b053dddf96
Bug 1046101 part.6 Rename NS_COMPOSITION_EVENT to eCompositionEventClass r=smaug
2014-08-04 14:28:48 +09:00
aaf3fac323
Bug 1046101 part.5 Rename NS_KEY_EVENT to eKeyboardEventClass r=smaug
2014-08-04 14:28:48 +09:00
1082b45ab4
Bug 1046101 part.4 Rename NS_UI_EVENT to eUIEventClass r=smaug
2014-08-04 14:28:47 +09:00
f34a1f1968
Bug 1046101 part.3 Rename NS_INPUT_EVENT to eInputEventClass r=smaug
2014-08-04 14:28:47 +09:00
3d2163899e
Bug 1046101 part.2 Rename NS_GUI_EVENT to eGUIEventClass r=smaug
2014-08-04 14:28:46 +09:00
00f048e7cd
Bug 1046101 part.1 Rename nsEventStructType to mozilla::EventClassID r=smaug
2014-08-04 14:28:46 +09:00
e72f8e97bc
Bug 1045096. Make sure initial about:blank in iframes have a nsDOMNavigationTiming object, so we don't end up with window.performance throwing when accessed on their window. r=smaug
2014-08-04 01:25:30 -04:00
48406d15bd
Bug 1047220 - Shrink the static atoms table. r=bz.
...
--HG--
extra : rebase_source : a7f8d54cce7f20d784371348f0d0cabdacf43c0a
2014-07-31 22:57:13 -07:00
965e3f8650
b=1045482 make XErrorTrap installation and removal thread-safe r=jesup
...
Error handling is now applied to the Display using async_handlers, instead of
replacing and trying to reinstate the XSetErrorHandler() global handler for
all Xlib Displays.
Inspired by use of the same Xlibint.h structures at
https://git.gnome.org/browse/gtk+/tree/gdk/x11/gdkasync.c?id=0e1a4248#n252
https://git.gnome.org/browse/gtk+/tree/gdk/x11/gdkasync.c?id=0e1a4248#n150
Compare use of _XAsyncErrorHandler in libX11.
--HG--
extra : rebase_source : 8bb16f09912b15e435441657e7275c9213b1a475
2014-08-01 17:58:24 +12:00
ddcaf3a3a7
b=1042508 enable neon resampler optimizations r=padenot
...
--HG--
extra : rebase_source : fd984143b78ecf25a5158715dd8bed8df7398945
2014-07-22 16:51:08 +12:00
ee8186f60a
b=1042508 remove : of empty Clobber asm argument r=padenot
...
https://gcc.gnu.org/onlinedocs/gcc/Extended-Asm.html#Extended-Asm says
asm [volatile] ( AssemblerTemplate : [OutputOperands] [ : [InputOperands] [ : [Clobbers] ] ] )
which implies that Clobbers is optional even after the third colon, but
the gcc used for b2g_try_emulator_dep builds says
resample_neon.c: In function 'saturate_32bit_to_16bit':
resample_neon.c:50: error: expected string literal before ')' token
--HG--
extra : rebase_source : 6ca7768d5b30f67ff6827acc9a4fdc51a672bc1e
2014-07-22 20:10:36 +12:00
5a1e783dae
b=1042508 move resampler simd optimizations to separate translation units r=padenot
...
This ensures that code in resample.c will run on Intel x86 cpus even when SSE
support has been compiled, and will provide similarly for neon support when
enabled.
--HG--
rename : media/libspeex_resampler/sse-detect-runtime.patch => media/libspeex_resampler/simd-detect-runtime.patch
rename : media/libspeex_resampler/src/resample_neon.h => media/libspeex_resampler/src/resample_neon.c
rename : media/libspeex_resampler/src/resample_sse.h => media/libspeex_resampler/src/resample_sse.c
rename : media/libspeex_resampler/src/sse_detect.cpp => media/libspeex_resampler/src/simd_detect.cpp
rename : media/libspeex_resampler/src/sse_detect.h => media/libspeex_resampler/src/simd_detect.h
extra : rebase_source : bdf1ef791129dedeadce7369354a5992729a99b7
2014-07-23 21:49:04 +12:00
75884b1238
b=1042508 tie resampler sample type choice to MOZ_SAMPLE_TYPE_S16 r=padenot
...
When built with FIXED_POINT speex_resampler_process_float() assumes the
samples are in the range ±0x7fff and so Gecko instead uses ConvertAudioSamples
and speex_resampler_process_int(). This decision is based on
MOZ_SAMPLE_TYPE_S16.
--HG--
extra : rebase_source : 1e67861478026dfa65b61a31e3272970a7094094
2014-07-22 15:41:50 +12:00
cc0c4ee1b1
b=1042508 include resample_neon.h from speexdsp 305e54ea r=padenot
...
