On 25.02.2019 11:23, Martin Husemann wrote: > On Mon, Feb 25, 2019 at 01:19:02PM +0300, Valery Ushakov wrote: >> On Mon, Feb 25, 2019 at 06:23:33 +0100, Martin Husemann wrote: >> >>> On Sun, Feb 24, 2019 at 09:43:52PM +0100, Kamil Rytarowski wrote: >>>> I consider that this is just GCC specific behavior to make some warnings >>>> fatal depending on driver configuration. This is typical behavior of GCC >>>> that we keep observing all over again. >>> >>> No, this is very different to optimizer and target specific warnings. >>> >>> There must be a fundamental difference in includes, it should all be pretty >>> obvious from comparing .depend / *.d files. >>> >>> IMO this should be analyzed clearly instead of papered over. >> >> GENERIC does not enable SCTP, but ALL does. Do we build ALL as part >> of autobuilds? > > Duh - no, we don't. I guess we should fix that... > > Martin > This is right, SCTP build wasn't tested and I got it enabled locally. $ grep -r SCTP sys/arch/i386/conf/ sys/arch/i386/conf/ALL:options SCTP # Stream Control Transmission Protocol sys/arch/i386/conf/ALL:options SCTP_DEBUG The same with amd64. Robert Swindells mentioned that he had a patch for the same reason to unbreak build locally. I propose to build ALL by default as we now depend on external services to have functional 24/7 build with sanitizers (syzkaller).
Attachment:
signature.asc
Description: OpenPGP digital signature