Re: [PATCH] mm: compaction: skip memory hole rapidly when isolating migratable pages

[Baolin Wang]

On 6/12/2023 2:39 PM, Huang, Ying wrote:
> Baolin Wang <baolin.wang@xxxxxxxxxxxxxxxxx> writes:
>
> > On some machines, the normal zone can have a large memory hole like
> > below memory layout, and we can see the range from 0x100000000 to
> > 0x1800000000 is a hole. So when isolating some migratable pages, the
> > scanner can meet the hole and it will take more time to skip the large
> > hole. From my measurement, I can see the isolation scanner will take
> > 80us ~ 100us to skip the large hole [0x100000000 - 0x1800000000].
> >
> > So adding a new helper to fast search next online memory section
> > to skip the large hole can help to find next suitable pageblock
> > efficiently. With this patch, I can see the large hole scanning only
> > takes < 1us.
> >
> > [    0.000000] Zone ranges:
> > [    0.000000]   DMA      [mem 0x0000000040000000-0x00000000ffffffff]
> > [    0.000000]   DMA32    empty
> > [    0.000000]   Normal   [mem 0x0000000100000000-0x0000001fa7ffffff]
> > [    0.000000] Movable zone start for each node
> > [    0.000000] Early memory node ranges
> > [    0.000000]   node   0: [mem 0x0000000040000000-0x0000000fffffffff]
> > [    0.000000]   node   0: [mem 0x0000001800000000-0x0000001fa3c7ffff]
> > [    0.000000]   node   0: [mem 0x0000001fa3c80000-0x0000001fa3ffffff]
> > [    0.000000]   node   0: [mem 0x0000001fa4000000-0x0000001fa402ffff]
> > [    0.000000]   node   0: [mem 0x0000001fa4030000-0x0000001fa40effff]
> > [    0.000000]   node   0: [mem 0x0000001fa40f0000-0x0000001fa73cffff]
> > [    0.000000]   node   0: [mem 0x0000001fa73d0000-0x0000001fa745ffff]
> > [    0.000000]   node   0: [mem 0x0000001fa7460000-0x0000001fa746ffff]
> > [    0.000000]   node   0: [mem 0x0000001fa7470000-0x0000001fa758ffff]
> > [    0.000000]   node   0: [mem 0x0000001fa7590000-0x0000001fa7ffffff]
> >
> > Signed-off-by: Baolin Wang <baolin.wang@xxxxxxxxxxxxxxxxx>
> > ---
> >   include/linux/mmzone.h | 10 ++++++++++
> >   mm/compaction.c        | 23 ++++++++++++++++++++++-
> >   2 files changed, 32 insertions(+), 1 deletion(-)
> >
> > diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
> > index 5a7ada0413da..87e6c535d895 100644
> > --- a/include/linux/mmzone.h
> > +++ b/include/linux/mmzone.h
> > @@ -2000,6 +2000,16 @@ static inline unsigned long next_present_section_nr(unsigned long section_nr)
> >   	return -1;
> >   }
> >   
> > +static inline unsigned long next_online_section_nr(unsigned long section_nr)
> > +{
> > +	while (++section_nr <= __highest_present_section_nr) {
> > +		if (online_section_nr(section_nr))
> > +			return section_nr;
> > +	}
> > +
> > +	return -1UL;
> > +}
> > +
> >   /*
> >    * These are _only_ used during initialisation, therefore they
> >    * can use __initdata ...  They could have names to indicate
> > diff --git a/mm/compaction.c b/mm/compaction.c
> > index 3398ef3a55fe..3a55fdd20c49 100644
> > --- a/mm/compaction.c
> > +++ b/mm/compaction.c
> > @@ -229,6 +229,21 @@ static void reset_cached_positions(struct zone *zone)
> >   				pageblock_start_pfn(zone_end_pfn(zone) - 1);
> >   }
> >   
> > +static unsigned long skip_hole_pageblock(unsigned long start_pfn)
> > +{
> > +	unsigned long next_online_nr;
> > +	unsigned long start_nr = pfn_to_section_nr(start_pfn);
> > +
> > +	if (online_section_nr(start_nr))
> > +		return -1UL;
>
> Define a macro for the maigic "-1UL"?  Which is used for multiple times
> in the patch.

I am struggling to find a readable macro for these '-1UL', since the 
'-1UL' in next_online_section_nr() indicates that it can not find an 
online section. However the '-1' in skip_hole_pageblock() indicates that 
it can not find an online pfn.

So after more thinking, I will change to return 'NR_MEM_SECTIONS' if can 
not find next online section in next_online_section_nr(). And in 
skip_hole_pageblock(), I will change to return 0 if can not find next 
online pfn. What do you think?

static unsigned long skip_hole_pageblock(unsigned long start_pfn)
{
        unsigned long next_online_nr;
        unsigned long start_nr = pfn_to_section_nr(start_pfn);

        if (online_section_nr(start_nr))
                return 0;

        next_online_nr = next_online_section_nr(start_nr);
        if (next_online_nr < NR_MEM_SECTIONS)
                return section_nr_to_pfn(next_online_nr);

        return 0;
}

> > +
> > +	next_online_nr = next_online_section_nr(start_nr);
> > +	if (next_online_nr != -1UL)
> > +		return section_nr_to_pfn(next_online_nr);
> > +
> > +	return -1UL;
> > +}
> > +
> >   /*
> >    * Compound pages of >= pageblock_order should consistently be skipped until
> >    * released. It is always pointless to compact pages of such order (if they are
> > @@ -1991,8 +2006,14 @@ static isolate_migrate_t isolate_migratepages(struct compact_control *cc)
> >   
> >   		page = pageblock_pfn_to_page(block_start_pfn,
> >   						block_end_pfn, cc->zone);
> > -		if (!page)
> > +		if (!page) {
> > +			unsigned long next_pfn;
> > +
> > +			next_pfn = skip_hole_pageblock(block_start_pfn);
> > +			if (next_pfn != -1UL)
> > +				block_end_pfn = next_pfn;
> >   			continue;
> > +		}
> >   
> >   		/*
> >   		 * If isolation recently failed, do not retry. Only check the
>
> Do we need to do similar change in isolate_freepages()?

Yes, it's in my todo list with some measurement data.

Thanks for your comments.