Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Client: don't use hipDeviceSynchronize for blocking CPU to get CPU timer #753

Merged
merged 1 commit into from
May 3, 2024
Merged

Client: don't use hipDeviceSynchronize for blocking CPU to get CPU timer #753

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
4 changes: 2 additions & 2 deletions clients/common/utility.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -119,9 +119,9 @@ double get_time_us_sync_device(void)
/*! \brief CPU Timer(in microsecond): synchronize with given queue/stream and return wall time */
double get_time_us_sync(hipStream_t stream)
{
if(hipDeviceSynchronize() != hipSuccess)
if(hipStreamSynchronize(stream) != hipSuccess)
{
hipblaslt_cerr << "Synchronizing device failed" << std::endl;
hipblaslt_cerr << "Synchronizing stream failed" << std::endl;
}

auto now = std::chrono::steady_clock::now();
Expand Down
9 changes: 0 additions & 9 deletions clients/include/testing_matmul.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2460,7 +2460,6 @@ void testing_matmul_with_bias(const Arguments& arg)
CHECK_HIP_ERROR(hipEventRecord(event_gpu_time_start, stream));
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
flush_time_used = get_time_us_sync(stream);
}
for(int i = 0; i < flush_iter; i++)
Expand All @@ -2476,7 +2475,6 @@ void testing_matmul_with_bias(const Arguments& arg)
}
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
flush_time_used = get_time_us_sync(stream) - flush_time_used;
}
flush_time_used /= flush_iter;
Expand All @@ -2503,7 +2501,6 @@ void testing_matmul_with_bias(const Arguments& arg)
CHECK_HIP_ERROR(hipEventRecord(event_gpu_time_start, stream));
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
gpu_time_used = get_time_us_sync(stream);
}

Expand Down Expand Up @@ -2552,7 +2549,6 @@ void testing_matmul_with_bias(const Arguments& arg)
CHECK_HIP_ERROR(hipEventRecord(event_gpu_time_start, stream));
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
gpu_time_used = get_time_us_sync(stream);
}
for(int i = 0; i < number_hot_calls; i++)
Expand Down Expand Up @@ -2598,7 +2594,6 @@ void testing_matmul_with_bias(const Arguments& arg)
}
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
gpu_time_used = get_time_us_sync(stream) - gpu_time_used;
}
}
Expand Down Expand Up @@ -2635,7 +2630,6 @@ void testing_matmul_with_bias(const Arguments& arg)
CHECK_HIP_ERROR(hipEventRecord(event_gpu_time_start, stream));
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
gpu_time_used = get_time_us_sync(stream);
}

Expand All @@ -2654,7 +2648,6 @@ void testing_matmul_with_bias(const Arguments& arg)
}
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
gpu_time_used = get_time_us_sync(stream) - gpu_time_used;
}
}
Expand All @@ -2679,7 +2672,6 @@ void testing_matmul_with_bias(const Arguments& arg)
CHECK_HIP_ERROR(hipEventRecord(event_gpu_time_start, stream));
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
gpu_time_used = get_time_us_sync(stream);
}

Expand All @@ -2697,7 +2689,6 @@ void testing_matmul_with_bias(const Arguments& arg)
}
else
{
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
gpu_time_used = get_time_us_sync(stream) - gpu_time_used;
}
}
Expand Down