Magic-Mirror/gecko-dev
1
0
Fork 0
mirror of https://github.com/mozilla/gecko-dev.git synced 2024-11-08 20:47:44 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
372,887 Commits 615 Branches 98 Tags 5.9 GiB
8028bbe23b
Commit Graph

178 Commits

Author SHA1 Message Date
Robert O'Callahan
d49884afea Bug 1022612. Part 31: Perform layer-level occlusion culling in FrameLayerBuilder. r=mattwoodrow 
We need this to avoid constructing and painting unncecessarily large
ThebesLayers.

--HG--
extra : rebase_source : 22b1fbbd8b561bd68831ac1fb957d2d1901f3297
 2014-06-23 16:24:00 +12:00
Robert O'Callahan
f454ea8959 Bug 1022612. Part 10: Implement merging and flattening in ProcessDisplayItems. r=mattwoodrow 
BuildContainerLayerFor now has to be able to mutate the passed-in display item
list.

--HG--
extra : rebase_source : a3cf9106703d4946a49a73ef60a605f72b2afee3
 2014-06-09 16:48:00 +12:00
Ryan VanderMeulen
df23fd9dde Backed out 46 changesets (bug 1022612) for B2G mochitest permafails on a CLOSED TREE. 
Backed out changeset 34b3014a3112 (bug 1022612)
Backed out changeset 6ae9316fd909 (bug 1022612)
Backed out changeset b8f3749c95eb (bug 1022612)
Backed out changeset caab10bf6ca3 (bug 1022612)
Backed out changeset 0c57c620c898 (bug 1022612)
Backed out changeset fac64141a00a (bug 1022612)
Backed out changeset bf0df1c9d68b (bug 1022612)
Backed out changeset b42054800020 (bug 1022612)
Backed out changeset 667793b21194 (bug 1022612)
Backed out changeset f14ada64fe1b (bug 1022612)
Backed out changeset 75b837686bdf (bug 1022612)
Backed out changeset 66de53183a22 (bug 1022612)
Backed out changeset 0ff86ced4d46 (bug 1022612)
Backed out changeset 18eecc5b1ef7 (bug 1022612)
Backed out changeset 2763c4878de5 (bug 1022612)
Backed out changeset b72413ecc385 (bug 1022612)
Backed out changeset b23f1081afb8 (bug 1022612)
Backed out changeset f7e2c6a72043 (bug 1022612)
Backed out changeset 959917c9027d (bug 1022612)
Backed out changeset 0268a46f4880 (bug 1022612)
Backed out changeset 3388856a80ad (bug 1022612)
Backed out changeset e4b17cf0f806 (bug 1022612)
Backed out changeset 2f4e9da0e4b6 (bug 1022612)
Backed out changeset 489f6a7c0c03 (bug 1022612)
Backed out changeset 8369d9ad7ad3 (bug 1022612)
Backed out changeset 0758d2a06002 (bug 1022612)
Backed out changeset f2ae9cb22edb (bug 1022612)
Backed out changeset 9c48c6ee5dc2 (bug 1022612)
Backed out changeset fe7134400f08 (bug 1022612)
Backed out changeset cc2c5397ca8b (bug 1022612)
Backed out changeset a3d1a3e8b39d (bug 1022612)
Backed out changeset 8974b74b0eb0 (bug 1022612)
Backed out changeset 75f7dbb5a2a6 (bug 1022612)
Backed out changeset 2aa04a071e60 (bug 1022612)
Backed out changeset f2ab1bcd4c39 (bug 1022612)
Backed out changeset da9152b6ea29 (bug 1022612)
Backed out changeset 58abf5b0e148 (bug 1022612)
Backed out changeset 797058a09ad2 (bug 1022612)
Backed out changeset ea3e99a92ff0 (bug 1022612)
Backed out changeset adc4a4a7aa73 (bug 1022612)
Backed out changeset 7b18dedd1505 (bug 1022612)
Backed out changeset 055dd1921e8e (bug 1022612)
Backed out changeset 42fa2c97e989 (bug 1022612)
Backed out changeset cd594236388f (bug 1022612)
Backed out changeset 9eadc5fee43d (bug 1022612)
Backed out changeset 5cc8d30ff7c9 (bug 1022612)
 2014-07-17 11:24:47 -04:00
Robert O'Callahan
eff9ade093 Bug 1022612. Part 31: Perform layer-level occlusion culling in FrameLayerBuilder. r=mattwoodrow 
We need this to avoid constructing and painting unncecessarily large
ThebesLayers.

