Saturday, 2016-04-16

*** tpb has joined #timvideos00:00
mithroxfxf / CarlFK: IE Permanently put new firmware onto the board03:32
shenkimithro: which tree should i be using?04:42
mithroshenki: well, I haven't rebased my or1k stuff recently04:42
mithroshenki: are you planning on using the Atlys/Opsis or the minispartan6+?04:43
shenkimithro: minispartan604:43
shenkibecause i can have it on the couch with me04:43
mithroshenki: well, do you want to start with trying to replicate our stuff?04:46
mithroshenki: actually, do you want to start with fixing that framebuffer problem in our current firmware?04:49
shenkimithro: so i can use master for that?04:49
mithroshenki: Well, you should be able to do that without any hardware right? It's just about making the .elf output correct?04:50
mithrobut I just pushed my latest stuff at
tpbTitle: GitHub - mithro/HDMI2USB-misoc-firmware at shenki-minispartan6+ (at
mithroshenki: so you could try downloading that and seeing if it builds and runs on your minispartan6+04:51
mithroshenki: My hack to fix the frame buffer problem -
tpbTitle: firmware/lm32: Move the framebuffers down. · mithro/[email protected] · GitHub (at
mithro_florent_: It really does look like it's either the ethernet or jpeg encoder which is causing the routing problems04:55
*** Bertl_zZ is now known as Bertl_oO04:58
mithroshenki: I'm going to see if HDMI out works04:59
mithroshenki: if HDMI out works, then I'm happy to merge the minispartan6+ support and flip to trying to get the or1k working04:59
mithroshenki: after the or1k stuff, I think I'm going to go back to doing the ISE determinism hacks and then the timing refactor stuff05:01
mithrowell, my screen detects some type of signal - but is saying "not supported"05:04
mithroI think this means that the signal being output is incorrect05:05
mithroActually - this makes sense because I did some "incorrect" hacks to the HDMI core to get it to even compile05:06
mithroxfxf: ping?05:11
* mithro looks in the logs for how _florent_ mentioned to fix it properly05:15
mithroshenki: any luck in getting that to build?05:15
shenkimithro: yeah, a few false starts05:18
shenkimithro: xilinx is doing it's thing now05:18
mithroshenki: It should take less than 5 minutes05:24
shenkiok, i think that's coz it tried to do a parrelel make05:24
shenkimithro: nice05:25
mithroshenki: It looks better on the video then in real life05:25
mithroshenki: That output is very weird05:25
shenkimithro: once i power cycle, what's the command to fire up flterm?05:26
mithroshenki: Did you fix the "Error: libusb_claim_interface() failed with LIBUSB_ERROR_BUSY05:26
mithroError: unable to open ftdi device with vid 0403, pid 6010, description '*' and serial '*'" error?05:26
shenkimithro: yeah, by doing 'make -j1'05:26
mithroNothing in our makefile is safe for usage with "make -jX" where X > 105:27
shenkiwe should disable that in the makefile05:27
shenkiive got ccontrol installed, so it defaults to -j4 whenever i a05:27
mithroHow do we do that?05:27
mithroChanging to 800x600 seems to make the output more stable05:28
tpbTitle: GNU make (at
mithrothat would be good05:29
shenkithat is a handy option05:29
shenkimithro: so how do i bring up flterm without having it load the firmwrae?05:29
mithroshenki: after booting you can either hit the push button or type "reboot" and then hit "s" to load via serial05:30
mithroshenki: I haven't added anything like "make flash-clear" yet05:31
shenkimithro:          built: Apr 16 2016 14:51:1805:32
shenkiis it supposed to stop at that -- ?05:32
mithroshenki: no05:32
mithroit should stop at a HDMI2USB> prompt05:32
mithroIf you hit enter a couple of times what happens?05:33
shenkithis doesn't have the hdmi device blatting over the firwmare, does it?05:33
mithroshenki: shouldn't05:33
mithroshenki: well, it will be blating over stuff, just above were the firmware is being loaded05:34
mithroshenki: Can you pastebin the output05:34
shenkimithro: it does it at different points each time i loda05:35
shenkiboth when i load from flash, and from serial05:36
shenkimithro: ok, flashboot just got to a prompt05:36
shenkibut it's not reliable05:36
shenkiMiSoC BIOS   http://m-labs.hk05:36
shenki(c) Copyright 2007-2014 Sebastien Bourdeauducq05:36
shenkiRevision 61a14086 built Apr 16 2016 14:46:5005:36
shenkiBIOS CRC passed (ffdd022c)05:36
tpbTitle: M-Labs » open source experimental physics (at
shenkiRunning on MiSoC rev. 61a14086 (sysid:UN) at 80MHz05:36
shenkiInitializing SDRAM...05:36
shenkiMemtest failed: 5/532736 words incorrect05:36
shenkiMemory initialization failed05:36
mithroshenki: can you power cycle the board and then connect to it really quickly so you get the full booting?05:37
shenkiserialboot just worked05:38
shenkithis is from a fresh plugin:05:39
shenki[FLTERM] Starting...05:39
shenkiMiSoC BIOS   http://m-labs.hk05:39
shenki(c) Copyright 2007-2014 Sebastien Bourdeauducq05:39
shenkiRevision 61a14086 built Apr 16 2016 14:46:5005:39
tpbTitle: M-Labs » open source experimental physics (at
shenkiBIOS CRC passed (ffdd022c)05:39
shenkiRunning on MiSoC rev. 61a14086 (sysid:UN) at 80MHz05:39
shenkiInitializing SDRAM...05:39
shenkiMemtest failed: 33/532736 words incorrect05:39
shenkiMemory initialization failed05:39
mithroYeah, doing "reboot" doesn't seem to be working here either05:40
