Re: [PATCH 0/2] Acceptance test: update kernel used on m68k/q800 test

[Laurent Vivier]

Le 07/11/2019 à 19:00, Philippe Mathieu-Daudé a écrit :
> On 11/7/19 6:18 PM, Laurent Vivier wrote:
>> Le 07/11/2019 à 17:38, Cleber Rosa a écrit :
>>> ----- Original Message -----
>>>> From: "Eric Blake" <address@hidden>
>>>> To: "Cleber Rosa" <address@hidden>, address@hidden
>>>> Cc: "Peter Maydell" <address@hidden>, "Eduardo Habkost"
>>>> <address@hidden>, "Philippe Mathieu-Daudé"
>>>> <address@hidden>, "Wainer dos Santos Moschetta"
>>>> <address@hidden>, "Laurent Vivier" <address@hidden>,
>>>> "Willian Rampazzo" <address@hidden>, "Philippe Mathieu-Daudé"
>>>> <address@hidden>
>>>> Sent: Thursday, November 7, 2019 10:43:08 AM
>>>> Subject: Re: [PATCH 0/2] Acceptance test: update kernel used on
>>>> m68k/q800 test
>>>>
>>>> On 10/29/19 6:23 PM, Cleber Rosa wrote:
>>>>> The boot_linux_console.py:BootLinuxConsole.test_m68k_q800 was very
>>>>> recently merged, but between its last review and now, the Kernel
>>>>> package used went missing.
>>>>>
>>>>
>>>> meta-question: Why was this series posted in-reply-to the pull request,
>>>> rather than as a new top-level thread? I nearly missed it because I
>>>> don't expect to see unreviewed patches buried in threading like that.
>>>> My workflow would have been to post the series in isolation, then
>>>> manually reply to the pull request to mention the message-id of the
>>>> related series proposed as a followup.
>>>>
>>>
>>> Hi Eric,
>>>
>>> That was my attempt to signal that it was a fix to something which
>>> had *just*
>>> being merged as part of that pull request (though now caused by it).
>>>
>>> I basically did not know how to act properly, so I thank you for the
>>> workflow
>>> suggestion.  I'll certainly follow it next time.
>>
>> IMHO, you should send your series and then replies to the pull request
>> to tell you have sent your series that fixes the patch in the pull
>> request, or vice-versa.
>>
>> But your series has been queued by Alex, so there is no problem...
> 
> I prepared a different fix around the same time, but closed my laptop
> before the patch was sent and noticed the next day:
> https://lists.gnu.org/archive/html/qemu-devel/2019-10/msg08120.html
> 
> Laurent, are you OK with the new kernel being tested?
> 

I'm fine. We could have problems with 5.4 because the address mapping
has been changed for SWIM (aee6bff1c325 "m68k: mac: Revisit floppy disc
controller base addresses), but this has been fixed by my patch that has
been merged today in QEMU (653901ca2b  "q800: fix I/O memory map").

Thanks,
Laurent