Jun 08 11:14:10 * vvu shares drinks with everyone...received the BSc diplima
Jun 08 11:19:18 * anujdeshpande sets PARTYMODE flag to 1, also congratulates vvu 
Jun 08 11:31:54 <av500> vvu: congrats!!
Jun 08 11:39:20 <vvu> danke!
Jun 08 11:39:52 <vvu> i gave the keys from my college room and i am homless...at least homeless with a bsc
Jun 08 11:40:17 <av500> vvu: just buy a bahncard 100
Jun 08 11:40:23 <av500> then you can live anywhere
Jun 08 11:41:24 <vvu> i have my flight tomorrow to .ro
Jun 08 11:41:56 <vvu> had 3 tickets at lufthansa and one of my graduation guests could not make it and lufthansa did not want to change my flight from tomorrow to today
Jun 08 11:42:06 <vvu> parents left today, me leaving tomorrow
Jun 08 14:22:03 <Abhishek_> vvu: Heartiest Congratulations!
Jun 08 14:22:13 * Abhishek_ picks up the glass and looks for vvu
Jun 08 14:33:10 <alexanderhiam> congrats vvu!
Jun 08 19:11:30 <Abhishek_> rcn-ee: What does "cannot find pctldev for /ocp intc, deferring probe mean" ?
Jun 08 19:12:17 <rcn-ee> Abhishek_, it's "deferred".... it just means a hardware dependicy wasn't avaiable at that time slot, so it'll try again later
Jun 08 19:12:33 <Abhishek_> rcn-ee: And I don't get backlight on the LCD ...
Jun 08 19:13:18 <rcn-ee> which kernel? uname -r
Jun 08 19:13:31 <Abhishek_> 4.1.0 rc6 debian
Jun 08 19:13:46 <Abhishek_> 4.1.0-rc6-bone6
Jun 08 19:14:15 <rcn-ee> well, it works for my boards/lcd's... what's your dts look like?
Jun 08 19:14:37 <Abhishek_> rcn-ee: I saw the dts's you use gpio-backlight, I'm trying pwm backlight
Jun 08 19:15:01 <rcn-ee> as of last week, not anymore... they are all pwm-backlight..
Jun 08 19:16:04 <Abhishek_> I remember seeing them gpio-backlight?
Jun 08 19:16:48 <rcn-ee> https://github.com/beagleboard/bb.org-overlays/blob/master/src/arm/BB-BONE-LCD4-01-00A1.dts#L172
Jun 08 19:18:27 <Abhishek_> the ones in dtb-rebuilder are still gpio-backlight
Jun 08 19:19:23 <Abhishek_> rcn-ee: This was the error message: "could not find pctldev for node /ocp/interrupt-controller@48200000, deferring probe"
Jun 08 19:19:50 <rcn-ee> https://github.com/RobertCNelson/dtb-rebuilder/blob/4.1.x/src/arm/am335x-cape-bbb-exp-c.dtsi#L133
Jun 08 19:21:38 <rcn-ee> Abhishek_, "deferring" =/= error... it seemly stating it's "waiting" for xyz...
Jun 08 19:35:51 <Abhishek_> rcn-ee: trying to compile a new device tree for my cape using dtb-rebuilder, will let you know the results
Jun 08 19:40:18 * shubhangi wonders as to why she still uses irccloud when proxy issues no longer exists. Great tool though. Thanks to abhishek :)
Jun 08 19:41:43 * Abhishek_ still uses IRCCloud as well
Jun 08 19:43:49 <Abhishek_> shubhangi: did Suman reply to your last queries (the one which you cc'd to the group)?
Jun 08 19:44:21 <shubhangi> Abhishek_: i have a function pruss_write(memory_type, wordoffset, data_buf, buf_size ) working well. shall i write something for the pru to send an ACK signal after reading it so that subsequent writes to pru has to wait for this ACK.
