| *** tpb has joined #symbiflow | 00:00 | |
| *** _whitelogger has quit IRC | 01:21 | |
| *** _whitelogger has joined #symbiflow | 01:23 | |
| *** clacktronics has quit IRC | 01:23 | |
| *** citypw has joined #symbiflow | 03:59 | |
| *** Bertl_zZ is now known as Bertl_oO | 05:33 | |
| *** clacktronics has joined #symbiflow | 08:33 | |
| *** OmniMancer has joined #symbiflow | 10:42 | |
| *** craigo has quit IRC | 11:08 | |
| *** Bertl_oO is now known as Bertl | 12:13 | |
| OmniMancer | How would one find out how to help out with Project Tang? | 14:23 |
|---|---|---|
| *** somlo has quit IRC | 14:57 | |
| *** somlo has joined #symbiflow | 15:12 | |
| hackerfoo | OmniMancer: You could talk to the people working on it. | 15:19 |
| hackerfoo | OmniMancer: If you're talking about the Anlogic FPGA in the Lichee Tang, I don't know of such a project. | 15:58 |
| OmniMancer | I have found this: https://github.com/mmicko/prjtang | 16:13 |
| tpb | Title: GitHub - mmicko/prjtang: Documenting the Anlogic FPGA bit-stream format. (at github.com) | 16:13 |
| OmniMancer | I will see if I can get in tough with the author then | 16:13 |
| hackerfoo | OmniMancer: Thanks. | 16:14 |
| hackerfoo | Many of the projects are by different people or groups, for example prjmistral (Cyclone V.) | 16:15 |
| OmniMancer | Indeed | 16:17 |
| *** citypw has quit IRC | 16:20 | |
| *** OmniMancer has quit IRC | 16:26 | |
| litghost | acomodi: Any results on which murax/picosoc nets are taking so long in the ROI-less graph? | 16:56 |
| litghost | The results from my run last night were showing that 2 or 3 nets were massively misbehaving, so I think the hypothesis that certain nets are taking much more time may be borne out | 16:57 |
| sf-slack | <acomodi> @litghost I have not been able to get any meaningful results. I am currently running another build with the latest changes in arch-defs. It should end in a couple of hours. Did the net you found have to do with clocking trees? | 17:10 |
| litghost | acomodi: I haven't found the net specifically, only that I saw the router take a long time (e.g. ~1 hr) to reroute 3 nets | 17:11 |
| litghost | acomodi: That seems super excessive, and likely points to something going wrong | 17:11 |
| litghost | acomodi: I haven't investigated any further than that, hence why I was asking if you had found anything yet | 17:12 |
| *** adjtm has quit IRC | 18:31 | |
| mithro | https://twitter.com/fpga_dave/status/1187069745046142977 | 19:43 |
| *** adjtm has joined #symbiflow | 19:57 | |
| *** craigo has joined #symbiflow | 20:07 | |
| *** Bertl is now known as Bertl_zZ | 20:53 | |
| *** mdv29 has joined #symbiflow | 21:01 | |
| *** kuldeep has quit IRC | 21:22 | |
| sf-slack | <max> Does anyone have any experience with post routing simulation? I am doing work with asynchronous circuits as a research project, and having good delay estimates in simulation would be a huge help but I am having trouble finding anything in nextpnr or project trellis (using an ecp5). | 21:26 |
| daveshah | max: I'm working on it, see https://github.com/YosysHQ/nextpnr/pull/345 | 21:27 |
| tpb | Title: Improve handling of top level IO and add SDF support by daveshah1 · Pull Request #345 · YosysHQ/nextpnr · GitHub (at github.com) | 21:27 |
| daveshah | I haven't tried it for ECP5 yet, some sim model changes will be needed but the nextpnr side should work for ECP5 too | 21:27 |
| sf-slack | <max> thats awesome! I'll have a look | 21:28 |
| sf-slack | <max> Ah, I was under the impression that icarus could reasonably handle sdf files. Not having an open source option may be an issue. | 21:30 |
| daveshah | Someone just pointed out cvc as a possible option | 21:32 |
| daveshah | I haven't tried it myself yet | 21:32 |
| sf-slack | <max> I might give that a shot. Im using cocotb to generate tests just because it saves some time, but I reckon I will only need some basic test benches for post routing simulations. | 21:36 |
| *** mdv29 has quit IRC | 21:39 | |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!