Refactor AES context to be shallow-copyable

Replace RK pointer in AES context with a buffer offset, to allow
shallow copying. Fixes #2147.

Signed-off-by: Werner Lewis <werner.lewis@arm.com>
diff --git a/library/aesni.c b/library/aesni.c
index be226c9..87d818a 100644
--- a/library/aesni.c
+++ b/library/aesni.c
@@ -127,7 +127,7 @@
          "3:                        \n\t"
          "movdqu    %%xmm0, (%4)    \n\t" // export output
          :
-         : "r" (ctx->nr), "r" (ctx->rk), "r" (mode), "r" (input), "r" (output)
+         : "r" (ctx->nr), "r" (ctx->buf + ctx->rk_offset), "r" (mode), "r" (input), "r" (output)
          : "memory", "cc", "xmm0", "xmm1" );