Re: [Qemu-devel] [PATCH v4 0/4] Move target- and device specific code from monitor

["Denis V. Lunev" <den@openvz.org>]

On 09/10/2015 07:18 PM, Peter Crosthwaite wrote:
> On Thu, Sep 10, 2015 at 8:38 AM, Denis V. Lunev <den@openvz.org> wrote:
>> The monivation of this set is simple. Recently we have proposed patch
>> to monitor.c with specific x86 APIC HMP commands. The patchset was denied
>> with the main motivation "No more arch specific code in monitor.c"
>> This patchset is the first step to move arch specific code from
>> monitor.c targets.
>>
>> So, monitor.c already contains a lot of generic code, as well as the target
>> specifics code and eventually monitor.c volume will only grow. This trend leads
>> to a variety of fouling code ifdeffery(and combinations thereof),
>> poor readability, and entanglement of architecture of the project.
>> If someone wants to improve processing logic commands at the monitor,
>> it isn't necessarily must differentiate amongst the implementation of some ARM
>> or x86_64 specific commands, because the project already has separation of
>> target specific code on directories.
>>
>> The presented solution is not the best, but it is quite simple
>> (PATCH doesn't add more code!) and decides the above mentioned issue.
>> Subsequently it will not prevent the introduction of more advanced mechanism
>> that can more effectively resolve the issue.
>>
>> There is a issue with the placement of code for multiple architectures
>> (isn't for everyone), but this code is very small. This patch is a step towards
>> solving the issue associated with maintaining the purity of the code and
>> structure of the project, which solves not all, but doing a little better
>> than it is.
>>
>> Changes from v3:
>> - removed weak symbol, because it does not work on all platforms
>> - qemu-monitor-info.texi moved to patch 4
>>
>> Changes from v2:
>> - target_monitor_defs provides a target-specific array monitor_defs[]
>>    This call is added through stub to all archs as a weak symbol. It should
>>    be redefined on each specific architecture
>> - fix end of table info (information about the system state)
>> - ported to current HEAD (iothreads & skeys commands descriptions moved)
>>
>> Changes from v1:
>> - ported to new head
>>
>> Signed-off-by: Pavel Butsykin <pbutsykin@virtuozzo.com>
>> Signed-off-by: Denis V. Lunev <den@openvz.org>
>> CC: Paolo Bonzini <pbonzini@redhat.com>
>> CC: Peter Maydell <peter.maydell@linaro.org>
>>
> I need this!
>
> This is a piece of the puzzle for multi arch. I'll add it to my review
> queue, but can I take a CC on this in future incase it spins again
> before I get a chance?
>
> Regards,
> Peter

Sure! though I hope that this is a last re-spin...

Den