I wouldn't have thought so no. The more garbage that gets written the
better, provided it gets written to the right place that is.

FWIW. There's a measurable performance increase reverting that to memse=
t()
and adding tests to ensure that the compiler hasn't in fact optimized i=
t
away.

Peter




=

Gerrit =

Bruchh=E4user =

To
ro.com> openssl-dev@openssl.org =

Sent by: =
cc
owner-openssl-dev =

@openssl.org Subj=
ect
crypto/mem_clr.c is not thread s=
afe
=

20/06/07 05:02 PM =

=

=

Please respond to =

openssl-dev =

=

=





Hello,

I made few tests with Intel Thread Check. The following result came u=
p:

crypto/mem_clr.c: [OPENSSL_cleanse]

Memory read at [...] conflicts with a prior memory write a=
t
[...] (flow
dependence)


Maybe 'OPENSSL_cleanse' should be protected with a mutex?

BR,
Gerrit
__________________________________________________ ____________________
OpenSSL Project http://www.openssl.org
Development Mailing List openssl-dev@openssl.org
Automated List Manager majordomo@openssl.org
=


__________________________________________________ ____________________
OpenSSL Project http://www.openssl.org
Development Mailing List openssl-dev@openssl.org
Automated List Manager majordomo@openssl.org