_florent_ changed the topic of #litex to: LiteX FPGA SoC builder and Cores / Github : https://github.com/enjoy-digital, https://github.com/litex-hub / Logs: https://freenode.irclog.whitequark.org/litex
tpb has quit [Remote host closed the connection]
tpb has joined #litex
CarlFK has quit [Quit: Leaving.]
Skip has quit [Remote host closed the connection]
Degi has quit [Ping timeout: 256 seconds]
Degi has joined #litex
peeps[zen] has joined #litex
peepsalot has quit [Ping timeout: 246 seconds]
_whitelogger has joined #litex
_whitelogger has joined #litex
CarlFK has joined #litex
CarlFK has quit [Client Quit]
CarlFK has joined #litex
kgugala__ has joined #litex
kgugala_ has quit [Ping timeout: 258 seconds]
kgugala__ has quit [Read error: Connection reset by peer]
kgugala has joined #litex
captain_morgan has quit [Quit: Ping timeout (120 seconds)]
captain_morgan has joined #litex
<xobs> gregdavill: I just merged the eptri cdc stuff. It passes the testbench now (patches pending) with a sysclk of 100 MHz. Still need to feed it a 48 MHz clock. And probably should feed it a 12 MHz clock as well, though I suppose you could just flop it.
_franck_ has quit [Ping timeout: 256 seconds]
<CarlFK> kgugala: how do I get to the console?
Skip has joined #litex
<_florent_> somlo: i did some testing on Mor1kx: https://github.com/enjoy-digital/litex/issues/544
<tpb> Title: Mor1kx support broken by #517 · Issue #544 · enjoy-digital/litex · GitHub (at github.com)
<tpb> Title: litex/common.h at master · enjoy-digital/litex · GitHub (at github.com)
<somlo> apologies for not catching it when I reviewed that patch :(
<somlo> MMPTR should always be `unsigned long`, i.e. `unsigned int` on 32-bit CPU/alignment, and `unsigned long long int` on 64-bit CPU/alignment
<somlo> those are the only options, regardless of CSR_DATA_WIDTH (which indicates how many of the bits within a MMPTR are actually *populated*)!
<somlo> I'll send a fixup PR in a minute
<tpb> Title: csr: fix simple accessor alignment by gsomlo · Pull Request #545 · enjoy-digital/litex · GitHub (at github.com)
<zyp> ttps://bin.jvnv.net/file/GG3NH.png
<zyp> sorry, wrong channel
<kgugala> CarlFK: what console you want to access?
<CarlFK> kgugala: the thing that lets me see status and configure things like what input is hooked to what output
<kgugala> there is no firmware running on the netv2 now
<CarlFK> kgugala: on the Opsis, tio /dev/ttyACM0 ... get a prompt, type "version"
<kgugala> everything is configured from the driver
<CarlFK> ah, got it. I think...
<kgugala> can you try different resolution on pi ?
<CarlFK> maybe, but I have better hdmi sources. like my laptop.
<kgugala> sure
<CarlFK> shoud the netv2's tx hdmi be sending anything ?
<kgugala> not this version you have
<kgugala> we have a wip version of the transmitter on the v4l2-output branch of the repo you're using
<CarlFK> I'll ignore that for now
peeps[zen] is now known as peepsalot
<kgugala> if you have e.g opsis board you can generate a pattern with it
<kgugala> opsis can also help with fixing video timings
<kgugala> so you can use opsis as repeater
<CarlFK> should netv2 (we need a shorter name for that) be sending out edid?
<kgugala> yes I think it does that
<CarlFK> opsis out to netv2 in: output0: 1280x720@50.00Hz from pattern (underflows: 0)
<CarlFK> woot. gst showing it no skew!
<CarlFK> 62fps
<kgugala> yaay
<kgugala> you can try forwarding some other video source through opsis
<kgugala> like output from a laptop
<kgugala> and watch a movie via netv2/pcie in gst ;)
<CarlFK> is there any compression or anything that might cause artifacts?
<kgugala> no
<kgugala> it takes the video as it comes
<kgugala> uses DMA to transfer the video data over pcie to host memory
<CarlFK> thats what I figured. so no point in testing that sort of thign
<kgugala> there is no compression there (maybe just rgb888 -> yuv422 conversion)
<CarlFK> mithro: ^^^ success!
