Commit Graph

13157 Commits

Author SHA1 Message Date
Mason Chang
5bd5e81b3d Bug 1125273 - Disable CompositorVsyncObserver destructor assertion. r=benwa 2015-01-28 10:36:04 -08:00
Ryan VanderMeulen
ee1dd2fec5 Backed out changeset a501183528f3 (bug 1124847) for Win7 crashes (again). r=tryserverkthxbye
CLOSED TREE
2015-01-28 10:58:55 -05:00
Jeff Muizelaar
9822c711da Bug 1124847. Track D3D11 shared texture usage in about:memory. r=bas
This will hopefully help us track down large amounts of write-combined
mappings that we see.
2015-01-27 20:09:34 -05:00
Jonathan Kew
c1c55a435c Bug 1108177 - Implement harfbuzz glyph-extents callback, so that fallback mark positioning works in legacy truetype fonts. r=jdaggett 2015-01-28 09:03:28 +00:00
Benoit Girard
ce118bc3bc Bug 1121871 - Properly compute the invalid region when painting using the tiling fast path. r=nical 2015-01-27 16:24:11 -05:00
Nicholas Nethercote
cf9236371a Bug 1124973 (part 5) - Use PL_DHashTableSearch() in gfxFT2FontList.cpp. r=froydnj,jkew.
Because PL_DHashTableLookup() never returns null, GetInfoForFile() features not
one but *two* can-never-fail null checks on its result.

Having said that, the code as written works, at least for non-zero-sized files,
because |entry->mFileSize| will always be zero if the lookup fails (thanks to
PLDHashTable always being zeroed at construction, and |mMap| using
PL_DHashClearEntryStub which also zeroes).

But for zero-sized files the current code will act like they don't exist. Maybe
this can't happen in practice, but it seems dangerous and so I've changed it so
the new code will treat zero-sized files just like non-zero-sized files.

--HG--
extra : rebase_source : c617862c23babb1726480faff771a8dfe586bfa1
2015-01-26 14:25:13 -08:00
Ryan VanderMeulen
6b143430fb Backed out changeset 1c324316ea7a (bug 1124847) for Win7 crashes.
CLOSED TREE
2015-01-27 21:59:52 -05:00
Wes Kocher
50351b0c8d Backed out changeset 774ff21aecb3 (bug 1120128) for build bustage 2015-01-27 17:48:43 -08:00
Seth Fowler
fe165433d0 Bug 1126146 - Disable the single-color optimization in reftests. r=tn 2015-01-27 17:43:21 -08:00
David Anderson
bf9db3498b Don't require paint resampling for desktop APZ, yet. (bug 1126045 part 1, r=mattwoodrow)
--HG--
extra : rebase_source : c942d32bc9570d1ba9375e9c8289dc7aefceaea3
2015-01-27 16:01:48 -08:00
Matt Woodrow
4f89591f0d Bug 1120128 - Implement blacklist for DXVA and blacklist AMD Radeon HD 5800. r=Bas
--HG--
extra : rebase_source : 782dc73a9f284c7a0d98061f2fe2dfb87e3bb149
2015-01-26 13:34:28 +13:00
Jeff Muizelaar
1baaab90c9 Bug 1124847. Track D3D11 shared texture usage in about:memory. r=bas
This will hopefully help us track down large amounts of write-combined
mappings that we see.

--HG--
extra : rebase_source : fbb5356fd81de3fa484639b955c3c4f806ce4f45
2015-01-27 20:09:34 -05:00
Bas Schouten
e7a45f235f Bug 1089454: Prevent usage of incompatible graphics objects after device reset. r=jrmuizel 2015-01-28 00:54:19 +00:00
Daniel Holbert
16d66373ef Bug 1118419 followup: Annotate macro-defined typeName functions as MOZ_OVERRIDE for remaining GLContext subclasses. 2015-01-27 11:45:57 -08:00
Sotaro Ikeda
46266124b8 Bug 1125913 - Use GraphicBuffer::lockAsync() on gonk LL r=nical 2015-01-27 10:05:35 -08:00
Jeff Muizelaar
3553ebb2b3 Bug 1125709. Avoiding using MacIOSurfaceTextureSourceOGL in MacIOSurfaceTextureHostOGL. r=nical
Using GLTextureSource gives us better performance because we don't need to call
CGLTexImageIOSurface2D everytime we draw.

