Tuesday, 2015-11-03

*** tpb has joined #timvideos00:00
*** springermac has quit IRC00:05
*** springermac has joined #timvideos00:06
CarlFKxfxf: where is your script for saving atlys stream?00:25
xfxfon road atm00:31
xfxfI'm.intending on hacking on it more tonight00:31
tpbTitle: HDMI Capture and Analysis FPGA Project warmcat.com (at warmcat.com)01:42
*** Bertl is now known as Bertl_zZ02:10
mithroG33KatWork / seaLne: ping?04:08
mithroshenki: I'm assuming you aren't around?04:47
*** springermac has quit IRC05:28
*** springermac has joined #timvideos05:50
*** travis-ci has joined #timvideos06:12
travis-ci[timvideos/HDMI2USB-misoc-firmware/master#279] (fb3d8ad): The build passed. (https://travis-ci.org/timvideos/HDMI2USB-misoc-firmware/builds/88943359)06:12
*** travis-ci has left #timvideos06:12
mithro_florent_: when your bored / eating lunch / etc, take a read of https://warmcat.com/hardware%20design/hdmi/fpga/2015/10/22/hdmi-capture-and-analysis-fpga-project-3.html - specifically the "Raster measurement and statistics collection" section06:32
tpbTitle: HDMI Capture and Analysis FPGA Project 3 warmcat.com (at warmcat.com)06:32
*** springermac has quit IRC06:35
mithro_florent_: not urgent at all06:35
*** springermac has joined #timvideos06:36
*** se6astian|away is now known as se6astian07:34
*** [d__d] has joined #timvideos08:03
*** rohitksingh has joined #timvideos08:04
*** rohitksingh has quit IRC08:14
*** rohitksingh has joined #timvideos08:26
seaLnemithro: pong09:02
*** Bertl_zZ is now known as Bertl09:25
*** [d__d] has quit IRC09:38
*** JBentley has joined #timvideos10:34
*** [d__d] has joined #timvideos11:08
thaytanmithro, hmm - the Opsis campaign page doesn't say anything about what the milkymist compatibility board actually does11:11
thaytanoh, it's in the updates tab11:11
JBentleyHey Guys, I'm James and I just got an Opsis board. I spoke with mithro over email and he said I should log on here for some help setting it up11:42
JBentleyif anyone around able to give me some pointers on where to start?11:42
seaLneJBentley: https://github.com/timvideos/HDMI2USB-misoc-firmware/tree/master/scripts gives the main instructions on getting the board up and runnig11:55
tpbTitle: HDMI2USB-misoc-firmware/scripts at master · timvideos/HDMI2USB-misoc-firmware · GitHub (at github.com)11:55
mithroI'll be around in like 15 minutes11:56
JBentleyAwesome thanks seaLne12:01
JBentleyand mithro :)12:01
JBentleySo to start I'll need to set up an Ubuntu 14.04 environment I guess12:02
seaLneworks fine on wily aswell if that helps12:03
seaLnebut trusty would be the same as mithro is using so most likely to work12:04
seaLnei think some people are using other distros but the simplest way is with *ubuntu12:04
JBentleyok cool, well I suppose I ought to get on that before I do anything else as I only have an osx machine in the office currently12:05
seaLneyou might get away with a virtual machine but usb stuff can be a bit more complicated12:09
mithroJBentley: I'll be working on improving Mac support for users, but there is no way you can do development on a Mac12:09
mithrothe FPGA tools only run on Linux and Windows12:09
JBentleyAlright Well I was planning on partitioning a mac-mini we have with an arch linux partition anyway12:13
JBentleyso I may just do another partition with ubuntu12:13
seaLnemithro: is it still advised to use your opsis-fixes or is all of that merged now?12:23
mithroseaLne: that is merged now12:24
mithroseaLne: but I still have more to do12:24
seaLnesomething that i only recently discovered was that get-env.sh could be rerun to get updates12:26
seaLneso might not be obvious to others either12:26
seaLnei didn't really know what conda was12:27
xfxfmithro: my hdmi2usb toggle script you added read stuff to, do you have that somewhere?12:28
mithroxfxf: https://gist.github.com/mithro/0717f9a0b39a538e1e6d12:31
tpbTitle: Modified Ryans script · GitHub (at gist.github.com)12:31
mithroxfxf: its not very reliable12:31
xfxfyeah i remember you saying12:31
xfxfdid you identify why?12:31
xfxfi.e. it is a pyserial problem, or something funky going on with hdmi2usb's serial interface?12:32
*** Bertl is now known as Bertl_oO12:48
mithroxfxf: its the stupid exart driver and not the hdmi2usb stuff12:54
xfxfright, so nothing i can fix in my script then?13:01
xfxfi'll hold off integrating your read changes for now then13:01
