Message-ID: <CAGsJ_4wK=Xw7g=FZgRBo4FRK11AeyS0Jyrd7D0UeDe3DLVUmfA@mail.gmail.com>
Date:   Tue, 17 May 2022 23:14:15 +1200
From:   Barry Song <21cnbao@...il.com>
To:     Rongwei Wang <rongwei.wang@...ux.alibaba.com>
Cc:     sj@...nel.org, Andrew Morton <akpm@...ux-foundation.org>,
        Linux-MM <linux-mm@...ck.org>,
        LKML <linux-kernel@...r.kernel.org>,
        Matthew Wilcox <willy@...radead.org>, shuah@...nel.org,
        Brendan Higgins <brendanhiggins@...gle.com>,
        foersleo@...zon.de, sieberf@...zon.com,
        Shakeel Butt <shakeelb@...gle.com>, sjpark@...zon.de,
        tuhailong@...il.com, Song Jiang <sjiang88@...il.com>,
        张诗明(Simon Zhang) 
        <zhangshiming@...o.com>,
        李培锋(wink) <lipeifeng@...o.com>,
        xhao@...ux.alibaba.com, vxc5208@...s.uta.edu
Subject: Re: DAMON VA regions don't split on an large Android APP

On Tue, May 17, 2022 at 3:00 AM Rongwei Wang
<rongwei.wang@...ux.alibaba.com> wrote:
>
>
>
> On 5/16/22 3:03 PM, Barry Song wrote:
> > On Thu, Apr 28, 2022 at 7:37 PM Barry Song <21cnbao@...il.com> wrote:
> >>
> >> On Thu, Apr 28, 2022 at 2:05 PM Rongwei Wang
> >> <rongwei.wang@...ux.alibaba.com> wrote:
> >>>
> >>>
> >>>
> >>> On 4/27/22 5:22 PM, Barry Song wrote:
> >>>> On Wed, Apr 27, 2022 at 7:44 PM Barry Song <21cnbao@...il.com> wrote:
> >>>>>
> >>>>> On Wed, Apr 27, 2022 at 6:56 PM Rongwei Wang
> >>>>> <rongwei.wang@...ux.alibaba.com> wrote:
> >>>>>>
> >>>>>>
> >>>>>>
> >>>>>> On 4/27/22 7:19 AM, Barry Song wrote:
> >>>>>>> Hi SeongJae & Andrew,
> >>>>>>> (also Cc-ed main damon developers)
> >>>>>>> On an Android phone, I tried to use the DAMON vaddr monitor and found
> >>>>>>> that vaddr regions don't split well on large Android Apps though
> >>>>>>> everything works well on native Apps.
> >>>>>>>
> >>>>>>> I have tried the below two cases on an Android phone with 12GB memory
> >>>>>>> and snapdragon 888 CPU.
> >>>>>>> 1. a native program with small memory working set  as below,
> >>>>>>> #define size (1024*1024*100)
> >>>>>>> main()
> >>>>>>> {
> >>>>>>>            volatile int *p = malloc(size);
> >>>>>>>            memset(p, 0x55, size);
> >>>>>>>
> >>>>>>>            while(1) {
> >>>>>>>                    int i;
> >>>>>>>                    for (i = 0; i < size / 4; i++)
> >>>>>>>                            (void)*(p + i);
> >>>>>>>                    usleep(1000);
> >>>>>>>
> >>>>>>>                    for (i = 0; i < size / 16; i++)
> >>>>>>>                            (void)*(p + i);
> >>>>>>>                    usleep(1000);
> >>>>>>>
> >>>>>>>            }
> >>>>>>> }
> >>>>>>> For this application, the Damon vaddr monitor works very well.
> >>>>>>> I have modified monitor.py in the damo userspace tool a little bit to
> >>>>>>> show the raw data getting from the kernel.
> >>>>>>> Regions can split decently on this kind of applications, a typical raw
> >>>>>>> data is as below,
> >>>>>>>
> >>>>>>> monitoring_start:             2.224 s
> >>>>>>> monitoring_end:               2.329 s
> >>>>>>> monitoring_duration:       104.336 ms
> >>>>>>> target_id: 0
> >>>>>>> nr_regions: 24
> >>>>>>> 005fb37b2000-005fb734a000(  59.594 MiB): 0
> >>>>>>> 005fb734a000-005fbaf95000(  60.293 MiB): 0
> >>>>>>> 005fbaf95000-005fbec0b000(  60.461 MiB): 0
> >>>>>>> 005fbec0b000-005fc2910000(  61.020 MiB): 0
> >>>>>>> 005fc2910000-005fc6769000(  62.348 MiB): 0
> >>>>>>> 005fc6769000-005fca33f000(  59.836 MiB): 0
> >>>>>>> 005fca33f000-005fcdc8b000(  57.297 MiB): 0
> >>>>>>> 005fcdc8b000-005fd115a000(  52.809 MiB): 0
> >>>>>>> 005fd115a000-005fd45bd000(  52.387 MiB): 0
> >>>>>>> 007661c59000-007661ee4000(   2.543 MiB): 2
> >>>>>>> 007661ee4000-0076623e4000(   5.000 MiB): 3
> >>>>>>> 0076623e4000-007662837000(   4.324 MiB): 2
> >>>>>>> 007662837000-0076630f1000(   8.727 MiB): 3
> >>>>>>> 0076630f1000-007663494000(   3.637 MiB): 2
> >>>>>>> 007663494000-007663753000(   2.746 MiB): 1
> >>>>>>> 007663753000-007664251000(  10.992 MiB): 3
> >>>>>>> 007664251000-0076666fd000(  36.672 MiB): 2
> >>>>>>> 0076666fd000-007666e73000(   7.461 MiB): 1
> >>>>>>> 007666e73000-007667c89000(  14.086 MiB): 2
> >>>>>>> 007667c89000-007667f97000(   3.055 MiB): 0
> >>>>>>> 007667f97000-007668112000(   1.480 MiB): 1
> >>>>>>> 007668112000-00766820f000(1012.000 KiB): 0
> >>>>>>> 007ff27b7000-007ff27d6000( 124.000 KiB): 0
> >>>>>>> 007ff27d6000-007ff27d8000(   8.000 KiB): 8
> >>>>>>>
> >>>>>>> 2. a large Android app like Asphalt 9
