References: <1573760407-20662-1-git-send-email-selva.nair@...277...> <108f5ba4-240b-5496-f37b-51ff2a6fb2fe@...2080...> From: Arne Schwabe Message-ID: <5bb0e3e1-e5ea-6385-a7e5-0a5dd23626d8@...1227...> Date: Sun, 17 Nov 2019 20:08:10 +0100 MIME-Version: 1.0 In-Reply-To: Content-Type: multipart/signed; micalg=pgp-sha256; protocol="application/pgp-signature"; boundary="g0r0FTbW4kq2X4GimhBuhid3xZRNli2Gv" Subject: Re: [Openvpn-devel] [PATCH] Fix ACL_CHECK_ADD_COMPILE_FLAGS to work with clang List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Selva Nair , David Sommerseth Cc: openvpn-devel This is an OpenPGP/MIME signed message (RFC 4880 and 3156) --g0r0FTbW4kq2X4GimhBuhid3xZRNli2Gv Content-Type: multipart/mixed; boundary="4Yug9fg8EmVQ9jNc8kGPWAzO0pKEFT5ZG" --4Yug9fg8EmVQ9jNc8kGPWAzO0pKEFT5ZG Content-Language: de-DE Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Am 14.11.19 um 22:58 schrieb Selva Nair: > Hi David >=20 > Thanks for the comments >=20 > My idea was just to add -Werror right in the line above, and not > extend the > ACL_CHECK_ADD_COMPILE_FLAGS macro with another argument. >=20 >=20 > I'm fine with that approach as well. Let me know if you want a v2. >=20 >=20 > I think you said it pretty well in your mail: >=20 > > Darn compilers and darn -Werror >=20 >=20 > So your change does improve Clang ... I dunno what we will do with GC= C. >=20 >=20 > It doesn't affect gcc as gcc does error (or ignore -- see below) > unsupported options with or without -Werror. So I think we are okay. > travis build on my fork confirms that: > https://travis-ci.org/selvanair/openvpn/builds/612019849 >=20 > However, gcc does have quirky behaviour when it comes to options like > -Wno-xx-yy. It just accepts them silently[*] with or without -Werror. > That works fine for us, but it does generate an error about that > unsupported option if any other error is encountered during the build.=C2= =A0 > As other errors are anyway a show-stopper, I think we can live with that. >=20 > Selva >=20 > [*] I think their reasoning is that -Wno-xx-yy can be thus used to > suppress warnings added to newer versions without breaking builds with > older ones. But clang doesn't share that view. I agree with Selva here. If we cannot get rid of the warnings themselves this seems to be the only sane way. So I am not totally with this stuff but it is better than the altneratives at the moment. And I really annoyed by all these new warnings of Clang Acked-By: Arne Schwabe Arne --4Yug9fg8EmVQ9jNc8kGPWAzO0pKEFT5ZG-- --g0r0FTbW4kq2X4GimhBuhid3xZRNli2Gv Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- iQIzBAEBCAAdFiEE88wmPb+Azu8xbws3FRhsZwKxRFQFAl3RmpoACgkQFRhsZwKx RFQr1A//YMOe82GIqI3ORu5CoIgs2FOnvpkYpuJjOpuCvklSvHnsk+5fPgKJ+cdS JXqXjpnEUbrzAvtJZ1ZVNyaNK9+Dz6OswB7pP71gOx9e1IeQBP7fdknQ6y7vi71y NqKf664tU9LwOvAAbIQDMUUrZ7GOYCYdB8stV4XpfkWFXQvVyW4JdU241/rCgdjZ Rmc7AsilrVjC/Gz8sXzqxiR6rw6T9mHxQC8XvZBAZZZsPJayDoakIksVkplP2AIp 581TGHk0DXbDaRVQjTX7VC3PWMmyGJfhH+l4KJClm9dJIG3cHFNkX5hMetWioGkv C5pzZRvTGQo2HesDM1eIiJrxDImpnD9clxw3cNm0So+/FvrzYsa8zMpFP6fe9jVC ZiURO52+n+zyRMFTNhFbubqpbd7Q34ww7Z2SgeRFuErLz8cDlATvKno7UsY4fOff iFofy9MziFkV8SDZiI2MaIUCEAq/wffzoqkBu0OQRDbXgKQGea1EX/Mjwn2bshJ0 tqAV+1wDK+difY4uATNaHepwxxCNHQgQ22ZlnoKUHqz7JJ0+Q4CZIINnGWnBRETR 92Bu6GvwHPhvnu9RMBDL021qqMaVdTHxj9uOkFXpAs8AgFmBUkAyx7witkcOM/wc xLoSFrJd92QtH+NH7zMlVCVQdPd/vPYfyJxAZbT06w2VqoKSBMM= =SCeB -----END PGP SIGNATURE----- --g0r0FTbW4kq2X4GimhBuhid3xZRNli2Gv--