Skip to contentSkip to conversation

s390x: add migration test for storage keys

PATCH Superseded Nico Boehr 3 participants 8 messages May 16, 2022 – May 17, 2022

Subsystems (2)

v2 of 6 Superseded Latest is v6 →

lore
Share & tools
lore

Nico Boehr

author
#0 lore raw

v2 cover s390x: add migration test for storage keys

v1->v2:
----
* As per discussion with Janis and Claudio, remove the actual access check from the test. This also allows us to remove the check_pgm_int_code_xfail() patch.
* Typos/Style suggestions (thanks Janis)

Upon migration, we expect storage keys being set by the guest to be preserved, so add a test for it.

We keep 128 pages and set predictable storage keys. Then, we migrate and check they can be read back and the respective access restrictions are in place when the access key in the PSW doesn't match.

Nico Boehr (1):
s390x: add migration test for storage keys

 s390x/Makefile         |  1 +
 s390x/migration-skey.c | 78 ++++++++++++++++++++++++++++++++++++++++++
 s390x/unittests.cfg    |  4 +++
 3 files changed, 83 insertions(+)
 create mode 100644 s390x/migration-skey.c
Signature · 1 line
2.31.1
author
#1 lore raw

PATCH v2 1/1 s390x: add migration test for storage keys

Upon migration, we expect storage keys being set by the guest to be preserved, so add a test for it.

We keep 128 pages and set predictable storage keys. Then, we migrate and check they can be read back and the respective access restrictions are in place when the access key in the PSW doesn't match.

---
 s390x/Makefile         |  1 +
 s390x/migration-skey.c | 78 ++++++++++++++++++++++++++++++++++++++++++
 s390x/unittests.cfg    |  4 +++
 3 files changed, 83 insertions(+)
 create mode 100644 s390x/migration-skey.c
s390x/Makefile +1 -0
index a8e04aa6fe4d..f8ea594b641d 100644
@@ -32,6 +32,7 @@ tests += $(TEST_DIR)/epsw.elf
 tests += $(TEST_DIR)/adtl-status.elf
 tests += $(TEST_DIR)/migration.elf
 tests += $(TEST_DIR)/pv-attest.elf
+tests += $(TEST_DIR)/migration-skey.elf
 
 pv-tests += $(TEST_DIR)/pv-diags.elf
 
s390x/migration-skey.c new file +78 -0

78 lines not on this page · raw message

s390x/unittests.cfg +4 -0
index b456b2881448..1e851d8e3dd8 100644
@@ -176,3 +176,7 @@ extra_params = -cpu qemu,gs=off,vx=off
 file = migration.elf
 groups = migration
 smp = 2
+
+[migration-skey]
+file = migration-skey.elf
+groups = migration
Signature · 1 line
2.31.1
Reviewed-by on v2 1/1
#2 lore raw
Quoted 48 lines · On Mon, 16 May 2022 11:07:02 +0200 Nico Boehr <[email protected]> wrote:

Upon migration, we expect storage keys being set by the guest to be preserved, so add a test for it.

We keep 128 pages and set predictable storage keys. Then, we migrate and check they can be read back and the respective access restrictions are in place when the access key in the PSW doesn't match.

---
 s390x/Makefile         |  1 +
 s390x/migration-skey.c | 78 ++++++++++++++++++++++++++++++++++++++++++
 s390x/unittests.cfg    |  4 +++
 3 files changed, 83 insertions(+)
 create mode 100644 s390x/migration-skey.c
s390x/Makefile +1 -0
index a8e04aa6fe4d..f8ea594b641d 100644
@@ -32,6 +32,7 @@ tests += $(TEST_DIR)/epsw.elf
 tests += $(TEST_DIR)/adtl-status.elf
 tests += $(TEST_DIR)/migration.elf
 tests += $(TEST_DIR)/pv-attest.elf
+tests += $(TEST_DIR)/migration-skey.elf
 
