<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body dir="auto"><div>Ok. If this occurs again, please update the threshold. <br></div><div><br>On May 17, 2017, at 12:03 PM, David P Grove <<a href="mailto:groved@us.ibm.com">groved@us.ibm.com</a>> wrote:<br><br></div><blockquote type="cite"><div><p><tt><font size="2"><a href="mailto:swift-dev-bounces@swift.org">swift-dev-bounces@swift.org</a> wrote on 05/17/2017 02:04:23 PM:</font></tt><br><tt><font size="2">> <br>> Failure in dispatch-starfish. This test looks timing-sensitive. <br>> Boris, is this unlikely to be related to your changes? Does anyone <br>> familiar with dispatch have suggestions, such as increasing the <br>> latency threshold?</font></tt><br><tt><font size="2">> <br>> [BEGIN] Latency</font></tt><br><tt><font size="2">> Actual: 1553</font></tt><br><tt><font size="2">> Expected: <1000</font></tt><br><tt><font size="2">> [FAIL] Latency (dispatch_starfish.c:82)</font></tt><br><tt><font size="2">> dispatch_starfish.c:82</font></tt><br><tt><font size="2">> lap: 9</font></tt><br><tt><font size="2">> count: 1000</font></tt><br><tt><font size="2">> delta: 6756400000 ns</font></tt><br><tt><font size="2">> math: 3374.825175 ns / lap</font></tt><br><tt><font size="2">> <br>> [BEGIN] Latency</font></tt><br><tt><font size="2">> Actual: 3374</font></tt><br><tt><font size="2">> Expected: <1000</font></tt><br><tt><font size="2">> [FAIL] Latency (dispatch_starfish.c:82)</font></tt><br><tt><font size="2">> dispatch_starfish.c:82</font></tt><br><tt><font size="2">> <br></font></tt><br><tt><font size="2">The test is subject to failing spuriously if the machine is badly overloaded. One could increase the threshold by changing the #define on line 43 of dispatch_starfish.c</font></tt><br><br><tt><font size="2">There's a trade-off between spurious failures due to overloads on build machines vs. missing performance regressions by making the latency thresholds much too lenient. </font></tt><br><br><tt><font size="2">We've been running with the current threshold in dispatch_starfish since we began running the dispatch tests in the CI system last summer. So if this isn't a one-off failure caused by CI load, it might be pointing to an underlying performance regression.</font></tt><br><br><tt><font size="2">--dave</font></tt><br><br>
</p></div></blockquote></body></html>