Closed Bug 958404 Opened 6 years ago Closed 6 years ago

Build mfbt, ffi and icu without horrible hacks in js/src

Categories

(Firefox Build System :: General, defect)

defect
Not set

Tracking

(Not tracked)

RESOLVED FIXED
mozilla29

People

(Reporter: glandium, Assigned: glandium)

References

Details

Attachments

(4 files)

No description provided.
Depends on: 958437
Depends on: 959508
Attachment #8360287 - Flags: review?(gps)
Blocks: 960505
Attachment #8360287 - Flags: review?(gps) → review+
Attachment #8360288 - Flags: review?(gps) → review+
Attachment #8360289 - Flags: review?(gps) → review+
Comment on attachment 8360290 [details] [diff] [review]
Recurse icu from top-level when building js

Review of attachment 8360290 [details] [diff] [review]:
-----------------------------------------------------------------

This should be a nice speed win!

I didn't do a super thorough job verifying the new Makefile.in matches the old content. Looks pretty similar and I trust you'd tell me if other changes were sneaking in here.

::: intl/icu/Makefile.in
@@ +74,5 @@
> +distclean clean::
> +ifdef CROSS_COMPILE
> +	$(call SUBMAKE,$@,host)
> +endif
> +	$(call SUBMAKE,$@,target)

I guess we're lucky and don't hit any order-only prerequisites in the clean targets (since we're not using $(ICU_MAKE) here).
Attachment #8360290 - Flags: review?(gps) → review+
(In reply to Gregory Szorc [:gps] from comment #5)
> Comment on attachment 8360290 [details] [diff] [review]
> Recurse icu from top-level when building js
> 
> Review of attachment 8360290 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> This should be a nice speed win!
> 
> I didn't do a super thorough job verifying the new Makefile.in matches the
> old content. Looks pretty similar and I trust you'd tell me if other changes
> were sneaking in here.

There is one change to make it a static directory (in that there's now only one rule to build stuff, instead of choosing between compile and libs depending on MOZ_PSEUDO_DERECURSE)

> ::: intl/icu/Makefile.in
> @@ +74,5 @@
> > +distclean clean::
> > +ifdef CROSS_COMPILE
> > +	$(call SUBMAKE,$@,host)
> > +endif
> > +	$(call SUBMAKE,$@,target)
> 
> I guess we're lucky and don't hit any order-only prerequisites in the clean
> targets (since we're not using $(ICU_MAKE) here).

Actually, iirc, pymake will barf if a makefile contains an order-only prereq at parse time, so we may actually hit this already. But since barely anyone runs make clean... want me to change this while here?
Please change SUBMAKE to ICU_MAKE. carry r+ forward.
Blocks: 968652
Depends on: 971426
Product: Core → Firefox Build System
You need to log in before you can comment on or make changes to this bug.