Join the chat at beagleboard.org/chat
IRC Log for 2008-07-22
Timestamps are in UTC.
- [00:00:00] <paul_pwsan>
mru: the dss1 portion of the clock tree needs some work, but you should be able to get effective rate changing in the interim by modifying clock34xx.h: dpll4_m4_ck in the following manner:
- [00:00:30] <paul_pwsan>
- add .set_rate = &omap2_clksel_set_rate
- [00:00:37] <paul_pwsan>
- add .round_rate = &omap2_clksel_round_rate
- [00:00:50] <mru>
there's a problem with that
- [00:01:00] <mru>
there's a bit of a loop in the structure
- [00:01:12] <mru>
which is why I'm somewhat confused
- [00:01:39] <mru>
first of all, the .clksel_reg and .clksel_mask fields of dss1_alwon_fck seem weird
- [00:03:41] <paul_pwsan>
mru: it's not exactly optimal, but should be functional. for the time being, you need to issue the clk_set_rate/clk_round_rate on the dpll4_m4_ck, and the enable/disable on dss1_alwon_fclk
- [00:04:05] <paul_pwsan>
the weirdness is because the dpll will emit sys_ck when in bypass
- [00:05:19] <mru>
I don't get it... dss1_alwon_fck.clksel_reg points at CM_IDLEST_CKGEN
- [00:05:30] <mru>
this doesn't make any sense to me
- [00:07:02] <mru>
ok, I think I see what's going on now
- [00:07:14] <paul_pwsan>
when CM_IDLEST_CKGEN.ST_PERIPH_CLK == 0, DPLL4 is bypassed with the sys_ck freq
- [00:07:31] <mru>
still strange to have three struct clk for a single clock
- [00:11:28] <paul_pwsan>
the dss code should only need to know about one of them: dss1_alwon_fclk.
- [00:11:40] <paul_pwsan>
not at this exact moment in time, unforunately.
- [00:20:52] * johnx (n=john@p4003-ipbf301hodogaya.kanagawa.ocn.ne.jp) Quit (Read error: 113 (No route to host))
- [00:24:49] <mru>
paul_pwsan: thanks, using clk_set_rate on dpll4_m4_ck seems to work with that change
- [00:25:29] <paul_pwsan>
mru: cool
- [00:27:49] <paul_pwsan>
sakoman_ and any other i2c testers: a tarball with the updated set of 9 twl4030 patches which incorporates Peter De Schrijver's feedback is at: http://www.pwsan.com/omap/twl4030_set_v2.tar.gz
- [00:28:03] <paul_pwsan>
seems to work fine here, will plan to post tomorrow to linux-omap.
- [00:34:35] <mru>
is there a reason there is no clk_get_parent for omap2/3?
- [00:36:44] <paul_pwsan>
not as far as i know... i guess no drivers have wanted it yet
- [00:36:49] <mru>
what I'm really after is the rate going into the dpll4 m4 divider
- [00:37:11] <mru>
the rate I'll want to request depends on what I can actually get
- [00:37:47] * johnx (n=john@p4003-ipbf301hodogaya.kanagawa.ocn.ne.jp) has joined #beagle
- [00:38:25] <paul_pwsan>
makes sense. a patch to add a get_parent would definitely be welcome...
- [00:38:50] <mru>
does it need to do anything other than return clk->parent?
- [00:38:57] <paul_pwsan>
don't think so.
- [00:39:33] <mru>
guess I'll add one below get_parent then
- [00:39:41] <mru>
or is there a better place for it?
- [00:40:51] <paul_pwsan>
there's already a stub for it in plat-omap/clock.c. should only have to add the function itself into arch/arm/mach-omap2/clock.c and modify the struct clk_functions assignments in clock24xx.c/clock34xx.c
- [00:41:41] <mru>
yes, that's what I was thinking
- [00:49:40] <mru>
seems to work
- [00:51:16] <mru>
is it preferred to keep the struct clk_functions assignments in a logical order, or do new entries simply get added at the end?
- [00:52:05] <paul_pwsan>
probably best to use the order in the struct clk_functions def in include/asm-arm/arch-omap/clock.h
- [00:52:16] <paul_pwsan>
dunno if they are currently in that order
- [00:52:37] <paul_pwsan>
if not, then anywhere would be fine
- [00:53:10] <mru>
they seem to have that order currently
- [00:53:24] <mru>
and it's what I'd consider the most logical order as well
- [00:54:43] <paul_pwsan>
good to hear
- [01:00:05] <paul_pwsan>
mru: perhaps you might be willing to try some untested patches? hehe...
- [01:00:08] <paul_pwsan>
this set: http://www.pwsan.com/omap/use_parent_rate_ops_dss1.tar.gz2
- [01:00:25] <paul_pwsan>
should hopefully let you call clk_round_rate/clk_set_rate on dss1_alwon_fclk
- [01:00:44] <paul_pwsan>
although sadly, will not help with the parent clock rate getting portion -
- [01:00:58] <paul_pwsan>
will have to consider that more closely
- [01:01:53] <mru>
I've committed a get_parent to my git
- [01:04:32] <paul_pwsan>
yes, but to use it, you'll still need to clk_get() that M4 divider clock, and that's what i'd like to avoid, at least in the medium to long term
- [01:04:45] <mru>
yes, of course
- [01:06:21] <mru>
I don't think your patches will work quite as intended
- [01:06:55] <mru>
adding .set_rate to dss1_alwon_fck will have it called immediately without invoking the parent-finding code
- [01:07:39] <paul_pwsan>
oops, indeed, please scratch the second patch
- [01:08:05] <mru>
those should go on the dpll4_m4_ck instead
- [01:08:17] <paul_pwsan>
yep
- [01:08:35] <mru>
isn't this a bit dangerous though?
