When iterating over the coord transforms or texture samplers of a
FP also have access to the owning FP.
Pass a coord transform range to GPs rather than a pointer to an
iterator.
Change-Id: If7c829a67dce6600d7f49e12d6f49f685dcace3a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/256216
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
The new c++17 only has a single level of destructuring;
append the index instead of nesting it.
Change-Id: I26bb1ca387d1de57233dbf7096dcf4bb7aaedecf
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254637
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Herb Derby <herb@google.com>
In order to be able to create GrProgramDescs pre-flush we would like as little reliance on the GrRenderTarget as possible (also GrPixelConfig is going away).
Bug: skia:9455
Change-Id: Ie35c87758bb1ed97c99c6e9e130eaeb3355f4139
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/256457
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Stephen White <senorblanco@chromium.org>
Commit-Queue: Robert Phillips <robertphillips@google.com>
This reverts commit 3e7af41224.
Change-Id: Id4f66b3956f4bdbe690db20fc478b7365ee89717
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/256676
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>
Auto-Submit: Mike Reed <reed@google.com>
To facilitate this, the GrQuadBuffer::Iter's local GrQuads that are
modified on each next() are now allowed to be operated on for the AA
inset/outsetting. Previously this required additional GrQuads on the
stack to hold the results, and additional guards for accessing localQuad()
when the entry didn't have actual coords.
With this change, a 2D op should have its device and src GrQuads' Ws
set to 1 once, and then they are completely ignored for all iteration
and tessellation, without any more redundant initialization. In all
likelihood we won't see the needle move on powerful platforms, but may
help lower end devices.
Change-Id: I457205786766403a760918e779d36ba056d69cde
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/256097
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
This is a reland of 997b37fdb9
Layout tests have been suppressed.
Original change's description:
> Revert "Revert "Make FP optimizations helpers use SkAlphaType not GrColorType""
>
> This reverts commit 078e8faa26.
>
> Change-Id: I67b2970584db5a1afbf9928b77c5444947ef71a3
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/255897
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Brian Salomon <bsalomon@google.com>
Change-Id: I6db4cff8b71ef1f29ac47f7729b9514d009f730a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/256225
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Change-Id: Ib29f41f6e71b176fec1ead26259ad1945a41e634
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254677
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
(also renames CreateTextureSetOps to AddTextureSetOps, to match naming of
GrFillRectOp::AddFillRectOps).
Now that GrTextureOp can add more than one op to the GrRTC, it can take
over ownership of its fallback code for the texture set. It already had
taken over the code for non src-over blends when drawing a single texture.
Besides consolidating where the logic of converting TextureSetEntries into
op data lives, this makes the fallback case more consistent in terms of
performance. Previously, it would go through GrRTC::drawTexturedQuad,
which attempts to merge the clip with the draw for correctness reasons.
A batch never attempted these optimizations, so now even when one op per
quad is required, there won't be the overhead of comparing clips.
Change-Id: I30883e3bd45ed4386f81584e3d68229c46b17e47
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/255781
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
This reverts commit 1792b19485.
Reason for revert: need to update legacy_convexity, still used by google3
Original change's description:
> Revert "Revert "Use flat version of path-direction enum""
>
> This reverts commit 0dacc6b7d3.
>
> Change-Id: Ie103e9f36b07e4ee256a3688a4decf3a6dd74314
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/255832
> Auto-Submit: Mike Reed <reed@google.com>
> Reviewed-by: Mike Reed <reed@google.com>
> Commit-Queue: Mike Reed <reed@google.com>
TBR=reed@google.com
Change-Id: I0ecea0eb8a237298c6b908cc4bfd1cacdfc5b900
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/255976
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>
This reverts commit 0dacc6b7d3.
Change-Id: Ie103e9f36b07e4ee256a3688a4decf3a6dd74314
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/255832
Auto-Submit: Mike Reed <reed@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>
This reverts commit e0fbe94351.
Reason for revert: need to add guard flag to flutter
Original change's description:
> Use flat version of path-direction enum
>
> Bug: skia:9663
> Change-Id: I00077d9f2b14b3e983e6a46ef6f560cabdb1678d
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/242557
> Commit-Queue: Mike Reed <reed@google.com>
> Reviewed-by: Florin Malita <fmalita@chromium.org>
TBR=fmalita@chromium.org,reed@google.com
Change-Id: If47173d9b203b2d3a175af290a15d986accb4703
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:9663
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/255831
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>
This CL is 100% plumbing. We ultimately want all GrPrimitiveProcessor-derived objects to not be refCounted. This will make several helper objects POD and, by putting them into an arena, will make managing their lifetime easier (e.g., for DDL prePreparing).
Note: the CCPR GrGeometryProcessor-derived classes only ever appear on the stack so aren't forced into arenas.
