Clean-up the logd tests.
Make it so all of the tests that send command all go through a
single function to avoid the command being sent incorrectly.
Modify the send_to_control function to avoid hanging forever,
and also read the entire buffer to avoid error messages from
logd. Also, update some errors so it's more obvious what went wrong.
General clean-up, adding TEMP_FAILURE_RETRY when appropriate.
Remove the benchmark test, it doesn't do anything since it depends
on another executable. The test itself is out-dated since its
checking for pruning, which is not enabled now.
Test: Ran unit tests in a loop.
Change-Id: I071431e200af4cdf0311daae46483d993b23528b
1 file changed