On 2/12/21 12:15 PM, Cornelia Huck wrote:
On Wed, 10 Feb 2021 14:20:14 +0100
Pierre Morel <pmorel@xxxxxxxxxxxxx> wrote:
Measurement block format 1 is made available by the extended
mesurement block facility and is indicated in the SCHIB by
s/mesurement/measurement/
50% good and yes, 50% bad, I change it thanks.
the bit in the PMCW.
The MBO is specified in the SCHIB of each channel and the MBO
defined by the SCHM instruction is ignored.
The test of the MB format 1 is just skipped if the feature is
not available.
Signed-off-by: Pierre Morel <pmorel@xxxxxxxxxxxxx>
---
lib/s390x/css.h | 14 ++++++++++++++
s390x/css.c | 36 ++++++++++++++++++++++++++++++++++++
2 files changed, 50 insertions(+)
(...)
+static void test_schm_fmt1(void)
+{
+ struct measurement_block_format1 *mb1;
+
+ report_prefix_push("Format 1");
+
+ mb1 = alloc_io_mem(sizeof(struct measurement_block_format1), 0);
+ if (!mb1) {
+ report_abort("measurement_block_format1 allocation failed");
+ goto end;
+ }
+
+ schm(NULL, 0); /* Clear previous MB address */
Same comment as for the last patch.
Yes,
+ schm(0, SCHM_MBU);
+
+ /* Expect error for non aligned MB */
+ report_prefix_push("Unaligned MB origin");
+ report_xfail(start_measure((u64)mb1 + 1, 0, true), mb1->ssch_rsch_count != 0,
+ "SSCH measured %d", mb1->ssch_rsch_count);
+ report_prefix_pop();
+
+ memset(mb1, 0, sizeof(*mb1));
+
+ /* Expect success */
+ report_prefix_push("Valid MB address and index");
+ report(start_measure((u64)mb1, 0, true) &&
+ mb1->ssch_rsch_count == SCHM_UPDATE_CNT,
+ "SSCH measured %d", mb1->ssch_rsch_count);
+ report_prefix_pop();
+
+ free_io_mem(mb1, sizeof(struct measurement_block_format1));
Also here, you need to stop the measurements before freeing the block.
yes, I will.
Thanks,
Pierre
--
Pierre Morel
IBM Lab Boeblingen
