ShuriZma
/
desmume
mirror of https://github.com/TASEmulators/desmume.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

GPU: 

- Fix possible memory corruption with display capture, at the cost of some performance. (Regression from r5243.)
- Add a couple more rules for determining if the 3D framebuffer will be read directly for display capture.
- Keep track of render states that are updated while rendering, even when the frame isn’t rendered.
This commit is contained in:
rogerman 2016-02-24 07:33:42 +00:00
parent 45f6ca71cc
commit ee5fcf6bd2
2 changed files with 114 additions and 30 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

136
desmume/src/GPU.cpp
View File

@ -927,7 +927,30 @@ void GPUEngineBase::_RenderLine_Clear(const u16 clearColor, const u16 l, u16 *ds
template<bool ISCUSTOMRENDERINGNEEDED>
void GPUEngineBase::RenderLine(const u16 l)
{
// By default, do nothing.
this->UpdatePropertiesWithoutRender(l);
}
void GPUEngineBase::UpdatePropertiesWithoutRender(const u16 l)
{
// Update BG2/BG3 parameters for Affine and AffineExt modes
if ( this->_enableLayer[GPULayerID_BG2] &&
((this->_BGLayer[GPULayerID_BG2].baseType == BGType_Affine) || (this->_BGLayer[GPULayerID_BG2].baseType == BGType_AffineExt)) )
{
IOREG_BG2Parameter &BG2Param = this->_IORegisterMap->BG2Param;
BG2Param.BG2X.value += BG2Param.BG2PB.value;
BG2Param.BG2Y.value += BG2Param.BG2PD.value;
}
if ( this->_enableLayer[GPULayerID_BG3] &&
((this->_BGLayer[GPULayerID_BG3].baseType == BGType_Affine) || (this->_BGLayer[GPULayerID_BG3].baseType == BGType_AffineExt)) )
{
IOREG_BG3Parameter &BG3Param = this->_IORegisterMap->BG3Param;
BG3Param.BG3X.value += BG3Param.BG3PB.value;
BG3Param.BG3Y.value += BG3Param.BG3PD.value;
}
}
void GPUEngineBase::FramebufferPostprocess()
@ -4093,13 +4116,13 @@ void GPUEngineA::SetCustomFramebufferSize(size_t w, size_t h)
bool GPUEngineA::WillRender3DLayer()
{
return ( this->_enableLayer[GPULayerID_BG0] && (this->_IORegisterMap->DISPCNT.BG0_Enable != 0) && (this->_IORegisterMap->DISPCNT.BG0_3D != 0) );
return ( this->_enableLayer[GPULayerID_BG0] && (this->_IORegisterMap->DISPCNT.BG0_3D != 0) );
}
bool GPUEngineA::WillCapture3DLayerDirect()
{
const IOREG_DISPCAPCNT &DISPCAPCNT = this->_IORegisterMap->DISPCAPCNT;
return ( (DISPCAPCNT.CaptureEnable != 0) && (vramConfiguration.banks[DISPCAPCNT.VRAMWriteBlock].purpose == VramConfiguration::LCDC) && (DISPCAPCNT.SrcA != 0) );
return ( (DISPCAPCNT.CaptureEnable != 0) && (DISPCAPCNT.SrcA != 0) && (DISPCAPCNT.CaptureSrc != 1) && (vramConfiguration.banks[DISPCAPCNT.VRAMWriteBlock].purpose == VramConfiguration::LCDC) );
}
template<bool ISCUSTOMRENDERINGNEEDED>
@ -4166,6 +4189,57 @@ void GPUEngineA::RenderLine(const u16 l)
}
}
void GPUEngineA::UpdatePropertiesWithoutRender(const u16 l)
{
GPUEngineBase::UpdatePropertiesWithoutRender(l);
// Update display capture properties
const IOREG_DISPCAPCNT &DISPCAPCNT = this->_IORegisterMap->DISPCAPCNT;
const u8 vramWriteBlock = DISPCAPCNT.VRAMWriteBlock;
if ((DISPCAPCNT.CaptureEnable != 0) && (vramConfiguration.banks[vramWriteBlock].purpose == VramConfiguration::LCDC) && (l < this->_dispCapCnt.capy))
{
const IOREG_DISPCNT &DISPCNT = this->_IORegisterMap->DISPCNT;
const u8 vramReadBlock = DISPCNT.VRAM_Block;
VRAM3DUsageProperties &vramUsageProperty = GPU->GetRenderProperties();
switch (DISPCAPCNT.CaptureSrc)
{
case 0: // Capture source is SourceA
vramUsageProperty.isCustomBlockUsed[vramWriteBlock] = this->isCustomRenderingNeeded;
break;
case 1: // Capture source is SourceB
{
switch (DISPCAPCNT.SrcB)
{
case 0: // Capture VRAM
vramUsageProperty.isCustomBlockUsed[vramWriteBlock] = vramUsageProperty.isCustomBlockUsed[vramReadBlock];
break;
case 1: // Capture FIFO
vramUsageProperty.isCustomBlockUsed[vramWriteBlock] = false;
break;
}
break;
}
default: // Capture source is SourceA+B blended
{
if ( (DISPCAPCNT.SrcB == 0) && vramUsageProperty.isCustomBlockUsed[vramReadBlock] )
{
vramUsageProperty.isCustomBlockUsed[vramWriteBlock] = true;
}
else
{
vramUsageProperty.isCustomBlockUsed[vramWriteBlock] = this->isCustomRenderingNeeded;
}
break;
}
}
}
}
template <bool ISCUSTOMRENDERINGNEEDED>
void GPUEngineA::_RenderLine_Layer(const u16 l, u16 *dstColorLine, const size_t dstLineWidth, const size_t dstLineCount)
