#timvideos

NB, we have a Digilent Analog Discovery 2 here, so I could probably use that to check the HSYNC and VSYNC for VGA...

Title: GitHub - dirjud/Nitro-Parts-lib-Xilinx: This is mainly a simulation library of xilinx primitives that are verilator compatible. (at github.com)

mithro: https://github.com/timvideos/litex-buildenv/iss...

Title: conda fails to install python=36 (verification error) · Issue #22 · timvideos/litex-buildenv · GitHub (at github.com)

mithro: It appears by far the fastest solution to "conda can't upgrade from python=3.5 to python=3.6" is just to move build/conda aside, and run `scripts/download-env.sh` to install a brand new conda environment. (That issue has a bunch of other things I tried unsuccessfully to hopefully avoid others having to repeat the same problems...)

mithro: I've just added a comment to the litex-buildenv commit changing to python-3.6 pointing at that issue, to help someone else find the quicker solution...

ewen: I found that too. I was doing a clean build for other reasons, so just ended up removing build completely.

futarisIRCcloud: Thanks for the confirmation. Removing build/conda seems sufficient, and faster than having to, eg, download Linux git again too.

But it's probably going to be a trap a bunch of the fpga-miniconf users fall into...

mithro: Maybe post something to fpga-miniconf list pointing out issue, and quickest work around?

_florent_: So in litesdcard, it's not possible to read SCR register if you attach SDCore to the BIST Module, correct?

err let me rephrase that

_florent_: So in litesdcard, it's not possible to read the data returned from the SD card* if you attach SDCore to the BIST Module, correct?

This complicates supporting old cards that don't support CMD23 (SET_BLOCK_LENGTH)... b/c reading the SCR register (ACMD55) is how you figure out which commands are supported.

And SCR register is a command that has a data payload.

_florent_: Current status is that I'm writing a Wishbone bridge module for litesdcard, but there's gonna be some conflicts w/ the existing BIST that I need help with once I submit a pull. I'll let you know.

cr1901_modern: litesdcard only has 1 sink/1 source. For now the user should provide the mux before that so that you can have BIST + another sink/source.

_florent_: Is there a way to access the memory port of a sink?

_florent_: Ignore last q

CarlFK: look at the comments in the github UI -- my comments would make more sense then

mithro: um.. I just looked. still not sure what you mean

CarlFK[m]: https://usercontent.irccloud-cdn.com/file/Fca9L...

mithro: still lost

CarlFK: That comment refers to the new lines inside the def find_last_rev function

oh. so you are top posting ;)

mithro: um.. my UI is a little different ... sec...

mithro: https://veyepar.nextdayvideo.com/static/temp/gh...

CarlFK: click the "Files changed" tab

mithro: still not sure which lines you are talking about

CarlFK: https://github.com/timvideos/HDMI2USB-mode-swit...

Title: finds last rev that built, and caches github by CarlFK · Pull Request #97 · timvideos/HDMI2USB-mode-switch · GitHub (at github.com)

mithro: umm... what comment and what lines are you talking about ?

CarlFK: That is the UI I'm looking at

CarlFK: Do my comments make more sense when looking at that UI?

nope

CarlFK: Okay, I'll try again

mithro: lets go though each one.

FIXME: Move cache to a class

do you want that to be added as a #comment or actually re coded now?

CarlFK: There - I removed my old comments and added more explicit requests

CarlFK: do they make more sense?

mithro: yes.

mithro: the UI doesn' t make it obvious if you are commenting on code above or below

CarlFK: Okay - the general idea is that you put the comment at the start of the scope the comment relates to (IE for functions that would be after the function definitions, for files that would be at the top/first line of the file, etc)

mithro: that seems odd. is it somehow dictated by something, or some convention ?

CarlFK: Common convention

CarlFK: IE -- It's were you would put a docstring for the object

mithro: huh. seems more like top posting to me.

CarlFK: Generally after fixing each comment you would reply with "Done" or "Fixed" on each one separately btw

mithro: ah - sure. next time

CarlFK: Better code review tools let you basically "mark comment as fixed"

mithro: I'm not sure this is important, but if you have a sec.. hdmi2usb-mode-switch -v --by-type atlys --flash-image ... Numato Opsis in 'operational' mode ... https://paste.ubuntu.com/26537347/

Title: Ubuntu Pastebin (at paste.ubuntu.com)

should it be telling me about Opsis when I say --by-type atlys ?

CarlFK: because you have -v

k - just making sure.

mithro: another if you have time. current udev rule package (no idea if it was ever tested for Atlys) $ hdmi2usb-mode-switch -v --by-type atlys --load-fx2-firmware hdmi2usb.hex (no sudo)

Found exart-uarts at [LsusbDevice(04e2:1410:0001 /dev/bus/usb/002/012)] associating with Atlys at []

Title: Ubuntu Pastebin (at paste.ubuntu.com)

vizzini-dkms Installed: 1.0.0-5~ppa3~ubuntu14.04

I forget if that kicks in before --load-fx2-firmware hdmi2usb.hex

dmesg shows [4885339.175478] usbcore: registered new interface driver vizzini

CarlFK: You don't have both Atlys USB plugged in?

CarlFK: Found exart-uarts at [LsusbDevice(04e2:1410:0001 /dev/bus/usb/002/012)] associating with Atlys at []

CarlFK: That means there is something wrong with the FX2 on the Atlys

mithro (IRC): huh. "yes". fixed by unplugging everything plugging it back in.. etc. odd that this is an atlys mouted in a case with usb cables attached to the side of the case, and they looked well seated.

Found exart-uarts at [LsusbDevice(04e2:1410:0001 /dev/bus/usb/002/064)] associating with Atlys at [Board(dev=LsusbDevice(1443:0007:0000 /dev/bus/usb/002/065), type='atlys', state='unconfigured')]

There could multiple things going wrong

CarlFK[m]: mode-switch is not well tested with an Atlys and the exart serial interface makes things even more complicated

mithro (IRC): this line line wasn't showing up in lsusb Bus 002 Device 068: ID 1443:0007 Digilent Development board JTAG

and I am sure it was before. and didn't touch cables.