• [00:02:48] * florian (n=fuchs@e179107064.adsl.alicedsl.de) Quit (Read error: 113 (No route to host))
  • [00:05:01] * kevinsc_ (n=kevinsc@pool-173-74-177-131.dllstx.fios.verizon.net) Quit (Read error: 110 (Connection timed out))
  • [00:12:05] * mib_xpumcw (i=47091882@gateway/web/ajax/mibbit.com/x-ee05e9e904d9bdda) has joined #beagle
  • [00:14:16] * fulgas is now known as FuL|OUT
  • [00:16:24] * bavison (n=ben@cpc2-cmbg9-0-0-cust635.cmbg.cable.ntl.com) Quit (Read error: 110 (Connection timed out))
  • [00:18:02] * mib_gkus70 (i=4ba892b5@gateway/web/ajax/mibbit.com/x-53b1b454d5864aec) has joined #beagle
  • [00:18:11] * mib_gkus70 (i=4ba892b5@gateway/web/ajax/mibbit.com/x-53b1b454d5864aec) Quit (Client Quit)
  • [00:31:55] <tomasg> fwiw I now have dss2 working with the omap3evm. turns out *both* panel-generic and panel-sharp... need to be loaded or the omapfb driver can't init with the sharp panel
  • [00:31:56] * midtown4 (n=krooked@68-189-104-94.dhcp.wtvl.ca.charter.com) Quit (Read error: 54 (Connection reset by peer))
  • [00:34:20] * midtown4 (n=krooked@68-189-104-94.dhcp.wtvl.ca.charter.com) has joined #beagle
  • [00:37:36] * midtown4 (n=krooked@68-189-104-94.dhcp.wtvl.ca.charter.com) Quit (Read error: 104 (Connection reset by peer))
  • [00:37:47] * midtown4 (n=krooked@68-189-104-94.dhcp.wtvl.ca.charter.com) has joined #beagle
  • [00:38:07] * Downix (n=downix@cpe-72-184-210-102.tampabay.res.rr.com) has joined #beagle
  • [00:50:57] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [00:59:37] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [01:01:01] * mib_hua54l (i=049a66d4@gateway/web/ajax/mibbit.com/x-daa6478325746b6e) has joined #beagle
  • [01:04:56] * mib_hua54l (i=049a66d4@gateway/web/ajax/mibbit.com/x-daa6478325746b6e) Quit (Client Quit)
  • [01:08:33] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [01:11:54] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit (Client Quit)
  • [01:16:54] * MrBIOS-_ (n=aperez@h65.149.51.74.cable.snbr.sanbrunocable.com) has joined #beagle
  • [01:18:21] <GregorR> tomba: By the way, did you get my message a few days ago re getting interlacing working properly? I just swapped the overlays and it worked (no shock), is this a PAL-vs-NTSC problem?
  • [01:18:52] <GregorR> (I had to send in my Beagle for repairs :( )
  • [01:25:36] * NishanthMenon (n=nmenon@nat/ti/x-a58fb81f0f3be7d7) Quit ("Leaving")
  • [01:27:53] * tharvey__ (n=tharvey@adsl-76-205-222-173.dsl.snlo01.sbcglobal.net) Quit (Read error: 110 (Connection timed out))
  • [01:31:48] * mib_xpumcw (i=47091882@gateway/web/ajax/mibbit.com/x-ee05e9e904d9bdda) Quit ("http://www.mibbit.com ajax IRC Client")
  • [01:32:30] * mib_xpumcw (i=47091882@gateway/web/ajax/mibbit.com/x-a9e07478a1760cf3) has joined #beagle
  • [01:39:50] * mib_xpumcw (i=47091882@gateway/web/ajax/mibbit.com/x-a9e07478a1760cf3) Quit ("http://www.mibbit.com ajax IRC Client")
  • [01:43:44] <tomasg> hello all. I've been reviewing recent threads on musb/OTG in 2.6.29. Looks like there is work to be done to get OTG stable again. Who is championing that development for beagle/omap3evm?
  • [01:49:48] * tharvey (n=tharvey@adsl-76-205-222-173.dsl.snlo01.sbcglobal.net) has joined #beagle
  • [01:50:53] * ScriptRipper_ (n=mmohring@p4FDBCBCC.dip.t-dialin.net) has joined #beagle
  • [01:56:23] * emeb_mac (n=ericb@ip72-223-90-212.ph.ph.cox.net) Quit ()
  • [01:57:02] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has left #beagle
  • [02:02:00] * emeb_mac (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [02:08:19] <MrBIOS-_> ds / ds2 ping
  • [02:08:37] <ds> ds != ds2
  • [02:08:53] <solar> Can the beagle boot off a usb storage device? I appear to have a bad sd/mmc slot
  • [02:09:40] <MrBIOS-_> ds: one can imagine why I'd draw that conclusion, however
  • [02:09:53] <ds> heh
  • [02:10:57] * ScriptRipper (n=mmohring@opensuse/member/MartinMohring) Quit (Read error: 110 (Connection timed out))
  • [02:12:02] <russ> and dss2 != ds2
  • [02:12:30] <russ> tomasg: david brownell I think
  • [02:13:09] * brolin (n=brolin@190.157.14.3) has joined #beagle
  • [02:19:09] * lcuk (i=lcuk@cpc3-oldh7-0-0-cust590.manc.cable.ntl.com) Quit ("Leaving")
  • [02:20:45] <russ> hmmm...I'm trying to disable the 26MHz osc by tying the ENABLE pin to VPLL1 instead of VIO and turning off VPLL1 during off mode
  • [02:21:03] <russ> but it looks like the pin has an internal pullup, so voltage is draining from VIO to VPLL1
  • [02:21:54] <russ> (I think)
  • [02:22:56] <tomasg> russ: thanx!
  • [02:23:10] <russ> tomasg: just skim the linux-usb list
  • [02:23:42] * BThompsonD (n=bernie@cpe-72-190-75-99.tx.res.rr.com) Quit (Remote closed the connection)
  • [02:24:43] * mib_s3kavp (i=60fc7a40@gateway/web/ajax/mibbit.com/x-421b9e5dfaf3b755) has joined #beagle
  • [02:24:48] <tomasg> it actually seems like i may have a working otg w/ 2.6.29. The Kconfigs are just a mess so various deps wern't being resolved..
  • [02:26:30] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [02:31:19] <russ> whats a good rework temp?
  • [02:32:54] <mib_s3kavp> @russ: Are you asking for a rework station?
  • [02:33:25] <russ> I have a weller temperature controlled iron
  • [02:35:33] <mib_s3kavp> I dont have the temps in mind. As long as the solder starts melting, I take it as good temp. You will notice if the tip is too hot if the solder forms a blob sticks around the tip. its difficult to explain
  • [02:39:10] * fenn_ (n=fenn@cpe-72-177-52-180.austin.res.rr.com) has joined #beagle
  • [02:39:26] * fenn (n=fenn@cpe-72-177-52-180.austin.res.rr.com) Quit (Nick collision from services.)
  • [02:40:57] * fenn_ is now known as fenn
  • [02:44:21] <russ> hmmm...I'm starting to think Y1 is glued to the board
  • [02:51:47] <russ> :( Its like I need two irons
  • [02:53:53] * timchen1` is now known as nasloc__
  • [02:55:13] * mib_s3kavp (i=60fc7a40@gateway/web/ajax/mibbit.com/x-421b9e5dfaf3b755) Quit ("http://www.mibbit.com ajax IRC Client")
  • [03:01:07] * openfree (i=cb6ea385@gateway/web/ajax/mibbit.com/x-eacedc65b88ed44f) has joined #beagle
  • [03:02:12] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [03:02:43] <russ> ok, replaced my clock
  • [03:05:18] * Abraxas3D (n=Abraxas3@76.194.190.126) has joined #beagle
  • [03:06:14] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [03:09:47] * mib_xpumcw (i=47091882@gateway/web/ajax/mibbit.com/x-ef053bf4e2197b70) has joined #beagle
  • [03:14:17] * mib_xpumcw (i=47091882@gateway/web/ajax/mibbit.com/x-ef053bf4e2197b70) Quit (Client Quit)
  • [03:17:00] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) Quit ()
  • [03:28:50] * Abraxas3D (n=Abraxas3@76.194.190.126) has left #beagle
  • [03:36:58] <russ> hmmm...cutting power to the rs232 level transceivers causes a wakeup
  • [03:44:58] * Wowbagger_ (n=wowbagge@d154-20-156-61.bchsia.telus.net) has joined #beagle
  • [03:45:50] * mib_5u6cx7 (i=43bbb3ea@gateway/web/ajax/mibbit.com/x-d4d8586569ae30c6) has joined #beagle
  • [03:46:28] * mrc3_ (n=mrc3@nat/ti/x-569513360d29e60a) Quit (Remote closed the connection)
  • [03:46:39] * mrc3_ (n=mrc3@nat/ti/x-3e5fe8aa54b7bb54) has joined #beagle
  • [03:49:55] * mib_e9mk8i (i=555dcd66@gateway/web/ajax/mibbit.com/x-9520d24e7d87b1a9) has joined #beagle
  • [03:54:50] * ali_as (n=ali_as@ambix.plus.com) Quit (Read error: 60 (Operation timed out))
  • [03:55:47] * favor (n=user@221.10.25.6) has joined #beagle
  • [03:56:15] * vishnupsn (i=ca58ea15@gateway/web/ajax/mibbit.com/x-b7fdf56e77bee0eb) has joined #beagle
  • [03:56:54] * ali_as (n=ali_as@ambix.plus.com) has joined #beagle
  • [03:57:24] <favor> hi. can you give me some help. I am strugge with compiler error. arm-none-linux-gnueabi-gcc -c -MD -pipe -nostdinc -isystem /home/user/myoe/tmp/cross/armv7a/lib/gcc/arm-angstrom-linux-gnueabi/4.3.1/include -fomit-frame-pointer -Wall -Wno-trigraphs -Werror-implicit-function-declaration -fno-strict-aliasing -mapcs -mno-sched-prolog -mabi=aapcs-linux -mno-thumb-interwork -march=armv7a -msoft-float -Uarm -D"KBUILD_STR(s)=#s" -D"KBUILD_BASENAME=KBUILD_STR(libbri
  • [03:57:24] <favor> dge)" -D"KBUILD_MODNAME=KBUILD_STR(libbridge)" -DMODULE -D__LINUX_ARM_ARCH__=7 -fno-common -DLINUX -DGT_TRACE -DDEBUG -DOMAP_3430 -fpic -I. -I/home/user/dspbridge/mpu_driver/inc -I/home/user/dspbridge/mpu_driver/inc/linux -include /home/user/felipec-linux-omap/include/linux/autoconf.h -o .obj/DSPManager.o DSPManager.c DSPManager.c:1: error: bad value (armv7a) for -march= switch
  • [03:57:25] <favor> make[1]: *** [.obj/DSPManager.o] Error 1
  • [03:57:27] <favor> make[1]: Leaving directory `/home/user/dspbridge/mpu_api/src/bridge' make: *** [bridge] Error 2
  • [04:00:33] * mib_14scna (i=ad1142b9@gateway/web/ajax/mibbit.com/x-d7e1059a0767e073) has joined #beagle
  • [04:02:30] * midtown4 (n=krooked@68-189-104-94.dhcp.wtvl.ca.charter.com) has left #beagle
  • [04:03:37] * kona (i=96fa724b@gateway/web/ajax/mibbit.com/x-c1b81e610946cedc) has joined #beagle
  • [04:05:21] <kona> hey, anyone know of a chip that will convert the 1.8V pins to 5 volt signals, i need to run a 5V UART from the chip
  • [04:05:47] <vishnupsn> hi.. can u help me. i am trying to connect to beagle board using Code Composer Studio 3.3 with spectrum digital XDS560 USB emulator. But it is showing some error and i am not able to connect to the beagle board. Can anyone help to solve this problem? what may be the possible reasons why i am not able to connect to the board?
  • [04:12:05] <russ> kona: look at the beagle schematics
  • [04:12:17] <russ> kona: they convect to 3.3V, then to rs232 levels
  • [04:13:50] <mib_5u6cx7> hi, can anyone tell me how to change the display resolutions?
  • [04:17:57] * mib_14scna (i=ad1142b9@gateway/web/ajax/mibbit.com/x-d7e1059a0767e073) Quit ("http://www.mibbit.com ajax IRC Client")
  • [04:19:27] * kona (i=96fa724b@gateway/web/ajax/mibbit.com/x-c1b81e610946cedc) Quit ("http://www.mibbit.com ajax IRC Client")
  • [04:20:36] * mib_e9mk8i (i=555dcd66@gateway/web/ajax/mibbit.com/x-9520d24e7d87b1a9) Quit ("http://www.mibbit.com ajax IRC Client")
  • [04:26:29] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) Quit (Read error: 54 (Connection reset by peer))
  • [04:26:32] * ZeZu_ (n=null@c-98-227-56-17.hsd1.in.comcast.net) has joined #beagle
  • [04:37:06] * mib_5u6cx7 (i=43bbb3ea@gateway/web/ajax/mibbit.com/x-d4d8586569ae30c6) Quit ("http://www.mibbit.com ajax IRC Client")
  • [04:39:13] * shenki (n=joel@202.174.42.5) Quit (Read error: 60 (Operation timed out))
  • [04:49:39] <russ> neato, I'm down to 7.5mW sleep
  • [04:52:35] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [05:02:16] <raster> not too bad at all
  • [05:02:25] <raster> memory in self-refresh?
  • [05:02:30] <raster> and thats the whole beagle?
  • [05:02:40] <russ> raster: thats the whole beagle
  • [05:02:46] <russ> only 1/4 of memory is refreshing
  • [05:03:58] <raster> whats the rest doing?
  • [05:03:59] <raster> off?
  • [05:04:03] <raster> or runing full?
  • [05:04:26] <russ> the rest of memory?
  • [05:04:32] <russ> its not refreshing
  • [05:04:41] <raster> its off?
  • [05:04:54] <raster> (ie u turned off 75% of the ram - has no contents)
  • [05:04:56] <russ> well, there is only one memory chip
  • [05:05:01] <russ> right
  • [05:05:28] <raster> ok
  • [05:05:35] <raster> and the 1/4 is runing full
  • [05:05:45] <raster> or on "suspend" (low power self-refresh) ?
  • [05:05:51] <russ> self refresh
  • [05:05:56] <raster> and is cpu core running or suspended?
  • [05:06:10] <raster> or using ti's power patches for 0hz?
  • [05:06:29] <russ> the cpu is off
  • [05:06:35] <russ> except the SRAM
  • [05:06:36] <raster> totally off?
  • [05:06:41] <raster> or in suspend
  • [05:06:44] <raster> ie it can be woken up?
  • [05:06:47] <russ> VDD1, VDD2, and VPLL are all turned off
  • [05:06:56] <raster> hmm ok
  • [05:07:00] <russ> only VIO is on, so only the wakeup domain is on
  • [05:07:03] <raster> so u'd have to fully boot up
  • [05:07:16] <russ> the wakeup domain can sense a change on a pad and start the wakeup process
  • [05:07:54] <raster> where is most of the 7mw going
  • [05:08:06] <russ> good question
  • [05:08:25] <russ> probably the TWL4030 and the OMAP wakeup domain/SRAM
  • [05:10:19] <raster> hmmm
  • [05:10:23] <raster> thats not too bad
  • [05:10:39] <russ> my paper is on a PASR (partial array self refresh) implementation
  • [05:10:43] <raster> i suspect you would not be too much worse if u used 0hz
  • [05:11:08] <russ> at full refresh, the SDRAM uses 450uA@1.8V, and at 1/8th refresh it uses 200uA@1.8V
  • [05:11:33] * shenki (n=joel@202.174.42.5) has joined #beagle
  • [05:11:34] <russ> so getting sleep power usage down into the range was even noticable was necessary
  • [05:11:49] <russ> er, into the range where that was even noticable
  • [05:12:13] <raster> oh sure
  • [05:12:27] <raster> so thats 1/4 of the sdram
  • [05:12:35] <raster> (32m or 64m? a revb or revc?)
  • [05:12:38] * jsync (n=jess@207.88.137.2.ptr.us.xo.net) has joined #beagle
  • [05:12:39] <russ> revb
  • [05:12:43] <raster> ok so 32m
  • [05:12:54] <raster> and u have it on 1/8th refresh?
  • [05:13:02] <russ> 1/8th is 16M
  • [05:13:09] <raster> ooh
  • [05:13:11] <raster> 18th the ram
  • [05:13:21] <raster> as opposed to 1/8th the refresh cycle freq
  • [05:13:48] <raster> i'm just really curious - what the power consumption of the 35xx is
  • [05:13:53] <raster> with all its ram on
  • [05:13:55] <raster> (128m)
  • [05:13:58] <raster> but on self-refresh
  • [05:13:58] <russ> there is an option to do a 1/16th refresh, but I have to mark all the pages in the non-refreshed regions as movable, so the kernel can't make allocations there
  • [05:14:02] <raster> and the cpu in 0hz mode
  • [05:14:10] <russ> the kernel doesn't like booting with only 8MB
  • [05:14:13] <raster> (that means not off or suspended but clock-stopped with registers in self-refresh)
  • [05:14:37] <raster> basically i'm not keen on suspend/resume :)
  • [05:14:44] <raster> so 0-clock is much more attractive
  • [05:15:52] <russ> I'm not sure, I don't have a config loaded to enter the lower C states
  • [05:16:09] <russ> but I think its in the 90mA @ 5V range
  • [05:16:24] <raster> my understanding is that its is not dissimilar to a full suspend in power usage
  • [05:16:29] <raster> (overall)
  • [05:16:43] <russ> maybe for what other people are doing
  • [05:18:26] <russ> but I'm actually turning off supply rails during suspend
  • [05:26:01] * Wowbagger_ (n=wowbagge@d154-20-156-61.bchsia.telus.net) Quit ("Leaving")
  • [05:27:21] * ogra (n=ogra@ubuntu/member/ogra) Quit (Read error: 104 (Connection reset by peer))
  • [05:28:08] * ogra (n=ogra@ubuntu/member/ogra) has joined #beagle
  • [05:29:49] <favor> hi. can you give me some help. I have a compiler error. DSPManager.c:1: error: bad value (armv7a) for -march= switch, the detail is at http://pastebin.ca/1392495. thnanks .
