[RFCv2 0/6] introduce memory hinting API for external process
From: Minchan Kim
Date: Fri May 31 2019 - 02:47:06 EST
- Background
The Android terminology used for forking a new process and starting an app
from scratch is a cold start, while resuming an existing app is a hot start.
While we continually try to improve the performance of cold starts, hot
starts will always be significantly less power hungry as well as faster so
we are trying to make hot start more likely than cold start.
To increase hot start, Android userspace manages the order that apps should
be killed in a process called ActivityManagerService. ActivityManagerService
tracks every Android app or service that the user could be interacting with
at any time and translates that into a ranked list for lmkd(low memory
killer daemon). They are likely to be killed by lmkd if the system has to
reclaim memory. In that sense they are similar to entries in any other cache.
Those apps are kept alive for opportunistic performance improvements but
those performance improvements will vary based on the memory requirements of
individual workloads.
- Problem
Naturally, cached apps were dominant consumers of memory on the system.
However, they were not significant consumers of swap even though they are
good candidate for swap. Under investigation, swapping out only begins
once the low zone watermark is hit and kswapd wakes up, but the overall
allocation rate in the system might trip lmkd thresholds and cause a cached
process to be killed(we measured performance swapping out vs. zapping the
memory by killing a process. Unsurprisingly, zapping is 10x times faster
even though we use zram which is much faster than real storage) so kill
from lmkd will often satisfy the high zone watermark, resulting in very
few pages actually being moved to swap.
- Approach
The approach we chose was to use a new interface to allow userspace to
proactively reclaim entire processes by leveraging platform information.
This allowed us to bypass the inaccuracy of the kernelâs LRUs for pages
that are known to be cold from userspace and to avoid races with lmkd
by reclaiming apps as soon as they entered the cached state. Additionally,
it could provide many chances for platform to use much information to
optimize memory efficiency.
To achieve the goal, the patchset introduce two new options for madvise.
One is MADV_COLD which will deactivate activated pages and the other is
MADV_PAGEOUT which will reclaim private pages instantly. These new options
complement MADV_DONTNEED and MADV_FREE by adding non-destructive ways to
gain some free memory space. MADV_PAGEOUT is similar to MADV_DONTNEED in a way
that it hints the kernel that memory region is not currently needed and
should be reclaimed immediately; MADV_COLD is similar to MADV_FREE in a way
that it hints the kernel that memory region is not currently needed and
should be reclaimed when memory pressure rises.
This approach is similar in spirit to madvise(MADV_WONTNEED), but the
information required to make the reclaim decision is not known to the app.
Instead, it is known to a centralized userspace daemon, and that daemon
must be able to initiate reclaim on its own without any app involvement.
To solve the concern, this patch introduces new syscall -
struct pr_madvise_param {
int size; /* the size of this structure */
int cookie; /* reserved to support atomicity */
int nr_elem; /* count of below arrary fields */
int __user *hints; /* hints for each range */
/* to store result of each operation */
const struct iovec __user *results;
/* input address ranges */
const struct iovec __user *ranges;
};
int process_madvise(int pidfd, struct pr_madvise_param *u_param,
unsigned long flags);
The syscall get pidfd to give hints to external process and provides
pair of result/ranges vector arguments so that it could give several
hints to each address range all at once. It also has cookie variable
to support atomicity of the API for address ranges operations. IOW, if
target process changes address space since monitor process has parsed
address ranges via map_files or maps, the API can detect the race so
could cancel entire address space operation. It's not implemented yet.
Daniel Colascione suggested a idea(Please read description in patch[6/6])
and this patchset adds cookie a variable for the future.
- Experiment
We did bunch of testing with several hundreds of real users, not artificial
benchmark on android. We saw about 17% cold start decreasement without any
significant battery/app startup latency issues. And with artificial benchmark
which launches and switching apps, we saw average 7% app launching improvement,
18% less lmkd kill and good stat from vmstat.
A is vanilla and B is process_madvise.
