On Mon, Jul 10, 2017 at 08:27:11AM +0300, Leon Romanovsky wrote: > On Sun, Jul 09, 2017 at 10:42:03AM -0600, Jason Gunthorpe wrote: > > There is nothing special about bnxt_re - it should build in all cases, > > unless your source tree is older than the patch. Are you sure your > > build step is not failing for some reason? > > > > > Adding Jason, Seemingly a bug in the make-install path. > > > > Looks to me like a build problem, install assumes a successfull build, > > if it expects bnxt_re then it should have been built. > > > > The other option is your build output tree got confused when you > > switched between a tree before/after bnxt_re introduction, but even that is > > supposed to work automatically. > > Jason, > > They are building with build.sh script with -DIN_PLACE=1. > > It is documentation issue on the rationale of building _in_place_ and why > installation is not needed for the developers and how RPM/DEB files > of rdma-core are generated. Let me see if I can make 'make install' fail with a more sensible error in this case. As you said, 'make install' with '-DIN_PLACE' is not allowed or sane. Jason -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
