Ticket #151 (closed bug: fixed)
t/pmc/exceptionhandler.t: File terminates early without explanation, disrupting plan
Reported by: | jkeenan | Owned by: | |
---|---|---|---|
Priority: | normal | Milestone: | |
Component: | core | Version: | trunk |
Severity: | medium | Keywords: | exceptionhandler plan |
Cc: | Language: | ||
Patch status: | Platform: | linux |
Description
Running make smoldertest on Linux at r35324 tonight, I got this one test failure (see first attachment). Note that there is no explanation as to why tests 7 and 8 failed to run.
When I checked the uploaded Smolder report for this file, I saw something very peculiar (second attachment). Even though the plan specifically calls for 8 tests, only 6 complete -- yet the file displays on Smolder as a 100% pass.
I wonder if the fact that some tests are in a subroutine invoked by a gotoo could have confused the plan count.
53: goto more_tests 54- 55- nonfatal_handler_one: 56- .local pmc e, c 57- .get_results (e) 58- ok(1, 'Min and Max severity for exception handlers') -- 70: more_tests: 71- 72- eh = new ['ExceptionHandler'] 73- set_addr eh, typed_handler_one 74- eh.'handle_types'(.CONTROL_OK, .CONTROL_BREAK) 75- push_eh eh
There are other recent Smolder reports (on FreeBSD, for instance) where the file DWIMs, i.e., it runs the full 8 tests but reports one as skipped.
Can anyone shed any light on what's going on?
Thank you very much.
kid51