Observed AE layer parenting semantics:
* layers are flagged as either 2D or 3D
* camera applies to 3D layers, but not to 2D layers
* parented 3D layers treat their ancestor transform chain as 3D (SkMatrix44)
* parented 2D layers treat their ancestor transform chain as 2D (SkMatrix, ignoring 3D components)
This means that for a given layer, we may need to build two distinct transform chains - depending
on the type of descendant layer being considered.
Furthermore, transforms are animatable and their animators are scoped to a layer controller. Since
we're potentially building two version of the transform node, we need to ensure all animators for
both of them are transferred to controller object (we still want to only instantiate a single layer
controller and render tree to avoid duplication). IOW, all dependent layer transforms need to be
considered before "sealing off" a given layer controller.
In order to avoid a layer dependency/topological sort, we can split off the transform tree
construction into a separate pass. High-level changes:
-- replace existing LayerAttachContext with CompositionBuilder
(holds LayerBuilders and other Composition-wide state)
-- replace LayerRec with LayerBuilder
(holds Layer-wide state and also caches transform nodes)
-- pass 1: for each LayerBuilder, transitively build and cache a transform chain
of a type (2d/3d) determined by the leaf (entry point) layer
-- pass 2: for each LayerBuilder, build the actual layer content render tree
and instantiate the layer controller objects
Bug: skia:8914
Change-Id: I9f7efcf4819424282fd3dda98f5621ba12fd001b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/251001
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Use `extra_cflags=["-DSK_CAPTURE_DRAW_TEXT_BLOB"]` to enable.
Change-Id: I1d6db478ee91696cdce090647b889c17a83a2718
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/250259
Commit-Queue: Hal Canary <halcanary@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Rewrite program instructions so that each value becomes available as
late as possible, just before it's used by another instruction. This
reorders blocks of instructions to reduce them number of temporary
registers in flight.
Take this example of the sort of program that we naturally write,
noting the registers needed as we progress down the right:
src = load32 ... (1)
sr = extract src ... (2)
sg = extract src ... (3)
sb = extract src ... (4)
sa = extract src ... (4, src dies)
dst = load32 ... (5)
dr = extract dst ... (6)
dg = extract dst ... (7)
db = extract dst ... (8)
da = extract dst ... (8, dst dies)
r = add sr dr (7, sr and dr die)
g = add sg dg (6, sg and dg die)
b = add sb db (5, sb and db die)
a = add sa da (4, sa and da die)
rg = pack r g ... (3, r and g die)
ba = pack b a ... (2, b and a die)
rgba = pack rg ba ... (1, rg and ba die)
store32 rgba ... (0, rgba dies)
That original ordering of the code needs 8 registers (perhaps with a
temporary 9th, but we'll ignore that here). This CL will rewrite the
program to something more like this by recursively issuing inputs only
once needed:
src = load32 ... (1)
sr = extract src ... (2)
dst = load32 ... (3)
dr = extract dst ... (4)
r = add sr dr (3, sr and dr die)
sg = extract src ... (4)
dg = extract dst ... (5)
g = add sg dg (4, sg and dg die)
rg = pack r g (3, r and g die)
sb = extract src ... (4)
db = extract dst ... (5)
b = add sb db (4, sb and db die)
sa = extract src ... (4, src dies)
da = extract dst ... (4, dst dies)
a = add sa da (3, sa and da die)
ba = pack b a (2, b and a die)
rgba = pack rg ba ... (1, rg and ba die)
store32 rgba ... (0)
That trims 3 registers off the example, just by reordering!
I've added the real version of this example to SkVMTest.cpp.
(Its 6th register comes from holding the 0xff byte mask used
by extract, in case you're curious).
I'll admit it's not exactly easy to work out how this reordering works
without a pen and paper or trial and error. I've tried to make the
implementation preserve the original program's order as much as makes
sense (i.e. when order is an otherwise arbitrary choice) to keep it
somewhat sane to follow.
