Re: [Qemu-devel] [PATCH v3 21/21] iotests: Test new blkdebug/blkverify interface

[Max Reitz <mreitz@redhat.com>]

On 12.12.2013 11:41, Fam Zheng wrote:
> On 2013年12月12日 02:11, Max Reitz wrote:
>> Add a test for the new blkdebug/blkverify interface.
>>
>> Signed-off-by: Max Reitz <mreitz@redhat.com>
>> ---
>>   tests/qemu-iotests/071     | 201 
>> +++++++++++++++++++++++++++++++++++++++++++++
>>   tests/qemu-iotests/071.out |  73 ++++++++++++++++
>>   tests/qemu-iotests/group   |   1 +
>>   3 files changed, 275 insertions(+)
>>   create mode 100755 tests/qemu-iotests/071
>>   create mode 100644 tests/qemu-iotests/071.out
>>
>> diff --git a/tests/qemu-iotests/071 b/tests/qemu-iotests/071
>> new file mode 100755
>> index 0000000..4be525e
>> --- /dev/null
>> +++ b/tests/qemu-iotests/071
>> @@ -0,0 +1,201 @@
>> +#!/bin/bash
>
> Is is intended to use hand coded json instead of iotests.py for QMP 
> test here?

I first tried to use Python, but I hit problems with 
human-monitor-command and qemu-io. qemu-io prints errors to stderr, 
therefore they are not returned via QMP. iotests.py discards stderr, 
therefore there was no way to actually check for the qemu-io return status.

I could have added a function to iotests.py, but then I would have to 
manually compare and filter the qemu-io output with a reference. All 
these functions are already there for shell script tests.

I guess I should add this explanation to the commit message, though…

Max