On Fri, 2022-05-13 at 09:50 +0000, Janosch Frank wrote:
> Let's check if the UV really protected all the memory we donated.
>
> Signed-off-by: Janosch Frank <frankja@xxxxxxxxxxxxx>
Reviewed-by: Nico Boehr <nrb@xxxxxxxxxxxxx>
One suggestion below for you to consider.
> ---
> s390x/uv-host.c | 42 ++++++++++++++++++++++++++++++++++++++++--
> 1 file changed, 40 insertions(+), 2 deletions(-)
>
> diff --git a/s390x/uv-host.c b/s390x/uv-host.c
> index a1a6d120..0f0b18a1 100644
> --- a/s390x/uv-host.c
> +++ b/s390x/uv-host.c
> @@ -142,7 +142,8 @@ static void test_cpu_destroy(void)
> static void test_cpu_create(void)
> {
> int rc;
> - unsigned long tmp;
> + unsigned long tmp, i;
> + uint8_t *access_ptr;
>
> report_prefix_push("csc");
> uvcb_csc.header.len = sizeof(uvcb_csc);
> @@ -194,6 +195,18 @@ static void test_cpu_create(void)
> report(rc == 0 && uvcb_csc.header.rc == UVC_RC_EXECUTED &&
> uvcb_csc.cpu_handle, "success");
>
> + rc = 1;
> + for (i = 0; i < uvcb_qui.cpu_stor_len / PAGE_SIZE; i++) {
> + expect_pgm_int();
> + access_ptr = (void *)uvcb_csc.stor_origin + PAGE_SIZE
> * i;
> + *access_ptr = 42;
> + if (clear_pgm_int() !=
> PGM_INT_CODE_SECURE_STOR_ACCESS) {
> + rc = 0;
> + break;
> + }
> + }
> + report(rc, "Storage protection");
All of these for loops look pretty similar, would it make sense to move
them to their own function like:
assert_range_write_protected(void *start, size_t len, int
pgm_int_code)?
