• [00:00:10] <jkridner> odesus: lots of different levels of "experience". just ask your question.
  • [00:02:36] <odesus> I want to produce an application..
  • [00:02:51] * rsalveti (n=salveti@189.70.119.44) has joined #beagle
  • [00:02:57] <odesus> which running over an rtos
  • [00:03:03] <odesus> like armgstrong or maemo
  • [00:03:10] <odesus> but I'm confused..
  • [00:03:28] <harperrr> jkridner: thanks. sorry for the n00bness
  • [00:03:31] <davr> monoprice.com is great for buying cables (within the US at least)
  • [00:04:07] <odesus> in how produce the application..
  • [00:04:20] <odesus> with Open embedded seems to build your own RTOS..
  • [00:04:41] <odesus> and with code sourcery compiler you produce your own applications..
  • [00:04:45] <odesus> for example..
  • [00:04:48] <odesus> with ECOS rtos..
  • [00:04:49] <davr> eg 6ft HDMI / DVI cable for $4.03: http://tinyurl.com/hdmi2dvi
  • [00:05:18] <odesus> you running the operative system..
  • [00:05:27] <odesus> and don't worry about if rtos works or not..
  • [00:05:40] <odesus> you only produce applications which running over ECOS..
  • [00:05:50] <odesus> and if you have some problems with any driver..
  • [00:06:08] <odesus> just create your driver on ECOS and voil??!! everything is solved..
  • [00:06:17] <odesus> but how about OE or code sourcery compiler..
  • [00:06:58] <odesus> do you construct a monotonic kernel and recompile
  • [00:07:04] <odesus> everytime you need?
  • [00:07:16] <odesus> someone can help me please???
  • [00:07:24] <jkridner> thinking how best to answer.
  • [00:07:40] <jkridner> OE is for building everything.
  • [00:08:05] <ds2> OE is not the only way to go and might be less painful to avoid it entirely
  • [00:08:49] <jkridner> I won't disagree with that, though others might.
  • [00:08:56] <jkridner> still, it solves a lot of problems for me.
  • [00:09:16] <jkridner> it is an approach to building lots of open source packages.
  • [00:09:24] <jkridner> the kernel is just one of those.
  • [00:09:35] <odesus> ok OE works like windows ce then?
  • [00:09:36] <odesus> I mean..
  • [00:09:37] <jkridner> the approach has lots of benefits and a few pitfalls.
  • [00:09:46] <odesus> when you work with windows ce..
  • [00:09:52] <dcordes_> odesus: it's a normal linux, like the kernel you run on your desktop
  • [00:09:59] <dcordes_> real time?
  • [00:10:16] <odesus> yes real time..
  • [00:10:20] <odesus> all special RTOS..
  • [00:10:22] <ds2> no, OE is a tool to automatically down load and build things on a list and then assemble it into a system
  • [00:10:27] <odesus> need to have this characteristics..
  • [00:10:28] <jkridner> the linux kernel could be monotonic or could be modular.
  • [00:10:44] <ds2> problem is like all automation systems, it is not perfect
  • [00:11:01] <odesus> that's the problem!!
  • [00:11:06] <odesus> the automation systems..
  • [00:11:06] <ds2> RT in the linux-omap tree is still relatively immature, IMO
  • [00:11:06] <dcordes_> odesus: openembedded is like gentoo, just for cross compiling
  • [00:11:13] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) has joined #beagle
  • [00:11:15] <jkridner> if you really need an RTOS, Linux may not be the best answer, but it RT performance is relative.
  • [00:11:16] <dcordes_> odesus: it automates the build process.
  • [00:11:28] <odesus> what about ECOS or qnix?
  • [00:11:30] <dcordes_> odesus: but in the end, you have a normal linux distro and kernel which you can also make manually
  • [00:11:34] <ds2> it all comes down to what kind of response you need
  • [00:11:52] <odesus> ok..
  • [00:11:56] <odesus> the principal issue here..
  • [00:12:04] <odesus> is to find something stable..
  • [00:12:06] <odesus> as I mentioned..
  • [00:12:09] <ds2> for more RT performance, you can apply some patches or go with a commercial solution that has it all worked out
  • [00:12:13] <odesus> I have worked with windows CE..
  • [00:12:21] <ds2> CE is not RT
  • [00:12:22] <odesus> in windows CE the methodology was..
  • [00:12:41] <odesus> ds2: CE is an RTOS indeed..
  • [00:12:46] <ds2> I would go further and say CE is not an OS either ;)
  • [00:12:50] <NishanthM> http://en.wikipedia.org/wiki/Comparison_of_open_source_operating_systems might be of some interest -> see the RTOS stuff
  • [00:12:53] <dcordes_> ds2: :D
  • [00:13:00] <odesus> :)
  • [00:13:02] <ds2> VxWorks is more RT
  • [00:13:02] <dcordes_> with that I would agree
  • [00:13:13] <odesus> ok guys let me explain you this..
  • [00:13:14] <odesus> ok.
  • [00:13:18] <odesus> as I told you..
  • [00:13:26] <odesus> when I worked with CE..
  • [00:13:38] <odesus> for example if you need to add some special functionallity..
  • [00:13:47] <odesus> like touchscreen support..
  • [00:13:57] <odesus> you need to recompile the complete image
  • [00:14:07] <odesus> that should be tranfered to SD card..
  • [00:14:14] <ds2> in Linux you can do that or you can have modules
  • [00:14:27] <odesus> and insert in your evaluation system
  • [00:14:40] <jkridner> I disagree with ds2 about CE not being an RTOS, but it isn't one I'd prefer.
  • [00:14:50] <odesus> every time you need to add something special..
  • [00:14:53] <odesus> you need to recompile this..
  • [00:14:54] <jkridner> also, it is rather nice to run Linux on the ARM and an RTOS on the DSP.
  • [00:15:12] <odesus> like DaVinci..l
  • [00:15:19] <odesus> Beagle can run windows CE!!
  • [00:15:24] <odesus> but Windows CE is so expensive!!
  • [00:15:31] <odesus> royalty fees are a nightmare..!!
  • [00:15:35] <odesus> ok..
  • [00:15:51] <odesus> when you want to create..
  • [00:15:54] <odesus> something new.
  • [00:16:01] <odesus> with a stable RTOS like winCE..
  • [00:16:14] <dcordes_> odesus: touchscreen example in linux: you could make the kernel with module support enabled. then you can load the driver during the operating system is running on linux
  • [00:16:20] <dcordes_> beagle I mean
  • [00:16:22] <odesus> you only need to add the programm as an executable..
  • [00:16:24] <NishanthM> http://en.wikipedia.org/wiki/RTEMS -> support for C4x at least..
  • [00:16:51] <odesus> and running over WINCE..
  • [00:16:58] <odesus> withouth no recompilation..
  • [00:17:07] <odesus> 'cos your image is complete..
  • [00:17:18] <odesus> you have an stable rtos in your SD..
  • [00:17:32] <ds2> if WinCE is a RTOS... it should not have the huge jitter underload as shown here:
  • [00:17:37] <ds2> http://www.windowsfordevices.com/articles/AT6761039286.html
  • [00:17:54] <odesus> ok..
  • [00:17:58] <NishanthM> odesus: still cant figure out why u state rtos! u just need an OS which allows u for run time driver/app load and unload
  • [00:18:04] <jkridner> if you were to add new support to a WinCE BSP, then you'd still have to rebuild the BSP.
  • [00:18:23] <odesus> the Real time concept is quite polemic
  • [00:18:28] <odesus> ok forget RTOS..
  • [00:18:35] <jkridner> Linux isn't any worse about that, though the kernel isn't as isolated from the BSP as the drivers can make use of most kernel features.
  • [00:18:36] <odesus> think it just a label..
  • [00:18:43] <odesus> I prefer linux than wince..
  • [00:18:47] <odesus> 'cos is free..
  • [00:18:50] <odesus> and widely support..
  • [00:18:52] <NishanthM> u could think netbsd too
  • [00:19:01] <dcordes_> odesus: I think you can do waht you are planning with beagle
  • [00:19:02] <ds2> just wanted to make sure we were talking about the same thing
  • [00:19:12] <odesus> WINCE is a monopolium
  • [00:19:19] <NishanthM> it is free, bsd license and it has arm support too.. probably not cortex-a8
  • [00:19:34] <odesus> ok ds2..
  • [00:19:37] <odesus> let me explain..
  • [00:19:45] <odesus> I want to have an stable system..
  • [00:20:01] <odesus> and don't worry about if has x, y or z characteristics..
  • [00:20:10] <odesus> just a platform for develpment any stuff..
  • [00:20:14] <odesus> like linux..
  • [00:20:22] <odesus> you have your linux stable in your desketop..
  • [00:20:37] <odesus> you need something like gnomebaker for burning your cd's..
  • [00:20:53] <odesus> gnome baker developers know the linux kernel architecture..
  • [00:20:55] <ds2> not to be a trouble maker but, define stable - a system that doesn't crash (i.e. a non Windows system) or a system that always behaves reasonably regardless of inputs (i.e. a FIR filter is unconditional stable)
  • [00:21:05] <odesus> ok..
  • [00:21:27] <odesus> stable is a system like fedora, ubuntu, suse tthat don't crash and tell me..
  • [00:21:39] <ds2> ah
  • [00:21:45] <odesus> your system is unstable..
  • [00:21:55] <odesus> and it must shut down..
  • [00:22:01] <odesus> like terrible windows!!
  • [00:22:08] <odesus> windows is very unstable..
  • [00:22:11] <odesus> and primitive..
  • [00:22:17] <odesus> linux is strong..
  • [00:22:17] <dcordes_> opkg install bluescreen :)
  • [00:22:29] <odesus> and stable.
  • [00:22:53] <odesus> Idon't have to reboot my pc in a month..
  • [00:22:54] * NishanthM ponders about the drivers which oopss... does that make system stable? in that case probably openbsd claims for stability? hmmm.. stability might be a state of mind.. :D
  • [00:23:21] <odesus> ok.. that's way.. we need an stable system..
  • [00:23:26] <odesus> imagine you're using a beagle board..
  • [00:23:33] <odesus> for data monitoring..
  • [00:23:37] <odesus> or something like ..
  • [00:23:47] <odesus> working with wavelets and stockastic systemes..
  • [00:24:00] <odesus> specially imaging stocastick systems
  • [00:24:14] <odesus> then if you loose a frame ..
  • [00:24:19] <odesus> you loose everything..
  • [00:24:26] <odesus> if your system is unstable..
  • [00:24:37] <odesus> the missing frame will be lost..
  • [00:24:55] <odesus> and many systems works with add and match algorithms..
  • [00:25:07] <odesus> you need to have an stable system here..
  • [00:25:13] <odesus> beagle can do that..
  • [00:25:15] <odesus> and more!!!
  • [00:25:34] <odesus> then ..
  • [00:25:44] <ds2> and hard realtime will help a lot there
  • [00:26:00] * t_s_o (n=tso@173.84-49-129.nextgentel.com) has joined #beagle
  • [00:26:01] <odesus> EXACTLY!!! ds2!! exactly!!! :)
  • [00:26:03] <NishanthM> ok lets see: a) should not loose periodic data-> hard realtime, b) should not crash-> stability.
  • [00:26:11] <odesus> yeah..
  • [00:26:16] <odesus> well concept or real time.
  • [00:26:18] <odesus> is not a problem..
  • [00:26:27] <odesus> as satellital systems..
  • [00:26:36] <odesus> you can use a kind of backup..
  • [00:26:43] <odesus> and if you loose a frame for processing..
  • [00:26:49] <odesus> like WAVELET dsp processing.
  • [00:26:51] <odesus> no matter..
  • [00:26:52] <ds2> for video rates, there might be patches you can add to Linux to give you sufficient HRT ability
  • [00:27:00] <odesus> yeah..
  • [00:27:08] <odesus> linux is faster than windows..
  • [00:27:11] <ds2> for controlling a vector thrust plane, I doubt if there is any patches that will make Linux sufficient
  • [00:27:23] <odesus> but no matter at last..
  • [00:27:26] <odesus> you can have..
  • [00:27:30] <odesus> a backup system..
  • [00:27:32] <odesus> a backup data..
  • [00:27:35] <odesus> this backup data..
  • [00:27:38] <odesus> only receives the frames..
  • [00:27:42] <odesus> if you loose a frame..
  • [00:27:45] * NishanthM gotta go..interesting system.. but ingo molnar's rt patches might be of interest..
  • [00:27:48] <odesus> you will pick up the frame..
  • [00:27:55] <dcordes_> odesus: you are looking to develop medical devices or actually use them with the beagleboard?
  • [00:28:12] <odesus> Beagle board is HUGE..
  • [00:28:21] <odesus> the applications are ENOURMOUS
  • [00:28:28] <odesus> many applications there..
  • [00:28:30] <odesus> medical..
  • [00:28:37] <odesus> with piezoelectric transductors..
  • [00:28:47] <odesus> even in art...
  • [00:28:55] <odesus> beagle board could be useful in art..
  • [00:29:06] <odesus> I can0t imagine the final with beagle board..
  • [00:29:11] <odesus> I'm not exagerated..
