| *** tpb has joined #symbiflow | 00:00 | |
| *** futarisIRCcloud has quit IRC | 00:07 | |
| *** TMM has quit IRC | 00:20 | |
| *** TMM has joined #symbiflow | 00:22 | |
| *** maartenBE has quit IRC | 00:22 | |
| *** maartenBE has joined #symbiflow | 00:23 | |
| *** krogozinski_ has joined #symbiflow | 00:40 | |
| *** krogozinski has quit IRC | 00:41 | |
| *** sorear has quit IRC | 00:48 | |
| *** davidlattimore has quit IRC | 00:51 | |
| *** sorear has joined #symbiflow | 00:52 | |
| *** jopdorp has quit IRC | 00:52 | |
| *** davidlattimore has joined #symbiflow | 00:53 | |
| *** jopdorp has joined #symbiflow | 00:53 | |
| *** Niklas[m] has quit IRC | 01:11 | |
| *** gruetzkopf has quit IRC | 01:11 | |
| *** flokli has quit IRC | 01:11 | |
| *** Niklas[m] has joined #symbiflow | 01:12 | |
| *** gruetzkopf has joined #symbiflow | 01:12 | |
| *** flokli has joined #symbiflow | 01:12 | |
| *** unrznbl[m] has quit IRC | 01:14 | |
| *** Niklas[m] has quit IRC | 01:14 | |
| *** abeljj[m] has quit IRC | 01:14 | |
| *** xobs has quit IRC | 01:14 | |
| *** promach3 has quit IRC | 01:15 | |
| *** unrznbl[m] has joined #symbiflow | 01:35 | |
| *** futarisIRCcloud has joined #symbiflow | 01:37 | |
| *** xobs has joined #symbiflow | 01:45 | |
| *** Niklas[m] has joined #symbiflow | 01:48 | |
| *** abeljj[m] has joined #symbiflow | 01:50 | |
| *** citypw_ has joined #symbiflow | 01:54 | |
| *** Niklas[m] has quit IRC | 02:03 | |
| *** unrznbl[m] has quit IRC | 02:03 | |
| *** xobs has quit IRC | 02:03 | |
| *** abeljj[m] has quit IRC | 02:04 | |
| *** citypw_ has quit IRC | 02:12 | |
| *** hansfbaier has joined #symbiflow | 02:26 | |
| *** xobs has joined #symbiflow | 02:32 | |
| *** promach3 has joined #symbiflow | 02:33 | |
| *** Niklas[m] has joined #symbiflow | 02:36 | |
| *** bjorkintosh has joined #symbiflow | 02:37 | |
| *** unrznbl[m] has joined #symbiflow | 02:42 | |
| *** abeljj[m] has joined #symbiflow | 03:04 | |
| *** cr1901_modern1 has joined #symbiflow | 03:29 | |
| *** HackerFoo1 has joined #symbiflow | 03:31 | |
| *** TMM has quit IRC | 03:38 | |
| *** rvalles has quit IRC | 03:38 | |
| *** HackerFoo has quit IRC | 03:38 | |
| *** cr1901_modern has quit IRC | 03:38 | |
| *** Lofty has quit IRC | 03:38 | |
| *** ssb has quit IRC | 03:38 | |
| *** cr1901_modern1 has quit IRC | 03:39 | |
| *** cr1901_modern has joined #symbiflow | 03:39 | |
| *** Niklas[m] has quit IRC | 03:39 | |
| *** unrznbl[m] has quit IRC | 03:40 | |
| *** Degi has quit IRC | 03:42 | |
| *** Degi has joined #symbiflow | 03:42 | |
| *** hosana has quit IRC | 03:42 | |
| *** TMM has joined #symbiflow | 03:43 | |
| *** Lofty has joined #symbiflow | 03:43 | |
| *** ssb has joined #symbiflow | 03:43 | |
| *** TMM has quit IRC | 03:43 | |
| *** Niklas[m] has joined #symbiflow | 03:43 | |
| *** TMM has joined #symbiflow | 03:43 | |
| *** hosana has joined #symbiflow | 03:45 | |
| *** rvalles has joined #symbiflow | 03:45 | |
| *** unrznbl[m] has joined #symbiflow | 03:49 | |
| sf-slack | <jgoeders> @litghost Thanks. "make install" with capnproto-java did the trick. Now I'm running into: ```AttributeError: capnp/schema.c++:493: failed: struct has no such member; name = name``` I'm guessing my RapidWright version is somehow out of sync with something. I'll check into it later. Also, I tried setting up fasm2bels on my office computer today, but ran into the same `Could not find a version that satisfies the | 03:59 |
|---|---|---|
