I've added rows to the spreadsheet for the results of tests generated by EvoSuite on pairs of GAVs that are not source equivalent (#34, #35). These new rows are coloured peach and have "N" in the newly added leftmost column ("From source-equivalent GAV pair?").
None of the 10 obfs test classes had any tests fail, but gaoss had 11 test classes with one or more failures, 4 of which had exactly the same failures on mvnc itself and are thus effectively false positives. This leaves 7 test classes with failures on gaoss only (marked in green on the spreadsheet):
run/gaoss/io/undertow/undertow-servlet/2.2.23.Final/io.undertow.servlet.spec.UpgradeServletInputStream_ESTest.out
run/gaoss/io/undertow/undertow-servlet/2.2.24.Final/io.undertow.servlet.spec.UpgradeServletInputStream_ESTest.out
run/gaoss/io/undertow/undertow-servlet/2.2.25.Final/io.undertow.servlet.spec.UpgradeServletInputStream_ESTest.out
run/gaoss/io/undertow/undertow-servlet/2.2.26.Final/io.undertow.servlet.spec.UpgradeServletInputStream_ESTest.out
run/gaoss/io/undertow/undertow-servlet/2.2.28.Final/io.undertow.servlet.spec.UpgradeServletInputStream_ESTest.out
run/gaoss/io/undertow/undertow-servlet/2.2.31.Final/io.undertow.servlet.spec.UpgradeServletInputStream_ESTest.out
run/gaoss/io/undertow/undertow-servlet/2.2.32.Final/io.undertow.servlet.spec.UpgradeServletInputStream_ESTest.out
Investigate.
I've added rows to the spreadsheet for the results of tests generated by EvoSuite on pairs of GAVs that are not source equivalent (#34, #35). These new rows are coloured peach and have "N" in the newly added leftmost column ("From source-equivalent GAV pair?").
None of the 10 obfs test classes had any tests fail, but gaoss had 11 test classes with one or more failures, 4 of which had exactly the same failures on mvnc itself and are thus effectively false positives. This leaves 7 test classes with failures on gaoss only (marked in green on the spreadsheet):
Investigate.