Fix build warning for uint64_t

[Lewuathe]

Summary

A format specifier used in SparkUtils causes the warning which formats uint64_t with %lu. We can use %llu instead of uint64_T.

SparseUtils.cpp:169:36: warning: format specifies type 'unsigned long' but the argument has type 'uint64_t' (aka 'unsigned long long') [-Wformat]
  sprintf(var, "SPARSE_MATRIX%lu", id);
                             ~~~   ^~
                             %llu

[jpienaar]

The pre-merge differs here - if you want to appease both you'd probably need to use PRIu64 (https://en.cppreference.com/w/c/types/integer) it seems

[ftynse]

mlir/lib/ExecutionEngine/SparseUtils.cpp

169

Can we rather use a portable macro PRIu64 (https://en.cppreference.com/w/cpp/types/integer#Format_macro_constants)

[aartbik]

Yes please use PRIu64 as I did above. I actually had noticed this omission, but was too late.
Sorry for the lint breakage.

[aartbik]

Note that this has been fixed as part of D89005, which also changes the name of this method for consistency.

[Lewuathe]

Thanks for fixing in https://reviews.llvm.org/D89005

mlir/lib/ExecutionEngine/SparseUtils.cpp

extern "C" void closeMatrix() {
   sparseFile = nullptr;
   sparseFilename = nullptr;
 }
 
 // Helper method to read sparse matrix filenames from the environment, defined
 // with the naming convention ${SPARSE_MATRIX0}, ${SPARSE_MATRIX1}, etc.
 extern "C" char *getSparseMatrix(uint64_t id) {
   char var[80];
-  sprintf(var, "SPARSE_MATRIX%lu", id);
+  sprintf(var, "SPARSE_MATRIX%llu", id);
   char *env = getenv(var);
   return env;
 }