xfxfdid i send you a PR with changes to the flash script? (and i assume they're probably obselete now given openocd branch is now merged?)13:02
ysionneaumithro: for debug, is there a way to ask FX2 to reprogram itself from I2C? like a PROG pin or something? or maybe setting some config pin + asserting reset?13:13
mithroysionneau: there is a reset pin13:14
mithroysionneau: and you can reboot the fx2 via the fx2 firmware13:14
ysionneaufor dev I think I will just add a CLI command to the LM32 fw : debug flash_fx213:15
ysionneauthat will reset the fx2 and let it reprogram13:15
ysionneauis there a PDF of the schematics of the Opsis board somewhere? Or just the kicad files?13:16
tpbTitle: HDMI2USB-numato-opsis-docs/pdfs at gh-pages · timvideos/HDMI2USB-numato-opsis-docs · GitHub (at github.com)13:21
mithroysionneau: http://opsis.hdmi2usb.tv/features/usb-peripheral.html13:21
tpbTitle: Numato Opsis : USB Peripheral (at opsis.hdmi2usb.tv)13:21
*** wanig_ has quit IRC13:24
*** wanig___ has joined #timvideos13:24
*** JBentley has quit IRC13:44
*** se6astian is now known as se6astian|away16:05
*** CarlFK has quit IRC16:23
*** CarlFK has joined #timvideos16:42
*** ChanServ sets mode: +v CarlFK16:42
*** se6astian|away is now known as se6astian16:52
CarlFKhttp://youtu.be/H_og9lRyfxg  look at that sync!16:57
CarlFKhd camera pointed at the monitor hooked up to Atlys output 116:58
CarlFKI am a little disturbed that I don't have a sync problem16:58
*** rohitksingh has quit IRC17:29
*** wanig___ has quit IRC17:38
*** wanig___ has joined #timvideos17:39
*** rohitksingh has joined #timvideos17:48
*** rohitksingh has quit IRC17:55
*** rohitksingh has joined #timvideos18:10
*** rohitksingh has quit IRC18:16
*** rohitksingh has joined #timvideos18:31
*** rohitksingh has quit IRC19:02
*** rohitksingh has joined #timvideos19:04
*** springermac has quit IRC19:36
*** springermac has joined #timvideos19:36
*** rohitksingh has joined #timvideos20:19
*** rohitksingh has quit IRC20:25
ysionneau_florent_: I cannot find where you platform.request the i2c pins on Opsis board (for i2c.c bit banging)20:26
ysionneauah, found it https://github.com/timvideos/HDMI2USB-misoc-firmware/commit/197d81f67fa0314e2762681d762b0463c7a0eff120:35
tpbTitle: targets: add dna to base targets and remove i2c from opsis (not used … · timvideos/[email protected] · GitHub (at github.com)20:35
*** CarlFK has quit IRC21:16
mithroysionneau: how did you go?21:31
ysionneauI didn't write much today in fact, mostly documenting (I'm not accounting it)21:37
ysionneaulooking at where is what in the repo :)21:37
mithroFair enough21:38
ysionneaudo you have an example of usage of the i2c.c ?21:38
mithroRohit had used it as a master21:38
mithroI'm on my phone right now so can't link it21:39
ysionneauok no problem, you can link it later21:39
ysionneauI built the bios/fw/bitstream, can't flash it, but my dev env is ready21:40
ysionneauI'm pretty sure of the answer but I need to work on the misoc firmware, right?21:40
mithroOkay, cool. I expect you'll want to do most of the i2c stuff at a high level at first?21:41
ysionneauso that it's clear21:41
mithroMight want to put it in a little ccan library or something?21:42
ysionneauI'll write some fx2_flash() function which resets the fx2 and then waits for fx2 to do the start bit and so on21:42
ysionneauthis could run at lm32 fw boot for instance21:42
ysionneauor on demand via debug command21:42
mithroYeah that is a good start21:42
ysionneauccan, I don't know what this is21:43
mithroEventually we should have an interrupt on this i2c bus21:43
ysionneauhttp://ccodearchive.net? ?21:43
tpbTitle: ccan (at ccodearchive.net)21:44
ysionneauI was thinking about irq ...21:44
ysionneaubut maybe this is cleaner and uses less cpu with polling?21:44
ysionneauah ... but then you can't reboot *from* the fx221:44
mithroActually you can21:44
mithroBut don't worry about that to much21:45
ysionneauyes, but I was just thinking that if I just let lm32 send the fw to fx2 when the lm32 boots or on uart command, then if fx2 reboots by itself it won't get any fw from lm3221:45
ysionneauso irq stuff would be needed there21:46
mithroBTW you should make sure your code has the right abstraction so that we can replace the bit banging with an i2c UART in the future21:46
ysionneauexcept if there is a way to probe the fx2 boot from the FPGA21:46
ysionneaui2c UART?21:47
mithroThe fx2 booting is detected by i2c commands happening on the i2c bus21:47
