From mboxrd@z Thu Jan 1 00:00:00 1970 From: Adolf Belka To: development@lists.ipfire.org Subject: Re: pmacct marked as SKIP during IPFire build Date: Thu, 15 Jul 2021 19:48:34 +0200 Message-ID: <3d71249a-3c95-c8c9-0c6f-5ce78cf64fed@ipfire.org> In-Reply-To: <05DFA074-F1F8-4603-865B-FA76CEA596C3@ipfire.org> MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="===============6329245037970832692==" List-Id: --===============6329245037970832692== Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Hi Michael, On 15/07/2021 18:40, Michael Tremer wrote: > Hello, >=20 > This is just a warning that is being treated as an error. >=20 > GCC seems to get the size calculation wrong, but it generally doesn=E2=80= =99t look like good coding style. >=20 > Could you try this patch? Tried it and it worked but then the build of pmacct stopped for the exact sam= e error. >=20 > https://gitweb.gentoo.org/repo/gentoo.git/tree/dev-libs/libcdada/files/libc= dada-0.3.4-Werror.patch Looked in gentoo for a pmacct werror patch and found one for 1.7.4 but the po= sitioning is too far away from the lines in 1.7.6 for the patch to work. So I= am creating a new patch that removes the Werror parts from the configure.ac = in pmacct-1.7.6. Hopefully that should then work as it did for libcdada. Regards, Adolf. >=20 > -Michael >=20 >> On 15 Jul 2021, at 17:10, Adolf Belka wrote: >> >> Hi Michael, All, >> >> On 15/07/2021 14:53, Adolf Belka wrote: >>> Hi Michael, >>> >>> On 15/07/2021 13:55, Michael Tremer wrote: >>>> Hey, >>>> >>>> I am not the export here, but I believe this was disabled because libcda= da does not build with GCC 11. >>>> >>>> Is there an update maybe? >>> I have had a look and there is an update for libcdada. Doesn't mention an= ything about problems with gcc in its changelog. >>> Will give a go to testing with existing and also with the new libcdada ve= rsion and see how it goes. >>>> >> So I took out the SUP_ARCH lines and did a build with both libcdada-0.3.4 = and libcdada-0.3.5 and got the same error with both. >>>> You can just enable it again (remove the SUP_ARCHES line) and see what h= appens. Feel free to post the log here. >> >> The log for the 0.3.4 version is at the end of this email. >> >> There is a line about all warnings being treated as errors and searching o= n that it seems that might be a consequence of GCC 11. My internet search see= med to indicate that will happen with GCC 11 if you have Werror in the flags = and that is present in both CFLAGS and CXXFLAGS. The suggestion in those post= s I read was to remove the Werror entry from the two sets of flags but as I a= m not experienced with those types of things I thought I would ask here. >> >> The actual error or warning appears to be related to a variable that is be= ing flagged that it might exceed the maximum object size. You can find the er= ror message near the end of the log below. >> >> There were also a lot of warnings in the autogen.sh part of the build with= obsolete macros being flagged up and suggestion to run autoupdate. I tried p= utting that after the autogen.sh line but it made no difference to the error = that stopped the build, although it might be good to also have in place. >> >> Any suggestions for things to try out will be gladly welcomed. >> >> Regards, >> >> Adolf. >> >>> Will do so if I have any problems. >>> >>> Thanks for the info. >>> >>> Regards, >>> Adolf. >>>> >>>> Best, >>>> -Michael >>>> >>>>> On 15 Jul 2021, at 11:58, Adolf Belka wrote: >>>>> >>>>> Hi All, >>>>> >>>>> I have been doing some update builds and while doing that with "next" I= noticed that pmacct and its dependency libcdada were both marked as SKIP and= so didn't get built. >>>>> >>>>> What is the reason for this? Has there been some problem that needs to = be fixed? >>>>> >>>>> >>>>> Regards, >>>>> >>>>> Adolf. >>>>> >>>> >> >=20 --===============6329245037970832692==--