  • [05:32:31] <kulve> armv7-a?
  • [05:32:59] * shenki (n=joel@202.174.42.5) Quit (Read error: 110 (Connection timed out))
  • [05:37:43] * ZeZu_ is now known as ZeZu
  • [05:38:57] <favor> no, I think the newest arm-angstrom- cross gcc dont support the -march options, don't it?
  • [05:39:06] <favor> doesn't it?
  • [05:47:59] * fenn (n=fenn@cpe-72-177-52-180.austin.res.rr.com) Quit ("this computer has been assimilated")
  • [05:48:46] <favor> any ides?
  • [05:48:48] <favor> any idea?
  • [05:51:17] * emeb_mac (n=ericb@ip72-223-90-212.ph.ph.cox.net) Quit ()
  • [06:03:23] * eFfeM (n=nly91006@neo-y1a.ehv.campus.philips.com) has joined #beagle
  • [06:05:07] * pcgeil (n=steffen@p5B17FCAB.dip.t-dialin.net) has joined #beagle
  • [06:06:00] * ArteK (n=Artur@81.15.241.96) has joined #beagle
  • [06:09:05] <raster> sakoman: boo!
  • [06:09:06] * arne-unicap (n=arne@p5481AC2F.dip0.t-ipconnect.de) has joined #beagle
  • [06:09:49] <raster> sakoman: wakie! push your kernel mods for overo to oe.dev!
  • [06:09:50] <raster> :)
  • [06:11:55] * mib_dba94f (i=5640c7e4@gateway/web/ajax/mibbit.com/x-eaa91583d0f0d795) has joined #beagle
  • [06:18:38] * tsjsieb (n=tsjsieb@dejongbeheer.nl) has joined #beagle
  • [06:18:48] * favor (n=user@221.10.25.6) Quit (Remote closed the connection)
  • [06:25:58] * tsjsieb (n=tsjsieb@dejongbeheer.nl) Quit (Remote closed the connection)
  • [06:26:19] * shenki (n=joel@202.174.42.5) has joined #beagle
  • [06:29:43] * tsjsieb (n=tsjsieb@dejongbeheer.nl) has joined #beagle
  • [06:33:21] <koen> good morning all
  • [06:33:37] <ds2> morning koen
  • [06:34:42] <koen> denix: heh, I was planning to tackle udev caches this morning :)
  • [06:40:11] * ScriptRipper_ is now known as ScriptRipper
  • [06:46:19] * jsync (n=jess@207.88.137.2.ptr.us.xo.net) has left #beagle
  • [06:53:14] * tsjsieb (n=tsjsieb@dejongbeheer.nl) Quit (Remote closed the connection)
  • [06:54:09] * tsjsieb (n=tsjsieb@dejongbeheer.nl) has joined #beagle
  • [06:55:01] <raster> koen: dutchman!
  • [06:55:50] <ds2> solar: USB boot is possible
  • [06:56:01] * pcgeil (n=steffen@p5B17FCAB.dip.t-dialin.net) Quit ()
  • [06:58:28] <koen> raster: crazyman!
  • [06:59:43] <russ> then its time to make a rescue beagle
  • [06:59:59] <raster> koen: boo
  • [07:00:02] <russ> plug its otg port into the other beagle and the rescue beagle acts as a mass storage device
  • [07:00:12] <russ> that boots the other beagle with its own xloader/u-boot
  • [07:03:24] <ds2> rescue beagle? shouldn't that be the st bernard?
  • [07:05:13] * raster (n=raster@enlightenment/developer/raster) Quit ("Gettin' stinky!")
  • [07:10:06] * mckoan|away is now known as mckoan
  • [07:31:16] * raster (n=raster@enlightenment/developer/raster) has joined #beagle
  • [07:37:42] <raster> koen: ogooood koen!
  • [07:37:49] <raster> (moving my patches from or-dev)
  • [07:37:52] <raster> err oe.dev
  • [07:49:28] * ceyusa (n=ceyusa@cm216016.red83-165.mundo-r.com) has joined #beagle
  • [07:50:31] * lcuk (i=lcuk@cpc3-oldh7-0-0-cust590.manc.cable.ntl.com) has joined #beagle
  • [07:54:21] <eFfeM> ds2, wrt usb boot: is there indeed a std usb boot facility (so just boot a kernel from a usb device)? I thought there was quite some trickery involved
  • [07:54:36] * vishnupsn (i=ca58ea15@gateway/web/ajax/mibbit.com/x-b7fdf56e77bee0eb) Quit ("http://www.mibbit.com ajax IRC Client")
  • [07:54:43] * vishnupsn (i=ca58ea15@gateway/web/ajax/mibbit.com/x-1eae593f73214b9d) has joined #beagle
  • [07:55:41] * MrBIOS-_ (n=aperez@h65.149.51.74.cable.snbr.sanbrunocable.com) Quit (Read error: 110 (Connection timed out))
  • [08:03:19] <ds2> eFfeM: std for OMAP, IIRC
  • [08:08:40] * mib_dba94f (i=5640c7e4@gateway/web/ajax/mibbit.com/x-eaa91583d0f0d795) Quit ("http://www.mibbit.com ajax IRC Client")
  • [08:11:36] * ArteK (n=Artur@81.15.241.96) has left #beagle
  • [08:16:14] * rupeshgujare (n=rupesh@59.160.172.220) has joined #beagle
  • [08:16:55] * Xenion (n=robert@p579FCC6B.dip.t-dialin.net) has joined #beagle
  • [08:18:24] * XorA (n=XorA@93-97-174-2.zone5.bethere.co.uk) has joined #beagle
  • [08:25:15] * abitos (n=nixgibts@p5B2E7A17.dip.t-dialin.net) has joined #beagle
  • [08:40:27] * XorA_ (n=XorA@93-97-174-2.zone5.bethere.co.uk) has joined #beagle
  • [08:41:40] * PhastPhrog (n=chatzill@194.193.86.112) has joined #beagle
  • [08:54:58] * XorA (n=XorA@93-97-174-2.zone5.bethere.co.uk) Quit (Read error: 110 (Connection timed out))
  • [09:04:13] <methril|work> good morning
  • [09:04:46] <koen> hey methril|work
  • [09:05:48] <methril|work> the most active oe commiter
  • [09:05:54] <methril|work> :)
  • [09:06:35] <AV500> gm
  • [09:06:35] <koen> most active typo maker
  • [09:06:39] <koen> hey AV500
  • [09:07:12] <AV500> koen: http://www.h-online.com/open/Intel-wants-to-co-develop-the-GCC--/news/113065
  • [09:07:16] <AV500> thats what I meant
  • [09:08:22] <koen> AV500: nice!
  • [09:08:51] <AV500> yes, where is the same from ARM?
  • [09:09:24] <koen> AV500: if only ARM would change the terms of their CSL contract...
  • [09:09:49] <koen> "make if f***** work on real hardware and get it upstream"
  • [09:10:51] <methril|work> hehehe
  • [09:12:43] * guillaum1 (n=gl@AMontsouris-153-1-1-6.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [09:13:22] * guillaum1 (n=gzba4143@AMontsouris-153-1-1-6.w86-212.abo.wanadoo.fr) has joined #beagle
  • [09:13:37] <AV500> koen: this is not enuf
  • [09:14:05] <AV500> getting crap upstream is not enough
  • [09:14:19] <mru> removing crap from upstream would be better
  • [09:16:15] <koen> I'm having trouble forming an opinion on gcc
  • [09:16:25] <mru> let me help you: it sucks
  • [09:17:02] <koen> since most of the people complaining about it are offering either an alternative (RVCT, tms370) or commercial support (CSL)
  • [09:17:33] <mru> csl sucks every bit as much as fsf gcc
  • [09:18:02] <AV500> RVCT and tms370 both cannot compile kernel -> fail
  • [09:18:12] <koen> my main problem with gcc is the fsf
  • [09:18:14] <AV500> icc can -> win
  • [09:18:23] <mru> that's because the kernel uses gcc-specific features all over the place
  • [09:18:25] <koen> or rather copyright assingment
  • [09:18:43] <AV500> if gcc suck then arm needs to make RVCT gcc compatible and give if for free for os use
  • [09:19:00] <mru> rcvt is mostly gcc compatible
  • [09:19:07] <mru> only inline asm is missing
  • [09:19:14] <AV500> "only" :-)
  • [09:19:16] <mru> and some other misfeatures nobody should be using
  • [09:19:23] <AV500> and this threads issue, no?
  • [09:19:28] <mru> what threads issues?
  • [09:19:46] <AV500> it support only one of the 2 threads models
  • [09:19:52] <AV500> AFAIK
  • [09:19:53] <mru> which two?
  • [09:20:21] <mru> compilers don't need to know or care about threads
  • [09:20:30] <mru> it's all hidden behind the pthreads api
  • [09:20:42] <AV500> "Finally, the target must be running a Linux kernel with support for NPTL..."
  • [09:20:50] <AV500> maybe this is a non-issue.
  • [09:20:58] <mru> that's totally a non-issue
  • [09:21:07] <AV500> ok, then we missunderstood
  • [09:21:18] <mru> the old linuxthreads was a temporary hack until the kernel got fixed properly
  • [09:22:19] <mru> nptl has been around for 6 years or so
  • [09:22:37] <mru> *nobody* uses the old stuff now
  • [09:22:45] <mru> except my employer...
  • [09:22:49] <koen> nobody relevant
  • [09:23:03] <AV500> I just found this: LINUXTHREADS_OLD=y :-)
  • [09:23:13] <koen> when we started angstrom we said "only support kernel 2.6, use gcc4, eabi and nptl"
  • [09:23:27] <mru> good decision
  • [09:23:34] <koen> that was a HUGE break, but all neatly at the same time
  • [09:23:50] <mru> linux 2.6 supports everything older kernels did and more
  • [09:23:56] <AV500> damn, cannot run most lovely 2.2 kernels any more
  • [09:23:58] * pcgeil (n=steffen@HFUVPN475.HFU-VPN.HS-Furtwangen.DE) has joined #beagle
  • [09:24:04] <mru> so there's no reason, ever, to use linux 2.4
  • [09:24:49] <mru> koen: huge break against what?
  • [09:25:23] <koen> the current status back them (hardfpa oabi, linuxthreads, mixed bag of 2.4.19-rmk1, 2.6.x, etc)
  • [09:26:01] <mru> uh, I fail to see the problem
  • [09:26:11] <mru> it's not like you need to run random binaries off the net
  • [09:26:15] <koen> right :)
  • [09:26:19] <koen> exactly
  • [09:26:30] <mru> your system is exactly what you make it
  • [09:26:48] <AV500> mru: you would be surprised how many people ask us how to install office on our units
  • [09:27:02] <AV500> so random binaries for a different cpu arch and OS :-)
  • [09:27:08] <mru> AV500: no, I wouldn't
  • [09:27:16] <AV500> IE is also popular
  • [09:27:39] <mru> and that doesn't change the fact that there *are no* binaries to download and run
  • [09:27:46] * DJWillis has just recently escaped from a project that had 2.4.2* built on top of 2.4.19-rmk with patches and backports from 2.6 on a shipping ARM setup :-o. With a glibc 2.2 userspace.
  • [09:27:51] <mru> apart from distro packages, but they don't count
  • [09:28:13] * florian_kc (n=fuchs@port-217-146-132-69.static.qsc.de) has joined #beagle
  • [09:28:34] * florian_kc is now known as florian
  • [09:29:09] <mru> the libc6 switch on x86 was ugly
  • [09:29:20] <mru> I tried to upgrade a system in-place once
  • [09:29:26] <mru> rendered it unusable
  • [09:31:16] * ant_work (n=andrea@host214-85-static.34-85-b.business.telecomitalia.it) has joined #beagle
  • [09:33:56] * vsr (i=7aa60de8@gateway/web/ajax/mibbit.com/x-707ef1f4cf3f54f9) Quit ("http://www.mibbit.com ajax IRC Client")
  • [09:35:20] * Xenion (n=robert@p579FCC6B.dip.t-dialin.net) Quit (Read error: 101 (Network is unreachable))
  • [09:35:42] * Xenion (n=robert@p579FCA8E.dip.t-dialin.net) has joined #beagle
  • [09:48:37] * french_user (i=c1317c6b@gateway/web/ajax/mibbit.com/x-2bd12dcf357a4180) has joined #beagle
  • [09:51:14] * mib_690hko (i=7aa60de8@gateway/web/ajax/mibbit.com/x-91ddc257565d47f5) Quit ("http://www.mibbit.com ajax IRC Client")
  • [10:08:42] * fenn (n=fenn@cpe-72-177-52-180.austin.res.rr.com) has joined #beagle
  • [10:10:44] * pcgeil (n=steffen@HFUVPN475.HFU-VPN.HS-Furtwangen.DE) Quit (Read error: 113 (No route to host))
  • [10:14:35] * dipak (i=c22f0bce@gateway/web/ajax/mibbit.com/x-545571e7e36c1c35) has joined #beagle
  • [10:17:48] <koen> florian: openpad?
  • [10:22:59] <florian> Amazing how many prople look at the ARM machine registry.
  • [10:25:21] <dipak> Hi all! Can I use beagleboard as a set top box ?
  • [10:26:09] <dipak> What are the ways by which I can connect this board to internet ? like ethernet port etc
  • [10:29:01] <AV500> dipak: usb ethernet
  • [10:30:20] <russ> its late
  • [10:30:43] <russ> can some one else look at omap3_gpio_pads_init in arch/arm/plat-omap/gpio.c and tell me if the code is whack?
  • [10:30:57] <russ> the j > 31 part seems wrong
  • [10:31:08] <russ> actually, no
  • [10:31:11] <dipak> thanks av500
  • [10:31:13] <russ> I'm on crack
  • [10:35:58] * dipak (i=c22f0bce@gateway/web/ajax/mibbit.com/x-545571e7e36c1c35) Quit ("http://www.mibbit.com ajax IRC Client")
  • [10:37:03] * raster (n=raster@enlightenment/developer/raster) Quit ("Gettin' stinky!")
  • [10:38:14] <koen> florian: I had to check whether TI/logic had registered their DA830 board (which they didn't)
  • [10:40:17] <florian> koen: There are still a lot of people/companies around nnot using the registry :-/ I had to register the Topas myself as well.
  • [10:40:53] * XorA_ is now known as XorA
  • [10:42:10] * royerfa (n=fabroy01@fw-tnat.cambridge.arm.com) has joined #beagle
  • [10:42:35] * javaJake (n=javaJake@unaffiliated/javajake) has joined #beagle
  • [10:42:42] * javaJake (n=javaJake@unaffiliated/javajake) has left #beagle
  • [10:49:14] * raster (n=raster@enlightenment/developer/raster) has joined #beagle
  • [10:50:18] * geckosenator (n=sean@c-71-237-94-78.hsd1.co.comcast.net) has joined #beagle
  • [10:53:25] * PhastPhrog (n=chatzill@194.193.86.112) Quit (Remote closed the connection)
  • [10:57:24] * geckosenator (n=sean@c-71-237-94-78.hsd1.co.comcast.net) Quit (Read error: 60 (Operation timed out))
  • [11:00:37] * PhastPhrog (n=chatzill@194.193.86.112) has joined #beagle
  • [11:21:04] * ghaz`` (i=theghaz@pomm.es) Quit (Remote closed the connection)
  • [11:22:27] * brolin (n=brolin@190.157.14.3) Quit ("Lost terminal")
  • [11:27:02] * bavison (n=ben@cpc2-cmbg9-0-0-cust635.cmbg.cable.ntl.com) has joined #beagle
  • [11:29:29] * ghaz`` (i=theghaz@pomm.es) has joined #beagle
  • [11:31:53] <koen> raster: what's the status of the e<>connman integration?
  • [11:33:51] <MrMarv> what do i need to do to have the Rev B6 miniusb out to be used as host? When i connect a pendrive its build-in LED does not lid up but on any other usb power source it does. So i thing there is no power coming out of the beagleboard...
  • [11:34:29] <koen> MrMarv: you need a proper OTG host cable and connect a hub
  • [11:36:40] <MrMarv> koen: this is what i did. I connected a mini-male to normal size female cable between the board and a active powered usb hub. When i plug in my pendirve to the hub its led is powered.... well it is as well if i does not connect hub and boad. So this only indicates electric power.
  • [11:37:32] <MrMarv> it is a OS dependent feature to use the USB stuff, right?
  • [11:37:37] <adj> what kind of mini connector does your cable have? is it mini-A?
  • [11:38:01] <ShadowJK> There's a difference between OTG host cable and a normal mini-male to normal female cable, right
  • [11:39:53] <MrMarv> hmm... the cables Box says its "USB-Mini-A"
  • [11:40:14] <MrMarv> checking wikipedia says that it looks like Mini-B...