ysionneauyes but if we put an irq on, say, the scl pin, the lm32 will get interrupted quite often21:48
mithroi2c hardware which does the bit banging, so you just write whole cmds/bytes rather then every scl21:49
ysionneauah, an i2c hw core21:49
ysionneauright ok21:49
ysionneauI'll keep that in mind21:50
mithroYou'd interrupt on the first SCL transition then go into fx2 booting mode21:50
ysionneauthen I mask this irq21:51
ysionneauupload the fw, unmask it?21:51
ysionneauwhat if the fx2 does some other i2c commands after booting?21:51
mithroIt's okay if the lm32 is busy just doing the fx2 upload for now21:51
mithroIE it ignores all other tasks21:52
ysionneauhmmm, is the fx2 supposed to do i2c transactions during the product life time, except for fetching its fw?21:54
ysionneauok, so those would also wake up the lm32 which will think the fx2 is booting and wants its firmware fed, right?21:55
mithroMultiple i2c busses21:55
ysionneauah ok21:56
ysionneauok so not on the i2c bus MAC_*21:56
ysionneauabout ccan, I didn't know about that, but I was planning on working your repo and working in my branch and just adding a fx2_flash.c/.h file in the lm32 fw directory21:57
mithroI think there will be like ~10 i2c busses in total eventually21:57
ysionneauwow =)21:57
ysionneaui2c highway21:57
mithroCcan is useful if you want a quick unit testing library or something21:57
mithroYou can think of it like a "useful C code snippets"21:58
mithroEach HDMI port has an i2c bus on it21:58
mithroBTW I forgot to mention it21:59
ysionneauah yes the DDC or something21:59
mithroI believe EDID is actually an i2c EEPROM of the same sort the fx2 expects21:59
mithroSo you might be able to crib a bunch of stuff from the HDMI in core22:00
ysionneauah interesting22:02
mithroBTW you saw https://github.com/timvideos/HDMI2USB-misoc-firmware/issues/21 right?22:02
tpbTitle: Add I2C EEPROM gateware and software · Issue #21 · timvideos/HDMI2USB-misoc-firmware · GitHub (at github.com)22:02
ysionneauthis one is entirely in gateware it seems22:02
*** se6astian is now known as se6astian|away22:02
ysionneauthe edid is in an sram and is output via an fsm22:03
mithroysionneau: yeah, but there isn't a good reason for it to be if I understand correctly22:03
ysionneauI'll have a look22:03
mithroysionneau: your @fallen on GitHub right?22:03
ysionneauyes I am22:04
mithroThe EDID is often stored in the monitor in amemory device called a serial PROM(programmable read-only memory) orEEPROM (electrically erasable PROM) and is accessible via the I²C-bus at address 0x50.[2]The EDID PROM can often be read by the host PC even if the display itself is turned off.22:04
tpbTitle: Extended Display Identification Data - Wikipedia, the free encyclopedia (at en.m.wikipedia.org)22:04
ysionneauok good22:04
ysionneauso I can definetely have a look at the gateware FSM to make sure I understand what I'm supposed to do in software22:04
ysionneau22:59 <@mithro> BTW you saw https://github.com/timvideos/HDMI2USB-misoc-firmware/issues/21 right? < yes saw it22:05
tpbTitle: Add I2C EEPROM gateware and software · Issue #21 · timvideos/HDMI2USB-misoc-firmware · GitHub (at github.com)22:05
mithroYeah, eventually I expect we'll rip out the EEPROM gateware in the HDMI in and replace it with software22:07
mithroGot to run22:08
ysionneauwhy do you want to fix something which is working ? :p22:08
ysionneauok se you !22:08
mithroBe back in an hour or two22:08
ysionneauah I'll be sleeping22:08
mithroYeah, I figured as much22:08
mithroOtherwise I wouldn't be typing so much on my phone :-)22:09
*** CarlFK has joined #timvideos22:10
*** ChanServ sets mode: +v CarlFK22:10
ysionneaulinking it for the i2c "protocol" ?22:10
ysionneauI already found some 24A* datasheet22:11
ysionneauthe TMC one22:12
ysionneaubut yours is interesting as it's clearly branded as a VESA EDID eeprom =)22:12
ysionneauinteresting how in fact the protocole does not seem to be I2C at all22:16
ysionneaujust clocking the bits out of it22:17
ysionneauok I have some more reading to do then =)22:23
CarlFKmithro: put something in testing please23:07
mithroysionneau: this might be useful too https://github.com/groeck/i2c-tools/tree/master/eeprom ?23:16
tpbTitle: i2c-tools/eeprom at master · groeck/i2c-tools · GitHub (at github.com)23:16
*** Bertl_oO is now known as Bertl_zZ23:24

Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!