Bug: skia:9455
Change-Id: Ib9be503d2fbf8c2578642df93fc301156629829d
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/255304
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
We've just been getting lucky that the char[] was 4-byte aligned.
Change-Id: I550be24b0e6f074a38ef002b6b10fbdc5cdfcf2b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/255557
Commit-Queue: Mike Klein <mtklein@google.com>
Commit-Queue: Ben Wagner aka dogben <benjaminwagner@google.com>
Auto-Submit: Mike Klein <mtklein@google.com>
Reviewed-by: Ben Wagner aka dogben <benjaminwagner@google.com>
These are really designed around x86, so forcing them
on ARM where our existing non-immediate ops work better
is kind of silly.
Change-Id: I6b66ed0b0a71b335becdcb1d67dec471620542b0
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254440
Reviewed-by: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Change-Id: I2619784eca0f7a4dd66f2db0104cb746d9266b4e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/244369
Commit-Queue: John Rosasco <rosasco@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Hopefully this makes it clearer that the GrProgramInfo isn't copying these
objects.
Bug: skia:9455
Change-Id: Id43b7978d63b6d53d531601f3446745f33afcf2d
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254970
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
It is, arguably, not a good metric but the max number of AA Quads that are merged or chained together across all Skia's GMs and SKPs is 276.
This should also reduce the memory consumed from 245,760 bytes to 30,720 bytes.
Bug: b/143572065 skia:9601
Change-Id: I86f3ae2830d8a64ebcc6f5b7da8c0a4dac8761d5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254916
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
This is required before we can lower the max AA quad count (again).
Bug: b/143572065 skia:9601
Change-Id: Ibd029a59f6589e1ae6c66d072e5039f0e9de06dc
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254581
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
This is required before we can lower the max AA quad count (again).
Bug: b/143572065 skia:9601
Change-Id: Iea5884ffff9bd62f62527fcb597dac97d33ecd76
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254438
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
This fixes an issue where fragment processors which were not written
in pure SkSL did not interact properly with coordinate overrides.
Change-Id: I960173986fd8d6be7322f55ab0d81a533b6b89e5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/252916
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
This all comes together as
uminv tmp, condition
fmov gp, tmp
cbnz gp, all_true
brk 0
all_true:
...
The key idea is uminv(vec) will return 0 if any of the inputs are 0,
and non-zero if all of the inputs are non-zero, namely 0xffffffff.
fmov moves that minimum from a vector register to a general purpose
register where we can test it with cbnz, compare and branch if non-zero.
This jumps over the `brk 0` debug trap when all inputs are true.
Change-Id: If5deb77a77f52221d0649e537179743c45eb9cc5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254479
Auto-Submit: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
This reverts commit dd947ceb55.
Reason for revert: CL isn't getting any respect from Android
Original change's description:
> Respect the max indexBuffer limits in the bulk texture draw API
>
> This is required before we can lower the max AA quad count (again).
>
> Bug: b/143572065 skia:9601
> Change-Id: Id34123476ad49a57dc9ce7fe13f941c06f721b74
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/252603
> Commit-Queue: Robert Phillips <robertphillips@google.com>
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>
TBR=robertphillips@google.com,michaelludwig@google.com
Change-Id: Ibe396fe5a0351190c214ed54611ce72b7928cadd
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: b/143572065 skia:9601
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254437
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
This is required before we can lower the max AA quad count (again).
Bug: b/143572065 skia:9601
Change-Id: Id34123476ad49a57dc9ce7fe13f941c06f721b74
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/252603
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Enance SkBulkGlyphMetrics, SkBulkGlyphMetricsAndPaths, and SkBulkGlyphMetricsAndImages
with single glyph calls. In addtion, add calls needed to have the rest of the system
work with these interfaces.
As a resulte move the glyph, prepareImage, and preparePath calls to private.
Change-Id: I8d383b649390e45f621dcb9d62fb8367a55cee02
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254056
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Herb Derby <herb@google.com>
Change-Id: I90f40467be145537cc08461fd3805606432f0cfd
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254240
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: John Rosasco <rosasco@google.com>
Auto-Submit: John Rosasco <rosasco@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Lots of x86 instructions can take their right hand side argument from
memory directly rather than a register. We can use this to avoid the
need to allocate a register for many constants.
The strategy in this CL is one of several I've been stewing over, the
simplest of those strategies I think. There are some trade offs
particularly on ARM; this naive ARM implementation means we'll load&op
every time, even though the load part of the operation can logically be
hoisted. From here on I'm going to just briefly enumerate a few other
approaches that allow the optimization on x86 and still allow the
immediate splats to hoist on ARM.
1) don't do it on ARM
A very simple approach is to simply not perform this optimization on
ARM. ARM has more vector registers than x86, and so register pressure
is lower there. We're going to end up with splatted constants in
registers anyway, so maybe just let that happen the normal way instead
of some roundabout complicated hack like I'll talk about in 2). The
only downside in my mind is that this approach would make high-level
program descriptions platform dependent, which isn't so bad, but it's
been nice to be able to compare and diff debug dumps.
