[AD] [ alleg-Bugs-2951650 ] sanity checks for allegro-4.9.9.1 API

[ Thread Index | Date Index | More lists.liballeg.org/allegro-developers Archives ]


Bugs item #2951650, was opened at 2010-02-14 18:56
Message generated for change (Comment added) made by elias
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=105665&aid=2951650&group_id=5665

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Linux
Group: 4.9
Status: Open
Resolution: Invalid
Priority: 5
Private: No
Submitted By: Andrey Ponomarenko (andrey-moscow)
Assigned to: Nobody/Anonymous (nobody)
Summary: sanity checks for allegro-4.9.9.1 API

Initial Comment:
Hello,
I have performed some sanity checks of the library API and have found that it is not 100% safe. I have used API-Sanity-Autotest tool and have detected segfaults in the 122/408 functions (30%). Test results are in the attachment. See "Received signal SEGV" section of the report.

Environment: Mandriva 2009.1, KDE-4.2, x86, allegro-4.9.9.1

----------------------------------------------------------------------

>Comment By: Elias Pschernig (elias)
Date: 2010-02-14 21:25

Message:
All the cases I looked at are still invalid calls which are supposed to
segfault (or assert in debug mode). Can you run it in debug mode, and
ignore all tests which have SIGABRT instead of SIGSEGV?

Also, 4.9.9 is so severely outdated that any genuine results would be
meaningless, we're at 4.9.17 by now.

----------------------------------------------------------------------

Comment By: Andrey Ponomarenko (andrey-moscow)
Date: 2010-02-14 21:11

Message:
Test results have been updated. Added preamble call of al_install_system()
function to all test cases.

----------------------------------------------------------------------

Comment By: Elias Pschernig (elias)
Date: 2010-02-14 20:21

Message:
Functions are supposed to crash when called before al_install_system().

That was an early design decision - the mailing list would be a better
place to discuss changing this.

However maybe you can make the test generator create somewhat more
meaningful tests? E.g. calling al_install_system() before using an Allegro
function. Or creating a display before calling a function which requires an
active display. Or if you wait on a condition variable, actually have the
condition change somewhere.


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=105665&aid=2951650&group_id=5665




Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/