This reordering naturally skips dead code, so pour one out for ☠️ .
We lose our cute dead code emoji marker, but on the other hand all code
downstream of Builder::done() can assume every instruction is live.
Change-Id: Iceffcd10fd7465eae51a39ef8eec7a7189766ba2
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/249999
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Change-Id: I6d29290eb2962262bb080a86dc829c39986cae4f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/249226
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
- Copy effect state to particle uniforms before each script, so changes
from spawn or update are visible.
- Guard path binding against out of range access
- New effect that actually stresses both of these conditions
Change-Id: Ice6112793099e515438af8bb863e9e1bf03d08b1
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/249125
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Most hoisted values are used in the loop body (and that's really the
whole point of hoisting) but some are just temporaries to help produce
other hoisted values. This used_in_loop bit helps us distinguish the
two, and lets us recycle registers holding temporary hoisted values not
used in the loop.
The can-we-recycle logic now becomes:
- is this a real value?
- is it time for it to die?
- is it either not hoisted or a hoisted temporary?
The set-death-to-infinity approach for hoisted values is now gone. That
worked great for hoisted values used inside the loop, but was too
conservative for hoisted temporaries. This lifetime extension was
preventing us from recycling those registers, pinning enough registers
that we run out and fail to JIT.
Small amounts of refactoring to make this clearer:
- move the Instruction hash function definition near its operator==
- rename the two "hoist" variables to "can_hoist" for Instructions
and "try_hoisting" for the JIT approach
- add ↟ to mark hoisted temporaries, _really_ hoisted values.
There's some redundancy here between tracking the can_hoist bit, the
used_in_loop bit, and lifetime tracking. I think it should be true, for
instance, that !can_hoist && !used_in_loop implies an instruction is
dead code. I plan to continue refactoring lifetime analysis (in
particular reordering instructions to decrease register pressure) so
hopefully by the time I'm done that metadata will shake out a little
crisper.
Change-Id: I6460ca96d1cbec0315bed3c9a0774cd88ab5be26
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/248986
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Gives enough information to locate variables by name (using the same
scheme as glGetUniformLocation), and provide hints about type and size.
Bug: skia:9513
Change-Id: I9444f1042471967a79c9f05167dcdb78eca41bad
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/244502
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Simplify burst handling. Scripts should just add to burst (if
they want to handle programmatic bursting, as well).
Update most effects to handle dynamic updates to position better,
and add a sample effect meant to be used with mouse tracking.
Change-Id: Ia302e1d04e62e2b07974807c44067786cc10a8ad
Bug: skia:9513
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/248798
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
This was only being used in one effect (and for no good reason). SkSL is
plenty powerful to re-implement something similar if required, at no
real performance cost.
Re-implemented the one effect that used it with simpler math in the
script, updated the copy of that effect in the gallery.
Docs-Preview: https://skia.org/?cl=247040
Change-Id: I68c86d6550dd4f003f6ba5ecd0febab37b86540b
Bug: skia:9513
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/247040
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Confetti mimics the look of a standard skottie asset
Change-Id: Iffeedeb24182c4ac2d3ec390614bc1861b821376
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/246518
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Also removed some older effects that weren't interesting, improved others,
cleaned up the unused functions in several, and renamed most of them to
reflect which feature they're demonstrating.
Change-Id: Ib44a00ec3d25e852a1d1661918137ba13d30c86b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/244119
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
* Added a new binding type, SkEffectBinding. This stores another
entire effect params structure (so the JSON is just nested).
The name is a callable value that spawns a new instance of
that effect, inheriting the parameters of the spawning effect
or particle (depending on which kind of script made the call).
* Broke up the monolithic update function into some helpers,
got some code reuse with the script calling logic.
* Unlike particle capacity, there is no upper limit on child
effects (yet), so it's easy to trigger runaway memory and
CPU consumption. Be careful.
