qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [Qemu-ppc] [PATCH 6/8] target-ppc: Rework get_physical_


From: Alexander Graf
Subject: Re: [Qemu-devel] [Qemu-ppc] [PATCH 6/8] target-ppc: Rework get_physical_address()
Date: Sat, 23 Feb 2013 13:21:19 +0100

On 23.02.2013, at 12:56, David Gibson wrote:

> On Sat, Feb 23, 2013 at 11:38:23AM +0100, Alexander Graf wrote:
>> 
>> 
>> Am 23.02.2013 um 07:28 schrieb David Gibson <address@hidden>:
>> 
>>> On Fri, Feb 22, 2013 at 05:06:52PM +0100, Alexander Graf wrote:
>>>> 
>>>> On 12.02.2013, at 03:00, David Gibson wrote:
>>>> 
>>>>> Currently get_physical_address() first checks to see if translation is
>>>>> enabled in the MSR, then in the translation on case switches on the mmu
>>>>> type.  Except that for BookE MMUs, translation is always on, and so it
>>>>> has to switch in the "translation off" case as well and do the same thing
>>>>> as the translation on path for those MMUs.  Plus, even translation off
>>>>> doesn't behave exactly the same on the various MMU types so there are
>>>>> further mmu type checks in the "translation off" path.
>>>>> 
>>>>> As a first step to cleaning this up, this patch moves the switch on mmu
>>>>> type to the top level, then makes the translation on/off check just for
>>>>> those mmu types where it is meaningful.
>>>> 
>>>> Eventually the mmu translation should just be a class dispatched
>>>> indirect function call. I don't think it's within the scope of this
>>>> first round of cleanup though.
>>> 
>>> I agree.  My plan was to first push the switching all up to the top
>>> level as I've done, then replace that with a method dispatch
>>> afterwards.
>> 
>> If you're convinced the path you're taking is already correct, don't
>> post patches as RFC :).
> 
> Well, I'm not sure about all of it.
> 
> More to the point, I'm really not sure what you're suggesting I do
> differently.  We can't switch to a method invocation when there are
> mmu type dependencies scattered all through the path.  So that change
> more or less has to go after what I have already.

I never disagreed. I think your patches are fine, nothing really needs to be 
done differently. I just don't want the conversion to be done half-way.

I can understand if you don't feel comfortable not converting non-book3s mmus, 
but I want to have it done at least completely for all book3s ones then. I'm 
not a big fan of aborting a race right before the goal :)


Alex




reply via email to

[Prev in Thread] Current Thread [Next in Thread]