Closed Bug 832063 Opened 8 years ago Closed 8 years ago

IonMonkey: tidy up flush/finish

Categories

(Core :: JavaScript Engine, defect)

defect
Not set
normal

Tracking

()

RESOLVED FIXED
mozilla21

People

(Reporter: luke, Assigned: luke)

Details

Attachments

(1 file)

Attached patch patchSplinter Review
No description provided.
Attachment #703635 - Flags: review?(mrosenberg)
Comment on attachment 703635 [details] [diff] [review]
patch

Review of attachment 703635 [details] [diff] [review]:
-----------------------------------------------------------------

Is this to align the definitions of flush() and finish() on arm and x86, since arm has both, and x86 should only have one of the two?

::: js/src/ion/IonLinker.h
@@ -29,5 @@
>      }
>  
>      IonCode *newCode(JSContext *cx, IonCompartment *comp) {
>          AssertCanGC();
> -#ifndef JS_CPU_ARM

Shouldn't this be #ifdef JS_CPU_ARM?
Attachment #703635 - Flags: review?(mrosenberg) → review+
I don't understand that question; are you asking why the code was removed or why it was originally the way it was?
(In reply to Marty Rosenberg [:mjrosenb] from comment #1)
> Is this to align the definitions of flush() and finish() on arm and x86,
> since arm has both, and x86 should only have one of the two?

Yeah, the goal is to have one "finish()" called in one place, rather than two finish-like calls called in 2 different places, one surrounded by an #ifdef.
https://hg.mozilla.org/mozilla-central/rev/879677a0aa37
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla21
You need to log in before you can comment on or make changes to this bug.