  • [00:29:23] <odesus> I'm not overreacting..
  • [00:29:42] <odesus> beagle is flexible platform for creation.
  • [00:29:46] <odesus> well..
  • [00:29:52] <odesus> lets talk about stability..
  • [00:30:03] <odesus> then if I want to download satellite signals..
  • [00:30:06] <odesus> with a downconverter..
  • [00:30:14] <odesus> and other additional DSP processing..
  • [00:30:25] <odesus> and let beagle process the frames.
  • [00:30:30] <odesus> I need and stable system..
  • [00:30:39] <ds2> someone is working on soft radio stuff using the beagle
  • [00:30:49] <odesus> oh..
  • [00:31:04] <odesus> ds2: how do you create your applications?
  • [00:31:12] <odesus> what are the methodology you use?
  • [00:31:12] <dcordes_> gnu radio?
  • [00:31:25] <ds2> I just build it with make
  • [00:31:33] <ds2> dcordes: I think so.
  • [00:31:41] <odesus> Are you using code sourcery?
  • [00:31:43] <ds2> dcordes: ask Crofton when he is around
  • [00:32:06] <ds2> I am using a flavor of GCC based on the CS stuff
  • [00:32:24] <ds2> but I also use the CS stuff when it is handy
  • [00:32:46] <odesus> after you finish your applications..
  • [00:33:01] <odesus> you upload them to rtos?
  • [00:33:03] <odesus> I mean..
  • [00:33:10] <odesus> did you running over it??
  • [00:33:20] <odesus> for example armstrong?
  • [00:34:10] <odesus> did you upload it over armstrong?
  • [00:34:12] <ds2> upload? I guess you can say that... usually for dev, I have a network so I just scp stuff over
  • [00:34:18] <ds2> I don't use angstrom
  • [00:34:29] <ds2> I use my own kernel + a basic root disk
  • [00:34:54] <odesus> oh..
  • [00:34:56] <odesus> ok.
  • [00:44:13] * [-ip-] (n=ts@mnhm-590e08cb.pool.einsundeins.de) has left #beagle
  • [01:14:09] <jkridner> is www.angstrom-distribution.org down?
  • [01:15:57] <Crofton|work> yep
  • [01:16:05] <Crofton|work> linuxtogo is also out
  • [01:17:15] * lgentili (i=bd2fa60e@gateway/web/ajax/mibbit.com/x-921303c5d09d766a) has joined #beagle
  • [01:23:13] * t_s_o (n=tso@173.84-49-129.nextgentel.com) Quit ("Konversation terminated!")
  • [01:26:30] * odesus (n=Widmung@201.165.141.79) has left #beagle
  • [01:32:50] * odesus (n=Widmung@201.165.141.79) has joined #beagle
  • [01:34:34] <odesus> what jtag interface will be useful for working with DSP module of OMAP?
  • [01:34:45] <odesus> xds510 or xds560
  • [01:34:59] <odesus> even for working with ARM..
  • [01:35:05] <odesus> any suggestion?
  • [01:35:36] <lgentili> both works, 560 has higher speed and it's alot more expensive
  • [01:35:43] <odesus> very expensive..
  • [01:35:46] <odesus> any alternative?
  • [01:35:55] <lgentili> www.00ic.com
  • [01:36:07] <lgentili> cheap tools from china ;)
  • [01:36:18] <lgentili> compatible with xds510
  • [01:36:58] <odesus> oh.. but is it CE certified?
  • [01:37:16] <odesus> also were to find it in america?
  • [01:37:41] <lgentili> I bought some from ebay
  • [01:37:51] <lgentili> just search "jtag TI" in ebay
  • [01:38:00] <lgentili> I payed 200 for a 510 compatible
  • [01:38:00] <odesus> and how much?
  • [01:38:08] <odesus> oh my gosh..
  • [01:38:09] <lgentili> and worked perfect
  • [01:38:11] <odesus> very affordable..
  • [01:38:14] <lgentili> yep
  • [01:38:16] <odesus> speed?
  • [01:38:40] <lgentili> jtag 12Mhz I think
  • [01:38:43] <lgentili> let me check
  • [01:39:04] <odesus> ok..
  • [01:40:17] <jkridner> Crofton|work: not sure where the dialog ended on mirroring/hosting the site. these outages are maddening.
  • [01:41:06] <lgentili> http://cgi.ebay.com/NEW-USB-XDS510-TI-DSP-JTAG-TMS320-Emulator-Programmer_W0QQitemZ170241579408QQihZ007QQcategoryZ4661QQtcZphotoQQcmdZViewItemQQ_trksidZp1742.m153.l1262
  • [01:42:01] <lgentili> The best performance/cost tool is this: http://cgi.ebay.com/NEW-Softronics-ACE-USB-2-0-Emulation-Emulator-Card-JTAG_W0QQitemZ290217823047QQihZ019QQcategoryZ4663QQrdZ1QQssPageNameZWD1VQQcmdZViewItemQQ_trksidZp1638Q2em118Q2el1247
  • [01:42:08] <lgentili> 20Mhz jtag speed
  • [01:42:28] <lgentili> sorry for the long link :O
  • [01:42:31] <odesus> no problem..
  • [01:42:32] <odesus> oh..
  • [01:42:48] <odesus> what do use the second one?
  • [01:42:56] <odesus> is similar like xds560
  • [01:43:11] <lgentili> yep, I will buy 1 next month
  • [01:43:23] <lgentili> but for work with F28335, not the beagle
  • [01:43:43] <lgentili> try to surf the manufaturer webpage
  • [01:43:46] <odesus> f28335? what is that?
  • [01:43:49] <lgentili> that tool is really good
  • [01:44:43] <lgentili> TMS320F28335 it's a DSC (digital signal controller) with float point running at 150 mhz.
  • [01:45:09] <odesus> oh..
  • [01:45:10] <odesus> nice..
  • [01:45:12] * Olipro__ is now known as Olipro
  • [01:45:30] <lgentili> It's like a DSP but the analogy of a MPU (microprocessor unit) vs MCU (microcontroller unit).
  • [01:45:36] <lgentili> has flash peripherials ,etc
  • [01:45:43] <lgentili> but with a FLOAT POINT!! :)
  • [01:46:21] <lgentili> TI is the best :)
  • [01:46:39] <odesus> yes.. there's no better thing that TI..
  • [01:46:44] <lgentili> all return to the root, my first computer was a TI 99 when I was 8
  • [01:46:46] <odesus> cost/functinoalibty..
  • [01:46:59] <lgentili> I agree
  • [01:47:09] <odesus> but if you're thinking in complex systems..
  • [01:47:16] <odesus> I prefer FPGA..
  • [01:47:26] <odesus> but they're not suitable for commercial applications..
  • [01:47:33] <lgentili> well I don't have the excuse to use 1 :(
  • [01:47:49] <lgentili> All I need can be done with a DSP or MCU
  • [01:47:51] <lgentili> so :(
  • [01:47:56] <lgentili> don't have a chance
  • [01:48:03] <odesus> yes you have a chance..
  • [01:48:15] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [01:48:20] <odesus> virtex III
  • [01:48:22] <odesus> not expensive..
  • [01:48:35] <lgentili> how much in 100 units? for example
  • [01:48:49] <odesus> well.. not so much affordable.
  • [01:48:50] <odesus> hahahah
  • [01:48:52] <odesus> ;)
  • [01:48:53] <lgentili> hahah
  • [01:49:02] <odesus> It's for special things..
  • [01:49:08] <odesus> like artificial intelligence..
  • [01:53:04] <odesus> lgentili
  • [01:53:17] <lgentili> yep
  • [01:53:18] <odesus> but It cant support an OS ?
  • [01:53:23] <odesus> your board..
  • [01:53:29] <odesus> F28335
  • [01:53:43] <lgentili> yep, dsp-bios
  • [01:53:52] <lgentili> it's the TI real time os
  • [01:54:02] <odesus> with human-machine interface?
  • [01:54:49] <lgentili> the dspbios it's for controll or that kind of things
  • [01:54:54] <lgentili> not for GUIs
  • [01:55:09] <lgentili> but can be done of course
  • [01:55:18] <lgentili> but it's not off the shelf
  • [01:55:56] <odesus> but what about omap..
  • [01:56:03] <odesus> OMAP3 can do that..
  • [01:56:13] <odesus> well It doesn't have a ADC's
  • [01:56:16] <lgentili> I like the idea of linux for GUIs, and MCU or DSP for control
  • [01:56:43] <lgentili> Linux with a ARM or PPC
  • [01:56:56] <odesus> yep is possible..
  • [01:57:00] <odesus> OMAP3
  • [01:57:02] <odesus> but..
  • [01:57:22] <odesus> why choose F28XXX
  • [01:57:43] <lgentili> because I need a low cost standalone module
  • [01:58:27] <odesus> yes.. the implementation of your own board..
  • [01:58:29] <odesus> mmm
  • [01:58:32] <lgentili> you have to put the PCB in your cost :P
  • [01:58:45] <lgentili> omap3 is expensive for manufature the board and test
  • [01:58:47] <odesus> yes.. that's I was thinking..
  • [01:59:06] <odesus> well the board manufacturing..
  • [01:59:11] <odesus> the principal problem..
  • [01:59:19] <odesus> specially if you want to use VIP technology..
  • [01:59:35] <odesus> but soldering is not a problem..
  • [01:59:59] <lgentili> yes if you have to solder a pop chip or a microBGA
  • [02:00:22] <odesus> nop.. with IR rework stations..!!
  • [02:00:28] <odesus> just a kiddo play..
  • [02:00:39] <odesus> very easily..
  • [02:00:42] <odesus> and trustable..
  • [02:00:49] <lgentili> I don't have those tools
  • [02:01:10] <odesus> you can purchase it!!
  • [02:01:14] <odesus> very cheap..
  • [02:01:18] <odesus> for all comunity..
  • [02:01:25] <odesus> tornado infra!!
  • [02:01:28] <odesus> check it!!
  • [02:01:34] <lgentili> give me a link
  • [02:01:35] * nooomem (n=chatzill@ip68-100-201-32.dc.dc.cox.net) has joined #Beagle
  • [02:02:00] <odesus> http://www.tornadoinfra.com/
  • [02:02:02] * grummel69 (n=christia@cl-13.dus-01.de.sixxs.net) has joined #beagle
  • [02:02:04] * grummel (n=christia@cl-13.dus-01.de.sixxs.net) Quit (Network is unreachable)
  • [02:02:23] <odesus> just $600 for a complete equipment..
  • [02:02:30] <odesus> talking about this..
  • [02:02:30] * grummel69 is now known as grummel
  • [02:02:54] <odesus> Does anybody knows which company can manufacture a PCB with VIP via technology?
  • [02:02:58] <odesus> at affordable prices?
  • [02:03:15] <odesus> Do you know jkridner?
  • [02:03:18] <lgentili> looks nice, but I think it's not possible to center a microBGA chip
  • [02:03:27] <odesus> yes is possible..
  • [02:03:34] <odesus> remember the thermal tension..
  • [02:03:35] <lgentili> have you done it?
  • [02:03:40] <odesus> yes..!!
  • [02:03:45] <jkridner> VIP?
  • [02:03:54] <jkridner> via-in-pad?
  • [02:03:59] <odesus> yes!!
  • [02:04:02] <odesus> via inpad..
  • [02:04:09] <odesus> Is it expensive?
  • [02:04:25] <odesus> lgentili..
  • [02:04:26] <jkridner> yes, it is a bit expensive, as far as I know.
  • [02:04:42] <odesus> any company jkridner?
  • [02:05:03] <ds2> how does a via in pad work? doesn't solder get sucked through the via?
  • [02:05:05] <lgentili> I think the cost of a single board is between 500 and 1000 :O
  • [02:05:25] <jkridner> can be.
  • [02:05:30] <odesus> oh my gosh..
  • [02:05:32] <odesus> ds2
  • [02:05:33] <odesus> nop..
  • [02:05:43] <odesus> its filled with epoxic
  • [02:05:51] <odesus> there's no vaccum in there..
  • [02:06:07] <ds2> capillary action, the same force that self aligns bgas
  • [02:06:12] <jkridner> we've had some quoted by a hand-ful of PCB shops in Texas.
  • [02:06:21] <lgentili> the only chance we have to play with OMAP3 is beagle board
  • [02:06:35] <lgentili> if not be ready to pay $1000
  • [02:06:45] <odesus> well not like that..
  • [02:06:51] <ds2> and did that quote had to be approved by the entire board of directors? ;)
  • [02:06:56] <odesus> I think there should be other companies..
  • [02:06:58] <lgentili> or wait openpandora but will be 350
  • [02:07:11] <ds2> Zooms are about 800ish, IIRC
  • [02:07:13] <lgentili> ds2: hehe
  • [02:07:30] <jkridner> You can also use the SOM out of the Zoom.
  • [02:07:31] <nooomem> lgentili: where can you get more info on openpandora?
  • [02:07:40] <jkridner> which should be much less than the Zoom itself.
  • [02:07:41] <odesus> openpandora.org
  • [02:07:47] <lgentili> I have a IMX31 zoom board, but I don't like their support
  • [02:07:57] <jkridner> you can also wait for the Gumstix or Cogent boards.