{
@ -4573,7 +4647,7 @@ void GPUEngineA::_RenderLine_DispCapture_FIFOToBuffer(u16 *fifoLineBuffer)
}
template<int SOURCESWITCH, size_t CAPTURELENGTH, bool CAPTUREFROMNATIVESRC, bool CAPTURETONATIVEDST>
void GPUEngineA::_RenderLine_DispCapture_Copy(const u16 *__restrict src, u16 *__restrict dst, const size_t captureLengthExt, const size_t captureLineCount)
void GPUEngineA::_RenderLine_DispCapture_Copy(const u16 *src, u16 *dst, const size_t captureLengthExt, const size_t captureLineCount)
{
const u16 alphaBit = (SOURCESWITCH == 0) ? 0x8000 : 0x0000;
@ -4752,7 +4826,7 @@ __m128i GPUEngineA::_RenderLine_DispCapture_BlendFunc_SSE2(__m128i &srcA, __m128
#endif
template<bool CAPTUREFROMNATIVESRCA, bool CAPTUREFROMNATIVESRCB>
void GPUEngineA::_RenderLine_DispCapture_BlendToCustomDstBuffer(const u16 *__restrict srcA, const u16 *__restrict srcB, u16 *__restrict dst, const u8 blendEVA, const u8 blendEVB, const size_t length, size_t l)
void GPUEngineA::_RenderLine_DispCapture_BlendToCustomDstBuffer(const u16 *srcA, const u16 *srcB, u16 *dst, const u8 blendEVA, const u8 blendEVB, const size_t length, size_t l)
{
#ifdef ENABLE_SSE2
const __m128i blendEVA_vec128 = _mm_set1_epi16(blendEVA);
@ -4799,7 +4873,7 @@ void GPUEngineA::_RenderLine_DispCapture_BlendToCustomDstBuffer(const u16 *__res
}
template<size_t CAPTURELENGTH, bool CAPTUREFROMNATIVESRCA, bool CAPTUREFROMNATIVESRCB, bool CAPTURETONATIVEDST>
void GPUEngineA::_RenderLine_DispCapture_Blend(const u16 *__restrict srcA, const u16 *__restrict srcB, u16 *__restrict dst, const size_t captureLengthExt, const size_t l)
void GPUEngineA::_RenderLine_DispCapture_Blend(const u16 *srcA, const u16 *srcB, u16 *dst, const size_t captureLengthExt, const size_t l)
{
const u8 blendEVA = GPU->GetEngineMain()->_dispCapCnt.EVA;
const u8 blendEVB = GPU->GetEngineMain()->_dispCapCnt.EVB;
@ -5728,6 +5802,9 @@ void GPUSubsystem::SetWillAutoResolveToCustomBuffer(const bool willAutoResolve)
void GPUSubsystem::RenderLine(const u16 l, bool isFrameSkipRequested)
{
const bool isFramebuffeRenderNeeded[2] = { CommonSettings.showGpu.main && !(this->_engineMain->GetIsMasterBrightFullIntensity() && (this->_engineMain->GetIORegisterMap().DISPCAPCNT.CaptureEnable == 0)),
CommonSettings.showGpu.sub && !this->_engineSub->GetIsMasterBrightFullIntensity() };
if (l == 0)
{
CurrentRenderer->SetFramebufferFlushStates(this->_engineMain->WillRender3DLayer(), this->_engineMain->WillCapture3DLayerDirect());
@ -5760,43 +5837,48 @@ void GPUSubsystem::RenderLine(const u16 l, bool isFrameSkipRequested)
}
}
if (!isFrameSkipRequested)
if (isFramebuffeRenderNeeded[GPUEngineID_Main] && !isFrameSkipRequested)
{
if (CommonSettings.showGpu.main && !(this->_engineMain->GetIsMasterBrightFullIntensity() && (this->_engineMain->GetIORegisterMap().DISPCAPCNT.CaptureEnable == 0)))
if (this->_engineMain->isCustomRenderingNeeded)
{
if (this->_engineMain->isCustomRenderingNeeded)
{
this->_engineMain->RenderLine<true>(l);
}
else
{
this->_engineMain->RenderLine<false>(l);
}
this->_engineMain->RenderLine<true>(l);
}
if (CommonSettings.showGpu.sub && !this->_engineSub->GetIsMasterBrightFullIntensity())
else
{
if (this->_engineSub->isCustomRenderingNeeded)
{
this->_engineSub->RenderLine<true>(l);
}
else
{
this->_engineSub->RenderLine<false>(l);
}
this->_engineMain->RenderLine<false>(l);
}
}
else
{
this->_engineMain->UpdatePropertiesWithoutRender(l);
}
if (isFramebuffeRenderNeeded[GPUEngineID_Sub] && !isFrameSkipRequested)
{
if (this->_engineSub->isCustomRenderingNeeded)
{
this->_engineSub->RenderLine<true>(l);
}
else
{
this->_engineSub->RenderLine<false>(l);
}
}
else
{
this->_engineSub->UpdatePropertiesWithoutRender(l);
}
if (l == 191)
{
if (!isFrameSkipRequested)
{
if (CommonSettings.showGpu.main && !(this->_engineMain->GetIsMasterBrightFullIntensity() && (this->_engineMain->GetIORegisterMap().DISPCAPCNT.CaptureEnable == 0)))
if (isFramebuffeRenderNeeded[GPUEngineID_Main])
{
this->_engineMain->ApplyMasterBrightness<false>();
}
if (CommonSettings.showGpu.sub && !this->_engineSub->GetIsMasterBrightFullIntensity())
if (isFramebuffeRenderNeeded[GPUEngineID_Sub])
{
this->_engineSub->ApplyMasterBrightness<false>();
}

