QEMU iotest 267 failure / assertion in migration code

Posted by Thomas Huth 1 year, 3 months ago


  Hi!

I just ran "make check" in a build directory where I did
not compile qemu-system-x86_64, and got a failure with
iotest 267.

Re-running the "check" script directly got me some more
information:

$ ./check -qcow2 267
QEMU          -- "/tmp/qemu/tests/qemu-iotests/../../qemu-system-alpha" -nodefaults -display none -accel qtest
QEMU_IMG      -- "/tmp/qemu/tests/qemu-iotests/../../qemu-img"
QEMU_IO       -- "/tmp/qemu/tests/qemu-iotests/../../qemu-io" --cache writeback --aio threads -f qcow2
QEMU_NBD      -- "/tmp/qemu/tests/qemu-iotests/../../qemu-nbd"
IMGFMT        -- qcow2
IMGPROTO      -- file
PLATFORM      -- Linux/x86_64 thuth 4.18.0-425.3.1.el8.x86_64
TEST_DIR      -- /tmp/qemu/tests/qemu-iotests/scratch
SOCK_DIR      -- /tmp/tmp_qbcjhsu
GDB_OPTIONS   --
VALGRIND_QEMU --
PRINT_QEMU_OUTPUT --

267   fail       [18:39:41] [18:39:44]   3.5s                 output mismatch (see /tmp/qemu/tests/qemu-iotests/scratch/267.out.bad)
--- .../qemu/tests/qemu-iotests/267.out
+++ /tmp/qemu/tests/qemu-iotests/scratch/267.out.bad
@@ -31,23 +31,23 @@
  Testing: -drive driver=IMGFMT,file=TEST_DIR/t.IMGFMT,if=none
  QEMU X.Y.Z monitor - type 'help' for more information
  (qemu) savevm snap0
-(qemu) info snapshots
-List of snapshots present on all disks:
-ID        TAG               VM SIZE                DATE     VM CLOCK     ICOUNT
---        snap0                SIZE yyyy-mm-dd hh:mm:ss 00:00:00.000
-(qemu) loadvm snap0
-(qemu) quit
+QEMU_PROG: ../../qemu/migration/ram.c:874: pss_find_next_dirty: Assertion `pss->host_page_end' failed.
+./common.rc: line 181: 1903770 Aborted                 (core dumped) ( if [ -n "${QEMU_NEED_PID}" ]; then
+    echo $BASHPID > "${QEMU_TEST_DIR}/qemu-${_QEMU_HANDLE}.pid";
+fi; GDB=""; if [ -n "${GDB_OPTIONS}" ]; then
+    GDB="gdbserver ${GDB_OPTIONS}";
+fi; VALGRIND_QEMU="${VALGRIND_QEMU_VM}" _qemu_proc_exec "${VALGRIND_LOGFILE}" $GDB "$QEMU_PROG" $QEMU_OPTIONS "$@" )


Looks like this test does not work if the main machine
of the selected QEMU binary does not support migration?

Should we remove this test from the "auto" group?

Anyway, QEMU should also not trigger an assertion, so this
sounds like another bug?

  Thomas

Re: QEMU iotest 267 failure / assertion in migration code

Posted by Dr. David Alan Gilbert 1 year, 3 months ago

* Thomas Huth (thuth@redhat.com) wrote:
> 
>  Hi!
> 
> I just ran "make check" in a build directory where I did
> not compile qemu-system-x86_64, and got a failure with
> iotest 267.
> 
> Re-running the "check" script directly got me some more
> information:
> 
> $ ./check -qcow2 267
> QEMU          -- "/tmp/qemu/tests/qemu-iotests/../../qemu-system-alpha" -nodefaults -display none -accel qtest
> QEMU_IMG      -- "/tmp/qemu/tests/qemu-iotests/../../qemu-img"
> QEMU_IO       -- "/tmp/qemu/tests/qemu-iotests/../../qemu-io" --cache writeback --aio threads -f qcow2
> QEMU_NBD      -- "/tmp/qemu/tests/qemu-iotests/../../qemu-nbd"
> IMGFMT        -- qcow2
> IMGPROTO      -- file
> PLATFORM      -- Linux/x86_64 thuth 4.18.0-425.3.1.el8.x86_64
> TEST_DIR      -- /tmp/qemu/tests/qemu-iotests/scratch
> SOCK_DIR      -- /tmp/tmp_qbcjhsu
> GDB_OPTIONS   --
> VALGRIND_QEMU --
> PRINT_QEMU_OUTPUT --
> 
> 267   fail       [18:39:41] [18:39:44]   3.5s                 output mismatch (see /tmp/qemu/tests/qemu-iotests/scratch/267.out.bad)
> --- .../qemu/tests/qemu-iotests/267.out
> +++ /tmp/qemu/tests/qemu-iotests/scratch/267.out.bad
> @@ -31,23 +31,23 @@
>  Testing: -drive driver=IMGFMT,file=TEST_DIR/t.IMGFMT,if=none
>  QEMU X.Y.Z monitor - type 'help' for more information
>  (qemu) savevm snap0
> -(qemu) info snapshots
> -List of snapshots present on all disks:
> -ID        TAG               VM SIZE                DATE     VM CLOCK     ICOUNT
> ---        snap0                SIZE yyyy-mm-dd hh:mm:ss 00:00:00.000
> -(qemu) loadvm snap0
> -(qemu) quit
> +QEMU_PROG: ../../qemu/migration/ram.c:874: pss_find_next_dirty: Assertion `pss->host_page_end' failed.