--HG--
extra : rebase_source : effe9aa434298b0b49099b6809d506b92075ad1c
2014-07-21 17:08:00 +12:00
a8370d640a
Bug 1037328 Enable TSF mode in default settings of Nightly builds r=jimm+emk+m_kato
2014-08-04 10:31:25 +09:00
0efc1afca6
Fixup bug 1043268 because of b2g build failures after the landing of bug 1047791. r=me
2014-08-04 10:30:21 +09:00
d6833db2a9
Bug 1046529 (part 2) - Ensure static atoms always end up in gStaticAtomTable. r=bz,hsivonen.
2014-07-30 23:22:42 -07:00
3a306a87c3
Bug 1046529 (part 1) - Split and simplify AtomTableKey's constructors. r=bz.
2014-07-30 23:22:40 -07:00
4d34febc3f
Bug 901208 - Fix Skia for ARM v4t. r=derf
2014-08-04 09:04:06 +09:00
3107b612d5
Bug 1047791 - Use integers for audio when on Android, or when using ARM on other OSes, and disable webm encoding. r=padenot
2014-08-04 09:03:59 +09:00
408bc9dae9
Bug 1046644 - Honor OpenH264 autoupdate pref for the initial GMP install check. r=gfritzsche
2014-08-01 16:46:45 -04:00
854041b4fe
Bug 1046549 - Ignore mdat unless we've already found a media track. r=kentuckyfriedtakahe
2014-07-31 18:54:12 +12:00
90d4d003d2
Bug 996796 patch 22 - Make restyling exact - only call RestyleBeforePseudo and RestyleAfterPseudo when we have an eRestyle_Subtree hint. r=heycam
2014-08-03 13:11:55 -07:00
638868ad96
Bug 996796 patch 21 - Make restyling exact - Only call MaybeRecreateFramesForElement when we have a nonzero restyle hint. r=heycam
2014-08-03 13:11:55 -07:00
9c23ec99b9
Bug 996796 patch 20 - Make restyling exact - Avoid rerunning selector matching on everything when the basis of rem units changes. r=heycam
2014-08-03 13:11:55 -07:00
1f200c6e83
Bug 996796 patch 19 - Avoid reresolution for extra style contexts when the restyle hint is not eRestyle_Self or eRestyle_Subtree. r=heycam
...
This is part of making reresolution not touch style that it was not
asked to touch, so that we can use it for the animation-only restyle.
2014-08-03 13:11:55 -07:00
f7374aad6e
Bug 996796 patch 18 - Fix RestyleTracker to handle restyle hints exactly rather than pessimistically when restyling continuations with varying styles (e.g., spans inside ::first-line or ::first-letter). r=heycam
...
This will be necessary when we use the restyle tracker for the
animation-only style flush, because animation-only style flushes need to
update *only* the animation style data and no other style data. Thus
using the RestyleTracker for animation-only style flushes requires that
we do this.
2014-08-03 13:11:55 -07:00
0ccaf0860c
Bug 1043041: Use mozilla::pkix::Time instead of PRTime, r=keeler
...
--HG--
extra : rebase_source : 2cc39d3c322c1355aad003f2497659a091febac2
2014-08-02 08:49:12 -07:00
de725ae5ef
Bug 1047792: Rely on mozilla::pkix to filter out expired certs instead of CERT_CreateSubjectCertList, r=keeler
...
--HG--
extra : rebase_source : 5182147037b69f0ac3c3cd060d6e2af71bfde2e7
2014-08-01 23:16:21 -07:00
25a73829af
Bug 1041344: Refactor mozilla::pkix::CheckCertificatePolicies, r=cviecco
...
--HG--
extra : rebase_source : d40184b986e9c6ed44c0b39a485292a91f924f13
2014-07-19 18:51:10 -07:00
02778ff4f1
Bug 1048017: pref on WebRTC screensharing - whitelist blocks access by default rs=jesup
2014-08-03 11:30:06 -04:00
9564759214
Bug 1031407 - DataStoreService doesn't need to use IDBCursor to know if a revision exists., r=janv
2014-08-03 15:48:19 +01:00
9a69dcaf2c
Bug 1047778 - Fix some bad implicit constructors in xpcprivate.h; r=bholley
2014-08-03 10:40:44 -04:00
fac69fce74
Bug 1046582 - mochitest for AudioChannel changes in MediaStream objects, r=ehsan, r=bz
2014-08-03 14:46:17 +01:00
a39da5881c
Back out bug 996796 patch 18 through patch 25 (changesets fbe97c2db729 through 9719c08c3144) to fix 50%-of-the-time Android 4.0 debug orange from single assertion ("Why did this not get handled while processing mRestyleRoots?", layout/base/RestyleTracker.cpp, line 87) in gfx/tests/crashtests/815489.html .
