[00:01:59] <mike_morrison> when i boot up angstrom, it never shows a login prompt on the monitor. it just seems to be looping through a screen/color test. any idea why this might be?
[00:02:46] <Crofton|irssi> that's one I haven't heard of ....
[00:03:15] * flo_lap (n=fuchs@f049001134.adsl.alicedsl.de) Quit ("Verlassend")
[00:03:47] <jkridner> thanks ds2
[00:04:15] <Crofton|irssi> ds2: the pinmux page looks great
[00:08:38] <mike_morrison> it cycles through the following 5 screens: blue, black, white, grey, color bar gradients (red, green blue, grayscale)
[00:09:22] <mike_morrison> also this starts before booting angstrom
[00:09:43] * jrmuizel (n=jrmuizel@wycliffe03.wycliffe.utoronto.ca) has joined #beagle
[00:11:15] * robclark (n=robclark@nat/ti/x-394b891415c74c12) Quit ()
[00:15:20] * gduncan (n=spec@pool-98-113-113-148.nycmny.east.verizon.net) has joined #beagle
[00:26:17] * mike_morrison (n=mmorriso@24.68.237.228) Quit (Remote closed the connection)
[00:31:52] * mike_morrison (n=mmorriso@24.68.237.228) has joined #beagle
[00:31:54] * jrmuizel (n=jrmuizel@wycliffe03.wycliffe.utoronto.ca) Quit (Read error: 110 (Connection timed out))
[00:37:30] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
[00:52:00] * _apk (n=kows@85-18-136-98.fastres.net) has joined #beagle
[00:53:14] * harry-plenware (n=harry@221.10.25.6) has joined #beagle
[00:59:39] * likewise_ (n=chatzill@82-171-51-231.ip.telfort.nl) Quit ("ChatZilla 0.9.84 [Firefox 3.0/2008061015]")
[01:26:37] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) Quit (Read error: 110 (Connection timed out))
[01:32:40] * Ikarus (n=ikarus@dhcp-077-249-080-210.chello.nl) Quit (Connection timed out)
[01:33:13] * rsalveti (n=salveti@200.184.118.132) Quit (Remote closed the connection)
[01:42:35] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) has joined #beagle
[01:51:11] * mike_morrison (n=mmorriso@24.68.237.228) Quit ("Ex-Chat")
[01:54:28] * khasim (n=a0393720@192.163.20.231) has joined #beagle
[02:06:19] * SpaceAceX (i=Rakso@CPE001310d5071d-CM0011ae01368a.cpe.net.cable.rogers.com) has joined #beagle
[02:11:53] * Eanmig (n=zaphod@dsl-67-204-16-239.acanac.net) Quit (Read error: 104 (Connection reset by peer))
[02:12:18] * Eanmig (n=zaphod@dsl-67-204-16-239.acanac.net) has joined #beagle
[02:14:16] * rsalveti (n=salveti@189.70.53.187) has joined #beagle
[02:25:04] * Ikarus (n=ikarus@i227209.upc-i.chello.nl) has joined #beagle
[02:28:05] * Crofton (n=balister@12.234.232.66) Quit ("Leaving")
[02:47:12] * cbrake is now known as cbrake_away
[02:55:20] * robtow (n=rtow@nat/montavista/x-470efab8d4d066e2) Quit ("Leaving.")
[03:03:36] * Crofton (n=balister@12.234.232.66) has joined #beagle
[03:10:45] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) has joined #beagle
[03:17:02] * tulpe (n=ask@p54AF7C75.dip.t-dialin.net) Quit (Read error: 60 (Operation timed out))
[03:17:14] * tulpe (n=ask@p54AF7A50.dip.t-dialin.net) has joined #beagle
[03:21:07] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
[03:27:00] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) Quit (Client Quit)
[03:28:31] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) Quit ("open.neurostechnology.com")
[03:42:13] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) has joined #Beagle
[03:48:27] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (kubrick.freenode.net irc.freenode.net)
[03:48:27] * Stskeeps (n=chobits@84.238.11.16) Quit (kubrick.freenode.net irc.freenode.net)
[03:48:27] * bazbell (n=a0192809@nat/ti/x-91885a886d983aba) Quit (kubrick.freenode.net irc.freenode.net)
[03:48:27] * shenki (n=joel@202.174.42.5) Quit (kubrick.freenode.net irc.freenode.net)
[03:48:27] * ldesnogu (n=ldesnogu@fw-tnat.cambridge.arm.com) Quit (kubrick.freenode.net irc.freenode.net)
[03:48:27] * oyvindje (i=oyvindje@158.38.48.10) Quit (kubrick.freenode.net irc.freenode.net)
[03:48:27] * keesj (n=keesj@ip49-193-210-87.adsl2.static.versatel.nl) Quit (kubrick.freenode.net irc.freenode.net)
[03:48:27] * tobb (i=phikz@user152.77-105-212.netatonce.net) Quit (kubrick.freenode.net irc.freenode.net)
[03:48:54] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
[03:48:54] * Stskeeps (n=chobits@84.238.11.16) has joined #beagle
[03:48:54] * bazbell (n=a0192809@nat/ti/x-91885a886d983aba) has joined #beagle
[03:48:54] * shenki (n=joel@202.174.42.5) has joined #beagle
[03:48:54] * ldesnogu (n=ldesnogu@fw-tnat.cambridge.arm.com) has joined #beagle
[03:48:54] * oyvindje (i=oyvindje@158.38.48.10) has joined #beagle
[03:48:54] * tobb (i=phikz@user152.77-105-212.netatonce.net) has joined #beagle
[03:48:54] * keesj (n=keesj@ip49-193-210-87.adsl2.static.versatel.nl) has joined #beagle
[03:54:46] * [X]Spot (n=stancho@78.90.99.168) has joined #beagle
[03:55:43] * Stasi (n=stancho@78.90.99.168) Quit (Read error: 104 (Connection reset by peer))
[03:56:16] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) has joined #beagle
[04:00:55] * emeb (n=ericb@ip72-223-95-46.ph.ph.cox.net) has left #beagle
[04:06:18] * emeb_mac (n=ericb@ip72-223-95-46.ph.ph.cox.net) has joined #beagle
[04:06:24] * Stasi (n=stancho@78.90.99.168) has joined #beagle
[04:07:08] * Beagle8 (n=Beagle8@208-100-140-231.bendbroadband.com) has joined #beagle
[04:07:13] * Beagle8 is now known as _altered
[04:08:09] * [X]Spot (n=stancho@78.90.99.168) Quit (Read error: 113 (No route to host))
[04:14:36] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) Quit ("Trillian (http://www.ceruleanstudios.com")
[04:33:34] * [X]Spot (n=stancho@78.90.99.168) has joined #beagle
[04:34:05] * Stasi (n=stancho@78.90.99.168) Quit (Read error: 113 (No route to host))
[04:37:39] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
[04:44:33] * rsalveti (n=salveti@189.70.53.187) Quit (Read error: 113 (No route to host))
[04:52:53] * chelli_ (n=chelli@p54B865CE.dip.t-dialin.net) has joined #beagle
[05:07:45] * khasim (n=a0393720@192.163.20.231) has joined #beagle
[05:09:51] * chelli (n=chelli@debian/developer/tschmidt) Quit (Read error: 110 (Connection timed out))
[05:17:21] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) Quit (Read error: 104 (Connection reset by peer))
[05:29:20] * gduncan (n=spec@pool-98-113-113-148.nycmny.east.verizon.net) Quit ()
[05:30:54] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
[05:34:05] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
[05:38:12] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
[05:40:18] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
[05:41:14] * aanishn (n=anish@59.160.172.220) has joined #beagle
[05:49:49] * khasim (n=a0393720@192.163.20.231) has joined #beagle
[05:59:17] * Stasi (n=stancho@78.90.99.168) has joined #beagle
[06:01:59] * emeb_mac (n=ericb@ip72-223-95-46.ph.ph.cox.net) Quit ()
[06:13:08] * [X]Spot (n=stancho@78.90.99.168) Quit (Read error: 110 (Connection timed out))
[06:26:13] * Beagle8 (n=Beagle8@s-inf-pc117.oulu.fi) has joined #beagle
[06:34:20] * jsync (n=jess@59.160.172.220) Quit (Remote closed the connection)
[06:39:40] * Eanmig (n=zaphod@dsl-67-204-16-239.acanac.net) Quit (Remote closed the connection)
[06:44:50] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) Quit ("Leaving")
[07:09:39] * jsync (n=jess@59.160.172.220) has joined #beagle
[07:20:12] * ds2 bangs head on table
[07:29:58] * wbrown_ (n=wbrown@c-67-175-7-145.hsd1.il.comcast.net) has joined #beagle
[07:29:58] * wbrown (n=wbrown@c-67-175-7-145.hsd1.il.comcast.net) Quit (Read error: 104 (Connection reset by peer))
[07:47:08] <_altered> is anyone else using the MV Linux demo image from ESC?