- [01:09:09] <mru>
if someone tries to set_rate on a clock that doesn't have one, it could end up changing one of the main dplls
- [01:09:22] <mru>
and that's probably not good
- [01:11:39] <paul_pwsan>
in practice, i don't think it will be much of a problem, since the only drivers that are likely to attempt to set their rate are drivers where the author knows that they have a divider upstream
- [01:11:57] <mru>
yes, of course
- [01:12:29] <mru>
I just have habit of trying to avoid making really bad mistakes
- [01:13:04] <mru>
and accidentally messing with a clock could have consequences all over the place
- [01:14:45] <paul_pwsan>
in the medium- to long-term, i personally think that any driver that uses a clock that is sensitive to changes or glitches needs to have a pre- and post-rate-change notifier registered
- [01:15:07] <paul_pwsan>
obviously, we are not there yet.
- [01:15:49] <paul_pwsan>
but the general problem is similar to ones faced by drivers that depend on cpu frequency: something else on the system can come along and change it
- [01:16:05] <mru>
I was thinking of clocks that are never meant to change
- [01:16:32] <paul_pwsan>
from the device's perspective or the system's?
- [01:20:02] <mru>
well, there are a number of clocks that are assumed to have a fixed rate, such as the 96MHz clock feeding all sorts of things
- [01:20:39] <mru>
obviously the plls will have to be programmed depending on the external clock used
- [01:20:59] <mru>
but I'd expect that to only happen once during bootup
- [01:23:36] <paul_pwsan>
in the TI OPP model, that is definitely true (with the exception of DPLL5 i believe.) but some applications may want to adjust DPLL frequency to save power if some downstream device is disabled.
- [01:24:23] <mru>
I knew there was something I wasn't thinking of ;-)
- [01:25:54] <paul_pwsan>
apparently in TI's DPLL blocks, the lower the divider, the more power is saved
- [01:30:09] <mru>
well, it's getting late here... if you think of a good solution to the dss1 situation, let me know
- [01:30:26] <paul_pwsan>
will do...
- [01:30:37] <paul_pwsan>
thanks for talking through the problems with me.
- [01:39:06] * rsalveti (n=salveti@189.70.143.203) Quit (Read error: 104 (Connection reset by peer))
- [01:39:16] * rsalveti (n=salveti@189.70.143.203) has joined #beagle
- [03:04:00] * MattJ (n=MattJ@88-111-173-221.dynamic.dsl.as9105.com) Quit (Read error: 110 (Connection timed out))
- [04:06:13] * rsalveti (n=salveti@189.70.143.203) Quit ("Bye :)")
- [04:14:06] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) Quit ("Trillian (http://www.ceruleanstudios.com")
- [04:33:35] * paul_pwsan (n=paul@utopia.booyaka.com) has left #beagle
- [05:05:02] * KSoze (n=Miles@67.107.206.170.ptr.us.xo.net) has joined #beagle
- [05:06:06] * KeyserSoze (n=Miles@67.107.206.170.ptr.us.xo.net) Quit (Read error: 54 (Connection reset by peer))
- [05:36:10] * Viral_Sachde (n=Viral_Sa@122.167.133.54) has joined #beagle
- [05:41:04] * Olipro (i=Olipro@unaffiliated/olipro) has joined #beagle
- [06:10:39] * bazbell (n=a0192809@nat/ti/x-718db01e8e9c72c1) Quit (Remote closed the connection)
- [06:37:24] * Viral_Sachde (n=Viral_Sa@122.167.133.54) Quit (Read error: 110 (Connection timed out))
- [07:31:23] * ciskje (n=francesc@131.114.33.209) has joined #beagle
- [08:11:46] * ciskje (n=francesc@131.114.33.209) Quit (Remote closed the connection)
- [08:29:43] * ciskje (n=francesc@131.114.33.209) has joined #beagle
- [08:34:43] <koen>
good morning all
- [08:34:56] <mru>
morning koen
- [08:38:59] * ciskje (n=francesc@131.114.33.209) Quit (Remote closed the connection)
- [08:42:56] * Viral_Sachde (n=Viral_Sa@122.167.147.142) has joined #beagle
- [08:57:15] * Olipro_ (i=Olipro@unaffiliated/olipro) has joined #beagle
- [08:58:22] * royerfa (n=x0091539@nat/ti/x-5cf6f0759ece9484) has joined #beagle
- [09:02:07] * Olipro (i=Olipro@unaffiliated/olipro) Quit (Nick collision from services.)
- [09:02:12] * Olipro_ is now known as Olipro
- [09:14:00] * ciskje (n=francesc@131.114.33.209) has joined #beagle
- [09:16:30] * ciskje (n=francesc@131.114.33.209) Quit (Remote closed the connection)
- [09:19:32] <koen>
*sigh*
- [09:19:49] <koen>
mythtv configure looks in /usr/lib to see if one was a lib
- [09:20:06] <koen>
of course my arm crosscompiler can link with x86 libs
- [09:25:39] * MattJ (n=MattJ@88-111-173-221.dynamic.dsl.as9105.com) has joined #beagle
- [09:39:08] * royerfa (n=x0091539@nat/ti/x-5cf6f0759ece9484) Quit (Remote closed the connection)
- [09:45:53] * bjdooks tries to persuade boss to order beagle
- [09:57:20] <koen>
janneg: http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=blob_plain;f=packages/mythtv/mythtv/configure.patch;hb=HEAD
- [10:42:49] <koen>
man, mythtv is nasty
- [10:43:00] <koen>
arm has no cmov
- [10:43:09] <koen>
why does myth enable it?
- [10:43:57] <koen>
why does mythtv configure does first check gcc foo.c -o bar -lmp3lame and when it succeed check if the lib in in LD_PATH?
