Thursday, 2015-09-03

*** tpb has joined #timvideos00:00
mithro_florent_: sorry need to concentrate on finishing this stuff - maybe tomorrow?01:07
*** rohitksingh has joined #timvideos01:25
CarlFKmithro: you proly mean me ;P01:47
CarlFKI packaged gstreamer1.0-dvswitch  and dvsource-v4l2-other and they are in my PPA01:48
CarlFKbb in 2001:48
tpbTitle: EDID Related Code · timvideos/HDMI2USB Wiki · GitHub (at
mithroCarlFK: actually -
tpbTitle: EDID Related Code · timvideos/HDMI2USB Wiki · GitHub (at
mithroEnable that and then plug / unplug the connector02:11
mithroCarlFK: that might give more information about what is going on with the hotplug line02:11
mithroCarlFK: If that doesn't shine any light on the issue you are going to need an oscilloscope I think02:12
CarlFKthe info is in dmesg, right ?02:13
mithroCarlFK: yeah02:13
mithroCarlFK: you might see something about disabling / enabling interrupts or something02:13
CarlFKmithro: every 10 seconds I see:02:16
CarlFK[11042.816502] [drm:intel_tv_detect], [CONNECTOR:25:SVIDEO-1] force=002:16
CarlFKand now this just one: [11056.985696] [drm:i965_irq_handler], pipe A underrun02:16
CarlFKun/plugging now02:16
CarlFKunplug, dmesg new:
tpbTitle: Ubuntu Pastebin (at
CarlFKmithro: pluged in, dmesg:  [11274.559615] [drm:i915_hotplug_work_func], Connector HDMI-A-1 (pin 4) received hotplug event.02:21
tpbTitle: Ubuntu Pastebin (at
mithroCarlFK: so - I think the lines which say "NACK" are the cause of the issue02:22
mithroand actually - it might be something similar to issue described at - "The main bug I had was not having implemented support for I2C writes; I'd assumed the PC would simply read the EEPROM once and be done with it, but it looks like it does a couple 1-byte reads, then a write to set the EDID EEPROM02:24
mithroaddress followed by a read of the full content"02:24
mithro_florent_: ping - are you able to confirm that the EDID code in misoc deals with the address write issue mentioned above?02:25
mithro_florent_: I also thing really need to get a message echoed on the serial console when EDID operations occur on a port.02:26
mithroCarlFK: brb02:27
mithroCarlFK: Can you document somewhere your laptop information and how it is connected to the computer?02:29
CarlFKmithro: github issue?02:30
mithroYes, you can use one if you want02:31
*** rohitksingh has quit IRC02:40
*** rohitksingh has joined #timvideos02:54
CarlFKmithro: having a monitor hooked up to the output shouldn't matter right ?02:58
tpbTitle: laptop not seeing valid display device · Issue #38 · timvideos/HDMI2USB-misoc-firmware · GitHub (at
mithroCarlFK: correct03:44
CarlFKmithro: given that is the only thing I can think of... should I bother unpacking one, or spend the time poking other things>03:46
mithroCarlFK: You could go and have a poke around the EDID code in the firmware and see if you see anything03:47
mithroCarlFK: you've tried a whole lot of other devices right?03:48
CarlFKmithro: I did last night.  including this laptop03:49
mithroDo you have a different adapter you can try? This is a mini-DP to HDMI right?04:27
mithroCarlFK: can you try the other HDMI input too?04:28
mithroBTW Did you get the raspberry pi working in the end?04:29
mithroCarlFK: If you get a moment, you should see if you can fix the fpgalink error that comes up when flashing. That is just some simple Python code that should be pretty easy to fix by adding the right try/except bits...04:31
CarlFKmithro: hdmi0, no work, but different dmesg I think?