[08:13:13] <ds2> _altered: what do you need? =)
[08:13:49] * dignus (n=dignus@bastion.jkit.nl) has joined #beagle
[08:17:13] <dignus> morning all
[08:40:28] * chelli_ is now known as chelli
[08:44:34] * adj__ (i=antti@hervanta.com) has joined #beagle
[08:45:41] * valhalla (n=valhalla@81-174-21-31.dynamic.ngi.it) has joined #beagle
[08:52:37] * adj_ (i=antti@hervanta.com) Quit (Read error: 110 (Connection timed out))
[08:52:49] * koen is getting a bit lost in all the evil binary packages
[08:55:00] * methril|gone (n=Methril@213.27.233.98) Quit (Read error: 113 (No route to host))
[09:07:14] <geckosenator> there are binary packages?
[09:07:26] <geckosenator> it's not like x86
[09:09:44] <koen> I'm trying to get binaries to run that I don't have the source for
[09:10:18] <koen> I downloaded the sgx SDK from the ti website
[09:10:48] <kulve> is that publicly downloadable? For what hw?
[09:11:17] <koen> "send a mail and pray" publicly downloadable
[09:11:32] <koen> the kernel drivers for the evm work with the beagle .22 kernel
[09:11:41] <koen> well "work"
[09:11:54] <kulve> I asked that for the beagleboard, but I only got a "wait awhile" -email
[09:11:56] <koen> it seems that .22 relies on uboot to initialize the DSS properly
[09:13:14] <koen> and I still have the demo fs from LRL
[09:13:53] <kulve> open source kernel driver for that would help a lot even if the opengl es stack would be closed binary
[09:13:58] <koen> just to be annoying TI/imgtek has packaged the linux drivers inside a windows installer
[09:14:42] <kulve> sure, that one would have to accept a license, or something..
[09:14:54] <kulve> I'm sure that would be hard to do in linux
[09:23:33] <koen> every other TI packages is a selfextracting shellscript
[09:23:38] <koen> with the license
[09:26:59] * dignus (n=dignus@bastion.jkit.nl) has left #beagle
[09:37:35] * vu3rdd (n=user@117.196.133.162) has joined #beagle
[09:38:24] * abitos (n=nixgibts@dslb-084-057-207-140.pools.arcor-ip.net) has joined #beagle
[09:39:21] * kozak|work (i=cbc8c84a@gateway/web/ajax/mibbit.com/x-e9ac84445d36110f) has joined #beagle
[09:49:15] * kaa (n=kaa@62.70.2.252) Quit (Remote closed the connection)
[10:11:19] * pundiramit (n=amit@122.166.2.243) has joined #beagle
[10:11:59] * guillaum1 (n=gl@AMontsouris-153-1-67-44.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
[10:12:24] <pundiramit> hi
[10:12:40] <pundiramit> is is a right channel to ask omap3-evm related queries
[10:22:00] <kulve> partly (but I guess there's no better place either)
[10:26:17] <pundiramit> thanks... did anyone try porting X11+Gnome-desktop on beagle or omap3-evm?
[10:28:27] <kulve> everybody
[10:28:37] <kulve> ?ngstrom is probably the easiest way to test it
[10:29:00] * harry-plenware (n=harry@221.10.25.6) Quit (Read error: 110 (Connection timed out))
[10:34:04] * florian (n=fuchs@217.146.132.69) has joined #beagle
[10:40:20] * vu3rdd` (n=user@117.196.128.188) has joined #beagle
[10:40:31] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
[10:42:38] * vu3rdd (n=user@117.196.133.162) Quit (Read error: 60 (Operation timed out))
[10:49:49] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Read error: 110 (Connection timed out))
[10:56:19] <koen> wtf?
[10:56:23] <koen> it's snowing outside
[11:04:06] <geckosenator> where are you?
[11:04:13] <geckosenator> it's snowing here too
[11:04:17] <geckosenator> or, about to
[11:09:06] <koen> .nl
[11:09:57] <koen> sadly it's 5 degrees outside so the snow melts when it hits the ground
[11:10:28] <AV500> you're not in the "dutch mountains" :-)
[11:12:09] <geckosenator> a month ago I got caught in a foot of snow
[11:12:28] <geckosenator> then I moved down the the flat areas.. I was camping in the rockies
[11:12:46] <geckosenator> well, I was living in a tent up there
[11:21:04] <_altered> ds2: sorry, was afk for a while.. I was wondering if anyone else was having trouble with the beagleboard as a USB Gadget with the MV Linux image from ESC.
[11:28:03] * koen reboots again
[11:28:21] <koen> this TI 2.6.22 kernel is being extremely unstable
[11:28:32] <koen> I suspect it relies on a specific uboot
[11:29:37] * kozak|work (i=cbc8c84a@gateway/web/ajax/mibbit.com/x-e9ac84445d36110f) Quit ("http://www.mibbit.com ajax IRC Client")
[11:31:37] * methril|gone (n=Methril@213.27.233.98) has joined #beagle
[11:37:05] * pundiramit (n=amit@122.166.2.243) has left #beagle
[11:38:44] * methril|gone (n=Methril@213.27.233.98) Quit (Read error: 104 (Connection reset by peer))
[11:51:12] * likewise_ (n=chatzill@82-171-51-231.ip.telfort.nl) has joined #beagle
[11:56:38] * methril|gone (n=Methril@213.27.233.98) has joined #beagle
[12:01:52] <koen> gah
[12:01:58] * koen stabs TI uboot 1.1.4
[12:04:22] <koen> does anyone have a TI uboot 1.3.3 binary?
[12:04:37] <koen> http://code.google.com/p/beagleboard/downloads/list only have source or uboots that flash uboots
[12:05:32] * methril|gone (n=Methril@213.27.233.98) Quit (Read error: 104 (Connection reset by peer))
[12:05:54] * methril|gone (n=Methril@213.27.233.98) has joined #beagle
[12:18:54] * SpaceAceX (i=Rakso@CPE001310d5071d-CM0011ae01368a.cpe.net.cable.rogers.com) Quit (Read error: 60 (Operation timed out))
[12:21:49] * SpaceAceX (i=Rakso@CPE001310d5071d-CM0011ae01368a.cpe.net.cable.rogers.com) has joined #beagle
[12:22:17] * SpaceAceX (i=Rakso@CPE001310d5071d-CM0011ae01368a.cpe.net.cable.rogers.com) Quit (Client Quit)
[12:30:26] <koen> likewise_: ping
[12:31:53] <likewise_> koen: pong
[12:32:05] <likewise_> koen: hi
[12:35:19] * geckosenator (n=sean@c-24-8-193-190.hsd1.co.comcast.net) Quit (Read error: 110 (Connection timed out))
[12:46:11] * Beagle7 (n=Beagle7@213.139.154.7) has joined #beagle
[12:46:17] <koen> likewise_: I already have Xora's SOB for that revert, but thanks for responding :)
[12:46:27] <Beagle7> hi there
[12:47:00] <Beagle7> anybody here who can tell me why my angstrom demo distro is starting in serial but my LCD stays blank ?
[12:47:09] <likewise_> koen: revert?
[12:47:40] <koen> li [Blikew http://gitweb.openembedded.net/?p=openembedded.git;a=commit;h=13d35d55272113b1cac95c826c08c5ce99f7f440
[12:48:03] <koen> Beagle7: probably because your monitor doesn't support that sync range
[12:48:35] <Beagle7> but my monitor tells me in red when sync range is to high ...
[12:48:38] <Beagle7> but there is nothing
[12:48:41] <Beagle7> only blank
[12:49:03] <koen> Beagle7: could you try booting http://dominion.thruhere.net/koen/OE/uImage-2.6.27+2.6.28-rc3+r2+gitrf7429fd378a29cf6947c2613e0fd6e6e36165167-r2-beagleboard.bin ?