  • [11:40:15] <MrMarv> oh
  • [11:40:27] <raster> koen: errrr..... dunno
  • [11:40:49] <adj> MrMarv: mini-a and mini-b are very close to each other
  • [11:41:29] <MrMarv> yeay: regarnding to http://en.wikipedia.org/wiki/File:Types-usb_new.svg it is Mini-B
  • [11:41:40] <adj> MrMarv: http://www.hervanta.com/wiki/images/7/70/Minit.jpg
  • [11:41:56] <adj> left one is mini-b, right one is mini-a
  • [11:42:04] <MrMarv> adj: the left one. Mini-b
  • [11:42:15] <MrMarv> hmmm... this ucks
  • [11:44:03] <MrMarv> so.... which is the significant difference?
  • [11:44:31] <adj> mini-a has ID pin shorted to ground
  • [11:44:34] <MrMarv> ... besides that a works an b does not
  • [11:44:53] <adj> that way OTG device will know that it should act as a host
  • [11:44:59] * XorA_ (n=XorA@93-97-174-2.zone5.bethere.co.uk) has joined #beagle
  • [11:45:52] <adj> mini-B to big female A shouldn't actually even exist... but unfortunately they do exist. They are against every USB specifications
  • [11:46:26] <russ> adj: I like having enough different types of cables in order to "hang myself with"
  • [11:47:36] <MrMarv> i thing i am going to hang the shop guy who sold me this as mini-a with my mini-b cable ;)
  • [11:48:11] <adj> yeah, i have those cables as well, but when dealing only with usb compliant devices one shouldn't ever even have need for those cables
  • [11:49:13] <MrMarv> hmmm i wonder if there is a adapter for mini-b to mini-a?
  • [11:49:49] <adj> both male? then yes
  • [11:50:52] <MrMarv> might be cheaper than buying another cable
  • [11:51:33] <adj> ah, hmm, that kind of an adapter. That might not exist. At least shouldn't
  • [11:52:13] * XorA_ has the right cables to make any <-> any for USV
  • [11:52:15] <XorA_> USB
  • [11:52:25] <adj> the _only_ USB compliant cable/adapter with female connector on the other end is Micro-A (male) <--> Std-A (female)
  • [11:53:08] <russ> mini-A to mini-B exists, but why wouldn't it be compliant?
  • [11:53:21] * __alanc__ (n=a-campbe@nat/ti/x-486cdc4e5e7c65f5) has joined #beagle
  • [11:53:25] * rsalveti (n=salveti@189.70.45.223) Quit (Remote closed the connection)
  • [11:53:28] <adj> Mini-A (male) <--> Std-A (female) was (okay, it still is)
  • [11:53:59] <adj> mini-A to mini-B has male connectors at both ends
  • [11:54:21] <ShadowJK> hm, is it just me or is mini-a somewhat hard to find
  • [11:54:38] <MrMarv> well i would need mini-b female <-> mini-a male
  • [11:55:08] <XorA_> oddly mini-A to mini-B would be required for the tosa :-D
  • [11:55:09] <MrMarv> ShadowJK: no you are not alone.... i went to 5 stores but not a single mini-a to whateer cable
  • [11:55:27] <XorA_> mini-a I found loads of on ebay
  • [11:55:32] <XorA_> from chinese sellers
  • [11:56:24] <adj> beware of those chinese cables, many of them are actually mini-b, not mini-a. I learnt this the hard way
  • [11:56:55] <XorA_> adj: mine defineately is
  • [11:57:12] * arne-unicap (n=arne@p5481AC2F.dip0.t-ipconnect.de) Quit (Read error: 113 (No route to host))
  • [11:57:18] * XorA (n=XorA@93-97-174-2.zone5.bethere.co.uk) Quit (Read error: 110 (Connection timed out))
  • [11:57:24] * XorA_ is now known as XorA
  • [11:58:03] * mib_s5emrm (i=d470a0d2@gateway/web/ajax/mibbit.com/x-b7f8d7108d7af3b4) has joined #beagle
  • [11:59:26] <adj> http://cgi.ebay.co.uk/USB-OTG-Cable-Transparent-Type-Mini-A-to-Type-B-2m_W0QQitemZ400032887456QQcmdZViewItemQQptZUK_Computing_USB_Cables?hash=item400032887456&_trksid=p4634.c0.m14.l1262
  • [11:59:34] <adj> http://cgi.ebay.co.uk/USB-OTG-Cable-Transparent-Type-Mini-A-to-Mini-B-0-5m_W0QQitemZ400033739679QQcmdZViewItemQQptZUK_Computing_USB_Cables?hash=item400033739679&_trksid=p4634.c0.m14.l1262
  • [12:01:53] <ShadowJK> hm, I guess digikey has them too
  • [12:05:37] * Openfree` (n=Openfree@222.65.169.254) has joined #beagle
  • [12:05:43] <MrMarv> the text says it is mini-a but the pictures are mini-b, right? http://cgi.ebay.de/USB-Adapter-Buchse-A-Stecker-MINI-A-5-polig_W0QQitemZ310125250303QQcmdZViewItemQQptZDE_Technik_Computerzubeh%C3%B6r_Kabel_Adapter?hash=item310125250303&_trksid=p3286.c0.m14
  • [12:09:22] * arne-unicap (n=arne@p5481AC2F.dip0.t-ipconnect.de) has joined #beagle
  • [12:09:31] * rsalveti (n=salveti@200.184.118.130) has joined #beagle
  • [12:10:19] <kulve> MrMarv: looks like mini-b to me
  • [12:10:28] <MrMarv> ya
  • [12:12:17] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) has joined #beagle
  • [12:14:21] * ddompe_ (n=ddompe@ip221-27-10-190.ct.co.cr) has joined #beagle
  • [12:15:10] * ddompe__ (n=ddompe@ip221-27-10-190.ct.co.cr) has joined #beagle
  • [12:15:18] * ddompe_ (n=ddompe@ip221-27-10-190.ct.co.cr) Quit (Read error: 104 (Connection reset by peer))
  • [12:16:10] * ddompe_ (n=ddompe@ip221-27-10-190.ct.co.cr) has joined #beagle
  • [12:16:10] * ddompe__ (n=ddompe@ip221-27-10-190.ct.co.cr) Quit (Read error: 104 (Connection reset by peer))
  • [12:16:27] * ddompe_ (n=ddompe@ip221-27-10-190.ct.co.cr) Quit (Client Quit)
  • [12:18:51] * pcgeil (n=steffen@lnht-l1.lnht.FH-Furtwangen.DE) has joined #beagle
  • [12:19:47] * pcgeil (n=steffen@lnht-l1.lnht.FH-Furtwangen.DE) Quit (Client Quit)
  • [12:23:05] * vishnupsn (i=ca58ea15@gateway/web/ajax/mibbit.com/x-1eae593f73214b9d) Quit ("http://www.mibbit.com ajax IRC Client")
  • [12:23:15] <russ> ever notice that mini-a looks a lot like DVI?
  • [12:23:37] <koen> like hdmi?
  • [12:23:48] <mru> mini-hdmi?
  • [12:24:13] <russ> ya, sorry, hdmi
  • [12:26:53] <MrMarv> hmmm... is there no way to use a mini-b cable? a friend just told me he is using such a cable with his beagle (rev b7)
  • [12:27:12] <MrMarv> he told me that even otg worked
  • [12:28:20] <russ> I think you can solder down the id pin on the board
  • [12:28:32] <russ> or was that for using a mini-a cable when a mini-b was needed...
  • [12:28:48] <ShadowJK> isn't there a way to force it into host mode regardless of cable?
  • [12:29:34] <koen> there's a sysfs node for it
  • [12:30:02] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) Quit (Read error: 110 (Connection timed out))
  • [12:31:54] * arne-unicap (n=arne@p5481AC2F.dip0.t-ipconnect.de) Quit (Read error: 113 (No route to host))
  • [12:39:22] * hrw|gone is now known as hrw
  • [12:39:27] <hrw> morning
  • [12:40:19] <hrw> koen: which EP93xx you have?
  • [12:40:26] <koen> hrw: the olimex one
  • [12:40:35] <koen> ehm
  • [12:40:35] <hrw> koen: which cpu I mean
  • [12:40:37] <koen> sorry
  • [12:40:55] <hrw> koen: I got edb9301 (cirruslogic) today
  • [12:40:56] <koen> the wone with crunch and fb, 9312
  • [12:43:04] <koen> hrw: IIRC (it's been a while ago) the 9301 lacks the FPU according to the docs, but it does have one
  • [12:44:14] <hrw> interesting ;D
  • [12:44:46] <koen> kinda like the celerons and amd triplecore cpus
  • [12:45:00] <hrw> economic cpu
  • [12:45:37] <mru> buying a triple-core is buying something you *know* is broken
  • [12:45:42] <koen> I think they stopped producing 9301s and when demand rose just relabeled 9302s
  • [12:46:02] <koen> mru: buying a ep93xx cpu amounts to the same
  • [12:46:23] <mru> I'm not familiar with that cpu
  • [12:46:24] <koen> stepping A through D don't boot
  • [12:47:10] <koen> Stepping E needs an RC network to catch a reset bug and reset the CPU voltage source
  • [12:47:28] <hrw> and this board uses redboot which suxx
  • [12:47:51] <koen> there's a reason why every ep93xx board has 2 boot leds :)
  • [12:48:03] <koen> if the red led is on for too long -> RC network kicks in
  • [12:48:41] <koen> it's redeeming feature is an omap3 like bootrom that can load a steppingstone from serial in case you brick it
  • [12:51:15] * russ http://www.flickr.com/photos/31208937@N06/3443906129/
  • [12:51:24] * russ http://www.flickr.com/photos/31208937@N06/3443906135/
  • [12:51:33] * russ http://www.flickr.com/photos/31208937@N06/3443906143/
  • [12:51:43] * russ http://www.flickr.com/photos/31208937@N06/3443906137/
  • [12:53:31] * koen is having a bit of trouble spotting where the cut trace is
  • [12:53:49] <russ> on which one?
  • [12:54:05] * moldy (n=rene@unaffiliated/moldy) has joined #beagle
  • [12:54:09] * moldy (n=rene@unaffiliated/moldy) has left #beagle
  • [12:54:50] * moldy (n=rene@unaffiliated/moldy) has joined #beagle
  • [12:54:54] * moldy (n=rene@unaffiliated/moldy) has left #beagle
  • [12:55:22] <koen> http://www.flickr.com/photos/31208937@N06/3443906135/in/photostream/
  • [12:55:40] <russ> that one doesn't have any cut traces
  • [12:55:41] * moldy (n=rene@unaffiliated/moldy) has joined #beagle
  • [12:55:43] * moldy (n=rene@unaffiliated/moldy) has left #beagle
  • [12:56:16] <koen> I meant http://www.flickr.com/photos/31208937@N06/3443906129/
  • [12:56:21] <koen> and you've added a note :)
  • [12:56:32] * moldy (n=rene@unaffiliated/moldy) has joined #beagle
  • [12:56:40] * hrw is now known as hrw|gone
  • [12:56:54] * moldy (n=rene@unaffiliated/moldy) has left #beagle
  • [12:56:56] * jkridner|work (n=a0321898@nat/ti/x-0bbee3380c1292ef) has joined #beagle
  • [12:57:16] <russ> that brings me down to 7mW
  • [12:58:11] <XorA> hi jkridner|work
  • [12:58:48] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [13:03:11] <XorA> jkridner|work: any chance of a brief priv chat
  • [13:03:37] * prestonw (i=18634e1d@gateway/web/ajax/mibbit.com/x-27812c39e13a4214) Quit ("http://www.mibbit.com ajax IRC Client")
  • [13:13:55] * cbrake_away is now known as cbrake
  • [13:14:54] * BThompson (n=a0193480@nat/ti/x-468ff1546b81edfa) has joined #beagle
  • [13:16:34] * JuanG (n=Juan@nat/ti/x-ea3c9f889d37ed2e) has joined #beagle
  • [13:16:37] * JuanG (n=Juan@nat/ti/x-ea3c9f889d37ed2e) has left #beagle
  • [13:17:01] * rbelem (n=rodrigo@pdpc/supporter/active/rbelem) has joined #beagle
  • [13:18:39] * robclark (n=robclark@nat/ti/x-6a57a466745d4d27) has joined #beagle
  • [13:20:33] * emeb_mac (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [13:21:19] <Crofton> anyone built qt for beagle in OE?
  • [13:22:25] * nielsbohr (n=piespy@75.149.106.130) Quit (Remote closed the connection)
  • [13:22:55] * ddompe (n=ddompe@200.122.155.113) has joined #beagle
  • [13:23:11] <koen> Crofton: which qt?
  • [13:23:23] <koen> Crofton: 2,3,4, x11,embedded,directfb?
  • [13:23:37] <Crofton> er, the one you wold use on a pc
  • [13:23:43] * Crofton knew that was coming
  • [13:24:06] <Crofton> gnuradio is adding a qt based gui
  • [13:24:32] <koen> Crofton: opkg list | grep qt4
  • [13:24:58] <Crofton|work> koen, you spell yes funny :)
  • [13:25:18] <koen> hey, you're the one complaining about terseness :)
  • [13:26:42] <Crofton|work> how would you handle an Android image needing a specific kernel?
  • [13:28:14] <XorA> by listening to Beastie Boys
  • [13:28:16] <Crofton|work> why did I agree to review this paper
  • [13:28:53] * arne-unicap (n=arne@p5481AC2F.dip0.t-ipconnect.de) has joined #beagle
  • [13:29:37] * sam` (n=sam@frog.bulix.org) has joined #beagle
  • [13:32:26] * Yuvi_ (n=yuvi@resnet-241-44.resnet.umbc.edu) has joined #beagle
  • [13:33:18] <russ> and do the qt4 dance
  • [13:35:51] * ant_work (n=andrea@host214-85-static.34-85-b.business.telecomitalia.it) Quit ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]")
  • [13:36:00] * russ http://www.youtube.com/watch?v=NbTEVbQLC8s
  • [13:36:18] <russ> thats crofton banging on the keyboard
  • [13:36:26] * Yuvi__ (n=yuvi@resnet-241-44.resnet.umbc.edu) has joined #beagle
  • [13:43:22] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [13:45:32] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [13:45:47] * twebb (i=d1e1a1fa@gateway/web/ajax/mibbit.com/x-efa5a9638031bab7) has joined #beagle
  • [13:46:24] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit (Client Quit)
  • [13:46:39] <Crofton|work> actually, someone else I know
  • [13:46:40] * twebb (i=d1e1a1fa@gateway/web/ajax/mibbit.com/x-efa5a9638031bab7) Quit (Client Quit)
  • [13:48:03] <koen> Crofton|work: by patching the regular kernel
  • [13:48:07] * Yuvi (n=yuvi@resnet-241-44.resnet.umbc.edu) Quit (Read error: 110 (Connection timed out))
  • [13:48:29] <koen> Crofton|work: most android kernel stuff is in the staging tree and modular
  • [13:48:33] <Crofton|work> yeah, but how to you select the correct kernel for the image?
  • [13:49:13] <koen> 15:47 < koen> Crofton|work: by patching the regular kernel
  • [13:49:41] <Crofton|work> heh
  • [13:50:02] <Crofton|work> I still think we should keep android out of Angstrom :)
  • [13:50:19] <Crofton|work> All your DISTRO are beling to Angstrom ...
  • [13:51:17] * Yuvi (n=yuvi@resnet-241-44.resnet.umbc.edu) has joined #beagle
  • [13:51:41] <koen> keeping android out of angstrom has nothing todo with your kernel question
  • [13:52:24] <Crofton|work> so you can make one kernel rule them all
  • [13:52:24] * Yuvi_ (n=yuvi@resnet-241-44.resnet.umbc.edu) Quit (Read error: 110 (Connection timed out))
  • [13:53:31] <koen> Crofton|work: how do you manage rebuilding things like kernel modules and tightly coupled userspace when building a different image (needing a different kernel)?
  • [13:53:53] <AV500> koen: you just need the android patches in your kernel
  • [13:54:05] * hrw|gone is now known as hrw
  • [13:54:06] <koen> AV500: that's what I've been saying
  • [13:54:13] <Crofton|work> the dozens of android diffs were frightening me
  • [13:54:15] <keesj> how about generic hardware ain't that a better solution?
  • [13:55:03] <raster> hrrrrrm
  • [13:55:18] * raster suspects some oe b0rkedness with fribidi
  • [13:55:41] <koen> XorA: did the postdude show up already?
  • [13:56:05] <raster> oooh no
  • [13:56:08] <raster> me wrongz
  • [13:56:10] <XorA> koen: wont know until I get home
  • [13:56:13] * Batko_Marto (n=Batko_Ma@CPE001346f996d2-CM001e6b1a8d1e.cpe.net.cable.rogers.com) has joined #beagle
  • [13:56:14] <AV500> koen: so, patch em in :-)
  • [13:56:25] <koen> AV500: that's what I've been saying
  • [13:56:36] <AV500> I guess in a couple of week Android is going to dethrone Mini-A here :-)
  • [13:56:58] <koen> I don't think so
  • [13:57:06] <koen> there's a developer crowd here
  • [13:57:19] <koen> angstrom is something only users and PHBs seem to want
  • [13:57:39] <Crofton> freudian slip
  • [13:58:17] <AV500> I knew it :-)
  • [13:58:32] * AV500 must tell boss to grow PH
  • [13:58:56] <koen> heh
  • [13:59:03] <koen> that was a nice typo
  • [13:59:12] <keesj> angstrom laks an accecible end user sdk like mamona had
  • [13:59:20] * koen puts on dunce cap
  • [13:59:37] <mru> sdk, bleh
  • [13:59:40] <AV500> manamana?