tpbTitle: Ubuntu Pastebin (at
mithro9:25 pm <@mithro> Do you have a different adapter you can try? This is a mini-DP to HDMI right?04:32
CarlFKmithro: I never got the pi to do 720p04:33
CarlFKno - full size hdmi04:33
mithroWas the HDMI2USB detecting the low resolution?04:33
CarlFKnot tonight.04:34
CarlFKlet me try again to be sure04:34
CarlFKmithro: pi xrandr default connected 640x480+0+0 0mm x 0mm04:36
CarlFKhdmi_in0:  0x004:36
CarlFKhdmi_in1:  0x004:36
CarlFKhdp_toggle doesn't have any affect on the Atlys right?04:37
mithroIt shouldn't04:55
mithroMake sure you try both inputs to be certain.04:55
mithroBTW you made very sure that your jumpers on the board are configured correctly, right?04:56
CarlFKmithro: would it work if they were not?04:57
CarlFKI checked a few weeks ago. I was surprised that the 2 jumpers were next to each other.  I expected one by each hdmi port04:58
mithroIt would work unreliable04:59
CarlFKoh oh.. I only checked 2.  I see there are 3..05:02
CarlFKmithro: is JP2 next to 6 and 7?05:05
CarlFKmithro: any idea where a map of the board is so I can look up where JP2 is?05:19
mithroCarlFK: I sure there was a screenshot or something in the wiki05:20
CarlFKmithro: k - silly me was looking at all the docs like
mithroCarlFK: the docs should also have it05:20
CarlFKmithro:  best picture I have found.. any idea where JP2 is?05:22
tpb<> (at
mithroIIRC the jumpers you care about are05:24
mithro * Up in the top corner near the HDMI In 0 and the Ethernet port05:24
mithro * Down in the bottom corner near HDMI In 2 and the PMOD header05:25
mithroThey should be marked on the board05:26
CarlFKthe hdmi in 0 and 1 labels are reversed of what the misoc firmware calls them right ?05:27
mithroNo idea, sorry05:28
CarlFKis JP2 one of the 3 in the lower left?05:29
CarlFKon my board, all 3 jumpers are in place, just like the picture05:30
CarlFKmithro:  what is "HDMI In 2" ?05:31
CarlFKfound a better image.  Still can't find JP2
CarlFKmithro: do you have an Atlys handy?05:39
mithroCarlFK: not at this moment05:39
CarlFKI can not find JP205:39
CarlFKnor can I find any doc describing what all the jumpers do.05:40
CarlFKmithro: JP2/SDA (which connects the EDID lines from J1/HDMI IN to JA/HDMI OUT).05:41
CarlFKI am 1/2 thinking that might be causing the problem I am having tonight with no lcd hooked up05:42
*** rohitksingh has quit IRC05:42
CarlFKmithro: I give up.05:52
mithroCarlFK: So, the only two devices you haven't been able to get working are your RPi and your laptop?05:53
mithroCarlFK: can you try other devices you have?05:53
CarlFKmithro: I have 4 hdmi sources:05:54
CarlFKnegk , pi, twist (my laptop) yogi (Sheila's laptop)05:54
CarlFKnegk is what I have been successfully testing with the last 3 days05:55
CarlFKlast night when I was trying people's laptops, these two didn't work right away, so I gave up and went on to the next:  twist (my laptop) yogi (Sheila's laptop)05:56
mithronegk doesn't work now?05:56
*** rohitksingh has joined #timvideos05:56
CarlFKmithro: correct05:57
mithroBut it worked yesterday?05:57
CarlFKmithro: correct05:58
mithroAnd the Pi worked yesterday?05:58
CarlFKnot really - I think I saw hdmi_in1:  640x480 but never set that vga_mode and for sure didn't see it displayed on the hdmi out or usb uvc05:59
CarlFKI am not sure when I will get another chance to do talks, so I think you should help me get this working tomorrow else there may not be another good chance right away06:03
CarlFKmithro: I found SDA... "JP2/SDA (which connects the EDID lines from J1/HDMI IN to JA/HDMI OUT)."06:04
CarlFKbut it is 2 jumpers, and no JPx label anywhere06:04
CarlFKdo you know where the "JP2/SDA" instructions came from?06:05
mithroThere is one jumper which is a 2x206:05
mithroThere are two 1x2 jumpers06:06
CarlFKis there some reason you don't say "there are 2 jumpers" ?06:07
CarlFK"one jumper" to me sounds like there is only 1. not 2.  just 1.06:08
mithroRyan wrote those lines I think06:08
CarlFKwhat were you just trying to tell me about how many jumpers there are?06:09
mithroThere are three sets of pins which should have a total of 4 little jumpers06:39