  • [02:07:58] <odesus> ah.. motorola..
  • [02:08:00] <lgentili> check their forum
  • [02:08:03] <odesus> IMX31
  • [02:08:09] <odesus> i prefer Texas.
  • [02:08:10] <lgentili> there are like 10 post only about thery boar
  • [02:08:16] <lgentili> their board
  • [02:08:24] <ds2> doesn't Cogent require buying their official SDK before they will sell you any boards?
  • [02:08:33] <odesus> nop..
  • [02:08:39] <odesus> I have to try to by one board.
  • [02:08:47] <odesus> you need to belong to a big company..
  • [02:08:53] <lgentili> because the fault of TI, I have my imx31 board with dust
  • [02:08:55] <odesus> and buy them many boards..
  • [02:08:58] <odesus> ds2
  • [02:09:10] <lgentili> after saw the omap3, I loose my motivation with the imx31 :(
  • [02:09:21] <odesus> I work with xilinx..
  • [02:09:26] <nooomem> openpandora is 350?
  • [02:09:36] <odesus> and after saw omap..
  • [02:09:39] <odesus> not anymore..
  • [02:09:40] <odesus> hahaahah!
  • [02:09:54] <ds2> the iMX31 is a 'fine' chip =) and their PDK is closer to a useable finish device
  • [02:09:56] <lgentili> nooomem: yep
  • [02:10:04] <lgentili> nooomem: 3530
  • [02:10:05] <nooomem> lgentili: where do you see that?
  • [02:10:34] <lgentili> ds2: yep but it's not so fast only 400 bogomips
  • [02:10:53] <lgentili> www.openpandora.org
  • [02:11:02] <lgentili> nooomem: check the forum and blog
  • [02:11:39] <nooomem> lgentili: so how much slower is it compared to the latest beagleboard?
  • [02:12:05] <lgentili> dunno, I think like omap3 x3 faster
  • [02:12:21] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [02:12:52] <odesus> why don't think in made your own board?
  • [02:13:05] <lgentili> with what cpu?
  • [02:14:01] <lgentili> I can't affort to "try" to make work an omap3
  • [02:14:07] <odesus> OMAP3!!
  • [02:14:11] <lgentili> need tools, time, money
  • [02:14:19] <odesus> well think in one thing..
  • [02:14:20] <lgentili> stop dreaming :P
  • [02:14:26] <odesus> ahahahh :)
  • [02:14:32] <odesus> lgentili :(
  • [02:14:34] <odesus> :)
  • [02:15:12] <Crofton|work> jkridner, angstrom/linuxtogo is another machine
  • [02:15:13] <jkridner> plan on spending $25,000+ to build a PCB.
  • [02:15:35] <lgentili> I tho was expensive, but that much?
  • [02:15:37] <lgentili> wow
  • [02:15:48] <odesus> OH MY GOD!!
  • [02:15:52] <odesus> 25000!!
  • [02:15:55] <odesus> :((
  • [02:15:57] <jkridner> for a few revisions, etc.
  • [02:15:57] <odesus> :(
  • [02:16:05] <lgentili> ic
  • [02:16:05] <odesus> and without VIP technolgoy?
  • [02:16:09] <Crofton|work> paying the people working on the board
  • [02:16:15] <Crofton|work> gn
  • [02:16:35] <odesus> Beagle board is a piece of art..:)
  • [02:16:36] <jkridner> $500-1000 for individual protos (1 board) should be expected.
  • [02:16:45] <jkridner> gn Crofton|work. thanks for the tip.
  • [02:17:07] <odesus> but thinking a bigger board without vip technoolgy?
  • [02:17:26] <odesus> manufacturing with e-teknet?
  • [02:17:32] <odesus> affordable prices..:)
  • [02:17:33] <ds2> unless there are PCB deigns that can be borrowed from the chip vendor ;)
  • [02:17:47] <ds2> biggest problem I see is reliable assembly
  • [02:17:47] <jkridner> don't forget that the OMAP3530 is available with a larger ball pitch (0.65mm with "via channel array")
  • [02:18:09] <lgentili> jkridner: do you have any plan when the new omap3530 package get done?
  • [02:18:27] <jkridner> the via-channel-array allows you to use 0.8mm design rules and escape the device in just 2 signal layers + power and ground (4 layers).
  • [02:18:41] <ds2> does the 0.65 version take accept POP memory on top?
  • [02:18:44] <jkridner> there are samples now. not sure when they hit distribution.
  • [02:19:04] <jkridner> no. the 0.65 mm (16x16 package) does not accept PoP.
  • [02:19:19] <ds2> yuck, SDRAM routing
  • [02:19:22] * ds2 runs in horror
  • [02:19:39] <jkridner> why, oh why, does OE keep rebuilding the tools every time!
  • [02:20:03] <jkridner> ds2 is wise, but there is a reference layout available.
  • [02:20:18] <jkridner> POP is a beautiful thing.
  • [02:20:20] <lgentili> but If I'm not wrong, you said something about make another board with that package?
  • [02:20:37] <odesus> that could be useful..
  • [02:20:38] <odesus> :)
  • [02:20:45] <odesus> I think is possible!!
  • [02:20:47] <jkridner> at some point, we will re-spin Beagle with that package to reduce cost.
  • [02:21:11] <lgentili> I would like to help with something, is there any chance?
  • [02:21:23] <jkridner> we are waiting for the TWL/TPS device to be available in a similar package.
  • [02:21:25] <lgentili> I have a milling machine :)
  • [02:21:41] <odesus> I have a PDR infrarred station if you need it jason!!
  • [02:22:01] <jkridner> I don't know that the milling machine would help all that much, but we will try to make it as much a community effort as possible.
  • [02:22:19] <jkridner> it'll definitely be *after* Rev C (which is still scheduled for November).
  • [02:22:40] <odesus> if you need help just tell me jkridner..
  • [02:23:06] <odesus> I dream to have an omap3 on many devices..
  • [02:23:12] <odesus> medical, communications..
  • [02:24:08] <odesus> home appliances..
  • [02:24:09] <odesus> etc, etc..
  • [02:24:24] <jkridner> OE is killing me by not recognizing what has already been built.
  • [02:24:35] <lgentili> jkridner: this is my machine http://www.lpkfusa.com/protomat/s100.htm
  • [02:24:57] <lgentili> If you wanna try something count with that :)
  • [02:25:05] <ds2> at the risk of repeating myself, i really wish someone made a 1"x1" or so carrier board with pads on the side so it can be assembled onto other things w/o dealing with BGA; just have the OMAP3 + POP memory and maybe the TWL/TPS
  • [02:25:06] <odesus> wow lgentili!!
  • [02:25:29] <jkridner> nice!
  • [02:25:32] <odesus> interesting idea ds2!!!
  • [02:25:38] <odesus> very interesting idea ds2!!
  • [02:26:07] <lgentili> I have a plating bench too, but don't have a press for multilayers :(
  • [02:26:17] <lgentili> I'm saving money to buy one
  • [02:26:27] <odesus> oh my gosh lgentili..
  • [02:26:37] <odesus> but what about copper pour..
  • [02:26:39] <odesus> interconectiviy..
  • [02:27:03] <odesus> do you use cooper for conectivity?
  • [02:27:12] <odesus> is a kind of paste..
  • [02:27:17] <odesus> but I don't remember the name..
  • [02:27:20] <odesus> cooperset!!!
  • [02:27:24] <odesus> coperset!!
  • [02:27:25] <odesus> yes..
  • [02:27:34] <lgentili> nope it's a chemical process what we do
  • [02:27:39] <lgentili> with the plating bench
  • [02:27:46] <odesus> blackshadow?
  • [02:27:53] <odesus> novoringarth?
  • [02:27:56] <odesus> prioringarth?
  • [02:28:03] <odesus> seleo?
  • [02:28:05] <lgentili> don't remember the name
  • [02:28:12] <odesus> ah.. blackhole!!
  • [02:28:26] <odesus> blackhole is the lpkf method..
  • [02:28:41] <odesus> do you have blackhole machine
  • [02:28:42] <lgentili> I'm the programmer guy and my partner is the hw guy :P
  • [02:28:52] <lgentili> but before I used to make the board
  • [02:29:02] <odesus> have you tried..
  • [02:29:12] <odesus> the method of printer transference..
  • [02:29:18] <odesus> is it real good..
  • [02:29:23] <lgentili> yep
  • [02:29:29] <odesus> I have get tracks of .5 mm
  • [02:29:37] <odesus> way thiner!!
  • [02:29:40] <lgentili> I don't hehe
  • [02:29:44] <lgentili> I suxed with that
  • [02:29:56] <odesus> oh..
  • [02:30:04] <odesus> how much do you change the tips?
  • [02:30:11] <lgentili> but I did the first mp3 player free over the net back in 1998
  • [02:30:22] <lgentili> and the hw was a mess
  • [02:30:24] <lgentili> but worked
  • [02:30:40] <lgentili> I don't have the page anymore, was www.mp3ar.com
  • [02:30:43] <odesus> wow!!!
  • [02:30:48] <jkridner> gn all
  • [02:30:54] <lgentili> gn
  • [02:30:58] <odesus> gn
  • [02:31:06] <lgentili> let me check if there is some webpage history
  • [02:31:39] <odesus> how much did cost your protomat?
  • [02:32:07] <lgentili> $20000
  • [02:32:23] <odesus> oh my gosh...
  • [02:32:26] <lgentili> but you know you need, acessories and the vacumm, etc
  • [02:32:35] <lgentili> the movile cart
  • [02:32:42] <odesus> oh my gosh..
  • [02:32:47] <odesus> and do you produce many pcb's?
  • [02:32:52] <odesus> or is fo ryour personal use?
  • [02:32:56] <lgentili> inhouse only
  • [02:33:08] <odesus> for your personal use then..:o
  • [02:33:14] <odesus> :O
  • [02:33:21] <lgentili> nope for the company
  • [02:33:30] <lgentili> but don't sell pcbs
  • [02:33:43] <odesus> then you sell complete solutions..
  • [02:33:47] <lgentili> we do our prototypes
  • [02:34:23] <odesus> oh my gosh..
  • [02:34:27] <lgentili> if you do prototypes often, the machine it pay itself
  • [02:34:29] <odesus> amazing thing..
  • [02:34:32] <lgentili> time and money
  • [02:34:39] <odesus> well I use other method.
  • [02:34:46] <odesus> quite rudementary..
  • [02:34:49] <odesus> but efectively..
  • [02:34:57] <odesus> I Have get .4 mm tracks..
  • [02:35:06] <odesus> the connectivity matter..
  • [02:35:12] <odesus> i s solved with coperset..
  • [02:35:24] <lgentili> you pro!
  • [02:35:25] <odesus> and the plating with tin paste..
  • [02:35:26] <lgentili> ;)
  • [02:35:37] <odesus> but dry the coperset..
  • [02:35:43] <lgentili> I used to solder the vias 1 by 1
  • [02:35:43] <odesus> is difficult..
  • [02:35:48] <lgentili> :(
  • [02:35:59] <odesus> oh my gosh..
  • [02:36:01] <odesus> interesting.
  • [02:36:15] <odesus> which is your company?
  • [02:37:10] <lgentili> other day i will tell you
  • [02:37:12] <lgentili> :P
  • [02:37:47] <odesus> in private.
  • [02:37:51] <odesus> lets talk in private..
  • [02:38:52] <odesus> hahaha
  • [02:38:54] <odesus> :)
  • [02:38:57] <lgentili> hehe
  • [02:39:23] <lgentili> this was me in 1999 haha: http://web.archive.org/web/20000408180856/http://www.mp3ar.com/
  • [02:39:23] <odesus> very intersting..
  • [02:40:28] <odesus> wow!!
  • [02:40:33] <odesus> my respects!!
  • [02:40:50] <odesus> :O
  • [02:41:12] * Beagle9 (n=Beagle9@pool-72-74-27-62.bstnma.east.verizon.net) has joined #beagle
  • [02:41:45] <odesus> oh...
  • [02:41:48] <odesus> excelent..
  • [02:42:32] <lgentili> shitty webpage
  • [02:42:39] <lgentili> but helped me alot in that time
  • [02:43:02] <lgentili> and was really the first CD mp3 player open
  • [02:43:12] <odesus> oh..
  • [02:43:21] <odesus> pretty nice indeed..
  • [02:43:53] <lgentili> thanks
  • [02:45:00] <odesus> all engineers must be practical..
  • [02:45:25] <odesus> specially when they're working to bring to comunity the advantages of embedded ssytems..
  • [02:47:19] <odesus> but just I have worked with xilinx for a long time.
  • [02:47:23] <odesus> neural networks..
  • [02:47:40] <odesus> multicore..
  • [02:47:47] <odesus> etc..
  • [02:47:56] <odesus> well..
  • [02:48:06] <odesus> good nite for all..
  • [02:48:09] <odesus> bye..
