[Qemu-devel] Re: commit e09a5267

[Jan Kiszka <jan.kiszka@siemens.com>]

Anthony Liguori wrote:
> Jan Kiszka wrote:
>> Avi Kivity wrote:
>>   
>>> On 09/10/2009 04:56 PM, Anthony Liguori wrote:
>>>     
>>>> The problem is patch volume.  We often see hundreds of patches a day. 
>>>> If typing a mail for each patch takes 2 minutes, that's potentially
>>>> hours spent just on sending these mails.
>>>>
>>>>       
>>> You exaggerate.  The average rate is 13 patches per calendar day.  The
>>> bulk of the patches are in patchsets which can be acked as a set, not
>>> once per patch.
>>>
>>>     
>>>> What I really need is some way to automatically generate these
>>>> notifications.  It's pretty easy to send a mail when a patch enters
>>>> the queue but it's more difficult to send a mail when a patch is
>>>> removed from the queue via a rebase.  Often times, I remove patches
>>>> from the queue simply because I'm not the right path for the patches
>>>> to be committed from (like linux-user).
>>>>       
>>> I think more per-patch attention is needed, not less, for example see
>>> this commit:
>>>
>>> commit e09a5267adf0af25b55d2abaf06e288b2d9537ea
>>> Author: Dustin Kirkland <kirkland@canonical.com>
>>> Date:   Thu Sep 3 12:31:33 2009 -0500
>>>
>>>     qemu-kvm: fix segfault when running kvm without /dev/kvm, falling
>>> back to non-accelerated mode
>>>
>>>     qemu-kvm: fix segfault when running kvm without /dev/kvm, falling back
>>>     to non-accelerated mode
>>>
>>>     We're seeing segfaults on systems without access to /dev/kvm.  It
>>>     looks like the global kvm_allowed is being set just a little too late
>>>     in vl.c.  This patch moves the kvm initialization a bit higher in the
>>>     vl.c main, just after options processing, and solves the segfaults.
>>>     We're carrying this patch in Ubuntu 9.10 Alpha.  Please apply
>>>     upstream, or advise if and why this might not be the optimal solution.
>>>
>>>     Signed-off-by: Dustin Kirkland <kirkland@canonical.com>
>>>
>>>     Move the kvm_init() call a bit higher to fix a segfault when
>>>     /dev/kvm is not available.  The kvm_allowed global needs
>>>     to be set correctly a little earlier.
>>>
>>>     Signed-off-by: Dustin Kirkland <kirkland@canonical.com>
>>>     Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
>>>
>>> There are many examples like this in the tree which is a pity.  Others
>>> include parts of an email conversation.  I'd like history to look better
>>> than this.
>>>     
>> Even worse, I think this patch does not belong into upstream as it fixed
>> a qemu-kvm-only bug. I think this was caused by Dustin CC'ing qemu, right?
>>
>> Did anyone test properly if the change has no side effects on upstream
>> kvm (which has a different initialization scheme)?
>>   
> 
> It doesn't break upstream qemu's -enable-kvm.

Yep, that's what I can confirm now too - after unbreaking -enable-kvm again.

Jan

-- 
Siemens AG, Corporate Technology, CT SE 2
Corporate Competence Center Embedded Linux