22:02:21] * eFfeM-away is now known as eFfeM
  • [22:03:39] <eFfeM> aargh, the <br> was there because I copied from the web page source, noticed that during the test run and fixed it and I thought I uploaded the fixed version, but apparently something went wrong and as I was in a time squeeze last friday I did not re-check.
  • [22:04:06] <eFfeM> bedtime here, catch you tomorrow
  • [22:04:14] * eFfeM is now known as eFfeM-zzz
  • [22:05:00] <Crofton> gn
  • [22:05:59] <eFfeM-zzz> ty, (firing up a fresh rebuild of oe, to see if that fixes some issues, then I'm gone; nice recompile job for the night :-) )
  • [22:06:44] * Xenion (n=robert@p579FCE76.dip.t-dialin.net) Quit ("Verlassend")
  • [22:08:33] <eFfeM-zzz> ty, (firing up a fresh rebuild of oe, to see if that fixes some issues, then I'm gone; nice recompile job for the night :-)
  • [22:09:17] * ceyusa (n=ceyusa@cm40051.red83-165.mundo-r.com) Quit (Read error: 110 (Connection timed out))
  • [22:24:01] * geckosenator (n=sean@71.237.94.78) Quit (Read error: 104 (Connection reset by peer))
  • [22:30:51] * mib9go (i=421b50c2@gateway/web/ajax/mibbit.com/x-5fc2839a5ec8bb8e) has joined #beagle
  • [22:32:57] * ScriptRipper (n=martin@opensuse/member/MartinMohring) Quit (Remote closed the connection)
  • [22:35:30] * mib9go (i=421b50c2@gateway/web/ajax/mibbit.com/x-5fc2839a5ec8bb8e) Quit (Client Quit)
  • [22:35:36] * ssvb (n=ssvb___@a88-114-221-132.elisa-laajakaista.fi) Quit ("Leaving")
  • [22:40:35] * hli_ (i=chaton@vig91-2-82-232-97-149.fbx.proxad.net) Quit ("ChatZilla 0.9.84 [Firefox 3.0.6/2009011913]")
  • [22:42:49] <vlad_> hrm, to get the sgx stuff going, do I want to use the kernel that's in the package? or roll my own 2.6.28?
  • [22:43:07] <vlad_> I guess there's a good chance that that 2.6.28 doesn't have the usb network adapter modules and stuff that I need
  • [22:49:13] * ant_home (n=andrea@host37-249-dynamic.4-87-r.retail.telecomitalia.it) has left #beagle
  • [22:59:36] * likewise (n=likewise@p5B39EE40.dip.t-dialin.net) Quit ()
  • [23:03:18] * alecrim (n=alecrim@201008196183.user.veloxzone.com.br) has joined #beagle
  • [23:11:30] * rsalveti (n=salveti@200.222.2.234) has joined #beagle
  • [23:19:18] * geckosenator (n=sean@71.237.94.78) has joined #beagle
  • [23:19:50] * mrc3 (n=ddiaz@189.157.115.100) Quit ("Leaving")
  • [23:24:11] * pH5_ (n=ph5@e178236226.adsl.alicedsl.de) Quit ("bye")
  • [23:25:30] * bmxr (n=bmxr@d66-183-44-76.bchsia.telus.net) has joined #beagle
  • [23:26:23] * ali_as is now known as as_leep
  • [23:29:40] * ScriptRipper (n=martin@host-82-135-36-197.customer.m-online.net) has joined #beagle
  • [23:30:41] * Robot_ (n=robot@rb5bx242.net.upc.cz) has left #beagle
  • [23:38:18] * CmosReloaded (n=netdevil@cpc4-glfd1-0-0-cust691.glfd.cable.ntl.com) Quit ()
  • [23:41:05] * TAK2004 (n=thomas@dslb-088-072-106-205.pools.arcor-ip.net) Quit (Remote closed the connection)
  • [23:49:58] * alecrim (n=alecrim@201008196183.user.veloxzone.com.br) Quit (Remote closed the connection)
  • [23:54:32] * flameman (n=toor@net-93-65-1-167.cust.dsl.vodafone.it) Quit (Read error: 60 (Operation timed out))
  • [23:55:14] <vlad_> damn, this was -almost- a completely painless setup
  • [23:55:30] <vlad_> except the sgx kernel's config has the driver for my usb ethernet thing as a module, and the modules don't come in the package =/
  • [23:57:37] * guillaum1 (n=Guillaum@AMontsouris-153-1-22-157.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [23:57:59] * guillaum1 (n=gl@AMontsouris-153-1-22-157.w86-212.abo.wanadoo.fr) has joined #beagle