Hello, I've decided to get a look at the memory handling in worms, and check typical initialization problems, using valgrind [1] 3.1.0 (Debian) tool named memcheck [2], which I find very effective, at least it is 100% when I code. So far, I just loaded the game, let it display the menu then exit.
Well, it seems there are lots of leaks and uninitialized data access, mainly due to autoFree not set most of the time in Surface objects, which I find rather strange though. This ends up with those typical results: ==28256== 13,959,556 bytes in 94 blocks are possibly lost in loss record 139 of 140 ==28256== at 0x401B41A: malloc (vg_replace_malloc.c:149) ==28256== by 0x40F9DAB: SDL_CreateRGBSurface (in /usr/lib/libSDL-1.2.so.0.7.2) ==28256== by 0x402A7E9: rotozoomSurfaceXY (in /usr/lib/libSDL_gfx.so.4.9.0) ==28256== by 0x8060E7C: Wormux::Surface::RotoZoomXY(double, double, double, int) (surface.cpp:212) ==28256== by 0x8063210: Sprite::EnableFlippingCache() (sprite.cpp:210) ==28256== by 0x80B1DCB: Skin::LoadManySkins(xmlpp::Element*, Profile*) (skin.cpp:165) ==28256== by 0x80B2349: Skin::Charge(std::string const&, std::string const&) (skin.cpp:115) ==28256== by 0x80B2674: LoadOneSkin(std::string const&, std::string const&) (skin.cpp:280) ==28256== ==28256== ==28256== 14,177,432 bytes in 1,833 blocks are indirectly lost in loss record 140 of 140 ==28256== at 0x401B41A: malloc (vg_replace_malloc.c:149) ==28256== by 0x40F9DAB: SDL_CreateRGBSurface (in /usr/lib/libSDL-1.2.so.0.7.2) ==28256== by 0x402A7E9: rotozoomSurfaceXY (in /usr/lib/libSDL_gfx.so.4.9.0) ==28256== by 0x8060E7C: Wormux::Surface::RotoZoomXY(double, double, double, int) (surface.cpp:212) ==28256== by 0x8063210: Sprite::EnableFlippingCache() (sprite.cpp:210) ==28256== by 0x80B1DCB: Skin::LoadManySkins(xmlpp::Element*, Profile*) (skin.cpp:165) ==28256== by 0x80B2349: Skin::Charge(std::string const&, std::string const&) (skin.cpp:115) ==28256== by 0x80B2674: LoadOneSkin(std::string const&, std::string const&) (skin.cpp:280) OK, I think I've already pasted a huge chunk so I let you examine more in depth the log available at [3]. I don't think it is unreadable, but if you need help on reading the log, I can try. Best regards, Kurosu [1] http://valgrind.org/ [2] http://valgrind.org/docs/manual/mc-manual.html [3] http://kurosu.free.fr/wormux/wormux.log.bz2
