AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Support IPC sharing of non paged system memory

Browse Source 
Non paged system memory is allocated with node id 0. However, since a
gpu node is required for allocating system memory via KFD, the first
dgpu is used. In hsaKmtShareMemory() if system memory use the same
(first) dgpu.

Change-Id: I85789a89a4e4f7888e3826826401ea89ce4d1718
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>


[ROCm/ROCR-Runtime commit: 186527d0b7]
This commit is contained in:
Harish Kasiviswanathan
2017-07-24 15:12:24 -04:00
parent 8581bb285d
commit faa5102340
1 changed files with 8 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/rocr-runtime/src/fmm.c
+8 -1
View File
@@ -2597,7 +2597,14 @@ HSAKMT_STATUS fmm_share_memory(void *MemoryAddress,
r = validate_nodeid(obj->node_id, &gpu_id);
if (r != HSAKMT_STATUS_SUCCESS)
return r;
if (!gpu_id && is_dgpu) {
/* Sharing non paged system memory. Use first dgpu which was
* used during allocation. See fmm_allocate_host_gpu()
*/
r = find_first_dgpu(&gpu_id);
if (r != HSAKMT_STATUS_SUCCESS)
return r;
}
exportArgs.handle = obj->handle;
exportArgs.gpu_id = gpu_id;