  • [13:59:50] <koen> keesj: you don't need an SDK for a distro
  • [13:59:55] <russ> do-do do do doo
  • [13:59:57] <XorA> AV500: n8x0 distro built from OE
  • [14:00:02] <AV500> manamana!
  • [14:00:08] <russ> do-do do doo
  • [14:00:16] <AV500> manamana!!!!!
  • [14:00:26] * Yuvi_ (n=yuvi@resnet-241-44.resnet.umbc.edu) has joined #beagle
  • [14:00:33] <koen> keesj: you only need an SDK if you have a funky system with weird libs
  • [14:01:02] <Crofton> koen, one day, I hope you have to actually write some code for an OE target, rather than build existing packages .....
  • [14:01:08] * davidm1 (n=david@nat/ti/x-8b1a1cee45368d6e) has joined #beagle
  • [14:01:27] <koen> Crofton: I already did :)
  • [14:01:41] <koen> it's really not different from writing code in general
  • [14:01:54] <XorA> OE makes a great SDK :-)
  • [14:02:02] <koen> write code, compile, test
  • [14:02:06] <keesj> you should be able to compile things on the host pc targeting the target without the need to OE
  • [14:02:09] <koen> then write code, x-compile, test
  • [14:02:23] <koen> keesj: why?
  • [14:02:24] <Crofton> using bitbake to do the compile can be annoying in that loop
  • [14:02:39] <koen> keesj: if you are going to install an SDK, why not OE?
  • [14:03:01] <keesj> to be more accecible to revelopers (they can not and do not want to package for every cycle)
  • [14:03:08] * Yuvi__ (n=yuvi@resnet-241-44.resnet.umbc.edu) Quit (Read error: 110 (Connection timed out))
  • [14:03:15] <russ> keesj: its really bad in a development environment to build stuff for the target against stuff installed on your own machine
  • [14:03:34] <koen> keesj: nobody is forcing you to package anything
  • [14:03:36] <russ> keesj: then each developer ends up building against different versions of things and there are lots of broken builds
  • [14:03:41] * davidm2 (n=David@nat/ti/x-8ea09117634331e8) has joined #beagle
  • [14:03:54] <Crofton> russ, you can build against the OE libs without using OE
  • [14:04:59] <Crofton> what does adding LD_FLAGS to TARGET_ARCH_CC do?
  • [14:06:19] <koen> Crofton: it forced LDFLAGS into the $CC var
  • [14:06:36] <koen> Crofton: so buildsystems have a harder time ignoring it
  • [14:06:46] <Crofton> ok
  • [14:06:48] <koen> Crofton: codec-engine will still ignore LDFLAGS with that, though
  • [14:07:05] <keesj> russ of course there are som drawbacks (like that not everybody has a different version :p)
  • [14:07:07] <koen> since for some *$(*$&#$*@(#$(@ reason it calls /full/path/to/compiler/gcc
  • [14:07:47] <Crofton> I've noticed we seem to be on autopilot approving that change, but I wanted to have some idea before I joined the crowd
  • [14:08:10] <keesj> but indeed I was thinking of gcc+stating that can be downloaded
  • [14:08:18] * Crofton really needs to look into the OE sdk targets, but he already figured out how to use his tmp stuff we he needed to got that route
  • [14:09:23] * prpplague (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [14:09:35] <koen> Crofton: there are 2 ways to solve the ldflags issue: patch the buildsystem (as Jeremy is doing) and maintain those patches) or force LDFLAGS into TARGET_CC_ARCH
  • [14:10:09] <prpplague> sakoman: morning
  • [14:10:15] <prpplague> sakoman: whats the scoup?
  • [14:11:40] <russ> hey prpplague
  • [14:12:00] <prpplague> russ: hey bud, how the heck are you? it has been ages since i've seen you around
  • [14:12:25] * Yuvi (n=yuvi@resnet-241-44.resnet.umbc.edu) Quit (Read error: 110 (Connection timed out))
  • [14:13:01] * bar_foo (n=falseadd@194.197.235.70) has joined #beagle
  • [14:13:06] <russ> been good
  • [14:13:19] <russ> back in school for my masters
  • [14:16:10] <koen> russ: are you able to measure the impact of PASR now?
  • [14:16:14] * arne_ (n=arne@p5481AC2F.dip0.t-ipconnect.de) has joined #beagle
  • [14:16:26] <russ> koen: ya, its the difference between 6mW and 7mW
  • [14:16:44] <russ> er, 7mW and 8mW
  • [14:17:36] * Leon_Nardella (n=Leon@200-161-14-111.dsl.telesp.net.br) has joined #beagle
  • [14:17:36] <prpplague> russ: ahh dandy
  • [14:17:41] <koen> did you do the powerled hack already?
  • [14:17:51] <russ> koen: I've done all my powerhacks
  • [14:18:33] * arne_ is now known as acasp
  • [14:18:48] <koen> the 10% drop for pasr is acceptable?
  • [14:18:57] <keesj> but koen do you understand why I think "pure" oe is not the best choice for a developer while developing?
  • [14:19:05] <koen> 7mW?
  • [14:19:13] <prpplague> russ: i still have a bunch of your jtag/mtd code in use around the office, hehe
  • [14:19:34] <prpplague> keesj: hey bud, did you get your nail board back ok?
  • [14:19:43] <koen> keesj: I do, but as I said, SDKs don't make sense for a distro
  • [14:19:56] <koen> russ: are you planning to go below 7mW?
  • [14:20:00] <keesj> prpplague: nope :`````(
  • [14:20:07] <russ> koen: I don't see anything else I can cut
  • [14:20:28] <prpplague> keesj: doh, let me call the boss, can you email me your shipping address so that i can confirm
  • [14:25:24] * Leon_Nardella (n=Leon@200-161-14-111.dsl.telesp.net.br) Quit ("Leaving.")
  • [14:26:36] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [14:26:39] * bar_foo (n=falseadd@194.197.235.70) has left #beagle
  • [14:29:45] * arne-unicap (n=arne@p5481AC2F.dip0.t-ipconnect.de) Quit (Read error: 113 (No route to host))
  • [14:33:22] * emeb_mac (n=ericb@ip72-223-90-212.ph.ph.cox.net) Quit ()
  • [14:33:31] <sakoman> prpplague: not much new - serial and rtc work, mmc does not
  • [14:34:00] <koen> prpplague: for a moment I was fearing you gone over to the darkside: http://linuxdevices.com/news/NS8166029500.html
  • [14:34:06] <sakoman> the spi-ethernet code micrel provided is an abomination
  • [14:34:22] <prpplague> sakoman: i suspected as much
  • [14:34:37] <sakoman> I really don't know that I will have the time anytime soon to beat it into submission
  • [14:34:51] <sakoman> Have to focus on the paying work :-)
  • [14:34:57] <prpplague> koen: ew that is ugly
  • [14:35:02] <prpplague> sakoman: understood
  • [14:36:25] <sakoman> I'll create a tincan branch for my u-boot and kernel work so others can take a crack at it
  • [14:36:36] <prpplague> sakoman: thanks
  • [14:37:56] * eFfeM (n=nly91006@neo-y1a.ehv.campus.philips.com) Quit ("Leaving.")
  • [14:38:18] <sakoman> prpplague: if Rusty has to spin the board for mmc stuff, you might want to try to convince him to use an ethernet chip that already has a driver in the linus tree
  • [14:38:41] <prpplague> sakoman: problem is there are only two SPI based ethernet chips
  • [14:38:48] <sakoman> It is not fun trying to whip a grungy vendor driver into something reasonable
  • [14:39:09] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [14:39:09] <prpplague> sakoman: the enc28j60, which i wrote the initial rev of and is now in the main tree, and the micrell part
  • [14:39:14] <sakoman> And then you need to maintain it . . .
  • [14:39:39] <prpplague> sakoman: the drawback is that enc28j60 is 3.3v only and is limited to 10BaseT
  • [14:39:50] * tsjsieb (n=tsjsieb@dejongbeheer.nl) Quit ("good night")
  • [14:39:53] <prpplague> sakoman: the micrell is 1.8v capable and can do 10/100 BaseT
  • [14:40:06] <prpplague> sakoman: the enc28j60 has a maintainer now
  • [14:40:14] <sakoman> yeah, I understand the "why"
  • [14:40:57] <sakoman> the signals available on the BB expansion connector really tie your hands
  • [14:41:04] <prpplague> sakoman: indeed
  • [14:41:24] * MrBIOS-_ (n=aperez@h65.149.51.74.cable.snbr.sanbrunocable.com) has joined #beagle
  • [14:41:33] <prpplague> sakoman: i was really for the usage of the enc28j60, but so many more of the beagle crowd complained that they wanted to have 100BaseT
  • [14:42:16] * mru wants 1G ;-)
  • [14:42:22] <sakoman> perhaps they will be willing to step up to the work of getting the driver in shape and maintaining it ;-)
  • [14:43:05] <emeb> what's the advantage of 100baseT when everything gets squeezed down through a SPI interface to the OMAP?
  • [14:43:11] <prpplague> sakoman: yea, i've got a proto board we did setup here, so i'll see what i can do about getting working with my s3c24xx stuff
  • [14:43:47] <prpplague> emeb: yea you don't get the full 100mbit, you should get roughly around 40mbit
  • [14:44:24] <sakoman> prpplague: have the micrel folks benchmarked their driver on beagle?
  • [14:44:38] <sakoman> it would be good to know what they were getting with their hacked driver
  • [14:44:40] <emeb> I suppose that even 40Mbit would be better than 10baseT, but 40 seems optimistic for SPI when you factor in overhead.
  • [14:44:58] * Yuvi_ (n=yuvi@resnet-241-44.resnet.umbc.edu) Quit (Read error: 104 (Connection reset by peer))
  • [14:45:03] * Yuvi (n=yuvi@resnet-241-44.resnet.UMBC.EDU) has joined #beagle
  • [14:45:17] <prpplague> emeb: based on my experience with the enc28j60, you get about 20% overhead
  • [14:45:32] <prpplague> sakoman: yea, from what i understand that are only getting about 5mbit
  • [14:45:38] <prpplague> sakoman: which is silly
  • [14:45:43] <sakoman> indeed
  • [14:46:38] * Yuvi_ (n=yuvi@resnet-241-44.resnet.umbc.edu) has joined #beagle
  • [14:47:07] * dillera (n=dillera@75.149.106.130) has joined #beagle
  • [14:47:07] <emeb> I'm using a Trendnet USB dongle for my BB - does 100baseT through a full-speed USB ifc. Dumb (but cheap)
  • [14:47:20] <emeb> wonder what it benchmarks at.
  • [14:47:30] <hrw> my USB ethernet do 40Mbps on beagleboard (via OTG port)
  • [14:47:57] <emeb> How is that possible? OTG is full-speed only (12Mbps)
  • [14:48:19] <koen> otg is all speeds
  • [14:48:24] <koen> ehci is highspeed only
  • [14:48:32] <emeb> Ah.
  • [14:48:35] <emeb> dopey me
  • [14:48:42] <XorA> koen: high speed only?
  • [14:48:50] <ShadowJK> hrw, what chipset?
  • [14:48:55] <hrw> ShadowJK: dm9601
  • [14:49:01] <ShadowJK> wow
  • [14:49:01] <koen> you get about 70Mbit/s when using g_ether with an old kernel
  • [14:49:07] <hrw> ops.. 40Mbps was other device
  • [14:49:15] <hrw> 7Mbps in beagleboard
  • [14:49:20] <hrw> sorry.. too many boards in range
  • [14:49:20] <koen> XorA: yes, 480MBit only, for everything else -> hub
  • [14:49:32] <XorA> koen: so none compliant then :-(
  • [14:49:50] <koen> the asix gets around 7MByte/s
  • [14:50:03] <ShadowJK> hm, I forget if mine was dm9601, but dm-something... and it gave me about 4Mbit/s in a desktop system, with frequent stalls.. :-)
  • [14:50:47] <XorA> remeber with ethernet on the wire on a real network you can never get about 50MBit
  • [14:50:54] <XorA> s/about/above/
  • [14:51:23] <hrw> koen: I got 860Mbps from my ethernet in dekstop<>laptop
  • [14:51:53] <ShadowJK> Uh, I get ~12megabytes/sec with NFS, 100mbit/s ethernet between two desktops going through a cheapcrap d-link switch
  • [14:52:22] <hrw> koen: r8169 -> dlink 1GbE switch -> tg3
  • [14:52:46] <denix> morning
  • [14:53:33] <denix> koen: I was actually using/testing those changes for couple weeks...
  • [14:56:32] <sakoman> prpplague: because of the pinout change from rev b to rev c the beagle board file will need to do some magic to decide whether to use gpio140 or gpio144 for mmc cd
  • [14:57:05] <sakoman> the uart diffs should be hidden from the kernel by u-boot
  • [14:57:31] <prpplague> sakoman: yea i figured as much, but i was under the impression that there was something in uboot to detect the board rev
  • [14:57:43] <Crofton> whatever happened to "one pin, one function"?
  • [14:58:18] <sakoman> prpplague: there is, but u-boot can't hide the gpio140/144 change
  • [14:59:35] <prpplague> sakoman: right, what i meant was if uboot can detect the board rev, you should be able to add a module argument to the kernel command line to specify which gpio to use
  • [14:59:43] <sakoman> sadly it looks like folks who use the expansion connector are pretty much doomed to having to use a a custom kernel
  • [15:00:46] <sakoman> prpplague: yes, or alternatively the kernel can look at the 2 gpio pins that indicate revision
  • [15:01:14] * CanyonMan (i=81155ca2@gateway/web/ajax/mibbit.com/x-9f91b5087df78172) has joined #beagle
  • [15:01:31] * dirk2 (n=dirk@p5B041961.dip0.t-ipconnect.de) has joined #beagle
  • [15:01:43] <CanyonMan> Can you buy the BeagleBoard's CPU from them, with the pop memory etc. ?
  • [15:02:02] <prpplague> sakoman: ahh, so its two gpio's used to generate a board rev bit?
  • [15:02:08] <koen> denix: I was planning to move to a recent udev as well, but it seems that will be some more work, udev finally has a sane buildsystem :)
  • [15:02:12] <sakoman> prpplague: yes
  • [15:02:35] <prpplague> CanyonMan: iirc, you cannot as the memory pop is done at the same time as the board population
  • [15:02:37] <emeb> Crofton: whoever said "one pin, one function" was living in a time before pins cost more than functions.
  • [15:02:44] <CanyonMan> ah
  • [15:04:42] <WAHa_06x36> anybody around who's written a custom IRQ handler for the omap3?
  • [15:04:45] <hrw> koen: how long does narcissus need to assemble image? and does it give link after that?
  • [15:05:06] <WAHa_06x36> I'm trying to copy the code from the data sheet and modifying it a little to call C functions but it's breaking in strange ways.
  • [15:06:09] <koen> hrw: the "assembling" step is using bzip2, which is usually CPU bound
  • [15:06:13] <hrw> ok
  • [15:06:15] <koen> hrw: and yes, it gives you a link
  • [15:07:43] <koen> hrw: it runs with nice -n10 bzip2, though
  • [15:08:14] * ScriptRipper (n=mmohring@opensuse/member/MartinMohring) Quit (Read error: 104 (Connection reset by peer))
  • [15:08:26] * koen tries
  • [15:08:59] <koen> http://amethyst.openembedded.net/~koen/narcissus/deploy/atngw100/6b3c5d/hrw-test-image-atngw100.tar.bz2
  • [15:09:03] <koen> there we go
  • [15:09:31] <hrw> koen: no, I did build for om-gta02 ;D
  • [15:09:47] <hrw> ep9301 is armv4t and I have armv5te and avr32 in /srv/nfs ;(
  • [15:10:30] * Yuvi (n=yuvi@resnet-241-44.resnet.UMBC.EDU) Quit (Read error: 110 (Connection timed out))
  • [15:11:01] * CanyonMan (i=81155ca2@gateway/web/ajax/mibbit.com/x-9f91b5087df78172) Quit ("http://www.mibbit.com ajax IRC Client")
  • [15:11:19] <dirk2> prpplague, sakoman: I haven't looked at the expansion connector recently, but would it be possible to use e.g. some gpio's at expansion connector to detect if (a) a peripheral board is attached and (b) which one?
  • [15:12:34] <WAHa_06x36> dirk2: sure.
  • [15:12:35] <dirk2> prpplague, sakoman: E.g if a expansion board is connected, two gpio's are driven low by expansion board. If nothing is connected, pins are default high.
  • [15:12:56] <WAHa_06x36> dirk2: if you configure the pull resistors right.
  • [15:13:37] <dirk2> With this we would be able to detect expansion board and do correct pin mux for each expansion automatically. Same way as it is done now with the two gpio's used for Beagle rev dectection
  • [15:17:02] <sakoman> dirk2: sadly the board in question uses every single pin on the expansion connector for some function!
  • [15:18:49] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit (Read error: 104 (Connection reset by peer))
  • [15:18:58] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [15:19:04] <WAHa_06x36> sakoman: could you use a known idle state of some pins that are in use?
  • [15:19:10] <dirk2> sakoman: ok. Maybe there would be a chance that there is one pin used as OMAP3 input which might have some different state if expansion is connected. But sure, this has to be tested.
  • [15:19:37] <dirk2> WAHa_06x36: yes, same idea here
  • [15:21:11] <dirk2> Maybe for rev D we should add some 'expansion detect pins'
  • [15:21:31] <sakoman> dirk2: that is a dangerous method to use :-)
  • [15:21:51] * PhastPhrog (n=chatzill@194.193.86.112) Quit ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]")
  • [15:21:56] <sakoman> could be lots of unintended consequences for people using the expansion port!
