The tests are now using mutation observer to check
when the dom is updated instead.
Differential Revision: https://phabricator.services.mozilla.com/D57499
--HG--
extra : moz-landing-system : lando
CLOSED TREE
Backed out changeset 27cec4c76272 (bug 1601688)
Backed out changeset 9aa05612f10f (bug 1601688)
--HG--
extra : rebase_source : 8be2ada196777716cf7a1a47112033b8d1afd481
The tests are now using mutation observer to check
when the dom is updated instead.
Differential Revision: https://phabricator.services.mozilla.com/D57499
--HG--
extra : moz-landing-system : lando
* talos-powers-content.js has to use `wantUntrusted` to receive untrusted
events from page scripts unless enablePrivilege is used.
* Removed `contentMarker` because it is unused and it depends on
enablePrivilege to call `Services.profiler.AddMarker`.
Differential Revision: https://phabricator.services.mozilla.com/D57240
--HG--
extra : moz-landing-system : lando
While I'm here, I sorted the manifest, in order to match the sorted PGO list, so that it's easier to see when there's a discrepancy.
Differential Revision: https://phabricator.services.mozilla.com/D58968
--HG--
extra : moz-landing-system : lando
In brief, this just includes two out-of-process iframes in the
tresize test page, so that when we are resizing the main window,
we will also be resizing in each of the other fission processes.
The current tresize test is registered at a chrome:// URL, so I
had to make it an http URL to have OOP http iframes, which meant
I had to adjust a bit of the framescriptery to account for this.
Differential Revision: https://phabricator.services.mozilla.com/D55572
--HG--
extra : moz-landing-system : lando
Depends on D49569
We can stop using the exact host:port URL and rely on the proxy for all DAMP calls
Differential Revision: https://phabricator.services.mozilla.com/D53317
--HG--
extra : moz-landing-system : lando
This adds a tp5n-multiproc output directory which contains postprocessed
versions of the tp5n set. They are processed to take iframes with relative
URLs and output absolute URLs with domains similar to their original
domains. This will help us better simulate a real-world case for Fission,
which will split processes based on the ETLD+1 of the domain.
Differential Revision: https://phabricator.services.mozilla.com/D53065
--HG--
extra : moz-landing-system : lando
This will cause us to actually launch a new content process, which
we were previously not doing.
Differential Revision: https://phabricator.services.mozilla.com/D50708
--HG--
extra : moz-landing-system : lando
This will simply allow tests to use alternate domains if they want to ensure
they're getting different processes. This is important for realisitcally
measuring the impact of Fission.
Differential Revision: https://phabricator.services.mozilla.com/D50707
--HG--
extra : moz-landing-system : lando
nsXULWindow is no longer XUL specific and is somewhat confusing name.
Differential Revision: https://phabricator.services.mozilla.com/D51486
--HG--
rename : xpfe/appshell/nsXULWindow.cpp => xpfe/appshell/AppWindow.cpp
rename : xpfe/appshell/nsXULWindow.h => xpfe/appshell/AppWindow.h
rename : xpfe/appshell/nsIXULWindow.idl => xpfe/appshell/nsIAppWindow.idl
extra : moz-landing-system : lando
nsXULWindow is no longer XUL specific and is somewhat confusing name.
Differential Revision: https://phabricator.services.mozilla.com/D51486
--HG--
rename : xpfe/appshell/nsXULWindow.cpp => xpfe/appshell/AppWindow.cpp
rename : xpfe/appshell/nsXULWindow.h => xpfe/appshell/AppWindow.h
rename : xpfe/appshell/nsIXULWindow.idl => xpfe/appshell/nsIAppWindow.idl
extra : moz-landing-system : lando
nsXULWindow is no longer XUL specific and is somewhat confusing name.
Differential Revision: https://phabricator.services.mozilla.com/D51486
--HG--
rename : xpfe/appshell/nsXULWindow.cpp => xpfe/appshell/AppWindow.cpp
rename : xpfe/appshell/nsXULWindow.h => xpfe/appshell/AppWindow.h
rename : xpfe/appshell/nsIXULWindow.idl => xpfe/appshell/nsIAppWindow.idl
extra : moz-landing-system : lando
The codesize impact is approximately 0 (actually positive by a a few KB),
because AssignJSString is not inlined.
If we were to force-inline AssignJSString, that would actually _save_ ~115KB
based on bloaty measurements for linux64 shippable opt builds (so with LTO),
because then ConvertJSValueToString stops being inlined in bindings.
Given that ConvertJSValueToString pretty much always calls AssignJSString, so
there is no slowdown from changing which of them is inlined, this is _really_
tempting to do, but I don't want to mess with compiler inlining heuristics...
Inlining both AssignJSString and ConvertJSValueToString increases codesize by
~2MB, so that's definitely _not_ desirable.
Differential Revision: https://phabricator.services.mozilla.com/D50627
--HG--
extra : moz-landing-system : lando
This is the smallest change that helps having a clean baseline when we start using "Inspect Element" in the inspector custom test
Differential Revision: https://phabricator.services.mozilla.com/D50589
--HG--
extra : moz-landing-system : lando
This is needed to release a new mozlog with the PRECONDITION_FAILED
test and subtest status for use in web-platform-tests.
Update all in-tree dependencies on mozlog to >=5.0. These were found
with `hg grep 'mozlog.*[0-9]'`.
Only testing/web-platform/tests/tools/wptrunner/requirements.txt
remains on 4.2.0, and it will be updated in upstream wpt after mozlog
5.0 has been released.
Differential Revision: https://phabricator.services.mozilla.com/D50456
--HG--
extra : moz-landing-system : lando
The current limit of 2 million entries only allows
to record part of a DAMP run. The profiler team will investigate
if the number of entries can/should be reduced, but for the time
being we bump the limit to 10 million in order to record full
profiles when running DAMP on try.
Differential Revision: https://phabricator.services.mozilla.com/D49555
--HG--
extra : moz-landing-system : lando