On Tue, Jul 12, 2016 at 07:02:01PM -0700, Kevin J. McCarthy wrote: > On Tue, Jul 12, 2016 at 01:44:33PM -0500, Will Fiveash wrote: > > Today I did some memory leak testing of mutt v1.6.2 and found that when > > sending email via mutt->libsasl->libcrypto (OpenSSL) there were a number > > of memory leaks. Basically I see about 10KB leaked per message sent. > > Has anyone else observed this? > > > > Note that my configuration is using smtp_url="smtps://...". > > > > I've attached a memory leak report of the four email sends. BTW, I also > > see similar leaks with mutt v1.5.23. > > Well, I did clean up one of those leaks in our code, and spotted another > while I was in the same region. The OpenSSL ones look to be internal to > their library.
Yes, I see a similar issue with fetchmail and OpenSSL leaking. > It looks like the SASL password is leaked, but it's not clear to me how > we're supposed to free that. I saw one of the example clients used a > static variable and reused it for each authentication. I'm reluctant to > make that change though: I'd rather leak the memory than have an > implementation try to free it and then we realloc the freed memory. When I have more time I will investigate this. -- Will Fiveash
