| Message ID | 20210312124700.142269-1-imbrenda@linux.ibm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [kvm-unit-tests,v1,1/1] s390x: mvpg: add checks for op_acc_id | expand |
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>
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>
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!
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();
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(-)