[12:49:06] <Beagle7> the led says it is aktivated ... the monitor i mean but only blank ... no info for to high frequ
[12:50:08] <Beagle7> only this image as uImage on vat32 part ?
[12:50:18] * john280z (n=jmitchel@static-71-111-254-119.rlghnc.dsl-w.verizon.net) has joined #beagle
[12:50:30] <koen> Beagle7: yes
[12:50:40] <Beagle7> what bootparams i have to take ?
[12:50:40] <koen> Beagle7: that should give you 1024x768@60Hz
[12:50:43] <Beagle7> can you tell me plz
[12:50:51] <Beagle7> little bit new at beagleboarding
[12:50:51] <Beagle7> ;-)
[12:51:08] <koen> Beagle7: see http://www.angstrom-distribution.org/demo/beagleboard/README.txt for bootargs
[12:51:21] <jkridner> good morning all.
[12:51:27] <koen> hey jkridner
[12:52:11] <Beagle7> so same i use now
[12:52:12] <Beagle7> ;-)
[12:52:15] <Beagle7> thx
[12:52:17] <likewise_> morning jkridner
[12:52:21] <Beagle7> tell you feedback in some sec
[12:53:07] <likewise_> koen: ok now I got it :-)
[12:53:16] <likewise_> koen: +1
[12:53:27] <jkridner> perhaps beagleboard.org traffic is peaked. I thought that the Ars Technica article on Android would drive a lot of new traffic.
[12:53:40] <Beagle7> so i use the same mmc ? so only replace the angstrom uimage with your ?
[12:53:44] <koen> maybe the android hype has died down
[12:53:54] <koen> Beagle7: yes
[12:53:57] <Beagle7> ;-)
[12:53:58] <Beagle7> thx
[12:54:22] <koen> if that doesn't give output there's something else wrong
[12:54:30] <jkridner> Maybe.
[12:54:46] <jkridner> I did see a roughly 15% jump.
[12:55:04] <koen> so 7 hits yesterday?
[12:55:06] * koen hides
[12:56:48] <jkridner> maybe all the people that read Ars Technica have already been to BeagleBoard.org.
[12:57:02] <jkridner> There are over 1,000 unique visitors a day.
[12:58:43] <jkridner> I'm surprised that Alexa doesn't seem to pick up on that.
[12:59:21] <jkridner> any good 3rd party traffic estimation sites to verify claims?
[13:00:45] * robclark (n=robclark@nat/ti/x-78ec326194a13225) has joined #beagle
[13:00:59] <koen> not that I know of
[13:01:08] <Beagle7> koen with your image i get this output and monitor stay in stanby mode ... system in serial can not start too with these error: http://rafb.net/p/N6IYhE58.html
[13:01:10] <koen> I had google statistic on a few sites
[13:01:32] <koen> ah
[13:01:39] <koen> I gave you the wrong kernel, sorry
[13:01:44] <Beagle7> ;-)
[13:01:45] <Beagle7> ok
[13:02:28] * codepope (n=codepope@ip-217.146.111.161.merula.net) has joined #beagle
[13:02:33] <koen> http://dominion.thruhere.net/koen/OE/uImage-2.6.27+2.6.28-rc4+r0+gitr5ecf98b76fa95078277c9037bb01640fd3de5e2c-r0-beagleboard.bin
[13:02:40] <koen> that shouldn't have the div by zero
[13:04:00] <Beagle7> ok
[13:04:02] <Beagle7> i try
[13:04:03] <Beagle7> thx
[13:05:56] * aanishn (n=anish@59.160.172.220) has left #beagle
[13:08:49] <Beagle7> koen same problem
[13:08:54] <koen> hmmm
[13:08:56] <Beagle7> system start in serial until login
[13:09:06] <Beagle7> so everything fine i think
[13:09:21] <Beagle7> monitor comes up from standy in aktive mode
[13:09:26] <Beagle7> black screen
[13:09:38] <Beagle7> but no to high frequ error from lcd
[13:09:53] <koen> could you type 'fbset' in the serial console?
[13:10:11] <Beagle7> yes
[13:10:39] <Beagle7> http://rafb.net/p/8hhBCw63.html
[13:10:40] <koen> the V: should report something to 60Hz
[13:10:57] <Beagle7> 59.870 Hz
[13:11:16] <koen> yeah, so it does a proper 1024x768@60Hz
[13:11:33] <Beagle7> seems so
[13:11:46] <Beagle7> should i try another monitor ?
[13:11:58] <koen> if you have access to one
[13:12:07] <Beagle7> sec
[13:13:07] * jsync (n=jess@59.160.172.220) has left #beagle
[13:13:21] <koen> my monitor has a wide sync range, so I tend not to get any problems with that :)
[13:14:20] * koen implements a uboot cascade script
[13:15:35] <Beagle7> ok other monitor same problem
[13:16:07] <Beagle7> can i set an fb mode qhat should work 100%
[13:16:15] <Beagle7> ?!
[13:17:43] <koen> no idea on that, sorry
[13:17:50] <koen> maybe someone on the ml has a suggestion
[13:18:31] <Beagle7> should a newser xloader help ?
[13:18:40] <Beagle7> or something like that
[13:18:41] * rsalveti (n=salveti@201.82.128.154) has joined #beagle
[13:19:04] <koen> that would have been my suggestion is fbset hadn't returned the close-to-60Hz value
[13:19:48] * koen notices how slow the SD driver is in .22 compared to >=.26
[13:20:19] <Beagle7> so the output from fbset i gave you is fine right ?
[13:20:23] <koen> yes
[13:20:26] <Beagle7> ok
[13:20:49] <jkridner> koen: what is a cascade script? is that like what sakoman has done?
[13:21:06] <jkridner> with 'if' statements/
[13:21:07] <jkridner> ?
[13:21:22] <koen> no
[13:21:36] <koen> 'run sd-uboot' will load and run a uboot from SD
[13:21:39] * vu3rdd` (n=user@117.196.128.188) Quit ("ERC Version 5.3 (IRC client for Emacs)")
[13:22:30] <Beagle7> how can i upgrade my 1.41 xloader ?!
[13:22:32] <koen> since I suspect uboot is causing the DSS problems with the .22 kernel
[13:22:35] * ClaudeQC (n=claude@bas1-quebec03-1279636213.dsl.bell.ca) has joined #beagle
[13:22:49] <koen> Beagle7: http://www.angstrom-distribution.org/demo/beagleboard/README.txt
[13:23:27] <Beagle7> MLO is the xloader right ?
[13:23:32] <koen> yes
[13:23:50] <koen> technically MLO is a signed xloader, but they are both the same in that dir
[13:23:50] <Beagle7> u boot i upgraded allready
[13:25:55] <Beagle7> so when i only want to upgrade MLO i only need these lines right ? : http://rafb.net/p/Xdo5BM31.html
[13:27:27] <koen> I would leave out line 7
[13:27:33] <koen> just in case you mis-pasted the commands
[13:27:36] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
[13:28:38] <jkridner> koen: we should actually be creating redundant copies of x-loader in the RAM in case there is an error. I don't think that is done anywhere yet.
[13:28:47] <jkridner> er, flash. RAM=flash.
[13:28:51] * abitos (n=nixgibts@dslb-084-057-207-140.pools.arcor-ip.net) Quit (Nick collision from services.)
[13:29:07] * abitos (n=nixgibts@dslb-084-057-206-230.pools.arcor-ip.net) has joined #beagle
[13:29:49] <jkridner> I don't recall the addresses checked, but the ROM bootloader will attempt to read an image from I think 4 different addresses.
[13:30:06] <koen> how big is x-load?
[13:30:10] <koen> more than 16k?
[13:30:13] * _altered (n=Beagle8@208-100-140-231.bendbroadband.com) Quit (Read error: 54 (Connection reset by peer))
[13:30:28] <koen> or is the PoP NAND safe for the first 32kB?
[13:32:33] * koen compiles cairo on the beagle
[13:34:09] * Openfree (n=df@222.65.143.118) has joined #beagle
[13:41:23] <Beagle7> koen so after i have done all lines without line 7 i have to poweroff the beagle and power it on again ?
[13:41:30] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
[13:41:35] <Beagle7> and everything works fine after that like before ?
[13:41:54] <koen> it everything went OK, the 'reset' command is also ok
[13:42:25] <Beagle7> so i inly need the MLO file on the mmc right ?
