Ignore EAGAIN reading timerfd in message_pump_android.

EAGAIN should only happen when no timer has expired, but the function
should only be called when a timer expires. And yet, very rarely during
tests, we error with EAGAIN. I have no theories as to why this is
happening. Possibly a platform bug.

Ignoring EAGAIN should be safe, we only read the timer to clear it so
that we don't keep getting called back. If it's already cleared then
we're fine too.

Bug: 872044
Change-Id: I6eb29ab99402d90140bd5b89a26df2aa5e53b1c6
Reviewed-on: https://chromium-review.googlesource.com/1168154
Commit-Queue: Michael Thiessen <mthiesse@chromium.org>
Reviewed-by: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#582315}

base/message_loop/message_pump_android.cc

@@ -118,7 +118,18 @@ void MessagePumpForUI::OnDelayedLooperCallback() {
   // Clear the fd.
   uint64_t value;
   int ret = read(delayed_fd_, &value, sizeof(value));
-  DCHECK_GE(ret, 0);
+
+  // TODO(mthiesse): Figure out how it's possible to hit EAGAIN here.
+  // According to http://man7.org/linux/man-pages/man2/timerfd_create.2.html
+  // EAGAIN only happens if no timer has expired. Also according to the man page
+  // poll only returns readable when a timer has expired. So this function will
+  // only be called when a timer has expired, but reading reveals no timer has
+  // expired...
+  // Quit() and ScheduleDelayedWork() are the only other functions that touch
+  // the timerfd, and they both run on the same thread as this callback, so
+  // there are no obvious timing or multi-threading related issues.
+  DPCHECK(ret >= 0 || errno == EAGAIN);
+
   delayed_scheduled_time_ = base::TimeTicks();
 
   base::TimeTicks next_delayed_work_time;
@@ -168,7 +179,7 @@ void MessagePumpForUI::OnNonDelayedLooperCallback() {
   // See http://man7.org/linux/man-pages/man2/eventfd.2.html
   uint64_t value = 0;
   int ret = read(non_delayed_fd_, &value, sizeof(value));
-  DCHECK_GE(ret, 0);
+  DPCHECK(ret >= 0);
 
   // If we read a value > 1, it means we lost the race to clear the fd before a
   // new task was posted. This is okay, we can just re-schedule work.
@@ -272,7 +283,7 @@ void MessagePumpForUI::ScheduleWork() {
   // EFD_SEMAPHORE.
   uint64_t value = 1;
   int ret = write(non_delayed_fd_, &value, sizeof(value));
-  DCHECK_GE(ret, 0);
+  DPCHECK(ret >= 0);
 }
 
 void MessagePumpForUI::ScheduleDelayedWork(const TimeTicks& delayed_work_time) {
@@ -294,7 +305,7 @@ void MessagePumpForUI::ScheduleDelayedWork(const TimeTicks& delayed_work_time) {
   ts.it_value.tv_nsec = nanos % TimeTicks::kNanosecondsPerSecond;
 
   int ret = timerfd_settime(delayed_fd_, TFD_TIMER_ABSTIME, &ts, nullptr);
-  DCHECK_GE(ret, 0);
+  DPCHECK(ret >= 0);
 }
 
 void MessagePumpForUI::QuitWhenIdle(base::OnceClosure callback) {