8
desmume/src/GPU.h
View File

@ -1304,6 +1304,7 @@ public:
template<GPUEngineID ENGINEID> void ParseAllRegisters();
template<bool ISCUSTOMRENDERINGNEEDED> void RenderLine(const u16 l);
void UpdatePropertiesWithoutRender(const u16 l);
void FramebufferPostprocess();
bool isCustomRenderingNeeded;
@ -1377,7 +1378,7 @@ protected:
void _RenderLine_DispCapture_FIFOToBuffer(u16 *fifoLineBuffer);
template<int SOURCESWITCH, size_t CAPTURELENGTH, bool CAPTUREFROMNATIVESRC, bool CAPTURETONATIVEDST>
void _RenderLine_DispCapture_Copy(const u16 *__restrict src, u16 *__restrict dst, const size_t captureLengthExt, const size_t captureLineCount);
void _RenderLine_DispCapture_Copy(const u16 *src, u16 *dst, const size_t captureLengthExt, const size_t captureLineCount); // Do not use restrict pointers, since src and dst can be the same
u16 _RenderLine_DispCapture_BlendFunc(const u16 srcA, const u16 srcB, const u8 blendEVA, const u8 blendEVB);
@ -1386,10 +1387,10 @@ protected:
#endif
template<bool CAPTUREFROMNATIVESRCA, bool CAPTUREFROMNATIVESRCB>
void _RenderLine_DispCapture_BlendToCustomDstBuffer(const u16 *__restrict srcA, const u16 *__restrict srcB, u16 *__restrict dst, const u8 blendEVA, const u8 blendEVB, const size_t length, size_t l);
void _RenderLine_DispCapture_BlendToCustomDstBuffer(const u16 *srcA, const u16 *srcB, u16 *dst, const u8 blendEVA, const u8 blendEVB, const size_t length, size_t l); // Do not use restrict pointers, since srcB and dst can be the same
template<size_t CAPTURELENGTH, bool CAPTUREFROMNATIVESRCA, bool CAPTUREFROMNATIVESRCB, bool CAPTURETONATIVEDST>
void _RenderLine_DispCapture_Blend(const u16 *__restrict srcA, const u16 *__restrict srcB, u16 *__restrict dst, const size_t captureLengthExt, const size_t l);
void _RenderLine_DispCapture_Blend(const u16 *srcA, const u16 *srcB, u16 *dst, const size_t captureLengthExt, const size_t l); // Do not use restrict pointers, since srcB and dst can be the same
template<bool ISCUSTOMRENDERINGNEEDED> void _HandleDisplayModeVRAM(u16 *__restrict dstColorLine, const size_t l, const size_t dstLineWidth, const size_t dstLineCount);
void _HandleDisplayModeMainMemory(u16 *dstColorLine);
@ -1410,6 +1411,7 @@ public:
bool WillCapture3DLayerDirect();
template<bool ISCUSTOMRENDERINGNEEDED> void RenderLine(const u16 l);
void UpdatePropertiesWithoutRender(const u16 l);
void FramebufferPostprocess();
};