* [kvm-unit-tests PATCH v1 1/1] s390x: mvpg: add checks for op_acc_id
@ 2021-03-12 12:47 Claudio Imbrenda
2021-03-12 13:10 ` David Hildenbrand
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: Claudio Imbrenda @ 2021-03-12 12:47 UTC (permalink / raw)
To: kvm; +Cc: david, thuth, frankja, cohuck, pmorel
Check the operand access identification when MVPG causes a page fault.
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
---
s390x/mvpg.c | 28 ++++++++++++++++++++++++++--
1 file changed, 26 insertions(+), 2 deletions(-)
diff --git a/s390x/mvpg.c b/s390x/mvpg.c
index 5743d5b6..2b7c6cc9 100644
--- a/s390x/mvpg.c
+++ b/s390x/mvpg.c
@@ -36,6 +36,7 @@
static uint8_t source[PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
static uint8_t buffer[PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
+static struct lowcore * const lc;
/* Keep track of fresh memory */
static uint8_t *fresh;
@@ -77,6 +78,21 @@ static int page_ok(const uint8_t *p)
return 1;
}
+/*
+ * Check that the Operand Access Identification matches with the values of
+ * the r1 and r2 fields in the instruction format. The r1 and r2 fields are
+ * in the last byte of the instruction, and the Program Old PSW will point
+ * to the beginning of the instruction after the one that caused the fault
+ * (the fixup code in the interrupt handler takes care of that for
+ * nullifying instructions). Therefore it is enough to compare the byte
+ * before the one contained in the Program Old PSW with the value of the
+ * Operand Access Identification.
+ */
+static inline bool check_oai(void)
+{
+ return *(uint8_t *)(lc->pgm_old_psw.addr - 1) == lc->op_acc_id;
+}
+
static void test_exceptions(void)
{
int i, expected;
@@ -201,17 +217,25 @@ static void test_mmu_prot(void)
report(clear_pgm_int() == PGM_INT_CODE_PROTECTION, "destination read only");
fresh += PAGE_SIZE;
+ report_prefix_push("source invalid");
protect_page(source, PAGE_ENTRY_I);
+ lc->op_acc_id = 0;
expect_pgm_int();
mvpg(0, fresh, source);
- report(clear_pgm_int() == PGM_INT_CODE_PAGE_TRANSLATION, "source invalid");
+ report(clear_pgm_int() == PGM_INT_CODE_PAGE_TRANSLATION, "exception");
unprotect_page(source, PAGE_ENTRY_I);
+ report(check_oai(), "operand access ident");
+ report_prefix_pop();
fresh += PAGE_SIZE;
+ report_prefix_push("destination invalid");
protect_page(fresh, PAGE_ENTRY_I);
+ lc->op_acc_id = 0;
expect_pgm_int();
mvpg(0, fresh, source);
- report(clear_pgm_int() == PGM_INT_CODE_PAGE_TRANSLATION, "destination invalid");
+ report(clear_pgm_int() == PGM_INT_CODE_PAGE_TRANSLATION, "exception");
+ report(check_oai(), "operand access ident");
+ report_prefix_pop();
fresh += PAGE_SIZE;
report_prefix_pop();
--
2.26.2
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [kvm-unit-tests PATCH v1 1/1] s390x: mvpg: add checks for op_acc_id
2021-03-12 12:47 [kvm-unit-tests PATCH v1 1/1] s390x: mvpg: add checks for op_acc_id Claudio Imbrenda
@ 2021-03-12 13:10 ` David Hildenbrand
2021-03-12 15:56 ` Thomas Huth
2021-03-15 10:18 ` Janosch Frank
2 siblings, 0 replies; 4+ messages in thread
From: David Hildenbrand @ 2021-03-12 13:10 UTC (permalink / raw)
To: Claudio Imbrenda, kvm; +Cc: thuth, frankja, cohuck, pmorel
On 12.03.21 13:47, Claudio Imbrenda wrote:
> Check the operand access identification when MVPG causes a page fault.
>
> Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
> ---
> s390x/mvpg.c | 28 ++++++++++++++++++++++++++--
> 1 file changed, 26 insertions(+), 2 deletions(-)
>
> diff --git a/s390x/mvpg.c b/s390x/mvpg.c
> index 5743d5b6..2b7c6cc9 100644
> --- a/s390x/mvpg.c
> +++ b/s390x/mvpg.c
> @@ -36,6 +36,7 @@
>
> static uint8_t source[PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
> static uint8_t buffer[PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
> +static struct lowcore * const lc;
>
> /* Keep track of fresh memory */
> static uint8_t *fresh;
> @@ -77,6 +78,21 @@ static int page_ok(const uint8_t *p)
> return 1;
> }
>
> +/*
> + * Check that the Operand Access Identification matches with the values of
> + * the r1 and r2 fields in the instruction format. The r1 and r2 fields are
> + * in the last byte of the instruction, and the Program Old PSW will point
> + * to the beginning of the instruction after the one that caused the fault
> + * (the fixup code in the interrupt handler takes care of that for
> + * nullifying instructions). Therefore it is enough to compare the byte
> + * before the one contained in the Program Old PSW with the value of the
> + * Operand Access Identification.
