[PATCH -v4 00/10] THP swap: Delay splitting THP during swapping out
From: Huang, Ying
Date: Thu Sep 29 2016 - 02:34:16 EST
From: Huang Ying <ying.huang@xxxxxxxxx>
This patchset is to optimize the performance of Transparent Huge Page
Hi, Andrew, could you help me to check whether the overall design is
Hi, Hugh, Shaohua, Minchan and Rik, could you help me to review the
swap part of the patchset? Especially [01/10], [04/10], [05/10],
[06/10], [07/10], [10/10].
Hi, Andrea and Kirill, could you help me to review the THP part of the
patchset? Especially [02/10], [03/10], [09/10] and [10/10].
Hi, Johannes, Michal and Vladimir, I am not very confident about the
memory cgroup part, especially [02/10] and [03/10]. Could you help me
to review it?
And for all, Any comment is welcome!
Recently, the performance of the storage devices improved so fast that
we cannot saturate the disk bandwidth when do page swap out even on a
high-end server machine. Because the performance of the storage
device improved faster than that of CPU. And it seems that the trend
will not change in the near future. On the other hand, the THP
becomes more and more popular because of increased memory size. So it
becomes necessary to optimize THP swap performance.
The advantages of the THP swap support include:
- Batch the swap operations for the THP to reduce lock
acquiring/releasing, including allocating/freeing the swap space,
adding/deleting to/from the swap cache, and writing/reading the swap
space, etc. This will help improve the performance of the THP swap.
- The THP swap space read/write will be 2M sequential IO. It is
particularly helpful for the swap read, which are usually 4k random
IO. This will improve the performance of the THP swap too.
- It will help the memory fragmentation, especially when the THP is
heavily used by the applications. The 2M continuous pages will be
free up after THP swapping out.
- It will improve the THP utilization on the system with the swap
turned on. Because the speed for khugepaged to collapse the normal
pages into the THP is quite slow. After the THP is split during the
swapping out, it will take quite long time for the normal pages to
collapse back into the THP after being swapped in. The high THP
utilization helps the efficiency of the page based memory management
This patchset is based on 9/20 head of mmotm/master.
This patchset is the first step for the THP swap support. The plan is
to delay splitting THP step by step, finally avoid splitting THP
during the THP swapping out and swap out/in the THP as a whole.
As the first step, in this patchset, the splitting huge page is
delayed from almost the first step of swapping out to after allocating
the swap space for the THP and adding the THP into the swap cache.
This will reduce lock acquiring/releasing for the locks used for the
swap cache management.
With the patchset, the swap out throughput improves 12.1% (from about
1.12GB/s to about 1.25GB/s) in the vm-scalability swap-w-seq test case
with 16 processes. The test is done on a Xeon E5 v3 system. The swap
device used is a RAM simulated PMEM (persistent memory) device. To
test the sequential swapping out, the test case uses 16 processes,
which sequentially allocate and write to the anonymous pages until the
RAM and part of the swap device is used up.
The detailed compare result is as follow,
%stddev %change %stddev
\ | \
1118821 Â 0% +12.1% 1254241 Â 1% vmstat.swap.so
2460636 Â 1% +10.6% 2720983 Â 1% vm-scalability.throughput
308.79 Â 1% -7.9% 284.53 Â 1% vm-scalability.time.elapsed_time
1639 Â 4% +232.3% 5446 Â 1% meminfo.SwapCached
0.70 Â 3% +8.7% 0.77 Â 5% perf-stat.ipc
9.82 Â 8% -31.6% 6.72 Â 2% perf-profile.cycles-pp._raw_spin_lock_irq.__add_to_swap_cache.add_to_swap_cache.add_to_swap.shrink_page_list
>From the swap out throughput number, we can find, even tested on a RAM
simulated PMEM (Persistent Memory) device, the swap out throughput can
reach only about 1.1GB/s. While, in the file IO test, the sequential
write throughput of an Intel P3700 SSD can reach about 1.8GB/s
steadily. And according the following URL,
The sequential write throughput of Intel P3608 SSD can reach about
3.0GB/s, while the random read IOPS can reach about 850k. It is clear
that the bottleneck has moved from the disk to the kernel swap
The improved storage device performance should have made the swap
becomes a better feature than before with better performance. But
because of the issues of kernel swap component itself, the swap
performance is still kept at the low level. That prevents the swap
feature to be used by more users. And this in turn causes few kernel
developers think it is necessary to optimize kernel swap component.
To break the loop, we need to optimize the performance of kernel swap
component. Optimize the THP swap performance is part of it.
- Per Johannes' comments, simplified swap cgroup array accessing code.
- Per Kirill and Dave Hansen's comments, used HPAGE_PMD_NR instead of
- Per Anshuman's comments, used HPAGE_PMD_NR instead of 512 in patch
- Per Andrew's suggestion, used a more systematical way to determine
whether to enable THP swap optimization
- Per Andrew's comments, moved as much as possible code into
#ifdef CONFIG_TRANSPARENT_HUGE_PAGE/#endif or "if (PageTransHuge())"
- Fixed some coding style warning.
- Original [1/11] sent separately and merged
- Use switch in 10/10 per Hiff's suggestion