I've been trying to verify an alternative fix to issue 338793 to see if my original workaround could be rolled back, but I've noticed that top of tree v3.18 still is flaky
I'm running the following on a chromeos-3.18 board :
test_that -b <board> <ip addr> network_DhcpStaticIP network_DhcpWpadNegotiation network_DhcpNonAsciiParameter network_DhcpRenew network_DhcpClasslessStaticRoute
---------------------------------------------------------------------------------------------------------------------
/tmp/test_that_results_6YM2nt/results-1-network_DhcpStaticIP [ FAILED ]
/tmp/test_that_results_6YM2nt/results-1-network_DhcpStaticIP FAIL: Test server didn't get all the messages it was told to expect during negotiation.
/tmp/test_that_results_6YM2nt/results-1-network_DhcpStaticIP/network_DhcpStaticIP [ FAILED ]
/tmp/test_that_results_6YM2nt/results-1-network_DhcpStaticIP/network_DhcpStaticIP FAIL: Test server didn't get all the messages it was told to expect during negotiation.
/tmp/test_that_results_6YM2nt/results-1-network_DhcpStaticIP/network_DhcpStaticIP 03/15 14:09:01.772 ERROR| dhcp_test_server:0225| Test in progress has timed out.
/tmp/test_that_results_6YM2nt/results-2-network_DhcpWpadNegotiation [ PASSED ]
/tmp/test_that_results_6YM2nt/results-2-network_DhcpWpadNegotiation/network_DhcpWpadNegotiation [ PASSED ]
/tmp/test_that_results_6YM2nt/results-3-network_DhcpNonAsciiParameter [ PASSED ]
/tmp/test_that_results_6YM2nt/results-3-network_DhcpNonAsciiParameter/network_DhcpNonAsciiParameter [ PASSED ]
/tmp/test_that_results_6YM2nt/results-4-network_DhcpRenew [ PASSED ]
/tmp/test_that_results_6YM2nt/results-4-network_DhcpRenew/network_DhcpRenew [ PASSED ]
/tmp/test_that_results_6YM2nt/results-5-network_DhcpClasslessStaticRoute [ PASSED ]
/tmp/test_that_results_6YM2nt/results-5-network_DhcpClasslessStaticRoute/network_DhcpClasslessStaticRoute [ PASSED ]
---------------------------------------------------------------------------------------------------------------------
Total PASS: 8/10 (80%)
Doesn't fail every time. Need to dig into this. The original bug issue 338793 was very deterministic, but this seems less so.
Comment 1 by bleung@chromium.org
, Mar 15 2016