| *** tpb has joined #vtr-dev | 00:00 | |
| *** digshadow has quit IRC | 00:43 | |
| *** digshadow has joined #vtr-dev | 01:01 | |
| mithro | kem_: Thanks -- will test out the pass_gate stuff again soon | 02:54 |
|---|---|---|
| *** digshadow has quit IRC | 05:04 | |
| *** digshadow has joined #vtr-dev | 05:47 | |
| mithro | daveshah: If you happen to be around, I have some questions about ice40 routing fabric and wanted to know if you thought this looked valid or not.... | 16:12 |
| mithro | daveshah: https://github.com/SymbiFlow/symbiflow-arch-defs/issues/131 | 16:12 |
| tpb | Title: Investigate weird routing in ice40 arch · Issue #131 · SymbiFlow/symbiflow-arch-defs · GitHub (at github.com) | 16:12 |
| daveshah | mithro: have limited time this evening to look at stuff | 16:15 |
| daveshah | But at a glance it seems dodgy to me | 16:15 |
| daveshah | What is being connected to on the second image? | 16:15 |
| daveshah | This could only happen using neighbour links, as far as I know, and there is no direct connection between them and the span12s | 16:17 |
| mithro | daveshah: Yeah seems a bit weird, I'm going to write another icebox tool because I just can't get the info I need from the current docs | 16:33 |
| daveshah | mithro: https://knielsen.github.io/ice40_viewer/ice40_viewer.html is quite nice because it shows the net number, but does not show all span <-> logic/local connectivity | 16:51 |
| tpb | Title: ICE40 layout viewer (at knielsen.github.io) | 16:51 |
| mithro | https://www.irccloud.com/pastebin/KRrE54lL/ | 17:07 |
| tpb | Title: Snippet | IRCCloud (at www.irccloud.com) | 17:07 |
| daveshah | mithro: neat! | 17:07 |
| mithro | daveshah: I just need to work out how to get the permanent connections now... | 17:08 |
| daveshah | mithro: look at https://github.com/daveshah1/icestorm/blob/tiny_ice40/icebox/icebox_chipdb.py#L195 and https://github.com/daveshah1/icestorm/blob/tiny_ice40/icebox/icebox_chipdb.py#L358 | 17:09 |
| tpb | Title: icestorm/icebox_chipdb.py at tiny_ice40 · daveshah1/icestorm · GitHub (at github.com) | 17:09 |
| daveshah | that's how they are output into the chipdb | 17:09 |
| daveshah | beware that they will vary depending on location | 17:09 |
| daveshah | due to the edges and corners of the device being different with IO etc | 17:10 |
| mithro | daveshah: I started with icebox_maps.py | 17:13 |
| benreynwar | If I run "make wire.patched.disp ARCH=testarch" in symbiflow-test-archs, I get an error "TypeError: Pin doesn't have port_name BLK_IG-IBUF(0)->None[None]". | 17:19 |
| benreynwar | Is there a branch of vpr I should be using that outputs port names in the rr_graph xml? | 17:20 |
| mithro | benreynwar: Yes - upstream master does | 17:20 |
| mithro | benreynwar: you might want to take a look at my 4mcmaster branch | 17:21 |
| mithro | benreynwar: I need to clean it up a bit | 17:21 |
| benreynwar | mithro: Will do. Thanks. | 17:21 |
| mithro | benreynwar: But I did a major refactor of the tests | 17:22 |
| mithro | benreynwar: I have a weekend project for you if you want to learn more about building rr_graphs :-P | 17:22 |
| benreynwar | You're welcome to point me in a specific direction, but I'm sneaking in bits of work between entertaining kids so I can't guarantee much progress :p. | 17:24 |
| mithro | benreynwar: I need a tool which checks that you can get from every SOURCE to every SINK | 17:40 |
| mithro | benreynwar: It should be a pretty easy graph traversal problem... | 17:44 |
| mithro | daveshah: Do you know if there is an easy way to map between pin numbers and io_tile coordinates? | 17:45 |
| daveshah | mithro: No, you have to use the mapping in icebox | 17:45 |
| mithro | daveshah: btw - what are .sym values in an .asc file? | 17:46 |
| mithro | daveshah: They seem to be names for things - but I don't get the numbers? | 17:46 |
| daveshah | They map net names to icebox net numbers | 17:46 |
| daveshah | For diagnostics in icetime etc | 17:47 |
| mithro | daveshah: In what structure do I look up the numbers? | 17:48 |
| daveshah | mithro: they are the net numbers in the chipdb. Don't know more than that and not at computer so can't give any further info | 17:49 |
| daveshah | Have a look at how they end up in the chipdb | 17:49 |
| mithro | daveshah: Thanks, the icebox_chipdb is what I needed to look at | 17:51 |
| benreynwar | mithro: Sounds like a nice little project for me. | 18:33 |
| benreynwar | mithro: Implemented it with method on the RoutingGraph object. No test yet, so likely buggy. | 21:13 |
| benreynwar | mithro: https://github.com/SymbiFlow/symbiflow-arch-defs/pull/132 | 21:13 |
| tpb | Title: Add methods to RoutingGraph to determine whether any sink nodes are i… by benreynwar · Pull Request #132 · SymbiFlow/symbiflow-arch-defs · GitHub (at github.com) | 21:13 |
| mithro | benreynwar: I'd love to see some tests :-P | 21:15 |
| mithro | benreynwar: I'm not smart enough to write this type of function without writing tests :-P | 21:18 |
| benreynwar | Yep. Hence the comment not to merge it yet :). | 21:20 |
| mithro | benreynwar: If you put a WIP in the title of the pull request then the WIP check will prevent me accidently merging it :-P | 21:21 |
| mithro | be back in an 30 minutes, going to find some caffeine and go for a walk... | 21:23 |
| benreynwar | I'm disassembling and recovering a couch now, but will probably get on my laptop again later today. | 21:27 |
| mithro | I need to go buy a couch... | 21:58 |
| mithro | kem_: Hopefully you are not around, but on the of chance you are -- have you used IO pad placement recently? | 22:09 |
| mithro | https://www.irccloud.com/pastebin/10URhJWx/ | 22:15 |
| tpb | Title: Snippet | IRCCloud (at www.irccloud.com) | 22:15 |
| mithro | Hrm... | 22:22 |
| mithro | https://www.irccloud.com/pastebin/xhU4U54b/ | 22:22 |
| tpb | Title: Snippet | IRCCloud (at www.irccloud.com) | 22:22 |
| mithro | daveshah / jhol: Just place-and-routed the iceblink example with real pin locs! | 22:26 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!