1253553 - Move --enable-artifact-builds, --disable-compile-environment and --enable-build-backend to moz.configure

Status: RESOLVED FIXED

(Firefox Build System :: General, defect)

Description

[Mike Hommey [:glandium]]

They are all tied together.

Comment 1

[Mike Hommey [:glandium]]

Attachment: MozReview Request: Bug 1253553 - Move --enable-artifact-builds, --disable-compile-environment and --enable-build-backend to moz.configure

Note I'm using raw primitives until patterns emerge and we decide on
some helpers.

Review commit: https://reviewboard.mozilla.org/r/38137/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/38137/

Comment 2

[Nick Alexander :nalexander [he/him]]

Comment on attachment 8726649 [details]
MozReview Request: Bug 1253553 - Move --enable-artifact-builds, --disable-compile-environment and --enable-build-backend to moz.configure

https://reviewboard.mozilla.org/r/38137/#review34881

::: build/moz.configure/old.configure:66
(Diff revision 1)
> -                      python):
> +                      python, compile_env):

It would help grep if the name of the @depends function and the name of the variable were the same, although perhaps that obscures more than clarifies.

::: moz.configure:33
(Diff revision 1)
> +def build_backends_choices(help):

Again, why `help`?  Worth calling the pattern out until it's standard.

::: moz.configure:38
(Diff revision 1)
> +option('--enable-build-backend', nargs='+', choices=build_backends_choices,

This choices=decorated function strikes my eye as odd.

::: moz.configure:48
(Diff revision 1)
> +    set_config('BUILD_BACKENDS', unique_list(all_backends))

Why doesn't this produce `['FM+RM', 'RM', 'FM']`?  Is that not a problem?

Comment 3

[Chris Manchester (limited bugmail, email directly)]

Comment on attachment 8726649 [details]
MozReview Request: Bug 1253553 - Move --enable-artifact-builds, --disable-compile-environment and --enable-build-backend to moz.configure

https://reviewboard.mozilla.org/r/38137/#review35125

Great stuff.

Comment 4

[Chris Manchester (limited bugmail, email directly)]

https://reviewboard.mozilla.org/r/38137/#review35205

::: moz.configure:26
(Diff revision 1)
> +    if value:
> +        set_config('COMPILE_ENVIRONMENT', '1')

I'm finding this breaks an artifact enabled build locally, because of this check, which relies on COMPILE_ENVIRONMENT being set to an empty value:

https://dxr.mozilla.org/mozilla-central/rev/be593a64d7c6a826260514fe758ef32a6ee580f7/python/mozbuild/mozbuild/frontend/emitter.py#184

Comment 5

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/addd5fc00284

Comment 6

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/addd5fc00284