|
From: Mark L. <mar...@gm...> - 2009-07-06 19:29:48
|
Is there a way to persuade Memcheck to treat leak warnings as errors, and exit with an error status? (The "--error-exitcode" option doesn't seem to do that.) Such a feature would be very helpful for automatic regression testing. It would be unfortunate to resort to scraping the warning output to determine pass/fail status. Thanks, Mark Leone (mar...@gm...) |
|
From: Nicholas N. <n.n...@gm...> - 2009-07-06 22:18:35
|
On Tue, Jul 7, 2009 at 4:57 AM, Mark Leone<mar...@gm...> wrote: > Is there a way to persuade Memcheck to treat leak warnings as errors, > and exit with an error status? (The "--error-exitcode" option doesn't > seem to do that.) > > Such a feature would be very helpful for automatic regression testing. > It would be unfortunate to resort to scraping the warning output to > determine pass/fail status. You can't do it currently, but it's on the "wanted" list for 3.5.0, the next release: https://bugs.kde.org/show_bug.cgi?id=152393 Nick |
|
From: Ashley P. <as...@pi...> - 2009-07-06 22:26:50
|
On Mon, 2009-07-06 at 11:57 -0700, Mark Leone wrote: > > Such a feature would be very helpful for automatic regression testing. > It would be unfortunate to resort to scraping the warning output to > determine pass/fail status. When I faced this problem I resorted to using the XML output of valgrind and writing my own wrapper which could return an exit code on various conditions being true, it wasn't particularly quick to do but once implemented it was very versatile. I probably still have the code for it around somewhere if you wanted a copy. Ashley Pittman, -- Ashley Pittman, Bath, UK. Padb - A parallel job inspection tool for cluster computing http://padb.pittman.org.uk |