  • [15:22:27] <russ> dirk2: maybe a i2c rom on each expansion board?
  • [15:22:37] <sakoman> the only thing that makes sense to me is to fix the definition of the i2c pins and require the use of a cheap i2c rom to indicate pinmux config
  • [15:22:46] <russ> heh
  • [15:22:52] <sakoman> russ: great minds :-)
  • [15:26:08] * nabax (n=nabax@89.129.96.71) has joined #beagle
  • [15:26:25] * rupeshgujare (n=rupesh@59.160.172.220) has left #beagle
  • [15:30:38] * dirk2 discovers OMAP4 machine ID '2160' for 'OMAP_4430SDP' in ARM machine registry http://www.arm.linux.org.uk/developer/machines/
  • [15:30:46] * jrmuizel_ (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [15:34:02] * jrmuizel_ (n=jrmuizel@mozca02.ca.mozilla.com) Quit (Client Quit)
  • [15:34:14] * jrmuizel_ (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [15:37:01] * Openfree` (n=Openfree@222.65.169.254) Quit ("Leaving")
  • [15:37:29] * XorA_ (n=XorA@93-97-174-2.zone5.bethere.co.uk) has joined #beagle
  • [15:38:30] * raster (n=raster@enlightenment/developer/raster) Quit ("Gettin' stinky!")
  • [15:40:34] <koen> dirk2: sadly SDPs are hard to get
  • [15:40:37] <koen> dirk2: and bulky
  • [15:40:50] * luigid (i=629a2d1d@gateway/web/ajax/mibbit.com/x-e0e9cbcf49d83e34) has joined #beagle
  • [15:41:16] * XorA_ (n=XorA@93-97-174-2.zone5.bethere.co.uk) Quit (Client Quit)
  • [15:42:11] <dirk2> koen: But seems that there is already something going on. This is the good news :)
  • [15:42:27] <koen> dirk2: SDPs are mostly internal TI boards
  • [15:42:40] * mib_xpumcw (i=47091882@gateway/web/ajax/mibbit.com/x-33e30749377ddf27) has joined #beagle
  • [15:42:54] <AV500> koen: or for large WTBU customers :-(
  • [15:42:58] <AV500> koen: or for large WTBU customers :-)
  • [15:42:58] <luigid> Anyone know how the beagleboard DSP bridgedriver can be built from OpenEmbedded?
  • [15:43:11] <dirk2> koen: ... most probably to create WTBU kernels ;)
  • [15:43:28] <koen> dirk2: WTBU kernel already have omap4 support
  • [15:43:41] <koen> luigid: no, but the dsplink driver can be built with OE
  • [15:44:04] <luigid> Can you point me to info on that?
  • [15:44:37] <koen> luigid: see recipes/dsplink, there's a readme in there
  • [15:44:52] <koen> luigid: http://angstrom-distribution.org/demo/beagleboard/ already has all the needed binaries inside
  • [15:45:43] <luigid> Yes, but that kernel uses no PM
  • [15:46:08] <koen> so build the PM kernel with OE and the dsplink stuff against that
  • [15:46:47] <luigid> i have built the pm kernel, but do not know how to ADD to it.
  • [15:47:51] <koen> attention deficit disorder?
  • [15:48:08] <luigid> possibly
  • [15:48:09] <koen> dsplink is all built as external modules, no need to patch stuff
  • [15:48:49] <luigid> i guess i need to understand the build system more - am used to "make"
  • [15:48:53] <dirk2> jkridner|work: OpenOCD Cortex A8 support: Would a project 'add cortex A8 support to OpenOCD' have a chance to get approved in Sponsored Project Program (to get a BeagleBoard) ? See "I dont have any A8 hardware." in https://lists.berlios.de/pipermail/openocd-development/2009-April/005305.html
  • [15:49:23] * jrmuizel_ (n=jrmuizel@mozca02.ca.mozilla.com) Quit ()
  • [15:49:24] <luigid> Do you happen to know the milliwatt delta w/wo the dsp in use?
  • [15:49:40] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit (Read error: 110 (Connection timed out))
  • [15:49:56] * french_user (i=c1317c6b@gateway/web/ajax/mibbit.com/x-2bd12dcf357a4180) Quit ("http://www.mibbit.com ajax IRC Client")
  • [15:50:16] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [15:50:54] <dirk2> jkridner, jkridner|work: And btw, when do you think would be the first week to give two boards to approved projects?
  • [15:50:55] * XorA (n=XorA@93-97-174-2.zone5.bethere.co.uk) Quit (Read error: 110 (Connection timed out))
  • [15:51:41] <koen> dirk2: there need to be projects worthy of a beagle first
  • [15:52:08] <koen> dirk2: atm only the ogre3d project seems big enough to warrant a real beagle
  • [15:53:14] <luigid> After i build a kernel in oe how can i get the source tree for the build?
  • [15:53:17] <dirk2> koen: That was my question: If somebody adds a project 'implement OpenOCD Cortex A8 support', would this be 'enough' to get a beagle first
  • [15:55:03] <dirk2> koen: Maybe we then should already start to fill table http://elinux.org/BeagleBoard/contest#Approved_projects with 'not approved'?
  • [15:55:05] <koen> dirk2: no idea, I don't really care about jtag, since the omap3 has a bootrom :)
  • [15:55:45] <koen> dirk2: I'd like to get some more projects first before starting to (dis)approve
  • [15:57:41] * jrmuizel_ (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [15:59:36] * jrmuizel__ (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [15:59:53] * luigid (i=629a2d1d@gateway/web/ajax/mibbit.com/x-e0e9cbcf49d83e34) Quit ("http://www.mibbit.com ajax IRC Client")
  • [16:00:15] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit (Read error: 104 (Connection reset by peer))
  • [16:00:30] * vsr (i=7aa60de8@gateway/web/ajax/mibbit.com/x-187537581371270f) has joined #beagle
  • [16:02:59] * mib_xpumcw (i=47091882@gateway/web/ajax/mibbit.com/x-33e30749377ddf27) Quit ("http://www.mibbit.com ajax IRC Client")
  • [16:08:30] * JuanG (n=Juan@nat/ti/x-35f8141ce96ed70b) has joined #beagle
  • [16:08:40] * JuanG (n=Juan@nat/ti/x-35f8141ce96ed70b) has left #beagle
  • [16:11:43] * PhastPhrog (n=chatzill@87.113.75.147.plusnet.pte-ag2.dyn.plus.net) has joined #beagle
  • [16:12:33] <hrw> http://www.flickr.com/photos/hrwandil/3444358903/
  • [16:14:17] <prpplague> hrw: hehe
  • [16:14:27] <prpplague> hrw: maybe we should add one to the 8050
  • [16:14:36] <hrw> prpplague: the fun is that this button has use
  • [16:14:49] <hrw> prpplague: and its use is really "Don't care" :D
  • [16:15:20] <prpplague> hrw: our first rev of the keypad for the 8050, for one of the user defined keys we had "AnyKey"
  • [16:15:47] * jrmuizel_ (n=jrmuizel@mozca02.ca.mozilla.com) Quit (Read error: 110 (Connection timed out))
  • [16:16:07] * russ (i=foobar@ip70-176-253-20.ph.ph.cox.net) Quit (Read error: 104 (Connection reset by peer))
  • [16:16:12] * pcgeil (n=steffen@p5B17FCAB.dip.t-dialin.net) has joined #beagle
  • [16:18:27] * russ (i=foobar@ip70-176-253-20.ph.ph.cox.net) has joined #beagle
  • [16:21:15] * colla (n=chatzill@host198-216-static.52-88-b.business.telecomitalia.it) has joined #beagle
  • [16:26:12] * Wowbagger_ (n=wowbagge@d154-20-156-61.bchsia.telus.net) has joined #beagle
  • [16:26:29] * vsr (i=7aa60de8@gateway/web/ajax/mibbit.com/x-187537581371270f) has left #beagle
  • [16:27:58] * vijay (n=quassel@203.199.213.3) has joined #beagle
  • [16:29:51] * vijay (n=quassel@203.199.213.3) Quit (Client Quit)
  • [16:30:10] * mckoan is now known as mckoan|away
  • [16:33:12] * montamer (n=quassel@203.199.213.3) has joined #beagle
  • [16:34:38] * guillaum1 (n=gzba4143@AMontsouris-153-1-1-6.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [16:35:06] * guillaum1 (n=gl@AMontsouris-153-1-1-6.w86-212.abo.wanadoo.fr) has joined #beagle
  • [16:39:01] * colla (n=chatzill@host198-216-static.52-88-b.business.telecomitalia.it) Quit ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]")
  • [16:44:29] * montamer (n=quassel@203.199.213.3) Quit (Remote closed the connection)
  • [16:44:53] * montamer (n=quassel@203.199.213.3) has joined #beagle
  • [16:47:30] * rsalveti (n=salveti@200.184.118.130) Quit (Remote closed the connection)
  • [16:50:34] <hrw> have a nice rest of day
  • [16:50:57] * hrw is now known as hrw|gone
  • [16:51:19] * luigid (i=629a2d1d@gateway/web/ajax/mibbit.com/x-002a33dd8afa9516) has joined #beagle
  • [16:54:50] * rsalveti (n=salveti@200.184.118.130) has joined #beagle
  • [16:55:03] * luigid (i=629a2d1d@gateway/web/ajax/mibbit.com/x-002a33dd8afa9516) Quit (Client Quit)
  • [16:55:24] * abitos (n=nixgibts@p5B2E7A17.dip.t-dialin.net) Quit (Nick collision from services.)
  • [16:55:30] * Wiedi (n=wiedi@newton-air.w.fruky.net) has joined #beagle
  • [16:55:41] * abitos (n=nixgibts@p5B2E49C0.dip.t-dialin.net) has joined #beagle
  • [16:56:01] * florian (n=fuchs@port-217-146-132-69.static.qsc.de) Quit (Remote closed the connection)
  • [17:05:40] <ds2> morning
  • [17:06:34] <prpplague> ds2: greetings
  • [17:10:31] <ds2> hey prpplague
  • [17:16:04] * beagle_dev543 (i=52011590@gateway/web/ajax/mibbit.com/x-f25d8c3259924209) has joined #beagle
  • [17:18:08] <beagle_dev543> Has anyone here got glxinfo or any other glx program to work on the beagle, using mesa (forget SGX for now)
  • [17:19:26] <beagle_dev543> I've installed libGL & friends (including /usr/lib/dri/swrast_dri.so for the render) but in my Xlog I get "(EE) AIGLX error: dlopen of /usr/lib/dri/swrast_dri.so failed (/usr/lib/dri/swr) (EE) GLX: could not load software renderer"
  • [17:19:39] <beagle_dev543> which is strange as it is there :S
  • [17:22:01] <beagle_dev543> I also have /usr/lib/xorg/modules/extensions/libglx.so which I had to manually copy across as OE didn't seem to package it
  • [17:23:21] * mib_s5emrm (i=d470a0d2@gateway/web/ajax/mibbit.com/x-b7f8d7108d7af3b4) Quit ("http://www.mibbit.com ajax IRC Client")
  • [17:27:11] * ceyusa (n=ceyusa@cm216016.red83-165.mundo-r.com) Quit (Remote closed the connection)
  • [17:31:35] * montamer (n=quassel@203.199.213.3) Quit (Read error: 104 (Connection reset by peer))
  • [17:33:17] <russ> hmm...arm needed to commonly allocate 16kB in one physical unit is starting to become a problem for me
  • [17:34:55] * mib_e9mk8i (i=555dcd66@gateway/web/ajax/mibbit.com/x-7e97fdc91fc86b8c) has joined #beagle
  • [17:34:56] * robtow (n=rtow1@nat/montavista/x-9494d2044bae8a6b) Quit (Read error: 104 (Connection reset by peer))
  • [17:35:31] * robtow (n=rtow1@nat/montavista/x-79cf0a99d94a6be4) has joined #beagle
  • [17:37:21] * Wowbagger_ (n=wowbagge@d154-20-156-61.bchsia.telus.net) Quit ("Leaving")
  • [17:38:06] <beagle_dev543> huh? ARM works on 4K pages right? or are you talking about something else?
  • [17:38:59] <russ> when arm allocates a new pgd, it has to be 4 contiguous pages
  • [17:39:12] <russ> so, anytime there is a fork()
  • [17:39:57] <russ> and I'm simultaneously fragmenting my memory and reducing available memory
  • [17:42:43] <beagle_dev543> But it's virtual memory, although I guess the mm doesn't defrag
  • [17:44:03] <russ> the processor looks at pgd tables as physical things
  • [17:46:26] <beagle_dev543> Right, are you saying you can't allocate 16kb of physically contiguous memory?
  • [17:47:49] <russ> yes
  • [17:48:01] <beagle_dev543> ouch!
  • [17:48:31] <russ> 228*4kB 181*8kB 1*16kB 0*32kB 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 2376kB
  • [17:48:41] <russ> like that, but one less order 2 allocation available
  • [17:50:59] <koen> russ: now I think of it, your ram resizing approach will interest the dsplink folks, since they need to allocate physically contiguous memory, which linux doesn't let you do easily
  • [17:51:10] <beagle_dev543> Hmm, I thought that if the allocator ran out of one size it would spilt a larger allocation block or join a small one.
  • [17:51:52] <beagle_dev543> It may depend on the allocator and it would also require some of your 8K blocks to be next to each other
  • [17:53:37] <beagle_dev543> Either way your running with very little free RAM
  • [17:55:30] * PhastPhrog (n=chatzill@87.113.75.147.plusnet.pte-ag2.dyn.plus.net) Quit (Read error: 104 (Connection reset by peer))
  • [17:56:42] <russ> I've setup a 16MB zone and a 112MB zone
  • [17:56:54] * dirk2 (n=dirk@p5B041961.dip0.t-ipconnect.de) Quit ("Konversation terminated!")
  • [17:57:00] <russ> the 16MB zone is the only zone that non-movable allocations can be made in
  • [17:57:40] <russ> during suspend, I call shrink_all_memory() which causes user memory to be compressed with ramzswap
  • [17:57:41] <russ> ramzswap can only allocate pages in the kernel zone
  • [17:58:00] <russ> so I get a ton of free memory in the movable zone, but very little free memory in the kernel zone
  • [17:58:21] <russ> but the allocator doesn't know that a way to free up kernel zone memory is to unswap pages from ramzswap
  • [17:58:30] * pcgeil (n=steffen@p5B17FCAB.dip.t-dialin.net) Quit ()
  • [17:58:58] <russ> maybe upon wakeup I just need to kick ramzswap into puking up a bunch of pages
  • [18:01:08] <beagle_dev543> I'm afraid you've gone my level of knowledge :). You want to dynamically resize your memory?
  • [18:01:38] <russ> thats what I've been doing, thats why I need to set the upper 112MB offlimits for unmovable pages
  • [18:01:50] <AV500> koen: they don't, the DSP could use its mmu to see that same virt mem as linux
  • [18:02:10] <russ> I think getting ramzswap to puke up pages will be a good solution for now, just need to figure out how to do that
  • [18:02:43] <AV500> --puke-up?
  • [18:03:12] <beagle_dev543> ok. I've thought about this in the past, but never looked into it. At the time I wondered about using hotplug memory but I'm sure you've found a better way :)
  • [18:03:36] <russ> AV500: take pages that it has compressed and uncompress them
  • [18:04:04] <russ> beagle_dev543: I'm utilize portions of memory hotplug to do online_memory and offline_memory
  • [18:04:10] <AV500> russ: yes, I was joking, sorry, bad joke.
  • [18:04:23] <AV500> russ: you got anything online about what you are doing?
  • [18:04:40] <beagle_dev543> Would be very useful for the display controler driver as it needs large contiguous blocks of memory which will change at run time (for new video surfaces of different sizes)
  • [18:05:34] <AV500> beagle_dev543: the current approach is CMEM which is phys ram taken away from the kernel, but I agree this is not so flexible as one would like it to be
  • [18:06:07] * rbelem is now known as rbelem-lunch
  • [18:08:23] * midtown4 (n=krooked@68-189-104-94.dhcp.wtvl.ca.charter.com) has joined #beagle
  • [18:09:48] <koen> AV500: if they can use the MMU, then why is cmem still needed?
  • [18:11:14] <AV500> coz they did not implement it yet :-)
  • [18:11:28] <AV500> dsplink is from omap1/2 and davinci, they did not have that
  • [18:12:11] <AV500> we asked them for it, they said "eventually" :-)
  • [18:12:16] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [18:12:55] <AV500> but now that dsplink is BSD, it does not need TI to get that changes... hint hint
  • [18:12:59] <beagle_dev543> For the OMAP5 :)?
  • [18:13:34] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit (Remote closed the connection)
  • [18:13:35] <AV500> well, the omap3 will be around for some time
  • [18:13:47] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [18:14:07] <beagle_dev543> I'm sure it will, what I meant is it will be release when the OMAP5 is ;p
  • [18:14:46] <beagle_dev543> Hell, I've got a OMAP5910 :)
  • [18:15:05] * Xenion (n=robert@p579FCA8E.dip.t-dialin.net) Quit (Read error: 101 (Network is unreachable))
  • [18:15:23] * Xenion (n=robert@p579FCF6C.dip.t-dialin.net) has joined #beagle
  • [18:16:16] <AV500> i've got some nice innocator board lying around
  • [18:16:20] <AV500> i've got some nice innovator board lying around
  • [18:16:39] * guillaum1 (n=gl@AMontsouris-153-1-1-6.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [18:17:04] <beagle_dev543> Mines an E3 (http://www.earth.li/~noodles/hardware-e3.html)
  • [18:17:07] * guillaum1 (n=Guillaum@AMontsouris-153-1-1-6.w86-212.abo.wanadoo.fr) has joined #beagle
  • [18:18:39] <russ> beagle_dev543, AV500: I can send you some WIP stuff if you like
  • [18:19:21] <beagle_dev543> WIP?