[13:45:32] * guillaum1 (n=GZBA4143@AMontsouris-153-1-67-44.w90-2.abo.wanadoo.fr) has joined #beagle
[13:53:47] * montamer (n=vijay@203.199.213.3) has joined #beagle
[13:54:41] * montamer (n=vijay@203.199.213.3) Quit (Client Quit)
[13:55:16] * montamer (n=vijay@203.199.213.3) has joined #beagle
[13:58:35] * emeb (n=ericb@ip72-223-95-46.ph.ph.cox.net) has joined #beagle
[13:59:50] <koen> hmmm
[14:00:00] <koen> windows ce can play an mp3 and show a jpg
[14:00:12] <likewise_> koen: is that good?
[14:00:14] * koen unimpressed
[14:00:38] <koen> likewise_: I just watched the "windows ce on beagle" video
[14:03:33] <likewise_> koen: my from-scratch build gives this, where it seems a tar ball is tried to be made before the fetch is finished? http://www.pastebin.ca/1263483
[14:03:58] * montamer (n=vijay@203.199.213.3) Quit ("leaving")
[14:04:01] * likewise_ wonders if he is using the wrong bitbake again
[14:04:16] <koen> likewise_: it should try to get a tarball from the server :(
[14:04:30] <koen> that rev vanished from mrus git server
[14:05:05] <likewise_> so much for revision *history* :-)
[14:05:18] <koen> git_git.mansr.com.ffmpeg.mru_afb98868e19e63cbca6c9f0ed9e6cfa48d40277d.tar.gz
[14:05:56] <koen> likewise_: git is a glorified patch tracker, anyone that cares about history or integrity wouldn't use git
[14:06:25] <koen> and git stores inline conflict markers in files...
[14:06:37] * koen hates it when stuff breaks because git can't merge stuff
[14:06:41] * Beagle8 (n=Beagle8@s-inf-pc117.oulu.fi) Quit ()
[14:07:27] * likewise_ throws away the lock file (which IS there), cleans, removes git/ and retries
[14:08:04] <koen> likewise_: get the above tarball and place it in your downloads folder (with the .md5)
[14:09:36] <likewise_> tnx. can it be I have a server overridden somewhere
[14:09:58] * JuanG (n=Juan@nat/ti/x-0aab8409fbea0189) has joined #beagle
[14:10:04] * JuanG (n=Juan@nat/ti/x-0aab8409fbea0189) has left #beagle
[14:10:46] <likewise_> I have a modified SOURCE_MIRROR_URL...
[14:11:16] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Connection timed out)
[14:11:40] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
[14:12:14] <Beagle7> koen that my 2nd beagleboard now which doen not work after writing this MLO
[14:12:36] <Beagle7> the beagleboard onlystarts by pressing user button at power on
[14:12:52] <Beagle7> and the keyboard in serial does not work an longer
[14:13:26] <koen> hmmm
[14:13:50] <Beagle7> so i can throw these two board away or not ?
[14:14:01] <Beagle7> dont see any possiblity to use it anymore
[14:14:23] <Beagle7> something like factory reset isnt implmented or ?
[14:14:27] <DJWillis> Beagle7: sounds like your flash is bad rather then a fault at a guess/
[14:14:54] <Beagle7> so what could i make wrong ?
[14:14:59] <Beagle7> wrote the MLO on mmc
[14:15:15] <Beagle7> ande have done the 6 lines like angstrom describes
[14:15:23] <Beagle7> without any errors
[14:16:13] <Beagle7> so anybody an idea how i can come back to the origin ?
[14:16:55] <DJWillis> Beagle7: well if you can boot with User then your in a good state anyway to boot the board and just try again.
[14:17:12] <DJWillis> Your MMC is a bootable one I assume?
[14:17:31] <Beagle7> yes i can boot with user botton ... but my keyboard doe not work after that flushing
[14:22:24] <sakoman_> jkridner: the factory scripts I did for gumstix put 4 copies of x-load in nand
[14:22:57] <jkridner> smart man. :)
[14:25:00] * rsalveti (n=salveti@201.82.128.154) Quit (Read error: 110 (Connection timed out))
[14:26:28] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
[14:27:40] * prpplague (n=dave@mail.americanmicrosystems.com) has joined #beagle
[14:39:21] <jkridner> hurray, https://www-a.ti.com/downloads/sds_support/targetcontent/LinuxDspTools/index.html is live again.
[14:39:35] <prpplague> jkridner: morning
[14:39:39] * prpplague looks at the url
[14:41:51] <jkridner> it is a bit prettier and more comprehensive than previous.
[14:45:11] <jkridner> double hurray, codec engine is released: https://www-a.ti.com/downloads/sds_support/targetcontent/CE/ce_2_21/index.html
[14:45:52] <tomasg> is this the release that supports DSPLINK 1.60 so we can turn on PREEMPT?
[14:46:06] <jkridner> it is the one that supports Link 1.60.
[14:46:34] <jkridner> and is BSD-licensed for most ARM user-space code.
[14:50:54] * JimDonova1 (n=jimd@c-66-30-221-249.hsd1.ma.comcast.net) has joined #beagle
[14:52:52] * JimDonovan (n=jimd@c-66-30-221-249.hsd1.ma.comcast.net) Quit (Read error: 110 (Connection timed out))
[14:56:13] * gcohler (n=gcohler@nat/ti/x-ebc63821fb7fb195) has joined #beagle
[14:56:59] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
[15:02:02] * john280z (n=jmitchel@static-71-111-254-119.rlghnc.dsl-w.verizon.net) has left #beagle
[15:02:18] * NishanthMenon (n=nMenon@nat/ti/x-aa361e1249789e07) has joined #beagle
[15:02:20] * NishanthMenon (n=nMenon@nat/ti/x-aa361e1249789e07) Quit (Client Quit)
[15:02:34] <Crofton> We need a link for dsp/bios :)
[15:04:28] * kevinsc (n=kevinsc@nat/ti/x-0c847ce7fc2178c8) Quit (".")
[15:04:36] * kevinsc (n=kevinsc@nat/ti/x-f48bbaf8f278c211) has joined #beagle
[15:04:43] <DJWillis> jkridner: good news. Licence changes are very welcome.
[15:05:26] <DJWillis> sakoman_: don't suppose you would share those scripts (out of pure interest)?
[15:13:28] <AV500> I am booting Angstrom for the 1st time on my BB and I get an empty screen that asks me to select a language?!?
[15:17:47] * cbrake_away is now known as cbrake
[15:21:44] <sakoman_> DJWillis: I'd have to get approval from Gumstix to publish the complete scripts
[15:34:47] * BThompson (n=BThompso@nat/ti/x-e65c0d64881dba57) has joined #beagle
[15:37:38] * gduncan (n=spec@rrcs-208-105-88-237.nyc.biz.rr.com) has joined #beagle
[15:38:21] <DJWillis> sakoman_: ahh, ok, no worries.
[15:40:49] <sakoman_> koen: you were right -- switching to metacity is going to require a chunk of work
[15:42:09] <koen> sakoman_: my beagle is busy running a X benchmark to see whether Xsgx has accell for Xrender and xlib stuff, so I can like into that after that finishes
[15:42:20] <Beagle7> anybody knows the probleme after flushin MLO that the board only starts by pressing user button and keyboard does not work in serial console ?
[15:42:39] <Beagle7> is there any possibility to make something like factory reset
[15:42:50] <koen> sakoman_: basically we want to launch gdm and use its scripts instead of xserver-(kdrive-)common
[15:42:59] * jkridner|work (n=a0321898@nat/ti/x-034d62fdd6e2cda2) has joined #beagle
[15:43:17] <koen> sakoman_: gdm should take care of lauching gnome-session, which will launch metacity
[15:43:49] <sakoman_> koen: I tried the hack that you suggested
[15:43:55] <sakoman_> no joy
[15:43:58] <koen> hmmm
[15:44:00] <sakoman_> black screen
[15:44:20] <sakoman_> tried to launch gdm by hand and got a slew of error messages
[15:44:29] <jkridner|work> Digi-Key finally has a serial cable: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=BBC01-ND.
[15:44:44] <sakoman_> at that point it became no longer a "pleasant diversion" :-)
[15:44:56] <AV500> how do I get past the profile selection when booting angstrom?
