| *** tpb has joined #vtr-dev | 00:00 | |
| *** tpb` has joined #vtr-dev | 00:10 | |
| *** tpb has quit IRC | 00:10 | |
| *** tpb` is now known as tpb | 00:10 | |
| *** digshadow has joined #vtr-dev | 04:32 | |
| *** digshadow has quit IRC | 04:47 | |
| *** digshadow has joined #vtr-dev | 05:15 | |
| daveshah | mithro: yes, although I think in a larger design the packing issues turn out to be less significant | 06:29 |
|---|---|---|
| daveshah | But it's something to experiment with once we have a more repeatable PnR flow than arachne-pnr | 06:30 |
| mithro | daveshah: morning! | 06:30 |
| mithro | daveshah: I'm trying to figure out carry chains / packing stuff | 06:30 |
| daveshah | mithro: yep. have you tested the basic routing yet though (with icebox_vlog)? probably best before moving on imo | 06:43 |
| mithro | daveshah: I was trying to figure out how the internal block routing stuff should work | 06:43 |
| daveshah | mithro: sure, but you don't need carry chains - or flipflops at all - to start doing some basic tests | 06:43 |
| mithro | daveshah: take a look at the 4mcmaster branches of the arch-defs and vpr | 06:47 |
| mithro | daveshah: I could use some help adding a "make check" or similar type rule to look at the HLC output... | 06:50 |
| mithro | daveshah: https://github.com/mithro/symbiflow-arch-defs/blob/4mcmaster/tests/make/test-common.mk#L204-L210 | 06:50 |
| tpb | Title: symbiflow-arch-defs/test-common.mk at 4mcmaster · mithro/symbiflow-arch-defs · GitHub (at github.com) | 06:50 |
| daveshah | mithro: yes, that shouldn't be too hard to implement | 06:51 |
| daveshah | just run icebox_vlog and an equivalence check against the original source | 06:51 |
| mithro | make clean && (mkdir build-ice40-top-routing-virt-HX1K; cd build-ice40-top-routing-virt-HX1K; ln -s ../../../build/ice40-top-routing-virt-HX1K/ice_node_id.csv .) && make route.echo | 06:51 |
| mithro | daveshah: and rm -f build-ice40-top-routing-virt-HX1K/example.route && make route.echo | 06:52 |
| mithro | I'm thinking of heading to bed soon | 06:52 |
| daveshah | mithro: I need to sort out plans for today, but I may get a chance to look at it | 06:53 |
| *** digshadow has quit IRC | 17:32 | |
| mithro | morning | 17:54 |
| *** digshadow has joined #vtr-dev | 18:08 | |
| mithro | digshadow: So it looks like there are invalid routes in the rr_graph you are generating... | 21:11 |
| mithro | Parse error in line 2108: | 21:12 |
| mithro | span12_y4_g14_0 -> span4_y4_g11_7 <-> span4_x7_g4_0 | 21:12 |
| mithro | No bit pattern for ['buffer', 'sp12_h_r_11', 'sp4_h_r_7'] in LogicTile(1k, 7, 4) | 21:12 |
| mithro | digshadow: Or I'm screwing up the translation of coordinates somewhere... | 21:12 |
| *** digshadow has left #vtr-dev | 21:14 | |
| *** digshadow has joined #vtr-dev | 21:14 | |
| digshadow | mithro: did you get feedback from jhol about why the coordinates were inverted in the one case | 21:15 |
| mithro | digshadow: https://www.irccloud.com/pastebin/NauaLYwd/ | 21:15 |
| tpb | Title: Snippet | IRCCloud (at www.irccloud.com) | 21:15 |
| mithro | digshadow: It was a mistake | 21:16 |
| mithro | digshadow: any thoughts on how to debug that issue? | 21:26 |
| mithro | benreynwar: Looks like check_graph takes quite a long time on a ice40 1K graph | 21:53 |
| mithro | benreynwar: Give http://hopper.mithis.com/~tim/rr_graph.real.xml a try when you get a chance | 21:55 |
| benreynwar | mithro: I'll take a look tonight. | 21:56 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!