  • [02:48:13] <lgentili> gn
  • [02:48:56] * odesus (n=Widmung@201.165.141.79) has left #beagle
  • [02:57:16] * Beagle9 (n=Beagle9@pool-72-74-27-62.bstnma.east.verizon.net) Quit ()
  • [03:04:33] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) Quit ("Trillian (http://www.ceruleanstudios.com")
  • [03:08:52] * lgentili (i=bd2fa60e@gateway/web/ajax/mibbit.com/x-921303c5d09d766a) Quit ("http://www.mibbit.com ajax IRC Client")
  • [03:20:53] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) Quit ("Aloha")
  • [03:23:12] * lgentili (i=bd2fa60e@gateway/web/ajax/mibbit.com/x-be93dd3b7697c12a) has joined #beagle
  • [03:27:06] * nooomem (n=chatzill@ip68-100-201-32.dc.dc.cox.net) Quit (Read error: 104 (Connection reset by peer))
  • [03:29:15] * nooomem (n=ravisy@ip68-100-201-32.dc.dc.cox.net) has joined #beagle
  • [03:33:13] * dcramer (n=davec@dcdsl.ebox.com) Quit (Read error: 60 (Operation timed out))
  • [03:33:53] * dcramer (n=davec@dcdsl.ebox.com) has joined #beagle
  • [03:36:25] * lgentili (i=bd2fa60e@gateway/web/ajax/mibbit.com/x-be93dd3b7697c12a) Quit ("http://www.mibbit.com ajax IRC Client")
  • [03:41:33] * grummel (n=christia@cl-13.dus-01.de.sixxs.net) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * guillaum1 (n=gl@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * ds2 (i=noinf@netblock-66-245-251-24.dslextreme.com) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * ahu (i=ahu@outpost.ds9a.nl) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * zaphand1 (n=Z@85.64.205.76.dynamic.barak-online.net) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * JoeBorn (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * chakie_work (n=jeh@vpn-pub2.hard.ware.fi) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * Pavlov (n=pavlov@shell.off.net) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * shoragan (n=shoragan@debian/developer/shoragan) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * diego__ (n=diego@host-84-222-19-249.cust-adsl.tiscali.it) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * zarquin (n=zarquin@bright-snat.ucc.asn.au) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:33] * j24 (n=j24@ns32433.ovh.net) Quit (brown.freenode.net irc.freenode.net)
  • [03:41:50] * JoeBorn (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) has joined #beagle
  • [03:41:50] * chakie_work (n=jeh@vpn-pub2.hard.ware.fi) has joined #beagle
  • [03:41:50] * Pavlov (n=pavlov@shell.off.net) has joined #beagle
  • [03:41:54] * dcordes (n=dcordes@f049190045.adsl.alicedsl.de) has joined #beagle
  • [03:42:15] * shoragan (n=shoragan@debian/developer/shoragan) has joined #beagle
  • [03:42:15] * grummel (n=christia@cl-13.dus-01.de.sixxs.net) has joined #beagle
  • [03:42:15] * guillaum1 (n=gl@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) has joined #beagle
  • [03:42:15] * zaphand1 (n=Z@85.64.205.76.dynamic.barak-online.net) has joined #beagle
  • [03:42:15] * diego__ (n=diego@host-84-222-19-249.cust-adsl.tiscali.it) has joined #beagle
  • [03:42:15] * ds2 (i=noinf@netblock-66-245-251-24.dslextreme.com) has joined #beagle
  • [03:42:15] * ahu (i=ahu@outpost.ds9a.nl) has joined #beagle
  • [03:42:15] * zarquin (n=zarquin@bright-snat.ucc.asn.au) has joined #beagle
  • [03:42:15] * j24 (n=j24@ns32433.ovh.net) has joined #beagle
  • [03:44:04] <sakoman> ds2: you are describing something like a gumstix
  • [03:46:04] <chakie_work> why would anyone want a gumstix instead of a beagle? except for the size?
  • [03:46:28] * lgentili (i=bd2fa60e@gateway/web/ajax/mibbit.com/x-85241f18d28f3d8c) has joined #beagle
  • [03:49:03] <sakoman> chakie_work: I think ds2 has applications in mind where you want to build custom hw, but don't want to deal with the complexities of BGA
  • [03:49:12] * JJS is now known as JJS[out]
  • [03:49:41] <chakie_work> sakoman: ok, so for manufacturing reasons?
  • [03:49:50] <sakoman> yes
  • [03:49:57] <chakie_work> those i have no idea about :)
  • [03:50:01] <chakie_work> i'm a software guy
  • [03:50:05] <lgentili> sakoman: how long have you been working with linux (drivers)?
  • [03:50:14] <chakie_work> the hardware just arrives on my table :)
  • [03:50:49] <sakoman> lgentili: linux drivers, not long at all
  • [03:51:23] <sakoman> drivers in general? hmm . . . more than 30 years I guess
  • [03:51:34] <lgentili> I have read your comments and you looks very familiar with the linux driver model
  • [03:51:36] <sakoman> makes me feel really old!
  • [03:51:40] <lgentili> haha
  • [03:51:44] <chakie_work> heh
  • [03:52:50] <chakie_work> and i thought i was old... some channels are full of teenagers and there i feel like a dinosaur
  • [03:53:46] * JoeBorn__ (n=jborn@h-69-3-101-194.chcgilgm.dynamic.covad.net) has joined #beagle
  • [03:54:05] * dcordes_ (n=dcordes@unaffiliated/dcordes) Quit (Read error: 110 (Connection timed out))
  • [03:54:12] <sakoman> heh, dinosaurs ruled the earth for a long time ;-)
  • [03:54:24] <lgentili> I have been working with drivers alot too, but always with in-house os or small rtos
  • [03:54:43] * JuanG (n=Juan@nat/ti/x-7e20a1ed99a53f8f) has joined #beagle
  • [03:54:45] <lgentili> I'm trying to learn how they works in linux
  • [03:55:04] * JuanG (n=Juan@nat/ti/x-7e20a1ed99a53f8f) has left #beagle
  • [03:55:11] <sakoman> best way is to just sit down and start writing one
  • [03:55:43] <lgentili> sakoman: I did a driver for 2.6 but just a led driver :(
  • [03:55:43] <lgentili> hehe
  • [03:55:46] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [03:55:59] <chakie_work> lgentili: that's a start
  • [03:56:08] <lgentili> true
  • [03:56:27] <sakoman> starting out with really simple hw is a realy good idea
  • [03:56:39] <lgentili> hey did you know that the second edition of "building embedded linux systems" came out?
  • [03:57:07] <lgentili> I'm thinking to buy that book
  • [03:57:17] <lgentili> what do you think guys?
  • [03:59:02] <sakoman> I bought the first edition, but to be honest after browsing through it right after the purchase I haven't found myself making too much use of it
  • [03:59:27] <chakie_work> lgentili: link?
  • [03:59:38] <lgentili> sakoman: You looks like advanced linux programmer
  • [03:59:44] <lgentili> maybe this is for rookies like me
  • [03:59:45] <chakie_work> i had the first edition at a previous job
  • [03:59:58] <lgentili> chakie: 1 sec
  • [04:00:02] <sakoman> lgentili: looks can be deceiving!
  • [04:00:38] <lgentili> sorry my english kinda sux
  • [04:00:50] <lgentili> deceiveing what that means?
  • [04:00:51] <sakoman> beagle can be a great learning tool
  • [04:01:03] <lgentili> I hope something good :)
  • [04:01:21] <sakoman> umm . . . it means that things aren't always what they look like
  • [04:01:34] <lgentili> ah, ok
  • [04:01:42] <lgentili> chakie: http://oreilly.com/catalog/9780596529680/
  • [04:01:43] <sakoman> i.e. I'm not an advanced linux programmer :-)
  • [04:01:58] <chakie_work> lgentili: thanks
  • [04:02:08] <lgentili> chakie: np
  • [04:02:35] <chakie_work> i found the first edition to be full of promise but a bit messy
  • [04:02:48] <lgentili> sakoman: how often a bdi3000 or bdi2000 woulda accelerate your driver creation process?
  • [04:02:54] <lgentili> or debug
  • [04:03:17] <chakie_work> i'm trying to move more into the embedded linux field, there is no work in the desktop/backend field
  • [04:03:31] <sakoman> since I've never had such a thing I have no idea how to answer that question!
  • [04:03:43] <lgentili> I trying to figure out If it is really necesary
  • [04:03:52] <lgentili> but looks like it does not
  • [04:03:54] <lgentili> :)
  • [04:03:59] <sakoman> being a dinosaur I do things the old fashioned way
  • [04:04:46] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [04:04:49] <lgentili> sorry but desktop never liked to me
  • [04:04:51] <sakoman> think hard, use prinfs, and occassionally pull out the scope or logic analyzer
  • [04:05:06] <chakie_work> ah, you code by 32 switches for the address and 8 for the byte value and one for "submit"?
  • [04:05:07] <sakoman> printfs -- can't spel
  • [04:05:21] <sakoman> yup, just like that!
  • [04:06:09] <chakie_work> sometimes i miss the old c64 days where you mostly coded like that. then again, those days were quite frustrating too
  • [04:06:42] <sakoman> to be honest it doesn't feel all that different to me
  • [04:07:10] <lgentili> and that works when you are trying to make an audio driver?
  • [04:07:12] <lgentili> hmm
  • [04:07:44] <lgentili> how you handle that in linux?
  • [04:08:10] <chakie_work> i've found that i rarely use a debugger to find errors. if i need one then i'm in really deep shit
  • [04:09:00] <sakoman> chakie_work: same here. I mostly use the think hard/printf method
  • [04:09:44] <lgentili> when you say printf you mean printk if you are in kernel space?
  • [04:09:53] <sakoman> yup
  • [04:09:56] <lgentili> ok
  • [04:09:58] <chakie_work> i have tremendous respect for guys that work on low level stuff like bootloaders and similar where you really can't shove in a debugger even if you wanted to
  • [04:11:35] <lgentili> so i don't have excuses I just have to put my hand in the code and get dirty :)
  • [04:11:49] <lgentili> I was waiting to buy a jtag tool
  • [04:11:52] <lgentili> haha :(
  • [04:12:58] <chakie_work> jtag and onwards is where my knowledge just does't cut it anymore :)
  • [04:13:57] <lgentili> sakoman: when you need to stop something what you do? just "for (;;);" hehe
  • [04:14:18] <lgentili> I didn't put that face there
  • [04:14:26] <lgentili> for ( ; ; ) ;
  • [04:14:49] <lgentili> it's a joke, but what's the solution
  • [04:15:58] <lgentili> chakie: did you buy a beagle?
  • [04:16:22] <chakie_work> lgentili: not yet, but i will in a week or so
  • [04:16:41] * chakie_work is just a wannabe hangin' out with the big boys
  • [04:16:49] <lgentili> haha
  • [04:17:04] <lgentili> we can learn alot, from the big boys
  • [04:17:09] <chakie_work> yup
  • [04:17:26] <lgentili> and what's your main plan with the board
  • [04:17:33] <lgentili> I like the 3d part :)
  • [04:17:35] <chakie_work> i look forward to the increased freedom to do whatever i want that a beagle would give me
  • [04:17:51] <sakoman> lgentili: It is quite rare that I run into a situation where I want things to just hang!
  • [04:18:13] <chakie_work> i've uses some linux based "hackable" wifi boxes in the past for own projects, but those are quite limited
  • [04:18:56] <chakie_work> lgentili: for a first test i'll try to replace our media pc with a beagle, then i'll see if i can make a wifi enabled car music player
  • [04:19:26] <lgentili> sakoman: but for example, when you need to see some kind of hw output and don't change the state, like when you need a "step by step"
  • [04:19:28] <chakie_work> i like the idea of having the car sync music when you drive it home :)
  • [04:19:44] <lgentili> chakie: well that's a good use
  • [04:20:14] <chakie_work> lgentili: i guess. but there are quite a few question marks that need to be straightened out
  • [04:20:56] <chakie_work> as in "how to get an ir receiver working?" and "how to connect a some 4-5 keys to the beagle for input?"
  • [04:21:32] <lgentili> chakie: well you gonna enter to the embedded world tru the big door
  • [04:21:41] <chakie_work> i can solder up simple stuff if i find instructions, but i can't figure hw solutions out on my own
  • [04:21:55] <chakie_work> heh
  • [04:22:20] * chakie_work kicks in the front door and yells "hey, where's my candy?"
  • [04:22:26] <lgentili> lol
  • [04:22:38] <sakoman> You weren't far off if you just want to hang using C, you could use something like while(1);
  • [04:23:17] <ds2> NMI signaling imminent power loss would seem like a valid reason to do a for (;;);
  • [04:23:18] <ds2> ;)
  • [04:24:33] <lgentili> sakoman: are you doing the porting for other omap3 board?
  • [04:24:42] <sakoman> yes
  • [04:24:46] <lgentili> sakoman: I don't remember the name...
  • [04:24:59] <lgentili> sakoman: but it's wide use for hobbists
  • [04:25:02] <sakoman> Overo, for Gumstix
  • [04:25:10] <lgentili> yeah that gumstix
  • [04:25:24] <lgentili> it's like you have the monopoly of omap3
  • [04:25:25] <lgentili> :P
  • [04:25:51] <sakoman> hardly! lots of omap3's in cell phones!
  • [04:26:07] <ds2> know any off hand?