--HG--
extra : rebase_source : fd56072f7932fa9c39f314c4d95cdd0d1ad13de3
 2014-06-23 16:24:00 +12:00
Robert O'Callahan
d143dbdffc Bug 1022612. Part 10: Implement merging and flattening in ProcessDisplayItems. r=mattwoodrow 
BuildContainerLayerFor now has to be able to mutate the passed-in display item
list.

--HG--
extra : rebase_source : c242006445d90372f6c7941d909f30945422321b
 2014-06-09 16:48:00 +12:00
Carsten "Tomcat" Book
26a5361c4e Backed out changeset 3bf33a32bfe0 (bug 1022612) 2014-07-15 10:30:38 +02:00
Carsten "Tomcat" Book
d9ba3f2e23 Backed out changeset f8b488c389e9 (bug 1022612) 2014-07-15 10:30:04 +02:00
Robert O'Callahan
d43195c660 Bug 1022612. Part 31: Perform layer-level occlusion culling in FrameLayerBuilder. r=mattwoodrow 
We need this to avoid constructing and painting unncecessarily large
ThebesLayers.

--HG--
extra : rebase_source : 38cbd0b0dd7724cfdb98f3e215bd0f220a687f3c
 2014-06-23 16:24:00 +12:00
Robert O'Callahan
88afa39f1c Bug 1022612. Part 10: Implement merging and flattening in ProcessDisplayItems. r=mattwoodrow 
BuildContainerLayerFor now has to be able to mutate the passed-in display item
list.

--HG--
extra : rebase_source : 6a9727998a5ffb01896a4a2bbdd8d30a24c89dcd
 2014-06-09 16:48:00 +12:00
Benoit Girard
19effe56b7 Bug 1027496 - Stream display list and layers dump. r=mattwoodrow 2014-06-19 18:36:25 -04:00
Benoit Jacob
15b208d7f3 Bug 1027251 - Fix or whitelist dangerous public destructors in layout/ - r=dbaron 2014-06-18 22:28:59 -04:00
Kyle Huey
384121563d Bug 991812: Remove uses of RefCounted<T> that live in Gecko. r=ehsan 
--HG--
extra : rebase_source : 239ff77ec3765a3f190ed75b1645c6011e5d1dbd
 2014-04-14 12:04:24 -07:00
Ryan VanderMeulen
d3859c597c Backed out changesets ddbac34527fe and fa82f32d0c39 (bug 991812) for B2G bustage. 
CLOSED TREE
 2014-04-14 16:16:18 -04:00
Kyle Huey
675439ca71 Bug 991812: Remove uses of RefCounted<T> that live in Gecko. r=ehsan 2014-04-14 12:04:24 -07:00
Kyle Huey
39e8ea0778 Back out bug 991812 for bustage on a CLOSED TREE. r=me 2014-04-08 17:26:33 -07:00
Kyle Huey
d098e8c228 Bug 991812: Remove uses of RefCounted in code that lives solely in Gecko. r=ehsan 2014-04-08 16:37:05 -07:00
Daniel Holbert
501666c627 Bug 984786 part 1: Add MOZ_FINAL and private destructors to some refcounted classes in layout. r=dbaron 2014-03-21 09:16:20 +08:00
Ehsan Akhgari
9a39fcfd00 Bug 935778 - Part 0.3: Spray some more MOZ_DECLARE_REFCOUNTED_TYPENAME across the tree 2014-02-20 23:07:24 -05:00
Matt Woodrow
a3449e66b6 Bug 946952 - Run a 'compressed' paint up to 1 second after skipping the ColorLayer optimization. r=roc 2014-01-30 18:41:17 +13:00
Robert O'Callahan
54a2b48649 Bug 945203. Part 8: Add FrameLayerBuilder support for capturing nsDisplayEventRegions data into ThebesLayers. r=mattwoodrow 
--HG--
extra : rebase_source : e61c1bcfa700ac88523fb34eab7c6afd470d0d68
 2013-12-17 01:16:24 +13:00
Robert O'Callahan
f994dec6df Bug 919144. Part 9: Move FrameLayerBuilder::ContainerParameters to mozilla::ContainerLayerParameters. r=mattwoodrow 
This is to avoid having to #include FrameLayerBuilder.h in nsLayoutUtils.h in the next patch.