 pv-tests += $(TEST_DIR)/pv-diags.elf
 
s390x/migration-skey.c new file +15 -0
new file mode 100644
index 000000000000..ee4622eb94ba
@@ -0,0 +1,78 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Storage Key migration tests
+ *
+ * Copyright IBM Corp. 2022
+ *
+ * Authors:
+ *  Nico Boehr <[email protected]>
+ */
+
+#include <libcflat.h>
+#include <asm/facility.h>
+#include <asm/page.h>
+#include <asm/mem.h>
+#include <asm/interrupt.h>

s390x/migration-skey.c line 25

Quoted 10 lines
s390x/migration-skey.c new file +10 -0
+#include <hardware.h>
+
+#define NUM_PAGES 128
+static uint8_t pagebuf[NUM_PAGES][PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
+
+static void test_migration(void)
+{
+	int i, key_to_set;
+	uint8_t *page;
+	union skey expected_key, actual_key;

please reverse Christmas tree ^

with that fixed:

Quoted 65 lines
s390x/migration-skey.c new file +53 -0
+
+	for (i = 0; i < NUM_PAGES; i++) {
+		/*
+		 * Storage keys are 7 bit, lowest bit is always returned as zero
+		 * by iske
+		 */
+		key_to_set = i * 2;
+		set_storage_key(pagebuf[i], key_to_set, 1);
+	}
+
+	puts("Please migrate me, then press return\n");
+	(void)getchar();
+
+	for (i = 0; i < NUM_PAGES; i++) {
+		report_prefix_pushf("page %d", i);
+
+		page = &pagebuf[i][0];
+		actual_key.val = get_storage_key(page);
+		expected_key.val = i * 2;
+
+		/* ignore reference bit */
+		actual_key.str.rf = 0;
+		expected_key.str.rf = 0;
+
+		report(actual_key.val == expected_key.val, "expected_key=0x%x actual_key=0x%x", expected_key.val, actual_key.val);
+
+		report_prefix_pop();
+	}
+}
+
+int main(void)
+{
+	report_prefix_push("migration-skey");
+	if (test_facility(169)) {
+		report_skip("storage key removal facility is active");
+
+		/*
+		 * If we just exit and don't ask migrate_cmd to migrate us, it
+		 * will just hang forever. Hence, also ask for migration when we
+		 * skip this test altogether.
+		 */
+		puts("Please migrate me, then press return\n");
+		(void)getchar();
+
+		goto done;
+	}
+
+	test_migration();
+
+done:
+	report_prefix_pop();
+	return report_summary();
+}
s390x/unittests.cfg +4 -0
index b456b2881448..1e851d8e3dd8 100644
@@ -176,3 +176,7 @@ extra_params = -cpu qemu,gs=off,vx=off
 file = migration.elf
 groups = migration
 smp = 2
+
+[migration-skey]
+file = migration-skey.elf
+groups = migration
#3 lore raw
In reply to Nico Boehr
Quoted 2 lines · Nico Boehr

Upon migration, we expect storage keys being set by the guest to be preserved, so add a test for it.

"being set" implies that keys are set while the migration is going on. That's not the case, is it?

Quoted 3 lines

We keep 128 pages and set predictable storage keys. Then, we migrate and check they can be read back and the respective access restrictions are in place when

... check that they ...

Quoted 1 line

the access key in the PSW doesn't match.

The latter half of the sentence doesn't apply anymore, now that you simplified the test. So maybe something like: ... and check that they can be read back and match the value originally set.

Quoted 60 lines
---
 s390x/Makefile         |  1 +
 s390x/migration-skey.c | 78 ++++++++++++++++++++++++++++++++++++++++++
 s390x/unittests.cfg    |  4 +++
 3 files changed, 83 insertions(+)
 create mode 100644 s390x/migration-skey.c
s390x/Makefile +1 -0
index a8e04aa6fe4d..f8ea594b641d 100644
@@ -32,6 +32,7 @@ tests += $(TEST_DIR)/epsw.elf
 tests += $(TEST_DIR)/adtl-status.elf
 tests += $(TEST_DIR)/migration.elf
 tests += $(TEST_DIR)/pv-attest.elf
+tests += $(TEST_DIR)/migration-skey.elf
 
