aboutsummaryrefslogtreecommitdiff
path: root/PhysX_3.4/Samples/SampleFramework/renderer/src/RendererVertexBuffer.cpp
diff options
context:
space:
mode:
authorgit perforce import user <a@b>2016-10-25 12:29:14 -0600
committerSheikh Dawood Abdul Ajees <Sheikh Dawood Abdul Ajees>2016-10-25 18:56:37 -0500
commit3dfe2108cfab31ba3ee5527e217d0d8e99a51162 (patch)
treefa6485c169e50d7415a651bf838f5bcd0fd3bfbd /PhysX_3.4/Samples/SampleFramework/renderer/src/RendererVertexBuffer.cpp
downloadphysx-3.4-3dfe2108cfab31ba3ee5527e217d0d8e99a51162.tar.xz
physx-3.4-3dfe2108cfab31ba3ee5527e217d0d8e99a51162.zip
Initial commit:
PhysX 3.4.0 Update @ 21294896 APEX 1.4.0 Update @ 21275617 [CL 21300167]
Diffstat (limited to 'PhysX_3.4/Samples/SampleFramework/renderer/src/RendererVertexBuffer.cpp')
-rw-r--r--PhysX_3.4/Samples/SampleFramework/renderer/src/RendererVertexBuffer.cpp156
1 files changed, 156 insertions, 0 deletions
diff --git a/PhysX_3.4/Samples/SampleFramework/renderer/src/RendererVertexBuffer.cpp b/PhysX_3.4/Samples/SampleFramework/renderer/src/RendererVertexBuffer.cpp
new file mode 100644
index 00000000..c6bb4528
--- /dev/null
+++ b/PhysX_3.4/Samples/SampleFramework/renderer/src/RendererVertexBuffer.cpp
@@ -0,0 +1,156 @@
+// This code contains NVIDIA Confidential Information and is disclosed to you
+// under a form of NVIDIA software license agreement provided separately to you.
+//
+// Notice
+// NVIDIA Corporation and its licensors retain all intellectual property and
+// proprietary rights in and to this software and related documentation and
+// any modifications thereto. Any use, reproduction, disclosure, or
+// distribution of this software and related documentation without an express
+// license agreement from NVIDIA Corporation is strictly prohibited.
+//
+// ALL NVIDIA DESIGN SPECIFICATIONS, CODE ARE PROVIDED "AS IS.". NVIDIA MAKES
+// NO WARRANTIES, EXPRESSED, IMPLIED, STATUTORY, OR OTHERWISE WITH RESPECT TO
+// THE MATERIALS, AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT,
+// MERCHANTABILITY, AND FITNESS FOR A PARTICULAR PURPOSE.
+//
+// Information and code furnished is believed to be accurate and reliable.
+// However, NVIDIA Corporation assumes no responsibility for the consequences of use of such
+// information or for any infringement of patents or other rights of third parties that may
+// result from its use. No license is granted by implication or otherwise under any patent
+// or patent rights of NVIDIA Corporation. Details are subject to change without notice.
+// This code supersedes and replaces all information previously supplied.
+// NVIDIA Corporation products are not authorized for use as critical
+// components in life support devices or systems without express written approval of
+// NVIDIA Corporation.
+//
+// Copyright (c) 2008-2016 NVIDIA Corporation. All rights reserved.