| sf-slack | requirement rapidyaml (from python-fpga-interchange)` issue mentioned above. Is there a fix for that? Thanks! | 03:59 |
| litghost | pip install -r symbiflow-xc-fasm2bels/requirements.txt | 04:04 |
| sf-slack | <jgoeders> OK, yeah, that's what I thought I was doing. I'll double check next time I'm in the office. | 04:13 |
| *** tpb has joined #symbiflow | 05:45 | |
| *** hansfbaier has quit IRC | 05:53 | |
| *** HackerFoo1 has quit IRC | 06:58 | |
| *** m_hackerfoo has quit IRC | 06:59 | |
| nickoe | litghost: the ci was not completely happy, it seems https://github.com/SymbiFlow/symbiflow-arch-defs/pull/1987 | 07:02 |
| *** m_hackerfoo has joined #symbiflow | 07:03 | |
| nickoe | For examples it errors a bit about RapidWright and acapnp file not being found. https://source.cloud.google.com/results/invocations/133de234-2c0a-4092-8d1e-471081bdafbb/targets/foss-fpga-tools%2Fsymbiflow-arch-defs%2Fpresubmit%2Fxc7a200t_vendor/log | 07:04 |
| *** HackerFoo1 has joined #symbiflow | 07:10 | |
| *** hansfbaier has joined #symbiflow | 08:14 | |
| *** citypw has joined #symbiflow | 08:43 | |
| *** cr1901_modern has quit IRC | 08:58 | |
| *** cr1901_modern has joined #symbiflow | 09:00 | |
| *** smkz has quit IRC | 09:26 | |
| *** smkz has joined #symbiflow | 09:27 | |
| *** craigo has joined #symbiflow | 10:09 | |
| *** hansfbaier1 has joined #symbiflow | 10:26 | |
| *** hansfbaier has quit IRC | 10:27 | |
| lambda | symbiflow_pack fails with "Failed to find matching architecture model for 'PLLE2_BASE'", is that just not supported and I have to use PLLE2_ADV instead? | 10:51 |
| sf-slack | <kgugala> @acomodi @mkurc^^ | 10:52 |
| lambda | hm, it's in the techmap, guess yosys just didn't pick up on it for whatever reason? | 10:52 |
| sf-slack | <acomodi> At the moment the architecture does not include the PLLE2_BASE. I think it still needs to be added. PLLE2_ADV is supported though | 10:53 |
| sf-slack | <mkurc> PLLE2_ADV works for sure. If PLLE2_BASE didn't work then probably the techmap in SymbiFlow is incorrect. It should map BASE to ADV in the same way as Vivado does. | 10:55 |
| lambda | alright, I just expected yosys to a PLLE2_ADV and subsequently a PLLE2_ADV_VPR, that's what cells_map.v says | 10:55 |
| lambda | *to map it to a | 10:55 |
| lambda | ah yeah, there are mismatches in the PLLE2_BASE I/Os between yosys' cells_xtra.v and arch-defs' cells_map.v | 11:01 |
| sf-slack | <mkurc> @lambda Could you file an issue about that on SymbiFlow/symbiflow-arch-defs Github repo so that we won't forget about the mismatch? | 11:03 |
| lambda | will do | 11:03 |
| sf-slack | <mkurc> thanks | 11:04 |
| *** hansfbaier1 has quit IRC | 11:18 | |
| -_whitenotifier- [symbiflow-arch-defs] Xiretza opened issue #1991: PLLE2_BASE techmap doesn't match yosys' cells - https://git.io/Jt83Q | 11:30 | |
| lambda | I'm currently trying to get my design to build using only the symbiflow tools (used nextpnr-xilinx previously), so I may have a few more questions and issues in the next few days ;) | 11:32 |
| lambda | like just now where symbiflow_route fails to route a clock from a PLL to... somewhere (`BLK-TL-CMT_TOP_L_UPPER_T[0].CMT_TOP_R_UPPER_T_PLLE2_CLKOUT3[0]` to `BLK-TL-CLBLL_L[0].CLBLL_L_CLK[0]`) - are there any known issues with clock routing? | 11:37 |
| sf-slack | <acomodi> This might has to do with the fact that global clock buffers auto-inference is disabled. There needs to be an explicit BUFG connected to the clk output of the PLL instantiated at the moment | 11:40 |
