Safe mode doesn't need to write out an invalid compatibility version
Categories
(Toolkit :: Startup and Profile System, task)
Tracking
()
| Tracking | Status | |
|---|---|---|
| firefox69 | --- | fixed |
People
(Reporter: mossop, Assigned: mossop)
Details
Attachments
(1 file)
|
47 bytes,
text/x-phabricator-request
|
jcristau
:
approval-mozilla-beta-
|
Details | Review |
When running safe mode we write out a compatibility.ini that contains "Safe Mode" as the compatibility version in order to force the next startup to flush the caches. This is unnecessary since we already have a mechanism for telling the next startup to flush the caches.
| Assignee | ||
Comment 1•6 years ago
|
||
When launching into safe mode we write out a compatibility.ini with an invalid
compatibility version in it. This is to ensure that on the next startup we flush
the caches. This is unnecessary though, we already have a flag we can set in
compatibility.ini in order to request a flush on next startup. This patch
switches to doing that so the version check on startup continues to work.
Comment 2•6 years ago
|
||
Comment 5•6 years ago
|
||
Backed out changeset fc21b0bcbf12 (bug 1556831) for multiple reftest failures on a CLOSED TREE.
Backout link: https://hg.mozilla.org/integration/mozilla-inbound/rev/f4fe0d9e5130598febfb2ce89c96692e6d08efeb
Push with failures: https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&resultStatus=testfailed%2Cbusted%2Cexception&revision=fc21b0bcbf127471f27c92a10554ba27eba01d95&selectedJob=250441466
Log snippet:
19:14:57 INFO - REFTEST SUITE-START | Running 3761 tests
19:14:57 INFO - REFTEST TEST-START | data:text/html,
19:14:57 INFO - REFTEST TEST-LOAD | data:text/html, | 0 / 3761 (0%)
19:14:57 INFO - [Child 760, Main Thread] WARNING: No active window: file z:/build/build/src/js/xpconnect/src/XPCJSContext.cpp, line 662
19:14:57 INFO - ++DOMWINDOW == 7 (08ABB400) [pid = 760] [serial = 7] [outer = 012734C0]
19:14:57 INFO - [Parent 3856, StreamTrans #5] WARNING: 'NS_FAILED(rv)', file z:/build/build/src/modules/libjar/nsJARChannel.cpp, line 371
19:14:57 INFO - [Parent 3856, Main Thread] WARNING: NS_ENSURE_SUCCESS(rv, rv) failed with result 0x80520012: file z:/build/build/src/modules/libjar/nsJARChannel.cpp, line 994
19:14:57 INFO - REFTEST TEST-PASS | data:text/html, | (LOAD ONLY)
19:14:57 INFO - REFTEST TEST-END | data:text/html,
19:14:57 INFO - ++DOMWINDOW == 8 (08ABA400) [pid = 760] [serial = 8] [outer = 012734C0]
19:14:57 INFO - REFTEST TEST-UNEXPECTED-FAIL | data:text/html, | assertion count 1 is more than expected 0 assertions
| Assignee | ||
Updated•6 years ago
|
Comment 7•6 years ago
|
||
| bugherder | ||
| Assignee | ||
Comment 8•6 years ago
|
||
Comment on attachment 9069767 [details]
Bug 1556831: Have safe mode write out the normal compatibility version. r=froydnj
Beta/Release Uplift Approval Request
- User impact if declined: We may miss some profile downgrade cases if a newer profile is launched immediately after safe mode.
- Is this code covered by automated tests?: Yes
- Has the fix been verified in Nightly?: Yes
- Needs manual test from QE?: No
- If yes, steps to reproduce:
- List of other uplifts needed: None
- Risk to taking this patch: Low
- Why is the change risky/not risky? (and alternatives if risky): This essentially just changes the mechanism by which we invalidate the startup caches.
- String changes made/needed: None
Comment 9•6 years ago
|
||
Comment on attachment 9069767 [details]
Bug 1556831: Have safe mode write out the normal compatibility version. r=froydnj
I think I'd rather let this ride the trains. Missing a downgrade case is not the end of the world, it's not something we would have detected before anyway.
Description
•