1229233 - Add a close-to-end-to-end test of the build system

Status: RESOLVED FIXED

(Firefox Build System :: General, defect)

Description

[Mike Hommey [:glandium]]

We have many unit tests in the tree for some small parts of the build system pipeline, but we don't have anything that resembles an end to end test, and we kind of rely on the resulting Firefox not being broken by our changes.

With the Faster make backend growing, I want to ensure it produces the same thing as the recursive make backend, at least for the parts it supports. As a consequence, I want to add some kind of test that allows to check that.

It will start incomplete, but will grow with time.

Comment 1

[Mike Hommey [:glandium]]

Attachment: Use SourcePaths for CONFIGURE_SUBST_FILES and CONFIGURE_DEFINE_FILES

Comment 2

[Mike Hommey [:glandium]]

Attachment: Use SourcePaths for CONFIGURE_SUBST_FILES and CONFIGURE_DEFINE_FILES

Comment 3

[Mike Hommey [:glandium]]

Attachment: Skip building XPIDL when there aren't any in the FasterMake backend

The test I'm about to add doesn't have XPIDL files, and that currently avoids
the FasterMake backend to run properly. Also, in the future, when the FasterMake
grows the ability to build C++ files, it should be possible to build Spidermonkey
with the FasterMake backend, but it doesn't have XPIDL files either.

Comment 4

[Mike Hommey [:glandium]]

Attachment: Ensure the target directory for copies exists in the FasterMake backend

Comment 5

[Mike Hommey [:glandium]]

Attachment: Properly skip the top-level directory in the RecursiveMake backend

Comment 6

[Mike Hommey [:glandium]]

Attachment: Add a (incomplete) end-to-end build test

We have many unit tests in the tree for some small parts of the build system
pipeline, but we don't have anything that resembles an end to end test, and we
kind of rely on the resulting Firefox not being broken by our changes.

With the Faster make backend growing, I want to ensure it produces the same
thing as the recursive make backend, at least for the parts it supports.

This adds some kind of test that allows to check that.

Comment 7

[Mike Hommey [:glandium]]

FWIW, this test already allowed me to find a subtle regression in the patch queue I'm going to attach to a new, separate, bug.

Comment 8

[Gregory Szorc [:gps]]

Comment on attachment 8693906 [details] [diff] [review]
Ensure the target directory for copies exists in the FasterMake backend

Review of attachment 8693906 [details] [diff] [review]:
-----------------------------------------------------------------

Wish you had a better commit message for why this is required - I figured our install manifests had comprehensive enough coverage that this type of thing wasn't required any more.

Comment 9

[Gregory Szorc [:gps]]

Comment on attachment 8693908 [details] [diff] [review]
Add a (incomplete) end-to-end build test

Review of attachment 8693908 [details] [diff] [review]:
-----------------------------------------------------------------

Niiiice. I didn't look at the low-level test inputs and outputs too much, as changes are what we care about. Obviously there's more things we could test here. But you have to start somewhere.

Comment 10

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/cdcf951c2070
https://hg.mozilla.org/integration/mozilla-inbound/rev/183a19236984
https://hg.mozilla.org/integration/mozilla-inbound/rev/350085f66138
https://hg.mozilla.org/integration/mozilla-inbound/rev/603b6593c161
https://hg.mozilla.org/integration/mozilla-inbound/rev/e03b93cbb64d

Comment 11

[Mike Hommey [:glandium]]

Comment on attachment 8693908 [details] [diff] [review]
Add a (incomplete) end-to-end build test

Review of attachment 8693908 [details] [diff] [review]:
-----------------------------------------------------------------

::: python/mozbuild/mozbuild/test/backend/test_build.py
@@ +109,5 @@
> +
> +    def validate(self, config):
> +        self.maxDiff = None
> +        test_path = mozpath.join(mozpath.dirname(mozpath.abspath(__file__)),
> +                                 'data', 'build')

I folded a change here and in the corresponding format strings below to account for Windows when landing.

Comment 12

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/cf71b87eafcf
https://hg.mozilla.org/integration/mozilla-inbound/rev/958e2652c301

Comment 13

[Mike Hommey [:glandium]]

And a couple fixups:
https://hg.mozilla.org/integration/mozilla-inbound/rev/cf71b87eafcf
  for bustage on Windows because of the backslashes in the PYTHON variable
https://hg.mozilla.org/integration/mozilla-inbound/rev/958e2652c301
  for bustage on OSX Universal builds because of the objdir discrepancy. I copied bits from TestMozbuildObject.

Comment 14

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/cdcf951c2070
https://hg.mozilla.org/mozilla-central/rev/183a19236984
https://hg.mozilla.org/mozilla-central/rev/350085f66138
https://hg.mozilla.org/mozilla-central/rev/603b6593c161
https://hg.mozilla.org/mozilla-central/rev/e03b93cbb64d
https://hg.mozilla.org/mozilla-central/rev/cf71b87eafcf
https://hg.mozilla.org/mozilla-central/rev/958e2652c301