Scarfski/tracy
1
0
Fork 0
mirror of https://github.com/wolfpld/tracy.git synced 2026-06-08 08:33:48 +00:00
Code Issues Packages Projects Releases Wiki Activity

Remove memory API calls from cbid tracker maps

Browse Source 
cudaMalloc/cudaFree (and driver equivalents) were tracked in
cbidRuntimeTrackers/cbidDriverTrackers, creating a cudaCallSiteInfo
entry on each API call. But the MEMORY2 handler never calls
matchActivityToAPICall (and never calls EmitGpuZone) — it only needs
the address, size, and timestamp from the activity record itself. Since
no activity handler consumes these entries, they leaked indefinitely.

Remove the 6 memory API CBIDs from both tracker maps so no entry is
created. This eliminates the leak with no change in visible behavior:
the MEMORY2 handler already operates independently of cudaCallSiteInfo.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Basil Milanich
2026-04-13 09:05:58 -05:00
parent ec7805fd92
commit b22c8e86ba
1 changed files with 5 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
public/tracy/TracyCUDA.hpp
View File

@@ -819,9 +819,10 @@ namespace tracy
{ CUPTI_RUNTIME_TRACE_CBID_cudaLaunchKernel_ptsz_v7000, GET_STREAM_FUNC(cudaLaunchKernel_ptsz_v7000_params, stream) },
{ CUPTI_RUNTIME_TRACE_CBID_cudaLaunchKernelExC_v11060, GET_STREAM_FUNC(cudaLaunchKernelExC_v11060_params, config->stream) },
{ CUPTI_RUNTIME_TRACE_CBID_cudaLaunchKernelExC_ptsz_v11060, GET_STREAM_FUNC(cudaLaunchKernelExC_ptsz_v11060_params, config->stream) },
// Runtime: Memory
{ CUPTI_RUNTIME_TRACE_CBID_cudaMalloc_v3020, NON_STREAM_FUNC() },
{ CUPTI_RUNTIME_TRACE_CBID_cudaFree_v3020, NON_STREAM_FUNC() },
// Runtime: Memory — NOT tracked here. The MEMORY2 handler
// only needs address/size/timestamp from the activity record
// and never calls EmitGpuZone, so there is no consumer for
// the cudaCallSiteInfo entry. Tracking them would leak entries.
// Runtime: Memcpy
{ CUPTI_RUNTIME_TRACE_CBID_cudaMemcpy_v3020, NON_STREAM_FUNC() },
{ CUPTI_RUNTIME_TRACE_CBID_cudaMemcpyAsync_v3020, GET_STREAM_FUNC(cudaMemcpyAsync_v3020_params, stream) },
@@ -856,11 +857,7 @@ namespace tracy
{ CUPTI_DRIVER_TRACE_CBID_cuLaunchKernel_ptsz, GET_STREAM_FUNC(cuLaunchKernel_ptsz_params, hStream)} ,
{ CUPTI_DRIVER_TRACE_CBID_cuLaunchKernelEx, GET_STREAM_FUNC(cuLaunchKernelEx_params, config->hStream) },
{ CUPTI_DRIVER_TRACE_CBID_cuLaunchKernelEx_ptsz, GET_STREAM_FUNC(cuLaunchKernelEx_params, config->hStream) },
// Driver: Memory
{ CUPTI_DRIVER_TRACE_CBID_cuMemAlloc_v2, NON_STREAM_FUNC() },
{ CUPTI_DRIVER_TRACE_CBID_cuMemAllocManaged, NON_STREAM_FUNC() },
{ CUPTI_DRIVER_TRACE_CBID_cuMemAllocPitch_v2, NON_STREAM_FUNC() },
{ CUPTI_DRIVER_TRACE_CBID_cuMemFree_v2, NON_STREAM_FUNC() },
// Driver: Memory — NOT tracked (see Runtime: Memory comment above).
// Driver: Memcpy - Synchronous
{ CUPTI_DRIVER_TRACE_CBID_cuMemcpy, NON_STREAM_FUNC() },
{ CUPTI_DRIVER_TRACE_CBID_cuMemcpyHtoD_v2, NON_STREAM_FUNC() },