arm64: Align clear_page() to 16 bytes

With automatic zeroing of pages upon allocation enabled, time spent in clear_page() is significant according to perf top. Since clear_page() is used in loops, align it to 16 bytes, which significantly reduces CPU time spent in clear_page()
Ristovski · Nov 28, 2023 · 4a3be2c · 4a3be2c
1 parent 66d6acd
commit 4a3be2c
Showing 1 changed file with 1 addition and 0 deletions.
diff --git a/arch/arm64/lib/clear_page.S b/arch/arm64/lib/clear_page.S
@@ -14,6 +14,7 @@
  * Parameters:
  *	x0 - dest
  */
+    .p2align 4
 SYM_FUNC_START(clear_page)
 	mrs	x1, dczid_el0
 	and	w1, w1, #0xf