--HG--
extra : rebase_source : 5f3c3ae2785267949766ae094414781133d4ce44
 2013-09-27 18:01:16 +12:00
Robert O'Callahan
fd9ce678dc Bug 919144. Part 1: Rename identifiers from 'active scroll(ed) root' to 'animated geometry root'. r=mattwoodrow 
Also sneaks in a couple of other refactorings:

--HG--
extra : rebase_source : 8b0e0729f26231d98ed069b54dd4f7c1e51b10b5
 2013-09-26 09:07:26 +12:00
Matt Woodrow
1077db1b3c Bug 934860 - Don't paint display items that don't intersect the current invalid rects. r=roc 2013-11-07 08:10:50 +13:00
Matt Woodrow
43199694b8 Bug 934860 - Move paint region clipping into DrawThebesLayer. r=roc 2013-11-07 08:10:50 +13:00
Matt Woodrow
485862da70 Bug 934860 - Move some parts of FrameLayerBuilder::DrawThebesLayer into helper functions. r=roc 2013-11-07 08:10:49 +13:00
Matt Woodrow
d89f0d8a75 Bug 934860 - Start and finish transactions with temporary layer managers each time we use them to ensure that these are paired. r=roc 2013-11-07 08:10:49 +13:00
Robert O'Callahan
03e7828db2 Bug 910989. Remove nsTHashtable::Init, fallible allocation, and MT hashtables. r=ehsan,bsmedberg 
--HG--
extra : rebase_source : 0787130b1814c74bfb38dc178de94022f0b2e64e
 2013-09-02 20:41:57 +12:00
Ehsan Akhgari
2621b190f1 Bug 906790 - Minimize layout/base #includes; r=roc 2013-08-19 18:55:18 -04:00
Robert O'Callahan
613ea8adbd Bug 841192. Part 14: Convert all usage of nsDisplayClip(RoundedRect) to use DisplayListClipState/DisplayItemClip. r=mattwoodrow 
This patch does several things. Sorry.

In BuildDisplayList implementations, instead of wrapping display items in nsDisplayClip, we
push clip state onto the nsDisplayListBuilder and give the display items an
explicit clip when they're created.

In FrameLayerBuilder, we use the explicit clips we find on display items instead of
computing our own.

We remove nsDisplayClip and everything that depends on it.

We remove ExplodeAnonymousChildLists. With nsDisplayClip gone, and
nsDisplayOptionEventGrabber removed in a previous patch, there are no
anonymous child lists.

nsDisplayItem::TryMerge implementations need to make sure they have the same
clip before being merged.

I ripped out the part of PruneDisplayListForExtraPage that adjusts clip rects.
As far as I can tell, it isn't actually necessary.

--HG--
extra : rebase_source : 6f3988b385d0ac54ab26fad10b12173884441f48
 2013-03-04 22:56:02 +13:00
Robert O'Callahan
e6d799410e Bug 841192. Part 2: Move FrameLayerBuilder::Clip to DisplayItemClip. r=mattwoodrow 
--HG--
extra : rebase_source : 2950e5dd64e217a406ede2dc9f6682b5e0540975
 2013-03-04 22:55:59 +13:00
Robert O'Callahan
c4e65b7b56 Bug 852489. Part 9: Avoid temporarily setting a layer's visible region to a non-final value. r=mattwoodrow 
We currently set the visible region on most container layers twice: once in
BuildContainerLayerFor, where we don't take clipping from ancestors into
account, and then later ProcessDisplayItems calls RestrictVisibleRegionForLayer
which does take ancestor clipping into account. This causes unnecessary
Mutated()s.

In this patch we partially fix this by forcing transform layers to take
account of their ancestor clipping when they set the visible region in
BuildContainerLayerFor. For those layers we don't need to apply
ancestor clipping in ProcessDisplayItems. This is done in a rather
ugly way, by passing the ancestor clip rect as an extra field of
ContainerParameters. To preserve the optimization that prerendered
elements are treated as fully visible regardless of ancestor clipping,
we have to add a flag to turn that clipping off in BuildContainerLayerFor.

In bug 841192 we will be able to fix this in a much nicer way, because we can
get the ancestor clip directly off the nsDisplayItem passed to
BuildContainerLayerFor. But this approach is needed for the B2G18 branch.

