pxeboot: don't hang with TCP-only NFS server. Add useful error message and manpage hint.

[cracauer]

Summary

When pxeboot gets a 0 as a port number from portmapper (indicating an error), it currently happily sends NFS packets to the server's port 0 in an endless loop. Change this to instead, bail out with a useful message,

This happens, for example, with recent Linux NFS servers as many distributions switched to TCP only NFS serving by default. FreeBSD's pxeboot must have UDP. In this situation pxeboot asks the server's portmapper for the UDP NFS port and since there is none gets 0.

Also add a hint to the manpage explaining this and how to fix it.

Test Plan

This fixes the hang I had for a while.

Automated testing is not really feasible as you would have to build client and server and pxeboot can't be used from bhyve.

[cracauer]

Note: this is not related to the slowness that people experience with some but not all BIOSes.

[ziaee]

stand/i386/pxeldr/pxeboot.8

133–136

.Nm
requires NFS over UDP.
Many recent distributions of Linux only serve NFS over TCP.
Enable UDP by uncommenting
.Ql udp=y
in
.Pa /etc/nfs.conf .
.Pp

Notes:
+ one sentance per lin in roff.
+ the Pa macro makes this page go into the apropos database so that people can search nfs.conf and get here
+ no "you" in reference manual style

[cracauer]

Manpage edit as suggested.

[cracauer]

Fix next line after manpage change.

[ziaee]

Ok from manpages side; but I'd like a more senior programmer to review the code change. It might help if you upload the patch with more context, either using -U 9999 or git-arc.

[ziaee]

(adding senior colleagues who worked on stand/libsa lately)