Jun 08 19:44:28 <shubhangi> Abhishek_: not yet
Jun 08 19:45:21 <shubhangi> or overwrite without ensuring read from pru is left upto the user
Jun 08 19:45:30 <Abhishek_> shubhangi: Have you pushed that function to GitHub?
Jun 08 19:45:35 <shubhangi> no
Jun 08 19:45:52 <shubhangi> tomorrow
Jun 08 19:46:35 <shubhangi> finishing read right now .. one issue needs to be resolved .. reading sysfs from userspace is also causing show function to be triggered :(
Jun 08 19:46:46 <Abhishek_> you writing into the PRU memories (DRAM & shared)?
Jun 08 19:46:50 <shubhangi> sorry writing to sysfs
Jun 08 19:47:02 <shubhangi> DRAM0/1 and SHR RAM
Jun 08 19:47:09 <shubhangi> DDR is pending
Jun 08 19:47:15 <Abhishek_> shubhangi: isn't that how reading from sysfs is meant to be?
Jun 08 19:47:36 <shubhangi> Abhishek_: writing to sysfs is triggering show function
Jun 08 19:47:51 <shubhangi> both show and store ... only store is supposed to be trigerred
Jun 08 19:48:20 <Abhishek_> hmm, yes, what is the code that writes into the sysfs file?
Jun 08 19:48:37 <shubhangi> C .. fopen .. fwrite -> fclose
Jun 08 19:49:16 <shubhangi> i'll get it right .. was the last thing that came up when i last set work aside
Jun 08 19:49:25 <shubhangi> working on it now
Jun 08 19:49:31 <Abhishek_> no, don't use f-prefixed functions to access sysfs files
Jun 08 19:49:38 <shubhangi> then ?
Jun 08 19:49:47 <Abhishek_> use open(), read() and write()
Jun 08 19:49:55 <Abhishek_> directly
Jun 08 19:50:06 <shubhangi> okay .. i thought that too .. i guess fopen reads the file in the first place
Jun 08 19:50:14 <shubhangi> thats wht causing the issue
Jun 08 19:50:32 <shubhangi> ok .. trying it out .. thanks :)
Jun 08 19:50:35 <Abhishek_> shubhangi: That's because f-prefixed calls buffer the file
Jun 08 19:51:28 <shubhangi> okay
Jun 08 19:55:23 <Abhishek_> jkridner: you going to Shenzhen next week?
Jun 08 21:18:17 <Abhishek_> rcn-ee: successfully built a custom dtb for my LCD. However the display comes up quite late, about 5 seconds in. Is there a way to get it during early boot?
Jun 08 21:18:57 <Abhishek_> (does it have to do something with the "quiet" kernel cmdline?)
Jun 08 21:19:40 <rcn-ee> no, it's part of being a module... for overlays to work, i needed tilcd/pwm-backlight to be a module..
Jun 08 21:19:56 <rcn-ee> if you hard code everything in the *.dtb you can move those back to built-in..
Jun 08 21:20:02 <ds2> monolithic kernel + DT loaded by bootloader will solve that
Jun 08 21:22:15 <Abhishek_> rcn-ee: That's what I did just now, in my custom DTB I overclock i2c2 (cape EEPROM bus) to 400kHz, and then pwm backlight and panel timings for my display
Jun 08 21:22:33 <Abhishek_> and hardware setup
Jun 08 21:23:50 <Abhishek_> rcn-ee: to build the beaglebone kernel should I just clone http://github.com/beagleboard/kernel or set up my own tree using Linus' sources and patches?
Jun 08 21:24:18 <rcn-ee> "This repo is now deprecated" <- i think we need a louder font...
Jun 08 21:24:40 <rcn-ee> https://github.com/beagleboard/linux/tree/4.1
Jun 08 21:28:15 <Abhishek_> ah I got confused between "kernel" and "linux" repos
Jun 08 21:28:44 <ds2> rcn-ee: make the default Makefile target delete the tree
Jun 08 21:30:14 <rcn-ee> that would be too mean... some users setup scripts to use that repo. ;)
Jun 08 21:30:50 <Abhishek_> rcn-ee: If I were to install the ti-sgx modules from apt-get, would I be able to get GPU acceleration during normal desktop usage?