  • [04:26:56] <sakoman> umm . . . I was consulting for this company that is using one, so yes
  • [04:27:21] <sakoman> but I forget what is in my nda so I won't mention their name
  • [04:27:56] <lgentili> anyone has tryed to interface a lcd directly to the beagle's pcb?
  • [04:28:21] <ds2> the list of processors in use don't list many things for the OMAP3
  • [04:29:04] <lgentili> maybe I'm the 10000 person asking this :(
  • [04:29:18] <sakoman> lgentili: I wouldn't recommend trying that. It would not be easy mechanically
  • [04:29:31] <ds2> not recommended unless you are very familiar with hacking into SMT leads
  • [04:29:45] <sakoman> If you really need lcd you should look at some of the other omap3 products
  • [04:29:51] <ds2> in theory, the signals should be available at the DVI-D transmitter
  • [04:30:33] <ds2> and that is a leaded part so...
  • [04:31:18] <lgentili> I have an expert doing that kind of operations
  • [04:31:19] <sakoman> ds2: I've had to solder a few wires to that part -- it is not fun!
  • [04:31:50] <sakoman> lgentili: I would really advise against trying to do that
  • [04:32:06] <ds2> sakoman: I did say in theory =) think that's the same pitch as the maxim uQSOP package I think
  • [04:32:07] <lgentili> you are motivating me to do it
  • [04:32:48] <sakoman> You'll likely make a mess of the board and end up with a very suboptimal solution both electrically and mechanically
  • [04:33:12] <lgentili> ok :(
  • [04:33:19] <ds2> for the effort involved, wait for a better board or just blow the $$ on a zoom
  • [04:33:22] <sakoman> But it will make an interesting story and amusing pictures I am sure
  • [04:33:54] <lgentili> Maybe I will do it for the community
  • [04:33:59] <lgentili> ;)
  • [04:34:20] <lgentili> I love to interface lcds to things
  • [04:34:30] <sakoman> You can serve as an example of why they shouldn't take that project on!
  • [04:34:36] <lgentili> I think I have more than 40 different type of lcds
  • [04:34:47] <lgentili> haha
  • [04:34:50] <ds2> there are 3 basic problems you will need to solve
  • [04:35:11] <lgentili> what the bigger for u ds2?
  • [04:35:17] <ds2> 1) Getting to the pins, 2) getting a suitable connector on the LCD, 3) Getting the voltage translators in place since most surplus LCDs are not 1.8V
  • [04:35:36] <ds2> IMO, (2) is a biggest pain
  • [04:35:59] <sakoman> 4. mechanical stability once you have the wires connected
  • [04:36:00] <lgentili> hmm the bigger will be avoid the noise at high clock
  • [04:36:16] <lgentili> that happened to me
  • [04:36:25] <lgentili> the 4. you can glue the wires
  • [04:36:35] <lgentili> it's not nice but works
  • [04:37:22] <lgentili> I have interfaced a 800x480 at with a clock at 40 Mhz.
  • [04:38:20] <lgentili> and if you route different the wires the image change some colors :(
  • [04:38:39] <ds2> humbug to mechanical stability ;)
  • [04:40:01] <lgentili> I've never used a lvds interface... so maybe I'm wrong and can't be dont at 65 Mhz or more
  • [04:40:15] <lgentili> dont = done
  • [04:40:43] <ds2> LVDS? eh? you want to take a chip on there to handle (3)?
  • [04:41:20] <lgentili> yep
  • [04:42:03] <ds2> might I suggest a simplier starting project... transmuting the solder in led to gold
  • [04:42:11] <ds2> lead
  • [04:42:22] <ds2> lead in solder
  • [04:42:49] <lgentili> haha you kidding?
  • [04:42:56] <lgentili> what you talking
  • [04:43:44] <lgentili> you and sakoman are pushing me to do it
  • [04:44:03] <lgentili> to show you it can be done ;)
  • [04:44:39] <lgentili> I gonna see the pcb again to be sure the sizes
  • [04:44:49] <ds2> I know people that can do it, but these are the people that don't blink an eye when you ask them to swap out a 200+ pin QFP or to add 30-40 jumpers onto it so asignal can be probed
  • [04:46:58] <lgentili> do you know what pitch has the tpf4010?
  • [04:47:33] <lgentili> tfp4010
  • [04:47:35] <DDevine> Hey guys
  • [04:47:56] <DDevine> I was thinking last night that we should write a guide on screens for the beagle.
  • [04:48:29] <DDevine> what ya think?
  • [04:49:40] <lgentili> and what the guide will say?
  • [04:51:06] <DDevine> it should present the most viable options
  • [04:51:17] <DDevine> and then get more detailed on those.
  • [04:51:34] <DDevine> and have a section about the "no so viable but doable"
  • [04:52:31] <lgentili> for sure gonna help the newbie users
  • [04:53:17] <DDevine> exactly
  • [04:53:25] <ds2> if people think it is hard to get the current display driver working....
  • [04:53:26] <DDevine> it is something that everybody has to deal with.
  • [04:53:43] <ds2> good luck getting individual undocumented LCDs working
  • [04:54:15] <DDevine> http://elinux.org/BeagleBoard <--- I think the guide will go in there yes? (I will volunteer webspace if not)
  • [04:54:33] <DDevine> DS2 - we dont document how to get them working as such.
  • [04:54:50] <DDevine> we just point out which ones can be used and what would be involved in getting them to work
  • [04:55:40] <DDevine> so we have the types "USB LCD" "Serial LCD" "S-Video LCD(TFT)"
  • [04:56:26] <ds2> wonder if it is just cheaper to implement a USB LCD
  • [04:56:33] <DDevine> then just a summary of the ways they can be connected (and tell people to refer to the LCD's doco)
  • [04:56:48] <ds2> isn't there a MSP430 with both USB and an onboard LCD controller? or PIC or Atmel?
  • [04:56:53] <DDevine> DS2 it is easy (though i haven't done it)
  • [04:57:05] <ds2> DDevine: which is easy?
  • [04:57:12] <DDevine> USB LCDs
  • [04:57:28] <ds2> DDevine: graphical or alphanumeric?
  • [04:57:29] <DDevine> though it isnt an image LCD - text only.
  • [04:57:50] <DDevine> i knew you were going to say that :p
  • [04:58:01] <ds2> well, those are trivial
  • [04:58:15] <ds2> in fact those are easier to do using the GPIO's on the expansion header
  • [04:58:25] <lgentili> true
  • [04:59:33] <DDevine> I am a noob btw...
  • [05:00:53] <lgentili> ds2 where are you from?
  • [05:01:01] <DDevine> I dont know anything about the expansion header.
  • [05:01:15] <DDevine> But USB seems easy to me
  • [05:01:22] <DDevine> i could do Serial though.
  • [05:01:40] <lgentili> the other interfaces are easier
  • [05:01:44] <lgentili> i2c spi
  • [05:02:22] <ds2> lgentili: a blue marble orbiting a unary system =)
  • [05:02:39] <lgentili> ds2: :o
  • [05:03:19] <lgentili> I'm asking you because you handle kinda late here
  • [05:03:35] <DDevine> I saw something about i2c yesterday.
  • [05:04:09] <lgentili> i2c use 2 signals, spi use 3
  • [05:04:24] <lgentili> plus the chip select
  • [05:04:41] <lgentili> you can drive all kind of chips with those 2 interfaces
  • [05:04:55] <lgentili> spi handle higher speed than i2c
  • [05:05:01] * methril|laptop (n=methril@61.73.220.87.dynamic.jazztel.es) Quit (Read error: 110 (Connection timed out))
  • [05:05:27] * methril|laptop (n=methril@78.72.220.87.dynamic.jazztel.es) has joined #beagle
  • [05:05:38] <lgentili> ds2: what's the time where do you live?
  • [05:05:52] <lgentili> here is 2:08 am :O
  • [05:05:56] <ds2> lgentili: I'm in California, USA; it is 2208 PDT
  • [05:06:15] <lgentili> damn you, cali is the best place in the world to live
  • [05:06:36] <lgentili> SB?
  • [05:06:53] <lgentili> I love the weather
  • [05:07:00] <ds2> no, SJ
  • [05:07:07] <ds2> where are you?
  • [05:07:15] <lgentili> Argentina
  • [05:07:58] <lgentili> last time I was in cali was in 2003, learning english in UCSB
  • [05:07:59] <ds2> near penguin land?
  • [05:08:08] <lgentili> but looks like didn't work
  • [05:08:09] <lgentili> haha
  • [05:08:19] <lgentili> nope,
  • [05:08:36] <lgentili> in the center of the country, But right now I'm in brazil
  • [05:08:37] <DDevine> Im in Australia
  • [05:08:45] <Pavlov> australia is nice
  • [05:08:46] <DDevine> its 3:00PM
  • [05:09:51] <lgentili> I don't know these girls eat, here in brazil
  • [05:10:15] <lgentili> but they have the best ass :)
  • [05:10:19] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [05:11:01] <lgentili> well time to sleep
  • [05:11:09] <lgentili> gn all
  • [05:11:33] * lgentili (i=bd2fa60e@gateway/web/ajax/mibbit.com/x-85241f18d28f3d8c) has left #beagle
  • [05:14:48] * JoeBorn__ (n=jborn@h-69-3-101-194.chcgilgm.dynamic.covad.net) Quit (Read error: 110 (Connection timed out))
  • [05:43:36] * bazbell (n=a0192809@nat/ti/x-4d2fa6955f6385dd) has joined #beagle
  • [05:44:29] <DDevine> So, when you have a usb LCD screen - you have to send the data to it through /dev/usbtty0 right?
  • [05:44:40] <DDevine> and that sends the info to the display...
  • [05:45:21] <DDevine> and to do that do you need LCDproc?
  • [05:54:47] * nooomem (n=ravisy@ip68-100-201-32.dc.dc.cox.net) has left #beagle
  • [06:00:41] <DDevine> answered that myself - no.
  • [06:02:07] <DDevine> ...So it is my understanding that with a USB LCD that you just send commands to the controller...
  • [06:17:55] * JoeBorn__ (n=jborn@h-66-167-68-16.chcgilgm.dynamic.covad.net) has joined #beagle
  • [06:48:17] * geraint (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [06:53:45] * trickie (n=trickie@basesoft.xs4all.nl) has joined #beagle
  • [06:55:46] * methril|away is now known as methril
  • [07:03:03] * geraint (n=gnorth@tu008.demon.co.uk) Quit ()
  • [08:00:10] * t_s_o (n=tso@173.84-49-129.nextgentel.com) has joined #beagle
  • [08:08:39] * guillaum1 (n=gl@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
  • [08:11:29] * geraint (n=gnorth@217.207.128.218) has joined #beagle
  • [08:14:44] * valhalla (n=valhalla@81-174-21-16.dynamic.ngi.it) has joined #beagle
  • [08:15:35] * geraint_ (n=gnorth@217.207.128.219) has joined #beagle
  • [08:21:13] <koen> good morning all
  • [08:21:18] <mru> morning
  • [08:21:49] <valhalla> morning
  • [08:26:12] <koen> mru: do you have a link to the patch that fixes system headers?
  • [08:26:25] <mru> no
  • [08:26:37] <mru> but it's really easy to do
  • [08:26:39] <mru> it's one line
  • [08:26:47] <mru> find asm/types.h
  • [08:27:55] <mru> find the typedef for __u64
  • [08:29:02] <mru> it's under #ifndef __STRICT_ANSI__, but shouldn't be
  • [08:34:47] * geraint (n=gnorth@217.207.128.218) Quit (Read error: 110 (Connection timed out))
  • [08:38:08] * shrirama (i=3b6030f5@gateway/web/ajax/mibbit.com/x-3e27362473ced0c2) has joined #beagle
  • [08:45:54] * t_s_o (n=tso@173.84-49-129.nextgentel.com) Quit ("Konversation terminated!")
  • [08:50:07] * shrirama (i=3b6030f5@gateway/web/ajax/mibbit.com/x-3e27362473ced0c2) Quit ("http://www.mibbit.com ajax IRC Client")
  • [09:00:52] * DDevine (n=devine@123-243-187-85.static.tpgi.com.au) Quit (Remote closed the connection)
  • [09:11:00] * JoeBorn__ (n=jborn@h-66-167-68-16.chcgilgm.dynamic.covad.net) Quit (Read error: 104 (Connection reset by peer))
  • [10:06:36] * Foolean (n=emil@foolean.ros.sgsnet.se) has joined #beagle
  • [10:09:51] * Fragguel (n=rafa@240.114.218.87.dynamic.jazztel.es) has joined #beagle
  • [10:18:43] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Read error: 113 (No route to host))
  • [10:29:16] * Fragguel (n=rafa@240.114.218.87.dynamic.jazztel.es) Quit (Read error: 110 (Connection timed out))
  • [10:29:28] * Fragguel (n=rafa@alhakem.uco.es) has joined #beagle
  • [10:41:21] * sweetlilm (n=sweet@196.37.229.15) has joined #beagle
  • [10:47:08] * DDevine (n=devine@123-243-187-85.static.tpgi.com.au) has joined #beagle
  • [10:59:23] <sweetlilm> g'day, serial problem sorted. anyone got some hints on how to get OTG working with the test Angstrom image?