* Added death scripts to effects and particles, which are a
common place to want to spawn sub-effects. Like spawn,
these run on each loop, but for one-shots they play at the
end. Even with loops, this is helpful for timing sub-effects
(see fireworks2.json).
* Finally, added a much more comprehensive example effect,
raincloud.json. This includes a total of three effects, to
generate a cloud, raindrops, and splashes when those drops
hit "the ground".
Change-Id: I3d7b72bcbb684642cd9723518b67ab1c7d7a538a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/242479
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
This change adds another layer of complexity and control to
the particle system. There are now two code chunks: the old
code that's run per-particle, and new code that's run for
the effect itself. This allows for effect lifetime to be set
by the script (eg, randomly), as well as the emission rate.
Rate can vary over time (see pulse.json), and particles can
be emitted in bursts by setting the effect's burst field
(see fireworks.json).
Additionally, the effect has its own frame of reference and
color, which becomes the default state for newly emitted
particles. This allows synchronizing state across particles
in various interesting ways (see color in fireworks.json).
Change-Id: Iec2f7a3427ce1d6411ed7ef5b3023cbef2e8a134
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/240498
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
We're currently letting render context overrides (opacity, color
filters, blend mode, etc) spill down the descendent/mask content
tree.
This is not ideal, as mask content isolation breaks atomicity
assumptions for deferred overrides. Case in point: motion blur uses
SkBlendMode::kPlus to accumulate content "layers" - but since mask
content gets rendered into a separate layer, it fails to produce the
expected result.
The fix is to realize all context overrides on the top-level mask layer
(we already allocate this layer, so there's no reason to defer
downstream anyway).
Change-Id: Icbb7e403f90feecfae5846697f559a03d8aa4097
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/239036
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Change-Id: If99e1802c8187ebd98b67717d744c6695bb25900
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/238118
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Introduce a new hybrid valign extension, kVisualDownscaleToFit (sk_vj: 4):
- when the text shaped at the requested size fits within the box,
center vertically (same as kVisualCenter)
- otherwise, scale down until it fits (same as kVisualResizeToFit)
Change-Id: I8e096a49e2b87582e1bd42161657ec4ef561ebdf
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/235601
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
This makes the register recycling checks a bit more
precise. At head we never recycle a register that's
holding a hoisted value, which is overly conservative.
We really should never recycle a register that's still
needed. By extending the lifetime of any hoisted value
that's used in the loop, we prevent that, while still
allowing hoisted values that are only used in hoisted
computation to be reused.
This takes just a small tweak in the JIT code (removing
the !hoisted({x,y,z}) checks), and a somewhat larger
refactoring in the interpreter, making both hoisted and
non-hoisted code go through the same recycling register
assignment flow.
There's one diff in the existing cases where we now
reuse a hoisted register, and I've added a second test
just to make sure it's covered explicitly.
Change-Id: I25b37ab1f1fea3042d7fd167529abc8fed1dddff
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/233239
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Due to limitations in BodyMovin/AE JSX, full effect data is not
available (specifically the "channel range" property).
We only support static master hue, static master saturation and
static master lightness at this point.
This CL also introduces a new animation builder pattern:
DiscardableAdapterBase and attachDiscardableAdapter().
The former is a base class for adapters with full animator ownership.
This enables a) capturing raw adapter pointers in animator lambdas and
b) syncing to SG only once, after all local animators are updated).
The latter is a helper for managing adapter creation and optional
destruction (when all adapter properties are static we can discard it).
Change-Id: Iecc4b78830e5464e7958cb12cdfd75a61010aa25
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/231956
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Add missing comparison and selection ops, bit casts, 16-bit memory
operations, gathers, uniform loads, and fill in math holes where
reasonable. Update some names to be a bit more regular.
I think all instructions are implemented in the interpreter,
and many tested. More testing and JITs to follow.