 pv-tests += $(TEST_DIR)/pv-diags.elf
 
s390x/migration-skey.c new file +33 -0
new file mode 100644
index 000000000000..ee4622eb94ba
@@ -0,0 +1,78 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Storage Key migration tests
+ *
+ * Copyright IBM Corp. 2022
+ *
+ * Authors:
+ *  Nico Boehr <[email protected]>
+ */
+
+#include <libcflat.h>
+#include <asm/facility.h>
+#include <asm/page.h>
+#include <asm/mem.h>
+#include <asm/interrupt.h>
+#include <hardware.h>
+
+#define NUM_PAGES 128
+static uint8_t pagebuf[NUM_PAGES][PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
+
+static void test_migration(void)
+{
+	int i, key_to_set;
+	uint8_t *page;
+	union skey expected_key, actual_key;
+
+	for (i = 0; i < NUM_PAGES; i++) {
+		/*
+		 * Storage keys are 7 bit, lowest bit is always returned as zero
+		 * by iske
+		 */
+		key_to_set = i * 2;
+		set_storage_key(pagebuf[i], key_to_set, 1);

s390x/migration-skey.c line 43

Quoted 10 lines
s390x/migration-skey.c new file +10 -0
+	}
+
+	puts("Please migrate me, then press return\n");
+	(void)getchar();
+
+	for (i = 0; i < NUM_PAGES; i++) {
+		report_prefix_pushf("page %d", i);
+
+		page = &pagebuf[i][0];
+		actual_key.val = get_storage_key(page);

The page variable is kinda useless now, I'd just do get_storage_key(pagebuf[0]).

s390x/migration-skey.c line 46

Quoted 3 lines
s390x/migration-skey.c new file +3 -0
+		expected_key.val = i * 2;
+
+		/* ignore reference bit */

Why? Are there any implicit references I'm missing?

Quoted 44 lines
s390x/migration-skey.c new file +32 -0
+		actual_key.str.rf = 0;
+		expected_key.str.rf = 0;
+
+		report(actual_key.val == expected_key.val, "expected_key=0x%x actual_key=0x%x", expected_key.val, actual_key.val);
+
+		report_prefix_pop();
+	}
+}
+
+int main(void)
+{
+	report_prefix_push("migration-skey");
+	if (test_facility(169)) {
+		report_skip("storage key removal facility is active");
+
+		/*
+		 * If we just exit and don't ask migrate_cmd to migrate us, it
+		 * will just hang forever. Hence, also ask for migration when we
+		 * skip this test altogether.
+		 */
+		puts("Please migrate me, then press return\n");
+		(void)getchar();
+
+		goto done;
+	}
+
+	test_migration();
+
+done:
+	report_prefix_pop();
+	return report_summary();
+}
s390x/unittests.cfg +4 -0
index b456b2881448..1e851d8e3dd8 100644
@@ -176,3 +176,7 @@ extra_params = -cpu qemu,gs=off,vx=off
 file = migration.elf
 groups = migration
 smp = 2
+
+[migration-skey]
+file = migration-skey.elf
+groups = migration
author
#4 lore raw
Quoted 8 lines · Janis Schoetterl-Glausch
Quoted 3 lines · Nico Boehr

Upon migration, we expect storage keys being set by the guest to be preserved,
so add a test for it.

"being set" implies that keys are set while the migration is going on.
That's not the case, is it?

Fixed.

Quoted 6 lines
Quoted 4 lines

We keep 128 pages and set predictable storage keys. Then, we migrate and check
they can be read back and the respective access restrictions are in place when

... check that they ...

Added that.

Quoted 8 lines
Quoted 2 lines

the access key in the PSW doesn't match.

The latter half of the sentence doesn't apply anymore, now that you simplified the test.
So maybe something like: ... and check that they can be read back and match the value
originally set.

Fixed.

