Minchan Kim <minchan@xxxxxxxxxx> writes:
> Hi Huang,
>
> On Tue, Oct 31, 2017 at 01:32:32PM +0800, Huang, Ying wrote:
>> Hi, Minchan,
>>
>> Minchan Kim <minchan@xxxxxxxxxx> writes:
>>
>> > Hi Huang,
>> >
>> > On Fri, Oct 27, 2017 at 01:53:27PM +0800, Huang, Ying wrote:
>> >> From: Huang Ying <huang.ying.caritas@xxxxxxxxx>
>> >>
>> >> When a page fault occurs for a swap entry, the physical swap readahead
>> >> (not the VMA base swap readahead) may readahead several swap entries
>> >> after the fault swap entry. The readahead algorithm calculates some
>> >> of the swap entries to readahead via increasing the offset of the
>> >> fault swap entry without checking whether they are beyond the end of
>> >> the swap device and it relys on the __swp_swapcount() and
>> >> swapcache_prepare() to check it. Although __swp_swapcount() checks
>> >> for the swap entry passed in, it will complain with the error message
>> >> as follow for the expected invalid swap entry. This may make the end
>> >> users confused.
>> >>
>> >> swap_info_get: Bad swap offset entry 0200f8a7
>> >>
>> >> To fix the false error message, the swap entry checking is added in
>> >> swap readahead to avoid to pass the out-bound swap entries and the
>> >> swap entry reserved for the swap header to __swp_swapcount() and
>> >> swapcache_prepare().
>> >>
>> >> Cc: Tim Chen <tim.c.chen@xxxxxxxxxxxxxxx>
>> >> Cc: Minchan Kim <minchan@xxxxxxxxxx>
>> >> Cc: Michal Hocko <mhocko@xxxxxxxx>
>> >> Cc: <stable@xxxxxxxxxxxxxxx> # 4.11-4.13
>> >> Reported-by: Christian Kujau <lists@xxxxxxxxxxxxxxx>
>> >> Fixes: e8c26ab60598 ("mm/swap: skip readahead for unreferenced swap slots")
>> >> Signed-off-by: "Huang, Ying" <ying.huang@xxxxxxxxx>
>> >> ---
>> >> include/linux/swap.h | 1 +
>> >> mm/swap_state.c | 6 ++++--
>> >> mm/swapfile.c | 21 +++++++++++++++++++++
>> >> 3 files changed, 26 insertions(+), 2 deletions(-)
>> >>
>> >> diff --git a/include/linux/swap.h b/include/linux/swap.h
>> >> index 84255b3da7c1..43b4b821c805 100644
>> >> --- a/include/linux/swap.h
>> >> +++ b/include/linux/swap.h
>> >> @@ -476,6 +476,7 @@ extern int page_swapcount(struct page *);
>> >> extern int __swap_count(struct swap_info_struct *si, swp_entry_t entry);
>> >> extern int __swp_swapcount(swp_entry_t entry);
>> >> extern int swp_swapcount(swp_entry_t entry);
>> >> +extern bool swap_entry_check(swp_entry_t entry);
>> >> extern struct swap_info_struct *page_swap_info(struct page *);
>> >> extern struct swap_info_struct *swp_swap_info(swp_entry_t entry);
>> >> extern bool reuse_swap_page(struct page *, int *);
>> >> diff --git a/mm/swap_state.c b/mm/swap_state.c
>> >> index 6c017ced11e6..7dd70e77058d 100644
>> >> --- a/mm/swap_state.c
>> >> +++ b/mm/swap_state.c
>> >> @@ -569,11 +569,13 @@ struct page *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask,
>> >> /* Read a page_cluster sized and aligned cluster around offset. */
>> >> start_offset = offset & ~mask;
>> >> end_offset = offset | mask;
>> >> - if (!start_offset) /* First page is swap header. */
>> >> - start_offset++;
>> >>
>> >> blk_start_plug(&plug);
>> >> for (offset = start_offset; offset <= end_offset ; offset++) {
