Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 29 Jan 2015 10:30:40 +0100
From: Daniel Cegiełka <daniel.cegielka@...il.com>
To: musl@...ts.openwall.com
Subject: Re: thoughts on reallocarray, explicit_bzero?

The concept of safe memory cleaning was mostly promoted by
cryptographic libraries - eg. secure_memzero(). Unfortunately, we have
currently too many interfaces for the same functionality: memset_s(),
secure_memzero(), explicit_bzero(), memzero_explicit(). This is why I
believe that OpenBSD (and linux developers) goes bad path, introducing
yet another secure_memzero(). A better solution would be to promote a
single standard (eg. memset_s()) and the expectation that the compiler
will respect it.

summing up: we have several options:

* volatile based, but fails with LTO
http://openwall.com/lists/musl/2014/05/19/5

* weak symbols based (from Matthew Dempsky):
https://plus.google.com/+MatthewDempsky/posts/KQHFBouxurX
http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib/libc/string/explicit_bzero.c?rev=1.3&content-type=text/x-cvsweb-markup

* barrier based, but with asm inline:
http://openwall.com/lists/musl/2015/01/28/34

Is the musl will support this feature to improve compatibility with BSD?

Daniel

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.