2) split Op::splat up
The next less-simple approach to this problem could fix this by
splitting splats into two Ops internally, one inner Op::immediate that
guantees at least the constant is in memory and is compatible with
immediate-aware Ops like mul_f32_imm, and an outer Op::constant that
depends on that Op::immediate and further guarantees that constant has
been broadcast into a register to be compatible with non-immediate-aware
ops like div_f32. When building a program, immediate-aware ops would
peek for Op::constants as they do today for Op::splats, but instead of
embedding the immediate themselves, they'd replace their dependency with
the inner Op::immediate.
On x86 these new Ops would work just as advertised, with Op::immediate a
runtime no-op, Op::constant the usual vbroadcastss. On ARM
Op::immediate needs to go all the way and splat out a register to make
the constant compatible with immediate-aware ops, and the Op::constant
becomes a noop now instead. All this comes together to let the
Op::immediate splat hoist up out of the loop while still feeding
Op::mul_f32_imm and co. It's a rather complicated approach to solving
this issue, but I might want to explore it just to see how bad it is.
3) do it inside the x86 JIT
The conceptually best approach is to find a way to do this peepholing
only inside the JIT only on x86, avoiding the need for new
Op::mul_f32_imm and co. ARM and the interpreter don't benefit from this
peephole, so the x86 JIT is the logical owner of this optimization.
Finding a clean way to do this without too much disruption is the least
baked idea I've got here, though I think the most desirable long-term.
Cq-Include-Trybots: skia.primary:Test-Debian9-Clang-GCE-CPU-AVX2-x86_64-Debug-All-SK_USE_SKVM_BLITTER,Test-Debian9-Clang-GCE-CPU-AVX2-x86_64-Release-All-SK_USE_SKVM_BLITTER
Change-Id: Ie9c6336ed08b6fbeb89acf920a48a319f74f3643
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/254217
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
The logic implemented here is roughly
assert_true(v):
if any ~v {
int3()
}
in assembly as
```
vptest v, constant 0xffffffff mask
jc ok
int3
ok:
```
jc branches if (~v & mask) are all zero, with mask set fully, that's
branch if ~v are all zero, which is to say, v are all ~0, true. So we
jump over the int3 breakpoint if v are all true.
Cq-Include-Trybots: skia.primary:Test-Debian9-Clang-GCE-CPU-AVX2-x86_64-Debug-All-SK_USE_SKVM_BLITTER
Change-Id: Ie0fc1da15b1a0dba00c66af610ccde18f5985f8a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/253897
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Auto-Submit: Mike Klein <mtklein@google.com>
Replaces numSamples with numRasterSamples, and adds isMixedSampled.
The sample count that vulkan and metal actually want to know is how
many samples the rasterizer will compute, which may not match the
number of samples in the render target when we have mixed samples.
They will also need to know whether a program is mixed sampled in
order to set up coverage modulation.
Change-Id: I133c11f74b7dc6a7580818ef73d6deec1d201b64
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/253550
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Metal swapchain textures (equivalent to GL's FBO 0) can't be used with
MTLBlitCommandEncoder. For copySurface we need to detect this and fail.
Future changes will handle GrSurfaceProxy.
Change-Id: I9235d91cf31e96260c8d1da5938f5441d2e49d27
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/252045
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Will use these to implement assert_true on x86.
Change-Id: I9d2595a35518b6971dd8e418b583febd3960c7f6
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/253896
Commit-Queue: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Reed <reed@google.com>
Auto-Submit: Mike Klein <mtklein@google.com>
Reviewed-by: Mike Reed <reed@google.com>
This is an assert that is active in debug mode. For the moment it only
works in the interpreter, but I plan to follow up with JIT code too.
assert_true() is a data sink like a store() as far as lifetime goes,
though we take care to allow it to be hoisted if its inputs are. An
assert_true's existence will keep all its inputs alive, and in release
builds where we skip the instruction, those inputs will all drop away
automatically.
Tested locally by forcing the interpreter. It shouldn't be long before
I have at least x86 JIT asserts working too.
Change-Id: I7aba40d040436a57a6b930790f7b8962bafb1a8c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/253756
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
This plumbs through round but doesn't use it. I want that change to be
its own CL. It's nice to have assembler support and the name changes
even if I revert using round.
Change-Id: I6d67ec5c63546069eb7cc1c91599b599bafcda66
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/253724
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
No diffs.
Change-Id: Ia0b35c2787e27d74763f21b81072affa6caf1e5a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/253720
Commit-Queue: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Reed <reed@google.com>
Auto-Submit: Mike Klein <mtklein@google.com>
Reviewed-by: Mike Reed <reed@google.com>
