Skip to contentSkip to conversation

s390x/ioinst: Fix wrong MSCH alignment check on little endian

PATCH Applied David Hildenbrand 6 participants 10 messages Aug 5, 2021 – Aug 10, 2021

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 s390x/ioinst: Fix wrong MSCH alignment check on little endian

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

FAIL: Channel Subsystem: measurement block format1: Unaligned MB origin:
Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be aligned")
Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
---
 target/s390x/ioinst.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
target/s390x/ioinst.c +1 -1
index 4eb0a7a9f8..bdae5090bc 100644
@@ -123,7 +123,7 @@ static int ioinst_schib_valid(SCHIB *schib)
     }
     /* for MB format 1 bits 26-31 of word 11 must be 0 */
     /* MBA uses words 10 and 11, it means align on 2**6 */
-    if ((be16_to_cpu(schib->pmcw.chars) & PMCW_CHARS_MASK_MBFC) &&
+    if ((be32_to_cpu(schib->pmcw.chars) & PMCW_CHARS_MASK_MBFC) &&
         (be64_to_cpu(schib->mba) & 0x03fUL)) {
         return 0;
     }
Signature · 1 line
2.31.1
Reviewed-by on v1 1/1
#1 lore raw
Quoted 27 lines · David Hildenbrand

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

FAIL: Channel Subsystem: measurement block format1: Unaligned MB origin:

   Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be aligned")
Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
---
  target/s390x/ioinst.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
target/s390x/ioinst.c +0 -0
index 4eb0a7a9f8..bdae5090bc 100644

target/s390x/ioinst.c static int ioinst_schib_valid(SCHIB *schib) · line 129

Quoted 8 lines
target/s390x/ioinst.c +1 -1
@@ -123,7 +123,7 @@ static int ioinst_schib_valid(SCHIB *schib)
      }
      /* for MB format 1 bits 26-31 of word 11 must be 0 */
      /* MBA uses words 10 and 11, it means align on 2**6 */
-    if ((be16_to_cpu(schib->pmcw.chars) & PMCW_CHARS_MASK_MBFC) &&
+    if ((be32_to_cpu(schib->pmcw.chars) & PMCW_CHARS_MASK_MBFC) &&
          (be64_to_cpu(schib->mba) & 0x03fUL)) {
          return 0;
      }

It is obviously a 32bit.

Signature · 2 lines
Pierre Morel
IBM Lab Boeblingen
Reviewed-by on v1 1/1
#2 lore raw
In reply to David Hildenbrand
Quoted 27 lines · David Hildenbrand

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

FAIL: Channel Subsystem: measurement block format1: Unaligned MB origin:

   Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be aligned")
Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
---
  target/s390x/ioinst.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
target/s390x/ioinst.c +0 -0
index 4eb0a7a9f8..bdae5090bc 100644

target/s390x/ioinst.c static int ioinst_schib_valid(SCHIB *schib) · line 129

Quoted 8 lines
target/s390x/ioinst.c +1 -1
@@ -123,7 +123,7 @@ static int ioinst_schib_valid(SCHIB *schib)
      }
      /* for MB format 1 bits 26-31 of word 11 must be 0 */
      /* MBA uses words 10 and 11, it means align on 2**6 */
-    if ((be16_to_cpu(schib->pmcw.chars) & PMCW_CHARS_MASK_MBFC) &&
+    if ((be32_to_cpu(schib->pmcw.chars) & PMCW_CHARS_MASK_MBFC) &&
          (be64_to_cpu(schib->mba) & 0x03fUL)) {
          return 0;
      }
#3 lore raw
In reply to David Hildenbrand
Quoted 17 lines · David Hildenbrand

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

FAIL: Channel Subsystem: measurement block format1: Unaligned MB origin:
Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be aligned")
Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
Quoted 3 lines
---
 target/s390x/ioinst.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Same question here: still -rc worthy, or not?

#4 lore raw
Quoted 23 lines · Cornelia Huck
Quoted 17 lines · David Hildenbrand

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

FAIL: Channel Subsystem: measurement block format1: Unaligned MB origin:

   Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be aligned")
Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
Quoted 3 lines
---
  target/s390x/ioinst.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
Quoted 2 lines

Same question here: still -rc worthy, or not?

It's been broken forever, this can also wait I guess.

Signature · 3 lines
Thanks,

David / dhildenb
#5 lore raw
Quoted 26 lines · David Hildenbrand
Quoted 25 lines · Cornelia Huck
Quoted 22 lines · David Hildenbrand

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

   FAIL: Channel Subsystem: measurement block format1: Unaligned MB origin:
   Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be

aligned")

Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>

---
  target/s390x/ioinst.c | 2 +-

  1 file changed, 1 insertion(+), 1 deletion(-)
Quoted 4 lines
Quoted 2 lines

Same question here: still -rc worthy, or not?

