k210: fix k210_pll_calc_config()

The k210 driver is selected by sandbox_defconfig.
Building the sandbox on 32bit systems fails with:

test/dm/k210_pll.c: In function ‘dm_test_k210_pll_calc_config’:
include/linux/bitops.h:11:38: warning:
left shift count >= width of type [-Wshift-count-overflow]
   11 | #define BIT(nr)         (1UL << (nr))
      |                              ^~
test/dm/k210_pll.c:36:54: note: in expansion of macro ‘BIT’
   36 |                         error = abs((error - BIT(32))) >> 16;
      |                                              ^~~

Use the BIT_ULL() macro to create a u64 value.
Replace abs() by abs64() to get correct results on 32bit system
Apply the same for the unit test.

Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Sean Anderson <seanga2@gmail.com>
diff --git a/test/dm/k210_pll.c b/test/dm/k210_pll.c
index a0cc84c..354720f 100644
--- a/test/dm/k210_pll.c
+++ b/test/dm/k210_pll.c
@@ -33,7 +33,7 @@
 				error = DIV_ROUND_CLOSEST_ULL(f * inv_ratio,
 							      r * od);
 				/* The lower 16 bits are spurious */
-				error = abs((error - BIT(32))) >> 16;
+				error = abs64((error - BIT_ULL(32))) >> 16;
 				if (error < best_error) {
 					best->r = r;
 					best->f = f;