[15:45:09] <sakoman_> I'll look at it more later, but have some real work to do today
[15:45:36] <AV500> I can select "System Default" or "Ggz", then I am back to "language"... :-(
[15:45:49] <Crofton|irssi> stupid windoze
[15:47:25] <sakoman_> AV500: I haven't seen the loop back to language
[15:47:44] <sakoman_> I always get an empty language list
[15:48:17] <sakoman_> Clicking on next and then selecting Default or Standard gets me to the desktop
[15:48:48] <sakoman_> Haven't figured out how to make e happy with the linguas that are installed
[15:54:06] <koen> sakoman_: did you solve your mmc problems?
[15:58:13] <sakoman_> koen: nope, not yet
[15:58:41] <AV500> sakoman_: I get no profile at all, but somehow pressing "enter" gets me into a next menu, then back to language...
[15:58:56] <AV500> I am using: Angstrom-Beagleboard-demo-image-glibc-ipk-2008.1-test-20080920-beagleboard.rootfs.tar.bz2
[15:59:11] <AV500> but i will try now: Angstrom-Beagleboard-demo-image-glibc-ipk-2008.1-test-20081119-beagleboard.rootfs.tar.bz2
[15:59:36] <sakoman_> AV500: make sure your image recipe contains some config packages, i.e. e-wm-config-standard \
[16:02:15] <AV500> I am just using stuff from the website, I am not building myself...
[16:02:30] <AV500> I just want to "see" something :-(
[16:03:37] <sakoman_> AV500: Ah, OK. Then I can't be much help since I don't use pre-built images
[16:03:42] <koen> the 0920 image doesn't have all that fancy e-wm stuff
[16:03:57] <koen> so if you get a e17 selector, that's not the 0920 image :)
[16:04:05] <koen> fancy, new e-wm stuff
[16:04:16] <AV500> so what image should I use?
[16:04:50] <koen> one from http://www.angstrom-distribution.org/demo/beagleboard/
[16:04:59] <koen> or one from 'untested' if you're brave
[16:05:47] * gduncan (n=spec@rrcs-208-105-88-237.nyc.biz.rr.com) Quit ()
[16:10:29] <koen> cairoperf is still runnig
[16:10:36] <koen> Xsgx taking 100%
[16:10:44] <koen> sounds like heavy xlib work
[16:14:47] * Beagle2 (n=Beagle2@p57AB6BBE.dip.t-dialin.net) has joined #beagle
[16:16:51] * Beagle2 (n=Beagle2@p57AB6BBE.dip.t-dialin.net) Quit (Client Quit)
[16:19:50] <ldesnogu> koen: you got Xsgx to run reliably?
[16:21:33] * gduncan (n=spec@rrcs-208-105-88-237.nyc.biz.rr.com) has joined #beagle
[16:26:24] * gduncan (n=spec@rrcs-208-105-88-237.nyc.biz.rr.com) Quit (Client Quit)
[16:26:54] * gduncan (n=spec@rrcs-208-105-88-237.nyc.biz.rr.com) has joined #beagle
[16:29:25] <koen> ldesnogu: sort off
[16:29:33] * gduncan (n=spec@rrcs-208-105-88-237.nyc.biz.rr.com) Quit (Client Quit)
[16:30:00] <ldesnogu> koen: is that the October beta drop?
[16:30:02] <koen> ldesnogu: I got it to run using binaries from 3 different releases (for different boards!)
[16:30:08] <ldesnogu> ouch
[16:30:13] <Crofton|irssi> sort of does not sound "reliable" :)
[16:30:18] <ldesnogu> :)
[16:30:50] <Crofton|irssi> if you guys get an open source set of libraries for SGX
[16:31:05] <ldesnogu> Crofton|irssi: that won't happen soon :)
[16:31:06] <Crofton|irssi> We can award you the "rms medal of honor"
[16:31:22] <koen> kernel modules and libs from an april snapshot, Xsgx from an unknown snapshot (august?), headers from a beta sdk
[16:31:48] <ldesnogu> you like to live dangerously :)
[16:33:23] * abitos (n=nixgibts@dslb-084-057-206-230.pools.arcor-ip.net) Quit (Read error: 60 (Operation timed out))
[16:35:21] * dirk2 (n=dirk@92.117.85.152) has joined #beagle
[16:42:58] * methril|gone (n=Methril@213.27.233.98) Quit (Remote closed the connection)
[16:49:08] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has left #beagle
[16:49:58] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
[16:51:33] * Zeta_mobile (n=jacobsm@zetadell.dhcp.rose-hulman.edu) has joined #beagle
[17:01:38] * guillaum1 (n=GZBA4143@AMontsouris-153-1-67-44.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
[17:02:19] * guillaum1 (n=gl@AMontsouris-153-1-67-44.w90-2.abo.wanadoo.fr) has joined #beagle
[17:04:07] * rsalveti (n=salveti@201.82.128.154) has joined #beagle
[17:05:55] * kevinsc_ (n=kevinsc@nat/ti/x-21a00f2db005b75e) has joined #beagle
[17:08:58] * gregoiregentil (n=zonbu@adsl-71-135-103-88.dsl.pltn13.pacbell.net) has joined #beagle
[17:13:02] * kevinsc (n=kevinsc@nat/ti/x-f48bbaf8f278c211) Quit (Remote closed the connection)
[17:18:38] * dirk2 (n=dirk@92.117.85.152) has left #beagle
[17:20:26] * gduncan (n=spec@static-71-249-182-7.nycmny.east.verizon.net) has joined #beagle
[17:23:39] * robtow (n=rtow@nat/montavista/x-46a321692ade433c) has joined #beagle
[17:37:28] * kevinsc (n=kevinsc@nat/ti/x-e51f0f16e5d6381e) has joined #beagle
[17:41:25] * kevinsc_ (n=kevinsc@nat/ti/x-21a00f2db005b75e) Quit (Remote closed the connection)
[17:41:33] * Openfree (n=df@222.65.143.118) Quit (Remote closed the connection)
[17:43:51] * florian (n=fuchs@217.146.132.69) Quit (Remote closed the connection)
[17:45:57] <koen> kulve: it seems the delta is to high between experiments, so cairoperf is doing 100 iterations per test
[17:46:35] <kulve> "experiments"?
[17:46:59] <koen> [148] xlib-rgb paint-with-alpha_solid_rgba_over-512 126984 126.984 635.651 50.16% 100
[17:47:41] <koen> so test #148 ran 100 times with a sigma of 50%
[17:47:50] <kulve> sounds a bit odd..
[17:47:52] <koen> or sigma^2, I forget
[17:48:06] <koen> but:
[17:48:06] <koen> [152] xlib-rgb paint-with-alpha_image_rgb_over-512 2047516 2047.516 2048.310 0.02% 4
[17:48:16] <koen> #152 only ran 4 times
[17:51:43] * kevinsc (n=kevinsc@nat/ti/x-e51f0f16e5d6381e) Quit (Remote closed the connection)
[17:51:52] <kulve> "TI's new Linux CG tools site is now up and running."
[17:52:10] <kulve> https://www-a.ti.com/downloads/sds_support/targetcontent/LinuxDspTools/index.html
[17:52:24] <koen> "we did rm -rf and had to recreate it from scratch"
[17:52:58] <koen> is what I read between the lines :)
[17:55:15] <Crofton|irssi> heh
[17:55:46] <koen> jkridner|work: midday US as in "New York" or as in "San Francisco"?
[17:57:21] <jkridner|work> koen: no need to read between the lines.... that is exactly what happened.
[17:57:50] <jkridner|work> hopefully we have something better now recreated.
[17:57:54] * koen has multiple copies of almost everything
[17:57:56] <koen> <- accident prone
[17:57:57] <jkridner|work> more versions, etc.
[17:59:18] <jkridner|work> koen: you asking about my CE/Link meeting? That is 3 Central, so mid-day SFO.
[18:00:13] <koen> thanks
[18:00:28] <koen> fwiw, uboot 1.3.3 solved the colour issues on the .22 kernel
[18:00:43] <koen> and the irq -33 errors
[18:00:56] <prpplague> sakoman_: ping
[18:01:10] <sakoman_> prpplague: pong
[18:01:41] <prpplague> sakoman_: don't suppose you know the actual part number for the wi2wi chipset you guys are using with the serial interface on the bluetooth?
[18:07:04] <koen> wow
[18:07:19] <koen> cairoperf is running a lot faster
[18:07:23] <koen> and I only changed uboot
[18:07:44] * koen sets fire to TI 1.1.4 uboot
[18:08:04] <prpplague> koen: should try apex on the beagle
[18:09:30] <sakoman_> prpplague: w2cbw003-003 if I am reading it right
[18:10:20] <prpplague> sakoman_: that the part number off the chip or the actual part number?
