On 21 August 2017 at 12:59, Andrew Cagney <[email protected]> wrote: > On 20 August 2017 at 00:02, D. Hugh Redelmeier <[email protected]> wrote: >> This code in testing/cavp/cavp_ikev1.c looks fishy: >> >> if (prf == NULL) { >> print_line(prf->key); >> return; >> } >> >> My best guess is that the test is suposed to be: >> if (prf->prf == NULL) { > > Yes. Here, prf is never NULL, but prf->prf could be. > > If someone feeds SHA-224, which we don't support into CAVP, then it > will dump core instead of stumbling on. The test facility shouldn't > be doing that anyway :-) > > I'll change to code and get rid of the indirection.
This quickly got messy. So I renamed 'prf' to 'prf_entry' instead (and tweaked that check). Andrew _______________________________________________ Swan-dev mailing list [email protected] https://lists.libreswan.org/mailman/listinfo/swan-dev