mithroCarlFK: you understand that each jumper electrically connects one signal06:40
mithroCarlFK: there are 4 signals06:40
CarlFKsounds like 2 jumpers to me06:41
CarlFKanyway, I am too tired now.  I asked Ryan what JP2 was.  I suspect it is a typo.06:42
mithroNo, there are 4 of these ->
mithroone per signal06:46
mithroTwo of them go on one 2x2 pin header, the other two go on two separate 1x2 pin headers06:53
*** MaZderMind has quit IRC08:15
*** MaZderMind has joined #timvideos08:17
CarlFK"It's labelled SDA on our boards."  swell.13:22
CarlFKmithro: pulled jumpers, plugged in, negk sees display !14:42
CarlFKand now I can;t seem to get a /dev/video114:54
CarlFK[email protected]:~$ lsusb14:54
CarlFKBus 002 Device 011: ID 04e2:1410 Exar Corp.14:54
CarlFKBus 002 Device 016: ID 1d50:602b OpenMoko, Inc. FPGALink14:54
CarlFKencoder: 1280x720 @ 30fps (22Mbps) from hdmi_in1 (q: 85)14:56
CarlFKHDMI1 connected 1280x720+0+0 (normal left inverted right x axis y axis) 200mm x 112mm14:56
CarlFK_florent_: mithro ... help...   why am I not getting a /dev/video1?14:57
_florent_the fx2 firmware is loadedN14:57
CarlFK_florent_: I think so.  how can I check?14:58
_florent_try to do make load-fx2-firmware to be sure14:59
CarlFK[76763.270655] uvcvideo: Found UVC 1.00 device HDMI2USB (1d50:60b7)15:01
_florent_np, so for the issue with the video input, it was a bad configuration of the jumpers on the Atlys?15:09
_florent_ok thanks15:18
*** sb0 has joined #timvideos15:28
*** sb0 has quit IRC16:04
*** sb0 has joined #timvideos16:17
CarlFK_florent_: well.. something still wonky.16:23
CarlFKhdmi_in1:  0x016:24
CarlFKwith 3 different hdmi sources16:25
CarlFKnegk = hdmi laptop that seems fine when things are working fine.16:25
CarlFK_florent_: I want to power cycle and reflash.  anything I should check first?16:37
*** sb0 has quit IRC16:40
CarlFKrebooted all the boxes.17:04
CarlFKnow I cant get  /dev/ttyV*17:05
CarlFKsudo m-a b-i vizzini17:10
CarlFKgot it.17:10
CarlFKbut negk still no display device17:11
CarlFK_florent_: help ;)17:14
CarlFKwoo. pi ... hdmi_in1:  640x48017:21
CarlFKwoo, pi working.  mostly.17:22
CarlFKabout 4 hours left before I head to the meeting and try to use this17:30
CarlFKcurrently in bad shape17:31
CarlFK(08:06:42 PM) mithro: _florent_: sorry need to concentrate on finishing this stuff - maybe tomorrow?17:36
CarlFKmithro: do you have any time in the next 4 hours?17:36
mithroCarlFK: I'm just getting up now18:21
CarlFKlol - good morning mithro18:21
mithroCarlFK: I only got to bed 4 hours ago18:21
mithroCarlFK: you need to figure out what you are doing differently from the times it worked and times it didn't.18:22
CarlFKmithro: it would help if I knew what was expected so I could report when what I see deviates from expected18:23
mithroI don't understand what you mean18:24
CarlFKmithro: I do this: HDMI2USB>video_mode 918:25
CarlFKSetting video mode to 1280x720 @60Hz18:25
CarlFKun/plug pi18:25
CarlFKhdmi_in1:  640x48018:25
mithroIt is expected that the RPI would output 1280x72018:27
CarlFKthat's what I thought18:27
CarlFKwhat video mode is misoc in when it first boots?18:28
mithroI don't know of the top of my head. What does the status show when you boot it up?18:29
CarlFKhdmi_in0:  0x018:29
CarlFKhdmi_in1:  0x018:29
CarlFKhdmi_out0: 1280x720 from hdmi_in018:29
CarlFKhdmi_out1: 1280x720 from hdmi_in018:29
CarlFKencoder: off18:29
CarlFKddr: read: 1767Mbps  write:    0Mbps  all: 1767Mbps18:29
mithroLooks like 1280x72018:30
CarlFKso  1280x720 should be the only supported res in the EDID when I plug in a source, right?18:30
CarlFKok.  that is one less thing for me to set when I am experimenting18:31
mithroYou can even go look at the code which generates the EDID if you want to be very sure what the exact content should be.18:31
CarlFKI realize it may not be working that way.. but it lets me prioritize what tings I fiddle with18:31
mithroCarlFK: I don't know if setting the video mode works different from the state the board comes up in first boot18:32
CarlFKnothing I do at the HDMI2USB> prompt should cause the laptop to react, right ?18:33
CarlFKgiven Atlys18:33