mithrowhat B= and T= are you using?05:40
shenki(H2U B=minispartan6 T=base)05:41
mithrodoes it always work from pressing the reset button?05:44
mithroyou should be able to press it while still being connected05:44
shenkisame output05:46
shenkiInitializing SDRAM...05:46
shenkiMemtest failed: 33/532736 words incorrect05:46
shenkiMemory initialization failed05:46
mithroshenki: how often does it work verse fail?05:47
shenkiflashboot then gets me to a prompt05:47
mithroshenki: does the memtest *always* fail?05:47
shenkisome corruption in the text: ---------------------�-------------------------�05:47
shenkimithro: yeah, it appears so05:47
mithroshenki: can you do a "make clean; make gateware" again?05:47
mithroshenki: I think there are some missing timing constraints or something here05:50
mithrocan you check what it says the "maximum frequency" of the design is?05:50
mithroshenki: It seems like about 50% of the time it produces non-working gateware05:53
shenki   Minimum period:  11.430ns (Maximum frequency:  87.489MHz)05:53
mithrothat looks pretty good05:54
shenkicrashed first go05:55
shenki         built: Apr 16 2016�15:16:2805:55
shenkifroze there05:55
mithrodid memtest fail?05:56
shenkii loaded using jtag, so the console wasn't up05:56
mithromemtest shouldn't fail, if it does there is no point going any further05:56
shenkii will flash and try05:56
mithroI believe you can have the console up while flashing on the minispartan05:56
shenkidoes reset clear the gwateware, or just reset the system?05:56
mithrojust reset the system05:57
shenkiit appears it clears the contents of the fpga05:58
shenkiwhen i reset the old gateware version came up in the bios05:58
shenkiflashed, now it's new05:58
shenkiMemtest failed: 25/532736 words incorrect05:58
shenkiMemory initialization failed05:58
shenkistill failing05:58
mithroshenki: so, you did a "make flash" then pulled the USB plug out of the board and plugged it back in?05:59
shenkinah, tapped the reset button05:59
mithroyou can't do that06:00
shenkiin what sense?06:00
shenkiit ran the new gateware - the version changed06:00
mithroyou have to full power cycle it after the flash for some reason at the moment06:00
shenkievidence here says otherwise06:01
shenkianyway, i pulled it out and it has the same behaviour06:01
mithroshenki: that doesn't work here06:01
mithrobut my memtest does06:01
mithroshenki: do all your minispartan6+ LEDs come on during the "make flash" ?06:02
mithroshenki: This is why I want to work on the determinism stuff :)06:02
mithrocan you upload your .log file somewhere?06:03
mithroshenki: oh, and what is your git revision number?06:04
mithroshenki: Hrm, the reset button appears to be working in the spi flashing mode now - I tested that multiple times and it *definately* didn't work before!?06:06
mithroshenki: your revision number if your logs above looks wrong?06:07
mithroshenki: should be 46aa77d69700929e0eb9fc35019bfb586780ed44 or ad0ae48d915ba925cd94b8af6c308a9b0feee82d06:07
mithroshenki: poke?06:14
mithroshenki: any idea how I test the .NOTPARALLEL thing?06:18
_florent_mithro: nice for the video output, you are probably very close. It seems the DMA is not correctly synchronized with the VTG06:23
mithro_florent_: Yeah06:23
mithro_florent_: It seems pretty stable on 800x600 but there is some minor corruption on the output06:23
_florent_maybe that's the sdram that is not fast enough with there is the synchronization issue06:24
_florent_I'm just passing bye, I won't be there this weekend06:24
mithro_florent_: yeah - I would have thought that it would cause lots of fifo error output on the console06:25
mithrobut I'm not seeing any06:25
_florent_I'm not sure video output is printing some console message in case of underflow06:25
_florent_It can be worth adding that06:25
mithroAhh, was it only for video input?06:26
_florent_yes I think we only have overflow message for video input06:26
* mithro ponders what happened to shenki06:27
mithroeat by a couch maybe :P06:27
_florent_btw I almost got hdmi2usb working with my opsis-soc repo06:28
_florent_just need to do some fix06:28
mithro_florent_: okay06:28
_florent_I have to go, good hacking!06:28
mithro_florent_: have a good weekend06:29
*** tija has joined #timvideos08:13
mithrowell, heading to find some dinner09:07
mithroshenki: I figured out my confusion11:10
mithroshenki: the button on the pipistrello doesn't reset the device11:11
mithroit looks identical to the minispartan6 button11:11
mithrotija: can you take a look at ?11:11
tpbTitle: Reduced stages in RGB2YCbCr Module by Vikram9866 · Pull Request #2 · enjoy-digital/litejpeg · GitHub (at
*** Andrew____ has joined #timvideos11:57
mithroHi Andrew____12:18
*** Andrew____ has quit IRC12:23
xfxfmithro: delayed ack12:28
mithroxfxf: How are things going?13:33
CarlFKgod dam firewire still pissing me off...15:11
CarlFKbecause I want to bring a camera to Sheila's parents, and I won't be able to get it from the camera to my laptop... even if I record to tape first (that seemed like a solution)15:12
*** tija has quit IRC15:19
mithroshenki: I have a present for you tomorrow16:50
*** CarlFK has quit IRC18:25
*** Bertl_oO is now known as Bertl_zZ18:39
*** ssk1328 has joined #timvideos20:45
*** ssk1328 has quit IRC23:38

Generated by 2.13.1 by Marius Gedminas - find it at!