Closed
Bug 1163139
Opened 9 years ago
Closed 7 years ago
Do not set an async stack for Promises in Gecko 39
Categories
(Core :: DOM: Core & HTML, defect)
Tracking
()
RESOLVED
INVALID
People
(Reporter: Paolo, Unassigned)
References
(Blocks 1 open bug)
Details
Attachments
(1 file, 1 obsolete file)
We should back out bug 1140472 for Gecko 39 while we work on bug 1142577 and bug 1158133. This requires adjusting a few tests as well.
Reporter | ||
Comment 1•9 years ago
|
||
/r/8461 - Bug 1163139 - Do not set an async stack for Promises in Gecko 39. r=bz Pull down this commit: hg pull -r 833ce94f04f4c8d1891eb3eb0e8b54ca2f754ef4 https://reviewboard-hg.mozilla.org/gecko/
Attachment #8603812 -
Flags: review?(bzbarsky)
Reporter | ||
Comment 2•9 years ago
|
||
https://treeherder.mozilla.org/#/jobs?repo=try&revision=f9adf4b2292d
Reporter | ||
Comment 3•9 years ago
|
||
There are a few consistent failures in the try build but I'm not sure whether they are related to the patch, it's quite possible I missed something when pushing to try from the Aurora branch. These failures are not present for the same patch on mozilla-central.
Flags: needinfo?(lhenry)
Comment 4•9 years ago
|
||
Comment on attachment 8603812 [details] MozReview Request: bz://1163139/paolo https://reviewboard.mozilla.org/r/8459/#review7155 This seems ok for 39, but what about 40? It looks like the current nightly code will go to Aurora today or tomorrow.... What I think we should do is make the C++ bits #ifndef RELEASE_BUILD for now (or #ifdef NIGHTLY_BUILD, depending on whether we want async stacks on dev edition, and I suspect we do) and either adjust tests as needed on uplifts or make the tests automatically adjust based on channel or something. r=me with at least the C++ ifdef bits.
Attachment #8603812 -
Flags: review?(bzbarsky) → review+
Comment 5•9 years ago
|
||
(In reply to Not doing reviews right now from comment #4) > This seems ok for 39, but what about 40? It looks like the current nightly > code will go to Aurora today or tomorrow.... > > What I think we should do is make the C++ bits #ifndef RELEASE_BUILD for now > (or #ifdef NIGHTLY_BUILD, depending on whether we want async stacks on dev > edition, and I suspect we do) and either adjust tests as needed on uplifts > or make the tests automatically adjust based on channel or something. Yes, we are planning to merge 40 to aurora today as soon as possible.
Comment 6•9 years ago
|
||
The tree is closed, and we aren't going to build on 39 beta 1 until May 22, and shipping 39 updates until May 26th. For 40, we are turning off updates until this Friday. So you can land this after today's merge rather than try to land this today and force us to rebuild.
Flags: needinfo?(lhenry) → needinfo?(paolo.mozmail)
Reporter | ||
Comment 7•9 years ago
|
||
(In reply to Liz Henry (:lizzard) from comment #6) > we aren't going to build on 39 beta 1 until May 22, > and shipping 39 updates until May 26th. Good, this actually gives us some time to work on bug 1158133 instead of the stopgap in this bug. If we do the #ifdef we'll need to make the tests conditional anyways so we might as well place the disabling code in the right place. We can disable on mobile and add the #ifndef RELEASE_BUILD on Desktop until bug 1142577 is done.
tracking-firefox39:
? → ---
Flags: needinfo?(paolo.mozmail)
Reporter | ||
Comment 8•9 years ago
|
||
Attachment #8603812 -
Attachment is obsolete: true
Attachment #8620262 -
Flags: review+
Reporter | ||
Comment 9•9 years ago
|
||
Comment 10•7 years ago
|
||
Is this still open for a reason?
Updated•7 years ago
|
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → INVALID
Assignee | ||
Updated•5 years ago
|
Component: DOM → DOM: Core & HTML
You need to log in
before you can comment on or make changes to this bug.
Description
•