| *** tpb has joined #vtr-dev | 00:00 | |
| *** digshadow has quit IRC | 02:40 | |
| *** digshadow has joined #vtr-dev | 02:58 | |
| mithro | digshadow: FYI - I added a carry chain to the testarch for you | 03:59 |
|---|---|---|
| mithro | https://github.com/mithro/symbiflow-arch-defs/commit/c0cb9f8404eeb88e99469fef384607adc73af063 | 04:00 |
| tpb | Title: testarch: Adding a carry chain to lutff. · mithro/symbiflow-arch-defs@c0cb9f8 · GitHub (at github.com) | 04:00 |
| digshadow | mithro: thanks! | 04:02 |
| *** q3k has quit IRC | 06:24 | |
| *** q3k has joined #vtr-dev | 06:26 | |
| digshadow | mithro: re: ticket comment, I think the idea is that he is talking about a from scratch rr_graph vs us reparsing one | 16:51 |
| digshadow | but even then I'm not sure if its an issue, eh I'll look at his comments more soon | 16:51 |
| *** digshadow has quit IRC | 17:47 | |
| *** digshadow has joined #vtr-dev | 18:10 | |
| jhol | digshadow, mithro: https://github.com/SymbiFlow/symbiflow-arch-defs/pull/89 | 18:11 |
| tpb | Title: Various fixes to the PLB by jhol · Pull Request #89 · SymbiFlow/symbiflow-arch-defs · GitHub (at github.com) | 18:11 |
| jhol | so the PLB patches are in reasonable shape now | 18:11 |
| jhol | I had to make some more changes to the structure of the PLB | 18:11 |
| jhol | it turns out that VPR doesn't like the CARRYCHAIN <pack_pattern> crossing through mode tags | 18:12 |
| jhol | so I pulled the LUT+CARRY out of the modal DFF elements | 18:12 |
| jhol | the packer seems to do the right thing with a chained LUT4 - see the big_xor.v test | 18:13 |
| jhol | it also seems to do the right thing with a LUT->FF | 18:13 |
| jhol | testing counter.v it seems to fall over on routing SB_CARRY.O -> a LUT pin | 18:14 |
| jhol | there is a direct route with SB_CARRY.O -> LUT_CARRY.FCOUT -> LUT_CARRY.FCIN -> SB_LUT.I[2], but unless yosys selected I[2], VPR doesn't seem able to rearrange the LUT | 18:15 |
| jhol | the alternative route is via local tracks, but these are rather broken in the tile-routing PLB - lots of unconnected track warnings | 18:16 |
| jhol | so when we get the rr_graph in, this problem should go away | 18:16 |
| jhol | also it goes away if you comment out the carry_out+carry_in fc_override i.e. add some wires from the carry wire to the global routing | 18:17 |
| jhol | VERY VERY IMPORTANT: please can you accept these change into mainline, AND/OR rebase your branches on top of these patches | 18:18 |
| jhol | I spent all of this Monday fixing the previous divergence - and I don't want that to happen again | 18:19 |
| digshadow | jhol: okay will make it a priority to review right now | 18:19 |
| digshadow | out of curiosity, what diverged | 18:20 |
| jhol | this time though, patches are not a work-in-progress mess, so it's a reasonable ask for you to rebase on it | 18:20 |
| jhol | digshadow: last time both mithro and I made major conflicting changes to the PLB XML | 18:20 |
| digshadow | jhol: says DCO is missing | 18:20 |
| jhol | hopefully there won't be such problems this time | 18:21 |
| digshadow | can you fix that while I review? | 18:21 |
| jhol | what's DCO? | 18:21 |
| jhol | ok ok - I'll fix that | 18:22 |
| digshadow | did you find info? | 18:22 |
| digshadow | looks like mithro didn't sign his either | 18:22 |
| jhol | yeah | 18:22 |
| digshadow | or at least all of htem | 18:22 |
| digshadow | oh wait those might just be merges | 18:23 |
| jhol | hmm | 18:23 |