[18:10:38] <sakoman_> the one off the chip, which hopefully is also the actual!
[18:12:29] <prpplague> sakoman_: ok thanks, we were trying to double check and make sure that they didn't send us the wrong ones
[18:12:57] <sakoman_> prpplague: we had lots of issues with getting the wrong ones originally
[18:25:09] * gduncan (n=spec@static-71-249-182-7.nycmny.east.verizon.net) Quit ()
[18:25:17] <Crofton|irssi> I'm told threadX wouldn't be too hard to get running on the Beagle
[18:25:29] <Crofton|irssi> too bad I'm not great at the low level stuff
[18:35:15] * kevinsc (n=kevinsc@nat/ti/x-dc879c0ee1d0cf54) has joined #beagle
[18:35:50] * ldesnogu_ (n=ldesnogu@ven06-2-82-247-86-183.fbx.proxad.net) has joined #beagle
[18:36:20] * dirk2 (n=dirk@92.116.213.241) has joined #beagle
[18:36:33] * ClaudeQC (n=claude@bas1-quebec03-1279636213.dsl.bell.ca) Quit ("Leaving.")
[18:38:04] * ClaudeQC (n=claude@bas1-quebec03-1279636213.dsl.bell.ca) has joined #beagle
[18:40:06] <koen> Crofton|irssi: I'd rather try the linux-RT patch before going entirely RTOS
[18:41:06] <koen> sakoman_: your mmc problem was CD related?
[18:41:15] <Crofton|irssi> of course, but the recruiter dude is offering $100 per hour to do it
[18:41:22] <Crofton|irssi> that's a lot of beer
[18:42:16] <Beagle7> anybody knows the probleme after flushin MLO that the board only starts by pressing user button and keyboard does not work in serial console ? is there any possibility to make something like factory reset
[18:43:05] <koen> Beagle7: tried flashing back the old MLO?
[18:43:18] <koen> there's a coldflash util mentioned in the wiki
[18:47:28] <ldesnogu_> oh we now have access to TI ARM compiler it seems
[18:48:15] <Crofton|irssi> we should try a build with it :)
[18:48:27] <kulve> what's TI ARM compiler..?
[18:48:50] <ldesnogu_> kulve, it's TI compiler for ARM CPUs
[18:49:00] <kulve> or DSP cpus..?
[18:49:03] <ldesnogu_> the link jkridner|work posted is great :)
[18:49:13] <ldesnogu_> kulve, https://www-a.ti.com/downloads/sds_support/targetcontent/LinuxDspTools/download.html
[18:49:42] <ldesnogu_> Crofton|irssi, yes, we should convince koen to give it a try with OE :)
[18:49:57] <kulve> ldesnogu_: aren't those dsp compilers..?
[18:50:30] <kulve> TMS470 seems to be some arm..
[18:50:32] <koen> kulve: TMS470
[18:50:32] <ldesnogu_> kulve, you should really give this page a look...
[18:50:46] <ldesnogu_> 4.5.x Releases - for ARM 7,9,11, Cortex M3, R4, R4F, A8 Cores
[18:51:04] <koen> ldesnogu_: iirc mru said it's not 'gcc compatible', so you can't mix code AIUI
[18:51:13] <mru> that's correct
[18:51:22] <ldesnogu_> it doesn't support EABI?
[18:51:24] <mru> it won't even link without a mallet
[18:51:27] <ldesnogu_> or is that a file format issue
[18:51:29] <mru> it support eabi
[18:51:32] <koen> ldesnogu_: it does
[18:51:36] <koen> ldesnogu_: https://www-a.ti.com:443/downloads/sds_support/targetcontent/LinuxDspTools/doc/TMS470/TMS470CodeGenerationTools4.5.0ReleaseNotes.txt
[18:52:03] <robclark> hmm, don't know if it is still the case, but back in the day, CCS used COFF format, and I think these days GCC is all ELF
[18:52:13] <mru> it uses a different type for wchar_t, and tags the object files as such, so it won't link with normal gcc stuff
[18:52:24] <mru> if you never use that type, you can remove the attribute and have it link
[18:52:31] <kulve> I'm a bit lost, what's tms470?
[18:52:33] <ldesnogu_> and there's no option to change that?
[18:52:36] <sakoman_> koen: yes
[18:52:47] <mru> when it works, it (unsurprisingly) produces faster code than gcc
[18:52:53] <mru> when it doesn't work...
[18:52:54] <koen> kulve: the TI name for their arm compiler
[18:53:06] <mru> there's no option, no
[18:53:12] <mru> just like there wasn't with rvds 3
[18:53:13] <ldesnogu_> robclark, and COFF has its own TI flavor :)
[18:53:28] <mru> the TI ARM compiler uses ELF
[18:53:37] <ldesnogu_> tms470 is TI internal name for ARM microcontrollers IIRC
[18:53:37] <kulve> "TMS470 32-bit RISC ARM7TDMI?-based microcontroller"
[18:53:46] <kulve> they seem to have also cpus with that name..?
[18:54:07] <robclark> ok, could be... I've not actually used the TI compiler for, 5+ years
[18:54:29] <koen> -mv7A8 --neon --float_support=VFPv3 --abi=eabi
[18:54:43] <ldesnogu_> no vectorize? :)
[18:54:49] <mru> it tries
[18:54:51] <mru> and fails
[18:54:56] <mru> assembler chokes
[18:54:58] <ldesnogu_> hehe
[18:55:10] <mru> even in a 4.6 prerelease
[18:55:21] <ldesnogu_> this can probably worked around with sed'ing the produced asm file :p
[18:55:32] <mru> it's not that simple
[18:55:48] <koen> kulve: I limited cairoperf to 10 tries per test case, should speed things up a bit
[18:56:09] <kulve> koen: ok
[18:56:30] <mru> it tried to do vldr d0, [r0, #1]
[18:56:40] <mru> the offset has to be a multiple of 4
[18:57:26] <ldesnogu_> was it really expecting an offset of 1, or is it just a dump problem?
[18:57:42] <ldesnogu_> if it's the former then indeed it's unusable
[18:58:07] <mru> void bug(char *p, char *q){*(long long*)(p) = *(long long*)(q+1);}
[18:59:03] <mru> guess what I was compiling when I found it...
[18:59:08] <ldesnogu_> :)
[18:59:49] <ldesnogu_> so it fails in similar code as latest CSL gcc :)
[19:00:02] <mru> no
[19:00:21] <ldesnogu_> didn't you find a bug with NEON alignment issues?
[19:00:36] <mru> the csl gcc was creating valid instructions that failed at runtime
[19:00:48] <mru> the TI compiler outputs invalid instructions
[19:00:51] <ldesnogu_> bummer
[19:00:56] <ldesnogu_> I see now...
[19:01:03] <mru> the csl code was fine if the pointers were aligned
[19:02:08] <ldesnogu_> what's strange is that it tries to generated a load that isn't necessarily aligned
[19:02:27] <ldesnogu_> nothing gives information about q alignment in your example
[19:02:53] <mru> the compiler is free to assume that pointers have natural alignment
[19:02:54] <ldesnogu_> hum, can't remember: can NEON loads be unaligned?
[19:03:04] * geckosenator (n=sean@c-24-8-193-190.hsd1.co.comcast.net) has joined #beagle
[19:03:13] <mru> some neon loads can be unaligned but not that one
[19:03:33] <mru> VLDn without an alignment specifier can be unaligned
[19:03:52] * magnet (n=magnet@AMontpellier-259-1-40-228.w92-133.abo.wanadoo.fr) Quit (Read error: 110 (Connection timed out))
[19:04:08] <mru> but the C standard says if a pointer doesn't have the natural alignment for its type, results are undefined
[19:04:24] <mru> so a long long pointer can be assumed to be 8-byte aligned
[19:04:43] <ldesnogu_> mru, did you try the RVCT assembler option that gives information about NEON stall cycles?
[19:04:55] <mru> yes, I did
[19:05:01] <mru> got some hints from it
[19:05:02] <ldesnogu_> was it useful?