| lambda | ah, that makes, sense thanks | 11:42 |
| nickoe | litghost: Anything we can do to satisfy the ci? | 12:50 |
| *** FFY00 has quit IRC | 13:00 | |
| *** lopsided98 has quit IRC | 13:00 | |
| *** xobs has quit IRC | 13:00 | |
| *** sorear has quit IRC | 13:00 | |
| *** yeti has quit IRC | 13:00 | |
| *** FFY00 has joined #symbiflow | 13:01 | |
| *** lopsided98 has joined #symbiflow | 13:01 | |
| *** lopsided98 has quit IRC | 13:01 | |
| *** xobs has joined #symbiflow | 13:01 | |
| *** sorear has joined #symbiflow | 13:01 | |
| *** yeti has joined #symbiflow | 13:01 | |
| *** lopsided98 has joined #symbiflow | 13:01 | |
| *** futarisIRCcloud has quit IRC | 13:03 | |
| *** LoneTech has quit IRC | 13:03 | |
| *** futarisIRCcloud has joined #symbiflow | 13:03 | |
| *** LoneTech has joined #symbiflow | 13:03 | |
| *** abeljj[m] has quit IRC | 13:04 | |
| *** futarisIRCcloud has quit IRC | 13:04 | |
| *** Niklas[m] has quit IRC | 13:04 | |
| *** futarisIRCcloud has joined #symbiflow | 13:04 | |
| *** xobs has quit IRC | 13:04 | |
| *** infinite_recursi has joined #symbiflow | 13:05 | |
| *** promach3 has quit IRC | 13:07 | |
| *** lopsided98 has quit IRC | 13:08 | |
| *** unrznbl[m] has quit IRC | 13:08 | |
| *** lopsided98 has joined #symbiflow | 13:09 | |
| *** FFY00 has quit IRC | 13:22 | |
| *** abeljj[m] has joined #symbiflow | 13:31 | |
| *** Niklas[m] has joined #symbiflow | 13:32 | |
| *** unrznbl[m] has joined #symbiflow | 13:37 | |
| *** xobs has joined #symbiflow | 13:39 | |
| *** abeljj[m] has quit IRC | 13:48 | |
| *** xobs has quit IRC | 13:48 | |
| *** unrznbl[m] has quit IRC | 13:48 | |
| *** Niklas[m] has quit IRC | 13:49 | |
| *** FFY00 has joined #symbiflow | 14:05 | |
| *** promach3 has joined #symbiflow | 14:14 | |
| lambda | acomodi: it builds now with BUFGs on all the PLL outputs, just doesn't work - looking forward to some incredibly fun debugging | 14:27 |
| *** infinite_recursi has quit IRC | 14:29 | |
| *** abeljj[m] has joined #symbiflow | 14:40 | |
| *** unrznbl[m] has joined #symbiflow | 14:40 | |
| *** xobs has joined #symbiflow | 14:40 | |
| *** Niklas[m]1 has joined #symbiflow | 14:40 | |
| sf-slack | <acomodi> lmabda: you mean doesn't work on HW? | 14:40 |
| lambda | acomodi: yeah, there aren't really any life signs - will have to investigate and employ some debug LEDs | 14:50 |
| sf-slack | <acomodi> lambda: All right. One thing to check would be timing violations, and, if they exists, how bad they are. Just to be sure, have you applied time constraints in the XDC (or SDC)? | 14:56 |
| lambda | acomodi: I have a create_clock for all my PLL inputs and outputs, though I think the outputs may actually be induced automatically with this flow? | 15:07 |
| lambda | at least if `propagate_clocks` in synth.tcl does what I think it does | 15:08 |
| lambda | iirc routing failed setup timing massively, I assumed that was just a glitch since (1) timing reports with nextpnr were always pretty sketchy and (2) the counter example also failed in a similar fashion | 15:09 |
| sf-slack | <acomodi> Well, you can check if the propagate_clocks did something if, in the build directory you got an <design>.sdc file. That should include all propagated clock constraints for VPR | 15:10 |
| sf-slack | <acomodi> Also, you can check if timing constraints got correctly applied if the `route.log` log file contains a string such as `Applied <N> SDC commands from <SDC file name>` | 15:12 |
