On Sun, 2018-03-25 at 10:37 +0200, Niels Möller wrote: > [email protected] (Niels Möller) writes: > > > Nikos Mavrogiannopoulos <[email protected]> writes: > > > > > > is a bit strange. Maybe it's missing an #include > > > > <nettle/version.h> ? > > > > > > That's very strange. bignum.h already includes version.h, so that > > > should have been ok. I've also explicitly added it and re-run > > > with the > > > same results: > > > https://gitlab.com/gnutls/nettle/-/jobs/58340605 > > > > I'll try reproducing locally. > > It turns out it was a typo, with below fix, compilation works fine: > > --- a/lib/nettle/pk.c > +++ b/lib/nettle/pk.c > @@ -1062,7 +1062,7 @@ _wrap_nettle_pk_verify(gnutls_pk_algorithm_t > algo, > return ret; > } > > -#if NETTLE_MAJOR_VERSION < 3 || (NETTLE_MAJOR_VERSION == 3 && > NETTLE_MINOR_VERSION < 4) > +#if NETTLE_VERSION_MAJOR < 3 || (NETTLE_VERSION_MAJOR == 3 && > NETTLE_VERSION_MINOR < 4) > # ifdef ENABLE_NON_SUITEB_CURVES > # define nettle_get_secp_192r1() &nettle_secp_192r1 > # define nettle_get_secp_224r1() &nettle_secp_224r1
Thank you. I've committed a fix at: https://gitlab.com/gnutls/gnutls/merge_requests/614 > The compilation flag -Wundef would have caught this. I have never tried it, but it looks easy to fail on legal scenarios. If for example we compile with any version of nettle 3.1 or later (3.1 doesn't have version.h), that macro would make it fail. Extrapolating it to more dependencies would make a valid run extremely likely to fail compilation. regards, Nikos _______________________________________________ nettle-bugs mailing list [email protected] http://lists.lysator.liu.se/mailman/listinfo/nettle-bugs