| digshadow | no there are unsigned | 18:23 |
| jhol | maybe time to use git filter-branch | 18:23 |
| digshadow | jhol: maybe start by signing yours and we can get mithro to fix his | 18:23 |
| digshadow | anyway it can't get merged into mainline until that is fixed for both you and mithro | 18:23 |
| jhol | well I'm not so bothered about mainline, so long as it gets rebased off of for the coming the days | 18:27 |
| digshadow | jhol: also looks like travis is failing | 18:29 |
| digshadow | did you look into that? | 18:29 |
| digshadow | jhol: I'd like to get the signoff in there at least if I rebase | 18:29 |
| digshadow | although it sounds like you are conflicting more with mithro than me anyway | 18:30 |
| mithro | jhol: Yes, lets get your and my changes to the ice40 merged | 18:30 |
| jhol | ok... I'm rather out of time for today, so I can work on the DCO, and any other problems on Moday | 18:34 |
| jhol | for now, like I say it's more about divergence | 18:34 |
| jhol | mithro: this is annoying | 18:36 |
| mithro | jhol: Which is annoying? | 18:36 |
| jhol | the solution to my SB_CARRY -> SB_LUT4 routing issue might be related to the "class" of the SB_LUT4 pb_type | 18:36 |
| jhol | https://github.com/jhol/symbiflow-arch-defs/blob/modes-fixes3/ice40/primitives/sb_lut/sb_lut.pb_type.xml | 18:36 |
| tpb | Title: symbiflow-arch-defs/sb_lut.pb_type.xml at modes-fixes3 · jhol/symbiflow-arch-defs · GitHub (at github.com) | 18:37 |
| jhol | if SB_LUT4 could get class="lut", then presumably VPR would be able rearrange the LUT so it could be routed via FCOUT+FCIN | 18:37 |
| mithro | jhol: You need yosys to generate generic LUTs not SB_LUT4 | 18:38 |
| jhol | oh ok - ignore me, it's already doing that :( | 18:39 |
| jhol | so there's some other reason for why they're not getting wired up properly | 18:39 |
| jhol | well all the local tracks are messed up, so that's port of it | 18:39 |
| mithro | https://github.com/YosysHQ/yosys/blob/81a457c4a68937f8edb4c48ca5a5de86b5c05769/techlibs/ice40/synth_ice40.cc#L241-L246 | 18:39 |
| tpb | Title: yosys/synth_ice40.cc at 81a457c4a68937f8edb4c48ca5a5de86b5c05769 · YosysHQ/yosys · GitHub (at github.com) | 18:40 |
| jhol | but like I say, it should be possible to wire it with out locals, just using FCOUT/IN | 18:40 |
| mithro | jhol: Can you give me a better description of what you are seeing and what you are expecting? | 18:40 |
| jhol | yeah my eblif doesn't have any, so I don't know | 18:40 |
| jhol | just quick explanation | 18:40 |
| jhol | I've got to go... | 18:40 |
| jhol | basically if you run VPR on counter.v, it will fail to route - claiming it can't wire the SB_CARRY to a LUT | 18:41 |
| jhol | this can be "fixed" by commenting the carry_in/out fc_override | 18:41 |
| jhol | and presumably it will be fixed by the rr_graph | 18:41 |
| jhol | but really neither should be necessary | 18:42 |
| jhol | anyway got to go -- I'll check back later | 18:42 |
| mithro | jhol: So, go to tests directory and type "make ARCH=ice40 counter.echo" ? | 18:42 |
| jhol | yes | 18:42 |
| jhol | make ARCH=ice40 DEVICE_TYPE=tile-routing-virt DEVICE=test4 VPR_ARGS='--disp on' counter.gdb | 18:42 |
| mithro | jhol: Great, I'll see if I can get time to see what is going on | 18:43 |
| digshadow | mithro: do you need any help reviewing the patch | 19:44 |
| mithro | digshadow: jhol's patch? | 20:43 |
| mithro | digshadow: No | 20:43 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!