corecode | aha | 00:00 |
---|---|---|
*** tpb has joined #yosys | 00:00 | |
emeb | how well does that compare to what icecube estimates? | 00:01 |
corecode | icecube says 51MHz | 00:01 |
emeb | hmm... that's a fair difference for the same design. | 00:03 |
corecode | yes | 00:03 |
emeb | timing analysis has always seemed like it involves a lot of "art" and some luck though. | 00:04 |
*** danieljabailey has quit IRC | 00:07 | |
sorear | does icetime do multiclock? | 00:11 |
*** acksys_ has quit IRC | 00:12 | |
corecode | i doubt, how would it know | 00:17 |
*** ZipCPU|Laptop has joined #yosys | 00:59 | |
*** ZipCPU|Laptop has quit IRC | 01:19 | |
*** ZipCPU|Laptop has joined #yosys | 01:23 | |
*** ZipCPU|Laptop has quit IRC | 01:27 | |
lutsabound | sorear: nextpnr supports multiple clock timing. There's no longer any need for icetime | 01:30 |
sorear | lutsabound: mm, can you instruct corecode on how to use nextpnr to compute a timing analysis for a bitstream generated by iCEcube? I was under the impression that was not possible with nextpnr | 01:32 |
lutsabound | I'd have to go back to the nextpnr docs myself | 01:33 |
lutsabound | I know its in there, I've seen the capability, but havent done multiple clocks yet myself | 01:34 |
lutsabound | daveshah knows more about it than I do | 01:34 |
sorear | specifically I was under the impression icetime could ingest an externally-generated bitstream, but nextpnr was limited to timing bitstreams it itself generated | 01:35 |
lutsabound | Hmm... that may be the case | 01:36 |
*** _whitelogger has quit IRC | 01:46 | |
*** _whitelogger has joined #yosys | 01:48 | |
*** emeb has quit IRC | 01:50 | |
*** emeb_mac has joined #yosys | 01:52 | |
*** gsi__ has joined #yosys | 01:52 | |
*** gsi_ has quit IRC | 01:55 | |
*** citypw has joined #yosys | 03:22 | |
*** rohitksingh_work has joined #yosys | 03:39 | |
*** citypw has quit IRC | 03:39 | |
*** leviathanch has joined #yosys | 03:43 | |
*** citypw has joined #yosys | 03:45 | |
*** citypw has quit IRC | 03:48 | |
*** citypw has joined #yosys | 03:49 | |
*** lutsabound has quit IRC | 04:04 | |
*** danieljabailey has joined #yosys | 04:08 | |
*** pie___ has joined #yosys | 05:02 | |
*** pie___ has joined #yosys | 05:03 | |
*** pie__ has quit IRC | 05:06 | |
*** Cerpin has quit IRC | 05:16 | |
*** citypw has quit IRC | 05:19 | |
*** citypw has joined #yosys | 05:28 | |
*** Cerpin has joined #yosys | 05:33 | |
*** chaseemory has quit IRC | 05:50 | |
*** leviathanch has quit IRC | 05:58 | |
*** _whitelogger has quit IRC | 06:16 | |
*** _whitelogger has joined #yosys | 06:18 | |
*** emeb_mac has quit IRC | 07:18 | |
*** indy has quit IRC | 07:41 | |
*** leviathanch has joined #yosys | 07:47 | |
*** indy has joined #yosys | 07:49 | |
*** dys has quit IRC | 08:19 | |
*** m4ssi has joined #yosys | 08:23 | |
*** citypw has quit IRC | 10:06 | |
daveshah | sorear: yes, icetime can read bitstreams which is great for calibration against the vendor tools | 10:25 |
daveshah | micko started work on ingesting bitstreams into nextpnr, but I don't think that's really working yet | 10:26 |
corecode | icetime says 34MHz, icecube says 51MHz | 10:30 |
corecode | so i think icetime is a bit conservative | 10:30 |
daveshah | interesting | 10:30 |
daveshah | can you provide the asc? | 10:31 |
corecode | yes | 10:31 |
daveshah | and the icecube report? | 10:31 |
*** ec0 has quit IRC | 10:31 | |
corecode | should i run using icecube.sh? | 10:32 |
daveshah | No need, GUI is fine | 10:32 |
corecode | https://gist.github.com/corecode/7dd007a1c343b6666118b488f005a3e9 | 10:34 |
tpb | Title: gist:7dd007a1c343b6666118b488f005a3e9 · GitHub (at gist.github.com) | 10:34 |
corecode | https://gist.github.com/b700b546c64e1f0e051a8c3aba2d7caa | 10:35 |
tpb | Title: top_u4k_bitmap.asc · GitHub (at gist.github.com) | 10:35 |
daveshah | do you have the critical path report from icetime too? | 10:36 |
corecode | no | 10:37 |
daveshah | ah, it's in the middle, i see | 10:37 |
corecode | icecube yes | 10:37 |
daveshah | that looks like 51MHz before routing | 10:41 |
corecode | yes | 10:41 |
daveshah | whereas icetime runs on the routed bitstream, so the two aren't directly comparable | 10:42 |
daveshah | icecube should be able to run timing analysis after routing too | 10:42 |
corecode | hm | 10:42 |
corecode | aaah | 10:43 |
corecode | so why doesn't it print the estimated time after routing | 10:43 |
daveshah | no idea | 10:43 |
corecode | there is a report that says 39MHz | 10:44 |
daveshah | that sounds more like it | 10:44 |
corecode | yea | 10:44 |
corecode | thanks | 10:45 |
corecode | i guess i need to change my spi synchronizer to work at higher speeds | 10:45 |
*** ec0 has joined #yosys | 11:05 | |
*** lutsabound has joined #yosys | 12:31 | |
*** rohitksingh_work has quit IRC | 12:39 | |
*** _whitelogger has quit IRC | 12:46 | |
*** _whitelogger has joined #yosys | 12:48 | |
*** josi has quit IRC | 12:49 | |
*** proteusguy has joined #yosys | 13:32 | |
*** celadon_ has joined #yosys | 13:33 | |
*** rohitksingh has joined #yosys | 14:05 | |
*** leviathanch has quit IRC | 14:50 | |
*** lutsabound has quit IRC | 14:50 | |
*** leviathanch has joined #yosys | 14:53 | |
*** emeb has joined #yosys | 15:07 | |
*** maikmerten has joined #yosys | 15:29 | |
*** danieljabailey has quit IRC | 15:44 | |
*** seldridge has joined #yosys | 16:09 | |
*** leviathanch has quit IRC | 17:06 | |
*** m4ssi has quit IRC | 17:08 | |
*** rohitksingh has quit IRC | 18:36 | |
*** lutsabound has joined #yosys | 19:11 | |
*** m4ssi has joined #yosys | 19:26 | |
*** m4ssi has quit IRC | 20:23 | |
*** lutsabound has quit IRC | 21:25 | |
*** maikmerten has quit IRC | 21:44 | |
*** seldridge has quit IRC | 22:20 | |
*** seldridge has joined #yosys | 22:37 | |
*** vzcx has joined #yosys | 22:41 | |
*** vzcx has quit IRC | 23:11 | |
*** Nazara has quit IRC | 23:20 | |
*** seldridge has quit IRC | 23:45 | |
*** seldridge has joined #yosys | 23:49 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!