[19:05:06] <ldesnogu_> cool :)
[19:05:06] * ZeZu (n=null@c-24-14-8-40.hsd1.in.comcast.net) Quit (Connection reset by peer)
[19:05:37] <mru> although the first thing it prints is a warning that stall cycle counts are unreliable
[19:06:25] <ldesnogu_> it perhaps doesn't take into account fine details such as instruction alignment (which probably plays a role) and perhaps even pairing
[19:06:33] <mru> it would be neat if rvct could be told to accept gnu-style assembler
[19:06:52] <mru> I read somewhere it had to do with dual-issue
[19:07:03] <ldesnogu_> so it's pairing
[19:07:29] <mru> but any documentation released to the public is probably not to be trusted in such detail
[19:07:35] * gcohler (n=gcohler@nat/ti/x-ebc63821fb7fb195) Quit (Remote closed the connection)
[19:07:52] <ldesnogu_> well it even happen sometimes that internal documentation is not accurate
[19:08:05] <mru> I wouldn't be surprised if it's just a made-up excuse
[19:08:11] <ldesnogu_> it's like software, hardware design moves much quicker than doc :)
[19:08:48] <mru> I once asked for some documentation from 3dlabs, and they said they hadn't written it
[19:09:05] <Crofton|irssi> at least they were honest ...
[19:09:10] <ldesnogu_> :)
[19:09:11] <mru> apparently someone with money also asked for it, because it got written a few months later
[19:09:21] <mru> and they were kind enough to send me a copy
[19:09:58] <mru> my videos played much better after that
[19:10:07] <ldesnogu_> I have heard people saying noone should ever put any comment in code (C, Verilog, ...) because they become obsolete
[19:10:25] <mru> I say code shouldn't need comments
[19:10:33] * dirk2 (n=dirk@92.116.213.241) has left #beagle
[19:10:38] <ldesnogu_> I think the same, but there's a limit to that
[19:10:57] <ldesnogu_> not describing state machines with comments is crazy
[19:11:29] <mru> state machines are evil
[19:11:34] <mru> and there's far too many of them
[19:11:43] <ldesnogu_> well for hardware... :)
[19:11:53] <mru> hardware is different of course
[19:12:20] <mru> software state machines are frequently just trying to track the state of something else that could equally well be queried when it's needed
[19:12:48] <mru> and when the states inevitably go out of sync, all hell breaks loose
[19:13:37] <ldesnogu_> I find some state machines easier to read than other forms of code
[19:14:54] <mru> sometimes a state machine is the proper solution
[19:15:21] <mru> but every company has at least one Mr State Machine that solves *every* problem with one
[19:15:36] <ldesnogu_> pfff I was going to answer your patch for ftrv lacked warning about FPU exceptions :)
[19:15:52] <ldesnogu_> big state machines are evil for sure
[19:16:05] <ldesnogu_> unless generated by a tool
[19:16:07] <mru> qemu is a state machine...
[19:16:39] <ldesnogu_> hum what part of it would you call a state machine?
[19:16:52] <mru> all of it
[19:16:57] * ZeZu (n=null@c-24-14-8-40.hsd1.in.comcast.net) has joined #beagle
[19:17:21] <ldesnogu_> funny I don't see it as a state machine at all
[19:18:05] <mru> the contents of registers and memory are the state
[19:18:31] <mru> instructions executed change the state
[19:18:31] <ldesnogu_> you can reduce any program to a state machine :)
[19:18:50] * ldesnogu_ wonder why he wrote "reduce"...
[19:19:37] <robclark> you can reduce any processor to a state machine (or, put another way, a turing machine)... the program is just the data fed into the state machine
[19:19:43] <robclark> ;-)
[19:19:46] <ldesnogu_> :)
[19:20:10] <mru> that's kind of what I said
[19:20:13] <mru> or meant
[19:20:25] <robclark> yup, I understood it that way
[19:20:41] <ldesnogu_> and so did I
[19:21:06] <mru> my computer has 4GB of ram and 3TB disk space, how many states does it have?
[19:21:29] <ldesnogu_> I would say it has too many slow state
[19:21:30] <robclark> lots
[19:21:53] <ldesnogu_> I'd rather have 8 GB and 200 GB disk
[19:22:34] <mru> guess you don't have many HD films stored
[19:22:42] <ldesnogu_> none
[19:22:47] <ldesnogu_> I never download anything
[19:22:59] <mru> I didn't say I do...
[19:23:07] <ldesnogu_> did I imply this?
[19:23:08] <ldesnogu_> :)
[19:23:19] <ldesnogu_> you work on ffmpeg, you need films
[19:23:37] <mru> yep, it's all for research purposes
[19:24:04] <mru> maybe I shouldn't mention why I started working on ffmpeg in the first place...
[19:24:37] <ldesnogu_> you said too much, go on :)
[19:24:57] <mru> to optimise it for my computer of course
[19:25:26] <ldesnogu_> which was??
[19:25:27] <ldesnogu_> ?
[19:25:32] <mru> alpha
[19:25:51] <ldesnogu_> did it have the multimedia instruction (264 ?)?
[19:26:02] <mru> pca56
[19:26:14] <mru> so some, but not all
[19:26:57] <ldesnogu_> 21164PC then
[19:27:01] <mru> right
[19:27:25] <ldesnogu_> can't remember when they introduced byte ld/st
[19:27:31] <ldesnogu_> was it 264 or earlier?
[19:27:40] <mru> ev5
[19:27:43] <mru> 164
[19:27:58] <mru> maybe even ev45, I can't remember
[19:27:58] <ldesnogu_> really? I thought it was later
[19:28:11] <mru> sparse mappings are fun...
[19:28:14] <ldesnogu_> ev56 in fact
[19:28:17] <ldesnogu_> so 164A
[19:28:30] <ldesnogu_> if wikipedia has it right of course...
[19:29:14] <mru> I didn't do much low-level hacking on anything before ev56
[19:29:17] <ldesnogu_> do you run Linux or Tru64?
[19:29:27] <mru> linux
[19:29:40] <mru> I have a tru64 disc somewhere
[19:29:40] <ldesnogu_> and do you have Compaq compilers for Linux?
[19:29:45] <mru> yes
[19:29:50] <ldesnogu_> ha :)
[19:29:54] <mru> tru64 doesn't support any modern hardware
[19:30:16] <mru> video cards, sata controllers and the like
[19:30:47] <ldesnogu_> s your alpha PCI based?
[19:30:53] <mru> yes
[19:31:49] <ldesnogu_> (to be even more off-topic, your ftrv implementation needs a comment about inaccuracies and perhaps would be better implemented as a helper)
[19:32:51] <mru> when should something be a helper?
[19:33:06] <ldesnogu_> when it exceeds a given number of tcg ops
[19:33:12] <ldesnogu_> Fabrice said about 20
[19:33:22] <ldesnogu_> but that probably depends on host and compiler
[19:33:38] <ldesnogu_> I never measured that
[19:34:19] * likewise_ (n=chatzill@82-171-51-231.ip.telfort.nl) Quit (Read error: 104 (Connection reset by peer))
[19:34:50] * mru didn't look at qemu source until about 2 days ago
[19:35:05] <ldesnogu_> you're good at that :)
[19:35:15] <ldesnogu_> like NEON stuff
[19:35:42] <ldesnogu_> BTW did you try to enable the DEBUG define in tcg.h? It can spot some typing erros
[19:35:59] <mru> I noticed it, didn't try it
[19:36:09] <ldesnogu_> but I don't think you have any such error
[19:36:31] <mru> I'm checking it now just in case
[19:38:02] * likewise_ (n=chatzill@82-171-51-231.ip.telfort.nl) has joined #beagle
[19:39:41] <mru> hmm... seems there were a couple of mistakes
[19:42:56] <mru> is the qemu mailing list moderated?
[19:46:50] <ldesnogu_> I don't know
[19:47:20] <mru> I sent those patches last night and they showed up only recently
[19:47:41] <ldesnogu_> so it's probably moderated
[19:52:37] <mru> I'll subscribe and send updated patches
[19:57:24] * felipec (n=felipec@a91-153-251-222.elisa-laajakaista.fi) has joined #beagle
[19:57:35] * magnet (n=magnet@AMontpellier-259-1-72-134.w92-133.abo.wanadoo.fr) has joined #beagle
[19:58:36] * uwe_ (n=uwe_@dslb-084-056-045-071.pools.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
[19:58:59] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit ()
[19:59:46] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
[20:00:13] <ldesnogu_> mru, your mails arrived
[20:07:48] * abitos (n=nixgibts@p549E483C.dip.t-dialin.net) has joined #beagle
[20:09:38] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Remote closed the connection)
[20:14:15] * likewise_ (n=chatzill@82-171-51-231.ip.telfort.nl) Quit ("ChatZilla 0.9.84 [Firefox 3.0/2008061015]")
[20:14:53] * FuL|OUT is now known as fulgas
[20:29:10] * geckosenator (n=sean@c-24-8-193-190.hsd1.co.comcast.net) Quit (Read error: 110 (Connection timed out))
[20:30:36] <koen> likewise: new swfdec in OE
[20:32:40] * abitos (n=nixgibts@p549E483C.dip.t-dialin.net) Quit (Read error: 110 (Connection timed out))
[20:35:40] * uwe_ (n=uwe_@dslb-088-064-073-255.pools.arcor-ip.net) has joined #beagle
[20:52:06] * fulgas is now known as FuL|OUT
[20:52:56] <sakoman_> koen: does swfdec work with Firefox?