+#include <RendererVertexBuffer.h>
+#include <RendererVertexBufferDesc.h>
+
+using namespace SampleRenderer;
+
+PxU32 RendererVertexBuffer::getFormatByteSize(Format format)
+{
+ PxU32 size = 0;
+ switch(format)
+ {
+ case FORMAT_FLOAT1: size = sizeof(float) * 1; break;
+ case FORMAT_FLOAT2: size = sizeof(float) * 2; break;
+ case FORMAT_FLOAT3: size = sizeof(float) * 3; break;
+ case FORMAT_FLOAT4: size = sizeof(float) * 4; break;
+ case FORMAT_UBYTE4: size = sizeof(PxU8) * 4; break;
+ case FORMAT_USHORT4: size = sizeof(PxU16) * 4; break;
+ case FORMAT_COLOR_BGRA: size = sizeof(PxU8) * 4; break;
+ case FORMAT_COLOR_RGBA: size = sizeof(PxU8) * 4; break;
+ case FORMAT_COLOR_NATIVE: size = sizeof(PxU8) * 4; break;
+ default: break;
+ }
+ RENDERER_ASSERT(size, "Unable to determine size of Format.");
+ return size;
+}
+
+RendererVertexBuffer::RendererVertexBuffer(const RendererVertexBufferDesc &desc) :
+RendererInteropableBuffer(desc.registerInCUDA, desc.interopContext),
+ m_hint(desc.hint),
+ m_deferredUnlock(desc.interopContext == NULL)
+{
+ m_maxVertices = 0;
+ m_stride = 0;
+ m_lockedBuffer = 0;
+ m_numSemanticLocks = 0;
+
+ for(PxU32 i=0; i<NUM_SEMANTICS; i++)
+ {
+ Format format = desc.semanticFormats[i];
+ if(format < NUM_FORMATS)
+ {
+ SemanticDesc &sm = m_semanticDescs[i];
+ sm.format = format;
+ sm.offset = m_stride;
+ m_stride += getFormatByteSize(format);
+ }
+ }
+}
+
+RendererVertexBuffer::~RendererVertexBuffer(void)
+{
+ RENDERER_ASSERT(m_numSemanticLocks==0, "VertexBuffer had outstanding locks during destruction!");
+}
+
+PxU32 RendererVertexBuffer::getMaxVertices(void) const
+{
+ return m_maxVertices;
+}
+
+RendererVertexBuffer::Hint RendererVertexBuffer::getHint(void) const
+{
+ return m_hint;
+}
+
+RendererVertexBuffer::Format RendererVertexBuffer::getFormatForSemantic(Semantic semantic) const
+{
+ RENDERER_ASSERT(semantic < NUM_SEMANTICS, "Invalid VertexBuffer Semantic!");
+ return m_semanticDescs[semantic].format;
+}
+
+void *RendererVertexBuffer::lockSemantic(Semantic semantic, PxU32 &stride)
+{
+ void *semanticBuffer = 0;
+ RENDERER_ASSERT(semantic < NUM_SEMANTICS, "Invalid VertexBuffer Semantic!");
+ if(semantic < NUM_SEMANTICS)
+ {
+ SemanticDesc &sm = m_semanticDescs[semantic];
+ RENDERER_ASSERT(!sm.locked, "VertexBuffer Semantic already locked.");
+ RENDERER_ASSERT(sm.format < NUM_FORMATS, "VertexBuffer does not use this semantic.");
+ if(!sm.locked && sm.format < NUM_FORMATS)
+ {
+ if(!m_lockedBuffer && !m_numSemanticLocks)
+ {
+ m_lockedBuffer = lock();
+ }
+ RENDERER_ASSERT(m_lockedBuffer, "Unable to lock VertexBuffer!");
+ if(m_lockedBuffer)
+ {
+ m_numSemanticLocks++;
+ sm.locked = true;
+ semanticBuffer = ((PxU8*)m_lockedBuffer) + sm.offset;
+ stride = m_stride;
+ }
+ }
+ }
+ return semanticBuffer;
+}
+
+void RendererVertexBuffer::unlockSemantic(Semantic semantic)
+{
+ RENDERER_ASSERT(semantic < NUM_SEMANTICS, "Invalid VertexBuffer Semantic!");
+ if(semantic < NUM_SEMANTICS)
+ {
+ SemanticDesc &sm = m_semanticDescs[semantic];
+ RENDERER_ASSERT(m_lockedBuffer && m_numSemanticLocks && sm.locked, "Trying to unlock a semantic that was not locked.");
+ if(m_lockedBuffer && m_numSemanticLocks && sm.locked)
+ {
+ if(m_lockedBuffer && semantic == SEMANTIC_COLOR)
+ {
+ swizzleColor(((PxU8*)m_lockedBuffer)+sm.offset, m_stride, m_maxVertices, sm.format);
+ }
+ sm.locked = false;
+ m_numSemanticLocks--;
+ }
+ if(m_lockedBuffer && m_numSemanticLocks == 0 && m_deferredUnlock == false)
+ {
+ unlock();
+ m_lockedBuffer = 0;
+ }
+ }
+}
+
+void RendererVertexBuffer::prepareForRender(void)
+{
+ if(m_lockedBuffer && m_numSemanticLocks == 0)
+ {
+ unlock();
+ m_lockedBuffer = 0;
+ }
+ RENDERER_ASSERT(m_lockedBuffer==0, "Vertex Buffer locked during usage!");
+}