Re: [kvm-unit-tests PATCH v2 2/2] s390x: add migration TOD clock test

[Claudio Imbrenda <imbrenda@linux.ibm.com>]

On Tue, 11 Oct 2022 17:14:33 +0200
Nico Boehr <nrb@linux.ibm.com> wrote:

> On migration, we expect the guest clock value to be preserved. Add a
> test to verify this:
> - advance the guest TOD by much more than we need to migrate
> - migrate the guest
> - get the guest TOD
> 
> After migration, assert the guest TOD value is at least the value we set
> before migration.
> 
> Signed-off-by: Nico Boehr <nrb@linux.ibm.com>
> ---
>  s390x/Makefile        |  1 +
>  s390x/migration-sck.c | 44 +++++++++++++++++++++++++++++++++++++++++++
>  s390x/unittests.cfg   |  4 ++++
>  3 files changed, 49 insertions(+)
>  create mode 100644 s390x/migration-sck.c
> 
> diff --git a/s390x/Makefile b/s390x/Makefile
> index 649486f2d4a0..fba09bc2df3a 100644
> --- a/s390x/Makefile
> +++ b/s390x/Makefile
> @@ -36,6 +36,7 @@ tests += $(TEST_DIR)/migration-cmm.elf
>  tests += $(TEST_DIR)/migration-skey.elf
>  tests += $(TEST_DIR)/panic-loop-extint.elf
>  tests += $(TEST_DIR)/panic-loop-pgm.elf
> +tests += $(TEST_DIR)/migration-sck.elf
>  
>  pv-tests += $(TEST_DIR)/pv-diags.elf
>  
> diff --git a/s390x/migration-sck.c b/s390x/migration-sck.c
> new file mode 100644
> index 000000000000..286a693b4858
> --- /dev/null
> +++ b/s390x/migration-sck.c
> @@ -0,0 +1,44 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * SET CLOCK migration tests
> + *
> + * Copyright IBM Corp. 2022
> + *
> + * Authors:
> + *  Nico Boehr <nrb@linux.ibm.com>
> + */
> +
> +#include <libcflat.h>
> +#include <asm/time.h>
> +
> +static void test_sck_migration(void)
> +{
> +	uint64_t now_before_set = 0, now_after_migration, time_to_set, time_to_advance;
> +	int cc;
> +
> +	stckf(&now_before_set);
> +
> +	/* Advance the clock by a lot more than we might ever need to migrate (600s) */
> +	time_to_advance = (600ULL * 1000000) << STCK_SHIFT_US;
> +	time_to_set = now_before_set + time_to_advance;
> +
> +	cc = sck(&time_to_set);
> +	report(!cc, "setting clock succeeded");

I would also check here that the clock is set. This way we can
differentiate between a migration issue or a more general (and more
severe) clock issue. Even if we have this test somewhere else, this
makes it easier when debugging to narrow down the issue to migration.

> +
> +	puts("Please migrate me, then press return\n");
> +	(void)getchar();
> +
> +	cc = stckf(&now_after_migration);
> +	report(!cc, "clock still set");
> +
> +	report(now_after_migration >= time_to_set, "TOD clock value preserved");
> +}
> +
> +int main(void)
> +{
> +	report_prefix_push("migration-sck");
> +
> +	test_sck_migration();
> +	report_prefix_pop();
> +	return report_summary();
> +}
> diff --git a/s390x/unittests.cfg b/s390x/unittests.cfg
> index f9f102abfa89..2c04ae7c7c15 100644
> --- a/s390x/unittests.cfg
> +++ b/s390x/unittests.cfg
> @@ -197,3 +197,7 @@ file = panic-loop-pgm.elf
>  groups = panic
>  accel = kvm
>  timeout = 5
> +
> +[migration-sck]
> +file = migration-sck.elf
> +groups = migration