generic-library/opencv
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

fixes for gpu module:

Browse Source 
- fixed printCudaDeviceInfo for new CC
- fixed some compilation errors and warnings
- removed unset command from CMake script
- removed unused std imports
This commit is contained in:
Vladislav Vinogradov
2013-01-23 13:59:14 +04:00
parent b7e6b5af1b
commit ae6266e101
32 changed files with 232 additions and 144 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
modules/core/src/gpumat.cpp
View File

@@ -315,18 +315,6 @@ void cv::gpu::DeviceInfo::queryMemory(size_t& free_memory, size_t& total_memory)
namespace
{
template <class T> void getCudaAttribute(T *attribute, CUdevice_attribute device_attribute, int device)
{
*attribute = T();
//CUresult error = CUDA_SUCCESS;// = cuDeviceGetAttribute( attribute, device_attribute, device ); why link erros under ubuntu??
CUresult error = cuDeviceGetAttribute( attribute, device_attribute, device );
if( CUDA_SUCCESS == error )
return;
printf("Driver API error = %04d\n", error);
cv::gpu::error("driver API error", __FILE__, __LINE__);
}
int convertSMVer2Cores(int major, int minor)
{
// Defines for GPU Architecture types (using the SM version to determine the # of cores per SM
@@ -335,7 +323,7 @@ namespace
int Cores;
} SMtoCores;
SMtoCores gpuArchCoresPerSM[] = { { 0x10, 8 }, { 0x11, 8 }, { 0x12, 8 }, { 0x13, 8 }, { 0x20, 32 }, { 0x21, 48 }, {0x30, 192}, { -1, -1 } };
SMtoCores gpuArchCoresPerSM[] = { { 0x10, 8 }, { 0x11, 8 }, { 0x12, 8 }, { 0x13, 8 }, { 0x20, 32 }, { 0x21, 48 }, {0x30, 192}, {0x35, 192}, { -1, -1 } };
int index = 0;
while (gpuArchCoresPerSM[index].SM != -1)
@@ -344,7 +332,7 @@ namespace
return gpuArchCoresPerSM[index].Cores;
index++;
}
printf("MapSMtoCores undefined SMversion %d.%d!\n", major, minor);
return -1;
}
}
@@ -382,22 +370,13 @@ void cv::gpu::printCudaDeviceInfo(int device)
printf(" CUDA Driver Version / Runtime Version %d.%d / %d.%d\n", driverVersion/1000, driverVersion%100, runtimeVersion/1000, runtimeVersion%100);
printf(" CUDA Capability Major/Minor version number: %d.%d\n", prop.major, prop.minor);
printf(" Total amount of global memory: %.0f MBytes (%llu bytes)\n", (float)prop.totalGlobalMem/1048576.0f, (unsigned long long) prop.totalGlobalMem);
printf(" (%2d) Multiprocessors x (%2d) CUDA Cores/MP: %d CUDA Cores\n",
prop.multiProcessorCount, convertSMVer2Cores(prop.major, prop.minor),
convertSMVer2Cores(prop.major, prop.minor) * prop.multiProcessorCount);
int cores = convertSMVer2Cores(prop.major, prop.minor);
if (cores > 0)
printf(" (%2d) Multiprocessors x (%2d) CUDA Cores/MP: %d CUDA Cores\n", prop.multiProcessorCount, cores, cores * prop.multiProcessorCount);
printf(" GPU Clock Speed: %.2f GHz\n", prop.clockRate * 1e-6f);
// This is not available in the CUDA Runtime API, so we make the necessary calls the driver API to support this for output
int memoryClock, memBusWidth, L2CacheSize;
getCudaAttribute<int>( &memoryClock, CU_DEVICE_ATTRIBUTE_MEMORY_CLOCK_RATE, dev );
getCudaAttribute<int>( &memBusWidth, CU_DEVICE_ATTRIBUTE_GLOBAL_MEMORY_BUS_WIDTH, dev );
getCudaAttribute<int>( &L2CacheSize, CU_DEVICE_ATTRIBUTE_L2_CACHE_SIZE, dev );
printf(" Memory Clock rate: %.2f Mhz\n", memoryClock * 1e-3f);
printf(" Memory Bus Width: %d-bit\n", memBusWidth);
if (L2CacheSize)
printf(" L2 Cache Size: %d bytes\n", L2CacheSize);
printf(" Max Texture Dimension Size (x,y,z) 1D=(%d), 2D=(%d,%d), 3D=(%d,%d,%d)\n",
prop.maxTexture1D, prop.maxTexture2D[0], prop.maxTexture2D[1],
prop.maxTexture3D[0], prop.maxTexture3D[1], prop.maxTexture3D[2]);
@@ -457,7 +436,12 @@ void cv::gpu::printShortCudaDeviceInfo(int device)
const char *arch_str = prop.major < 2 ? " (not Fermi)" : "";
printf("Device %d: \"%s\" %.0fMb", dev, prop.name, (float)prop.totalGlobalMem/1048576.0f);
printf(", sm_%d%d%s, %d cores", prop.major, prop.minor, arch_str, convertSMVer2Cores(prop.major, prop.minor) * prop.multiProcessorCount);
printf(", sm_%d%d%s", prop.major, prop.minor, arch_str);
int cores = convertSMVer2Cores(prop.major, prop.minor);
if (cores > 0)
printf(", %d cores", cores * prop.multiProcessorCount);
printf(", Driver/Runtime ver.%d.%d/%d.%d\n", driverVersion/1000, driverVersion%100, runtimeVersion/1000, runtimeVersion%100);
}
fflush(stdout);