It doesn't do any good, and it breaks the build on CentOS.

This is a useless warning on any architecture with a barrel shifter,
which includes Intel and ARM and basically everything nowadays. Even
so, we'd enable it out of perfectionism, but the GCC directives that
ought to be useful for forcing structure alignment in order to
suppress it locally don't seem to be working quite right.

Matt Nordhoff authored 7 years ago and Matt Selsky committed 7 years ago

Two of the headers generated a build warning and weren't displayed.

[ 79/166] Compiling docs/rollover.txt
asciidoc: WARNING: rollover.txt: line 210: missing section: [gps_pivots]
asciidoc: WARNING: rollover.txt: line 259: missing section: [ntp_pivots]

Also fix a typo.

Follow-up to "Address Gitlab issue #376: Need a symlink from ./tests/pylib/ntp"

It was already known that the AM_NEWPASS case could be hit during pool
processing.  Apparently AM_MANYCAST can too.  Now guarding all three
cases just to be on the safe side.

Intended to prevent 'cast increases required alignment of target type/
on 32-bit machines.

distclean and clean are different. "clean" is part of build and conditional
casing for "clean" therefore makes sense.

It will mean more to the user than the internal name

Members that used to be copied ad hoc from the peer_ctl structure into
separate slots in a new peer structure are now, instead, copied into
a struct peer_ctl that retains their grouping.

This commit does no logic changes.  It is preparation for cleaning up the
logic in newpeer() that does the copies to do the block copy directly.
The ultimate goal is to eliminate friction when defining new struct peer_ctl
members in the configurator.