CarlFKwith the exception of reboot - I can see that being  a special case18:34
mithroNot quite true18:35
CarlFKwhat might cause the laptop to react?18:35
mithroThings at the HDMI2USB> prompt change the internal state of the Atlya18:35
mithroThe laptop could decide to reread that state for any number of reasons outside the Atlys control18:36
CarlFKI would not consider that reacting18:37
mithroIn theory the laptop should only reread the state when the HPD line changes18:37
mithroThe HPD isn't under software control18:37
CarlFKand that is the only thing that would cause the laptop to do anything, right ?18:38
CarlFKlike changing res with HDMI2USB>video_mode   shouldn't cause the laptop to even be aware of it, right ?18:38
mithroIf you ran xrandr on the laptop that might force a reread of state18:38
mithroCarlFK: At the moment we don't know when a laptop rereads state - so we should assume it happens randomly without warning18:39
CarlFKright.  the laptop isn't reacting to the video_mode change.18:40
CarlFKI am trying to focus on doing things that should work18:40
CarlFKsheila just plugged her laptop in - no display device seen18:45
mithroThe laptop could appear to react to video-mode change if the OS was polling the EDID data18:46
CarlFKI am seeing the NAK lines: negk [ 6450.448588] [drm:gmbus_xfer], GMBUS [i915 gmbus dpb] NAK for addr: 0050 w(1)18:47
mithroOn Sheila's laptop?18:47
CarlFKno.  she only had time for a quick "plug in, nothing, ok done" and she got back to work18:48
CarlFKalso, moving the pi between hdmi0 and 1, I only see hdmi_in1:  640x48018:50
CarlFKno change on hdmi_in018:51
mithroCarlFK: please explain exactly what you did - I assume "I unpluged the cable going from the RPi to the Atlys port XX and move it to Atlys Port YY and rebooted the Pi"18:51
CarlFKwhen plugged into 0: hdmi_in0:  0x018:51
CarlFKhdmi_in1:  0x018:51
CarlFKno rebooting.18:52
CarlFK2 cables, one in 0, the other in 1.  move the pi from 0 to 1 back to 0, back to 1.. a bunch of times. maybe 5 times.18:54
CarlFKwhen connected to 1 I see hdmi_in1:  640x48018:54
CarlFKhdmi_in0:  0x018:54
CarlFKconnecting to hdmi_in0 or not connected: hdmi_in0:  0x018:55
CarlFKhdmi_in1:  0x018:55
mithroCarlFK: okay - Can you try using the *same* cable?18:56
mithroI'm just heading to work18:56
mithrowill be back in 15-2018:56
mithroCarlFK: when changing things, change exactly one thing at a time. In your above example the cable plugged into one of the ports could be dodgy.19:13
mithroBTW are you using your crazy in a box thing which has like a ton of cables and connectors?19:15
CarlFKmithro: yes.    It adds 2 hdmi connections to  hdmi_in119:18
CarlFKand it keeps the same cables held in place.19:18
CarlFKso I would call it my sane box thing ;)19:19
CarlFKI used the same cable that worked  from pi to the cables that lead to hdmi_in1, moved it to hdmi_in0 (on the Atlys) still hdmi_in0:  0x019:20
CarlFKmithro: I am pretty good about testing cables by plugging into hdmi lcd.   they all seem to be in good shape.19:22
CarlFKmithro: are you back?  (I need to spend 20 min now and then gathering gear for tonight19:23
mithroThe problem is if the signal integrity is poor or something the LCD might be better at picking it up then the Atlys19:23
mithroLess variables!19:23
mithroI'm at lunch19:24
mithroI also don't have a Atlys board on me19:25
mithroI'd really like to help you figure out the problem but it's probably going to take a very systematic approach with as minimal set up as possible until we can get the extra debug stuff I want.19:29
mithroI'd really like to help you figure out the problem but it's probably going to take a very systematic approach with as minimal set up as possible until we can get the extra debug stuff I want.19:29
mithroI'd really like to help you figure out the problem but it's probably going to take a very systematic approach with as minimal set up as possible until we can get the extra debug stuff I want.19:33
CarlFKnew cable - full size -- micro?  (whatever my dev laptop uses) getting similar NAKs when plugging in19:35
CarlFK  [187077.419743] [drm:gmbus_xfer] GMBUS [i915 gmbus dpb] NAK for addr: 0050 w(1)19:36
