05:36 <mithro> https://source.cloud.google.com/results/invocations/19862a66-f746-45ad-97e8-b3f9140f291d/targets;collapsed=

06:09 <sxpert> gah, I'm in the "where did all my gates go" hell again

06:10 <sxpert> yosys seems to optimize out the whole thing

07:24 <sxpert> ah, I know

07:58 <sxpert> nope, no go

08:44 <kgugala> mithro: I can access the build status page

08:44 <nats`> mithro, sure go ahead on that topic I really have a lot of work

09:25 <sxpert> ah, in fact, yosys blows up :-(

09:37 <mithro> https://source.cloud.google.com/results/invocations/fd827790-2611-4e09-97c8-93bcb74387d7/targets

09:40 <mithro> kgugala: Example I have been testing with -> https://github.com/SymbiFlow/prjxray-kokoro/pull/1

09:40 <tpb> Title: Update README.md by mithro · Pull Request #1 · SymbiFlow/prjxray-kokoro · GitHub (at github.com)

11:09 <mithro> kgugala / litghost / acomodi / elms: Just merged the GCP CI stuff

11:10 <mithro> kgugala / litghost / acomodi / elms: Feel free to ignore if you think it is giving bogus results for now

11:10 <mithro> but do log a bug

11:21 <mithro> nats`: FYI - https://github.com/SymbiFlow/prjxray/issues/494

11:21 <tpb> Title: Make each fuzzer output a test_results.xml file · Issue #494 · SymbiFlow/prjxray · GitHub (at github.com)

11:22 <nats`> cool mithro

11:22 <nats`> but you shouldn't be in bed already ? :D

11:37 <mithro> Yes

22:29 <elms> mithro: After sleeping on it, I think if we are keeping travis in place, then we should make it more than policy and make it required if possible. When kokoro fails is there an email to notify a list or something at least? Or can it comment on the PR? If it isn't highly visible I think it will be easy to miss or forget.

22:29 <mithro> elms: The kokoro stuff should appear at the bottom in the same area Travis does

22:30 <mithro> I haven't set up the emailing yet as I wasn't sure everything was working okay

22:31 <elms> Ok

22:37 <litghost> mithro: Do you have an example of the kokoro PR display? I haven't seen it yet

22:37 <litghost> Maybe only new PR's will have it?

22:38 <mithro> Hrm - it should be appearing on https://github.com/SymbiFlow/prjxray/pull/620

22:38 <tpb> Title: WIP: adding DSP for FAN_ALTs and BYPs to 050-pip-seed by acomodi · Pull Request #620 · SymbiFlow/prjxray · GitHub (at github.com)

22:38 <mithro> #620 opened 12 hours ago by acomodi

22:39 <mithro> Let me see if I broke something when moving it from my testing prjxray-kokoro repository to just prjxray

22:59 <mithro> Doah! it seems I forgot to update the hooks...

23:00 <mithro> litghost: https://github.com/SymbiFlow/prjxray/pull/622

23:00 <tpb> Title: WIP: Testing the Kokoro CI building... by mithro · Pull Request #622 · SymbiFlow/prjxray · GitHub (at github.com)

23:06 <litghost> What's the difference between presubmit, presubmit - tests, presubmit - database?

23:06 <litghost> Also, the diffing logic is not working on continuous database builds

23:06 <litghost> ----------------------------------------

23:06 <litghost> Generating pretty diff

23:06 <litghost> ----------------------------------------

23:06 <litghost> The input is empty. Try again.

23:06 <litghost> Usage: diff2html [options] -- [diff args]

23:23 <mithro> litghost: "Presubmit - Tests" does a "make test" and the "Presubmit - Database" targets actually run the fuzzers and generate database output

23:24 <litghost> Is the presubmit a quick run, or a full run?

23:24 <mithro> litghost: Full run

23:25 <mithro> btw the config files are found at https://github.com/SymbiFlow/prjxray/tree/master/.github/kokoro

23:25 <tpb> Title: prjxray/.github/kokoro at master · SymbiFlow/prjxray · GitHub (at github.com)

23:25 <litghost> I wouldn't expect to gate a PR for a full run, that's order ~6 hrs?

23:25 <mithro> litghost: A full run only takes ~2

23:25 <mithro> https://github.com/SymbiFlow/prjxray/blob/master/.github/kokoro/db-full.sh

23:25 <tpb> Title: prjxray/db-full.sh at master · SymbiFlow/prjxray · GitHub (at github.com)

23:25 <litghost> It's splitting each build onto another machine?

23:26 <mithro> Yes, each Artix 7, Kintex 7 and Zynq 7 build is running on a seperate machine

23:26 <litghost> omph

23:27 <mithro> omph?

23:27 <litghost> Wow, etc

23:27 <mithro> I wouldn't suggest we block merging a pull request on the Kokoro builds

23:27 <mithro> But we should effectively get results for how a pull request changes the database output

23:42 <mithro> litghost: Will see if https://github.com/SymbiFlow/prjxray/pull/624/files helps with the diff issue...

23:42 <tpb> Title: infra: Improve the diff output. by mithro · Pull Request #624 · SymbiFlow/prjxray · GitHub (at github.com)

23:43 <mithro> litghost: Did you mean to merge that? I was going to wait until the presubmit finished

23:44 <mithro> litghost: Both the presubmit and continuous runs use the same script

23:51 <mithro> litghost: What is the state of the "QUICK=Y" database builds?

23:52 <litghost> I haven't been leveraging them

23:52 <litghost> We should probably tune QUICK=Y again to a reasonable build time

23:52 <mithro> litghost: Originally I was planning on running QUICK=Y before full builds (and only running the full builds if the quick build succeeded) to try and get people feedback quicker