Jun 08 21:31:05 <ds2> it would be loud
Jun 08 21:31:11 <rcn-ee> define "normal".. ;)
Jun 08 21:31:15 <rcn-ee> normal x86 ?
Jun 08 21:31:27 <rcn-ee> normal 80's dos....
Jun 08 21:33:26 <ds2> so you get accelerated desqview in dosbox on the BBB with that?
Jun 08 21:33:55 <Abhishek_> I mean, when I get a display of the Linux desktop after boot.
Jun 08 21:34:15 <Abhishek_> and then I move the mouse around; is that GPU accelerated?
Jun 08 21:34:20 <rcn-ee> nope...
Jun 08 21:34:26 <rcn-ee> "framebuffer" only...
Jun 08 21:34:35 <rcn-ee> it'll draw on top of x....
Jun 08 21:34:38 <rcn-ee> literly on top...
Jun 08 21:35:07 <ds2> rcn-ee: is that ti-sgx working well enough to abuse the opengles stuff as a small parallel computer?
Jun 08 21:35:44 <rcn-ee> that's probally the only 'useful' thing. ;)
Jun 08 21:36:03 <rcn-ee> but with imgtec's compiler, i double it's acurate. ;)
Jun 08 21:36:16 <ds2> accuracy isn't that critical
Jun 08 21:36:22 <ds2> need soemthing faster then neon
Jun 08 21:37:20 <rcn-ee> looks up gl es2's list..
Jun 08 21:37:51 <ds2> think I have seen that
Jun 08 21:38:02 <ds2> it is no opencl
Jun 08 21:38:05 <rcn-ee> https://www.khronos.org/registry/gles/specs/2.0/es_full_spec_2.0.25.pdf
Jun 08 21:38:14 <rcn-ee> yeah, very early.. i'm thinkg what you could abuse..
Jun 08 21:38:27 <ds2> and certain things won't fire off IIRC due to some of the optimization/elimination it does
Jun 08 21:38:53 <ds2> face detection is painly slow (<1fps @ 640x480)
Jun 08 21:39:23 <ds2> if it can bump up the frame rate byreturning an approximation that needs to be verfied with NEON, it should still be helpful (for sparse data)
Jun 08 21:39:41 <Abhishek_> I ran guvcview over my new panel, managed some 12fps for 480x272
Jun 08 21:39:45 <rcn-ee> well, there is a readpixels option in egls 2... so you could read back, just not as efficent as es gl 3/opencv..
Jun 08 21:40:25 <rcn-ee> but there is fbo's...
Jun 08 21:41:25 <rcn-ee> Abhishek_, the blob wouldn't accerlate that..
Jun 08 21:42:14 <Abhishek_> that's what I was wondering, what would that blob actually do then?
Jun 08 21:42:44 <rcn-ee> years ago, before we figured out how crappy the sgx was... We found, xorg with fb was faster xorg with sgx 2d/3d acceleration...
Jun 08 21:42:54 <rcn-ee> thus it was known as sgx de-accelerator...
Jun 08 21:46:35 <rcn-ee> Abhishek_, https://ssvb.github.io/2012/05/04/xorg-drivers-and-software-rendering.html scroll down to IGEPv2..
Jun 08 21:49:25 <Abhishek_> lol
Jun 08 21:54:43 <Abhishek_> know of any application that I could use to test multitouch on the panel?
Jun 08 22:36:35 <nerdboy> where'd you get the display?
Jun 08 22:37:07 * nerdboy would like one
Jun 08 22:37:21 <nerdboy> too much sensor focus...
**** ENDING LOGGING AT Tue Jun 09 02:59:58 2015