Remove the workaround to perfetto calling CommitData on the wrong sequence.

The fix was submitted in
https://android-review.googlesource.com/c/platform/external/perfetto/+/874695

So we no longer need to special case this instead we can just
delegate the call directly.

Bug: https://bugs.chromium.org/p/chromium/issues/detail?id=839074
Change-Id: Ie6d6ee8d94b00f73b59dd95a90d2c29dd3f9a383
Reviewed-on: https://chromium-review.googlesource.com/c/1451942
Commit-Queue: Stephen Nusko <nuskos@chromium.org>
Commit-Queue: Eric Seckler <eseckler@chromium.org>
Auto-Submit: Stephen Nusko <nuskos@chromium.org>
Reviewed-by: Eric Seckler <eseckler@chromium.org>
Cr-Commit-Position: refs/heads/master@{#629999}

services/tracing/public/cpp/perfetto/producer_client.cc

@@ -241,19 +241,7 @@ void ProducerClient::NotifyDataSourceStopped(
 
 void ProducerClient::CommitData(const perfetto::CommitDataRequest& commit,
                                 CommitDataCallback callback) {
-  // TODO(oysteine): Remove the PostTask once Perfetto is fixed to always call
-  // CommitData on its provided TaskRunner, right now it'll call it on whatever
-  // thread is requesting a new chunk when the SharedMemoryBuffer is full. Until
-  // then this is technically not threadsafe on shutdown (when the
-  // ProducerClient gets destroyed) but should be okay while the Perfetto
-  // integration is behind a flag.
-  if (GetTaskRunner()->RunsTasksInCurrentSequence()) {
-    producer_host_->CommitData(commit);
-  } else {
-    GetTaskRunner()->PostTask(
-        FROM_HERE, base::BindOnce(&ProducerClient::CommitDataOnSequence,
-                                  base::Unretained(this), commit));
-  }
+  producer_host_->CommitData(commit);
 }
 
 void ProducerClient::CommitDataOnSequence(