Stream::BlockHostUntilDone now returns Status rather than bool.

The now-deprecated Stream::BlockHostUntilDoneWithStatus remains, to facilitate a
multi-CL renaming transition.  Once all callers have been renamed to
BlockHostUntilDone, *WithStatus will be removed.

The StreamExecutor (private) method has also been renamed to BlockHostUntilDone.
It's only used by Stream.

The StreamExecutorInterface method will be renamed in a separate atomic CL.
It's harder to perform that transition gradually, and we've already performed an
atomic change previously, so we might as well fix it up in one shot.

PiperOrigin-RevId: 178907807

4 files changed, 11 insertions, 14 deletions

diff --git a/tensorflow/stream_executor/stream.cc b/tensorflow/stream_executor/stream.cc
index de65038d17..0512f4c79a 100644
--- a/tensorflow/stream_executor/stream.cc
+++ b/tensorflow/stream_executor/stream.cc
@@ -5055,7 +5055,7 @@ Stream &Stream::ThenEnqueueOnBackgroundThread(
   });
 }
 
-port::Status Stream::BlockHostUntilDoneWithStatus() {
+port::Status Stream::BlockHostUntilDone() {
   VLOG_CALL();
 
   if (!ok()) {
@@ -5072,7 +5072,7 @@ port::Status Stream::BlockHostUntilDoneWithStatus() {
     mutex_lock lock{mu_};
     for (auto &stream : sub_streams_) {
       if (!stream.second) {
-        first_error.Update(stream.first->BlockHostUntilDoneWithStatus());
+        first_error.Update(stream.first->BlockHostUntilDone());
         // Set this sub-stream as available.
         stream.second = true;
       }
@@ -5081,13 +5081,13 @@ port::Status Stream::BlockHostUntilDoneWithStatus() {
 
   temporary_memory_manager_.DeallocateFinalizedTemporaries();
 
-  first_error.Update(parent_->BlockHostUntilDoneWithStatus(this));
+  first_error.Update(parent_->BlockHostUntilDone(this));
   CheckError(first_error.ok());
   return first_error;
 }
 
-bool Stream::BlockHostUntilDone() {
-  return BlockHostUntilDoneWithStatus().ok();
+port::Status Stream::BlockHostUntilDoneWithStatus() {
+  return BlockHostUntilDone();
 }
 
 }  // namespace gputools
diff --git a/tensorflow/stream_executor/stream.h b/tensorflow/stream_executor/stream.h
index 15a5a2b6cb..4c34452048 100644
--- a/tensorflow/stream_executor/stream.h
+++ b/tensorflow/stream_executor/stream.h
@@ -1905,15 +1905,12 @@ class Stream {
   //
   // Returns an OK status if the blocking was successful and the stream is ok().
   // Otherwise returns an error describing why the blocking failed.
-  //
-  // TODO(b/70298427): Rename to BlockHostUntilDone, once all callers have been
-  // converted from the bool form.
-  port::Status BlockHostUntilDoneWithStatus() LOCKS_EXCLUDED(mu_);
+  port::Status BlockHostUntilDone() LOCKS_EXCLUDED(mu_);
 
-  // DEPRECATED(b/70298427) - new code should use BlockHostUntilDoneWithStatus()
+  // DEPRECATED(b/70298427) - new code should use BlockHostUntilDone()
   //
-  // Equivalent to BlockHostUntilDoneWithStatus().ok().
-  bool BlockHostUntilDone() LOCKS_EXCLUDED(mu_);
+  // Equivalent to BlockHostUntilDone()
+  port::Status BlockHostUntilDoneWithStatus() LOCKS_EXCLUDED(mu_);
 
   // Warning! This method interacts with internal threads in
   // sometimes-unpredictable ways and is intended for GPU-Executor-internal
diff --git a/tensorflow/stream_executor/stream_executor_pimpl.cc b/tensorflow/stream_executor/stream_executor_pimpl.cc
index 719f292937..c4b248657e 100644
--- a/tensorflow/stream_executor/stream_executor_pimpl.cc
+++ b/tensorflow/stream_executor/stream_executor_pimpl.cc
@@ -432,7 +432,7 @@ bool StreamExecutor::Launch(Stream *stream, const ThreadDim &thread_dims,
   return implementation_->Launch(stream, thread_dims, block_dims, kernel, args);
 }
 
-port::Status StreamExecutor::BlockHostUntilDoneWithStatus(Stream *stream) {
+port::Status StreamExecutor::BlockHostUntilDone(Stream *stream) {
   port::Status result;
   SCOPED_TRACE(TraceListener::BlockHostUntilDone, &result, stream);
 
diff --git a/tensorflow/stream_executor/stream_executor_pimpl.h b/tensorflow/stream_executor/stream_executor_pimpl.h
index d2965dbfd7..a2a77218cb 100644
--- a/tensorflow/stream_executor/stream_executor_pimpl.h
+++ b/tensorflow/stream_executor/stream_executor_pimpl.h
@@ -481,7 +481,7 @@ class StreamExecutor {
   // Causes the host code to synchronously wait for operations entrained onto
   // stream to complete. Effectively a join on the asynchronous device
   // operations enqueued on the stream before this program point.
-  port::Status BlockHostUntilDoneWithStatus(Stream *stream);
+  port::Status BlockHostUntilDone(Stream *stream);
 
   // Synchronously allocates size bytes on the underlying platform and returns
   // an opaque void* representing that allocation. In the case of failure,