Wednesday, 2020-01-22

*** tpb has joined #yosys00:00
*** rohitksingh has joined #yosys01:47
*** bitmapper has quit IRC02:05
*** rohitksingh has quit IRC02:47
*** markus-k has quit IRC02:56
*** develonepi3 has quit IRC03:25
*** emily has quit IRC03:25
*** corecode has quit IRC03:26
*** corecode_ has joined #yosys03:26
*** markus-k has joined #yosys03:29
*** citypw has joined #yosys03:31
*** emily has joined #yosys03:34
*** ZirconiumX has quit IRC04:30
*** dh73 has quit IRC05:05
*** attie has joined #yosys05:06
*** tpb has joined #yosys05:11
*** attie has quit IRC05:11
*** X-Scale` has joined #yosys06:00
*** X-Scale has quit IRC06:01
*** X-Scale` is now known as X-Scale06:01
*** captain_morgan20 has joined #yosys06:06
*** X-Scale` has joined #yosys06:16
*** X-Scale has quit IRC06:17
*** X-Scale` is now known as X-Scale06:18
*** futarisIRCcloud has quit IRC07:41
*** az0re has joined #yosys08:08
*** dys has joined #yosys08:22
*** ZirconiumX has joined #yosys08:32
*** dys has quit IRC08:37
*** corecode_ has quit IRC08:42
*** vidbina_ has joined #yosys08:55
*** corecode has joined #yosys09:11
*** rohitksingh has joined #yosys09:14
*** az0re has quit IRC09:26
*** attie has joined #yosys09:46
*** vidbina_ has quit IRC09:49
*** attie has quit IRC10:10
*** fsasm has joined #yosys10:11
*** cr1901_modern has quit IRC11:01
*** cr1901_modern has joined #yosys11:05
*** vidbina_ has joined #yosys11:06
*** futarisIRCcloud has joined #yosys11:16
*** vidbina_ has quit IRC11:20
*** rohitksingh has quit IRC11:31
*** vidbina_ has joined #yosys12:08
*** attie has joined #yosys12:09
*** attie has quit IRC12:14
*** mjoldfield has quit IRC12:24
*** mjoldfield has joined #yosys12:26
Xiretzajust got to the point where my SoC successfully gets through ghdl and Yosys, but nextpnr-xilinx fails because I have too many global clocks. looking at them, there are 4 with cryptic names like '$auto$clkbufmap.cc:247:execute$237951' - any tips on tracing those back to their approximate context in the design?12:34
daveshahFind the BUFG in the netlist (perhaps do a write_ilang at the end as that should be more readable than the JSON) and have a look at the input name12:46
mwkXiretza: try synth_xilinx -noclkbuf12:54
mwkthis will disable the clkbufmap pass12:54
daveshahThen you will need to instantiate BUFGs manually as nextpnr-xilinx currently relies on Yosys to promote them12:56
mwkhmm right12:57
daveshahIn this case I think '$auto$clkbufmap.cc:247:execute$237951' are usually clock buffer inputs12:59
Xiretzaindeed12:59
daveshahsomewhat confusingly the current hacky clock router in nextpnr-xilinx also lists these as global clocks because it uses a similar algorithm to route them12:59
*** s_frit has quit IRC13:04
*** s_frit has joined #yosys13:05
*** gromero has quit IRC13:24
*** futarisIRCcloud has quit IRC13:46
*** develonepi3 has joined #yosys13:56
*** dys has joined #yosys14:07
*** vidbina_ has quit IRC14:29
*** vidbina_ has joined #yosys14:47
*** emeb has joined #yosys15:26
*** captain_morgan20 has quit IRC15:58
*** captain_morgan20 has joined #yosys16:03
*** dh73 has joined #yosys16:07
*** captain_morgan20 has quit IRC16:10
*** mjoldfield has quit IRC16:13
*** mjoldfield has joined #yosys16:13
*** captain_morgan20 has joined #yosys16:24
*** rohitksingh has joined #yosys17:01
*** vidbina_ has quit IRC17:04
*** vidbina_ has joined #yosys17:05
*** vidbina_ has quit IRC17:11
*** fsasm has quit IRC17:13
*** X-Scale` has joined #yosys17:32
*** rohitksingh has quit IRC17:33
*** X-Scale has quit IRC17:34
*** X-Scale` is now known as X-Scale17:34
*** citypw has quit IRC17:48
ZirconiumXSo, here's a question: how does one create an asynchronous load when you only have a synchronous load input?17:56
ZirconiumXThe Cyclone V LAB has 3 clock inputs, a synchronous-load line, a synchronous-clear line, and two asynchronous-clear lines18:02
*** attie has joined #yosys18:03
*** dys has quit IRC18:17
daveshah> So, here's a question: how does one create an asynchronous load when you only have a synchronous load input?18:17
daveshahI expect the same way you fake async set and reset on devices that only have one or the other18:18
daveshahmore or less two FFs one with an async set and one with an async reset; and a latch and mux to select the last one asserted18:18
sorearisn't "latch" the same thing as "asynchronous load" and thereby begging the question18:19
daveshahWell, a latch is easy enough to make out of a LUT18:19
daveshahI'm guessing asynchronous load was FF + async load18:20
*** attie has quit IRC18:56
*** rohitksingh has joined #yosys19:34
*** Jybz has joined #yosys19:42
*** rohitksingh has quit IRC19:49
*** rohitksingh has joined #yosys19:54
*** Jybz has quit IRC19:59
*** rohitksingh has quit IRC20:00
*** rohitksingh has joined #yosys20:11
*** dys has joined #yosys20:12
*** rohitksingh has quit IRC20:51
*** rohitksingh has joined #yosys22:26
Xiretzais there a pass in yosys that removes all $assert/$assume/$cover cells? VHDL doesn't have the same preprocessor mechanic as verilog, so turning off generation of verification blocks in the frontend isn't quite as straightforward.22:33
daveshahXiretza: chformal -remove22:35
Xiretzadaveshah: thanks, that's a useful command!22:36
*** rohitksingh has quit IRC22:50
*** ebb has quit IRC23:21
*** ebb has joined #yosys23:22
*** dys has quit IRC23:33
*** develonepi3 has quit IRC23:34
*** develonepi3 has joined #yosys23:39

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