Quoted 46 lines
Quoted 46 lines
s390x/migration-skey.c new file +19 -0
new file mode 100644
index 000000000000..ee4622eb94ba
@@ -0,0 +1,78 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Storage Key migration tests
+ *
+ * Copyright IBM Corp. 2022
+ *
+ * Authors:
+ *  Nico Boehr <[email protected]>
+ */
+
+#include <libcflat.h>
+#include <asm/facility.h>
+#include <asm/page.h>
+#include <asm/mem.h>
+#include <asm/interrupt.h>
+#include <hardware.h>
+
+#define NUM_PAGES 128
+static uint8_t pagebuf[NUM_PAGES][PAGE_SIZE]

__attribute__((aligned(PAGE_SIZE)));

s390x/migration-skey.c new file +10 -0
+
+static void test_migration(void)
+{
+       int i, key_to_set;
+       uint8_t *page;
+       union skey expected_key, actual_key;
+
+       for (i = 0; i < NUM_PAGES; i++) {
+               /*
+                * Storage keys are 7 bit, lowest bit is always

returned as zero
+                * by iske
+                */
+               key_to_set = i * 2;
+               set_storage_key(pagebuf[i], key_to_set, 1); +       }

s390x/migration-skey.c new file +4 -0
+
+       puts("Please migrate me, then press return\n");
+       (void)getchar();
+
Quoted 8 lines
Quoted 5 lines

+       for (i = 0; i < NUM_PAGES; i++) {
+               report_prefix_pushf("page %d", i); +
+               page = &pagebuf[i][0];
+               actual_key.val = get_storage_key(page);

The page variable is kinda useless now, I'd just do get_storage_key(pagebuf[0]).

Removed.

Quoted 5 lines
Quoted 3 lines

+               expected_key.val = i * 2;
+
+               /* ignore reference bit */

Why? Are there any implicit references I'm missing?

Since the PoP specifies (p. 5-122):

"The record of references provided by the reference bit is not necessarily accurate. However, in the major- ity of situations, reference recording approximately coincides with the related storage reference."

I don't really see a way to test this properly.

Maybe I missed something?

#5 lore raw
Quoted 2 lines · Nico Boehr
Quoted 1 line · Janis Schoetterl-Glausch

On 5/16/22 11:07, Nico Boehr wrote:

[...]

Quoted 8 lines
Quoted 5 lines
Quoted 3 lines

+               expected_key.val = i * 2;
+
+               /* ignore reference bit */

Why? Are there any implicit references I'm missing?

Since the PoP specifies (p. 5-122):

Quoted 8 lines

"The record of references provided by the reference bit is not necessarily accurate. However, in the major- ity of situations, reference recording approximately coincides with the related storage reference."

I don't really see a way to test this properly.

Maybe I missed something?

No I think you're right, although in practice the reference bits should match. Or did you observe a mismatch?

author
#6 lore raw
Quoted 3 lines · Janis Schoetterl-Glausch

No I think you're right, although in practice the reference bits should
match. Or did you observe a mismatch?

Yes, I can sometimes observe mismatches.

#7 lore raw
In reply to Janis Schoetterl-Glausch
Quoted 19 lines · On Tue, 17 May 2022 13:44:51 +0200 Janis Schoetterl-Glausch…
Quoted 2 lines · Nico Boehr
Quoted 1 line · Janis Schoetterl-Glausch

On 5/16/22 11:07, Nico Boehr wrote:

[...]

Quoted 13 lines
Quoted 5 lines
Quoted 3 lines

+               expected_key.val = i * 2;
+
+               /* ignore reference bit */

Why? Are there any implicit references I'm missing?

Since the PoP specifies (p. 5-122):

"The record of references provided by the reference bit is not necessarily accurate. However, in the major- ity of situations, reference recording approximately coincides with the related storage reference."

Quoted 6 lines
Quoted 3 lines

I don't really see a way to test this properly.

Maybe I missed something?

No I think you're right, although in practice the reference bits should match. Or did you observe a mismatch?

the point is that the architecture allows for mismatches (in particular I think it is allowed to overindicate changes)

ignoring that bit is the correct thing to do