  • [18:19:37] <russ> work in progress
  • [18:19:52] <beagle_dev543> ah
  • [18:20:21] <AV500> beagle_dev543: mine is this one http://en.wikipedia.org/wiki/Archos#PMA400 :-)
  • [18:20:27] <AV500> a years old though
  • [18:20:31] <AV500> a few years old though
  • [18:21:19] <beagle_dev543> Nice. Bit more portable then mine :)
  • [18:21:45] <beagle_dev543> reboot time, see you on the flip side!
  • [18:28:37] * beagle_dev543 (i=52011590@gateway/web/ajax/mibbit.com/x-f25d8c3259924209) Quit ("http://www.mibbit.com ajax IRC Client")
  • [18:29:15] * florian (n=fuchs@f049005166.adsl.alicedsl.de) has joined #beagle
  • [18:29:46] <russ> I feel odd setting up an oom_notifier to uncompress memory....but hey, whatever works
  • [18:31:28] * beagle_dev543 (i=52011590@gateway/web/ajax/mibbit.com/x-f6842e8c72d2dbed) has joined #beagle
  • [18:34:35] * luigid (i=629a2d1d@gateway/web/ajax/mibbit.com/x-29cb6376449a3d55) has joined #beagle
  • [18:34:46] * Xenion (n=robert@p579FCF6C.dip.t-dialin.net) Quit ("Verlassend")
  • [18:35:57] <luigid> I have the angstrom demo running to test dsplink; modules are loaded; now is there a test app to exercise the dsp?
  • [18:36:17] * k2340 (n=k2340@lawn-128-61-121-203.lawn.gatech.edu) has joined #beagle
  • [18:36:51] <beagle_dev543> I've forgot, which package contains mount.nfs?
  • [18:37:10] <russ> maybe portmap?
  • [18:39:10] <beagle_dev543> nope :(
  • [18:39:24] <koen> nfs-utils?
  • [18:39:46] * koen still intensely dislikes nfs
  • [18:40:19] <luigid> I have the angstrom demo running to test dsplink; modules are loaded; now is there a test app to exercise the dsp?
  • [18:40:55] <beagle_dev543> Not under angstrom
  • [18:41:17] <russ> now...I just need to find which pages are adjacent to holes
  • [18:41:30] * rbelem-lunch is now known as rbelem
  • [18:41:38] <luigid> Can you point me to one?
  • [18:42:35] <beagle_dev543> Deploying find :)
  • [18:42:45] <luigid> thanks
  • [18:42:46] <koen> luigid: have a look in /usr/share/ti-*
  • [18:43:31] <luigid> thanks koen
  • [18:43:45] <koen> luigid: also, join #gst_ti, more dsp people there
  • [18:44:15] <luigid> great - ty
  • [18:47:11] <beagle_dev543> nfs does make development easier
  • [18:48:12] <koen> a large SD card even more so :)
  • [18:48:21] <prpplague> sakoman: ping
  • [18:48:24] <prpplague> koen: ping
  • [18:49:48] * hrw|gone is now known as hrw
  • [18:49:48] <hrw> re
  • [18:49:53] * brijesh (n=bksingh@192.94.94.106) has joined #beagle
  • [18:49:55] <emeb> koen: question on narcissus - just built an image & that worked OK. The SD image had empty FAT partition though. WTF?
  • [18:50:21] <hrw> just for info: Samsung 2494HM monitor syncs perfectly with 1280x800MR-16@50
  • [18:50:26] <emeb> I just used dd if=(image) of=/dev/sdc to setup SD
  • [18:50:31] <prpplague> hrw: hey, question for you
  • [18:50:35] <hrw> prpplague: go
  • [18:51:09] <prpplague> hrw: the new addon board we are doing for the beagle is originally designed with the micrel 10/100 part, we are debating about going back to the enc28j60 which is just a 10baseT part
  • [18:51:23] <prpplague> hrw: big difference is the enc28j60 driver is mainline right now
  • [18:51:29] <prpplague> hrw: thoughts?
  • [18:51:32] <beagle_dev543> With the SD card how do I update my program that I've just rebuilt?
  • [18:51:55] <hrw> prpplague: it would be hard to get >10Mbps probably without hogging cpu too much
  • [18:52:18] <prpplague> hrw: that was my thought as well
  • [18:52:22] <hrw> prpplague: boards are fine with ethernet I think, nevermind how fast it is.
  • [18:52:45] <hrw> prpplague: I had devboard which always had 100Mbps link but was unable to get >1Mbps transfers
  • [18:52:47] <prpplague> koen: your thoughts?
  • [18:52:54] <beagle_dev543> still about twice as fast as USB
  • [18:53:04] <hrw> prpplague: and I was happy that I had that ethernet ;D
  • [18:53:18] <beagle_dev543> I've seen about 600kb as the max in my beagle
  • [18:53:49] <hrw> my current usb ethernet do 7Mbps on beagle (in OTG port)
  • [18:53:55] * luigid (i=629a2d1d@gateway/web/ajax/mibbit.com/x-29cb6376449a3d55) Quit ("http://www.mibbit.com ajax IRC Client")
  • [18:54:37] <beagle_dev543> sorry, I mean usb_gadget (g_ether)
  • [18:54:51] <hrw> beagle_dev543: g+ether? someone still use it?
  • [18:56:04] <beagle_dev543> We use it at work as it saves all the cables and stuff when running USB host (hub + USB ether)
  • [18:56:27] <beagle_dev543> just the 3 connections (USB, serial, DVI )
  • [18:56:55] <beagle_dev543> guess we don't really need serial, but I like to see what's going on :)
  • [18:57:35] <prpplague> hmm, okie dokie
  • [18:57:53] <prpplague> i guess that makes the decision for now
  • [18:57:58] <hrw> prpplague: what else on expansion?
  • [18:58:40] <prpplague> hrw: second serial port, 4-bit standard sd/mmc card slot, battery backed RTC, and enc28j60 10BaseT
  • [18:59:16] <hrw> prpplague: was not there 8bit mmc port available?
  • [18:59:28] <prpplague> hrw: yea there is, but.......
  • [18:59:39] <prpplague> hrw: problem is going to be the level shifting
  • [18:59:42] <hrw> ko
  • [18:59:48] <prpplague> hrw: which will require control signals
  • [19:02:12] * jrmuizel__ (n=jrmuizel@mozca02.ca.mozilla.com) Quit ()
  • [19:02:38] * rbelem (n=rodrigo@pdpc/supporter/active/rbelem) Quit (Remote closed the connection)
  • [19:03:00] * mib_0k3xww (i=47091882@gateway/web/ajax/mibbit.com/x-6fb0fc3fcab38a2e) has joined #beagle
  • [19:05:17] <k2340> can anyone help me get svideo working? Using DSS2 I believe.
  • [19:05:25] <k2340> on angstrom
  • [19:05:27] <prpplague> ds2: ping
  • [19:05:48] * XorA (n=XorA@94.193.93.235) has joined #beagle
  • [19:12:16] * Xenion (n=robert@p579FCF6C.dip.t-dialin.net) has joined #beagle
  • [19:16:12] <koen> prpplague: 10baseT would probably be ok, if I want more speed I can use usb
  • [19:16:44] <prpplague> koen: we'll have the 100baseT later when the driver is more usuable
  • [19:17:20] <hrw> prpplague: how much for board?
  • [19:17:23] <prpplague> koen: we are going to do a quick re-layout of the board, any last minute suggestions?
  • [19:17:26] <prpplague> hrw: $79.00
  • [19:18:02] <prpplague> hrw: we have two other expansion boards in the works, a LCD board and an arduino board
  • [19:18:47] <hrw> prpplague: excluding rtc I would rather go for usb... but first have to switch to C2 revision to get better usb port
  • [19:18:53] * mib_0k3xww (i=47091882@gateway/web/ajax/mibbit.com/x-6fb0fc3fcab38a2e) Quit ("http://www.mibbit.com ajax IRC Client")
  • [19:19:22] <koen> prpplague: arduino or sanguino?
  • [19:19:30] <prpplague> hrw: sorry i did not understand your statement
  • [19:19:39] <prpplague> koen: the plan was standard arduino
  • [19:19:47] <koen> emeb: could you try again with narcissus?
  • [19:20:22] <hrw> prpplague: I wanted to say that I will rather be not good client for such expansion board.
  • [19:21:22] <prpplague> hrw: ahh, well yea, there are various ones we plan to make
  • [19:21:25] <emeb> koen: sure!
  • [19:22:03] <koen> emeb: populating the fat partition required mounting, which doesnt always go right, so I added a cached version
  • [19:22:11] * josch (n=josch@unaffiliated/josch) has joined #beagle
  • [19:22:17] <josch> hello there!
  • [19:22:43] <josch> just to confirm: binaries i build on beagleboard arm cortex are compatible with, say arm9 ?
  • [19:22:52] <hrw> josch: no
  • [19:23:01] <josch> why? arent both armel?
  • [19:23:08] <beagle_dev543> I remember now, I didn't get nfs working from the SD card I went stright to booting over nfs
  • [19:23:19] <hrw> josch: binaries optimized for cortexa8 will not work on armv5te (arm926)
  • [19:23:28] <hrw> josch: is i586 same as i686?
  • [19:23:37] <josch> hrm.. okay
  • [19:23:42] <hrw> josch: will SSE3 application run on 486?
  • [19:23:49] <josch> i get it ;)
  • [19:23:53] <emeb> koen: using loopback to mount the fat partition?
  • [19:24:31] <hrw> koen: when you will have a time - check 2.6.29 on your ep93xx
  • [19:24:54] <josch> how then can debian release binaries that work on both?
  • [19:25:07] <hrw> josch: because debian builds for armv4t
  • [19:25:16] <josch> ah i see
  • [19:25:19] <hrw> which is arm920t core - the lowest supported
  • [19:26:34] <hrw> josch: basically "armel" (I prefer arm/eabi name) binaries can be done for at least 4 different sub archs
  • [19:27:25] <josch> which are armv4-7
  • [19:27:41] <hrw> yep
  • [19:27:53] <hrw> and armv6 can be with or without vfp
  • [19:28:16] <hrw> armv4t can have vfp (ep93xx ones) but it is nonstandard and most of time broken
  • [19:28:53] <josch> nice - thx for the info hrw
  • [19:30:50] <hrw> btw - is beagleboard capable of displaying on SVideo and HDMI at same time?
  • [19:30:54] <suihkulokki> hrw: the ep93xx fpu is not vfp, it is something completly different
  • [19:30:58] <koen> emeb: mount loop file, but in new uImage, umount, gzip
  • [19:31:31] <hrw> suihkulokki: ok, it was shortcut
  • [19:31:36] <koen> hrw: ep93xx is crunch :)
  • [19:32:14] <emeb> OK - got the images. Any better way to write the SD image to card besides dd if= of= ?
  • [19:32:27] * mib_c7ld5h (i=d470a0d2@gateway/web/ajax/mibbit.com/x-ed79266db5096942) has joined #beagle
  • [19:32:34] <koen> emeb: not that I know off, but I only use the tarballs
  • [19:32:37] * jkridner|work (n=a0321898@nat/ti/x-0bbee3380c1292ef) Quit ("Leaving.")
  • [19:33:08] <emeb> 'k - thanks. Will let you know how it goes. Writing 1GB SD image is kinda slow.
  • [19:33:41] <koen> you could try playing with bs=1M, bs=512K, etc
  • [19:34:02] * koen needs to write 4GB sd image from time to time
  • [19:34:18] <emeb> worth a try.
  • [19:34:23] <koen> even at 15MB/s it's going to take a long timet
  • [19:34:55] <k2340> anyone get S-Video working with Angstrom?
  • [19:35:07] <beagle_dev543> *koen should use NFS :D*
  • [19:38:18] <beagle_dev543> no, but have you modified your kernel commandline to allow more RAM? AFAIK the display controller only allocs enough for the primary display (singled buffered)
  • [19:38:30] <emeb> koen: looks like the fat partion has goodies in it now. Thx.
  • [19:38:44] <hrw> bye
  • [19:38:50] * hrw is now known as hrw|gone
  • [19:38:59] <beagle_dev543> afk for a while
  • [19:39:39] <emeb> looks like 2.5MB/s on my card + USB2.0 writer.
  • [19:41:32] <koen> the fat partition is only ~30MB or so, no need to write the whole card
  • [19:42:04] <koen> run mke3fs on the second partition again (maybe resize it)
  • [19:42:12] <emeb> Yeah - won't next time. Just wanted to make sure I had right uboot & uImage for the narcissus image.
  • [19:42:30] <emeb> (don't know if it matters tho)
  • [19:42:54] * rbelem (n=rodrigo@pdpc/supporter/active/rbelem) has joined #beagle
  • [19:43:27] <koen> I need to see if the uboot is still the right one
  • [19:43:43] <koen> I should probably upgrade the narcissus on to 2009.03
  • [19:54:13] * mckoan|away is now known as mckoan
  • [19:56:40] <k2340> is there anyway to read a boot.scr?
  • [19:58:04] <prpplague> koen / sakoman how important is Wake-On-Lan capability to you guys?
  • [19:58:05] <koen> strings
  • [20:02:12] * jkridner|work (n=a0321898@nat/ti/x-0d6cb79fc16b42c4) has joined #beagle
  • [20:02:59] * jconnolly (n=jconnoll@firebug.buglabs.net) Quit (Remote closed the connection)
  • [20:02:59] * HaveAHen1essy (n=misaacs@firebug.buglabs.net) has joined #beagle
  • [20:03:11] * jconnolly (n=jconnoll@firebug.buglabs.net) has joined #beagle
  • [20:03:49] * lrg (n=lrg@host81-136-218-57.in-addr.btopenworld.com) Quit (Read error: 113 (No route to host))
  • [20:05:11] <koen> prpplague: I have heard of WOL, never used it
  • [20:05:13] <sakoman> prpplague: to me personally, not so important
  • [20:06:27] <koen> jkridner|work: still on the road?
  • [20:06:30] * Xenion (n=robert@p579FCF6C.dip.t-dialin.net) Quit (Read error: 101 (Network is unreachable))
  • [20:06:39] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [20:06:51] <jkridner|work> no. back today.
  • [20:06:59] <jkridner|work> just feeling a little under the weather.
  • [20:07:36] <koen> hayfever?
  • [20:07:45] <solar> WOL is a great option in general I use to all day everyday. But for embedded devices like this, it's maybe not so fitting.
  • [20:08:02] <jkridner|work> more like a cold.
  • [20:08:57] <XorA> hi jkridner|work
  • [20:09:06] <jkridner|work> hi XorA
  • [20:09:42] * mckoan is now known as mckoan|away
  • [20:09:46] <jkridner|work> sorry for being so slow getting the boards given away.
  • [20:10:10] <jkridner|work> should start giving 2 a week this week.
  • [20:10:30] <russ> I imagine WoL could be important in sensor type applications
  • [20:11:01] <koen> jkridner|work: are there 2 applications worthy of a board yet?
  • [20:11:19] <jkridner|work> there are some old entries.
  • [20:11:25] <ZeZu> omap3 really overclocks well, has anyone fried one yet from overclocking ?
  • [20:11:31] <jkridner|work> I am fond of Gentoo ports.
  • [20:12:43] <ZeZu> I haven't done it to the beagle yet, but i overclocked the pandora to 800mhz w/o changing voltage and it doesn't even get any warmer
  • [20:13:49] <ZeZu> was good to know performance scaled nearly linear with processor clock
  • [20:13:56] <ZeZu> (on the emulator)
  • [20:14:10] <koen> jkridner|work: yes, but gentoo ports have already been done, yours and mrus were done a year ago
  • [20:14:30] <solar> koen: not really.
  • [20:14:34] <jkridner|work> I never published mine.
  • [20:14:43] <jkridner|work> not sure where mru's is at.
  • [20:14:54] <jkridner|work> there is the pandora thing.
  • [20:14:55] <solar> the initial port. Sure. Gentoo in tree support is a whole nother ballgame. Autobuilds etc.
  • [20:15:32] <mru> what is there to publish? my /var/lib/gentoo/world file?
  • [20:15:33] <jkridner|work> would love to see FreeBSD get off the ground.
  • [20:15:41] <ZeZu> me as well
  • [20:15:57] <jkridner|work> mru: probably more of a Beagle-oriented how-to.
  • [20:16:07] <jkridner|work> didn't need to modify any .ebuild files?
  • [20:16:09] * beagle_dev543 (i=52011590@gateway/web/ajax/mibbit.com/x-f6842e8c72d2dbed) Quit ("http://www.mibbit.com ajax IRC Client")
  • [20:16:17] <solar> mru: you know what the autobuild system is?
  • [20:17:04] <solar> a beagle would sit there 24x7 building pkgs for the proper chost. Sharing the .tbz2 as well as making what are known as stage3's
  • [20:17:04] <mru> solar: I suppose it's something that builds packages
  • [20:17:35] <jkridner|work> Contest #1 had some other fair entries that could be rewarded.
  • [20:17:49] <koen> jkridner|work: the only entry qualifying for a beagle is the ogre3d thing, but that one is really light on details
  • [20:17:57] <jkridner|work> I think once people actually start to hear about boards going out the door, that the project proposals will get better.
  • [20:18:19] <jkridner|work> yeah, we need to beef up the application requirements.
