Sent from my iPhone
> On May 8, 2014, at 1:29, Andrew Deason <[email protected]> wrote: > > On Wed, 7 May 2014 21:29:26 -0400 > Coy Hile <[email protected]> wrote: > >> gcc -I/home/hile/openafs/src/config -I/home/hile/openafs/include -I. -I. >> -m64 -o rmtsysd rmtsysd.o libsys.a /home/hile/openafs/lib/librx.a libsys.a >> /home/hile/openafs/lib/liblwp.a /home/hile/openafs/lib/util.a -lresolv >> -lsocket -lnsl -lintl -ldl >> Undefined first referenced >> symbol in file >> savecontext /home/hile/openafs/lib/liblwp.a(lwp.o) >> returnto /home/hile/openafs/lib/liblwp.a(lwp.o) >> ld: fatal: symbol referencing errors. No output written to rmtsysd >> collect2: error: ld returned 1 exit status >> >> Is there something simple that I’ve missed here? If it would be >> helpful, I can tar up this build tree and stuff it in AFS somewhere. >> (obviously, this box has no access to AFS itself.) > > Those symbols are very platform-specific, so I don't think it's a normal > simple mistake and it actually has to do with these changes. Those > symbols are supposed to be in liblwp.a, from process.o which gets built > from process.amd64.s. Can you look in src/lwp and see what symbols are > in process.o, or what happened when trying to build it? > > Also it would be helpful if we knew exactly what source you're building > from. I'm not clear on if you're building directly from the changeset > referenced, or if you added stuff or pull it to modern master, etc. > (It'll need to be pulled to current master at some The tip of my branch now is 8d338b5 Before my changes that is. I'll attempt against master this evening and get a diff into gerrit.