2014-08-03 01:22:07 -07:00
fe9339c972
Bug 1047824 - Make Apple PDM handle Shutdown() call on non-main thread. r=rillian
2014-08-03 19:29:19 +12:00
4b33c80b1c
Bug 1047824 - Call PlatformDecoderModule::Shutdown() in MP4Reader::Shutdown() on the decode task queue. r=kinetik
2014-08-03 19:28:39 +12:00
9991c6dfee
Bug 1046416: Shut down gfxPlatformFontList on all platforms. r=jfkthame
2014-08-01 12:37:53 -07:00
8cae71e7f0
Bug 1045161 patch 2 - Remove write-only member variables of nsStyleTable. r=bzbarsky
2014-08-02 19:37:49 -07:00
979a8abe0d
Bug 1045161 patch 1 - Stop implementing ParseAttribute for frame and rules (to parse them to enumerated values) since we no longer map them into style. r=bzbarsky
...
I don't think this change should be observable to Web content anymore.
(If it is, we should probably fix that as a general issue.)
2014-08-02 19:37:49 -07:00
0d56159446
Bug 996796 patch 25 - Remove the old (now-unused) miniflush code (preserving one of the header comments). r=heycam
2014-08-02 19:37:49 -07:00
3fc788efdb
Bug 996796 patch 24 - Use a RestyleTracker for the coalescing in the animation-only style flush (miniflush). r=heycam
...
This changes the coalescing behavior during the animation-only style
flush by doing tree-based coalescing between the style updates required
by animations and those required by transitions, rather than doing
animations and transitions separately.
Note that both the old and the new code update all
animating/transitioning styles rather than only the throttled styles,
though we should fix that eventually as noted in the FIXME comment in
the code (but only for the existing caller, and not for the new one to
be introduced in bug 960465).
Note that this depends for its correctness on the previous patches to
make the restyling process exact.
The test changes are because the effects of bug 1031688 are changed by
the change in coalescing. In the old code, we updated transition styles
before animation styles; in the new code we do a single pass over the
tree, which for the relevant test means updating animations on the
parent before transitions on the child, which changes the effects of the
bug.
2014-08-02 19:37:49 -07:00
986f645557
Bug 996796 patch 23 - Make transition manager ignore StyleContextChanged notifications during an animation-only style update. r=heycam
...
This will be necessary when we use the restyle tracker for the
animation-only style flush.
2014-08-02 19:37:48 -07:00
e7c9303e63
Bug 996796 patch 22 - Make restyling exact - only call RestyleBeforePseudo and RestyleAfterPseudo when we have an eRestyle_Subtree hint. r=heycam
2014-08-02 19:37:48 -07:00
8aa0a176e2
Bug 996796 patch 21 - Make restyling exact - Only call MaybeRecreateFramesForElement when we have a nonzero restyle hint. r=heycam
2014-08-02 19:37:48 -07:00
9babce67c7
Bug 996796 patch 20 - Make restyling exact - Avoid rerunning selector matching on everything when the basis of rem units changes. r=heycam
2014-08-02 19:37:47 -07:00
4ff105698b
Bug 996796 patch 19 - Avoid reresolution for extra style contexts when the restyle hint is not eRestyle_Self or eRestyle_Subtree. r=heycam
...
This is part of making reresolution not touch style that it was not
asked to touch, so that we can use it for the animation-only restyle.
2014-08-02 19:37:47 -07:00
d2fe2e8c79
Bug 996796 patch 18 - Fix RestyleTracker to handle restyle hints exactly rather than pessimistically when restyling continuations with varying styles (e.g., spans inside ::first-line or ::first-letter). r=heycam
...
This will be necessary when we use the restyle tracker for the
animation-only style flush, because animation-only style flushes need to
update *only* the animation style data and no other style data. Thus
using the RestyleTracker for animation-only style flushes requires that
we do this.
2014-08-02 19:37:47 -07:00
242b2eae80
Bug 996796 patch 17 - Add comment about how RuleNodeWithReplacement should interact with nsTransitionManager and nsAnimationManager. r=heycam
2014-08-02 19:37:46 -07:00
a313113d7a
Bug 996796 patch 16 - Add comment about potential performance impromevent to RuleNodeWithReplacement. r=heycam
2014-08-02 19:37:46 -07:00
Masayuki Nakano
Boris Zbarsky
Nicholas Nethercote
Karl Tomlinson
Mike Hommey
Brian R. Bondy
Jean-Yves Avenard
L. David Baron
Brian Smith
Randell Jesup
Andrea Marchesini
Ehsan Akhgari
Chris Pearce
Kyle Huey