aboutsummaryrefslogtreecommitdiffhomepage
path: root/src/record
diff options
context:
space:
mode:
authorGravatar commit-bot@chromium.org <commit-bot@chromium.org@2bbb7eff-a529-9590-31e7-b0007b416f81>2014-04-29 15:34:03 +0000
committerGravatar commit-bot@chromium.org <commit-bot@chromium.org@2bbb7eff-a529-9590-31e7-b0007b416f81>2014-04-29 15:34:03 +0000
commit705b1ff617ffe672e65d53bcaf3b690e8bb6b025 (patch)
tree7067ee31d459ec7c8857e6da48b3dbf412c603a5 /src/record
parent0a5c233e3b911232c0d6f9a88ded99ecf88b8a97 (diff)
Don't bother doing the empty clip check in SkRecordDraw.
On Mike's suggestion, I tested out not doing any empty-clip check at all in SkRecordDraw, given that mostly we'll do that again anyway inside SkCanvas. Most SKPs are identical to the status quo, whether bot or silk, played back in tiles or full. Average playback performance, both arithmetic and geometric mean, is also unchanged. A handful of SKPs do draw faster or slower reliably, particularly when tiled. E.g. a cnn tile draws about 40% faster, a cuteoverload tile about 20% slower. Their profiles look pretty much the same before and after, so I can't really explain the changes. I'd say, given that performance is mostly identical and very identical in bulk, we might as well remove this code. It's nice to keep SkRecordDraw as dumb as possible. BUG=skia:2378 R=reed@google.com, fmalita@chromium.org, mtklein@google.com, borenet@google.com Author: mtklein@chromium.org Review URL: https://codereview.chromium.org/258183002 git-svn-id: http://skia.googlecode.com/svn/trunk@14433 2bbb7eff-a529-9590-31e7-b0007b416f81
Diffstat (limited to 'src/record')
-rw-r--r--src/record/SkRecordDraw.cpp21
1 files changed, 7 insertions, 14 deletions
diff --git a/src/record/SkRecordDraw.cpp b/src/record/SkRecordDraw.cpp
index bc1bbf5fd4..4878b3f4f6 100644
--- a/src/record/SkRecordDraw.cpp
+++ b/src/record/SkRecordDraw.cpp
@@ -7,8 +7,6 @@
#include "SkRecordDraw.h"
-#include "SkRecordTraits.h"
-
namespace {
// This is an SkRecord visitor that will draw that SkRecord to an SkCanvas.
@@ -29,27 +27,22 @@ private:
// No base case, so we'll be compile-time checked that we implemented all possibilities below.
template <typename T> void draw(const T&);
- // skip() returns true if we can skip this command, false if not.
- // Update fIndex directly to skip more than just this one command.
-
- // If we're drawing into an empty clip, we can skip it. Otherwise, run the command.
- template <typename T>
- SK_WHEN(SkRecords::IsDraw<T>, bool) skip(const T&) { return fCanvas->isClipEmpty(); }
+ // skip() should return true if we can skip this command, false if not.
+ // It may update fIndex directly to skip more than just this one command.
- template <typename T>
- SK_WHEN(!SkRecords::IsDraw<T>, bool) skip(const T&) { return false; }
+ // Mostly we just blindly call fCanvas and let it handle quick rejects itself.
+ template <typename T> bool skip(const T&) { return false; }
- // Special versions for commands added by optimizations.
+ // We add our own quick rejects for commands added by optimizations.
bool skip(const SkRecords::PairedPushCull& r) {
if (fCanvas->quickReject(r.base->rect)) {
fIndex += r.skip;
return true;
}
- return this->skip(*r.base);
+ return false;
}
-
bool skip(const SkRecords::BoundedDrawPosTextH& r) {
- return this->skip(*r.base) || fCanvas->quickRejectY(r.minY, r.maxY);
+ return fCanvas->quickRejectY(r.minY, r.maxY);
}
SkCanvas* fCanvas;