leavelet/cutlass
1
0
Fork 0
Code Issues Pull Requests Actions 27 Packages Projects Releases Wiki Activity

CUTLASS 3.8 Release (#2059)

Browse Source 
* CUTLASS 3.8 Release

* update

* Update README.md

* Revert "Update README.md"

This reverts commit b353e36fe83e0815f99b44e46c0c95494c44726b.

* update

* update

---------

Co-authored-by: Haicheng Wu <57973641+hwu36@users.noreply.github.com>
Co-authored-by: Haicheng Wu <haichengw@nvidia.com>
This commit is contained in:
mihir-awatramani
2025-01-25 02:44:06 -05:00
committed by GitHub
co-authored by Haicheng Wu Haicheng Wu
parent 9eb01fa0b0
commit 389e493055
290 changed files with 91222 additions and 291 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/profiler/src/options.cu
+21 -4
View File
@@ -33,6 +33,7 @@
*/
#include <algorithm>
#include <fstream>
#include <set>
#include "cutlass/cutlass.h"
@@ -810,16 +811,27 @@ Options::Options(cutlass::CommandLine const &cmdline):
}
else if (cmdline.check_cmd_line_flag("kernels")) {
cmdline.get_cmd_line_arguments("kernels", operation_names);
profiling.error_on_no_match = cmdline.check_cmd_line_flag("error-on-no-match");
profiling.error_if_nothing_is_profiled = cmdline.check_cmd_line_flag("error-if-nothing-is-profiled");
}
if (cmdline.check_cmd_line_flag("kernels-file")) {
std::string filename;
cmdline.get_cmd_line_argument("kernels-file", filename, {});
std::ifstream input(filename);
if (!input.good()) {
throw std::runtime_error("failed to open: " + filename);
}
for (std::string line; getline(input, line);) {
operation_names.push_back(line);
}
}
if (cmdline.check_cmd_line_flag("ignore-kernels")) {
cmdline.get_cmd_line_arguments("ignore-kernels", excluded_operation_names);
profiling.error_on_no_match = cmdline.check_cmd_line_flag("error-on-no-match");
profiling.error_if_nothing_is_profiled = cmdline.check_cmd_line_flag("error-if-nothing-is-profiled");
}
profiling.error_on_no_match = cmdline.check_cmd_line_flag("error-on-no-match");
profiling.error_if_nothing_is_profiled = cmdline.check_cmd_line_flag("error-if-nothing-is-profiled");
// Prevent launches on the device for anything other than CUTLASS operation
// Allow verification only on host
if (execution_mode == ExecutionMode::kTrace) {
@@ -856,6 +868,11 @@ void Options::print_usage(std::ostream &out) const {
<< " (\"s1688\" and \"nt\") or (\"s844\" and \"tn\" and \"align8\") in their" << end_of_line
<< " operation name using --kernels=\"s1688*nt, s884*tn*align8\"\n\n"
<< " --kernels-file=<filename> "
<< " Same behavior as --kernels, but kernel names are specified in a file" << end_of_line
<< " with one kernel on each line. Set of profiled kernels is the union of kernels specified" << end_of_line
<< " here and those specified in `kernels`.\n\n"
<< " --ignore-kernels=<string_list> "
<< " Excludes kernels whose names match anything in this list.\n\n"
;