  • [20:18:39] <koen> I think we should pimp the contest a bit more on the mailinglist and website and wait a week
  • [20:19:08] <jkridner|work> k. I just haven't managed to finish my draft of the announcement.
  • [20:19:42] <jkridner|work> got started with some wiki edits. looks like dirk2 or someone else cleaned them up a bit.
  • [20:20:17] * acasp (n=arne@p5481AC2F.dip0.t-ipconnect.de) Quit (Read error: 113 (No route to host))
  • [20:21:16] * koen pokes russ to register his project
  • [20:23:24] * HaveAHennessy (n=misaacs@firebug.buglabs.net) Quit (Read error: 110 (Connection timed out))
  • [20:23:41] * ArteK (n=Artur@81.15.241.96) has joined #beagle
  • [20:25:04] * alecrim (n=alecrim@189.2.128.130) has joined #beagle
  • [20:26:17] <russ> koen: its kindof due today
  • [20:26:42] <ZeZu> criteria for sponsored projects is the same as for the contests ?
  • [20:28:50] * jix (n=jix@dyndsl-095-033-066-019.ewe-ip-backbone.de) has joined #beagle
  • [20:29:48] <jkridner|work> process is pretty similar, but criteria cannot involved judgment of completed work. instead, ability to convince the "judges" that you can actually complete the task is required.
  • [20:31:22] <koen> and that the task is "good" enough
  • [20:31:34] <jix> i just got a c2 beagle board and want to use the usb hs port...
  • [20:31:35] <koen> "making a led blink" isn't good enough
  • [20:31:45] <jix> ehci is compiled into my kernel...
  • [20:31:54] <jix> but when i plug in a device nothing happens
  • [20:31:57] <jix> no kernel messages at all
  • [20:32:09] * MrBIOS-_ (n=aperez@h65.149.51.74.cable.snbr.sanbrunocable.com) Quit (Read error: 110 (Connection timed out))
  • [20:33:33] * k2340 (n=k2340@lawn-128-61-121-203.lawn.gatech.edu) Quit ("Leaving")
  • [20:34:32] <jix> is there something i have to do to get the port working?
  • [20:34:37] <ZeZu> jkridner|work, koen: and is open source still a requisite? I have one or two things I could submit there, but the main project being the emulator is probably by far the most interesting, and the other developer does not have a board, he is in greece and there is an extremely similar board but they want almost 500 EU for it :| I'm sending him mine and I'll be without and having to rely on pandora board only
  • [20:35:18] * koen wonders why all those emulator folks are so anti open source
  • [20:35:21] <ZeZu> while pandora is nice it has no dvi out, and a few other issues
  • [20:35:30] <ZeZu> we're not anti open source, its just not that time yet
  • [20:35:33] <jkridner|work> I'm pretty reluctant to vote for a closed source project.
  • [20:35:54] <ZeZu> and there are other people involved, who also have their say in it
  • [20:35:56] <Crofton|work> that wouldn't be very beneficial to the community ....
  • [20:36:36] <ZeZu> well that about settles it :)
  • [20:37:31] <XorA> the world needs more opensource emulators, saves re-inventing the wheel
  • [20:37:47] <ZeZu> I will submit either the recompiler, which is open source under bsd license .. or the dvb app i'm working on
  • [20:37:56] <ZeZu> which will be bsd license as well
  • [20:38:19] <koen> it' also funny that the most used excuse is "I'm for opening it, but the *other* dude(s) not"
  • [20:38:29] * FuL|OUT is now known as fulgas
  • [20:38:51] <ZeZu> actually the other "dude(s)" is, but it needs a lot of cleaning, that no one wants to do right now
  • [20:39:30] <ZeZu> there are all sorts of amusing comments, and mismatching styles of programming, and even bits of copyrighted source code we legally have access to but needs to be removed ...
  • [20:39:31] <koen> so open it up, maybe someone will clean it for you
  • [20:40:06] <jkridner|work> one of the internal developers is recommending we get some of the V4L2 folks to submit a project entry.
  • [20:40:16] <jkridner|work> should probably promote this on linux-omap as well.
  • [20:40:22] <jix> where can i get an example kernel config for current beagle boards?
  • [20:40:29] <ZeZu> V4L2 ?
  • [20:40:41] * ZeZu queries google
  • [20:41:00] <ZeZu> ah video for linux
  • [20:41:02] <koen> jix: zcat /proc/config.gz
  • [20:41:15] <koen> jkridner|work: you TI guys and your v4l fixation :p
  • [20:41:41] <jix> koen: i wasn't talking about my current config... but a config that is for the current beagle revision... with usb ehci and that stuff in it
  • [20:42:13] <koen> jix: the kernel in the angstrom demo works on revc with ehci and all
  • [20:42:24] <koen> and has a /proc/config.gz
  • [20:43:16] <jix> hmm is the config file located somewhere online? because overwriting my current setup just to get the config isn't what i want to do
  • [20:43:45] <koen> it's in openembedded
  • [20:47:18] <XorA> koen: you should get a special button on keyboard for that answer :-D
  • [20:48:39] <jix> ok i'm going to clone the whole openembedded git just to get an example config... seems to be no other way...
  • [20:48:50] <likewise> jix: cgit.openembedded.org
  • [20:49:02] <jix> likewise: thanks :)
  • [20:51:31] * ArteK (n=Artur@81.15.241.96) has left #beagle
  • [20:53:32] * zedstar_ (n=john@77-99-69-49.cable.ubr16.haye.blueyonder.co.uk) has joined #beagle
  • [21:01:36] * ScriptRipper (n=mmohring@opensuse/member/MartinMohring) has joined #beagle
  • [21:03:27] * dillera (n=dillera@75.149.106.130) Quit ()
  • [21:05:06] <prpplague> jkridner|work: hey bud!
  • [21:05:53] <jkridner|work> howdy.
  • [21:06:12] <jkridner|work> you must have some sort of project worth getting a board. :) picking up OpenOCD again?
  • [21:06:26] <koen> oh dear, omap3 in torpedos now
  • [21:08:50] <koen> still no omapl1x entry in http://www.arm.linux.org.uk/developer/machines/
  • [21:09:24] <prpplague> jkridner|work: i already have a couple beagle's, but they sit in a box, hehe
  • [21:09:53] <prpplague> jkridner|work: trying to iron out some of the changes for the expansion boards
  • [21:10:17] * alecrim (n=alecrim@189.2.128.130) Quit (Remote closed the connection)
  • [21:10:27] <prpplague> jkridner|work: we decided to scrap the micrel part and go back with the enc82j60 even though it is 3.3v part and only 10mbit
  • [21:10:59] * koen wonders why the leopardboard guys keep uploading mvista filesystems instead of working on kernel drivers
  • [21:11:35] <prpplague> koen: leopardboard?
  • [21:11:39] <jkridner> prpplague: sounds like some going in circles.
  • [21:11:48] <jkridner> http://leopardboard.org
  • [21:12:11] <prpplague> jkridner: yea, the first rev of the board had a few issues that werre unexpected
  • [21:12:54] <prpplague> jkridner: interesting
  • [21:12:58] * prpplague looks at the schematic
  • [21:16:36] <prpplague> jkridner: interesting
  • [21:17:39] <prpplague> jkridner: the adoption of the beagle board has been nothing short of amazing, imho
  • [21:19:14] <prpplague> koen: looks like they have their work cut out for them with kernel support
  • [21:20:13] <koen> prpplague: I already pushed the board file upstream to kevins tree
  • [21:20:51] <koen> prpplague: I have 2.6.29rc7 running on it already, I'm waiting for TI/LI to write the glue for the image sensor
  • [21:21:05] <prpplague> koen: ahh interesting
  • [21:21:18] <prpplague> koen: the ethernet is working?
  • [21:21:22] <koen> yes
  • [21:21:23] <ds2> regarding the continous memory needs of the dsp, look in the ALSA driver modules
  • [21:21:32] <ds2> they had the same needs for allocating DMA memory
  • [21:22:02] <koen> prpplague: ethernet and SD work, but that's kinda lame for a camera :)
  • [21:22:21] <prpplague> koen: interesting
  • [21:22:51] <ds2> prpplague: what was the reason you are scrapping the micrel?
  • [21:23:16] <likewise> prpplague: enc82j60? is that the ethernet part purely driven by errata?
  • [21:23:18] <prpplague> ds2: several reasons, mainly due to the fact that the driver is an absolute mess
  • [21:23:31] <prpplague> likewise: ??
  • [21:23:44] <ds2> prpplague: I see.... guess I will have competition... I was going to be the odd man out and build with the enc82j60
  • [21:24:02] <prpplague> ds2: yea i spoke to rusty about your board
  • [21:24:05] <likewise> prpplague: enc28j60?
  • [21:24:06] <prpplague> ds2: he was interested
  • [21:24:11] <ds2> prpplague: did he get my mail?
  • [21:24:21] <prpplague> likewise: "driven by errata" ?
  • [21:24:33] <prpplague> ds2: he didn't mention it
  • [21:24:38] <likewise> prpplague: lots of errata on that part I remember
  • [21:24:38] <prpplague> ds2: i'll ask
  • [21:24:43] <prpplague> likewise: ahh
  • [21:24:56] <prpplague> likewise: i've not had any problems with our production stuff
  • [21:25:18] <ds2> prpplague: it is cool if he wants to ignore it... only thing I care about is if it got lost in a spam trap
  • [21:25:18] * geckosenator (n=sean@c-71-237-94-78.hsd1.co.comcast.net) has joined #beagle
  • [21:25:18] <likewise> prpplague: you use it in programming dongles?
  • [21:25:42] <prpplague> ds2: i'll check with him
  • [21:25:50] <prpplague> likewise: industrial devices
  • [21:26:24] <prpplague> ds2: are you keeping the schematic close at hand for now?
  • [21:26:47] <ds2> prpplague: what do you mean? as in is it private or do I have it on me?
  • [21:26:54] <prpplague> ds2: private
  • [21:27:30] <ds2> prpplague: there isn't anything really private... I am modeling it off the SFE eval board
  • [21:27:41] <ds2> in fact my debugging proto IS a SFE eval board
  • [21:28:01] <ds2> it is look like I need to change out my level shifter as it isn't cutting it for SPI speeds
  • [21:28:12] <prpplague> ds2: hehe, sorry i meant your MID design
  • [21:29:12] <Crofton|work> omap3 in torpedos?
  • [21:29:14] * josch (n=josch@unaffiliated/josch) has left #beagle
  • [21:29:42] <koen> Crofton|work: http://www.arm.linux.org.uk/developer/machines/list.php?id=2178
  • [21:30:30] <Crofton|work> my friend teaches at the Naval Academy, they have torpedos in the hall ways
  • [21:30:39] <koen> the torpedo should have decent linux support by now
  • [21:30:48] <koen> peter is posting a lot on l-o ml
  • [21:30:58] * nabax (n=nabax@89.129.96.71) Quit (Remote closed the connection)
  • [21:35:02] * mib_1jh7k1 (i=cebe4b08@gateway/web/ajax/mibbit.com/x-0c3c593f17efd05a) has joined #beagle
  • [21:36:07] * luigid (i=629a2d1d@gateway/web/ajax/mibbit.com/x-b02417b6025d25e8) has joined #beagle
  • [21:37:07] <luigid> attempting to run /usr/share/ti-dmai/apps/image_decode_io1_omap3530.x470MV in Angstrom demo - need to know what the available codec options are to pass to the "-c" switch
  • [21:40:34] * abitos (n=nixgibts@p5B2E49C0.dip.t-dialin.net) Quit (Read error: 110 (Connection timed out))
  • [21:40:53] <jix> when i plug a usb 2.0 hub into the usb ehci port... it should generate some kernel message?
  • [21:45:56] <geckosenator> yeah
  • [21:49:38] * as_leep (n=ali_as@ambix.plus.com) has joined #beagle
  • [21:56:22] * __alanc__ (n=a-campbe@nat/ti/x-486cdc4e5e7c65f5) Quit ()
  • [21:58:43] * XorA (n=XorA@94.193.93.235) Quit (Remote closed the connection)
  • [21:59:26] * ali_as (n=ali_as@ambix.plus.com) Quit (Read error: 110 (Connection timed out))
  • [21:59:37] * cbrake is now known as cbrake_away
  • [22:05:53] * BThompson (n=a0193480@nat/ti/x-468ff1546b81edfa) Quit (Remote closed the connection)
  • [22:06:44] * Wiedi (n=wiedi@newton-air.w.fruky.net) Quit ("^C")
  • [22:07:15] * simon42 (n=simon@89.238.65.10) Quit (Read error: 60 (Operation timed out))
  • [22:11:15] <russ> I hate getting into the mm weeds
  • [22:11:23] * simon42 (n=simon@89.238.65.10) has joined #beagle
  • [22:11:26] <russ> I feel like I'm just chanting incantations
  • [22:11:37] * ds2 hands russ a machette
  • [22:12:10] <russ> wait_on_page_writeback, then lock, then wait again...
  • [22:12:31] * Downix (n=downix@cpe-72-184-210-102.tampabay.res.rr.com) Quit (Remote closed the connection)
  • [22:12:36] * ds2 can relate... had to mess with that before, very unpleasant experience
  • [22:13:02] <ds2> russ: what are you trying to do in the MM?
  • [22:13:25] <russ> I'm using compressed ram, along with a large movable region
  • [22:13:46] <ds2> your own patch?
  • [22:14:03] <russ> so I drop the size of my movable region, and allow shrink_all_pages along with compcache reduce the number of pages
  • [22:14:12] <russ> I'm using compcache for the ram compression
  • [22:14:24] <russ> compcache only allocates memory in the non-movable region of ram
  • [22:14:51] <russ> after I open things back up again, the memory compcache is using in the non-movable region can't really be freed
  • [22:15:14] <russ> arm wants 16k contiguous pages for fork(), and the system doesn't have any in the non-movable region
  • [22:15:22] <ds2> compcache was the project I was thinking about...nevermind
  • [22:15:41] <russ> so I have two paths, make the compcache allocations migratable
  • [22:15:42] * Downix (n=downix@cpe-72-184-210-102.tampabay.res.rr.com) has joined #beagle
  • [22:16:03] <russ> or, install a oom callback in compcache that paradoxially uncompresses pages and thus reduces the total free ram
  • [22:16:23] <russ> but increases the number of non-movable pages
  • [22:16:42] <russ> turns out forcing pages out of swap and then reclaiming the extents is a bit difficult
  • [22:17:48] <ds2> that is very similar to the stuff I was looking at...never did quite solve it
  • [22:18:50] <Downix> hey ds2, how did you interface your lcd?
  • [22:18:53] <Downix> ment to ask
  • [22:19:08] <ds2> how? I put down wires...
  • [22:19:19] <ds2> not sure how to answer it in a meaningful way
  • [22:19:20] <russ> it looks like that even though compcache has support for block discarding, the kernel never utilizes it for cases where cached pages are no longer in the system anywhere
  • [22:19:48] <Downix> ds2: any guides you can suggest, never delt with a raw LCD before
  • [22:19:53] <ds2> russ: isn't that the lazy discard incase it is needed later?
  • [22:20:33] <russ> ds2: imagine the unshared pages of a user process that gets reamed
  • [22:20:46] <russ> ds2: they stay in the compcache until the extents are reused
  • [22:20:57] <ds2> Downix: sorry, nothing there... I got lucky and was forced to design a LCD controller from scratch back in school so it is all just stuff in my head
  • [22:21:25] <Downix> oh cool
  • [22:21:36] <ds2> russ: that is prefectly fine on a desktop system... but that behavior could be bad if you are trying to consolidate memory into a single bank to turn off other banks for PM purposes
  • [22:21:37] <Downix> I delt with CRTs so LCD's are new to me
  • [22:21:49] <Downix> however, I did find a 4" CRT
  • [22:22:10] <ds2> well the basics are -
  • [22:22:19] <ds2> clock - this goes off on ever pixel
  • [22:22:24] <russ> ds2: I'm thinking maybe a thread or trigger that scans the thing for unmapped extents
  • [22:22:52] <jix> hmm what can i do if i don't get any response from the kernel to stuff put into the ehci port?
  • [22:22:58] <ds2> you clock in things line by line; at the end of each line, there is another clock signal that goes off; once the entire screen is clocked in,t here is another signal that goes off
  • [22:23:05] <ds2> that's pretty all there is to LCDs
  • [22:23:34] * ddompe (n=ddompe@200.122.155.113) Quit ()
  • [22:23:52] <ds2> Downix: so it is a lot like an LCD but things are discrete
  • [22:24:09] <Downix> ok
  • [22:24:37] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
  • [22:24:51] <ds2> s/an LCD/a CRT/
  • [22:24:53] <ds2> blah
  • [22:25:55] * brijesh (n=bksingh@192.94.94.106) Quit ()
  • [22:28:51] * bar_foo (n=falseadd@194.197.235.70) has joined #beagle
  • [22:29:14] * luigid1 (n=jld@cpe-98-154-45-29.socal.res.rr.com) has joined #beagle
  • [22:29:49] * chorne (i=45b2df25@gateway/web/ajax/mibbit.com/x-61efe7ea87f86264) has joined #beagle
  • [22:31:07] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit (Read error: 110 (Connection timed out))
  • [22:31:22] * davidm2 (n=David@nat/ti/x-8ea09117634331e8) has left #beagle
  • [22:31:28] <luigid1> Anybody know what the codec options are for image_decode_io1_omap3530.x470MV in the Angstrom demo?
  • [22:33:37] <russ> ds2: I'm very tempted to try changing the compcache allocations to 16kB
  • [22:34:09] <russ> ds2: then at least it couldn't fragment things horribly below order 2
  • [22:35:18] <luigid1> "beagle" know not; "gst_ti" knows not; "angstrom" knows not....
