Jeff King [off-list ref] writes:

On Sun, Jul 14, 2024 at 12:17:53PM +0200, René Scharfe wrote:

quoted

+	do { \
+		const char *expect[] = { __VA_ARGS__ }; \
+		if (check_uint(ARRAY_SIZE(expect), >, 0) && \
+		    check_pointer_eq(expect[ARRAY_SIZE(expect) - 1], NULL) && \
+		    check_uint((vec)->nr, ==, ARRAY_SIZE(expect) - 1) && \
+		    check_uint((vec)->nr, <=, (vec)->alloc)) { \
+			for (size_t i = 0; i < ARRAY_SIZE(expect); i++) { \
+				if (!check_str((vec)->v[i], expect[i])) { \
+					test_msg("      i: %"PRIuMAX, i); \
+					break; \
+				} \
+			} \
+		} \
+	} while (0)

The linux32 CI job seems to complain about this, since the concrete type
of "i" (a size_t) is "unsigned int" there, but PRIuMAX is %llu.
Presumably you just need to cast to uintmax_t.

Yeah, I noticed it, too, but unfortunately it was after pushing it
out and seeing it break there X-<.