TF_EXPECT_OK(ReadCache(&cache1, "", 0, 1, &out));
  EXPECT_EQ(calls, 1);
  // Now advance the clock one second at a time and redo the read. The call
  // count should advance every 3 seconds (i.e. every time the staleness is
  // greater than 2).
  for (int i = 1; i <= 10; i++) {
    env->SetNowSeconds(i + 1);
    TF_EXPECT_OK(ReadCache(&cache1, "", 0, 1, &out));
    EXPECT_EQ(calls, 1 + i / 3);
  }

https://github.com/tensorflow/tensorflow/blob/master/tensorflow/tools/tf_sig_build_dockerfiles/devel.usertools/code_check_full.bats
Here are the affected tests:
EOF
    while read dep; do
      echo "For dependency $dep:"
      # For every missing dependency, find the tests which directly depend on
      # it, and print that list for debugging. Not really clear if this is
      # helpful since the only examples I've seen are enormous.

### Typical Pull Request Workflow -

**1. New PR**

- As a contributor, you submit a New PR on GitHub.
- We inspect every incoming PR and add certain labels to the PR such as `size:`,
  `comp:` etc.  At this stage we check if the PR is valid and meets certain
  quality requirements. For example, we check if the CLA is signed, PR has

// graph mode). Calling this on a context tells it to start a step.
TF_CAPI_EXPORT extern void TFE_ContextStartStep(TFE_Context* ctx);

// Ends a step. When there is no active step (that is, every started step has
// been ended) step containers will be cleared. Note: it is not safe to call
// TFE_ContextEndStep while ops that rely on the step container may be running.

#ifdef __cplusplus
extern "C" {
#endif

// Resets `op_to_reset` with `op_or_function_name` and `raw_device_name`. This
// is for performance optimization by reusing an exiting unused op rather than
// creating a new op every time. If `raw_device_name` is `NULL` or empty, it
// does not set the device name. If it's not `NULL`, then it attempts to parse
// and set the device name. It's effectively `TFE_OpSetDevice`, but it is faster

  // TODO(srbs): It seems like this is used only for performance optimization
  // and not for correctness. The only downside of keeping this 1 seems to be
  // that the gradient accumulation is unbounded and we will never
  // aggressively aggregate accumulated gradients to recover memory.
  // Revisit and fix.
  return 1;
}

// Consumes references to the tensors in the gradient_tensors list and returns

https://github.com/tensorflow/tensorflow/blob/master/tensorflow/tools/tf_sig_build_dockerfiles/devel.usertools/code_check_full.bats
Here are the affected tests:
EOF
    while read dep; do
      echo "For dependency $dep:"
      # For every missing dependency, find the tests which directly depend on
      # it, and print that list for debugging. Not really clear if this is
      # helpful since the only examples I've seen are enormous.

    License along with the GNU C Library; if not, see
-   <http://www.gnu.org/licenses/>.  */
+   <https://www.gnu.org/licenses/>.  */
 
 #ifndef _UTMP_H
 # error "Never include <bits/utmp.h> directly; use <utmp.h> instead."
@@ -24,11 +24,13 @@
 #include <time.h>
 
 
-#define	UT_NAMESIZE	8
-#define	UT_LINESIZE	8
-#define	UT_HOSTSIZE	16
+#define UT_LINESIZE	32

                              static_cast<off_t>(offset))};
    if (r > 0) {
      dst += r;
      offset += static_cast<uint64_t>(r);
      n -= r;  // safe as 0 < r <= n so n will never underflow
      read += r;
    } else if (r == 0) {
      TF_SetStatus(status, TF_OUT_OF_RANGE, "Read fewer bytes than requested");
      break;
    } else if (errno == EINTR || errno == EAGAIN) {
      // Retry

      immediate_execution_ctx_.get(), {x.get()}, UseFunction()));
}

TEST_P(CppGradients, TestMatMulGrad) {
  // TODO(vnvo2409): Figure out why `gradient_checker` does not work very
  // well with `MatMul` and remove `TestMatMul*` in
  // `mnist_gradients_test` when done.
  GTEST_SKIP();

  float A_vals[] = {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f, 9.0f};
  int64_t A_dims[] = {3, 3};