Re: [Qemu-devel] [PATCH] memory-internal.h: Remove obsolete claim that header is obsolete

[Philippe Mathieu-Daudé]

On 11/21/2017 01:00 PM, Peter Maydell wrote:
> On 21 November 2017 at 15:57, Philippe Mathieu-Daudé <address@hidden> wrote:
>> Hi Peter,
>>
>> On 11/21/2017 12:08 PM, Peter Maydell wrote:
>>> The memory-internal.h header claims that it is for "obsolete
>>> exec.c functions" which "will be removed soon". This statement
>>> was added in 2011, six years ago, but the header is still here.
>>> (Admittedly none of the prototypes added in commit 67d95c153bef55f6
>>> are still in the header.)
>>>
>>> It's convenient to have a place to put prototypes for functions
>>> which are used internally to the various .c files of the memory
>>> system or by the accel/tcg code, which is inevitably fairly
>>> closely coupled. So keep the header but update the comments to
>>> reflect what we're actually using it for.
>>
>> Until your NotDirtyInfo addition, the only prototype used was
>> memory_region_access_valid() (in s390-pci-inst.c).
>>
>> Since "none of the prototypes added in commit 67d95c153bef55f6 are still
>> in the header" we could restrict it out of include/exec/ (kinda 'revert'
>> 022c62cbbc) and only keep memory_region_access_valid() + NotDirtyInfo
>> exposed in include/exec/.
> 
> I'm not sure what you're suggesting. I definitely think the
> s390 usage is pretty nasty but I guess it would need some
> rework to get rid of. For everything else, it's nice
> to have somewhere to share these things. You could argue
> for splitting the header into two, one for 'between memory.c
> and exec.c' and one for 'between memory.c and cputlb.c',
> but is it worth the effort?

Why not move memory_region_access_valid() + NotDirtyInfo in
"exec/exec-all.h" and only use this to "share these things"?