Re: [Qemu-devel] [PATCH] v3: don't call reset functions on cpu initialization

[Blue Swirl]

On Fri, Nov 6, 2009 at 9:50 PM, Laurent Desnogues
<address@hidden> wrote:
> On Fri, Nov 6, 2009 at 8:36 PM, Blue Swirl <address@hidden> wrote:
>> On Fri, Nov 6, 2009 at 9:13 PM, Laurent Desnogues
>> <address@hidden> wrote:
>>> On Fri, Nov 6, 2009 at 8:01 PM, Blue Swirl <address@hidden> wrote:
>>>> On Fri, Nov 6, 2009 at 8:43 PM, Laurent Desnogues
>>>> <address@hidden> wrote:
>>> [...]
>>>>> I honestly don't care that much as long as all targets still work
>>>>> in user mode :-)
>>>>>
>>>>> The aim was to make Glauber's patch less intrusive.
>>>>
>>>> Given that only the new calls to cpu_reset are important and the
>>>> removals are much less so (double reset shouldn't be a problem), the
>>>> least intrusive version would be to just add the new calls and do the
>>>> clean up later.
>>>
>>> Multiple resets would result in memory leaks for MIPS for instance.
>>> So I don't think they are that innocuous.  Though that probably is
>>> more a problem of the MIPS reset than anything else :-)
>>
>> You are hinting at target-mips/translate_init.c: mmu_init()? Right, it
>> shouldn't do the allocation.
>
> No, I was talking about mvp_init.

Both are then buggy.

Why are these even allocated instead being members of CPUState?