  • [11:01:10] * JoeBorn__ (n=jborn@h-66-167-69-28.chcgilgm.dynamic.covad.net) has joined #beagle
  • [11:01:36] * irq__ (n=irko@62.142.78.245) has joined #beagle
  • [11:02:03] <DJWillis> sweetlilm: care to tell us what the serial issue was?
  • [11:02:55] <sweetlilm> USB<->Serial adapter, TX worked (I could see the boot output), RX didn't. Made up a direct cable and all was good
  • [11:05:17] <DJWillis> Ahhh, cool.
  • [11:06:51] <sweetlilm> what is this (message in the console): "musb_bus_suspend 2129: trying to suspend as a_wait_vrise is_active=1"
  • [11:09:07] * Viral_Sachde (n=Viral_Sa@122.167.152.102) has joined #beagle
  • [11:51:43] <Crofton|work> koen, ping
  • [11:51:53] <Crofton|work> I have gnuradio trunk compiling
  • [11:51:58] <Crofton|work> but there is something ugly
  • [11:52:05] <Crofton|work> PYTHON_CPPFLAGS=-I${STAGING_INCDIR}/python2.5
  • [11:52:11] <Crofton|work> in EXTRA_OECONF
  • [11:52:26] <Crofton|work> what I need to flow into the Makefile is
  • [11:52:41] <Crofton|work> PYTHON_CPPFLAGS=-I${STAGING_INCDIR}/python${PYTHON_VERSION}
  • [11:52:57] <Crofton|work> but I can't figure out the needed quoting ...
  • [11:54:37] * lgentili (i=bd1344dd@gateway/web/ajax/mibbit.com/x-3b8d01aa97905c21) has joined #beagle
  • [15:52:04] * BeagleBot2 (n=PircBot@ec2-75-101-156-174.compute-1.amazonaws.com) has joined #beagle
  • [15:52:04] * Topic is 'Welcome to #Beagle | Discussion about the OMAP3 Beagle Board - http://beagleboard.org | Beagle search tools are on #dashboard at irc.gimp.org, NOT here ;) | Log is at http://www.beagleboard.org/irclogs'
  • [15:52:04] * Set by jkridner on Tue Jun 10 00:33:18 CDT 2008
  • [15:53:14] <jkridner> started a new logbot
  • [15:53:25] * nutbeam (n=nutbeam@smtp.dh-global.com) has joined #beagle
  • [15:54:10] <jkridner> old bot is still running, so I should be able to recover the last 4 days if I could just get into that machine.
  • [15:56:49] <jkridner> all, please let me know if you see something missing/broken.
  • [15:58:03] <jkridner> dirk2: I seem to have lost the FAQ fixes.
  • [16:00:33] <nutbeam> Hi all
  • [16:04:57] <nutbeam> Can I ask about the TWL4030?
  • [16:05:48] <kulve> just ask and wait. Maybe somebody answers if he knows the answer..
  • [16:06:44] <kulve> although this is quite active channels nowadays, so people probably don't read backlogs..
  • [16:06:46] <nutbeam> According to TI the TWL4030 chip is now obsolete. Has there been a replacement picked for it yet, for use on the beagleboard?
  • [16:07:23] <nutbeam> I found this out trying to find a data sheet for the TWL4030
  • [16:07:33] <kulve> iirc, beagle uses twl4030
  • [16:07:44] <DJWillis> sakoman: cool, as those changes should shrink the board code down I will follow and do a patch set once they it is done then.
  • [16:09:01] <nutbeam> as that chip is obsolete has there been a replacement chip selected?
  • [16:09:28] <kulve> well, afaik not for the boards already on the market and not for the rev c boards
  • [16:09:29] <jkridner> tps65950 is a drop-in replacement
  • [16:10:12] <jkridner> as a user, there should be no difference to you whatsoever.
  • [16:10:13] <sakoman> DJWillis: yes, makes sense to wait. will be *much* less work for you :-)
  • [16:10:15] * NishanthMenon (n=gnat@nat/ti/x-ce6285da028f7ef5) has joined #beagle
  • [16:10:16] <nutbeam> Thanks.
  • [16:10:53] <nutbeam> Much appreciated.
  • [16:12:19] <nutbeam> I was wondering how power management might be affected. Evidently it won't be :-)
  • [16:16:04] * t_s_o (n=tso@173.84-49-129.nextgentel.com) Quit ("Konversation terminated!")
  • [16:44:35] * DDevine (n=devine@123-243-187-85.static.tpgi.com.au) Quit (Remote closed the connection)
  • [16:46:14] * nutbeam (n=nutbeam@smtp.dh-global.com) has left #beagle
  • [16:46:35] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [16:52:41] * geraint_ (n=gnorth@217.207.128.219) Quit (Read error: 113 (No route to host))
  • [16:59:53] * Olipro (i=Olipro@unaffiliated/olipro) Quit (Read error: 104 (Connection reset by peer))
  • [17:00:18] * Olipro (i=Olipro@unaffiliated/olipro) has joined #beagle
  • [17:00:52] * BeagleBot (n=PircBot@ec2-75-101-156-174.compute-1.amazonaws.com) Quit (Read error: 110 (Connection timed out))
  • [17:01:56] * Viral_Sachde (n=NetUser@122.167.202.18) has joined #beagle
  • [17:03:06] * guillaum1 (n=gl@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) has joined #beagle
  • [17:03:49] * max_fr (i=max@212-198-79-163.rev.numericable.fr) has joined #beagle
  • [17:03:58] <max_fr> Hi!
  • [17:26:51] * neelesh_gurjar (n=neel@123.236.29.189) has joined #beagle
  • [17:27:00] <neelesh_gurjar> Hi All
  • [17:27:16] <neelesh_gurjar> I am new to Beagle Board
  • [17:34:25] * khilman (n=khilman@166.129.170.45) has joined #beagle
  • [17:43:19] <neelesh_gurjar> I want to buy Beagle Board
  • [17:43:38] <neelesh_gurjar> I tried link from beagleboard.org... which is from Digi-key
  • [17:43:45] <neelesh_gurjar> they said that they dont have it in stock
  • [17:43:57] <neelesh_gurjar> do anybodya have idea... how can i get it?
  • [17:44:04] <chakie> backorder?
  • [17:44:21] <kulve> that number might not tell the truth. But on the other hand I ordered it a week ago and haven't got any news yet..
  • [17:45:29] <chakie> well, they apparently have ran out of boards
  • [17:46:05] <neelesh_gurjar> hmm
  • [17:46:30] <neelesh_gurjar> Actully I have customized and compile Linux for Firewalling and print server
  • [17:46:51] <neelesh_gurjar> and wanted to try it...
  • [17:47:27] <neelesh_gurjar> do you have any idea similar boards like beagle?
  • [17:47:58] <kulve> they are all different. Maybe gumstix would do? It's not as powerful though
  • [17:51:22] <neelesh_gurjar> ok thanks
  • [17:51:35] <neelesh_gurjar> I think BeagelBoards are awesome
  • [17:51:58] <neelesh_gurjar> I am Unlucky... :(
  • [17:52:15] <kulve> http://gumstix.com/store/catalog/product_info.php?cPath=27&products_id=178
  • [17:55:46] <prpplague> or the hammer: http://www.tincantools.com
  • [17:57:21] * ade_ (n=ade@80.177.64.36) Quit ()
  • [17:57:24] * esden is now known as esden`away
  • [17:57:34] <Crofton|work> neelesh_gurjar, there is prodcut flowing toward Digikey
  • [17:59:19] <chakie> prpplague: hammer time?
  • [18:02:12] <neelesh_gurjar> gr8... I was littlebit nervous... because I put too much efforts in costumizing and compiling linux for it... and just waste bcoz of no hardware
  • [18:04:49] <robclark> random question: anyone doing development with macosx as the host machine? I think TFTP and NFS server, I've got.. not sure about gcc cross compiler
  • [18:05:27] <Crofton|work> I think some people have run OE on OSX :)
  • [18:06:15] <robclark> ok, I'll check oe.net and see if I can muddle my way thru it
  • [18:06:41] <robclark> (ok, I see a Getting_Started page.. good sign)
  • [18:06:55] <robclark> (for a newbie)
  • [18:06:58] <Crofton|work> it works well for me :)
  • [18:07:24] <Crofton|work> but I have been through it a bunch and I am on regualr intel
  • [18:07:25] <NishanthMenon> interesting.. i dont see codesourcery for macos
  • [18:07:59] <feig> i rid OSX from my MacBook Air so that i could get the TI dev tools for linux
  • [18:08:13] <robclark> hmm, k, found http://wiki.openembedded.net/index.php/OEandYourDistro#Using_OpenEmbedded_on_Mac_OS_X .... I guess i need to read some more and then see if I still have question
  • [18:08:45] <robclark> feig: :(
  • [18:09:42] <sakoman> kulve: http://www.bizjournals.com/sanjose/prnewswire/press_releases/national/California/2008/07/28/LAM036
  • [18:10:31] <NishanthMenon> robclark, dunno if this: http://www.linuxdevices.com/news/NS9251691128.html is interesting for you if you are looking to cross compile kernel on macos
  • [18:11:13] <robclark> you think I can use that instead of http://www.codesourcery.com/gnu_toolchains/arm/releases/2007q3
  • [18:12:11] <robclark> seems based on much older gcc...
  • [18:12:29] <NishanthMenon> am not too sure.. zaurus probably was 1710 or so.. i dont recollect.. dunno if it has cortex support even
  • [18:12:51] <mru> 2007q3 is based on gcc 4.2.1
  • [18:12:51] <robclark> I need at least thumb-2 code gen
  • [18:13:18] <robclark> (yeah, sorry, I meant that other link NishanthMenon sent look much older... but maybe it was just the press release that was older)
  • [18:13:47] * guillaum1 (n=gl@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
  • [18:14:13] * guillaum1 (n=Guillaum@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) has joined #beagle
  • [18:16:22] <robclark> oh, ok the codesourcery link has a src code option which I didn't notice at first. Ok, back in business
  • [18:16:28] <neelesh_gurjar> ok team thank you very much for help.... bye n have a nice day
  • [18:17:24] * neelesh_gurjar (n=neel@123.236.29.189) has left #beagle
  • [18:17:46] * NishanthMenon (n=gnat@nat/ti/x-ce6285da028f7ef5) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:17:46] * methril|laptop (n=methril@78.72.220.87.dynamic.jazztel.es) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:17:46] * Guest62082 (n=root@rrcs-208-125-50-2.nys.biz.rr.com) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:17:46] * zoobab (i=zoobab@vic.ffii.org) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:17:46] * bjdooks (n=ben@trinity.fluff.org) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:17:56] * NishanthMenon (n=gnat@nat/ti/x-ce6285da028f7ef5) has joined #beagle
  • [18:17:56] * methril|laptop (n=methril@78.72.220.87.dynamic.jazztel.es) has joined #beagle
  • [18:17:56] * Guest62082 (n=root@rrcs-208-125-50-2.nys.biz.rr.com) has joined #beagle
  • [18:17:56] * zoobab (i=zoobab@vic.ffii.org) has joined #beagle
  • [18:17:56] * bjdooks (n=ben@trinity.fluff.org) has joined #beagle
  • [18:18:06] * Viral_Sachde (n=NetUser@122.167.202.18) has left #beagle
  • [18:18:58] <kulve> sakoman: yeah, I read about that. I already have a gumstix and I already have ordered a beagle :)
  • [18:19:14] * GeneralAntilles (n=GAN8001@pdpc/supporter/active/generalantilles) has joined #beagle
  • [18:31:03] <feig> robclark: it was only a matter of time =)
  • [18:32:13] <robclark> well, if I didn't have to use an exchange mail server, micro$soft word/powerpoint/excel/etc, then I suppose I'd be pretty happy with linux
  • [18:32:43] <robclark> (or, well, maybe freebsd ;-))
  • [18:32:59] <feig> and if i had a bigger drive than 80GB, i might have had space for OSX ;)
  • [18:33:25] <prpplague> chakie: yea we get that alot
  • [18:33:47] <robclark> well, there is that.. my 100gb drive is starting to feel a bit cozy, and I don't even have any mp3's on my laptop
  • [18:34:14] * Crofton|work grumbles about Xilinx FPGA tools needing more than 5G to install
  • [18:34:18] <robclark> but I don't think it is any worse than windoze ;-)
  • [18:35:01] <feig> no, but on my OSX install, I had about 2GB free, under linux, with the same content, i have around 12GB free
  • [18:36:29] <robclark> well, I just need to figure out how to expense a 256GB flash drive ;-) (and then I woke up from my dream)
  • [18:37:19] * Guest62082 (n=root@rrcs-208-125-50-2.nys.biz.rr.com) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:37:19] * bjdooks (n=ben@trinity.fluff.org) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:37:19] * NishanthMenon (n=gnat@nat/ti/x-ce6285da028f7ef5) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:37:19] * zoobab (i=zoobab@vic.ffii.org) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:37:19] * methril|laptop (n=methril@78.72.220.87.dynamic.jazztel.es) Quit (kubrick.freenode.net irc.freenode.net)
  • [18:39:00] * NishanthMenon (n=gnat@nat/ti/x-ce6285da028f7ef5) has joined #beagle
  • [18:39:00] * methril|laptop (n=methril@78.72.220.87.dynamic.jazztel.es) has joined #beagle
  • [18:39:00] * Guest62082 (n=root@rrcs-208-125-50-2.nys.biz.rr.com) has joined #beagle
  • [18:39:00] * zoobab (i=zoobab@vic.ffii.org) has joined #beagle
  • [18:39:00] * bjdooks (n=ben@trinity.fluff.org) has joined #beagle
  • [18:39:17] * dirk2 (n=dirk@F32a1.f.strato-dslnet.de) has left #beagle
  • [18:42:37] * GeneralAntilles patiently waits for his backorder at Digi-Key to ship.