I don't understand how that can trigger - it needs investigating as a
bug.

> +./common.rc: line 181: 1903770 Aborted                 (core dumped) ( if [ -n "${QEMU_NEED_PID}" ]; then
> +    echo $BASHPID > "${QEMU_TEST_DIR}/qemu-${_QEMU_HANDLE}.pid";
> +fi; GDB=""; if [ -n "${GDB_OPTIONS}" ]; then
> +    GDB="gdbserver ${GDB_OPTIONS}";
> +fi; VALGRIND_QEMU="${VALGRIND_QEMU_VM}" _qemu_proc_exec "${VALGRIND_LOGFILE}" $GDB "$QEMU_PROG" $QEMU_OPTIONS "$@" )
> 
> 
> Looks like this test does not work if the main machine
> of the selected QEMU binary does not support migration?

Why doesn't it support migration?

> Should we remove this test from the "auto" group?
> 
> Anyway, QEMU should also not trigger an assertion, so this
> sounds like another bug?

Yeh; that's a weird failure.

(Alpha page size seems to be 8k from what I can tell; which should be
fine, if you're running on an x86 host)

Dave

>  Thomas
> 
-- 
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

Re: QEMU iotest 267 failure / assertion in migration code

Posted by Thomas Huth 1 year, 3 months ago

On 18/01/2023 19.02, Dr. David Alan Gilbert wrote:
> * Thomas Huth (thuth@redhat.com) wrote:
>>
>>   Hi!
>>
>> I just ran "make check" in a build directory where I did
>> not compile qemu-system-x86_64, and got a failure with
>> iotest 267.
>>
>> Re-running the "check" script directly got me some more
>> information:
>>
>> $ ./check -qcow2 267
>> QEMU          -- "/tmp/qemu/tests/qemu-iotests/../../qemu-system-alpha" -nodefaults -display none -accel qtest
>> QEMU_IMG      -- "/tmp/qemu/tests/qemu-iotests/../../qemu-img"
>> QEMU_IO       -- "/tmp/qemu/tests/qemu-iotests/../../qemu-io" --cache writeback --aio threads -f qcow2
>> QEMU_NBD      -- "/tmp/qemu/tests/qemu-iotests/../../qemu-nbd"
>> IMGFMT        -- qcow2
>> IMGPROTO      -- file
>> PLATFORM      -- Linux/x86_64 thuth 4.18.0-425.3.1.el8.x86_64
>> TEST_DIR      -- /tmp/qemu/tests/qemu-iotests/scratch
>> SOCK_DIR      -- /tmp/tmp_qbcjhsu
>> GDB_OPTIONS   --
>> VALGRIND_QEMU --
>> PRINT_QEMU_OUTPUT --
>>
>> 267   fail       [18:39:41] [18:39:44]   3.5s                 output mismatch (see /tmp/qemu/tests/qemu-iotests/scratch/267.out.bad)
>> --- .../qemu/tests/qemu-iotests/267.out
>> +++ /tmp/qemu/tests/qemu-iotests/scratch/267.out.bad
>> @@ -31,23 +31,23 @@
>>   Testing: -drive driver=IMGFMT,file=TEST_DIR/t.IMGFMT,if=none
>>   QEMU X.Y.Z monitor - type 'help' for more information
>>   (qemu) savevm snap0
>> -(qemu) info snapshots
>> -List of snapshots present on all disks:
>> -ID        TAG               VM SIZE                DATE     VM CLOCK     ICOUNT
>> ---        snap0                SIZE yyyy-mm-dd hh:mm:ss 00:00:00.000
>> -(qemu) loadvm snap0
>> -(qemu) quit
>> +QEMU_PROG: ../../qemu/migration/ram.c:874: pss_find_next_dirty: Assertion `pss->host_page_end' failed.
> 
> I don't understand how that can trigger - it needs investigating as a
> bug.

I've created a ticket so we don't forget:

https://gitlab.com/qemu-project/qemu/-/issues/1456

>> +./common.rc: line 181: 1903770 Aborted                 (core dumped) ( if [ -n "${QEMU_NEED_PID}" ]; then
>> +    echo $BASHPID > "${QEMU_TEST_DIR}/qemu-${_QEMU_HANDLE}.pid";
>> +fi; GDB=""; if [ -n "${GDB_OPTIONS}" ]; then
>> +    GDB="gdbserver ${GDB_OPTIONS}";
>> +fi; VALGRIND_QEMU="${VALGRIND_QEMU_VM}" _qemu_proc_exec "${VALGRIND_LOGFILE}" $GDB "$QEMU_PROG" $QEMU_OPTIONS "$@" )
>>
>>
>> Looks like this test does not work if the main machine
>> of the selected QEMU binary does not support migration?
> 
> Why doesn't it support migration?

Hmm, maybe I just got that wrong and it's really just the problem that the 
qemu-system-alpha is crashing.

  Thomas