Re: [RFC PATCH v2 1/5] mm: make HPAGE_PxD_{SHIFT,MASK,SIZE} always available

[Mike Rapoport <rppt@kernel.org>]

Hi Hugh,

On Mon, Jul 06, 2020 at 10:07:34PM -0700, Hugh Dickins wrote:
> On Mon, 6 Jul 2020, Mike Rapoport wrote:
> > From: Mike Rapoport <rppt@linux.ibm.com>
> > 
> > The definitions of shift, mask and size for the second and the third level
> > of the leaf pages are available only when CONFIG_TRANSPARENT_HUGEPAGE is
> > set. Otherwise they evaluate to BUILD_BUG().
> > 
> > There is no explanation neither in the code nor in the changelog why the
> > usage of, e.g. HPAGE_PMD_SIZE should be only allowed with THP and forbidden
> > otherwise while the definitions of HPAGE_PMD_SIZE and HPAGE_PUD_SIZE
> > express the sizes better than ambiguous HPAGE_SIZE.
> > 
> > Make HPAGE_PxD_{SHIFT,MASK,SIZE} definitions available unconditionally.
> 
> Adding Andrea to Cc, he's the one who structured it that way,
> and should be consulted.
> 
> I'm ambivalent myself.  Many's the time I've been irritated by the
> BUILD_BUG() in HPAGE_etc, and it's responsible for very many #ifdef
> CONFIG_TRANSPARENT_HUGEPAGEs or IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE)s
> that you find uglily scattered around the source.
> 
> But that's the point of it: it's warning when you write code peculiar
> to THP, that is going to bloat the build of kernels without any THP.
> 
> So although I've often been tempted to do as you suggest, I've always
> ended up respecting Andrea's intention, and worked around it instead
> (sometimes with #ifdef or IS_ENABLED(), sometimes with
> PMD_{SHIFT,MASK_SIZE}, sometimes with a local definition).

I could do with a local definition as well, but I think HPAGE_PxD_SHIFT
is better and more descriptive than ambiguous HPAGE_SHIFT and I was
thinking about wider change to use "THP" defines rather than "hugetlb"
defines wherever possible. 

In the end, HPAGE_PMD_SIZE does not have to be associated with THP and
limited to it, it just says what is the size of a leaf page at PMD
level.

> Hugh
> 
> > 
> > Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
> > ---
> >  include/linux/huge_mm.h | 10 ++--------
> >  1 file changed, 2 insertions(+), 8 deletions(-)
> > 
> > diff --git a/include/linux/huge_mm.h b/include/linux/huge_mm.h
> > index 71f20776b06c..1f4b44a76e31 100644
> > --- a/include/linux/huge_mm.h
> > +++ b/include/linux/huge_mm.h
> > @@ -115,7 +115,6 @@ extern struct kobj_attribute shmem_enabled_attr;
> >  #define HPAGE_PMD_ORDER (HPAGE_PMD_SHIFT-PAGE_SHIFT)
> >  #define HPAGE_PMD_NR (1<<HPAGE_PMD_ORDER)
> >  
> > -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> >  #define HPAGE_PMD_SHIFT PMD_SHIFT
> >  #define HPAGE_PMD_SIZE	((1UL) << HPAGE_PMD_SHIFT)
> >  #define HPAGE_PMD_MASK	(~(HPAGE_PMD_SIZE - 1))
> > @@ -124,6 +123,8 @@ extern struct kobj_attribute shmem_enabled_attr;
> >  #define HPAGE_PUD_SIZE	((1UL) << HPAGE_PUD_SHIFT)
> >  #define HPAGE_PUD_MASK	(~(HPAGE_PUD_SIZE - 1))
> >  
> > +#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > +
> >  extern unsigned long transparent_hugepage_flags;
> >  
> >  /*
> > @@ -316,13 +317,6 @@ static inline struct list_head *page_deferred_list(struct page *page)
> >  }
> >  
> >  #else /* CONFIG_TRANSPARENT_HUGEPAGE */
> > -#define HPAGE_PMD_SHIFT ({ BUILD_BUG(); 0; })
> > -#define HPAGE_PMD_MASK ({ BUILD_BUG(); 0; })
> > -#define HPAGE_PMD_SIZE ({ BUILD_BUG(); 0; })
> > -
> > -#define HPAGE_PUD_SHIFT ({ BUILD_BUG(); 0; })
> > -#define HPAGE_PUD_MASK ({ BUILD_BUG(); 0; })
> > -#define HPAGE_PUD_SIZE ({ BUILD_BUG(); 0; })
> >  
> >  static inline int hpage_nr_pages(struct page *page)
> >  {
> > -- 
> > 2.26.2

-- 
Sincerely yours,
Mike.