You know, you're entirely right. That's a flaw and needs correction. Thanks for the notification.
Cheers, Richard In message <[email protected]> on Mon, 15 Feb 2016 15:43:28 -0800, "Erik Forsberg" <[email protected]> said: erik> I was talking about CRYPTO_set_mem_functions() erik> erik> It doesnt do anything, just saving some pointers that are never erik> used for anything. So a big NO-OP, seems like a TODO implementation ? erik> erik> >-- Original Message -- erik> > erik> >In message <[email protected]> on Mon, 15 Feb 2016 15:15:28 -0800, "Erik Forsberg" <[email protected]> said: erik> > erik> >erik> erik> >erik> just started do the 1.1 updates needed in my own software, in the past I had erik> >erik> use the 1.0.2 memory allocation hooks to link into my own mem leak tester. erik> >erik> erik> >erik> 1.1 has reworked this, by what I read in the past, sounded like I could still erik> >erik> get value out of if, but today when I looked at CRYPTO_set_mem_functions() erik> >erik> I see they simply save function pointers into 3 static _wrapper variables. erik> >erik> You can read them by calling CRYPTO_get_mem_functions but thats all. erik> >erik> No other use of them at all in the file crypto/mem.c erik> >erik> erik> >erik> Are we supposed to be able to hook into OPENSSL_malloc() etc in 1.1 ? erik> > erik> >You may have missed it, but there's a CRYPTO_set_mem_functions() as erik> >well. Can't you use that? erik> > erik> >Cheers, erik> >Richard erik> > erik> >-- erik> >Richard Levitte [email protected] erik> >OpenSSL Project http://www.openssl.org/~levitte/ erik> > erik> >-- erik> >openssl-dev mailing list erik> >To unsubscribe: https://mta.openssl.org/mailman/listinfo/openssl-dev erik> erik> -- erik> openssl-dev mailing list erik> To unsubscribe: https://mta.openssl.org/mailman/listinfo/openssl-dev erik> -- openssl-dev mailing list To unsubscribe: https://mta.openssl.org/mailman/listinfo/openssl-dev
