Re: [wxlua-users] Memory leak on checking for non-existing fields

2016-05-02 Thread John Labenski
On Wed, Nov 18, 2015 at 11:00 AM, Paul K wrote: > Hi John, > > I ran my sample script through valgrind and it confirmed that the > leaked memory is what's being used for strings in error messages. I'm > including relevant parts of valgrind reports for different error > conditions, but it turned o

Re: [wxlua-users] Memory leak on checking for non-existing fields

2015-11-18 Thread Paul K
Hi John, I ran my sample script through valgrind and it confirmed that the leaked memory is what's being used for strings in error messages. I'm including relevant parts of valgrind reports for different error conditions, but it turned out to be leaking more memory than I thought (1.6k per error m

Re: [wxlua-users] Memory leak on checking for non-existing fields

2015-11-16 Thread Paul K
Hi John, Any update on a fix for this leak? It's easy to reproduce and would be great if this could be fixed, but I can't figure out what resources are not being collected... One more detail on this; it appears to leak even on `ctrl[{}]` call, which is the shortest branch of all the checks in `wx