This will eventually let us drop MacIOSurfaceTextureSourceOGL once WebGL on OS
X is using MacIOSurfaceTextureHostOGL.

--HG--
extra : rebase_source : 95a9b0888d7ddfeae520141e7f932f8555fa58d5
2015-01-19 21:23:42 -05:00
Ehsan Akhgari
34ebbbebeb Bug 1117034 - Mark some overridden functions in the tree as override
--HG--
extra : rebase_source : a37ba6b69c02db90b8f3c440ac2dbca4d80c8bc5
2015-01-26 21:09:59 -05:00
Wes Kocher
62df818443 Backed out changeset 783d0fe2770b (bug 1124394) for test bustage on a CLOSED TREE 2015-01-26 18:57:36 -08:00
Ehsan Akhgari
b0ff6c50c9 Backed out changeset 45921e3d9773 (bug 1117034) because of build bustage on a CLOSED TREE 2015-01-26 21:52:40 -05:00
Ehsan Akhgari
cacfbf0a61 Bug 1117034 - Mark some overridden functions in the tree as override 2015-01-26 21:14:12 -05:00
Jeff Gilbert
b17f3529ee Bug 1124394 - Allow Core Profile GL contexts. - r=kamidphish 2015-01-26 17:18:45 -08:00
Ethan Lin
2bbea6c40e Bug 753010 - Part 2 - Refactor some neon intrinsic functions for blur. r=mstange 2015-01-22 18:57:00 +01:00
Ethan Lin
af7d85c741 Bug 753010 - Add a NEON optimized blur method. r=mstange 2015-01-18 22:08:00 +01:00
Masatoshi Kimura
0676786e20 Bug 1111290 - Part 3: Remove TypedEnum.h and fold TypedEnumInternal.h into TypedEnumBits.h. r=waldo 2015-01-26 07:22:11 +09:00
Masatoshi Kimura
ca6988b0d9 Bug 1111290 - Part 2: Non-mechanical changes. r=waldo 2015-01-26 07:22:08 +09:00
Masatoshi Kimura
2490dcfe04 Bug 1111290 - Part 1: Remove MOZ_(BEGIN|END)_ENUM_CLASS. r=waldo 2015-01-26 07:22:07 +09:00
Markus Stange
8780bcc049 Bug 1117304 - Make sure the tile filter doesn't call CopyRect on surfaces with different formats. r=Bas 2015-01-05 18:40:27 +01:00
Markus Stange
b7823af622 Bug 1117304 - Also do the checks at the start of CopyRect in release builds. r=Bas 2015-01-06 12:08:39 +01:00
Ryan VanderMeulen
8f41d89831 Bug 1042103 - Fix non-unified bustage. r=botond 2015-01-23 17:51:34 -05:00
Kartikaya Gupta
7fbb5b933d Bug 1124099 - Make sure that we skip over scrollinfo layers properly with event-regions disabled. r=botond 2015-01-23 15:12:10 -05:00
Mason Chang
05082a3276 Bug 1098701. Part 1: gtest framework and tests. r=kats 2015-01-23 11:19:54 -08:00
Nicolas Silva
c964f3cf52 Bug 1119019 - Avoid destroying a SharedSurface before its TextureClient/Host pair. r=sotaro 2015-01-23 15:35:16 +01:00
Nicholas Nethercote
3519d0640a Bug 1123527 - Fix syntax error when NS_COORD_IS_FLOAT is defined. r=dholbert.
cppcheck found this.

