AI/rocm-systems
2
0
Derivar 0
Código Questões Pedidos de integração Operações Pacotes Planeamentos Lançamentos Wiki Trabalho

Merge pull request #748 from mkuron/getsymboladdress

Ver a proveniência 
Implement hipGetSymbolAddress and hipGetSymbolSize

[ROCm/clr commit: 40d3184dd1]
Este cometimento está contido em:
Maneesh Gupta
2018-11-21 10:32:01 +05:30
cometido por GitHub
ascendente c73ef1b0b9 4b77b5c1ba
cometimento f0976b3c37
9 ficheiros modificados com 93 adições e 11 eliminações
Descarregar ficheiro patch Descarregar ficheiro diff Expandir todos os ficheiros Colapsar todos os ficheiros
projects/clr/hipamd/docs/markdown/CUDA_Runtime_API_functions_supported_by_HIP.md
+4 -4
Ver ficheiro
@@ -132,8 +132,8 @@
| `cudaFreeHost` | `hipHostFree` |
| `cudaFreeMipmappedArray` | |
| `cudaGetMipmappedArrayLevel` | |
| `cudaGetSymbolAddress` | |
| `cudaGetSymbolSize` | |
| `cudaGetSymbolAddress` | `hipGetSymbolAddress` |
| `cudaGetSymbolSize` | `hipGetSymbolSize` |
| `cudaHostAlloc` | `hipHostMalloc` |
| `cudaHostGetDevicePointer` | `hipHostGetDevicePointer` |
| `cudaHostGetFlags` | `hipHostGetFlags` |
@@ -383,8 +383,8 @@
| `cudaCreateChannelDesc` | `hipCreateChannelDesc` |
| `cudaFuncGetAttributes` | |
| `cudaFuncSetCacheConfig` | |
| `cudaGetSymbolAddress` | |
| `cudaGetSymbolSize` | |
| `cudaGetSymbolAddress` | `hipGetSymbolAddress` |
| `cudaGetSymbolSize` | `hipGetSymbolSize` |
| `cudaGetTextureAlignmentOffset` | |
| `cudaLaunch` | |
| `cudaLaunchKernel` | |
projects/clr/hipamd/docs/markdown/hip_kernel_language.md
+1 -1
Ver ficheiro
@@ -159,7 +159,7 @@ void callMyKernel()
## Variable-Type Qualifiers
### `__constant__`
The `__constant__` keyword is supported. The host writes constant memory before launching the kernel; from the GPU, this memory is read-only during kernel execution. The functions for accessing constant memory (hipGetSymbolAddress(), hipGetSymbolSize(), hipMemcpyToSymbol(), hipMemcpyToSymbolAsync, hipMemcpyFromSymbol, hipMemcpyFromSymbolAsync) are under development.
The `__constant__` keyword is supported. The host writes constant memory before launching the kernel; from the GPU, this memory is read-only during kernel execution. The functions for accessing constant memory (hipGetSymbolAddress(), hipGetSymbolSize(), hipMemcpyToSymbol(), hipMemcpyToSymbolAsync(), hipMemcpyFromSymbol(), hipMemcpyFromSymbolAsync()) are available.
### `__shared__`
The `__shared__` keyword is supported.