The compute renderer is a failed experiment: a better port of the
Vello vector renderer exists[0] and the upcoming Go 1.24 release
no longer builds the gioui.org/cpu module because of #60725.
Remove it.
[0] https://github.com/dominikh/jello
Signed-off-by: Elias Naur <mail@eliasnaur.com>
FBOs and window framebuffers generally have different pixel formats, and
so require separate pipeline configurations.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
It was a design mistake to make profiling data available to programs.
Rather, profiling should either be a user-configurable debug overlay,
reported through runtime/trace, or both.
This change drops the io/profile package because we're about to overhaul
event routing.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
The only remaining use of the cache is mapping handles to textures.
Using a concrete type for the key avoids the allocation caused by convT.
If we need more caches again in the future we can copy the type, or make
it generic.
Instead of updating the benchmark, we removed it outright. It suffered
from several flaws:
- The amount of work for each iteration of b.N wasn't constant, because
the same cache was reused, growing ever larger in size.
- It only tested the cost of insertions. The comment "half are the same
and half updated" wasn't true, as calling 'put' with the same key twice
would've resulted in a panic.
- It didn't simulate any particular workload or cache size, making the
benchmark useless for comparing different cache implementations. The
cost of insertions isn't particularly interesting.
Signed-off-by: Dominik Honnef <dominik@honnef.co>
This adds support for nearest neighbor filtering,
which can be useful in quite a few scenarios.
img := paint.NewImageOp(m)
img.Filter = paint.FilterNearest
img.Add(gtx.Ops)
Fixes: https://todo.sr.ht/~eliasnaur/gio/414
Signed-off-by: Egon Elbre <egonelbre@gmail.com>
This commit updates the text package to be able to load system fonts. As a consequence,
application authors may choose to provide no fonts manually, and it's
also possible that the system provides none (WASM, for instance, currently provides no
system fonts). As such, the text stack needed some minor tweaks to handle this case by
displaying blank spaces where text should be rather than crashing when no faces are
available.
Internally, we are dropping the old method of choosing faces and instead relying solely
on the new font matching logic in go-text. I chose to do this because maintaining two
different sets of logic with a hierarchical relationship proved to be really complex,
and also the go-text logic seems to produce higher-quality choices.
The breaking API change from this commit is the new way of constructing a text shaper
using text.ShaperOptions. Providing no options will result in a shaper that uses solely
system fonts. The various options can be used to disable system font loading and to
provide an already-parsed collection of fonts as per Gio's old API.
The material.NewTheme function now accepts no arguments instead of a font collection.
Users wanting to provide a collection can simply provide a new shaper configured how
they would like:
theme := material.NewTheme()
theme.Shaper = text.NewShaper(text.NoSystemFonts(), text.WithCollection(gofont.Regular()))
This commit touches many packages to fix up their construction of text shapers, mostly in
test code. The changes to the tests in package widget deserve special note:
Changing our font resolution logic caused the tofu characters within the
test strings to use a different font's tofu. This isn't a problem, but shifted
the layout of the shaped text a little bit. I've updated the numbers to expect
the new glyph positions.
Fixes: https://todo.sr.ht/~eliasnaur/gio/309
Fixes: https://todo.sr.ht/~eliasnaur/gio/184
Signed-off-by: Chris Waldon <christopher.waldon.dev@gmail.com>
When building GPU vertices from paths, we call stroke.SplitCubic once
per OpCubic. Before this change, each call to stroke.SplitCubic would
allocate a slice, which we would only use to iterate over.
This allocation can be easily avoided by reusing the slice. We can
conveniently store it in gpu.quadSplitter.
In a real application that renders hundreds of paths with tens of
rounded rectangles per path, this saved roughly 4500 allocations (or 1
MB worth) per frame.
Signed-off-by: Dominik Honnef <dominik@honnef.co>
When the line overlaps itself backtracking exactly, e.g.
path.MoveTo(0, 100)
path.LineTo(100, 0)
path.LineTo(0, 100)
then acos calculation is relatively unstable. By using atan2 it avoids
some of such problems in the calculation. Additionally, it simpliflies
the round join calculation.
Fixes: https://todo.sr.ht/~eliasnaur/gio/474
Signed-off-by: Egon Elbre <egonelbre@gmail.com>
Apparently, alphaClose has been overflowing and giving the wrong answer
for a while and hence some of the tests are broken. I currently disabled
those tests, because I'm not quite sure where and how they broke.
Also, bumped alpha tolerance to 8, to ignore false positives.
Signed-off-by: Egon Elbre <egonelbre@gmail.com>
All GPU APIs except OpenGL ES 2 can generate mipmaps for textures.
This trades 33% more GPU memory use for improved rendering quality
and speed for downscaled images.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
This commit switches the priority of EGL and Vulkan so that EGL is always
tried first. This is because our EGL backend performs significantly better
than the Vulkan one, and we want the most performant experience to be the
default.
Our hypothesis is that the EGL backend is benefitting from lots of optimization
within the OpenGL driver that Vulkan drivers expect applications to perform
themselves. Rather than invest a bunch of time optimizing the Vulkan backend
right now, this change lets us focus on other priorities.
Signed-off-by: Chris Waldon <christopher.waldon.dev@gmail.com>
The Nix version of the macOS toolchain has difficulties compiling
Objective-C modules; disable modules instead of figuring out why.
It also doesn't include any frameworks automatically; add them explicitly.
While here, move suppression of OpenGL deprecation to a GL-specific
file.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
By keeping all the information in a single map, we avoid multiple
lookups and can switch between frames more easily.
Before ~35us, after ~20us for adding 50 new+old and switching
the frame.
Signed-off-by: Egon Elbre <egonelbre@gmail.com>
splitTransform func was creating multiple copies of f32.Affine2D
due to not having access to the internal and passing around non-pointer.
This makes splitTransform from ~8ns to .Split ~2ns.
It seems that the non-pointer receiver was ~0.6ns slower in this case.
Signed-off-by: Egon Elbre <egonelbre@gmail.com>
Apparently, there is a rounding error somewhere in the pipeline
from clearing a FBO to downloading its contents on at least one Apple
M1 machine. Tweak the test colors a bit to make it pass.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
Direct3D requires GPU vertex attribute structs sizes be a multiple
16. A cleanup commit removed an unusued field, and broke that
assumption.
Fixes: https://todo.sr.ht/~eliasnaur/gio/422
Signed-off-by: Elias Naur <mail@eliasnaur.com>
There are no public API that uses f32.Rectangle anymore. Move Rectangle
to an internal package for internal use.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
Like the change to op.Offset before this, clip.RRect and UniformRRect
is usually used with integer coordinates. Change to integer coordinates
to eliminate many useless conversions to float32.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
op.Offset is a convenience function most often used by layouts. Layouts
usually operate in integer coordinates, and the float32 version of op.Offset
needlessly force conversions from int to float32. This change makes op.Offset
take integer coordinates, to better match its intended use.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
OpenGL ES 2.0 doesn't support GL_PACK_ROW_LENGTH, so this change implements
a fallback using a temporary buffer.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
Creating an image.Rectangle from a f32.Rectangle is used by two packages in Gio
and about to be used for a third. Add a Round method to f32.Rectangle to avoid
duplicating the implementation.
Signed-off-by: Elias Naur <mail@eliasnaur.com>
vkDeviceWaitIdle and vkQueueWaitIdle are expensive; a vkFence is cheaper
and the usual way to ensure a previous frame has completed before starting
another.
References: https://todo.sr.ht/~eliasnaur/gio/375
Signed-off-by: Elias Naur <mail@eliasnaur.com>
Walter Werner SCHNEIDER