On Mon, 2017-01-30 at 14:17 -0800, Tony Cook via RT wrote: > On Mon, 30 Jan 2017 09:33:44 -0800, john@nixnuts.net wrote: > > On Sun, 2017-01-29 at 21:12 -0800, Tony Cook via RT wrote: > > > On Sat, 28 Jan 2017 16:41:50 -0800, john@nixnuts.net wrote: > > > > An updated patch is attached. > > > > > > Won't the allocated buffer leak if the load_module() fails, no > > > STORABLE_thaw > > > is defined, or if STORABLE_thaw dies? > > > > > > > Yes, Storable has several codepaths where it leaks memory or crashes > > while > > allocating memory. I'd be happy to try and fix these problems more > > systematically, but IMHO they are separate issues from the stack > > overflow. > > There's no need to add more leaks though. > > You can use SAVEFREEPV() to free the allocated memory on scope exit. > It's not clear to me what memory you believe my patch is leaking that wasn't leaking in the original code. The two assertions I added are before the matching calls to New(). Which variable have I created a new memory leak on?Thread Previous | Thread Next