|
c26011b8
|
2023-04-11T15:16:29
|
|
Vulkan: Make advanced blend alpha div workaround permanent
This workaround turned out to affect pretty much every driver, and
numerous GL implementations were found to work around it similarly to
ANGLE. It seems like this workaround may only be necessary for
colorburn and colordodge, but for now ANGLE applies it to all modes.
Bug: b/274528004
Bug: b/277777623
Change-Id: Id555c981a9775f949a3022b7e92c755accea7cea
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4416158
Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
|
|
|
6f80f0f0
|
2022-08-06T02:29:19
|
|
Translator: Clean up the compile flag passing interface
Historically, compile flags were sent to the translator as a bitmask.
Recently, we were getting close to running out of bits. Additionally,
direct-to-metal work had started to introduce constants to be passed to
the translator, which were misplaced in ShBuiltInResources and Caps.
Recent work on Pixel Local Storage adds even more constants, aggravating
the situation.
In this change, the interface to passing compile flags is reworked. A
struct is passed (instead of a bitmask) that has one bit for each flag.
This can be indefinitely extended. Additionally, the constants needed
by metal and PLS are also placed in this struct. In turn, the backends
can set these options directly, and don't have to hack them into Caps to
further get hacked into ShBuiltInResources.
Bug: angleproject:7559
Change-Id: If93f1e1b8818ad3a0ac708ab04ab93b4b397d114
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3812562
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com>
Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com>
Auto-Submit: Shahbaz Youssefi <syoussefi@chromium.org>
|
|
861149c7
|
2022-08-03T15:43:29
|
|
Make PLS coherent on desktop OpenGL
Implements ANGLE_shader_pixel_local_storage_coherent using fragment
shader synchronization extensions:
NV_fragment_shader_interlock
INTEL_fragment_shader_ordering
ARB_fragment_shader_interlock
With these extensions combined, we get coherency all 3 big desktop
vendors: NVIDIA, Intel, and AMD.
Bug: angleproject:7279
Change-Id: Ie20b251fb772898e89994b799640f1f2806581eb
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3773990
Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
|
|
9070e279
|
2022-06-30T09:35:59
|
|
Vulkan: Handle platforms that have imprecise float division issues
Even when the dividend and divisor have the same value
some platforms do not return 1.0f. We need to emit sepcial
division code on such platforms.
Bug: angleproject:3586
Tests: KHR-GLES31.core.blend_equation_advanced.blend*HSL*
Change-Id: Ibb81f57c959ae2c8edb6d361bf6241c019dfb988
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3738439
Reviewed-by: Charlie Lao <cclao@google.com>
Commit-Queue: mohan maiya <m.maiya@samsung.com>
Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
|
|
|
2df17a12
|
2022-05-25T16:18:37
|
|
Vulkan: Pack driver uniforms
Previously 5 vec4s were used for driver uniforms + 2 vec4s if
specialization constants couldn't be supported.
The driver uniforms are rearranged and packed such that only 2 vec4s are
normally used, which include fallback for specialization constants as
well. In the future, most of the specialization constants may turn into
uniforms, and this change prepares for that.
Additional uniforms are used (3 vec4s) only if common extensions are
missing; transform feedback and bresenham lines.
This change makes it more practical for driver uniforms to be turned
into push constants. Additionally, these uniforms could potentially be
loaded and cached at the beginning of the shader for more efficient
memory access.
On Pixel6, with this change, the traces show no difference in wall time.
On most traces, CPU time shows up to ~7% improvement.
Bug: angleproject:7366
Change-Id: I0f47f863955af06a19c69d1f1d7c45b97d95476e
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3668151
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Reviewed-by: Charlie Lao <cclao@google.com>
|
|
|
98c2e169
|
2022-05-20T16:17:49
|
|
Vulkan: Reduce pre-rotation spec const to bool
The specialization constant now only dictates whether x and y should be
swapped. The complete 8 possible states of rotation and y-flip are
achieved by using this swap in combination with a driver uniform for x
and y flip.
Swapping is still a specialization constant to avoid degrading
performance of dFdx/dFdy which otherwise would need both to be evaluated
instead of one. On platforms which don't support pre-rotation, the
specialization constant will never change and driver uniforms entirely
govern y-flip. On platforms that do support pre-rotation, only two
variations of the pipeline are needed.
Bug: angleproject:7366
Change-Id: I73f84e89fa9349d2098fa5b21573aee57d93a30c
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3663151
Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Lingfeng Yang <lfy@google.com>
Reviewed-by: Charlie Lao <cclao@google.com>
|
|
|
62fe36d3
|
2022-05-04T22:46:52
|
|
Vulkan: Emulate YUV built-ins
In this change, the rgb_2_yuv and yuv_2_rgb built-ins are emulated with
normal functions that perform matrix multiplication based on the given
conversion function.
Bug: angleproject:6818
Change-Id: I67adb029109aaf6a674b1ee75105c1b352325eb2
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3630599
Reviewed-by: Shahbaz Youssefi <syoussefi@google.com>
Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
|
c5271e8e
|
2022-04-04T23:28:35
|
|
Vulkan: Emulate GL_KHR_blend_equation_advanced
Based on a change by Brandon Schade <b.schade@samsung.com>
In the translator, when advanced blend is enabled, an input attachment
is added. Based on the listed advanced blend equations, emulation code
is added that performs those equations' functions. The blend equation
itself is passed through a driver uniform.
Note that the advanced blend extension only allows a single output to
use advanced blend, and that should be at location 0.
In the Vulkan backend, when advanced blend is used, the driver uniform
to select the equation is updated and normal blending is disabled.
Bug: angleproject:3586
Change-Id: Icc42e8be238d34fca149087eb9cfe616a7643a6b
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3575738
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
|