Closed
Bug 1205621
Opened 9 years ago
Closed 9 years ago
arm64: propagate OOM from calls to buffer.allocEntry
Categories
(Core :: JavaScript Engine: JIT, defect)
Tracking
()
RESOLVED
FIXED
mozilla44
Tracking | Status | |
---|---|---|
firefox44 | --- | fixed |
People
(Reporter: lth, Assigned: jolesen)
References
Details
Attachments
(1 file)
15.62 KB,
patch
|
lth
:
review+
|
Details | Diff | Splinter Review |
See bug 1186982. Calls to allocEntry() need to check the result, if the result value is a dummy (unassigned) value then OOM should be signaled via propagateOOM(). On ARM32 there's now one call to allocEntry but ARM64 appears to have several.
Assignee | ||
Comment 1•9 years ago
|
||
This seems like a fine thing for me to work on once I get SM(arm64) into treeherder.
Assignee: nobody → jolesen
Assignee | ||
Comment 2•9 years ago
|
||
When inserting instructions that encode a pc-relative offset, don't use a method that depends on getting a pointer to the newly inserted method. Use the new nextinstrOffset() method when computing the encoding of the pc-relative offset, and only insert each instruction once. Propagate OOM from calls to buffer.allocEntry, folowing the approach in the ARM assembler.
Attachment #8665505 -
Flags: review?(lhansen)
Reporter | ||
Updated•9 years ago
|
Attachment #8665505 -
Flags: review?(lhansen) → review+
Assignee | ||
Updated•9 years ago
|
Keywords: checkin-needed
Comment 3•9 years ago
|
||
hi, can we get a try run here ?
Flags: needinfo?(jolesen)
Keywords: checkin-needed
Assignee | ||
Comment 4•9 years ago
|
||
https://treeherder.mozilla.org/#/jobs?repo=try&revision=5771c7d2e7e1
Flags: needinfo?(jolesen)
Keywords: checkin-needed
Comment 6•9 years ago
|
||
https://hg.mozilla.org/mozilla-central/rev/0f9b612eebb1
Status: NEW → RESOLVED
Closed: 9 years ago
status-firefox44:
--- → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla44
You need to log in
before you can comment on or make changes to this bug.
Description
•