uid-range: fix out-of-bounds write in uid_range_partition()

uid_range_partition() filled the grown entries[] buffer backwards in
place. The backward-fill invariant (the write cursor stays above the
read index) only holds when every source entry contributes at least
one partition; an entry with nr < size contributes zero, so the cursor
stalls while the read index keeps descending. A later multi-part
entry's writes then overwrite the still-live zero-part slot, the
corrupted slot is re-read as a one-part entry, and the next
range->entries[--t] underflows.

Add a forward compaction first pass that drops the zero-part entries
before the backward fill.

Follow-up for 025439faaa

Co-Authored-by: Paul Meyer <katexochen0@gmail.com>

src/basic/uid-range.c

@@ -395,9 +395,18 @@ int uid_range_partition(UIDRange *range, uid_t size) {
         if (n_new_entries > range->n_entries && !GREEDY_REALLOC(range->entries, n_new_entries))
                 return -ENOMEM;
 
-        /* Work backwards to avoid overwriting entries we still need to read */
+        /* Compact in place: drop entries that contribute zero partitions (nr < size). This forward pass
+         * reads each entry once and only writes to lower-or-equal indices, so it cannot alias an unread
+         * source entry. */
+        size_t n_src = 0;
+        for (size_t i = 0; i < range->n_entries; i++)
+                if (range->entries[i].nr >= size)
+                        range->entries[n_src++] = range->entries[i];
+
+        /* Pre-compaction guarantees every surviving entry contributes at least one partition slot, so the
+         * write cursor t stays ahead of the read index. */
         size_t t = n_new_entries;
-        for (size_t i = range->n_entries; i > 0; i--) {
+        for (size_t i = n_src; i > 0; i--) {
                 UIDRangeEntry *e = range->entries + i - 1;
                 unsigned n_parts = e->nr / size;