It can't be quite that simple. That fails on both Linux and OSX, because there is no libatomic (at least on my machines and with my versions of gcc/etc).
I can totally buy that adding atomic to the target_link_libraries just for atomic_test is the right thing to do, but it needs to be guarded by some kind of test so that it only does it on the systems that need it. > On Oct 4, 2019, at 3:10 AM, Matteo F. Vescovi <mfvesc...@gmail.com> wrote: > > Hi! > > Sorry for the late reply. > > On 2019-09-09 at 17:05 (-07), Larry Gritz wrote: >> Just pushed an update. Maybe try now? >> >> If that doesn't work, I'm not sure what else to try. I think that to >> get this fixed on these two niche architectures, it's going to take a >> contribution from somebody who has access to and expertise on that >> platform and knows what they're doing. I'm just making blind guesses. > > The solution was even easier. > Attached, you'll find a one-line change provided by my fellow DD Mattia > (Cc-ed) which I'd like you could merge to master branch so I can remove > the patching on next release. > > It's already part of Debian OIIO package and it has been verified to > build fine on amd64/i386/armel/mipsel, actually. > > Thanks a lot. > > > -- > Matteo F. Vescovi > > <0001-Fix_FTBFS_on_armel-mipsel.patch> -- Larry Gritz l...@larrygritz.com
_______________________________________________ Oiio-dev mailing list Oiio-dev@lists.openimageio.org http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org