--HG--
extra : rebase_source : 26fbe55db84ab96e1e358b8803b0563f42590836
 2013-03-22 15:17:29 +13:00
Matt Woodrow
ac128fe507 Backout changeset 3ba90f85653a (Bug 805343) for regressing android talos tests 2012-12-14 11:51:00 +13:00
Matt Woodrow
716946b5d9 Backout changeset e8f5182d94c5 (Bug 805343) for regressing android talos tests 2012-12-14 11:50:57 +13:00
Matt Woodrow
44e6df36cc Bug 820541 - Only call ProcessRemovedDisplayItems once per transaction. r=roc 2012-12-13 09:47:05 +13:00
Robert O'Callahan
a72930aa6a Bug 805343. Part 2: Update FrameLayerBuilder::HasVisibleRetainedDataFor to account for occluded display items. r=mattwoodrow 
--HG--
extra : rebase_source : 293895627ef5f37c508a23fe9e9a5639527aafcc
 2012-12-10 18:49:17 +13:00
Robert O'Callahan
f3332e603a Bug 805343. Identify DisplayItemDatas that are completely hidden by other opaque content in their ThebesLayer, and ignore them when we check to see whether there's an image we might need to invalidate. r=mattwoodrow 
--HG--
extra : rebase_source : 7e68dc33324d8ff26286723bdbaaea18060354ac
 2012-12-10 18:47:33 +13:00
Karl Tomlinson
2d225d5927 b=786100 consider display item snap when hunting uniform background color r=roc 
--HG--
extra : transplant_source : P%7CO%27%D3%E6%B6-eX%BF%AAw%EA%D1.%B8j%D9%2C
 2012-12-05 15:53:17 +13:00
Robert O'Callahan
c97490c3bd Bug 810470. Part 3: Remove code for invalidating background-attachment:fixed content when scrolling. r=mattwoodrow 
New invalidation logic takes care of this automatically.

--HG--
extra : rebase_source : 0112cfe02dbaaabf04e4010823a3cceeb6549fcd
 2012-11-09 18:14:14 -05:00
Matt Woodrow
ced424bb37 Bug 798964 - Don't invalidate nsDisplayBackgroundColor if only the image changed. r=roc 2012-11-07 11:04:53 +13:00
Chris Lord
889837ab78 Bug 803826 - Only reset the layer builder pointer in ~ClippedDisplayItem. r=mattwoodrow 
PaintInactiveLayer was resetting the layer manager's layer builder pointer, but
this was unnecessary as this was also being done on destruction of the
ClippedDisplayItem.
 2012-10-23 12:05:14 +01:00
Matt Woodrow
65d3ca6956 Bug 795674 - Handled merged frame changes with minimal invalidations. r=roc 2012-10-16 14:23:07 +13:00
Robert O'Callahan ext:(%2C%20Matt%20Woodrow%20%3Cmwoodrow%40mozilla.com%3E)
2d4d388af6 Bug 770001. When a display item has changed, don't bother invalidating areas that have been clipped out. r=mattwoodrow, roc 2012-10-16 14:10:43 +13:00
Matt Woodrow
e134487b0f Bug 795674 - Remove mNewDisplayItemData. r=roc 2012-10-12 15:39:46 +13:00
Matt Woodrow
ce3d64c0bf Bug 795674 - Add DisplayItemData debugging code. r=roc 2012-10-12 12:38:25 +13:00
Matt Woodrow
ef27f4b582 Bug 795674 - Refactor DisplayItemData storage in FrameLayerBuilder. r=roc 2012-10-12 12:38:24 +13:00
Robert O'Callahan
ed5ba29431 Bug 793065. Keep mRootPresContext alive in case something crazy happens. r=mattwoodrow 2012-10-05 16:20:35 +13:00
Isaac Aggrey
481e7dfb0b Bug 791906: Replace NSPR integer limit constants with stdint ones; r=ehsan 2012-09-28 01:57:33 -05:00
Robert O'Callahan
d2d71eb564 Bug 539356 - Fix mActiveScrolledRootOffset to be offset to the reference frame. r=mattwoodrow 2012-09-28 23:19:39 +12:00
Matt Woodrow
638b4002cb Bug 781053 - Part 2 - Trigger empty transactions when an animated image in an ImageLayer changes frame. r=roc 2012-09-25 08:31:30 +12:00
Matt Woodrow
68d1759f83 Bug 539356 - Add HasRetainedDataFor. r=roc 2012-08-29 17:48:45 +12:00