On Tue, Oct 17, 2017 at 11:01 AM, Andre McCurdy <[email protected]> wrote: > On Tue, Oct 17, 2017 at 10:23 AM, Joe MacDonald > <[email protected]> wrote: >> [Re: [oe] [meta-networking][PATCH 1/1] mdns: move from >> meta-intel-iot-middleware] On 17.10.17 (Tue 10:01) Khem Raj wrote: >>> I do world builds on master-next with musl on qemumips and rpi3 >>> continuously, with a known set of failures, if a new one shows up >>> I tend to report. >> >> I obviously haven't been building with musl either, but if it's not that >> difficult to add to my autobuilder configuration, I'd be happy to. Is >> https://github.com/kraj/meta-musl/blob/master/README.md the best source >> for how to get started with a musl build? > > Going slightly off topic, but even though sources can be patched to > fix build issues due to musl's lack of nss.h, it doesn't really make > any sense to build nss modules for musl, since these modules can only > ever be used by glibc. > > e.g. we have musl patches in recipes such as nss-myhostname, but the > real fix is probably just to not build the recipe at all for musl. > > In this particular case I don't know whether mDNSResponder is tied > tightly to nss or whether the nss module is an optional add-on but > that should probably be determined before coming up with fix.
yes musl does not have nss and its intentional. Although mDNS can be implemented external to libc which I believe is a good thing. I would be happy if the patches to add these recipes relying on nss just added musl to incompatible hosts for immediate OE concerns. Eventually, it could have a different fix upstream for the given component, if they want to make nss support conditional or otherwise, its best left to component -- _______________________________________________ Openembedded-devel mailing list [email protected] http://lists.openembedded.org/mailman/listinfo/openembedded-devel
