Select JSOP_POP not JSOP_POPV for non-tail expression statements

ASSIGNED
Assigned to

Status

()

Core
JavaScript Engine
P3
enhancement
ASSIGNED
8 years ago
4 years ago

People

(Reporter: brendan, Assigned: brendan)

Tracking

(Depends on: 1 bug)

Firefox Tracking Flags

(Not tracked)

Details

Attachments

(1 attachment)

(Assignee)

Description

8 years ago
Created attachment 445019 [details] [diff] [review]
proposed fix

Seems easy enough. Perf results in a few.

/be
Attachment #445019 - Flags: review?(mrbkap)
(Assignee)

Comment 1

8 years ago
Perf change is noise.

/be
So, I seem to recall the tail call proposal in harmony having a complex definition of what consists of "tail position", and that it involved determining whether you're in a catch block, etc. Do we not have to do that here?
(Assignee)

Comment 3

8 years ago
(In reply to comment #2)
> So, I seem to recall the tail call proposal in harmony having a complex
> definition of what consists of "tail position", and that it involved
> determining whether you're in a catch block, etc. Do we not have to do that
> here?

Done in patch for bug 565374, which blocks this and on which bug's patch this bug's patch is based (notice use of pn_tailpos, which is not in tm tip yet!). See the URL for bug 565374 which links to the ES4-era spec.

/be

Updated

8 years ago
Attachment #445019 - Flags: review?(mrbkap) → review+
(Assignee)

Updated

7 years ago
Target Milestone: mozilla1.9.3a5 → ---
Blake: is Brendan's r+'d patch (from 2010) still relevant?
Flags: needinfo?(mrbkap)
Not really. The code has changed dramatically since this patch was written (and we'd have to land bug 565374 first anyway).
Flags: needinfo?(mrbkap)
You need to log in before you can comment on or make changes to this bug.