Re: [PATCH v2 3/5] mm: shmem: add large folio support for tmpfs

[Baolin Wang <baolin.wang@xxxxxxxxxxxxxxxxx>]

On 2024/11/13 00:19, David Hildenbrand wrote:
> On 12.11.24 08:45, Baolin Wang wrote:
> > Add large folio support for tmpfs write and fallocate paths matching the
> > same high order preference mechanism used in the iomap buffered IO path
> > as used in __filemap_get_folio().
> >
> > Add shmem_mapping_size_orders() to get a hint for the orders of the folio
> > based on the file size which takes care of the mapping requirements.
> >
> > Traditionally, tmpfs only supported PMD-sized huge folios. However 
> > nowadays
> > with other file systems supporting any sized large folios, and extending
> > anonymous to support mTHP, we should not restrict tmpfs to allocating 
> > only
> > PMD-sized huge folios, making it more special. Instead, we should allow
> > tmpfs can allocate any sized large folios.
> >
> > Considering that tmpfs already has the 'huge=' option to control the huge
> > folios allocation, we can extend the 'huge=' option to allow any sized 
> > huge
> > folios. The semantics of the 'huge=' mount option are:
> >
> > huge=never: no any sized huge folios
> > huge=always: any sized huge folios
> > huge=within_size: like 'always' but respect the i_size
> > huge=advise: like 'always' if requested with fadvise()/madvise()
> >
> > Note: for tmpfs mmap() faults, due to the lack of a write size hint, 
> > still
> > allocate the PMD-sized huge folios if huge=always/within_size/advise 
> > is set.
> >
> > Moreover, the 'deny' and 'force' testing options controlled by
> > '/sys/kernel/mm/transparent_hugepage/shmem_enabled', still retain the 
> > same
> > semantics. The 'deny' can disable any sized large folios for tmpfs, while
> > the 'force' can enable PMD sized large folios for tmpfs.
> >
> > Co-developed-by: Daniel Gomez <da.gomez@xxxxxxxxxxx>
> > Signed-off-by: Daniel Gomez <da.gomez@xxxxxxxxxxx>
> > Signed-off-by: Baolin Wang <baolin.wang@xxxxxxxxxxxxxxxxx>
> > ---
> >   mm/shmem.c | 91 +++++++++++++++++++++++++++++++++++++++++++++---------
> >   1 file changed, 77 insertions(+), 14 deletions(-)
> >
> > diff --git a/mm/shmem.c b/mm/shmem.c
> > index 86b2e417dc6f..a3203cf8860f 100644
> > --- a/mm/shmem.c
> > +++ b/mm/shmem.c
> > @@ -549,10 +549,50 @@ static bool shmem_confirm_swap(struct 
> > address_space *mapping,
> >   static int shmem_huge __read_mostly = SHMEM_HUGE_NEVER;
> > +/**
> > + * shmem_mapping_size_orders - Get allowable folio orders for the 
> > given file size.
> > + * @mapping: Target address_space.
> > + * @index: The page index.
> > + * @write_end: end of a write, could extend inode size.
> > + *
> > + * This returns huge orders for folios (when supported) based on the 
> > file size
> > + * which the mapping currently allows at the given index. The index 
> > is relevant
> > + * due to alignment considerations the mapping might have. The 
> > returned order
> > + * may be less than the size passed.
> > + *
> > + * Return: The orders.
> > + */
> > +static inline unsigned int
> > +shmem_mapping_size_orders(struct address_space *mapping, pgoff_t 
> > index, loff_t write_end)
> > +{
> > +    unsigned int order;
> > +    size_t size;
> > +
> > +    if (!mapping_large_folio_support(mapping) || !write_end)
> > +        return 0;
> > +
> > +    /* Calculate the write size based on the write_end */
> > +    size = write_end - (index << PAGE_SHIFT);
> > +    order = filemap_get_order(size);
> > +    if (!order)
> > +        return 0;
> > +
> > +    /* If we're not aligned, allocate a smaller folio */
> > +    if (index & ((1UL << order) - 1))
> > +        order = __ffs(index);
> > +
> > +    order = min_t(size_t, order, MAX_PAGECACHE_ORDER);
> > +    return order > 0 ? BIT(order + 1) - 1 : 0;
> > +}
> > +
> >   static unsigned int shmem_huge_global_enabled(struct inode *inode, 
> > pgoff_t index,
> >                             loff_t write_end, bool shmem_huge_force,
> > +                          struct vm_area_struct *vma,
> >                             unsigned long vm_flags)
> >   {
> > +    unsigned long within_size_orders;
> > +    unsigned int order;
> > +    pgoff_t aligned_index;
> >       loff_t i_size;
> >       if (HPAGE_PMD_ORDER > MAX_PAGECACHE_ORDER)
>
> We can allow all orders up to MAX_PAGECACHE_ORDER, 
> shmem_mapping_size_orders() handles it properly.
>
> So maybe we should drop this condition and use instead below where we have
>
> return BIT(HPAGE_PMD_ORDER);
>
> instead something like.
>
> return HPAGE_PMD_ORDER > MAX_PAGECACHE_ORDER ? 0 : BIT(HPAGE_PMD_ORDER);
>
> Ideally, factoring it out somehow
>
>
> int maybe_pmd_order = HPAGE_PMD_ORDER > MAX_PAGECACHE_ORDER ? 0 : 
> BIT(HPAGE_PMD_ORDER);
>
> ...
>
> return maybe_pmd_order;