  • [18:48:10] * likewise (n=chatzill@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [18:48:19] <Crofton|work> banderson, how did you build the sdk?
  • [18:48:30] <likewise> hey all
  • [18:48:39] <Crofton|work> likewise, gm
  • [18:48:40] <mru> likewise
  • [18:49:56] <feig> i'm waiting to order... i guess i could now
  • [18:50:20] <Crofton|work> I suspect that getting the order places will get the board sooner
  • [18:50:34] <feig> i heard this week
  • [18:51:51] * geraint (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [18:55:41] <GeneralAntilles> I'm hoping it'll end up here before I have to go back to school
  • [18:55:45] <GeneralAntilles> Seems unlikely, though.
  • [19:00:25] <banderson> Crofton|work: bitbake meta-toolchain (something koen mention to me before)
  • [19:00:32] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [19:00:45] * Olipro (i=Olipro@unaffiliated/olipro) Quit (No route to host)
  • [19:00:54] * Olipro_ is now known as Olipro
  • [19:02:55] * lgentili (i=bd1344dd@gateway/web/ajax/mibbit.com/x-4373b69fd95d18a9) has joined #beagle
  • [19:06:30] <Crofton|work> banderson, I'll try and put some cycles into this
  • [19:06:30] * khilman (n=khilman@166.129.170.45) Quit (Read error: 104 (Connection reset by peer))
  • [19:06:46] <Crofton|work> normally I build inside OE, but I understand why it is important to create an sdk for people
  • [19:09:27] <max_fr> I have recompiled the kernel, but I have an issue at boot, it happens when it tries to remount the filesystem :
  • [19:09:27] <max_fr> Unhandled fault: external abort on non-linefetch (0x1028) at 0xe58c937c
  • [19:09:28] <max_fr> have you any idea?
  • [19:09:34] <max_fr> I just have disabled the console on framebuffer
  • [19:09:58] <mru> I've seen that a couple of times too
  • [19:10:00] <max_fr> I have recompiled the kernel, but I have an issue at boot, it happens when it tries to remount the filesystem :
  • [19:10:03] <max_fr> Unhandled fault: external abort on non-linefetch (0x1028) at 0xe58c937c
  • [19:10:05] <max_fr> have you any idea? I just have disabled the console on framebuffer
  • [19:10:36] <mru> does it do that consistently?
  • [19:10:40] <mru> or just once?
  • [19:10:43] <keesj> a sdk requires some level of stability
  • [19:10:58] <max_fr> mru : yes
  • [19:11:06] <mru> yes to what?
  • [19:11:13] <max_fr> consistently :-=
  • [19:11:16] <max_fr> :-)
  • [19:11:29] <mru> make clean, and rebuild the kernel
  • [19:11:39] <mru> maybe there's some dependency that isn't picked up proplery
  • [19:11:52] <max_fr> ok, I will try this, thank you
  • [19:12:09] * Beagle5 (n=Beagle5@dhcp-78-57.apl.washington.edu) has joined #beagle
  • [19:12:58] <banderson> Crofton|work: I would love to help I just don't know OE that well (ok hardly at all)...So if there anything I could do to help let me know or any suggestions on what to look for.
  • [19:13:03] <Beagle5> Hi, I noticed that the "Hardware Reference Manual" 404's
  • [19:13:06] <Beagle5> on the main page
  • [19:13:37] <Crofton|work> can you try on a 32 bit machine?
  • [19:13:44] <Crofton|work> my build machine is also 64 bit
  • [19:17:04] * Beagle5 (n=Beagle5@dhcp-78-57.apl.washington.edu) has left #beagle
  • [19:19:49] <jkridner> I've managed to get back into the old beagleboard.org machine, but it is showing a load of 45!
  • [19:20:03] <mru> whoa, what's going on there
  • [19:20:07] <jkridner> top doesn't show anything that is taking any time.
  • [19:20:35] <jkridner> it is all related to me running out of disk space due to running OE incorrectly.
  • [19:21:20] <jkridner> there is a /dev/sda2 that I issued a request to re-mount, but it hasn't completed the mount
  • [19:21:37] <max_fr> mru : thank you it works well now
  • [19:23:22] <jkridner> I can ssh in, but even 'ps auwx' hangs.
  • [19:23:42] <jkridner> 'reboot' never finishes
  • [19:23:52] * guillaum2 (n=Guillaum@AMontsouris-153-1-94-176.w90-2.abo.wanadoo.fr) has joined #beagle
  • [19:24:19] * lgentili (i=bd1344dd@gateway/web/ajax/mibbit.com/x-4373b69fd95d18a9) Quit ("http://www.mibbit.com ajax IRC Client")
  • [19:24:39] <jkridner> any known good ways to force a Linux machine to reboot when you cannot reach the power switch?
  • [19:24:51] <Crofton|work> where is the machine?
  • [19:24:59] <jkridner> Amazon.
  • [19:25:13] <Crofton|work> I think most of the remote reboot exploits have been fixed ....
  • [19:31:13] <jkridner> I have shell access.
  • [19:31:37] <jkridner> I get lots of notifications that my machine is going down for reboot, but it doesn't happen.
  • [19:31:41] <Crofton|work> can you kill anything?
  • [19:31:45] <jkridner> yes
  • [19:31:49] <jkridner> should I kill init?
  • [19:33:28] <mru> jkridner: echo b > /proc/sysrq-trigger
  • [19:33:51] <jkridner> what does that do?
  • [19:33:55] <jkridner> reboot?
  • [19:33:58] <mru> yes
  • [19:34:00] <max_fr> jkridner: have you tried options of reboot ?
  • [19:34:12] <jkridner> I was just looking up the options.
  • [19:34:16] <mru> it's equivalent of pressing sysrq+b on the keyboard
  • [19:34:23] <mru> if that doesn't work, nothing will
  • [19:35:03] <jkridner> it worked. :)
  • [19:36:49] <BThompson> quick question (since the docs are not up): which uart on the OMAP does the beagle have pinned out?
  • [19:37:04] <Crofton|work> 3?
  • [19:37:10] * Crofton|work is not sure
  • [19:37:12] <jkridner> I think he means on the expansion header.
  • [19:37:28] <jkridner> it is UART3 to the serial connector. (ttyS2)
  • [19:37:29] <BThompson> i mean which uart is spiting out the terminal
  • [19:37:35] <BThompson> thats good ok
  • [19:37:37] <jkridner> UART3.
  • [19:37:47] <BThompson> thanks
  • [19:38:34] <BThompson> debugging something that works on beagle but doesnt on the mistral evm, and when we boot the beagle kernel on the evm as soon as it gets to the kernel the terminal goes dead, if it is using a different uart that would do it :)
  • [19:39:53] <Crofton|work> I'm glad something works on the beagle and not on the others :)
  • [19:40:49] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [19:41:30] * guillaum1 (n=Guillaum@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) Quit (Read error: 110 (Connection timed out))
  • [19:41:46] * guillaum2 is now known as guillaum1
  • [19:42:24] * mru had exactly the same thought
  • [19:44:44] * BeagleBot (n=PircBot@ec2-75-101-156-174.compute-1.amazonaws.com) has joined #beagle
  • [19:44:44] * Topic is 'Welcome to #Beagle | Discussion about the OMAP3 Beagle Board - http://beagleboard.org | Beagle search tools are on #dashboard at irc.gimp.org, NOT here ;) | Log is at http://www.beagleboard.org/irclogs'
  • [19:44:44] * Set by jkridner on Tue Jun 10 00:33:18 CDT 2008
  • [19:45:24] <jkridner> last night's server is now back on-line. Thanks mru.
  • [19:46:54] * lgentili (i=bd1344dd@gateway/web/ajax/mibbit.com/x-9f2a709ba53a6f7d) has joined #beagle
  • [19:47:31] * BeagleLogBot (n=PircBot@ec2-75-101-156-174.compute-1.amazonaws.com) Quit (Read error: 110 (Connection timed out))
  • [19:50:18] <banderson> Crofton|worK: I tried that as well and got basically the same results (http://pastebin.com/m685ec286)
  • [19:50:58] * dcordes (n=dcordes@unaffiliated/dcordes) has joined #beagle
  • [19:53:14] * max_fr (i=max@212-198-79-163.rev.numericable.fr) has left #beagle
  • [19:58:58] <feig> I just put in my order with Digi-Key
  • [20:06:07] * Guest62082 (n=root@rrcs-208-125-50-2.nys.biz.rr.com) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:06:07] * bjdooks (n=ben@trinity.fluff.org) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:06:07] * NishanthMenon (n=gnat@nat/ti/x-ce6285da028f7ef5) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:06:07] * zoobab (i=zoobab@vic.ffii.org) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:06:07] * methril|laptop (n=methril@78.72.220.87.dynamic.jazztel.es) Quit (kubrick.freenode.net irc.freenode.net)
  • [20:06:35] * NishanthMenon (n=gnat@nat/ti/x-ce6285da028f7ef5) has joined #beagle
  • [20:06:35] * methril|laptop (n=methril@78.72.220.87.dynamic.jazztel.es) has joined #beagle
  • [20:06:35] * Guest62082 (n=root@rrcs-208-125-50-2.nys.biz.rr.com) has joined #beagle
  • [20:06:35] * zoobab (i=zoobab@vic.ffii.org) has joined #beagle
  • [20:06:35] * bjdooks (n=ben@trinity.fluff.org) has joined #beagle
  • [20:09:36] <keesj> pdfinfo Beagle_HW_Reference_Manual_A_5.pdf | grep Title
  • [20:09:48] <keesj> boo
  • [20:12:28] * dcordes (n=dcordes@unaffiliated/dcordes) Quit (Read error: 104 (Connection reset by peer))
  • [20:14:52] * likewise (n=chatzill@82-171-51-231.ip.telfort.nl) Quit ("ChatZilla 0.9.83 [Firefox 3.0/2008061015]")
  • [20:28:57] * valhalla (n=valhalla@81-174-21-16.dynamic.ngi.it) Quit ("Leaving")
  • [20:37:26] * lgentili (i=bd1344dd@gateway/web/ajax/mibbit.com/x-9f2a709ba53a6f7d) Quit ("http://www.mibbit.com ajax IRC Client")
  • [20:38:11] * DJWillis (n=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) Quit (Read error: 104 (Connection reset by peer))
  • [20:38:31] * DJWillis (i=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) has joined #beagle
  • [20:41:41] * NishanthMenon (n=gnat@nat/ti/x-ce6285da028f7ef5) Quit ("We be chillin - IceChat style")
  • [20:41:58] * NishanthMenon (n=gnat@nat/ti/x-7b5397e462a6f81c) has joined #beagle
  • [20:42:00] * NishanthMenon (n=gnat@nat/ti/x-7b5397e462a6f81c) Quit (Client Quit)
  • [20:42:41] * NishanthMenon (n=gnat@nat/ti/x-b9cabc6e45076725) has joined #beagle
  • [20:51:32] <robclark> btw, if anyone else is interested in using macosx box as host computer, I found a couple interesting links about getting gcc cross compiler:
  • [20:51:37] <robclark> http://www.mikrocontroller.net/articles/ARM_GCC_toolchain_for_Linux_and_Mac_OS_X
  • [20:52:12] <robclark> and http://wiki.forum.nokia.com/index.php/MacOSX which is more about symbian but has instructions for modifying codesourcery src distribution to build on macosx
  • [20:52:31] <robclark> I'm not sure yet which works better..
  • [21:00:21] * feig (n=ejf3@34.sub-75-223-21.myvzw.com) Quit ("Konversation terminated!")
  • [21:02:59] * likewise (n=chatzill@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [21:05:27] * lgentili (i=bd1344dd@gateway/web/ajax/mibbit.com/x-17f9e9a7cfea4e4e) has joined #beagle
  • [21:32:32] * lgentili (i=bd1344dd@gateway/web/ajax/mibbit.com/x-17f9e9a7cfea4e4e) Quit ("http://www.mibbit.com ajax IRC Client")
  • [21:43:51] <jkridner|work> mru, koen: do you know if anyone has built the EVM kernel for use with omapfbplay?
  • [21:45:30] <mru> no idea
  • [21:45:52] <mru> I don't have an evm, so I certainly haven't
  • [21:46:40] <jkridner|work> mru: k. I'm looking to see if sakoman has built it. I'm not sure what the patch looks like. I don't see it on sakoman's OE git tree. (http://www.sakoman.net/cgi-bin/gitweb.cgi?p=org.openembedded.dev-omap3.git;a=tree;f=packages/linux/linux-omap2-git/omap3evm;h=c8e2af8c80d877ec57f19b78269a25cb33f5bc52;hb=HEAD)
  • [21:46:54] <mru> the patch for what?
