Re: [PATCH] Fix for memory leak in xlate code.

On Wed, 20 Jul 2005 kfogel@collab.net wrote:

> Peter Lundblad found a thread-safety problem in the patch below.
>
> (<embarrassment> We did think about thread-safety issues for previous
> versions of the patch, versions that used a boolean instead of a
> struct pointer. But when we switched to the struct, those previous
> thoughts became irrelevant. Call it the mental equivalent of a cache
> invalidation problem.)
>
I committed my patch in r15379. I added a line setting the result of
apr_xlate_open to APR_ENOTIMPL to fake this situation and verified that
that no memory was allocated on subsequent calls.

I'll propose for backport later.

Regards,
//Peter

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Jul 21 21:07:50 2005