search

snowleopard / hadrian

Hadrian: a new build system for the Glasgow Haskell Compiler. Now merged into the GHC tree!
https://gitlab.haskell.org/ghc/ghc/tree/master/hadrian
MIT License
208 stars 39 forks source link

Added support for testsuite #602

Closed chitrak7 closed 6 years ago

chitrak7 commented 6 years ago

@snowleopard @angerman Validation result after this push: SUMMARY for test run started at Mon May 14 00:32:42 2018 IST 0:20:37 spent to go through 6359 total tests, which gave rise to 23523 test cases, of which 17231 were skipped

  42 had missing libraries
6098 expected passes
  94 expected failures

   0 caused framework failures
   0 caused framework warnings
   1 unexpected passes
  55 unexpected failures
   2 unexpected stat failures
snowleopard commented 6 years ago

@chitrak7 Thanks! I've added some comments.

alpmestan commented 6 years ago

@chitrak7 Do you have any idea at this point of why we're still skipping so many tests? (this is not a comment/request on the actual patch, I'm just curious and this still feels like the best place to ask)

chitrak7 commented 6 years ago

@alpmestan I do not exactly know that, Can you give me pointers on where to look. More importantly, why does testsuite decide to skip a test.

chitrak7 commented 6 years ago

@snowleopard Can you please see why recursive rule failure is coming?

alpmestan commented 6 years ago

@alpmestan I do not exactly know that, Can you give me pointers on where to look. More importantly, why does testsuite decide to skip a test.

You should see some information about this in the testsuite driver initialization part of the output (early on, when you look at the full log).

chitrak7 commented 6 years ago

@alpmestan I believe skipped tests are because of using "fast" option with testsuite.

alpmestan commented 6 years ago

@chitrak7 OK. If you ever get a chance to, it would be terrific to add a way to run the equivalent of ./validate --slow. If not, I'll do it. But not in this PR of course. Thanks for the reply and taking a look at what was going on :+1:

snowleopard commented 6 years ago

@chitrak7 Please resolve merge conflicts.

snowleopard commented 6 years ago

Can you please see why recursive rule failure is coming?

@chitrak7 The recursive rule failure is unrelated to your PR. I think the reason is #599.

snowleopard commented 6 years ago

@chitrak7 Something is wrong, looks like you've duplicated the recent commits instead of rebasing.

chitrak7 commented 6 years ago

@snowleopard I have reset my repository and made necessary changes. It is currently failing due to "llvm" issue. I will resend it once that PR Lands.

sighingnow commented 6 years ago

Should the title of this PR be "Add support for nofib testsuite" ? We already support hadrian test.

alpmestan commented 6 years ago

No, support for nofib was the goal of #599, not this PR. This PR is about supporting the execution of more tests, and about building things that are required by some of them which were not built previously. It still skips a whole bunch of tests but maybe @chitrak7 could provide an overview of the state of things with/without the commits from his branch? In terms of # of tests being run, # of broken test expectations.

sighingnow commented 6 years ago

I just noticed that this PR also add a new file src/Rules/Nofib.hs...

alpmestan commented 6 years ago

Right, my commits somehow did not get merged properly apparently. But you cann tell they've already been merged because they refer to the PR number.

chitrak7 commented 6 years ago

Hi @snowleopard @angerman, I have pushed the work done with validation implementation.

snowleopard commented 6 years ago

@chitrak7 Thanks! Let's come back to this PR once we fix Hadrian.

snowleopard commented 6 years ago

OK, I think we can merge this -- there are a few minor issues but we can fix them later.

@alpmestan Do you have any further comments?

alpmestan commented 6 years ago

No, let's merge :+1:

snowleopard commented 6 years ago

Merged. Many thanks @chitrak7, and everyone else for the review!