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

P4 to Git Change 1323915 by gandryey@gera-w8 on 2016/10/07 12:59:30

Browse Source 
SWDEV-104441 - [SSG] OpenCL has not implemented the asynchronous transfer
	- Use lock protection for multiple maps of persistent memory
	- Don't mark persistent as host mem
	- Implement file write for invisible memory

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpumemory.cpp#129 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuresource.cpp#236 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#408 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.cpp#13 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/command.cpp#77 edit
This commit is contained in:
foreman
2016-10-07 13:07:19 -04:00
parent e8488126a8
commit 9d739c200c
5 changed files with 67 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files
rocclr/runtime/device/gpu/gpuresource.cpp
+3 -1
View File
@@ -1567,7 +1567,9 @@ Resource::map(VirtualGPU* gpu, uint flags, uint startLayer, uint numLayers)
// guarantee that the address will be valid,
// since GSL could still process the first map
if (address_ == NULL) {
amd::Os::sleep(10);
for (uint i = 0; address_ == NULL && i < 10; ++i) {
amd::Os::sleep(1);
}
assert((address_ != NULL) && "Multiple maps failed!");
}