KJ16609/gcc
1
0
Fork 0
mirror of https://gcc.gnu.org/git/gcc.git synced 2026-02-21 19:35:28 -05:00
Code Issues Packages Projects Releases Wiki Activity

libgomp, amdgcn, nvptx: Improve omp_target_is_accessible [PR121813]

Browse Source 
This patch extends omp_target_is_accessible to check the actual device status
for the memory region, on amdgcn and nvptx devices (rather than just checking
if shared memory is enabled).

In both cases, we check the status of each 4k region within the given memory
range (assuming 4k pages should be safe for all the currently supported hosts)
and returns true if all of the pages report accessible.

The testcases have been modified to check that allocations marked accessible
actually are accessible (inaccessibility can't be checked without invoking
memory faults), and to understand that some parts of an array can be accessible
but other parts not (I have observed this intermittently for the stack memory
on amdgcn using the Fortran testcase, which can have the allocation span pages).

There's also new testcases for the various other memory modes, and for managed
memory.

include/ChangeLog:

	* cuda/cuda.h (CUpointer_attribute): New enum.
	(cuPointerGetAttribute): New prototype.

libgomp/ChangeLog:

	PR libgomp/121813
	PR libgomp/113213
	* libgomp-plugin.h (GOMP_OFFLOAD_is_accessible_ptr): New prototype.
	* libgomp.h
	(struct gomp_device_descr): Add GOMP_OFFLOAD_is_accessible_ptr.
	* libgomp.texi: Update omp_target_is_accessible docs.
	* plugin/cuda-lib.def (cuPointerGetAttribute): New entry.
	* plugin/plugin-gcn.c (struct hsa_runtime_fn_info): Add
	hsa_amd_svm_attributes_get_fn and hsa_amd_pointer_info_fn.
	(init_hsa_runtime_functions): Add hsa_amd_svm_attributes_get and
	hsa_amd_pointer_info.
	(enum accessible): New enum type.
	(host_memory_is_accessible): New function.
	(device_memory_is_accessible): New function.
	(GOMP_OFFLOAD_is_accessible_ptr): New function.
	* plugin/plugin-nvptx.c (GOMP_OFFLOAD_is_accessible_ptr): Likewise.
	* target.c (omp_target_is_accessible): Call is_accessible_ptr_func.
	(gomp_load_plugin_for_device): Add is_accessible_ptr.
	* testsuite/libgomp.c-c++-common/target-is-accessible-1.c: Rework
	to match more details of the GPU implementation.
	* testsuite/libgomp.fortran/target-is-accessible-1.f90: Likewise.
	* testsuite/libgomp.c-c++-common/target-is-accessible-2.c: New test.
	* testsuite/libgomp.c-c++-common/target-is-accessible-3.c: New test.
	* testsuite/libgomp.c-c++-common/target-is-accessible-4.c: New test.
	* testsuite/libgomp.c-c++-common/target-is-accessible-5.c: New test.
This commit is contained in:
Andrew Stubbs
2025-12-08 16:18:02 +00:00
parent 0ef6df1db1
commit a53c5c42c6
14 changed files with 423 additions and 41 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
include/cuda/cuda.h
View File

@@ -143,6 +143,13 @@ typedef enum {
CU_MEMORYTYPE_UNIFIED = 0x04
} CUmemorytype;
typedef enum {
CU_POINTER_ATTRIBUTE_CONTEXT = 0x01,
CU_POINTER_ATTRIBUTE_MEMORY_TYPE = 0x02,
CU_POINTER_ATTRIBUTE_DEVICE_POINTER = 0x03,
CU_POINTER_ATTRIBUTE_HOST_POINTER = 0x04
} CUpointer_attribute;
typedef struct {
size_t srcXInBytes, srcY;
CUmemorytype srcMemoryType;
@@ -300,6 +307,8 @@ CUresult cuModuleGetGlobal (CUdeviceptr *, size_t *, CUmodule, const char *);
CUresult cuModuleLoad (CUmodule *, const char *);
CUresult cuModuleLoadData (CUmodule *, const void *);
CUresult cuModuleUnload (CUmodule);
CUresult cuPointerGetAttribute (CUmemorytype *, CUpointer_attribute,
CUdeviceptr);
CUresult cuOccupancyMaxPotentialBlockSize(int *, int *, CUfunction,
CUoccupancyB2DSize, size_t, int);
typedef void (*CUstreamCallback)(CUstream, CUresult, void *);