[mlir][LLVM] Plumb range attributes on parameters and results through (#117801)

We've had the ability to define LLVM's `range` attribute through
 #llvm.constant_range for some time, and have used this for some GPU
intrinsics. This commit allows using `llvm.range` as a parameter or
result attribute on function declarations and definitions.

mlir/lib/Target/LLVMIR/ModuleTranslation.cpp

@@ -1581,7 +1581,12 @@ ModuleTranslation::convertParameterAttrs(LLVMFuncOp func, int argIdx,
           .Case<IntegerAttr>([&](auto intAttr) {
             attrBuilder.addRawIntAttr(llvmKind, intAttr.getInt());
           })
-          .Case<UnitAttr>([&](auto) { attrBuilder.addAttribute(llvmKind); });
+          .Case<UnitAttr>([&](auto) { attrBuilder.addAttribute(llvmKind); })
+          .Case<LLVM::ConstantRangeAttr>([&](auto rangeAttr) {
+            attrBuilder.addConstantRangeAttr(
+                llvmKind, llvm::ConstantRange(rangeAttr.getLower(),
+                                              rangeAttr.getUpper()));
+          });
     } else if (namedAttr.getNameDialect()) {
       if (failed(iface.convertParameterAttr(func, argIdx, namedAttr, *this)))
         return failure();