autodoc.pl: Properly handle a single column

The program attempts to make a table of some of the output for
compactness.  But if there is room for only a single column, it was
using the number of rows that it calculated were necessary should two
columns have worked; hence the output was only displaying the first half
of the rows in perlintern.pod.  What is needed is to do the loop a
partial final iteration, just enough to calculate the number of rows for
that final column.

1 files changed, 5 insertions, 1 deletions

diff --git a/autodoc.pl b/autodoc.pl
index 92a4a9c1fc..0b8f361eff 100644
--- a/autodoc.pl
+++ b/autodoc.pl
@@ -1560,12 +1560,16 @@ sub construct_missings_section {
     # can accommodate all the data.  This algorithm doesn't require the
     # resulting columns to all have the same width.  This can allow for
     # as tight of packing as the data will possibly allow.
-    for ($columns = 7; $columns > 1; $columns--) {
+    for ($columns = 7; $columns >= 1; $columns--) {
 
         # For this many columns, we will need this many rows (final row might
         # not be completely filled)
         $rows = (@missings + $columns - 1) / $columns;
 
+        # We only need to execute this final iteration to calculate the number
+        # of rows, as we can't get fewer than a single column.
+        last if $columns == 1;
+
         my $row_width = 1;  # For 1 space indent
         my $i = 0;  # Which missing element