<kgugala> movies are for testing thighs like e.g. tearing
<kgugala> *are good for testing
<CarlFK> https://github.com/CarlFK/clocky I made that to help test sync, and the fast moving sweep hand is good for that
<tpb> Title: GitHub - CarlFK/clocky: analog clock with sweep hand (at github.com)
<CarlFK> getting the right wx libs installed ... grumble
_franck_ has joined #litex
<CarlFK> kgugala: hmm, it seems to have stopped updating....
<CarlFK> opsis uptime on the pattern, isn't ticking any more
<CarlFK> gst frame counter stopped too
<kgugala> sth got disconnected?
<CarlFK> sth?
<kgugala> like HDMI cable?
<CarlFK> I did un/re plug it, and it kept going after that
<tpb> Title: Ubuntu Pastebin (at paste.ubuntu.com)
<kgugala> those prints with pll etc means that the device locked on signal (later it prints the delay settings)
<CarlFK> ^c the gst line, re-ran it. it did 45 frames and stopped
<CarlFK> oh boy. did that 2 more times, now it is working again.
<CarlFK> cpu - 4 cores, each between 10-20%
<kgugala> I'd say it's displaying what consumes the cores
<kgugala> data is transferred over pcie by DMA in fpga
<kgugala> so it does not consume the CPU
<CarlFK> the hangs or pause are concerning
<CarlFK> cpu goes to about 0 when that happens
<CarlFK> dmesg is full of [67712.275108] x86/PAT: v4l2src0:src:20261 map pfn RAM range req uncached-minus for [mem 0x35e00000-0x35fc1fff], got write-back
<CarlFK> and this 2x a second: [68401.017895] litepcie ph -5 1 0 // charsync 111 [8 6 6] // WER 0 0 0 // chansync 1 // res 1280x720
<kgugala> the latter means it does not see the video input
<kgugala> looks like your stream is no stable
<kgugala> can you try different cable ?
<CarlFK[m]> Swapped cables, and this happened
* CarlFK[m] uploaded an image: 20200529_123052_819113531581712419.jpg (6034KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/mdMzNebQOyIXBIJOcIXCFYdH >
<CarlFK> ^c, gst, 44 frames and froze
<CarlFK> swapped cable from short magic to 3' normal probably to spec, same thing in dmesg
_franck_ has quit [Ping timeout: 260 seconds]
<CarlFK> kgugala: fps has dropped to 42 - no clue why. nothing has changed as far as I know
<kgugala> unsure what happened, I was keeping it running for 24h and haven't noticed such issues
<CarlFK> kgugala: for output, pretty please don't have black be the default for when there is no input.
<CarlFK> either color bars or some unique color that makes it obvious where the signal is coming from
scanakci has quit [Quit: Connection closed for inactivity]
<CarlFK> that just landed in the Opsis, it's helpful. ans surprising because I forgot about it
<CarlFK> and by "just landed" I mean the patch was accepted in 2019, and someone was able to build it in Jan 2020, and I'm not exactly sure where a working bitstream file is.
_franck_ has joined #litex
kgugala_ has joined #litex
kgugala has quit [Ping timeout: 260 seconds]
kgugala_ has quit [Read error: Connection reset by peer]
kgugala has joined #litex
kgugala_ has joined #litex
kgugala has quit [Ping timeout: 272 seconds]
kgugala has joined #litex
kgugala_ has quit [Ping timeout: 272 seconds]
kgugala_ has joined #litex
kgugala has quit [Ping timeout: 256 seconds]
kgugala has joined #litex
kgugala_ has quit [Ping timeout: 258 seconds]
<tpb> Title: GitHub - mithro/HDMI2USB-dev-boards (at github.com)
kgugala_ has joined #litex
kgugala_ has quit [Read error: Connection reset by peer]
kgugala_ has joined #litex
kgugala has quit [Ping timeout: 246 seconds]
miek has quit [Quit: miek]
miek has joined #litex
mahmudov has joined #litex
mahmudov has left #litex ["Leaving"]
somlo has quit [Ping timeout: 260 seconds]
somlo has joined #litex