This bug report was migrated from our old Bugzilla tracker.

These attachments are available in the static archive:

• Fix for memory leak (bugfix.diff, text/plain, 2009-06-20 14:33:05 +0000, 1186 bytes)

Reported in version: 1.2.13
Reported for operating system, platform: Windows (All), x86

Comments on the original bug report:

On 2009-06-20 14:33:05 +0000, Steven Don wrote:

Created attachment 333
Fix for memory leak

Certain codepaths in the DirectDraw backend for SDL 1.2 SDL_malloc a buffer which is never SDL_free'd. Instead, the pointer is simply overwritten by a pointer to a DirectDraw surface when the surface is locked. That is a memory leak.

The patch attached stores the pointer for later proper release.

I also posted this to the mailing list, and was directed to submit it here.

On 2009-06-30 22:11:39 +0000, Sam Lantinga wrote:

When does surface->pixels get allocated prior to DX5_LockHWSurface()?

On 2009-07-01 11:01:15 +0000, Steven Don wrote:

In SDL_SetVideoMode. The reason I've left it in and don't just SDL_Free it immediately on a lock is on the off-chance that some badly behaved program grabs the pointer immediately after SDL_SetVideoMode and uses it. Freeing it immediately or not allocating it at all would break compatibility with such programs. Personally I think that should be avoided if it can be helped.

On 2009-09-13 16:33:01 +0000, Ryan C. Gordon wrote:

Tagging this bug with "target-1.2.14" so we can try to resolve it for SDL 1.2.14.

Please note that we may choose to resolve it as WONTFIX. This tag is largely so we have a comprehensive wishlist of bugs to examine for 1.2.14 (and so we can close bugs that we'll never fix, rather than have them live forever in Bugzilla).

--ryan.

On 2009-09-26 21:06:21 +0000, Sam Lantinga wrote:

Can you give me a test case that reproduces the memory leak? I'd like to fully understand what's going on before I apply this patch.

Thanks!

On 2009-10-17 11:08:40 +0000, Sam Lantinga wrote:

This is a risky change for SDL 1.2, so I'm going to hold off fixing it for now. If you have a test case, please re-open this bug.

Thanks!