@average-vr-56795: you should not, no. not if there is an actual rule that can satisfy

Y for X

(otherwise we'd have hundreds of them in our existing rules)

I’ll get a specific repro for you 🙂

Also, I had some fun debugging your rust 1.26 issues

thanks a ton for doing that!

yes, thanks!

sometimes rustc gives bad advice when there’s more than one thing wrong

Has anyone gotten around to working out options passing for a v2 world yet? As far as I can tell right now the engine only gets hard-coded bootstrap options…

I’ve seen a few travis shards failing because they’re stuck indefinitely

Blocking waiting for file lock on the registry index

, e.g. https://travis-ci.org/pantsbuild/pants/jobs/379683408 - anyone seen this before?

I haven’t seen that before

it looks like there’s a fetch process that’s also running, and not finishing, is that your hypothesis too?

[Waiting for /home/travis/.cache/pants/rust/cargo/bin/cargo fetch --manifest-path /home/travis/build/pantsbuild/pants/src/rust/engine/Cargo.toml (pid 7114) to complete]
    Blocking waiting for file lock on the registry index

I think the fetch isn’t starting because it’s blocked on a lock

I think you’re right. I hadn’t checked the shell script and I’d thought that the

Waiting

logs were from the same process as the

Blocking

one

@average-vr-56795 mind if I file a “local exec timeout” issue. When I added timeout to ExecuteProcessRequest, I tried to add a test for local exec and noticed that it won’t work as it is

I think I know how to fix it, but I think it should be a separate story

Please do!

It will probably get resolved by https://github.com/pantsbuild/pants/issues/5763 but will need a ~trivial amount of additional work on top of that 🙂

😍

heh, neat

Leave the GIL, take the cannoli?

What should the default timeout be for ExecuteProcessRequest? Or should I make it an Option instead of a required value

@fancy-queen-20734: 15 minutes maybe.

k

i found

src/docs/invoking.md

and can view the python call graph, but i can't seem to figure out how to get a profile of the native engine stack. i'm trying (after killing pantsd):

valgrind --tool=callgrind -v --dump-every-bb=10000000 --trace-children=yes -- ./pants --glob-expansion-failure=ignore filedeps src/python/pants/util:objects

right now, but this errors out with:

Exception message: Build graph construction failed: ExecutionError Received unexpected Throw state(s):
Computing Select(Specs(dependencies=(SingleAddress(directory=u'src/python/pants/util', name=u'objects'),)), =TransitiveHydratedTargets)
  Computing Task(transitive_hydrated_targets, Specs(dependencies=(SingleAddress(directory=u'src/python/pants/util', name=u'objects'),)), =TransitiveHydratedTargets)
    Computing Task(addresses_from_address_families, Specs(dependencies=(SingleAddress(directory=u'src/python/pants/util', name=u'objects'),)), =BuildFileAddresses)
      Computing Snapshot(PathGlobs(include=(u'src/python/pants/util/BUILD.*', u'src/python/pants/util/BUILD'), exclude=(u'.*/', '/dist/', u'bower_components/', u'node_modules/', u'*.egg-info/')))
        Computing DigestFile(File { path: "src/python/pants/util/BUILD", is_executable: false })
          Throw(Error making env for store at "/home/cosmicexplorer/.cache/pants/lmdb_store/files/0": Invalid argument)
            Traceback (no traceback):
              <pants native internals>
            Exception: Error making env for store at "/home/cosmicexplorer/.cache/pants/lmdb_store/files/0": Invalid argument

is there a "normal" way to profile the native engine? i can't immediately interpret this error (working on it!)

i've always used OSX Instruments for profiling

but maybe you mean debugging?

i meant profiling, i'm on a linux laptop right now but have my macbook pro right next to me