fix(xcat-core): trim the commentary around the Ubuntu NTP fix

setupntp explained systemd-timesyncd in four lines where two carry the reason nothing displaces
it, and repeated the hwclock rationale that the guard below already implies. NTP::Backend's
header narrated that the module exists to be unit-tested.

The test header restated all of it a third time, and said "the first three are behaviours of the
script" about a list of three items whose third is the packaging check. Corrected and shortened;
the note on why setupntp cannot simply be executed is kept, since that is not obvious from
reading it.

Interface documentation is left alone: choose()'s parameter list and default_backend()'s
per-family table are what a caller needs.

18 comment lines removed, no behaviour change.

Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>

perl-xCAT/xCAT/NTP/Backend.pm

@@ -1,9 +1,8 @@
 package xCAT::NTP::Backend;
 
-# Selector for the NTP daemon xCAT configures on a node (chrony vs. ntpd), in the same spirit as
-# xCAT::DHCP::Backend (ISC vs. Kea). makentp/setupntp support chronyd and ntpd only -- there is no
-# systemd-timesyncd path (timesyncd is an SNTP client and cannot serve time to compute nodes), so an
-# xCAT MN always needs chrony or ntpd. This module centralizes and unit-tests the choice.
+# Selector for the NTP daemon xCAT configures (chrony vs. ntpd), in the same spirit as
+# xCAT::DHCP::Backend. There is no timesyncd path: it is an SNTP client and cannot serve time to
+# compute nodes, so an MN always needs chrony or ntpd.
 
 use strict;
 use warnings;
@@ -33,8 +32,8 @@ sub normalize {
 #   requested       -- override (default: site.ntpbackend, else 'auto')
 #   os_name/version -- OS identity (default: detected)
 #   available       -- optional { chrony => 0/1, ntpd => 0/1 } to bypass command detection (tests)
-#   check_available -- when true, downgrade chrony->ntpd (or ntpd->chrony) if the chosen one is
-#                      absent but the other is present, and flag install=1 when neither is present.
+#   check_available -- when true, downgrade to whichever daemon is installed, and flag install=1
+#                      when neither is.
 sub choose {
     my ( $class, %args ) = @_;
 

xCAT-test/unit/makentp_ntp_deps.t

@@ -6,14 +6,10 @@ use FindBin;
 use File::Temp qw(tempdir);
 use Test::More;
 
-# setupntp configures a server-capable NTP daemon on the management node. On a stock Ubuntu MN
-# three things went wrong: hwclock had moved to util-linux-extra and was absent, so the fatal
-# executable check aborted the ENTIRE NTP setup including the clock step that does not use it;
-# systemd-timesyncd was left running and kept disciplining the clock against the daemon being
-# configured; and the xcat package did not pull in a daemon that can serve time at all.
-#
-# The first three are behaviours of the script, so run it. The last is a property of the
-# packaging manifests, so read those.
+# A stock Ubuntu MN broke three ways: a missing hwclock aborted the whole NTP setup,
+# systemd-timesyncd kept fighting the daemon being configured, and the xcat package pulled in no
+# daemon that can serve time. The first two are behaviours of the script, so run it; the third is
+# a property of the packaging manifests, so read those.
 
 my $repo = "$FindBin::Bin/../..";
 my $setupntp_path = "$repo/xCAT/postscripts/setupntp";
@@ -25,10 +21,9 @@ my @lines = <$fh>;
 close $fh;
 my $source = join '', @lines;
 
-# setupntp forces its own PATH and refuses to run unless UID is 0, so its commands cannot be
-# stubbed from outside and it cannot run as an ordinary user. Take the script's own helper
-# functions and the section that configures the daemon, and drive them with shell functions --
-# which bash resolves ahead of PATH, and which `type` and `command -v` both report as present.
+# setupntp forces its own PATH and exits unless UID is 0, so it cannot be stubbed from outside
+# or run unprivileged. Drive its own helper functions with shell functions instead: bash resolves
+# those ahead of PATH, and both `type` and `command -v` report them as present.
 my ($helpers) = $source =~ /\A(.*?)^\[ "\$\{UID\}" -eq "0" \]/ms;
 my ($body)    = $source =~ /^(check_exec_or_exit cp cat logger grep\n.*?)^CHRONY_CONF=/ms;
 BAIL_OUT('could not take the helper functions from setupntp')          unless $helpers;

xCAT/postscripts/setupntp

@@ -146,9 +146,9 @@ unset MASTER
 unset NTPSERVERS
 
 check_exec_or_exit cp cat logger grep
-# hwclock is NOT required: it only persists the (already chronyd-stepped) system clock to the RTC
-# below, and it is optional -- on Ubuntu 24.04+ it moved to util-linux-extra, which is absent on
-# minimal cloud images. Requiring it here aborted the whole NTP setup (and the clock step) on noble.
+# hwclock is NOT required: it only persists the already-stepped clock to the RTC below, and on
+# Ubuntu 24.04+ it moved to util-linux-extra, absent on minimal images. Requiring it here aborted
+# the whole NTP setup, including the clock step that does not use it.
 check_exec_or_exit systemctl timedatectl
 
 systemctl stop ntp.service 2>/dev/null
@@ -159,10 +159,8 @@ systemctl disable ntpd.service 2>/dev/null
 systemctl disable ntp-wait.service 2>/dev/null
 systemctl disable ntpdate.service 2>/dev/null
 
-# systemd-timesyncd is an SNTP client that also disciplines the system clock. It is part of systemd
-# (not a "time-daemon" package), so it coexists with chrony/ntpd at the package level but must be
-# stopped or it fights the NTP daemon we are configuring. This is the default time-sync stack on
-# Ubuntu, where it must yield to chrony (timesyncd cannot serve time to compute nodes).
+# systemd-timesyncd ships as part of systemd rather than a time-daemon package, so nothing
+# displaces it and it keeps disciplining the clock against the daemon being configured.
 systemctl stop systemd-timesyncd.service 2>/dev/null
 systemctl disable systemd-timesyncd.service 2>/dev/null