[mlir][sparse] Reverting invalid D135126

[wrengr]

Summary

The change D135126 should never have landed, since the warnings in
question were already resolved by D134925.

[aartbik]

mlir/lib/ExecutionEngine/SparseTensor/File.cpp

27

please revert BAZEL fix as part of this change too

[stella.stamenova]

D135126 was needed to fix the Windows mlir bot which was red for several days because of https://github.com/llvm/llvm-project/commit/e898be2f6edbb886af2f6b23e2f5db5210535620 which landed after D134925 and undid it.

[NathanielMcVicar]

mlir/lib/ExecutionEngine/SparseTensor/File.cpp

111

I think it's probably irrelevant after D135304, but putting this in the default produces a clang warning (see discussion in D135126).

[wrengr]

Bah, when D135126 came to my attention I thought it had already landed (not just been approved). But since it hasn't, there's no need for this differential— supposing we can get D135126 into a state that all three of MSVC, Clang, and Bazel all like it before it does land

In D135186#3834611, @stella.stamenova wrote:

D135126 was needed to fix the Windows mlir bot which was red for several days because of https://github.com/llvm/llvm-project/commit/e898be2f6edbb886af2f6b23e2f5db5210535620 which landed after D134925 and undid it.

I missed that D134925 got reverted. Thanks for the heads-up

[wrengr]

This differential is no longer necessary. The changes in D135304 got everything into a good state for MSVC, Clang, and our team's style goals. And D135401 will revert the Bazel dependencies.

mlir/lib/ExecutionEngine/SparseTensor/File.cpp

 // tensor manipulations.  The functionality of the support library is meant
 // to simplify benchmarking, testing, and debugging MLIR code operating on
 // sparse tensors.  However, the provided functionality is **not** part of
 // core MLIR itself.
 //
 //===----------------------------------------------------------------------===//
 
 #include "mlir/ExecutionEngine/SparseTensor/File.h"
-#include "llvm/Support/ErrorHandling.h"
 
 #include <cctype>
 #include <cstring>
 
 using namespace mlir::sparse_tensor;
 
 /// Opens the file for reading.
 void SparseTensorFile::openFile() {
case ValueKind::kComplex:
     // When the file is specified to store complex values, then we
     // require a complex primary-type.
     return isComplexPrimaryType(valTy);
   case ValueKind::kUndefined:
     // The "extended" FROSTT format doesn't specify a ValueKind.
     // So we allow implicitly converting the stored values to both
     // integer and floating primary-types.
     return isRealPrimaryType(valTy);
+  default:
+    MLIR_SPARSETENSOR_FATAL("Unknown ValueKind: %d\n",
+                            static_cast<uint8_t>(valueKind_));
   }
-  llvm_unreachable("unknown ValueKind");
 }
 
 /// Helper to convert C-style strings (i.e., '\0' terminated) to lower case.
 static inline void toLower(char *token) {
   for (char *c = token; *c; ++c)
     *c = tolower(*c);
 }