You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As a follow-up from kernelci/kernelci-core#445, there appears to be some issues with the kselftest log being mangled in the LAVA logs. Also, the output appear to be parsed by sed rather than a proper TAP13 parser. These 2 things may be related, if the mangled log output is a side effect of parsing results with sed.
A better approach would seem to be capturing the TAP13 results and parse it with something such as tappy. Then ideally, the output associated with each test case should be output before telling LAVA the test case name and pass/fail status.
The text was updated successfully, but these errors were encountered:
Right, parsing results with an inline Perl script. Would it still not be better to use an existing implementation? Also it's important to keep the output of each test before the test results are being reported in LAVA, so we can associate the test log fragments with each test case.
As a follow-up from kernelci/kernelci-core#445, there appears to be some issues with the kselftest log being mangled in the LAVA logs. Also, the output appear to be parsed by
sed
rather than a proper TAP13 parser. These 2 things may be related, if the mangled log output is a side effect of parsing results withsed
.A better approach would seem to be capturing the TAP13 results and parse it with something such as
tappy
. Then ideally, the output associated with each test case should be output before telling LAVA the test case name and pass/fail status.The text was updated successfully, but these errors were encountered: