In unuse_pte_range() we blindly swap-in pages without checking if the
swap entry is already present in the swap cache.
By doing this, the hit/miss ratio used by the swap readahead heuristic
is not properly updated and this leads to non-optimal performance during
swapoff.
Tracing the distribution of the readahead size returned by the swap
readahead heuristic during swapoff shows that a small readahead size is
used most of the time as if we had only misses (this happens both with
cluster and vma readahead), for example:
r::swapin_nr_pages(unsigned long offset):unsigned long:$retval
COUNT EVENT
36948 $retval = 8
44151 $retval = 4
49290 $retval = 1
527771 $retval = 2
Checking if the swap entry is present in the swap cache, instead, allows
to properly update the readahead statistics and the heuristic behaves in
a better way during swapoff, selecting a bigger readahead size:
r::swapin_nr_pages(unsigned long offset):unsigned long:$retval
COUNT EVENT
1618 $retval = 1
4960 $retval = 2
41315 $retval = 4
103521 $retval = 8
In terms of swapoff performance the result is the following:
Testing environment
===================
- Host:
CPU: 1.8GHz Intel Core i7-8565U (quad-core, 8MB cache)
HDD: PC401 NVMe SK hynix 512GB
MEM: 16GB
- Guest (kvm):
8GB of RAM
virtio block driver
16GB swap file on ext4 (/swapfile)
Test case
=========
- allocate 85% of memory
- `systemctl hibernate` to force all the pages to be swapped-out to the
swap file
- resume the system
- measure the time that swapoff takes to complete:
# /usr/bin/time swapoff /swapfile
Result (swapoff time)
======
5.6 vanilla 5.6 w/ this patch
----------- -----------------
cluster-readahead 22.09s 12.19s
vma-readahead 18.20s 15.33s
Signed-off-by: "Huang, Ying" <[email protected]>
Signed-off-by: Andrea Righi <[email protected]>
---
Changes in v3:
- properly update swap readahead statistics instead of forcing a
fixed-size readahead
mm/swapfile.c | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/mm/swapfile.c b/mm/swapfile.c
index 5871a2aa86a5..f8bf926c9c8f 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -1937,10 +1937,14 @@ static int unuse_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
pte_unmap(pte);
swap_map = &si->swap_map[offset];
- vmf.vma = vma;
- vmf.address = addr;
- vmf.pmd = pmd;
- page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE, &vmf);
+ page = lookup_swap_cache(entry, vma, addr);
+ if (!page) {
+ vmf.vma = vma;
+ vmf.address = addr;
+ vmf.pmd = pmd;
+ page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE,
+ &vmf);
+ }
if (!page) {
if (*swap_map == 0 || *swap_map == SWAP_MAP_BAD)
goto try_next;
--
2.25.1
Andrea Righi <[email protected]> writes:
> In unuse_pte_range() we blindly swap-in pages without checking if the
> swap entry is already present in the swap cache.
>
> By doing this, the hit/miss ratio used by the swap readahead heuristic
> is not properly updated and this leads to non-optimal performance during
> swapoff.
It's more important to describe why we need this patch in the patch
description. So, please add some information about your use case. And
please focus on the technical part instead of the business part.
> Tracing the distribution of the readahead size returned by the swap
> readahead heuristic during swapoff shows that a small readahead size is
> used most of the time as if we had only misses (this happens both with
> cluster and vma readahead), for example:
>
> r::swapin_nr_pages(unsigned long offset):unsigned long:$retval
> COUNT EVENT
> 36948 $retval = 8
> 44151 $retval = 4
> 49290 $retval = 1
> 527771 $retval = 2
>
> Checking if the swap entry is present in the swap cache, instead, allows
> to properly update the readahead statistics and the heuristic behaves in
> a better way during swapoff, selecting a bigger readahead size:
>
> r::swapin_nr_pages(unsigned long offset):unsigned long:$retval
> COUNT EVENT
> 1618 $retval = 1
> 4960 $retval = 2
> 41315 $retval = 4
> 103521 $retval = 8
>
> In terms of swapoff performance the result is the following:
>
> Testing environment
> ===================
>
> - Host:
> CPU: 1.8GHz Intel Core i7-8565U (quad-core, 8MB cache)
> HDD: PC401 NVMe SK hynix 512GB
> MEM: 16GB
>
> - Guest (kvm):
> 8GB of RAM
> virtio block driver
> 16GB swap file on ext4 (/swapfile)
>
> Test case
> =========
> - allocate 85% of memory
> - `systemctl hibernate` to force all the pages to be swapped-out to the
> swap file
> - resume the system
> - measure the time that swapoff takes to complete:
> # /usr/bin/time swapoff /swapfile
>
> Result (swapoff time)
> ======
> 5.6 vanilla 5.6 w/ this patch
> ----------- -----------------
> cluster-readahead 22.09s 12.19s
> vma-readahead 18.20s 15.33s
>
> Signed-off-by: "Huang, Ying" <[email protected]>
Thanks! But you don't need to do this. You can add my Reviewed-by after
we have finished the work on patch description.
Best Regards,
Huang, Ying
On Fri, 17 Apr 2020 11:01:22 +0800 "Huang\, Ying" <[email protected]> wrote:
> Andrea Righi <[email protected]> writes:
>
> > In unuse_pte_range() we blindly swap-in pages without checking if the
> > swap entry is already present in the swap cache.
> >
> > By doing this, the hit/miss ratio used by the swap readahead heuristic
> > is not properly updated and this leads to non-optimal performance during
> > swapoff.
>
> It's more important to describe why we need this patch in the patch
> description. So, please add some information about your use case. And
> please focus on the technical part instead of the business part.
Confused. I thought the changelog was quite good. If "business part"
means "end user effect of the patch" then that's a very important
thing.
> Thanks! But you don't need to do this. You can add my Reviewed-by after
> we have finished the work on patch description.
Can you be more specific about how you want this changed?
Andrew Morton <[email protected]> writes:
> On Fri, 17 Apr 2020 11:01:22 +0800 "Huang\, Ying" <[email protected]> wrote:
>
>> Andrea Righi <[email protected]> writes:
>>
>> > In unuse_pte_range() we blindly swap-in pages without checking if the
>> > swap entry is already present in the swap cache.
>> >
>> > By doing this, the hit/miss ratio used by the swap readahead heuristic
>> > is not properly updated and this leads to non-optimal performance during
>> > swapoff.
>>
>> It's more important to describe why we need this patch in the patch
>> description. So, please add some information about your use case. And
>> please focus on the technical part instead of the business part.
>
> Confused. I thought the changelog was quite good. If "business part"
> means "end user effect of the patch" then that's a very important
> thing.
Previously, Andrea has described their use case in the cloud environment
to hiberate the guest and swapoff after resuming. So swapoff
performance is important for them. I think that should be included.
For the business part, I mean something like "Ubuntu used in AWS EC2", I
think that isn't important for the patch description.
>> Thanks! But you don't need to do this. You can add my Reviewed-by after
>> we have finished the work on patch description.
>
> Can you be more specific about how you want this changed?
Please use
Reviewed-by: "Huang, Ying" <[email protected]>
Best Regards,
Huang, Ying
On Fri, Apr 17, 2020 at 01:18:37PM +0800, Huang, Ying wrote:
> Andrew Morton <[email protected]> writes:
>
> > On Fri, 17 Apr 2020 11:01:22 +0800 "Huang\, Ying" <[email protected]> wrote:
> >
> >> Andrea Righi <[email protected]> writes:
> >>
> >> > In unuse_pte_range() we blindly swap-in pages without checking if the
> >> > swap entry is already present in the swap cache.
> >> >
> >> > By doing this, the hit/miss ratio used by the swap readahead heuristic
> >> > is not properly updated and this leads to non-optimal performance during
> >> > swapoff.
> >>
> >> It's more important to describe why we need this patch in the patch
> >> description. So, please add some information about your use case. And
> >> please focus on the technical part instead of the business part.
> >
> > Confused. I thought the changelog was quite good. If "business part"
> > means "end user effect of the patch" then that's a very important
> > thing.
>
> Previously, Andrea has described their use case in the cloud environment
> to hiberate the guest and swapoff after resuming. So swapoff
> performance is important for them. I think that should be included.
> For the business part, I mean something like "Ubuntu used in AWS EC2", I
> think that isn't important for the patch description.
I just sent a v4 of this patch adding "conclusion" section in the
description to better explain the purpose of this patch. Let me know if
you have any comment on that.
Thanks,
-Andrea