> + */
> +static inline bool check_oai(void)
> +{
> + return *(uint8_t *)(lc->pgm_old_psw.addr - 1) == lc->op_acc_id;
> +}
> +
> static void test_exceptions(void)
> {
> int i, expected;
> @@ -201,17 +217,25 @@ static void test_mmu_prot(void)
> report(clear_pgm_int() == PGM_INT_CODE_PROTECTION, "destination read only");
> fresh += PAGE_SIZE;
>
> + report_prefix_push("source invalid");
> protect_page(source, PAGE_ENTRY_I);
> + lc->op_acc_id = 0;
> expect_pgm_int();
> mvpg(0, fresh, source);
> - report(clear_pgm_int() == PGM_INT_CODE_PAGE_TRANSLATION, "source invalid");
> + report(clear_pgm_int() == PGM_INT_CODE_PAGE_TRANSLATION, "exception");
> unprotect_page(source, PAGE_ENTRY_I);
> + report(check_oai(), "operand access ident");
> + report_prefix_pop();
> fresh += PAGE_SIZE;
>
> + report_prefix_push("destination invalid");
> protect_page(fresh, PAGE_ENTRY_I);
> + lc->op_acc_id = 0;
> expect_pgm_int();
> mvpg(0, fresh, source);
> - report(clear_pgm_int() == PGM_INT_CODE_PAGE_TRANSLATION, "destination invalid");
> + report(clear_pgm_int() == PGM_INT_CODE_PAGE_TRANSLATION, "exception");
> + report(check_oai(), "operand access ident");
> + report_prefix_pop();
> fresh += PAGE_SIZE;
>
> report_prefix_pop();
>
Thanks - works with my TCG implementation as well just fine.
Reviewed-by: David Hildenbrand <david@redhat.com>
--
Thanks,
David / dhildenb
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [kvm-unit-tests PATCH v1 1/1] s390x: mvpg: add checks for op_acc_id
2021-03-12 12:47 [kvm-unit-tests PATCH v1 1/1] s390x: mvpg: add checks for op_acc_id Claudio Imbrenda
2021-03-12 13:10 ` David Hildenbrand
@ 2021-03-12 15:56 ` Thomas Huth
2021-03-15 10:18 ` Janosch Frank
2 siblings, 0 replies; 4+ messages in thread
From: Thomas Huth @ 2021-03-12 15:56 UTC (permalink / raw)
To: Claudio Imbrenda, kvm; +Cc: david, frankja, cohuck, pmorel
On 12/03/2021 13.47, Claudio Imbrenda wrote:
> Check the operand access identification when MVPG causes a page fault.
>
> Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
> ---
> s390x/mvpg.c | 28 ++++++++++++++++++++++++++--
> 1 file changed, 26 insertions(+), 2 deletions(-)
Reviewed-by: Thomas Huth <thuth@redhat.com>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [kvm-unit-tests PATCH v1 1/1] s390x: mvpg: add checks for op_acc_id
2021-03-12 12:47 [kvm-unit-tests PATCH v1 1/1] s390x: mvpg: add checks for op_acc_id Claudio Imbrenda
2021-03-12 13:10 ` David Hildenbrand
2021-03-12 15:56 ` Thomas Huth
@ 2021-03-15 10:18 ` Janosch Frank
2 siblings, 0 replies; 4+ messages in thread
From: Janosch Frank @ 2021-03-15 10:18 UTC (permalink / raw)
To: Claudio Imbrenda, kvm; +Cc: david, thuth, cohuck, pmorel
On 3/12/21 1:47 PM, Claudio Imbrenda wrote:
> + * Check that the Operand Access Identification matches with the values of
Reviewed-by: Janosch Frank <frankja@linux.ibm.com>
Thanks, picked!
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2021-03-15 10:18 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-03-12 12:47 [kvm-unit-tests PATCH v1 1/1] s390x: mvpg: add checks for op_acc_id Claudio Imbrenda
2021-03-12 13:10 ` David Hildenbrand
2021-03-12 15:56 ` Thomas Huth
2021-03-15 10:18 ` Janosch Frank
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.