Commit b3859a3c by Charlie Lao Committed by Commit Bot

Vulkan: disable glFlush deferral optimization for QualComm GPUs.

It appears this optimization causing Manhattan performance regression on pixel4 with QualComm GPU. It does have a measurable performance improvement on S20+ that has ARM GPU. This disables this optimization for QualComm but still leaves it enabled on other GPUs. Bug: b/166475273 Change-Id: I9c9cd55ab169463fc8cc16d7d917be660cb2f363 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2441667Reviewed-by: 's avatarMohan Maiya <m.maiya@samsung.com> Reviewed-by: 's avatarJamie Madill <jmadill@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
parent e1d1b8b3
......@@ -1925,7 +1925,9 @@ void RendererVk::initFeatures(DisplayVk *displayVk, const ExtensionNameList &dev
ANGLE_FEATURE_CONDITION(&mFeatures, preferredLargeHeapBlockSize4MB, !isQualcomm);
ANGLE_FEATURE_CONDITION(&mFeatures, deferFlushUntilEndRenderPass, true);
// Defer glFLush call causes manhattan 3.0 perf regression. Let Qualcomm driver opt out from
// this optimization.
ANGLE_FEATURE_CONDITION(&mFeatures, deferFlushUntilEndRenderPass, !isQualcomm);
// Android mistakenly destroys the old swapchain when creating a new one.
ANGLE_FEATURE_CONDITION(&mFeatures, waitIdleBeforeSwapchainRecreation, IsAndroid() && isARM);
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment