Skip Menu |

This queue is for tickets about the Win32-IPC CPAN distribution.

Report information
The Basics
Id: 100875
Status: open
Priority: 0/
Queue: Win32-IPC

Owner: Nobody in particular
Requestors: bulk88 [...]

Bug Information
Severity: (no value)
Broken in: 1.11
Fixed in: (no value)

Subject: cleaning up Win32::IPC
Download (untitled) / with headers
text/plain 1.6k
I am refactoring the XS code in IPC to bring it into the 21st century in speed/optimization/machine code size. My first issue is, when I changed BOOL in the typemap from T_IV to T_TRUEFALSE. Test started to fail. ----------------------------------------------- t\30-Mutex.t ......... # This test should take no more than 10 seconds. # If it takes longer, please kill it with Ctrl-Break (Ctrl-C won't work right). t\30-Mutex.t ......... 1/15 # Failed test 'release 4 fails' # at t\30-Mutex.t line 37. # got: '' # expected: '0' # Failed test 'release unowned mutex 2 fails' # at t\30-Mutex.t line 45. # got: '' # expected: '0' # Looks like you failed 2 tests of 15. ----------------------------------------- It is because of is() being used like "is($m->release, 0, 'release unowned mutex 2 fails');" is() does a "eq" not "==". Immortal PL_sv_no is empty string PV, and 0 IV, not "0" PV and 0 IV, but eq will use the PV first if available obviously. The docs say ----------------------------------------- =item $mutex->release Release ownership of a C<$mutex>. You should have obtained ownership of the mutex through C<new> or one of the wait functions. Returns true if successful, or zero if it fails (additional error information can be found in C<$^E>). ----------------------------------------- Not returning 0 or "0" or "". What is you take? does it have to return "0" on failure or is "" fine? Both test identically in a if() so I think the impact to code should be none. I have a solution to make a fast return SVIV 0 but it will complicate the module vs using boolSV/&PL_sv_no so I'd rather not use.
Download (untitled) / with headers
text/plain 143b
I think it should be ok to return PL_sv_no, but I'm leaning towards making this version 2.000 and noting that as a potentially breaking change.

This service is sponsored and maintained by Best Practical Solutions and runs on infrastructure.

Please report any issues with to