D3D11Device: Fix GPU usage without vsync
This commit is contained in:
parent
b53d941c3a
commit
373721af79
|
@ -612,7 +612,7 @@ bool D3D11Device::BeginPresent(bool skip_present)
|
|||
if (!m_swap_chain)
|
||||
{
|
||||
// Note: Really slow on Intel...
|
||||
//m_context->Flush();
|
||||
m_context->Flush();
|
||||
TrimTexturePool();
|
||||
return false;
|
||||
}
|
||||
|
@ -650,7 +650,7 @@ void D3D11Device::EndPresent(bool explicit_present)
|
|||
DebugAssert(!explicit_present);
|
||||
DebugAssert(m_num_current_render_targets == 0 && !m_current_depth_target);
|
||||
|
||||
if (m_vsync_enabled && m_gpu_timing_enabled)
|
||||
if (!m_vsync_enabled && m_gpu_timing_enabled)
|
||||
PopTimestampQuery();
|
||||
|
||||
// DirectX has no concept of tear-or-sync. I guess if we measured times ourselves, we could implement it.
|
||||
|
|
Loading…
Reference in New Issue