Change-Id: I8cf377e8b72a86ac950e020892ce82b39e9d7277
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/229893
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Change-Id: If57fb79db8f8c5fd185fefaa202167c8082dd846
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/229921
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Change-Id: Ic81b3433b485ca9ce0e60bd10ec12706e673ee89
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/229917
Commit-Queue: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Auto-Submit: Brian Osman <brianosman@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
This removes all of the fixed-function particle affector classes.
Instead, each particle effect just has two SkSL snippets, one for
spawn logic, and one for update logic. Each one gets an inout copy
of the particle struct. Ultimately, this makes the effects much
simpler and smaller, while also being far more flexible (you can
do whatever you want with any values you want). Finally, because
the interpreter is vectorized and a particular effect's scripts
are usually tuned to the specific behaviors desired, it's faster
on basically every effect I compared.
I re-created all of the old effects in the new system. Many just
use pure SkSL (no curves or anything). Some of the old curve and
path/text stuff was very handy, though - so those are now exposed
as external values in the interpreter. Basically, an effect can
have any number of named "bindings" that are a callable thing.
This can be a path, text (shortcut for making fancy paths), curve,
or color curve. The path ones return a float4 with position and
normal, the curves return one or four floats.
... and this transposes all of the particle data storage into
SoA form, so that it can use the much faster interpreter entry
point.
Change-Id: Iebe711c45994c4201041b12d171af976bc5e758e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/222057
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Even if a JIT ultimately doesn't end up hoisting any values, it's going
to want this information while it decides. Writing it in one place also
ensures we only get it wrong in one place...
I'm no_ extending the lifetime of hoisted instructions here in Builder.
That's something to leave to the backend so they have the flexibility of
which of these values to hoist, if any. If they don't hoist, they'll
need to know when the value dies.
Moving this information back here lets the test expectation goldens
reflect the hoist bit again too. Kind of nice.
Change-Id: Ib165ca898a97c1d822cb28fe24f15bae4d570a17
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/229024
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
I'm slowly refactoring my way to where hoisting and register assignment
are done in backend-specific ways, but this liveness analysis is always
going to be useful for each backend.
Use deaths() to restore friendly ☠️ dead code markers in test dumps.
Change-Id: I3ab94665bbbbf0788b0b27e00d644eba927dff47
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/228113
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Unlike all other Skottie effects, motion blur requires sampling at multiple
points on the timeline.
To support this:
1) Introduce MotionBlurEffect - a custom SG render node which can drive
the timeline of its subtree using an sksg::Animator.
2) Introduce MotionBlurController to swap for a regular LayerController
when needed. MotionBlurController dispatches time ticks to
MotionBlurEffect instead of directly to the layer animators.
The actual motion blur impl is based on
https://skia-review.googlesource.com/c/skia/+/221416.
Motion blur requires Lottie files exported with this BodyMovin patch:
https://github.com/bodymovin/bodymovin-extension/pull/15
Change-Id: I075e101ea91ec9aa300bac35ee810fd539f1aced
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/225416
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Noticed we were only dumping the final register
programs for the integer code. Might as well also
track the value programs.
Change-Id: I417c5c655b632691557bbbb136dcbd3f3167af9a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/225324
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
We used to rely solely on visual bounds for vertical alignment. That
had the downside of leading/trailing empty lines being ignored.
Then https://skia-review.googlesource.com/c/skia/+/220916 switched to
using typographical bounds. This approach produces results in line
with AE, but allows some glyphs to overflow the alignment boundary.
This CL introduces a hybrid approach:
1) for standard AE text alignment, continue to use typographical bounds
2) for Skottie VAlign extensions (sk_vj), use the union of typographical
and visual bounds - this should mitigate both issues mentioned above
Change-Id: Ifd3ccae3d721728ce67942206160ebe92056d3a2
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/224188
Reviewed-by: Ben Wagner <bungeman@google.com>
Reviewed-by: Avinash Parchuri <aparchur@google.com>
I was just reading the ARM docs and realized that their BIC ("BIt
Clear") is the same as SSE's ANDN ("AND Not") instruction. It's kind of
a neat little tool to have laying around... comes up more than you'd
think, and it's sometimes the clearest way to express what you're doing,
as in the changed program here where the comment is "mask away the low
bits". That's a bit_clear with a mask for what you want to clear away!
