piwalker/dolphin
1
0
Fork 0
mirror of https://github.com/dolphin-emu/dolphin.git synced 2024-11-14 21:37:52 -07:00
Code Issues Actions Packages Projects Releases Wiki Activity

OGL: Force memory barrier when reading back bounding box values

Browse Source 
We also need to ensure the the CPU does not receive stale values
which have been updated by the GPU. Apparently the buffer here
is not coherent on NVIDIA drivers. Not sure if this is a driver
bug/spec violation or not, one would think that
glGetBufferSubData() would invalidate any caches as needed, but
this path is only used on NVIDIA anyway, so it's fine. A point
to note is that according to ARB_debug_report, it's moved from
video to host memory, which would explain why it needs the
cache invalidate.
This commit is contained in:
Connor McLaughlin 2021-05-29 16:50:33 +10:00 committed by Stenzek
parent 7b2b05cc21
commit 6393f157f0
1 changed files with 12 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
Source/Core/VideoBackends/OGL/OGLBoundingBox.cpp
View File

@ -87,12 +87,21 @@ void BoundingBox::Flush()
void BoundingBox::Readback()
{
glBindBuffer(GL_SHADER_STORAGE_BUFFER, s_bbox_buffer_id);
// Using glMapBufferRange to read back the contents of the SSBO is extremely slow
// on nVidia drivers. This is more noticeable at higher internal resolutions.
// Using glGetBufferSubData instead does not seem to exhibit this slowdown.
if (!DriverDetails::HasBug(DriverDetails::BUG_SLOW_GETBUFFERSUBDATA) &&
!static_cast<Renderer*>(g_renderer.get())->IsGLES())
{
// Using glMapBufferRange to read back the contents of the SSBO is extremely slow
// on nVidia drivers. This is more noticeable at higher internal resolutions.
// Using glGetBufferSubData instead does not seem to exhibit this slowdown.
// We also need to ensure the the CPU does not receive stale values which have been updated by
// the GPU. Apparently the buffer here is not coherent on NVIDIA drivers. Not sure if this is a
// driver bug/spec violation or not, one would think that glGetBufferSubData() would invalidate
// any caches as needed, but this path is only used on NVIDIA anyway, so it's fine. A point to
// note is that according to ARB_debug_report, it's moved from video to host memory, which would
// explain why it needs the cache invalidate.
glMemoryBarrier(GL_SHADER_STORAGE_BARRIER_BIT);
std::array<s32, NUM_BBOX_VALUES> gpu_values;
glGetBufferSubData(GL_SHADER_STORAGE_BUFFER, 0, sizeof(s32) * NUM_BBOX_VALUES,
gpu_values.data());