Re: [PATCH v3 2/2] mm: remove odd HAVE_PTE_SPECIAL

[Christophe LEROY <christophe.leroy@c-s.fr>]

Le 11/04/2018 A  11:03, Laurent Dufour a A(C)critA :
> 
> 
> On 11/04/2018 10:58, Christophe LEROY wrote:
>>
>>
>> Le 11/04/2018 A  10:03, Laurent Dufour a A(C)critA :
>>> Remove the additional define HAVE_PTE_SPECIAL and rely directly on
>>> CONFIG_ARCH_HAS_PTE_SPECIAL.
>>>
>>> There is no functional change introduced by this patch
>>>
>>> Signed-off-by: Laurent Dufour <ldufour@linux.vnet.ibm.com>
>>> ---
>>>  A  mm/memory.c | 19 ++++++++-----------
>>>  A  1 file changed, 8 insertions(+), 11 deletions(-)
>>>
>>> diff --git a/mm/memory.c b/mm/memory.c
>>> index 96910c625daa..7f7dc7b2a341 100644
>>> --- a/mm/memory.c
>>> +++ b/mm/memory.c
>>> @@ -817,17 +817,12 @@ static void print_bad_pte(struct vm_area_struct *vma,
>>> unsigned long addr,
>>>  A A  * PFNMAP mappings in order to support COWable mappings.
>>>  A A  *
>>>  A A  */
>>> -#ifdef CONFIG_ARCH_HAS_PTE_SPECIAL
>>> -# define HAVE_PTE_SPECIAL 1
>>> -#else
>>> -# define HAVE_PTE_SPECIAL 0
>>> -#endif
>>>  A  struct page *_vm_normal_page(struct vm_area_struct *vma, unsigned long addr,
>>>  A A A A A A A A A A A A A A A A A A  pte_t pte, bool with_public_device)
>>>  A  {
>>>  A A A A A  unsigned long pfn = pte_pfn(pte);
>>>  A  -A A A  if (HAVE_PTE_SPECIAL) {
>>> +A A A  if (IS_ENABLED(CONFIG_ARCH_HAS_PTE_SPECIAL)) {
>>>  A A A A A A A A A  if (likely(!pte_special(pte)))
>>>  A A A A A A A A A A A A A  goto check_pfn;
>>>  A A A A A A A A A  if (vma->vm_ops && vma->vm_ops->find_special_page)
>>> @@ -862,7 +857,7 @@ struct page *_vm_normal_page(struct vm_area_struct *vma,
>>> unsigned long addr,
>>>  A A A A A A A A A  return NULL;
>>>  A A A A A  }
>>>  A  -A A A  /* !HAVE_PTE_SPECIAL case follows: */
>>> +A A A  /* !CONFIG_ARCH_HAS_PTE_SPECIAL case follows: */
>>>  A  A A A A A  if (unlikely(vma->vm_flags & (VM_PFNMAP|VM_MIXEDMAP))) {
>>>  A A A A A A A A A  if (vma->vm_flags & VM_MIXEDMAP) {
>>> @@ -881,7 +876,8 @@ struct page *_vm_normal_page(struct vm_area_struct *vma,
>>> unsigned long addr,
>>>  A  A A A A A  if (is_zero_pfn(pfn))
>>>  A A A A A A A A A  return NULL;
>>> -check_pfn:
>>> +
>>> +check_pfn: __maybe_unused
>>
>> See below
>>
>>>  A A A A A  if (unlikely(pfn > highest_memmap_pfn)) {
>>>  A A A A A A A A A  print_bad_pte(vma, addr, pte, NULL);
>>>  A A A A A A A A A  return NULL;
>>> @@ -891,7 +887,7 @@ struct page *_vm_normal_page(struct vm_area_struct *vma,
>>> unsigned long addr,
>>>  A A A A A A  * NOTE! We still have PageReserved() pages in the page tables.
>>>  A A A A A A  * eg. VDSO mappings can cause them to exist.
>>>  A A A A A A  */
>>> -out:
>>> +out: __maybe_unused
>>
>> Why do you need that change ?
>>
>> There is no reason for the compiler to complain. It would complain if the goto
>> was within a #ifdef, but all the purpose of using IS_ENABLED() is to allow the
>> compiler to properly handle all possible cases. That's all the force of
>> IS_ENABLED() compared to ifdefs, and that the reason why they are plebicited,
>> ref Linux Codying style for a detailed explanation.
> 
> Fair enough.
> 
> Should I submit a v4 just to remove these so ugly __maybe_unused ?
> 

Most likely, unless the mm maintainer agrees to remove them by himself 
when applying your patch ?

Christophe