- [10:44:08] <koen>
(where LD_PATH was hardcoded to /usr/lib)
- [10:44:19] <koen>
if it links, the lib is there, no need to double check
- [10:53:34] * docelic (n=docelic@78.134.192.57) has joined #beagle
- [10:58:54] <janneg>
koen: probably since your host has cmov, looking at the patch now
- [11:00:38] <koen>
there's some tab/space mess in that patch, but it should get the general idea across
- [11:01:23] <koen>
configure finished now, but it still puts ARCH_X86 into config.{h,mak}
- [11:02:42] <koen>
janneg: btw, the current ffmpeg configure script support cross-compiling fine (thanks to mru) since a year or 2
- [11:04:33] <janneg>
koen: thanks for the patch, I don't know why we added the has_library() and has_header(). was before my involment
- [11:05:01] * Olipro_ (i=Olipro@unaffiliated/olipro) has joined #beagle
- [11:05:10] <janneg>
I'm trying with every merge to make it look more like ffmpeg's configure
- [11:05:45] <janneg>
last sync was over half a year, next sync is soon
- [11:06:06] * royerfa (n=x0091539@nat/ti/x-212e26557e8b7d5f) has joined #beagle
- [11:07:28] <janneg>
koen: did you use --arch=arm*? what's your configure commandline
- [11:12:40] * incandescant (n=incandes@foo.stupids.org) has joined #beagle
- [11:22:53] * Olipro (i=Olipro@unaffiliated/olipro) Quit (Read error: 110 (Connection timed out))
- [11:25:58] <Crofton>
koen, any chance you can try nfs mounting something onto the beagle?
- [11:28:11] * dannyBlue|DDd (n=chatzill@89.181.68.180) has joined #beagle
- [11:30:41] <koen>
./configure --prefix=/usr --mandir=/usr/man --cpu=armv5te --disable-altivec --disable-strip --enable-v4l --enable-audio-oss --enable-proc-opt --enable-dvb --enable-libmp3lame --cross-compile --dvb-path=/OE/angstrom-tmp/staging/arm-angstrom-linux-gnueabi/usr/include --enable-armv5te
- [11:33:16] <koen>
it's better with --arch
- [11:33:18] <koen>
but still:
- [11:33:19] <koen>
koen@dominion:/OE/angstrom-tmp/work/armv5te-angstrom-linux-gnueabi/mythtv-0.21+svnr17880-r0$ grep ARM mythtv/*
- [11:33:22] <koen>
mythtv/config.h:#define ENABLE_ARMV4L 0
- [11:33:24] <koen>
mythtv/config.h:#define ENABLE_ARMV5TE 0
- [11:33:27] <koen>
mythtv/config.h:#define ENABLE_ARMV6 0
- [12:01:53] <janneg>
koen: I fail to see where it fails now. I'll try it myself later today
- [12:03:05] <janneg>
I guess grep ARCH mythtv/config.h will show the host arch
- [12:03:21] <koen>
janneg: it doesn't fail :)
- [12:03:31] <koen>
janneg: but it also doesn't enable the armv5 optimization
- [12:04:00] <janneg>
koen: it fails to set the correct arch
- [12:05:17] <janneg>
otherwise at least ENABLE_ARMV4L would be defined as 1
- [12:07:20] * trickie|away (n=trickie@basesoft.xs4all.nl) Quit ("Leaving")
- [12:12:23] * trickie (n=trickie@basesoft.xs4all.nl) has joined #beagle
- [12:16:58] * docelic_ (n=docelic@78.134.198.171) has joined #beagle
- [12:25:41] <koen>
root@beagleboard:~# uptime 12:25:28 up 3:47, 1 user, load average: 0.00, 0.01, 0.00
- [12:25:46] <koen>
still no serial lockup :)
- [12:27:27] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
- [12:29:05] * ciskje (n=francesc@131.114.33.209) has joined #beagle
- [12:31:23] * BThompson (n=BThompso@nat/ti/x-92ec3ba18589b007) has joined #beagle
- [12:34:07] * docelic (n=docelic@78.134.192.57) Quit (Read error: 110 (Connection timed out))
- [12:45:05] * docelic_ is now known as docelic
- [12:53:40] * bazbell (n=a0192809@nat/ti/x-75a4ec9496017afb) has joined #beagle
- [12:56:51] * mugenkun (n=sddd@201.37.105.27) has joined #beagle
- [13:00:32] * prpplague^2 (n=dave@mail.americanmicrosystems.com) has joined #beagle
- [13:02:49] * prpplague^2 is now known as prpplague
- [13:04:47] * dannyBlue|DDd (n=chatzill@89.181.68.180) Quit ("I was here.")
- [13:04:51] <sakoman_>
koen: what fixed the serial lockup?
- [13:17:58] * dschaeffer (n=daniel@timesys-gw0.cust.expedient.net) has joined #beagle
- [13:22:09] <DJWillis>
jkridner: ping
- [13:22:13] <jkridner>
pong
- [13:22:23] <DJWillis>
mind if I PM you?
- [13:22:35] <jkridner>
don't ask.
- [13:22:54] <Crofton>
otherwise we know you are up to something :)
- [13:22:58] <jkridner>
(just do... I'll ignore if inappropriate.)
- [13:23:15] <jkridner>
hehe... just creates more noise.
- [13:25:14] <DJWillis>
True, it's just far enough off topic to not interest Beagle guys ;-)
- [13:26:37] * mickey|zzZZzz is now known as mickeyl
- [13:45:47] <ciskje>
Hi guys
- [13:47:44] * vsriram (i=7aa60de8@gateway/web/ajax/mibbit.com/x-a82db680912b6111) has joined #beagle
- [14:14:57] * methril (n=Methril@213.27.233.98) has joined #beagle
- [14:15:01] <methril>
hi
- [14:17:18] <Crofton>
koen, I suspect my nfs issue is related to util-linux-ng versus util-linux
- [14:21:18] * JoeBorn (n=jborn@conference/oscon/x-c6384fc039757989) has joined #beagle
- [14:22:25] * methril is now known as methril|away
- [14:59:26] * Viral_Sachde (n=Viral_Sa@122.167.147.142) Quit ("Leaving.")
