jtreg supports a test selection option to run tests
based on status from a previous run.
So for example you might specify
-status:error,failed
which would re-run failed and timed out tests.
When you re-run jtreg, it accumulates the results in
the existing JTreport and JTwork directories, replacing
the results of just the tests that were in the re-run,
and presenting a summary that is the accumulation
of previously run tests and the newly run tests, and
the only thing that is "lost" is the previous status of
the specific tests that were re-run.
However when using the selection syntax above,
the presented summary is for ONLY the selected and
re-run tests, and any tests that were ignored are
not reported in the summary, contrary to expecation
and normal jtreg operation.
so if you ran 1,000 tests, and 5 failed and you re-ran
those 5, you'll get a summary like "5 tests passed",
rather than "1000 tests passed".
based on status from a previous run.
So for example you might specify
-status:error,failed
which would re-run failed and timed out tests.
When you re-run jtreg, it accumulates the results in
the existing JTreport and JTwork directories, replacing
the results of just the tests that were in the re-run,
and presenting a summary that is the accumulation
of previously run tests and the newly run tests, and
the only thing that is "lost" is the previous status of
the specific tests that were re-run.
However when using the selection syntax above,
the presented summary is for ONLY the selected and
re-run tests, and any tests that were ignored are
not reported in the summary, contrary to expecation
and normal jtreg operation.
so if you ran 1,000 tests, and 5 failed and you re-ran
those 5, you'll get a summary like "5 tests passed",
rather than "1000 tests passed".
- duplicates
-
CODETOOLS-7107914 set scope for reports (performance)
-
- Closed
-
-
CODETOOLS-6734346 harness reads too many files
-
- Closed
-
- relates to
-
CODETOOLS-7903318 Make -report:executed the default
-
- Closed
-