| lambda | no, doesn't seem like clocks are actually propagated over PLLs, I'll just continue defining them manually | 15:16 |
| sf-slack | <acomodi> Ok, there is also another thing to do which is add a clock constraint in the XDC to the top level clock port (if you get the clock from there) | 15:18 |
| sf-slack | <acomodi> Sth like `create_clock -period <period> <input_clock_port>` | 15:19 |
| lambda | that port is only used as my PLL input and I have a create_clock on that | 15:21 |
| lambda | -50000ns sTNS and -20ns sWNS sure would break about everything if those numbers are to be believed ;) | 15:23 |
| lambda | oops, my bad, I'm actually using the input clock for logic that's shared with the core clock - no surprise that doesn't work | 15:26 |
| *** citypw has quit IRC | 15:27 | |
| lambda | can vtr show the critical path? I don't think these two clocks share any logic anymore, but they still show up as a huge sWNS after placement | 16:05 |
| sf-slack | <acomodi> I am not sure if this is by default, but VTR can emit a detailed report with the worst paths (report_timing.setup.rpt) | 16:07 |
| litghost | I believe the reports are generated by default | 16:08 |
| litghost | I believe it prints the 100 highest critical path, both setup and hold | 16:09 |
| lambda | symbiflow_place isn't showing it, I'll look into that option | 16:11 |
| sf-slack | <arvindsrinivasan> Hello! So I’m attempting to install symbiflow from source per this page (https://symbiflow.readthedocs.io/en/latest/symbiflow-arch-defs/docs/source/getting-started.html) and it seems the `make env` stage is failing due to conda not being able to find these packages: ``` - litex-hub::gcc-riscv64-elf-newlib==9.2.0=20201119_154229 - litex-hub::zachjs-sv2v==0.0.5_0025_ge9f9696=20201120_205532 - | 16:12 |
| sf-slack | litex-hub::vtr-optimized==v8.0.0_3049_g64d15e2dd=20201217_025303 - litex-hub::capnproto-java==0.1.5_0012_g44a8c1e=20201104_165332 - litex-hub::prjxray-tools==0.1_2697_g0f939808=20201120_145821 - litex-hub::prjxray-db==v0.0_243_g057e179=20210119_091808``` Is this to be expected or is there further configuration that would be necessary. On MacOS 10.14.6 doing a clean install. | 16:12 |
| tpb | Title: Getting Started SymbiFlow (at symbiflow.readthedocs.io) | 16:12 |
| sf-slack | <acomodi> Oh, I mis-read. I don't think that after placement you have this information, you would need to wait for routing | 16:13 |
| sf-slack | <acomodi> Hi @arvindsrinivasan, FYI if your purpose is to just use the toolchain, I suggest you to read from this guide instead: https://symbiflow-examples.readthedocs.io/en/latest/ | 16:16 |
| litghost | We don't have MacOS packages right now | 16:16 |
| tpb | Title: Welcome to SymbiFlow examples! SymbiFlow examples documentation (at symbiflow-examples.readthedocs.io) | 16:16 |
| sf-slack | <arvindsrinivasan> @acomodi atm, I’ll be doing some work on the toolchain so I was hoping to install it fully rather than the packages | 16:17 |
| sf-slack | <arvindsrinivasan> But yea, understandable if there aren’t MacOS packages. I’ll use a linux vm instead | 16:17 |
| lambda | litghost: no, nothing from symbiflow_route either - just the WNS per domain | 16:21 |
| lambda | oh wait, report_timing.setup.rpt is a file, I thought it was an option - that looks promising | 16:23 |
| litghost | Ya, it's a file | 16:24 |
| litghost | There is a way to increase the verbosity of the report if needed | 16:24 |
| litghost | But I remember the default being ok | 16:25 |
| lambda | yeah, I think I've found my problem | 16:26 |
| litghost | Great | 16:30 |
| *** kgugala_ has joined #symbiflow | 16:41 | |
| *** kgugala has quit IRC | 16:44 | |
| *** kgugala__ has joined #symbiflow | 16:45 | |
| *** kgugala_ has quit IRC | 16:45 | |
| *** infinite_recursi has joined #symbiflow | 16:50 | |
| *** infinite_recursi has quit IRC | 17:01 | |
| *** kraiskil has joined #symbiflow | 17:44 | |
| nickoe | how goes? | 17:48 |
| litghost | nickoe: I missing something in my fixed yesterday, I'll fix it today | 17:48 |
| *** SmutLord^ has quit IRC | 17:49 | |
| lambda | litghost: got it semi-working! still failing timing and something is wrong with interrupts, but that may just be because of the former | 17:50 |
| *** SmutLord^ has joined #symbiflow | 17:50 | |
| litghost | lambda: Almost certainly | 17:50 |
| sf-slack | <arvindsrinivasan> Alright, now I have further questions on the install process on Ubuntu 20.10. I’m getting an error at in the install process regarding the rapidyaml package. I can’t seem to find the package on pypi, so has its name changed? | 17:52 |
| litghost | Pull latest version | 17:56 |
| litghost | Should be fixed | 17:56 |
| nickoe | litghost: cool :) | 17:58 |
| *** FFY00 has quit IRC | 18:02 | |
| *** ym_ has quit IRC | 18:03 | |
| *** FFY00 has joined #symbiflow | 18:04 | |
| sf-slack | <arvindsrinivasan> Alright @litghost sorry to bother again, but now I’m getting a error building a wheel for rapidyaml since it uses PEP 517 and can’t be installed directly? | 18:10 |
| litghost | Error? | 18:10 |
| sf-slack | <arvindsrinivasan> Copy paste is being weird so image | 18:11 |
| *** nickoe has quit IRC | 18:23 | |
| *** kgugala has joined #symbiflow | 18:36 | |
| *** kgugala__ has quit IRC | 18:39 | |
| sf-slack | <arvindsrinivasan> @litghost | 18:44 |
| litghost | What? You didn't paste anything? | 18:44 |
| sf-slack | <mgielda> its visible on slack | 18:45 |
| sf-slack | <mgielda> maybe not on IRC though | 18:45 |
| sf-slack | <mgielda> perhaps pastebin or similar will work @arvindsrinivasan | 18:45 |
| litghost | Nothing on IRC | 18:46 |
| litghost | Anyways | 18:46 |
| litghost | arvindsrinivasan: The error is that the build failed | 18:46 |
| litghost | You need to keep scrolling up | 18:46 |
| litghost | You copied the wrong error | 18:46 |
| litghost | The error above say "Failed to build wheel", and then the log above that is why the wheel failed to build | 18:46 |
| sf-slack | <arvindsrinivasan> Alright sorry for the trouble, here is the pastebin https://pastebin.com/L93Dxu4v | 18:51 |
| tpb | Title: Failed to build rapidyamlPip subprocess error: ERROR: Command errored out - Pastebin.com (at pastebin.com) | 18:51 |
| sf-slack | <arvindsrinivasan> Seems like it fails to find SWIG (Could NOT find SWIG *(*missing: SWIG_EXECUTABLE SWIG_DIR*))* | 18:53 |
| litghost | Yep | 19:32 |
| *** maartenBE has quit IRC | 19:40 | |
| *** maartenBE has joined #symbiflow | 19:42 | |
| *** nickoe has joined #symbiflow | 20:35 | |
| *** kraiskil has quit IRC | 22:26 | |
| nickoe | litghost: yo, I am back. (In case you wrote something when I was offline) | 22:30 |
| *** QDX45 has joined #symbiflow | 22:58 | |
| *** FFY00 has quit IRC | 23:20 | |
| *** hansfbaier has joined #symbiflow | 23:49 | |
| *** FFY00 has joined #symbiflow | 23:57 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!