- [15:10:13] * vsriram (i=7aa60de8@gateway/web/ajax/mibbit.com/x-a82db680912b6111) Quit ("http://www.mibbit.com ajax IRC Client")
- [15:11:17] * JuanG (n=Juan@nat/ti/x-771e02fa37cf8797) has joined #beagle
- [15:11:46] * JuanG (n=Juan@nat/ti/x-771e02fa37cf8797) has left #beagle
- [15:14:49] * trickie (n=trickie@basesoft.xs4all.nl) Quit ("Leaving")
- [15:19:03] * JoeBorn (n=jborn@conference/oscon/x-c6384fc039757989) Quit ("open.neurostechnology.com")
- [15:20:51] * JoeBorn_ (n=jborn@conference/oscon/x-92b9d1aaafc34c51) has joined #beagle
- [15:21:38] * JoeBorn_ is now known as JoeBorn
- [15:21:38] * JoeBorn (n=jborn@conference/oscon/x-92b9d1aaafc34c51) Quit (Killed by douglas.freenode.net (Nick collision))
- [15:22:03] * JoeBorn (n=jborn@conference/oscon/x-0034ff8b10decb15) has joined #beagle
- [15:23:17] * royerfa (n=x0091539@nat/ti/x-212e26557e8b7d5f) has left #beagle
- [15:33:43] * dirk2 (n=dirk@F3154.f.strato-dslnet.de) has joined #beagle
- [15:34:34] <koen>
sakoman_: false alarm, it took 5 hours to lock up instead of 5 minutes
- [15:42:11] * rsalveti (n=salveti@200.184.118.132) Quit (Read error: 113 (No route to host))
- [15:47:00] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
- [16:18:05] * ciskje (n=francesc@131.114.33.209) Quit (Remote closed the connection)
- [16:26:53] * jkridner|work (n=a0321898@nat/ti/x-2c24ae56d60a14d7) has joined #beagle
- [16:29:51] <jkridner|work>
koen: :(
- [16:31:29] <koen>
jkridner|work: ?
- [16:31:34] * mickeyl is now known as mickey|sports
- [16:31:49] <jkridner|work>
serial lock-up after 5 hours.
- [16:32:15] <koen>
it used to be 5 minutes
- [16:32:16] <jkridner|work>
to try to help with wringing out any issues, I've setup a page to register beagle boards.
- [16:32:23] <koen>
it did get 60 times better
- [16:32:25] <jkridner|work>
it is totally optional.
- [16:32:39] <jkridner|work>
http://beagleboard.org/support/board
- [16:33:07] <jkridner|work>
just gives us one place to put information in.
- [16:33:50] <koen>
I finally got vlc to compile up to the point where it starts needing CE headers
- [16:34:07] <koen>
so I can look at that tomorrow
- [16:34:15] <koen>
and I think I have mythtv building as well
- [16:34:24] <denix>
jkridner|work: what the one in your backpack doing? is it battery powered? :)
- [16:35:29] <denix>
koen: you got myth building - this is cool! have you tried it out yet?
- [16:35:40] <koen>
it's still building
- [16:35:48] <koen>
c++ takes some serious compile time
- [16:36:22] <denix>
yeah, that's true. need any help testing it? :)
- [16:36:40] <koen>
when it's done building :)
- [16:36:51] <denix>
btw, didn't know Janne was also interested in arm/omap... paving the road for low-cost myth frontends? :)
- [16:36:54] <koen>
it's building it for armv5 atm, will try an armv7 build later
- [16:37:09] <koen>
qt will need patching for gcc 4.3
- [16:38:45] <denix>
koen: we ran some benchmark tests here and difference between armv5 and armv7 binaries is not that big...
- [16:40:08] <denix>
performance-wise
- [16:40:47] <denix>
used cs 2007q3 - results were quite surprising...
- [16:41:01] <janneg>
denix: not really. I doubt I would use my A780 for running mythtv on it. but cross-compiling should work
- [16:41:16] <koen>
denix: not having NEON hurts with ffmeg
- [16:41:24] <bjdooks>
there's not an lot of improvement for the general case, the only benefits, iirc are the extra NEON type stuff
- [16:42:54] <denix>
janneg: ha! :) what about omap3? should be good as a frontend...
- [16:44:16] <koen>
janneg: you a780 has a nice overlay in sram and iwmmxt
- [16:44:36] <koen>
janneg: not powerfull enough to play SD video, thoughl
- [16:45:43] <janneg>
not that it would really matter on a qvga screen
- [16:46:16] * janneg searches for someone to implement on the fly transcoding in the backend
- [16:47:01] <koen>
janneg: the indt have something like that working for gmyth
- [16:47:21] <koen>
janneg: this patch would be nice for myth svn: http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=blob_plain;f=packages/mythtv/mythtv/ifdef-qt.diff;hb=HEAD
- [16:47:36] <koen>
my qt4 is built without opengl support, so it doesn't install that header
- [16:47:50] * bjdooks was just wondering about doing an Eagle symbol/layout for OMAP as if you can stack the memory on top of the cpu, layout in 4 layers shouldn't be impossible
- [16:48:01] <denix>
koen: have you seen a n800 port of myth?
