As long as Directory objects can be constructed in the content process, trying
to access the file system conceptually does not make sense because of
sandboxing. The canary in the coal mine for this was tests which started
crashing on Directory construction as we further locked down read access in the
sandbox.
MozReview-Commit-ID: HZitALYpO5H
--HG--
extra : rebase_source : 1f7e0361e40e45a9c23d5d78bdcdc1696d44d298
If we don't increment the plugin epoch in the parent process before
the first call to LoadPlugins, and all plugin info is already cached,
the epochs will match and we'll never get a complete plugin list in
the child. This patch makes sure our first parent/child epoch check
never matches, so we always update correctly.
MozReview-Commit-ID: Jj1F8m3aeGx
Because the nsVideoFrame's creation is not synchronous to "append to dom tree", there is a small gap
that the TextTrackManager wants to render cue but no frame. Change the testcase to not hit the period.
MozReview-Commit-ID: 9xhMjRJnoDR
We use gpu for yuv color conversion now. There are some precision problems in gpu path. Mark them fails-if and update the fuzzy-test value.
MozReview-Commit-ID: 6tunOCj3OHB
ProfilerStackFrameRAII and ProfilerStackFrameDynamicRAII are very similar; the
latter lets a dynamic string be specified as well (and lacks the
MOZ_GUARD_OBJECT stuff, for no good reason).
This patch does the following.
- Removes ProfilerStackFrameDynamicRAII, and adds a dynamic string to
ProfilerStackFrameRAII. It also reorders the constructor's arguments to match
the field ordering of ProfileEntry. There aren't many usage sites so these
changes don't affect many places.
- With that done, there is only a single callsite for each of
profiler_call_enter() and profiler_call_exit(), so the patch also inlines and
removes them.
Bug 1286530 made TCPSocket ChromeOnly, eliminating both the
dom.mozTCPSocket.enabled preference check as well as the "tcp-socket"
permission check. The API is now always exposed in chrome contexts.
This patch removes the leftover (and confusing) dead code.
Renamed and why:
- test_tcpsocket_enabled_no_perm.html renamed to
test_tcpsocket_not_exposed_to_content.html because it's now just a
question of content never seeing the API.
Removed tests and why:
- test_tcpsocket_enabled_with_perm.html used to be a standalone
verification of our permission check. We have no permission check
now and both test_tcpsocket_jsm.html and
test_tcpsocket_client_and_server_basics.html serve as tests that we
affirmatively expose the API and there are no "late" failure (such
as a secondary check in the parent).
- test_tcpsocket_default_permissions.html duplicated what (now)
test_tcpsocket_not_exposed_to_content.html accomplishes. It tried to
use the API and expect an exception. This is just superstition in a
WebIDL. (TCPSocket was not originally WebIDL-y.)
--HG--
rename : dom/network/tests/test_tcpsocket_enabled_no_perm.html => dom/network/tests/test_tcpsocket_not_exposed_to_content.html
extra : rebase_source : d2231ed3b0fb00541cc266569c2a7908a4074f9c
The use of add_task.js was ever only intended to be temporary until
bug 1078657 landed.
--HG--
extra : rebase_source : d9570859b35691002cf7f4642603f6730ddab7ad
This hint will inform readers if caching is discouraged (e.g., for
SourceBufferResource) or recommmended (e.g., for MediaCache-backed
ChannelMediaResource).
MozReview-Commit-ID: 7hopNS0s5tE
--HG--
extra : rebase_source : 681646cc904229e8513adb148baa085254508049
This permits NSS to load libsoftokn3.dylib, even when the rest of the file
system's access is removed (as is the goal for the content sandbox level 3).
This is needed for WebCrypto.
MozReview-Commit-ID: Bh54b87zIjY
--HG--
extra : rebase_source : e1fa59648d683e97a3bc1310ac1c362009f657f8