--HG--
extra : rebase_source : 6a83556a8baaa24546eeb28d35ad63f4ffd275bd
2015-01-21 19:23:28 -08:00
Bas Schouten
bf7969ed0e Bug 1097699 - Part 4: Remove LayerManagerD3D10 code. r=jrmuizel 2015-01-23 03:41:21 +00:00
Bas Schouten
dea47282db Bug 1097699 - Part 2: Remove LayerManagerD3D9 code. r=jrmuizel 2015-01-23 03:41:20 +00:00
Ryan VanderMeulen
cff23a3b2b Backed out changeset 8b3815de49f8 (bug 1123654) for talos xperf permafail.
CLOSED TREE
2015-01-22 20:00:24 -05:00
Nicholas Nethercote
7d9cc5ecf2 Bug 1124545 - Avoid creating the mergedFeatures hash table when possible. r=jdaggett.
When scrolling all the way through
https://bugzilla.mozilla.org/show_bug.cgi?id=MNG, |mergedFeatures| gets
initialized more than 50,000 times -- each one involving a heap allocation for
the entry storage -- but it never has anything put into it.

This patch refactors MergeFontFeatures to move |mergedFeatures| inside it. It
now doesn't get created in the common case.

--HG--
extra : rebase_source : 5a8d44b30f90169b70e5f50a5e38bbe57f02b12b
2015-01-21 22:41:02 -08:00
Jonathan Kew
1308275257 Bug 1123654 - Replace use of [deprecated] GetStockObject(DEFAULT_GUI_FONT) with newer API; results in use of Tahoma in place of Microsoft Sans Serif in various contexts. r=jimm 2015-01-22 22:52:26 +00:00
Nathan Froyd
d5e4654494 Bug 1122235 - set HAVE_UINT64_T for cairo compilation to improve codegen; r=mshal 2015-01-15 16:17:55 -05:00
Nathan Froyd
838298aaff Bug 967300 - enable cairo's atomic support on gcc-esque compilers; r=mshal
Cairo has a number of nifty features predicated on support for atomic
operations on integer types.  Normally, such support would be determined
by cairo's configure script.  But since we don't run cairo's configure
script, we need to manually define HAVE_INTEL_ATOMIC_PRIMITIVES during
cairo's build.  That macro enables codepaths that depend on certain
SIZEOF_* variables being defined by autoconf, so we also need to add the
necessary code in moz.build to set those.
2015-01-15 15:53:28 -05:00
Milan Sreckovic
67354457c3 Bug 1088833 - A bit of a clean up of warnings, and catch bad draw target in the d3d11 canvas case. r=bschouten 2015-01-20 12:47:29 -05:00
Milan Sreckovic
a1bdbcfff9 Bug 1121835: Deal with non-finite values in Path calls. r=bschouten 2015-01-19 14:19:25 -05:00
Mason Chang
b3246a1f83 Bug 1122943 - Ensure the Compositor Vsync Observer Shutsdown correctly. r=benwa 2015-01-21 16:13:15 -08:00
Botond Ballo
3a8c56fc2d Bug 1096513 - Follow up to fix an incorrect assertion. r=Cwiiis
--HG--
extra : rebase_source : bab8417ba9133c0f8039251cbc22f3b6d19541fc
2015-01-21 12:35:49 -05:00
Bas Schouten
866f97145d Bug 1117925: Make sure to invalidate the entire window when the size changes. r=jrmuizel 2015-01-21 18:00:31 +01:00
Seth Fowler
b92cad3917 Bug 1124072 - Flip on decode-on-draw on APZ platforms. r=tn 2015-01-21 05:07:46 -08:00
Robert Longson
02c054d045 Bug 1122761 - Avoid dividing by zero when calculating turbulence gradients. r=mstange 2015-01-21 13:01:00 +00:00
peter chang
49c42b50f9 Bug 1113524 - Fix the crash at nsDeviceContext::FindScreen() because mScreenManager is nullptr. r=tn 2015-01-20 17:52:00 -05:00
Boris Chiou
afe6cee3de Bug 1079755 - Send composer info to Layerscope viewer. r=kamidphish
Add a composition info, composedByHwc, in META packet to
notify users whether this frame is composed by Hardware composer.

--HG--
extra : rebase_source : b3883178f41acb92fa91b0cab5905d22fcf38553
2015-01-16 15:00:25 +08:00
Matt Woodrow
4c9fdd62ba Bug 1112331 - Disable DXVA when using WARP. r=cpearce,Bas 2015-01-21 16:36:01 +13:00