- [16:48:41] <koen>
denix: only of gmyth
- [16:49:07] <koen>
denix: http://gmyth.sourceforge.net/wiki/index.php/MobileMythtv
- [16:50:05] <denix>
no, someone built myth fe/be for n800, it is a real myth running... let me dig it up
- [16:50:54] <koen>
well, OE has a chinook-compat mode, so if I get myth on the beagle we can run in on the n800 as well
- [16:53:18] <denix>
http://www.internettablettalk.com/forums/showthread.php?t=18116
- [16:53:24] <denix>
http://www.internettablettalk.com/forums/showthread.php?t=18270
- [16:53:47] <janneg>
koen: committed, thanks
- [16:54:14] <denix>
qt4 and qt3 respectively
- [17:03:11] <koen>
janneg: is svn trunk protocol compatible with 0.21?
- [17:05:27] <janneg>
koen: no
- [17:07:02] <koen>
hmmm
- [17:07:17] <jkridner|work>
denix: it is for presentations when I travel. :)
- [17:07:56] <koen>
janneg: any ETA for the .22 release?
- [17:08:05] <jkridner|work>
anyone seen a good remote control that has a USB interface?
- [17:08:32] <denix>
jkridner|work: is OpenID the only provider supported?
- [17:08:41] <koen>
jkridner|work: bluetooth remote + bluetooth dongle?
- [17:08:57] <jkridner|work>
I have a presenter, but it doesn't have the right media controls.
- [17:09:07] <jkridner|work>
BT remote recommendation would be good.
- [17:09:28] <janneg>
koen: no, not before the qt4 port is complete, i.e. all qt3support usage removed
- [17:09:49] <jkridner|work>
especially if it used regular keyboard entries to allow it to integrate with lots of applications.
- [17:09:50] <janneg>
could take a while
- [17:10:17] <jkridner|work>
denix: OpenID isn't a provider. myopenid is just one provider, you can use (almost) any OpenID provider.
- [17:10:21] <janneg>
koen: I'll backport the cross-compile fixes to the 0.21 branch
- [17:10:30] <koen>
janneg: ok, I'll finish the svn recipe and then try to fix the 0.21 release
- [17:10:54] <koen>
all my computers here run myth 0.21
- [17:11:15] * jkridner|work really wants a solar backpack (http://www.voltaicsystems.com/bag_converter.shtml) to keep that 3rd Beagle in. :)
- [17:11:28] <denix>
jkridner|work: yeah, been trying other OpenID providers w/o success... :(
- [17:12:31] <jkridner|work>
which ones.
- [17:12:45] <janneg>
koen: I don't expect much conflicts in the configure part, the other one patch is not needed for qt3
- [17:15:57] <koen>
NOTE: package mythtv-0.21+svnr17891-r0: task do_compile: completed
- [17:15:59] <koen>
yay!
- [17:16:41] <denix>
jkridner|work: finally, signon worked... not sure what was the problem before, when I tried it and also claimid...
- [17:25:45] * RogerMonk (n=a0740758@nat/ti/x-1f98f7d0f3a390ca) has joined #beagle
- [17:32:02] <koen>
RogerMonk: good morning
- [17:35:38] <RogerMonk>
hi koen!
- [17:58:53] * koen looks at mythtv_0.21+svnr17891-r0_armv5te.ipk
- [18:02:13] <RogerMonk>
koen - u r the man!!
- [18:03:57] <koen>
I haven't checked if it actually works :)
- [18:04:05] * Olipro_ is now known as Olipro
- [18:11:59] <jkridner|work>
looks like I need to fix the username parser to understand https.
- [18:15:32] <denix>
didn't mean to give you troubles... :)
- [18:18:41] * rxr (n=rene@85.183.32.143) has joined #beagle
- [18:19:12] <rxr>
hi, can the board already be ordered ?
- [18:21:32] <koen>
jkridner|work: http://www.theregister.co.uk/2008/06/17/cherrypal/
- [18:23:09] <BThompson>
that cherrypal is a recent topic here :) there is an article on tom's hardware for it as well
- [18:28:11] <suihkulokki>
hmh, it doesn't look very different from efika which has been available for quite a while?
- [18:30:51] * mugenkun (n=sddd@201.37.105.27) Quit (Remote closed the connection)
- [18:32:48] <mru>
hi all
- [18:34:04] * rxr (n=rene@85.183.32.143) has left #beagle
- [18:41:43] <koen>
hey mru
- [18:43:03] <koen>
jkridner|work: when's the next beagle demo?
- [18:57:08] * dannyBlue|DDd (n=chatzill@89.181.68.180) has joined #beagle
- [19:01:41] <koen>
RogerMonk: I'll poke you tomorrow about how to rebuild various CE .a files (e.g. osal_dsplink_linux.a) to get the neuros VLC port to build
- [19:09:42] * dirk2 (n=dirk@F3154.f.strato-dslnet.de) has left #beagle
- [19:12:26] * JoeBorn (n=jborn@conference/oscon/x-0034ff8b10decb15) Quit (Read error: 110 (Connection timed out))
- [19:24:06] * docelic (n=docelic@78.134.198.171) Quit ("http://www.spinlocksolutions.com/")
- [19:40:39] * Thanatos (n=rwvens@a83-132-76-156.cpe.netcabo.pt) has joined #beagle
- [20:01:57] * Thanatos is now known as Thanatos_Wibrain
- [20:10:15] * ciskje (n=francesc@host99-35-dynamic.11-87-r.retail.telecomitalia.it) has joined #beagle
- [20:11:45] <Crofton|work>
koen, opkg -force-overwrite install util-linux-mount solved my nfs mount problem
- [20:13:17] <koen>
weird
- [20:13:23] <Crofton>
yeah
- [20:13:28] <koen>
util-linux-ng is supposed to be the non-buggy one :(
- [20:13:35] <Crofton>
the mount from -ng looks for mount.nfs
- [20:13:36] <Crofton>
heh
- [20:14:00] <Crofton>
I went through the sources, and can't find where mount.nfs[4] is supposed to come from
- [20:14:17] <koen>
nfsutils?