Good point. Will do.

> > @@ -564,15 +604,41 @@ static unsigned int 
> > shmem_huge_global_enabled(struct inode *inode, pgoff_t index
> >       if (shmem_huge_force || shmem_huge == SHMEM_HUGE_FORCE)
> >           return BIT(HPAGE_PMD_ORDER);
>
> Why not force-enable all orders (of course, respecting 
> MAX_PAGECACHE_ORDER and possibly VMA)?

The ‘force’ option will affect the tmpfs mmap()'s huge allocation, which 
I intend to handle in a separate patch as we discussed. Additionally, 
for the huge page allocation of tmpfs mmap(), I am also considering the 
readahead approach for the pagecache.

> > +    /*
> > +     * The huge order allocation for anon shmem is controlled through
> > +     * the mTHP interface, so we still use PMD-sized huge order to
> > +     * check whether global control is enabled.
> > +     *
> > +     * For tmpfs mmap()'s huge order, we still use PMD-sized order to
> > +     * allocate huge pages due to lack of a write size hint.
> > +     *
> > +     * Otherwise, tmpfs will allow getting a highest order hint based on
> > +     * the size of write and fallocate paths, then will try each 
> > allowable
> > +     * huge orders.
> > +     */
> >       switch (SHMEM_SB(inode->i_sb)->huge) {
> >       case SHMEM_HUGE_ALWAYS:
> > -        return BIT(HPAGE_PMD_ORDER);
> > -    case SHMEM_HUGE_WITHIN_SIZE:
> > -        index = round_up(index + 1, HPAGE_PMD_NR);
> > -        i_size = max(write_end, i_size_read(inode));
> > -        i_size = round_up(i_size, PAGE_SIZE);
> > -        if (i_size >> PAGE_SHIFT >= index)
> > +        if (vma)
> >               return BIT(HPAGE_PMD_ORDER);
> > +
> > +        return shmem_mapping_size_orders(inode->i_mapping, index, 
> > write_end);
> > +    case SHMEM_HUGE_WITHIN_SIZE:
> > +        if (vma)
> > +            within_size_orders = BIT(HPAGE_PMD_ORDER);
> > +        else
> > +            within_size_orders = 
> > shmem_mapping_size_orders(inode->i_mapping,
> > +                                       index, write_end);
> > +
> > +        order = highest_order(within_size_orders);
> > +        while (within_size_orders) {
> > +            aligned_index = round_up(index + 1, 1 << order);
> > +            i_size = max(write_end, i_size_read(inode));
> > +            i_size = round_up(i_size, PAGE_SIZE);
> > +            if (i_size >> PAGE_SHIFT >= aligned_index)
> > +                return within_size_orders;
> > +
> > +            order = next_order(&within_size_orders, order);
> > +        }
> >           fallthrough;
> >       case SHMEM_HUGE_ADVISE:
> >           if (vm_flags & VM_HUGEPAGE)
>
> I think the point here is that "write" -> no VMA -> vm_flags == 0 -> no 
> code changes needed :)

Yes. Currently the fadvise() have no HUGEPAGE handling, so I will drop 
the 'fadvise' in the doc.