OSSForks/tensorflow
1
0
Fork 0
mirror of https://github.com/zebrajr/tensorflow.git synced 2025-12-06 12:20:11 +01:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

[XLA:GPU]: Add peer to peer copies for cupti tracing.

Browse Source 
Before this change peer to peer copies done using
cuMemcpyPeerAsync was not being tracked with the driver API.

PiperOrigin-RevId: 825040246
This commit is contained in:
Sohaib Iftikhar 2025-10-28 07:57:28 -07:00 committed by TensorFlower Gardener
parent 034d750525
commit 542ffe0410
1 changed files with 2 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
third_party/xla/xla/backends/profiler/gpu/cupti_tracer.cc vendored
View File

@ -1225,6 +1225,8 @@ CuptiTracer::CreateDefaultCallbackIds() {
CUPTI_DRIVER_TRACE_CBID_cuMemcpy3DAsync_v2,
CUPTI_DRIVER_TRACE_CBID_cuMemcpyHtoA_v2,
CUPTI_DRIVER_TRACE_CBID_cuMemcpyHtoAAsync_v2,
CUPTI_DRIVER_TRACE_CBID_cuMemcpyPeer,
CUPTI_DRIVER_TRACE_CBID_cuMemcpyPeerAsync,
// MemAlloc
CUPTI_DRIVER_TRACE_CBID_cuMemAlloc_v2,
CUPTI_DRIVER_TRACE_CBID_cuMemAllocPitch_v2,