A B delta ratio(%)
allocstall_dma 0 0 0 0.00
allocstall_movable 1464 457 -1007 -69.00
allocstall_normal 263210 190763 -72447 -28.00
allocstall_total 264674 191220 -73454 -28.00
compact_daemon_wake 26912 25294 -1618 -7.00
compact_fail 17885 14151 -3734 -21.00
compact_free_scanned 4204766409 3835994922 -368771487 -9.00
compact_isolated 3446484 2967618 -478866 -14.00
compact_migrate_scanned 1621336411 1324695710 -296640701 -19.00
compact_stall 19387 15343 -4044 -21.00
compact_success 1502 1192 -310 -21.00
kswapd_high_wmark_hit_quickly 234 184 -50 -22.00
kswapd_inodesteal 221635 233093 11458 5.00
kswapd_low_wmark_hit_quickly 66065 54009 -12056 -19.00
nr_dirtied 259934 296476 36542 14.00
nr_vmscan_immediate_reclaim 2587 2356 -231 -9.00
nr_vmscan_write 1274232 2661733 1387501 108.00
nr_written 1514060 2937560 1423500 94.00
pageoutrun 67561 55133 -12428 -19.00
pgactivate 2335060 1984882 -350178 -15.00
pgalloc_dma 13743011 14096463 353452 2.00
pgalloc_movable 0 0 0 0.00
pgalloc_normal 18742440 16802065 -1940375 -11.00
pgalloc_total 32485451 30898528 -1586923 -5.00
pgdeactivate 4262210 2930670 -1331540 -32.00
pgfault 30812334 31085065 272731 0.00
pgfree 33553970 31765164 -1788806 -6.00
pginodesteal 33411 15084 -18327 -55.00
pglazyfreed 0 0 0 0.00
pgmajfault 551312 1508299 956987 173.00
pgmigrate_fail 43927 29330 -14597 -34.00
pgmigrate_success 1399851 1203922 -195929 -14.00
pgpgin 24141776 19032156 -5109620 -22.00
pgpgout 959344 1103316 143972 15.00
pgpgoutclean 4639732 3765868 -873864 -19.00
pgrefill 4884560 3006938 -1877622 -39.00
pgrotated 37828 25897 -11931 -32.00
pgscan_direct 1456037 957567 -498470 -35.00
pgscan_direct_throttle 0 0 0 0.00
pgscan_kswapd 6667767 5047360 -1620407 -25.00
pgscan_total 8123804 6004927 -2118877 -27.00
pgskip_dma 0 0 0 0.00
pgskip_movable 0 0 0 0.00
pgskip_normal 14907 25382 10475 70.00
pgskip_total 14907 25382 10475 70.00
pgsteal_direct 1118986 690215 -428771 -39.00
pgsteal_kswapd 4750223 3657107 -1093116 -24.00
pgsteal_total 5869209 4347322 -1521887 -26.00
pswpin 417613 1392647 975034 233.00
pswpout 1274224 2661731 1387507 108.00
slabs_scanned 13686905 10807200 -2879705 -22.00
workingset_activate 668966 569444 -99522 -15.00
workingset_nodereclaim 38957 32621 -6336 -17.00
workingset_refault 2816795 2179782 -637013 -23.00
workingset_restore 294320 168601 -125719 -43.00
pgmajfault is increased by 173% because swapin is increased by 200% by
process_madvise hint. However, swap read based on zram is much cheaper
than file IO in performance point of view and app hot start by swapin is
also cheaper than cold start from the beginning of app which needs many IO
from storage and initialization steps.
Brian Geffon in ChromeOS team had an experiment with process_madvise(2)
Quote form him:
"What I found is that by using process_madvise after a tab has been back
grounded for more than 45 seconds reduced the average tab switch times by
25%! This is a huge result and very obvious validation that process_madvise
hints works well for the ChromeOS use case."
This patchset is against on next-20190530.
Minchan Kim (6):
[1/6] mm: introduce MADV_COLD
[2/6] mm: change PAGEREF_RECLAIM_CLEAN with PAGE_REFRECLAIM
[3/6] mm: introduce MADV_PAGEOUT
[4/6] mm: factor out madvise's core functionality
[5/6] mm: introduce external memory hinting API
[6/6] mm: extend process_madvise syscall to support vector arrary
arch/x86/entry/syscalls/syscall_32.tbl | 1 +
arch/x86/entry/syscalls/syscall_64.tbl | 1 +
include/linux/page-flags.h | 1 +
include/linux/page_idle.h | 15 +
include/linux/pid.h | 4 +
include/linux/swap.h | 2 +
include/linux/syscalls.h | 3 +
include/uapi/asm-generic/mman-common.h | 13 +
include/uapi/asm-generic/unistd.h | 4 +-
kernel/fork.c | 8 +
kernel/signal.c | 7 +-
kernel/sys_ni.c | 1 +
mm/madvise.c | 586 +++++++++++++++++++++----
mm/swap.c | 43 ++
mm/vmscan.c | 83 +++-
15 files changed, 691 insertions(+), 81 deletions(-)
* from RFCv1
* Dropped MADV_[ANONYMOUS|FILE]_FILTER option. The option gave several
hundered millisecond improvement via removing address range parsing
overhead. However, there is other suggestion to create general API
to provide processs information(process_getinfo(2)) which would be
very fast via binary form and get only necessary information.
* There was lots of discussion how to provide *consistency*,*atomicity*
against on target process's address space change. It needs more
discussion. Let's do that in [6/6] if you have still a argument.
* There was a concern about the vector support because it didn't show
great performance benefit. However, I still included it because
it could make us address space range operation's atomicity easier
in future without introducing other new syscall.
* Naming of process_madvise - there was request to change naming from
process_madvise to pidfd_madvise for the consistency of existing
pidfd syscall but some of people including me still want to have
syscall name focus on what it's doing rather than how it does.
* I limited hints new syscall supports as MADV_COLD|PAGEOUT at this
moment because I'm not sure all hints makes sense for external
processs and would be some lurked bugs which relies on the caller
context. Please see description in [5/6].
--
2.22.0.rc1.257.g3120a18244-goog