Re: Problem with Avocado and QEMU console

[Philippe Mathieu-Daudé]

On 6/1/21 5:40 AM, Philippe Mathieu-Daudé wrote:
> On 5/3/21 4:46 PM, Willian Rampazzo wrote:
>> On Mon, May 3, 2021 at 10:36 AM Philippe Mathieu-Daudé <f4bug@amsat.org> 
>> wrote:
>>>
>>> On 5/3/21 3:12 PM, Willian Rampazzo wrote:
>>>> Hi Philippe,
>>>>
>>>> On Mon, May 3, 2021 at 9:59 AM Philippe Mathieu-Daudé <f4bug@amsat.org> 
>>>> wrote:
>>>>>
>>>>> Hi Cleber, Willian,
>>>>>
>>>>> Avocado marked the test_mips64el_fuloong2e as failed while
>>>>> it succeeded:
>>>>> https://gitlab.com/qemu-project/qemu/-/jobs/1231363571
>>>>>
>>>>> Apparently the first part of the console output is missing.
>>>>> (Normal console output attached).
>>>>>
>>>>> Any idea what could be the cause?
>>>>>
>>>>
>>>> The log shows the following:
>>>>
>>>> 12:00:37 ERROR| avocado.core.exceptions.TestFail: Failure message
>>>> found in console: "Kernel panic - not syncing". Expected: "Kernel
>>>> command line: printk.time=0 console=ttyS0"
>>>>
>>>> Does your comment about "console output missing" means the expected
>>>> message was supposed to be on this missing part?
>>>
>>> Yes, "the first part is missing" so Avocado didn't see the expected
>>> message pass, but the messages emitted *after* it are logged.
>>>
>>
>> Okay, got it.
>>
>> The reason for it to fail is here:
>> https://gitlab.com/qemu-project/qemu/-/blob/master/tests/acceptance/boot_linux_console.py#L53.
>> The way the `wait_for_console_pattern` method is defined tells the
>> test to fail if that `failure_message` is found.
>>
>> I'm still checking why the test missed the first part.
> 
> Could this be related to the recent failures in N8x0Machine.test_n810?
> 
> https://gitlab.com/qemu-project/qemu/-/jobs/1280746618#L128

Yet another one:

https://gitlab.com/qemu-project/qemu/-/jobs/1390365314#L129