Hide Forgot
The question is whether we should diverge from upstream on this. Also should it be disabled also for CRL verification?
(In reply to Tomas Mraz from comment #4) > The question is whether we should diverge from upstream on this. Also should > it be disabled also for CRL verification? Yes, all three uses for MD5withRSA are vulnerable. OCSP, certificates and CRLs, all of them allow for arbitrary extensions, which are placed as last in the hashed data, so all of them are vulnerable to chosen-prefix collision attacks. MD5 is completely broken for chosen-prefix collision attacks (complexity 2^39).
Does NSS in RHEL-7 disable all these three uses?
(In reply to Tomas Mraz from comment #6) > Does NSS in RHEL-7 disable all these three uses? Yes, it doesn't trust them by default. See bug 1012656.
This request was resolved in Red Hat Enterprise Linux 7.0. Contact your manager or support representative in case you have further questions about the request.