tpbTitle: Ubuntu Pastebin (at
mithroYou didn't say my nick so I got distracted.19:56
CarlFKmithro: we should have the same boards to test with20:06
mithroCarlFK: I have multiple Atlys boards with me - but I didn't get home until 5am this morning so didn't get a chance to try and get things working20:11
CarlFKmithro:  it wold have saved me an hour or two trying to figure out what J2 meant20:12
CarlFK"there is no J2" and take a picture of what you consider correct20:12
tpbTitle: IMG_8688.JPG - Google Drive (at
CarlFKmithro: that is just the upper left, ?20:15
mithroHDMI Out - J220:16
CarlFKah.  sorry.  whatever the jumper*S* were that I was looking for20:16
CarlFKlittle late now20:17
CarlFKif you want to connect to the dc10b box that has the 2 usb connections:   ssh -p 22174  [email protected]20:18
CarlFKmithro:  ^^20:20
CarlFKmithro: JP2/SDA (which connects the EDID lines from J1/HDMI IN to JA/HDMI OUT).20:20
CarlFKthat line needs help.20:20
CarlFKI did find SDA, seems to refer to 2 jumpers, I removed them both.20:21
tpbTitle: IMG_8687.JPG - Google Drive (at
tpbTitle: IMG_8692.JPG - Google Drive (at
CarlFKmithro: you have SDA jumpered20:23
mithroCarlFK: Have you seen the diagram at ->
tpbTitle: atlys:atlys:refmanual [Reference.Digilentinc] (at
mithroCarlFK: I just found those pictures20:23
tpbTitle: atlys:atlys:hdmi.png [Reference.Digilentinc] (at
CarlFKmithro: JP2/SDA   I see JP2 in that last one.  I can't find it on the Atlys board20:25
mithroCarlFK: so looking at that diagram - for J1 (HDMI IN) to work correctly with EDID the two jumpers must be *on* JP220:29
CarlFKoh boy.20:31
CarlFKmithro: A) wher is JP2?  B) Ryans readme says:20:31
CarlFKAs the HDMI2USB firmware manipulates the EDID information the following jumpers must be removed;  JP2/SDA (which connects the EDID lines from J1/HDMI IN to JA/HDMI OUT).20:31
mithroCarlFK: so - there should be two groups of jumpers labelled something like SDA/SCL20:36
CarlFKmithro: no more time to work on this today20:37
CarlFKplease make sure Ryans' read me is correct, and add a picture showing where these jumpers are20:37
mithroCarlFK: I'm going to go through the exact same process you are to do the above20:38
CarlFKmithro:  the following jumpers must be removed;  JP2/SDA20:39
CarlFKmithro: you said "the two jumpers must be *on* JP2"20:39
CarlFKdo what I asked :)20:39
mithroCarlFK: so I'm going to look at the code and the diagram and get out a multimeter and try and figure out what the problem is20:39
CarlFKthat works too20:40
mithroCarlFK: but I'm not going to get a chance to do that right now20:40
mithroCarlFK: and your smart enough to do this too20:40
CarlFKthats fine.  I am guessing it is to late to learn anything for tonights talks20:40
CarlFKmithro: I am not smart enough to find JP2.20:41
CarlFKmaybe some day you will help me with that. ;)20:41
mithroCarlFK: You should *understand* what is going on - not just try and follow the instructions blindly20:41
mithroCarlFK: Otherwise you'll have to wait for me to do everything20:42
CarlFKyou told me to follow Ryans20:42
CarlFK I have been asking everyone "where is JP2>?"20:43
CarlFKeven Ryan couldn't really answer that.20:43
CarlFKso yes, I am wating on you.20:43
CarlFKif you can't fingre it out, then we should all just give up.20:43
mithroCarlFK: You are a smart guy - I *know that. Can you try and figure out how you think it *should* work and see if that matches how it does work and then ask questions where things don't match, cross check things in the source code, in the manual, with a multimeter. That is exactly what I will end up doing, I don't have some extra magic....20:51
CarlFKmithro: for tonight, should I put those jumpers back or leave them off?20:53
mithroCarlFK: I have NFI - the wiki says off, that diagram indicates on20:54
CarlFKI'll put them back.  I think I got beetter results20:54
mithroCarlFK: We need to figure out the way we think they should be20:55
mithroCarlFK: there also look to be other jumpers which might be relavent20:56
CarlFKtoo late.  box is packed.20:57
*** CarlFK has quit IRC21:30

Generated by 2.13.1 by Marius Gedminas - find it at!