// RUN: tf-opt %s -split-input-file -verify-diagnostics | FileCheck %s

// Tests for TensorFlow ops with custom verifiers.

// TODO(hinsu): Remove tests for ops without custom verifiers. These tests were
// added along with manual op definition and are obsolete now that the op
// definitions are auto-generated.

// TODO(hinsu): Move attribute and type tests to types.mlir file.
//===--------------------------------------------------------------------===//

  auto output_element_type = output_type.getElementType();

  // We only handle float and int element type in the verifier currently
  // TODO(hanxiongwang): we can plan to handle more element type checks besides
  // int and float in the verifier
  if (input_type.hasStaticShape() && output_type.hasStaticShape() &&
      input_element_type.isIntOrFloat() && output_element_type.isIntOrFloat()) {

			}
		}
		return true
	}

	// Every expected chain should match one (or more) returned chain. We tolerate multiple
	// matches, as due to root store semantics it is plausible that (at least on the system
	// verifiers) multiple identical (looking) chains may be returned when two roots with the
	// same subject are present.
	for _, expectedChain := range test.expectedChains {
		var match bool
		for _, chain := range chains {

		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Err), r.URL)
		return
	}

	switch rAuthType {
	case authTypeStreamingSigned, authTypeStreamingSignedTrailer:
		// Initialize stream signature verifier.
		rd, s3Err = newSignV4ChunkedReader(r, rAuthType == authTypeStreamingSignedTrailer)
		if s3Err != ErrNone {
			writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Err), r.URL)
			return
		}

		c.claimToResponseMap[claim] = replace(response, &v)
	}
	c.wantErr = replace(c.wantErr, &v)
	c.wantInitErr = replace(c.wantInitErr, &v)

	if !c.fetchKeysFromRemote {
		// Set the verifier to use the public key set instead of reading from a remote.
		c.options.KeySet = &staticKeySet{keys: c.pubKeys}
	}

	if c.optsFunc != nil {
		c.optsFunc(&c.options)
	}

  return false;
}

bool ShapeInference::InferShapeForXlaHostComputeMlir(
    _XlaHostComputeMlirOp host_compute_op) {
  // Extract the module and function.
  // The '_XlaHostComputeMlir` verifier verifies that `host_mlir_module`
  // attribute is well formed, so we just return in case of an error in
  // extracting the host function since it should never occur.
  StringAttr host_module =

}

//===----------------------------------------------------------------------===//
// BatchMatMulOp
//===----------------------------------------------------------------------===//

// Verifier to check if following is true. Will perform the verification only if
// the BatchMatMulOp inputs and output are RankedTensors
// 1. The operands of a BatchMatMulOp are broadcast compatible

#include "mlir/IR/MLIRContext.h"  // from @llvm-project
#include "mlir/IR/OpDefinition.h"  // from @llvm-project
#include "mlir/IR/SymbolTable.h"  // from @llvm-project
#include "mlir/IR/Types.h"  // from @llvm-project
#include "mlir/IR/Verifier.h"  // from @llvm-project
#include "mlir/IR/Visitors.h"  // from @llvm-project
#include "mlir/Pass/PassManager.h"  // from @llvm-project
#include "mlir/Support/LLVM.h"  // from @llvm-project