- [20:14:21] <Crofton>
not sure
- [20:14:27] <Crofton>
I went through things yesterday
- [20:14:40] <Crofton>
I suppose I'll add to bugzilla
- [20:14:59] <Crofton>
[balister@localhost packages]$ rpm -qf /sbin/mount.nfs
- [20:14:59] <Crofton>
nfs-utils-1.1.0-4.fc7
- [20:16:06] <koen>
debian also has it in nfsutils
- [20:16:45] <koen>
the nfsutils in OE probably needs to get updated to 1.1.x
- [20:16:50] <Crofton>
yeah
- [20:17:15] <Crofton>
I suspect there are updates for util-linux-ng
- [20:18:26] <koen>
I poked at -ng last month, so I wouldn't hold your breath :)
- [20:18:52] <Crofton>
hmmm, maybe it was nfs-utils then
- [20:19:45] * koen wonders why it is so hard to get decent oatmeal-raisin cookies in .nl
- [20:23:14] * JoeBorn (n=jborn@12.157.240.3) has joined #beagle
- [20:23:56] <Crofton|work>
crap, I'm hanging after uncompressing the kernel a lot
- [20:24:43] <koen>
Crofton|work: powercycling seems to work
- [20:24:51] <Crofton|work>
several times ......
- [20:25:29] <koen>
yeah, I get the problem as well
- [20:25:37] <koen>
cold-booting seems to 'fix' it for me
- [20:25:58] <Crofton|work>
I am hopeful I can get "make check" to run for gnu radio
- [20:26:11] <Crofton|work>
that should provide feedback on the code generated
- [20:26:21] <Crofton|work>
root
- [20:27:48] <koen>
btw: http://amethyst.openembedded.net/~koen/angstrom-2008.1-test-20080722-armv7a-linux-gnueabi-toolchain.tar.bz2
- [20:35:47] <Crofton|work>
http://www.pastebin.ca/1079903
- [20:36:18] <koen>
hmmm
- [20:36:28] <Crofton>
swig runtime I think
- [20:36:34] <koen>
pastebin.ca is using avian carriers it seems
- [20:36:48] <koen>
../../../../arm-angstrom-linux-gnueabi-libtool: eval: line 2: ccache: not found
- [20:38:01] <Crofton>
basically, I think something is missing, so it is trying to build it
- [20:39:02] <koen>
it's trying to run the same stuff as during the build
- [20:39:12] <koen>
so 'ccache arm-foo-gcc'
- [20:39:41] <Crofton>
yeah
- [20:39:54] <Crofton>
but make check shouldn't need to build anything
- [20:40:08] <koen>
not even the check stuff?
- [20:40:23] <Crofton>
It should have built all the tests
- [20:40:29] <Crofton>
make check runs the tests
- [20:40:37] <Crofton>
at least this is my understanding
- [20:40:47] * koen is a make test n00b
- [20:40:47] <Crofton>
time to review the notes for doing this on the PS2
- [20:40:51] <Crofton>
er PS3
- [20:41:06] <Crofton>
I think it just runs various cppunit tests
- [20:41:15] <Crofton>
or one of those unit tets pacakages
- [20:42:41] <Thanatos_Wibrain>
hey #beagle, does the Beagleboard support bluetooth by default?
- [20:42:53] <koen>
Thanatos_Wibrain: if you have a usb bluetooth dongle
- [20:43:03] * cian (n=cianh@cian.ws) has joined #beagle
- [20:43:22] <Thanatos_Wibrain>
hmm, is a usb hub possible?
- [20:43:28] <Crofton|work>
yes :)
- [20:43:35] <cian>
Has anyone here outside the US ordered a board from Digikey yet?
- [20:44:06] <Thanatos_Wibrain>
annd, is there a good way for the board to receive raw video data?
- [20:44:10] * JoeBorn (n=jborn@12.157.240.3) Quit ("open.neurostechnology.com")
- [20:44:20] <Thanatos_Wibrain>
usb webcams are too consumer oriented
- [20:44:51] <koen>
Thanatos_Wibrain: i2c, gpio or usb ar the option AFAIK
- [20:45:02] <bjdooks>
cian: i'm trying to get my boss to do so
- [20:45:08] <Thanatos_Wibrain>
gpio?
- [20:45:16] <Crofton>
I would guess the EVM supports better cameras ....
- [20:45:47] <cian>
bjdooks: ok, apparently Digikey won't sell them outside the US...
- [20:46:56] <Crofton|work>
hmm, I can't even find an entry on digikey ....
- [20:47:34] <koen>
only evm and mdk are on digikey, right?
- [20:47:41] <koen>
beagle isn't for sale yet
- [20:49:23] <cian>
ok, nm, thanks
- [20:57:53] <Thanatos_Wibrain>
would the Beagleboard accept input from a high speed CMOS sensor?
- [20:59:11] <Thanatos_Wibrain>
*with an 8 bit RGB serial output
- [21:03:03] <koen>
RogerMonk: http://www.angstrom-distribution.org/repo/?action=details&pnm=mythtv-frontend
- [21:03:43] <RogerMonk>
built?
- [21:04:19] <koen>
yes
- [21:04:28] <RogerMonk>
*SWEEET!*
- [21:04:34] <RogerMonk>
does it run?
- [21:04:38] <koen>
it is svn trunk, so it's using qt4 and a new version of the protocol
- [21:04:42] <koen>
haven't tried to run it yet
- [21:05:00] <RogerMonk>
(fingers crossed for you!)
- [21:06:37] <Thanatos_Wibrain>
..?