  • [21:47:23] <jkridner|work> support for whatever features you are using for omapfbplay. namely, /dev/fb1
  • [21:47:48] <jkridner|work> the PSP kernel doesn't seem to react to the bootargs. could be user error though.
  • [21:48:02] <jkridner|work> want to make sure whatever patch is required for omapfbplay is already there.
  • [21:48:07] <mru> you need to configure enough dma memory
  • [21:48:23] <jkridner|work> is that support in the git kernel as-is?
  • [21:48:28] <mru> yes
  • [21:48:39] <mru> but the default probably isn't enough
  • [21:48:45] <mru> iirc the default is 2M
  • [21:48:55] <jkridner|work> k.
  • [21:48:56] <mru> that's enough on a very small screen
  • [21:49:03] <jkridner|work> where does /dev/fb1 come from?
  • [21:49:46] <mru> I'm not sure what you're asking
  • [21:51:26] <jkridner|work> well, we don't normally have /dev/fb0, only /dev/fb1.
  • [21:54:53] <jkridner|work> we've got 'video=omapfb:vram:2M,vram:4M', but no /dev/fb1 exists.
  • [21:55:07] <mru> check dmesg
  • [21:55:19] <jkridner|work> er, previously I meant we don't normally have /dev/fb1, just /dev/fb0
  • [21:55:25] <mru> I guessed as much
  • [21:55:33] <mru> you probably don't have enough dma memory
  • [21:55:51] <mru> CONFIG_FB_OMAP_CONSISTENT_DMA_SIZE
  • [21:56:08] <jkridner|work> k. so, I should look at your defconfig?
  • [21:56:25] * cbrake is now known as cbrake_away
  • [21:56:29] <mru> you should look at whatever config you're actually using, of course
  • [21:56:37] * mru never uses defconfig...
  • [21:57:55] <jkridner|work> k.
  • [21:58:05] <jkridner|work> so, you don't keep a public .config?
  • [21:58:48] <jkridner|work> CONFIG_FB_OMAP_CONSISTENT_DMA_SIZE=4 ?
  • [21:58:52] <mru> too small
  • [21:58:56] <mru> 2+4 > 4
  • [21:59:10] <jkridner|work> 8? :)
  • [21:59:16] <mru> won't work
  • [21:59:20] <mru> 6 will do
  • [21:59:32] <mru> 8 fails because it's too big for something else
  • [21:59:35] <mru> not sure what
  • [22:00:00] <jkridner|work> so, where does /dev/fb1 come from?
  • [22:00:25] <mru> omapfb registers it
  • [22:00:46] <mru> if you ask for it (vram=x) and there's enough dma memory
  • [22:02:32] * jkridner|work (n=a0321898@nat/ti/x-37a8a48c9bfd327e) has left #beagle
  • [22:02:36] * jkridner|work (n=a0321898@nat/ti/x-37a8a48c9bfd327e) has joined #beagle
  • [22:04:51] <jkridner|work> koen seems to be building with 8 for the demo: http://www.sakoman.net/cgi-bin/gitweb.cgi?p=org.openembedded.dev-omap3.git;a=blob;f=packages/linux/linux-omap2-git/beagleboard/defconfig;h=b3ea5f95579359ef58b485402721f6ddf540b97b;hb=HEAD
  • [22:05:38] <mru> hmm, maybe that works then
  • [22:05:57] <mru> I remember something going wrong when I tried to use too much
  • [22:11:18] * fer_luck (n=fer_luck@201-88-32-138.cbace700.e.brasiltelecom.net.br) Quit ()
  • [22:15:08] * robclark (n=robclark@nat/ti/x-2b1ec874959ea10e) Quit ()
  • [22:17:59] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
  • [22:24:59] * koen notices a few highlights
  • [22:34:12] * geraint_ (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [22:35:59] <koen> mru: 8 works, but iirc 12 failed
  • [22:36:47] <mru> it may have been that a single plane >6 failed
  • [22:39:35] * jpao79 (n=a0193294@nat/ti/x-a9fd52bcf066eb57) has joined #beagle
  • [22:42:55] * methril|laptop (n=methril@78.72.220.87.dynamic.jazztel.es) Quit (Read error: 60 (Operation timed out))
  • [22:43:20] * methril|laptop (n=methril@252.73.220.87.dynamic.jazztel.es) has joined #beagle
  • [22:43:45] * geraint__ (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [22:43:47] <jpao79> mru: jkridner was asking on my behalf. i re-built the kernel with dma size =6 but /dev/fb1 is still not showing up. are there any other things you can think of to get /dev/fb1 to show up?
  • [22:44:02] <ds2> would anyone know of a framebuffer image viewer that does not use directfb libraries?
  • [22:44:15] <mru> jpao79: you need the right kernel command line
  • [22:44:35] <mru> jkridner pasted the required bit above
  • [22:44:40] <mru> video=omapfb...
  • [22:46:11] <jpao79> yeah, i've got the following command line: bootargs=console=ttyS0,115200n8 console=tty0 root=/dev/mmcblk0p2 rootdelay=2 rootfstype=ext3 video=omapfb:vram:2M,vram:4M ip=dhcp and it's showing up in the init sequence... is there a file i should look in to determine if my kernel supports the video bootarg?
  • [22:46:19] * Guest62082 (n=root@rrcs-208-125-50-2.nys.biz.rr.com) Quit (kubrick.freenode.net irc.freenode.net)
  • [22:46:19] * bjdooks (n=ben@trinity.fluff.org) Quit (kubrick.freenode.net irc.freenode.net)
  • [22:46:19] * zoobab (i=zoobab@vic.ffii.org) Quit (kubrick.freenode.net irc.freenode.net)
  • [22:48:34] * bjdooks (n=ben@trinity.fluff.org) has joined #beagle
  • [22:48:49] <mru> jpao79: look for lines further down starting with omapfb:
  • [22:49:06] <mru> you should see something like "omapfb: Framebuffer initialized. Total vram 5529600 planes 2"
  • [22:50:39] * geraint (n=gnorth@tu008.demon.co.uk) Quit (Read error: 110 (Connection timed out))
  • [22:51:01] <jpao79> hmm... nothing like that in my kernel init... i'm using the ti psp kernel. maybe it's not supported. i'll do a bit of grepping...
  • [22:53:09] <mru> that won't work
  • [22:54:06] * geraint__ (n=gnorth@tu008.demon.co.uk) Quit (Read error: 104 (Connection reset by peer))
  • [22:54:20] <koen> the TI kernel uses some dialect of v4l for overlays
  • [22:54:37] <mru> and nobody's figured out how to use it...
  • [22:54:46] * geraint (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [22:55:09] <koen> not even TI :)
  • [22:55:11] * koen hides
  • [22:56:10] <koen> jkridner|work: the sugar stuff is now stuck on two points: the artwork issue and python-numpy act of putting -L/usr/lib in cflags
  • [22:56:22] * koen hates python
  • [22:56:32] * koen hates crosscompiling python modules even more
  • [22:56:33] * Guest62082 (n=root@rrcs-208-125-50-2.nys.biz.rr.com) has joined #beagle
  • [22:56:47] * zoobab (i=zoobab@82.197.166.218) has joined #beagle
  • [22:56:53] <jkridner> numpy?
  • [22:57:14] <koen> the successor of numeric python
  • [22:57:21] <vlad_> 0x4042cedc: add lr, pc, #4 ; 0x4
  • [22:57:21] <vlad_> 0x4042cee0: bx r12
  • [22:57:21] <vlad_> 0x4042cee4: ldreqsh r0, [r8], -r0
  • [22:57:23] <koen> sugar needs it aat runtime
  • [22:57:27] <vlad_> stupid ^I's
  • [22:57:38] <vlad_> gcc is not decoding that ldreqsh correctly, I don't think -- any ideas what it should be?
  • [22:57:42] <vlad_> er, gdb
  • [22:58:11] <jkridner> so, source is not in python itself?
  • [22:58:12] <mru> what generated that code?
  • [22:58:25] * BThompson (n=BThompso@nat/ti/x-ba931a6600aa53fb) Quit (Remote closed the connection)
  • [22:58:32] <vlad_> a jit that i'm trying to bang into shape
  • [22:58:37] <koen> jkridner|work: sugar is a mix of 'native' libs and python code
  • [22:59:01] <koen> jkridner|work: so it builds ok, but you need to figure out the runtime python depends
  • [22:59:05] <mru> vlad_: that's a valid instruction, but it looks might weird
  • [22:59:47] <koen> jkridner|work: see all the sugar commits at http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=summary :)
  • [22:59:54] <koen> it's an iterative process
  • [23:00:05] * koen commits early and often
  • [23:00:26] <mru> vlad_: can you post a bit more of that code?
  • [23:00:34] <mru> before and after what you had there
  • [23:00:50] <mru> including actual instruction bytes
  • [23:00:56] <mru> use pastebin.com
  • [23:01:09] <vlad_> mru: yep, sec
  • [23:01:45] <koen> vlad_: ff3 built with your mozconfig, I'm now sorting out the -dev dependency issues
  • [23:02:02] <jkridner> keon: so, is the approach to patch all of the makefile instances where -L/usr/lib is used?
  • [23:02:12] <vlad_> koen: awesome; will probably still be dog slow, I found a horrible 24bpp/16bpp issue yesterday
  • [23:02:23] * geraint_ (n=gnorth@tu008.demon.co.uk) Quit (Read error: 101 (Network is unreachable))
  • [23:02:42] <vlad_> there were a horrifying mass of 16->24 and 24->16 converions happening, thanks to gtk native theme wackyness
  • [23:02:43] <koen> vlad_: and it's taking 50MB of resisent memory according to top
  • [23:03:00] <vlad_> yeah, firefox proper will do that, fennec's better
  • [23:03:01] <vlad_> mru: http://pastebin.ca/1180143
  • [23:06:44] <mru> vlad_: that's not an instruction
  • [23:06:51] <mru> it's a literal
  • [23:06:55] <vlad_> mru: immediate constant
  • [23:07:01] <vlad_> yeah, I just got there, heh
  • [23:07:14] <mru> the ldr on the line before is a giveaway
  • [23:08:11] <mru> it's more common to group literals where normal program flow won't reach them
  • [23:08:16] <mru> like between functions
  • [23:08:19] <jkridner> koen: thanks for git. low-hanging fruit, but saved me from even having to do it.
  • [23:08:24] <vlad_> nod
  • [23:08:43] <jkridner> very nice for keeping backups of items on my cards.
  • [23:08:45] <vlad_> this is just emitting a stream at runtime, so I guess this kinda makes sense, though it means a bunch of extra instructions
  • [23:09:18] <mru> function calls like that will mess with the branch predictor too
  • [23:09:35] <mru> bl and blx push the return value onto the predicted return stack
  • [23:09:52] <mru> which is used by bx lr and pop pc instructions
  • [23:10:01] <mru> bx r12 won't do that
  • [23:10:20] <mru> so when the function returns, it will almost certainly mispredict
  • [23:10:40] * geraint_ (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [23:10:59] <mru> a mispredicted branch costs 13 cycles on A8
  • [23:11:07] <vlad_> mmm
  • [23:11:25] <vlad_> well, i'll solve that problem after I get things working
  • [23:11:40] <mru> correctly predicted instructions take only 1 cycle
  • [23:15:02] * geraint (n=gnorth@tu008.demon.co.uk) Quit (Read error: 60 (Operation timed out))
  • [23:15:49] * dcordes (n=dcordes@unaffiliated/dcordes) has joined #beagle
  • [23:18:39] * lgentili (i=bd1344dd@gateway/web/ajax/mibbit.com/x-4e940d9459be81c1) has joined #beagle
  • [23:20:44] * rsalveti (n=salveti@200.184.118.132) Quit ("Bye :)")
  • [23:21:12] * geraint_ (n=gnorth@tu008.demon.co.uk) Quit (Read error: 104 (Connection reset by peer))
  • [23:21:40] * geraint (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [23:27:00] * geraint_ (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [23:36:59] * guillaum1 (n=Guillaum@AMontsouris-153-1-94-176.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
  • [23:37:17] * guillaum1 (n=gl@AMontsouris-153-1-94-176.w90-2.abo.wanadoo.fr) has joined #beagle
  • [23:39:02] * geraint__ (n=gnorth@tu008.demon.co.uk) has joined #beagle
  • [23:40:54] * geraint_ (n=gnorth@tu008.demon.co.uk) Quit (Read error: 60 (Operation timed out))
  • [23:42:56] * geraint (n=gnorth@tu008.demon.co.uk) Quit (Read error: 110 (Connection timed out))
  • [23:43:31] * likewise (n=chatzill@82-171-51-231.ip.telfort.nl) Quit ("ChatZilla 0.9.83 [Firefox 3.0/2008061015]")
  • [23:45:46] * geraint__ (n=gnorth@tu008.demon.co.uk) Quit ()
  • [23:47:45] * jpao79 (n=a0193294@nat/ti/x-a9fd52bcf066eb57) has left #beagle