  • [22:35:59] <luigid1> What channel would be most knowledgeable about Angstrom on the Beagle?
  • [22:36:06] * NishanthMenon (n=nmenon@nat/ti/x-08066d90beb9e74a) has joined #beagle
  • [22:38:09] <Crofton|work> likely here
  • [22:38:24] <Crofton|work> but that is a hard question
  • [22:38:30] <luigid1> Bummer!
  • [22:39:36] <luigid1> How about an easier question....?
  • [22:40:53] <luigid1> After bitbake console-image where is the source that was just built?
  • [22:41:05] <Crofton|work> you mean the image?
  • [22:41:14] <luigid1> No the source
  • [22:41:16] <Crofton|work> tmp/deploy/glibc/.......
  • [22:41:21] <Crofton|work> tmp/work
  • [22:41:25] <Crofton|work> sources :)
  • [22:41:31] * robtow (n=rtow1@nat/montavista/x-79cf0a99d94a6be4) Quit (Remote closed the connection)
  • [22:41:42] <jix> what can be wrong if the beagle board just doesn't detect any device in the ehci port?
  • [22:42:06] * robtow (n=rtow1@nat/montavista/x-7c5f56b4263d9195) has joined #beagle
  • [22:42:14] <Crofton|work> is the driver in the kernel?
  • [22:43:16] <russ> hmm...I think part of my problem is that I'm leaking pages
  • [22:44:10] <jix> Crofton|work: yeah just figured it out
  • [22:44:17] <jix> had an old u-boot on my sd card
  • [22:44:21] <Crofton|work> ah
  • [22:44:26] <jix> the current driver requires u-boot to do some setup
  • [22:44:33] <Crofton|work> likely pinmux
  • [22:44:36] <jix> yeah
  • [22:44:41] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [22:44:46] <jix> and i spend like 2 hours recompiling different kernel versions ^^
  • [22:44:57] <Crofton|work> I hate those problems
  • [22:45:10] <Crofton|work> ehci works well on my Overo
  • [22:45:13] <jix> and everytime i turned it on it clearly said... loading u-boot from mmc.... should have realized that earlier ^^
  • [22:45:30] <ds2> russ: try it but there might be side effects
  • [22:45:42] <ds2> EHCI works fine on the RevC beagle
  • [22:45:48] <luigid1> The source - It's not organized like a normal kernel tree?
  • [22:45:51] <ds2> over 100+ users have verified that 2 weeks ago
  • [22:46:00] <jix> ds2: i never questioned that
  • [22:46:12] <ds2> jix: try the ESC image?
  • [22:46:17] <jix> ds2: it works now
  • [22:46:21] <ds2> good
  • [22:46:21] <jix> as i said i had the wrong uboot getting loaded
  • [22:46:34] <jix> the old one i had on my sd card which i used with my b5 and b7 boards
  • [22:46:51] <ds2> the next thing I was going to say is - for some reason there seems to be a defect on some board... I had one and I suspect GregorR had one too
  • [22:47:31] <kleinjt> quit []
  • [22:47:45] <kleinjt> er...
  • [22:48:23] <GregorR> :(
  • [22:48:36] <GregorR> I didn't want to ship my poor beagle to the hospital.
  • [22:48:41] <ds2> GregorR: did you RMA your board?
  • [22:48:45] <GregorR> Yeah.
  • [22:49:11] <ds2> I think you did everything possible to verify it
  • [22:49:26] <ds2> be interesting to see the RCA on the EHCI issue
  • [22:49:37] <GregorR> I don't like hardware problems. They're just rare enough that I /always/ assume I'm wrong :P
  • [22:51:08] <ds2> it is the price of being an early user
  • [22:51:12] <luigid1> After building an OE kernel how do I get the kernel source tree so that I can modify it or develop my own driver code against it?
  • [22:51:28] <ds2> give it 3-4 months for everyone else to find the issues
  • [22:51:49] <GregorR> I bought mine the day it was available :P
  • [22:52:19] * rsalveti (n=salveti@200.184.118.130) Quit (Remote closed the connection)
  • [22:54:57] <chorne> any hardware knowledge out there?
  • [22:55:03] <ds2> ?
  • [22:55:17] <chorne> efficiency of the TPS65950?
  • [22:55:28] <mru> GregorR: well, I've found one bug in the cortex-a8 cpu and encountered already known ones
  • [22:55:30] <ds2> isn't that on the datasheet?
  • [22:55:50] <chorne> yes
  • [22:56:00] <ds2> then?
  • [22:56:02] * JuanG (n=Juan@nat/ti/x-dfc1b2180986187f) has joined #beagle
  • [22:56:07] * JuanG (n=Juan@nat/ti/x-dfc1b2180986187f) has left #beagle
  • [22:56:19] * luigid (i=629a2d1d@gateway/web/ajax/mibbit.com/x-b02417b6025d25e8) Quit ("http://www.mibbit.com ajax IRC Client")
  • [22:56:35] * luigid1 (n=jld@cpe-98-154-45-29.socal.res.rr.com) has left #beagle
  • [22:56:59] <chorne> i'm worried about battery length
  • [22:57:16] <ds2> any specific concerns?
  • [22:57:17] <chorne> and heat disapated
  • [22:57:30] <ds2> based on Russ's work... you got bigger fish to fry for battery life
  • [22:57:39] <chorne> explain?
  • [22:58:05] <ds2> there are other power consumers for idle that I think may shadow the efficiency issues of the TPS65950
  • [22:58:21] * BThompsonD (n=bernie@cpe-72-190-75-99.tx.res.rr.com) has joined #beagle
  • [22:59:47] <chorne> what other consumers? video, audio, etc?
  • [22:59:59] <ds2> LED
  • [23:00:02] <ds2> Xtal
  • [23:00:02] <Downix> I wish I could find a good battery system other than the AAA batteries
  • [23:00:08] <ds2> there might have been others
  • [23:00:12] <ds2> Downix: Li Ion
  • [23:00:24] <ds2> other then fire and safety issues, Li ion is pretty easy
  • [23:00:27] <ds2> ;)
  • [23:00:57] <Downix> ds2: trying to find a source of controllers and batteries for it that don't have patent issues
  • [23:00:59] * florian (n=fuchs@f049005166.adsl.alicedsl.de) Quit ("Verlassend")
  • [23:01:11] * raster (n=raster@enlightenment/developer/raster) has joined #beagle
  • [23:01:14] <ds2> Downix: which patent issues are you worried about?
  • [23:01:22] <chorne> any other experiences of powering the omap with another chip?
  • [23:01:42] <russ> ds2: the biggest surprise for me was the 3.3V rail
  • [23:02:02] <russ> ds2: the level transceivers, the TFP410 in power down mode, and the 3.3V LDO were using a combined ~30mW
  • [23:02:11] <russ> and I think it was mainly the level transceivers
  • [23:02:49] <ds2> russ: hmmmm do you know if the level translators were floating?
  • [23:03:06] <russ> ds2: they can't in normal configurations
  • [23:03:22] <russ> ds2: they have a 1.8V to 3.3V level converter plugged into one end
  • [23:03:37] <russ> ds2: and your computer plugged into the other end
  • [23:03:40] <ds2> russ: right but if both side is high-z....
  • [23:03:57] * chorne (i=45b2df25@gateway/web/ajax/mibbit.com/x-61efe7ea87f86264) Quit ("http://www.mibbit.com ajax IRC Client")
  • [23:03:59] <ds2> this is the tsx0201e right?
  • [23:04:00] <russ> unplugging the computer side does bring consumption down
  • [23:04:27] <russ> SN65C3221
  • [23:04:37] <russ> the txs0102 consumption is very small
  • [23:04:39] * midtown4 (n=krooked@68-189-104-94.dhcp.wtvl.ca.charter.com) has left #beagle
  • [23:04:52] <Downix> ds2: too many lithium control systems have patents, by Sony mostly
  • [23:05:15] <ds2> Downix: there are precanned chips that can avoid some of it
  • [23:05:21] <ds2> Oh the RS-232 translators
  • [23:05:32] <Downix> right
  • [23:06:00] * ds2 reads the data sheet
  • [23:06:33] <ds2> russ: is the package such that you can drive forceoff with a GPIO?
  • [23:06:41] <russ> hmm..I could probably save another half a milliwatt if I switched several smaller things from VIO to VPLL
  • [23:06:47] <russ> ds2: yes
  • [23:08:26] * jix (n=jix@dyndsl-095-033-066-019.ewe-ip-backbone.de) Quit ("...")
  • [23:08:42] <ds2> if that chip is like the maxim stuff, i think the auto power save is for when the PC is disconnected
  • [23:10:55] <Downix> of course running a beagle smartphone off of AAA's wouldn't be too bad
  • [23:11:12] <ds2> I am more interested in AA's
  • [23:11:27] <ds2> they are cheap, and have roughly the same capacity as the common Li prism cells
  • [23:12:04] <Downix> AAA and AA have the same voltage
  • [23:12:08] <Downix> just less total storage
  • [23:12:53] <ds2> I know but AAA's seem to top out at about 800-900mAH whereas AA's are about 2500mAH to 3000mAH
  • [23:13:17] <ds2> but AAA's seems to go closer to $0.50 each vs $0.25 or less each for AA's (alkalines)
  • [23:13:55] * mib_1jh7k1 (i=cebe4b08@gateway/web/ajax/mibbit.com/x-0c3c593f17efd05a) Quit ("http://www.mibbit.com ajax IRC Client")
  • [23:14:39] <GregorR> If you wanted decent battery life, you'd probably need to have 2x2 AAs at least, maybe 2x4.
  • [23:14:56] * robclark (n=robclark@nat/ti/x-6a57a466745d4d27) Quit (Remote closed the connection)
  • [23:15:02] <ds2> 2x4 is what I'd lean toward
  • [23:15:24] <GregorR> Is 4.8V sufficient to run a beagle reliably?
  • [23:15:36] <ds2> I think 4.8V is within spec
  • [23:15:59] <GregorR> But that's the lowerbound, whereas that's the average for AAs >_>
  • [23:16:11] <Downix> Well, AAA will work for ptototyping
  • [23:16:13] <ds2> not for NiMH/NiCd
  • [23:16:28] <Downix> once I prove the concept I'll set it up for AA/CR2V
  • [23:16:31] <ds2> NiMH/NiCd is close to flat for most of its life
  • [23:17:25] <GregorR> Oh really? Nifty.
  • [23:18:03] <ds2> Now... Alkaline is another story....I think for full discharge, you can bring it down to 0.9 or 0.7V
  • [23:18:31] * mib_1jh7k1 (i=cebe4b08@gateway/web/ajax/mibbit.com/x-f2f76936f3a5c433) has joined #beagle
  • [23:18:42] <GregorR> Y'know what would be good? A USB->barrel cable with a builtin 5V regulator, so you could buy a cheapo USB battery pack, wall socket, whatever and not fear random voltage spikes.
  • [23:19:31] <fenn> GregorR: voltage regulators have a minimum forward voltage drop of about 0.3V
  • [23:19:40] <fenn> so if you start out with 5V the best you can expect on the other side is 4.7V
  • [23:19:43] <GregorR> Oh really? Bleh.
  • [23:19:48] <ds2> fenn: you can get them with 0.050 max drop
  • [23:19:49] <GregorR> Didn't know that. Makes sense though.
  • [23:20:03] <fenn> you can use a zener diode to bleed off excess voltage (search for transient voltage suppressor)
  • [23:20:10] <fenn> ds2: how does that work?
  • [23:20:13] <ds2> but some of them may not be unconditional stable
  • [23:20:27] <ds2> fenn: I think that's the current state of art LDO's
  • [23:20:58] <fenn> did geckosenator ever get his board running well on 3.6V?
  • [23:21:17] <fenn> i'd rather run it straight off my lithium batteries than use a boost converter
  • [23:21:51] <ds2> keep in mind that you WILL loose your EHCI port
  • [23:22:01] <fenn> it doesnt like 3.6V?
  • [23:22:08] * GregorR looks up "transient voltage suppressor"
  • [23:22:10] <russ> GregorR: artwork created with a regulator, loose capacitors, a random heatsink, solder, and a bunch of hotglue is far better than some integrated barrel plug
  • [23:22:57] <fenn> you just mean some kind of adaptor cable right?
  • [23:23:11] <GregorR> That was the idea ;P
  • [23:23:12] * prestonw (i=18634e1d@gateway/web/ajax/mibbit.com/x-d0b59d5a016b3d8f) has joined #beagle
  • [23:27:45] * jrmuizel_ (n=jrmuizel@mozca02.ca.mozilla.com) has joined #beagle
  • [23:28:12] * fulgas is now known as FuL|OUT
  • [23:28:39] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit ()
  • [23:30:44] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) has joined #beagle
  • [23:30:49] <russ> GregorR: like this http://www.flickr.com/photos/31208937@N06/3446196700/
  • [23:34:38] <ds2> fenn: here's an example - 0.050V typ. - http://focus.ti.com/docs/prod/folders/print/tps79133-ep.html
  • [23:41:13] <MrBIOS-> ds2: http://www.flickr.com/photos/liberalex/3443641338/
  • [23:41:46] <ds2> MrBIOS-: was it screwed together?
  • [23:42:21] <ds2> and is that a battery on the lower corner away from the RJ45?
  • [23:42:48] <MrBIOS-> ds2: just four screws
  • [23:42:50] <MrBIOS-> yes, yes it is
  • [23:42:59] <MrBIOS-> half the size of a normal CR2032 or so
  • [23:43:03] <ds2> on board RTC I assume?
  • [23:43:23] <MrBIOS-> the two white connectors in the distance are JTAG and 3.3V serial for the u-boot console
  • [23:43:28] <MrBIOS-> the other one is for power
  • [23:43:39] <fenn> ds2 100mA won't power a beagleboard
  • [23:43:44] <ds2> 3.3V?! so there is no useable console initially?
  • [23:43:51] * jrmuizel (n=jrmuizel@mozca02.ca.mozilla.com) Quit (Read error: 110 (Connection timed out))
  • [23:43:57] <MrBIOS-> ds2: I didn't bother attaching to it this time
  • [23:44:09] <ds2> fenn: that is just a sample...there are other bigger parts
  • [23:44:29] <ds2> MrBIOS-: does it come with an adapter cable for those jacks?
  • [23:44:41] <MrBIOS-> ds2: nope, the OEM devkit might
  • [23:44:43] <ds2> esp. the JTAG, it doesn't look like a stock ARM jtag
  • [23:44:56] <MrBIOS-> \its a standard micro-connector of some sort
  • [23:44:59] <ds2> sigh... so their strategy is to nickle and dime you to death
  • [23:45:06] <MrBIOS-> ds2: whose?
  • [23:45:14] <ds2> the board designer
  • [23:45:21] <MrBIOS-> ds2: frankly you seem overtly negative
  • [23:45:27] <MrBIOS-> it's probably worth$0.50
  • [23:45:28] <ds2> at least the BB used standard connectors
  • [23:45:50] <ds2> MrBIOS-: I do not like it when people deviate from stock connectors and not provide it as part of it
  • [23:46:12] <ds2> MrBIOS-: does OpenOCD recogonize it?
  • [23:46:38] <MrBIOS-> ds2: I don't know yet. Chill.
  • [23:47:18] <ds2> MrBIOS: I am both excited and irritated at the same time :D
  • [23:47:23] * mib_1jh7k1 (i=cebe4b08@gateway/web/ajax/mibbit.com/x-f2f76936f3a5c433) Quit ("http://www.mibbit.com ajax IRC Client")
  • [23:47:29] <MrBIOS-> that's allowed :)
  • [23:48:22] <ds2> given what I know now, i am not sure how I would use it without shelling out more $$ or rummaging through stores to find a suitable connector
  • [23:48:47] <ds2> there is a lot of value in using a 0.1pitched connector
  • [23:49:22] <MrBIOS-> ds2: it'd be trivial enough to just solder-tack to the pins
  • [23:50:38] <ds2> MrBIOS-: well, I suppose... it doesn't make me happy to have to do that on a suppose devel platform
  • [23:50:53] <MrBIOS-> ds2: the product I'm using is an end user product
  • [23:50:56] <MrBIOS-> not the SheevaPlug
  • [23:51:13] <ds2> ah okay
  • [23:51:28] <MrBIOS-> it's based on the OEM design
  • [23:51:31] <ds2> I can give more slack in that case... I thought you were using something that was suppose to be a dev platform
  • [23:51:32] <MrBIOS-> more or less identical
  • [23:51:45] <MrBIOS-> nonono, it's the PogoPlug
  • [23:52:07] <MrBIOS-> I'd rather give my $99 to a small start-up than to a company that doesn't give a damn about $99
  • [23:52:24] * rsalveti (n=salveti@189.70.45.223) has joined #beagle
  • [23:52:29] <ds2> but isn't the other guys also selling it for $99?
  • [23:52:38] <MrBIOS-> correct
  • [23:52:56] <MrBIOS-> marvell and their partners don't need my $99
  • [23:53:09] <ds2> actually, I'd be happier if they had not stuffed those 2 connectors on there
  • [23:53:15] <ds2> just leave pads for 0.1" stuff
  • [23:54:12] <ds2> stuffing the connectors on the BB is actually getting to be an irritant
  • [23:54:30] <ds2> reworking stuff cuz the @#@#@%A@#$A%$@#$!@$@#$@#$@ jtag connector is in the way
  • [23:58:41] * guillaum1 (n=Guillaum@AMontsouris-153-1-1-6.w86-212.abo.wanadoo.fr) Quit ("Leaving.")