- [21:08:27] <denix>
koen: heh, I'd love to try it out, but all my be/fe run 0.21-fixes... :(
- [21:09:32] * denix needs to build qt4-based svn trunk for testing/debugging...
- [21:10:15] <koen>
denix: I'' try to have a look at building the .21 branch tomorrow
- [21:10:29] <koen>
depends on how fast we get codec-engine in shape :)
- [21:10:49] <RogerMonk>
koen - ok, ok...
- [21:15:05] * koen heads toward bed
- [21:15:06] <koen>
later all
- [21:15:24] <denix>
good night
- [21:15:36] <RogerMonk>
koen - later sir - another big day of progress from you!
- [21:16:25] <Thanatos_Wibrain>
would the beagleboard accept serial output from a CMOS image sensor?
- [21:17:51] <RogerMonk>
beagleboard doesn't have OMAP sensor interface pinned out to a header
- [21:18:20] * RogerMonk heads out
- [21:18:30] <ds2>
depends on if you can convince the McBSP to do it
- [21:18:39] <ds2>
Thanatos_WiBrain: are you really on a WiBrain?
- [21:18:42] <denix>
looks like you need EVM instead...
- [21:18:51] <Thanatos_Wibrain>
and yes im on a wibrain
- [21:19:02] <Thanatos_Wibrain>
mcBSP?
- [21:19:43] <RogerMonk>
Thanatos_Wibrain - pls send link to sensor data sheet - someone may have some ideas
- [21:20:37] * Olipro_ (i=Olipro@unaffiliated/olipro) has joined #beagle
- [21:20:59] <ds2>
which WiBrain do you have?
- [21:21:15] <Thanatos_Wibrain>
uhh, take this one for example http://www.aptina.com/products/image_sensors/mt9m413c36stc/#overview
- [21:21:34] <Thanatos_Wibrain>
i have the B1L which i put windows on
- [21:21:48] * Olipro (i=Olipro@unaffiliated/olipro) Quit (Nick collision from services.)
- [21:21:50] * Olipro_ is now known as Olipro
- [21:22:01] <Thanatos_Wibrain>
think thats the one...
- [21:22:06] <ds2>
but to be strictly legel, the B1L + Windows is more expensive then a B1 with Linux put in afterwards
- [21:22:30] <Thanatos_Wibrain>
i had xp already
- [21:23:01] <ds2>
I see. How much did the lack of bluetooth on the B1L mattered?
- [21:23:18] <ds2>
<-- wants to get a WiBrain
- [21:24:20] <Thanatos_Wibrain>
didnt really need it myself
- [21:24:32] <Thanatos_Wibrain>
it has everything on it already
- [21:24:42] <ds2>
Thanatos_WiBrain: are you at home, at a cafe or ?
- [21:24:56] <Thanatos_Wibrain>
the touchscreen and touchpad are all i need for a mouse
- [21:25:10] <Thanatos_Wibrain>
im sitting on mu patio in a lawn chair ATM
- [21:25:34] <ds2>
so WiFi networking from your own setup?
- [21:25:41] <Thanatos_Wibrain>
yeah
- [21:25:48] <ds2>
I think I might want to be able to do GPRS/EDGE/etc from a phone
- [21:26:18] <Thanatos_Wibrain>
i was thinking of getting a USB GPRS flr on the go
- [21:27:23] * mickey|sports is now known as mickeyl
- [21:30:22] <Thanatos_Wibrain>
so anyone know much concerning the CMOS thing?
- [21:33:06] * BThompson (n=BThompso@nat/ti/x-92ec3ba18589b007) Quit ("Trillian (http://www.ceruleanstudios.com")
- [21:34:04] <Thanatos_Wibrain>
which one im getting if flexible though, i want it to be compatible with the Beagleboard
- [21:57:08] * mickeyl is now known as mickey|zzZZzz
- [21:57:27] <Crofton|work>
ouch, sorry about the stock ti guys ....
- [22:13:01] * Thanatos_Wibrain is now known as Thanatos
- [22:14:25] * Thanatos (n=rwvens@a83-132-76-156.cpe.netcabo.pt) Quit ()
- [22:17:12] * Thanatos (n=rwvens@a83-132-76-156.cpe.netcabo.pt) has joined #beagle
- [22:27:29] * bazbell (n=a0192809@nat/ti/x-75a4ec9496017afb) Quit (Remote closed the connection)
- [22:30:30] * Thanatos is now known as Thanatos_Wibrain
- [22:30:41] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
- [22:36:57] * dannyBlue|DDd (n=chatzill@89.181.68.180) Quit ("I was here.")
- [22:43:31] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
- [22:58:13] * ciskje (n=francesc@host99-35-dynamic.11-87-r.retail.telecomitalia.it) Quit (Remote closed the connection)
- [23:09:38] <Crofton|work>
export controls, don't you love them ....
- [23:17:08] * orcas (n=orcas@udp011879uds.ucsf.edu) Quit (brown.freenode.net irc.freenode.net)
- [23:17:08] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) Quit (brown.freenode.net irc.freenode.net)
- [23:17:08] * ldesnogu (n=ldesnogu@fw-tnat.cambridge.arm.com) Quit (brown.freenode.net irc.freenode.net)
- [23:17:08] * jkridner|work (n=a0321898@nat/ti/x-2c24ae56d60a14d7) Quit (brown.freenode.net irc.freenode.net)
- [23:17:08] * oelewapperke (i=a619fded@85-158-215-1.powered-by.benesol.be) Quit (brown.freenode.net irc.freenode.net)
- [23:17:08] * Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) Quit (brown.freenode.net irc.freenode.net)
- [23:17:08] * Crofton|work (n=balister@66-207-66-26.black.dmt.ntelos.net) Quit (brown.freenode.net irc.freenode.net)
- [23:17:29] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
- [23:17:29] * jkridner|work (n=a0321898@nat/ti/x-2c24ae56d60a14d7) has joined #beagle
- [23:17:29] * oelewapperke (i=a619fded@85-158-215-1.powered-by.benesol.be) has joined #beagle
- [23:17:29] * Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) has joined #beagle
- [23:17:29] * Crofton|work (n=balister@66-207-66-26.black.dmt.ntelos.net) has joined #beagle
- [23:17:29] * ldesnogu (n=ldesnogu@fw-tnat.cambridge.arm.com) has joined #beagle
- [23:17:29] * orcas (n=orcas@udp011879uds.ucsf.edu) has joined #beagle
- [23:19:04] <jkridner|work>
yeah, rough day.