It's been broken forever, this can also wait I guess.

But the commit referenced is recent:

commit a54b8ac340c20531daa89929c5ce7fed89fa401d
Date: Fri Feb 19 14:39:33 2021 +0100

css: SCHIB measurement block origin must be aligned

#6 lore raw
Quoted 30 lines · Philippe Mathieu-Daudé
Quoted 29 lines · David Hildenbrand
Quoted 27 lines · Cornelia Huck
Quoted 19 lines · David Hildenbrand

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

   FAIL: Channel Subsystem: measurement block format1: Unaligned MB
origin:
   Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be

aligned")

Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
Quoted 3 lines
---
   target/s390x/ioinst.c | 2 +-
   1 file changed, 1 insertion(+), 1 deletion(-)

Same question here: still -rc worthy, or not?

Quoted 9 lines
Quoted 1 line

It's been broken forever, this can also wait I guess.

But the commit referenced is recent:

commit a54b8ac340c20531daa89929c5ce7fed89fa401d
Date: Fri Feb 19 14:39:33 2021 +0100

css: SCHIB measurement block origin must be aligned

Well, okay yes :) I don't think it's very urgent though.

Signature · 3 lines
Thanks,

David / dhildenb
#7 lore raw
Quoted 35 lines · David Hildenbrand
Quoted 34 lines · Philippe Mathieu-Daudé
Quoted 26 lines · David Hildenbrand
Quoted 25 lines · Cornelia Huck
Quoted 22 lines · David Hildenbrand

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

   FAIL: Channel Subsystem: measurement block format1: Unaligned MB
origin:
   Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be

aligned")

Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
---
   target/s390x/ioinst.c | 2 +-
   1 file changed, 1 insertion(+), 1 deletion(-)
Quoted 4 lines
Quoted 2 lines

Same question here: still -rc worthy, or not?

It's been broken forever, this can also wait I guess.

But the commit referenced is recent:

Quoted 7 lines
Quoted 5 lines

commit a54b8ac340c20531daa89929c5ce7fed89fa401d
Date: Fri Feb 19 14:39:33 2021 +0100

css: SCHIB measurement block origin must be aligned

Well, okay yes :) I don't think it's very urgent though.

I think I agree. Before the referenced commit, we did not do any alignment checks at all, now we still let some unaligned blocks pass. I don't think I've seen any issues before, the problem has only been uncovered by the kvm unit test. So fix-worthy, but probably not late-rc-worthy.

Reviewed-by on v1 1/1
#8 lore raw
In reply to David Hildenbrand
Quoted 8 lines · On Thu, 5 Aug 2021 16:37:53 +0200 David Hildenbrand <[email protected]> wrote:

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

FAIL: Channel Subsystem: measurement block format1: Unaligned MB origin:
Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Quoted 9 lines
Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be aligned")
Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
Quoted 17 lines
---
 target/s390x/ioinst.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
target/s390x/ioinst.c +1 -1
index 4eb0a7a9f8..bdae5090bc 100644
@@ -123,7 +123,7 @@ static int ioinst_schib_valid(SCHIB *schib)
     }
     /* for MB format 1 bits 26-31 of word 11 must be 0 */
     /* MBA uses words 10 and 11, it means align on 2**6 */
-    if ((be16_to_cpu(schib->pmcw.chars) & PMCW_CHARS_MASK_MBFC) &&
+    if ((be32_to_cpu(schib->pmcw.chars) & PMCW_CHARS_MASK_MBFC) &&
         (be64_to_cpu(schib->mba) & 0x03fUL)) {
         return 0;
     }
#9 lore raw
In reply to David Hildenbrand
Quoted 17 lines · David Hildenbrand

schib->pmcw.chars is 32bit, not 16bit. This fixes the kvm-unit-tests "css" test, which fails with:

FAIL: Channel Subsystem: measurement block format1: Unaligned MB origin:
Program interrupt: expected(21) == received(0)

Because we end up not injecting an operand program exception.

Fixes a54b8ac340c2 ("css: SCHIB measurement block origin must be aligned")
Cc Halil Pasic <[email protected]>
Cc Cornelia Huck <[email protected]>
Cc Christian Borntraeger <[email protected]>
Cc Richard Henderson <[email protected]>
Cc Thomas Huth <[email protected]>
Cc Pierre Morel <[email protected]>
Quoted 3 lines
---
 target/s390x/ioinst.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Thanks, applied.