And the real reason to write this up is that I want to have a CL to
point to that shows how to add an instruction top to bottom.
Change-Id: I99690ed9c1009427b3986955e7ae6264de4d215c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/223120
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Instruction is the fundamental data, and Analysis derived from it.
The fields in Analysis are only* needed in Builder::done(), and this
split seems to help clarify what done() can tweak (Analysis) and what
it cannot (fProgram, Instructions). done() is now const.
No speed change as far as I can tell.
* As you may notice looking at the test expectations, making analysis
ephemeral means that dump() can no longer print the skull for dead code
or the arrow for hoisted. The register program that's also in the
expectation file still reflects both of these optimizations, so we're
not really losing any information. Just maybe less demo-friendly.
Change-Id: I79feb57558525591baf3faadeb59c418c12793f3
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/223119
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
This cuts the overhead bench from about 19µs to about 15µs.
The key insight here is that the only registers that might become
available after any given instruction are the ones that hold that
instruction's inputs. We can check when they become available
directly from the original Builder::Program, without needing a
side death schedule data structure.
Marking hoisted instructions as having life == program size
helps make this logic a little simpler to reason through.
Change-Id: Ifb9957f2d0e323e0e5d07996a2cc988f7c8b4c3f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/223117
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Implement radial wipe with a sweep gradient shader mask filter.
The implementation is slightly convoluted because edge feathering requires a real blur, which in turn requires content layer isolation.
So there are two distinct operation modes:
- no feather -> draw the content directly into the dest buffer, with the mask filter
deferred in SG context
- feather -> draw the content into a separate layer, then blend (dstOut) the composed
blur+shader mask on top
Change-Id: I253701aff42db8010ce463762252c262e2c5d92b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/222596
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
- 32x8 i32 add,sub,mul
- add I32_Naive bench/test builder to get better i32 mul coverage
- minor refactoring all over
Change-Id: I13cc19ff37a2da0bcff289ba51baac08f456d6c5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/222485
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
The motion tile phase is a one-dimensional shift, applied to every other
row or column (based on a selector property).
Implement using a masking shader (covering the static rows/cols),
and blend mode shader composition (srcIn for static/pass-through
rows/cols, and srcOut for phased rows/cols).
TBR=
Change-Id: I336c150e5d4900962dc2de801a4e1572cf4b5d59
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/221339
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Florin Malita <fmalita@chromium.org>
I'm staring at this assembly,
vmovups (%rsi), %ymm3
vpsrld $24, %ymm3, %ymm4
vpslld $16, %ymm4, %ymm15
vorps %ymm4, %ymm15, %ymm4
vpsubw %ymm4, %ymm0, %ymm4
Just knowing that could be
vmovups (%rsi), %ymm3
vpshufb 0x??(%rip), %ymm3, %ymm4
vpsubw %ymm4, %ymm0, %ymm4
That is, instead of shifting, shifting, and bit-oring
to create the 0a0a scale factor from ymm3, we could just
byte shuffle directly using some pre-baked control pattern
(stored at the end of the program like other constants)
pshufb lets you arbitrarily remix bytes from its argument and
zero bytes, and NEON has a similar family of vtbl instructions,
even including that same feature of injecting zeroes.
I think I've got this working, and the speedup is great,
from 0.19 to 0.16 ns/px for I32_SWAR, and
from 0.43 to 0.38 ns/px for I32.
Change-Id: Iab850275e826b4187f0efc9495a4b9eab4402c38
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/220871
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Yet another way to transform a layer, disguised as a distort effect.
TBR=
Change-Id: Ic2d5479fa6ae27b460de60875924f73f77fc7f71
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/221001
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Now that we've got shr_16x2, extract(..., 8, splat(0x00ff00ff)) is
better done as shr_16x2(..., 8). This swaps a 16-bit shift in for
the 32-bit shift, a wash, but lets us drop the bit_and at the end,
saving one whole instruction.
This places I32_SWAR a tiny little bit faster than the code in Opts,
like .19 ns/px vs .20 ns/px for Opts.
Change-Id: I4160dc03ecc8b855c0773a927f1510ad5cbb4b87
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/220856
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
This is the final bunny I've got in my hat, I think...
Remembering that none of the s += d*invA adds can overflow,
we can use a single 32-bit add to add them all at once.
This means we don't have to unpack the src pixel into rb/ga
halves. We need only extract the alpha for invA.
This brings I32_SWAR even with the Opts code!
curr/maxrss loops min median mean max stddev samples config bench
36/36 MB 133 0.206ns 0.211ns 0.208ns 0.211ns 1% ▁▇▁█▁▇▁▇▁▇ nonrendering SkVM_4096_I32_SWAR
37/37 MB 152 0.432ns 0.432ns 0.434ns 0.444ns 1% ▃▁▁▁▁▃▁▁█▁ nonrendering SkVM_4096_I32
37/37 MB 50 0.781ns 0.794ns 0.815ns 0.895ns 5% ▆▂█▃▅▂▂▁▂▁ nonrendering SkVM_4096_F32
37/37 MB 76 0.773ns 0.78ns 0.804ns 0.907ns 6% ▄█▅▁▁▁▁▂▁▁ nonrendering SkVM_4096_RP
37/37 MB 268 0.201ns 0.203ns 0.203ns 0.204ns 0% █▇▆▆▆▆▁▆▆▆ nonrendering SkVM_4096_Opts
Change-Id: Ibf0a9c5d90b35f1e9cf7265868bd18b7e0a76c43
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/220805
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
I figure the easiest way to expose 16-bit operations
is to expose 16x2 pair operations... this means we
can continue to always work with the same size vector.
Switching from 32-bit multiplies to 16-bit multiplies
is going to deliver the most oomph... they cost roughly
half what 32-bit multiplies do on x86.
Speed now:
I32_SWAR: 0.27 ns/px
I32: 0.43 ns/px
F32: 0.76 ns/px
RP: 0.8 ns/px
Opts: 0.2 ns/px
Change-Id: I8350c71722a9bde714ba18f97b8687fe35cc749f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/220709
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
I just kind of remembered that if we're doing (xy+x)/256
and x is a destination channel and y is 255-sa, then you
can get the +x for free by multiplying by 256-sa instead.
(d * (255-sa) + d)
(d * (255-sa + 1))
(d * (256-sa) )
Duh. This is a trick we play in a lot of legacy code and
I've just now realized it's exactly equivalent to the trick
I want to play here... sigh.
Folding this math in kind of makes mul/mad_unorm8 moot.
Speed's getting good:
I32_SWAR: 0.3 ns/px
I32 : 0.55 ns/px
F32 : 0.8 ns/px
RP : 0.8 ns/px
Opts : 0.2 ns/px
Change-Id: I4d10db51ea80a3258c36e97b6b334ad253804613
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/220708
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
The mask-only special case for extract is wrong...
it never looked it its input!
This not only makes things correct-er, but oddly it also
makes them faster by breaking inter-loop data dependencies.
Disable tests for _I32... they're actually still broken
because of a much more systemic flaw in how I've evaluated
programs. The _F32 and _I32_SWAR JIT code and all interpreted
code is just getting lucky. o_O
While here, update the I32_SWAR code to use the same math as I32,
(x*y+x)/256 for unorm8 mul. This just helps keep me sane.
Change-Id: I1acc09adb84c426fca4b2be5ca8c2d46d9678dd8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/220577
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