- [23:26:57] <ds2>
has anyone tried using the expansion connector signals?
- [23:27:19] <jkridner|work>
not that I'm aware.
- [23:27:46] <jkridner|work>
g'night all.
- [23:27:53] * jkridner|work (n=a0321898@nat/ti/x-2c24ae56d60a14d7) has left #beagle
- [23:33:08] <denix>
night
- [23:35:10] * Olipro (i=Olipro@unaffiliated/olipro) Quit (No route to host)
- [23:36:15] <Thanatos_Wibrain>
uhh is he coming back?
- [23:36:28] <Thanatos_Wibrain>
like on his home account?
- [23:36:47] <ds2>
heh
- [23:36:50] <BThompson>
he may be, often is
- [23:37:21] <Thanatos_Wibrain>
good, he seems to be my best help here
- [23:37:24] <Thanatos_Wibrain>
lol
- [23:37:31] <ds2>
what help do you need? =)
- [23:38:07] <Thanatos_Wibrain>
i meant to bug him about the CMOS's 8 bit serial output's compatibility with the Beagleboard
- [23:38:33] <ds2>
the standard OMAP CMOS sensor interface is not available on the beagle
- [23:38:45] <ds2>
so you'll need to make due with other stuff
- [23:38:59] <BThompson>
yea not pinned out, normally usb based cameras are suggested
- [23:39:25] <ds2>
yep. it was due to cost reasons; bringing out all the pins means more layers and the more layers on the PCB, the higher the cost
- [23:39:50] <Thanatos_Wibrain>
hmm, usb cams are heavilly consumer based, might not be as smooth as it needs to be
- [23:40:17] <ds2>
sometimes you can hack around it... SPI/I2C/McBSP/UART are all available... maybe with some GPIO and bit banging you could do it
- [23:40:31] <Thanatos_Wibrain>
ppl are gonna see the cam's output and NOTHING else, like night vision
- [23:40:48] <Thanatos_Wibrain>
usb cams typically dont offer such speed
- [23:40:57] <ds2>
what do you want to do with the sensor?
- [23:41:46] <BThompson>
for beagle you are somewhat limited, if it is OMAP3 you are after than the Mistral EVM has the interface pinned out, but of course that is larger, costlier, and would still require some custom adapter hardware
- [23:42:10] <Thanatos_Wibrain>
output from a ioq3 based renderer is overlayed onto real life for augmented reality
- [23:42:21] <Thanatos_Wibrain>
thats what i wanted to do with it
- [23:42:42] <ds2>
so you don't need high rez
- [23:42:46] <Thanatos_Wibrain>
nope
- [23:42:54] <ds2>
why not a USB NTSC capture stick and a NTSC output camera?
- [23:43:12] <Thanatos_Wibrain>
capture stick?
- [23:43:25] <ds2>
sure... most NTSC TV "sticks" have an AUX in
- [23:44:06] <BThompson>
capture card that is usb based, hauppage makes some and there are others
- [23:44:23] <Thanatos_Wibrain>
how bulky are those and how small do NTSC cams get
- [23:44:39] <Thanatos_Wibrain>
it WILL be mounted on someone's head ya know
- [23:44:51] <BThompson>
as small or smaller than a usb cam
- [23:44:58] <ds2>
http://www.allelectronics.com/make-a-store/item/VC-33/MINIATURE-COLOR-PINHOLE-CAMERA/1.html
- [23:45:16] <ds2>
that's an example NTSC camera... the USB part is around the size of a thumb drive
- [23:45:28] <ds2>
you'll most likely need a hub for power
- [23:45:56] <Thanatos_Wibrain>
hey thats pretty nice
- [23:46:09] <Thanatos_Wibrain>
its 9v just like the beagle rite?
- [23:46:29] <ds2>
the beagle is 5V, AFAIK
- [23:46:34] <Thanatos_Wibrain>
crap
- [23:46:41] <Thanatos_Wibrain>
not hard to fix
- [23:46:43] <Thanatos_Wibrain>
lol
- [23:46:56] <ds2>
I believe the camera is really 5V or 6V
- [23:47:41] <Thanatos_Wibrain>
res isnt particularly impressive
- [23:47:47] <ds2>
it is NTSC
- [23:47:54] <ds2>
assume 320x200
- [23:48:01] <Thanatos_Wibrain>
ouch
- [23:48:25] <Thanatos_Wibrain>
the HMD will be about 1024x768
- [23:48:35] <Thanatos_Wibrain>
so it would look like crap
- [23:49:20] * cian (n=cianh@cian.ws) Quit ()
- [23:49:21] <Thanatos_Wibrain>
imagine seeing the world thru 32px200
- [23:49:34] <Thanatos_Wibrain>
*320x200
- [23:53:46] <Thanatos_Wibrain>
ill ask Aptina if i can get This http://www.aptina.com/products/wafer-level_cameras/mt9v113m02stc/#overview for USB
- [23:53:50] <Thanatos_Wibrain>
hah
These logs were automatically created by BeagleLogBot on
irc.freenode.net
using the Java IRC LogBot.
See http://microformats.org/wiki/mflogbot for more information.