[20:54:48] <koen> not with 3.0 and 3.0.1
[20:55:03] <koen> 3.0.2 and up work
[20:55:43] * gregoiregentil (n=zonbu@adsl-71-135-103-88.dsl.pltn13.pacbell.net) has left #beagle
[20:56:45] <sakoman_> koen: ah, OK. one more version bump and we are there
[20:58:42] <Crofton|irssi> anyone seen ID #1 respawning to fast messages?
[20:59:03] <koen> yes, it means your init is messed up
[20:59:14] <koen> I instructed khem to bug you next week :)
[21:00:21] <Crofton|irssi> heh
[21:00:28] <Crofton|irssi> yeah, I'll do a build
[21:00:43] <Crofton|irssi> I need to read one more document and match it against some code
[21:01:09] <koen> I once had a document match the code
[21:01:17] <koen> except that the same idiot wrote both
[21:01:46] <sakoman_> koen: I take it that that 3.0.4 and 3.1 firefox recipes are broken?
[21:01:50] <koen> not much fun going through IDL code that does fancy math for satelite image projections
[21:02:09] <koen> sakoman_: 3.1b works, 3.0.4 misses the wchar patch that makes it actually compile
[21:02:21] * koen is using 3.1b on his beagle
[21:02:51] <sakoman_> ah, ok. both have DEFAULT_PREFERENCE set to -1
[21:03:02] <koen> yes
[21:03:06] <sakoman_> I'll give it a try on overo
[21:03:09] <koen> since 3.1 is a beta and 3.0.4 is broken
[21:03:41] <koen> http://www.angstrom-distribution.org/feeds/2008/ipk/glibc/armv7a/base/firefox_3.0.1+3.1b1-r0_armv7a.ipk
[21:04:02] <sakoman_> any major issues with 3.1?
[21:04:29] <koen> you can't download extentions, since for some reason mozilla disallows 3.1 clients
[21:04:49] <koen> it's slower and uses more ram than webkit
[21:04:55] <sakoman_> I assume it is as slow as the previous versions?
[21:05:02] <koen> no, it's faster
[21:05:05] <koen> as in useable
[21:05:22] <sakoman_> cool!
[21:06:07] <koen> try http://people.mozilla.com/~vladimir/demos/photos.svg for fun
[21:10:56] <koen> it works in epiphany and firefox
[21:12:41] * guillaum1 (n=gl@AMontsouris-153-1-67-44.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
[21:14:32] <robclark> seems to work fine with safari, so I assume WebKit too
[21:16:38] <koen> ok, Xsgx cairo perf done, moving to Xorg
[21:20:41] * robclark_ (n=robclark@nat/ti/x-cca5a21accc77341) has joined #beagle
[21:24:40] * robclark (n=robclark@nat/ti/x-78ec326194a13225) Quit (Remote closed the connection)
[21:45:37] <Vegar> that is neat, koen
[21:48:16] <Stskeeps> Xsgx?
[21:52:24] * likewise_ (n=chatzill@82-171-51-231.ip.telfort.nl) has joined #beagle
[21:53:16] * sakoman_ (n=sakoman@static-74-41-60-154.dsl1.pco.ca.frontiernet.net) Quit ("Ex-Chat")
[21:56:09] * surki (n=surki@gek7.kyla.fi) Quit (Read error: 60 (Operation timed out))
[21:56:12] * surki (n=surki@gek7.kyla.fi) has joined #beagle
[22:00:07] * BThompson (n=BThompso@nat/ti/x-e65c0d64881dba57) Quit ("Trillian (http://www.ceruleanstudios.com")
[22:29:57] * kevinsc (n=kevinsc@nat/ti/x-dc879c0ee1d0cf54) Quit (".")
[22:35:30] * flo_lap (n=fuchs@f049004049.adsl.alicedsl.de) has joined #beagle
[22:40:39] * sakoman (n=sakoman@static-74-41-60-154.dsl1.pco.ca.frontiernet.net) has joined #beagle
[22:47:51] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) has joined #Beagle
[22:49:54] * codepope (n=codepope@ip-217.146.111.161.merula.net) Quit ()
[22:53:54] * robclark_ (n=robclark@nat/ti/x-cca5a21accc77341) Quit (Remote closed the connection)
[22:57:51] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit (Read error: 110 (Connection timed out))
[22:58:43] * cbrake is now known as cbrake_away
[22:58:51] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
[22:59:24] <koen> kulve: ping
[22:59:25] <koen> kulve: http://dominion.thruhere.net/koen/OE/cairo-perff-diff.txt
[23:00:04] * robtow (n=rtow@nat/montavista/x-46a321692ade433c) Quit ("Leaving.")
[23:00:16] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
[23:00:55] * koen is not impressed with Xsgx for regular usage
[23:00:58] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
[23:01:15] <ds2> koen: which Xsgx are you using?
[23:03:33] <mru> koen: server not talking
[23:04:14] * hli (i=chaton@vig91-2-82-232-97-149.fbx.proxad.net) has joined #beagle
[23:04:41] <koen> ds2: how do I get the version? my sgx install is a bit cobbled together from multiple sources
[23:05:05] <ds2> koen: the source of the package should be sufficient
[23:05:41] <ds2> is it from the reference kernel SGX driver package?
[23:06:26] <koen> the kernel driver is from an april snapshot, for .22
[23:06:34] <koen> libs are from the lrl demo
[23:06:45] <koen> headers from an sdk
[23:06:58] <koen> and iirc xsgx is from an sdk as well
[23:07:19] <koen> I'm not sure, since the originals were in a windows vm that I deleted
[23:07:40] <ds2> hmm
[23:07:58] <ds2> where is the choke points you are seeing?
[23:08:48] <koen> http://dominion.thruhere.net/koen/OE/cairo-perff-diff.txt
[23:08:57] <koen> I went "wow, 6x speedup"
[23:09:09] <koen> and then "wow!!!! 19x slowdown"
[23:10:18] * beewoolie (n=beewooli@66.147.194.238) Quit ("Leaving")
[23:11:52] <ds2> where you comparing it to stock Xfbdev on the same kernel?
[23:12:07] <koen> yes
[23:12:29] <koen> I think the differences are due to different version of things like e.g. pixman
[23:12:40] <koen> imgtec only seemed to have added eglx to kdrive
[23:12:58] <koen> and not acceleration for e.g. xrender or xlib primitives
[23:13:16] <ds2> trying to figure if using the reference omapfb vs the l-o omapfb had any factors there
[23:13:37] <koen> I'm comparing against 2.6.28rc4 now
[23:15:09] <ds2> has anyone gotten Xomap working on the beagle?
[23:15:34] <koen> no, the work is focused on xf86-video-omapfb
[23:15:42] <koen> Xomap is EOL'ed by nokia
[23:16:07] <ds2> oh... was the Xomap sources ever available?
[23:16:14] <koen> yes
[23:40:49] * likewise_ hmm, not beagle build for me today :-/ I'm off to bed, cya!
[23:41:31] * likewise_ (n=chatzill@82-171-51-231.ip.telfort.nl) Quit ("ChatZilla 0.9.84 [Firefox 3.0/2008061015]")
[23:42:48] <RobotGuy> My second build of console-image is much better for some reason. I have a complete toolchain now - C and C++. :)
[23:49:26] <ds2> what? no java or fortran toolchains? *gasp* the horror!
[23:50:47] <koen> fortran should be there
[23:50:57] <koen> I haven't tried to build gcj yet
[23:54:15] * valhalla (n=valhalla@81-174-21-31.dynamic.ngi.it) Quit ("Leaving")
[23:58:18] * flo_lap (n=fuchs@f049004049.adsl.alicedsl.de) Quit (Read error: 113 (No route to host))