Re: Failing QEMU iotests

[Thomas Huth <thuth@redhat.com>]

On 17/11/2021 11.59, Hanna Reitz wrote:
> On 17.11.21 11:07, Thomas Huth wrote:
>>
>>  Hi!
>>
>> I think it has been working fine for me a couple of weeks ago,
>> but when I now run:
>>
>>  make check SPEED=slow
>>
>> I'm getting a couple of failing iotests... not sure whether
>> these are known issues already, so I thought I'd summarize them
>> here:
...
>> --- /home/thuth/devel/qemu/tests/qemu-iotests/206.out
>> +++ 206.out.bad
>> @@ -99,55 +99,19 @@
>>
>>  {"execute": "blockdev-create", "arguments": {"job-id": "job0", "options": 
>> {"driver": "qcow2", "encrypt": {"cipher-alg": "twofish-128", 
>> "cipher-mode": "ctr", "format": "luks", "hash-alg": "sha1", "iter-time": 
>> 10, "ivgen-alg": "plain64", "ivgen-hash-alg": "md5", "key-secret": 
>> "keysec0"}, "file": {"driver": "file", "filename": 
>> "TEST_DIR/PID-t.qcow2"}, "size": 33554432}}}
>>  {"return": {}}
>> +Job failed: Unsupported cipher algorithm twofish-128 with ctr mode
>>  {"execute": "job-dismiss", "arguments": {"id": "job0"}}
>>  {"return": {}}
>>
>>  image: TEST_IMG
>>  file format: IMGFMT
>>  virtual size: 32 MiB (33554432 bytes)
>> -encrypted: yes
>>  cluster_size: 65536
>>  Format specific information:
>>      compat: 1.1
>>      compression type: zlib
>>      lazy refcounts: false
>>      refcount bits: 16
>> -    encrypt:
>> -        ivgen alg: plain64
>> -        hash alg: sha1
>> -        cipher alg: twofish-128
>> -        uuid: XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX
>> -        format: luks
>> -        cipher mode: ctr
>> -        slots:
>> -            [0]:
>> -                active: true
>> -                iters: XXX
>> -                key offset: 4096
>> -                stripes: 4000
>> -            [1]:
>> -                active: false
>> -                key offset: 69632
>> -            [2]:
>> -                active: false
>> -                key offset: 135168
>> -            [3]:
>> -                active: false
>> -                key offset: 200704
>> -            [4]:
>> -                active: false
>> -                key offset: 266240
>> -            [5]:
>> -                active: false
>> -                key offset: 331776
>> -            [6]:
>> -                active: false
>> -                key offset: 397312
>> -            [7]:
>> -                active: false
>> -                key offset: 462848
>> -        payload offset: 528384
>> -        master key iters: XXX
>>      corrupt: false
>>      extended l2: false
> 
> I doubt this worked a couple of weeks ago, but it’s definitely one that we 
> should just get around to fixing. :/

Hm, maybe I've did the successful run on a different system last time ... I 
even slightly remember now having seen this before in the past on my current 
system, so yes, likely not something new.


>> +++ 297.out.bad
>> @@ -1,2 +1,21 @@
>>  === pylint ===
>> +************* Module image-fleecing
>> +tests/image-fleecing:34:24: C0326: Exactly one space required after comma
>> +patterns = [('0x5d', '0',         '64k'),
>> +                        ^ (bad-whitespace)
>> +tests/image-fleecing:35:25: C0326: Exactly one space required after comma
>> +            ('0xd5', '1M',        '64k'),
>> +                         ^ (bad-whitespace)
>> +tests/image-fleecing:36:26: C0326: Exactly one space required after comma
>> +            ('0xdc', '32M',       '64k'),
>> +                          ^ (bad-whitespace)
>> +tests/image-fleecing:39:25: C0326: Exactly one space required after comma
>> +overwrite = [('0xab', '0',         '64k'), # Full overwrite
>> +                         ^ (bad-whitespace)
>> +tests/image-fleecing:48:32: C0326: Exactly one space required after comma
>> +remainder = [('0xd5', '0x108000',  '32k'), # Right-end of partial-left [1]
>> +                                ^ (bad-whitespace)
>> +tests/image-fleecing:49:27: C0326: Exactly one space required after comma
>> +             ('0xdc', '32M',       '32k'), # Left-end of partial-right [2]
>> +                           ^ (bad-whitespace)
> 
> This could be because your pylint is too old.  At least for the python/ 
> tests we at least require 2.8.0 
> (https://lists.nongnu.org/archive/html/qemu-block/2021-10/msg00768.html) and 
> bad-whitespace was removed in 2.6.

Thanks, updating pylint fixed this problem, indeed!

But maybe the iotests should check the pylint version before using it?

  Thomas