>> >> + swp_entry_t ent = swp_entry(swp_type(entry), offset);
>> >> +
>> >> + if (!swap_entry_check(ent))
>> >> + continue;
>> >> /* Ok, do the async read-ahead now */
>> >> page = __read_swap_cache_async(
>> >> swp_entry(swp_type(entry), offset),
>> >> diff --git a/mm/swapfile.c b/mm/swapfile.c
>> >> index 3074b02eaa09..b04cec29c234 100644
>> >> --- a/mm/swapfile.c
>> >> +++ b/mm/swapfile.c
>> >> @@ -1107,6 +1107,27 @@ static struct swap_info_struct *swap_info_get_cont(swp_entry_t entry,
>> >> return p;
>> >> }
>> >>
>> >> +bool swap_entry_check(swp_entry_t entry)
>> >> +{
>> >> + struct swap_info_struct *p;
>> >> + unsigned long offset, type;
>> >> +
>> >> + type = swp_type(entry);
>> >> + if (type >= nr_swapfiles)
>> >> + goto bad_file;
>> >> + p = swap_info[type];
>> >> + offset = swp_offset(entry);
>> >> + if (unlikely(!offset || offset >= p->max))
>> >> + goto out;
>> >> +
>> >> + return true;
>> >> +
>> >> +bad_file:
>> >> + pr_err("%s: %s%08lx\n", __func__, Bad_file, entry.val);
>> >> +out:
>> >> + return false;
>> >> +}
>> >> +
>> >> static unsigned char __swap_entry_free(struct swap_info_struct *p,
>> >> swp_entry_t entry, unsigned char usage)
>> >> {
>> >> --
>> >> 2.14.2
>> >
>> > Although it's better than old, we can make it simple, still.
>> >
>> > diff --git a/include/linux/swapops.h b/include/linux/swapops.h
>> > index 291c4b534658..f50d5a48f03a 100644
>> > --- a/include/linux/swapops.h
>> > +++ b/include/linux/swapops.h
>> > @@ -41,6 +41,13 @@ static inline unsigned swp_type(swp_entry_t entry)
>> > return (entry.val >> SWP_TYPE_SHIFT(entry));
>> > }
>> >
>> > +extern struct swap_info_struct *swap_info[];
>> > +
>> > +static inline struct swap_info_struct *swp_si(swp_entry_t entry)
>> > +{
>> > + return swap_info[swp_type(entry)];
>> > +}
>> > +
>>
>> Why not just use swp_swap_info()?
>
> Yeap, I forgot the one that even I added the function. :(
>
>>
>> > /*
>> > * Extract the `offset' field from a swp_entry_t. The swp_entry_t is in
>> > * arch-independent format
>> > diff --git a/mm/swap_state.c b/mm/swap_state.c
>> > index 378262d3a197..a0fe2d54ad09 100644
>> > --- a/mm/swap_state.c
>> > +++ b/mm/swap_state.c
>> > @@ -554,6 +554,7 @@ struct page *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask,
>> > struct vm_area_struct *vma, unsigned long addr)
>> > {
>> > struct page *page;
>> > + struct swap_info_struct *si = swp_si(entry);
>> > unsigned long entry_offset = swp_offset(entry);
>> > unsigned long offset = entry_offset;
>> > unsigned long start_offset, end_offset;
>> > @@ -572,6 +573,9 @@ struct page *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask,
>> > if (!start_offset) /* First page is swap header. */
>> > start_offset++;
>> >
>> > + if (end_offset >= si->max)
>> > + end_offset = si->max - 1;
>> > +
>>
>> I don't think two implementation has much difference. Both look OK for
>> me.
>
> The point is *readbility*. We always try to make code simple/understandable.
> Even, it's more important for stable candidate.
>
> We don't need to check the validation in loop so reviewer don't need to
> take care of it once it passes boundary check in above logic.
>
> As you say it's okay, please resend it.
OK.
Best Regards,
Huang, Ying