> >>>>>>> For this case, basically regions can't split very well, but monitor
> >>>>>>> works on small vma:
> >>>>>>>
> >>>>>>> monitoring_start:             2.220 s
> >>>>>>> monitoring_end:               2.318 s
> >>>>>>> monitoring_duration:        98.576 ms
> >>>>>>> target_id: 0
> >>>>>>> nr_regions: 15
> >>>>>>> 000012c00000-0001c301e000(   6.754 GiB): 0
> >>>>>>> 0001c301e000-000371b6c000(   6.730 GiB): 0
> >>>>>>> 000371b6c000-000400000000(   2.223 GiB): 0
> >>>>>>> 005c6759d000-005c675a2000(  20.000 KiB): 0
> >>>>>>> 005c675a2000-005c675a3000(   4.000 KiB): 3
> >>>>>>> 005c675a3000-005c675a7000(  16.000 KiB): 0
> >>>>>>> 0072f1e14000-0074928d4000(   6.510 GiB): 0
> >>>>>>> 0074928d4000-00763c71f000(   6.655 GiB): 0
> >>>>>>> 00763c71f000-0077e863e000(   6.687 GiB): 0
> >>>>>>> 0077e863e000-00798e214000(   6.590 GiB): 0
> >>>>>>> 00798e214000-007b0e48a000(   6.002 GiB): 0
> >>>>>>> 007b0e48a000-007c62f00000(   5.323 GiB): 0
> >>>>>>> 007c62f00000-007defb19000(   6.199 GiB): 0
> >>>>>>> 007defb19000-007f794ef000(   6.150 GiB): 0
> >>>>>>> 007f794ef000-007fe8f53000(   1.745 GiB): 0
> >>>>>>>
> >>>>>>> As you can see, we have some regions which are very very big and they
> >>>>>>> are losing the chance to be splitted. But
> >>>>>>> Damon can still monitor memory access for those small VMA areas very well like:
> >>>>>>> 005c675a2000-005c675a3000(   4.000 KiB): 3
> >>>>>> Hi, Barry
> >>>>>>
> >>>>>> Actually, we also had found the same problem in redis by ourselves
> >>>>>> tool[1]. The DAMON can not split the large anon VMA well, and the anon
> >>>>>> VMA has 10G~20G memory. I guess the whole region doesn't have sufficient
> >>>>>> hot areas to been monitored or found by DAMON, likes one or more address
> >>>>>> choose by DAMON not been accessed during sample period.
> >>>>>
> >>>>> Hi Rongwei,
> >>>>> Thanks  for your comments and thanks for sharing your tools.
> >>>>>
> >>>>> I guess the cause might be:
> >>>>> in case a region is very big like 10GiB, we have only 1MiB hot pages
> >>>>> in this large region.
> >>>>> damon will randomly pick one page to sample, but the page has only
> >>>>> 1MiB/10GiB, thus
> >>>>> less than 1/10000 chance to hit the hot 1MiB. so probably we need
> >>>>> 10000 sample periods
> >>>>> to hit the hot 1MiB in order to split this large region?
> >>>>>
> >>>>> @SeongJae, please correct me if I am wrong.
> >>>>>
> >>>>>>
> >>>>>> I'm not sure whether sets init_regions can deal with the above problem,
> >>>>>> or dynamic choose one or limited number VMA to monitor.
> >>>>>>
> >>>>>
> >>>>> I won't set a limited number of VMA as this will make the damon too hard to use
> >>>>> as nobody wants to make such complex operations, especially an Android
> >>>>> app might have more than 8000 VMAs.
> >>>>>
> >>>>> I agree init_regions might be the right place to enhance the situation.
> >>>>>
> >>>>>> I'm not sure, just share my idea.
> >>>>>>
> >>>>>> [1] https://github.com/aliyun/data-profile-tools.git
> >>>>>
> >>>>> I suppose this tool is based on damon? How do you finally resolve the problem
> >>>>> that large anon VMAs can't be splitted?
> >>>>> Anyway, I will give your tool a try.
> >>>>
> >>>> Unfortunately, data-profile-tools.git doesn't build on aarch64 ubuntu
> >>>> though autogen.sh
> >>>> runs successfully.
> >>>>
> >>>> /usr/bin/ld: ./.libs/libdatop.a(disp.o): in function `cons_handler':
> >>>> /root/data-profile-tools/src/disp.c:625: undefined reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/disp.c:625: undefined
> >>>> reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/disp.c:625: undefined
> >>>> reference to `wgetch'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_win_create':
> >>>> /root/data-profile-tools/src/reg.c:108: undefined reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:108: undefined
> >>>> reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:108: undefined
> >>>> reference to `subwin'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_erase':
> >>>> /root/data-profile-tools/src/reg.c:161: undefined reference to `werase'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_refresh':
> >>>> /root/data-profile-tools/src/reg.c:171: undefined reference to `wrefresh'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_refresh_nout':
> >>>> /root/data-profile-tools/src/reg.c:182: undefined reference to `wnoutrefresh'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_update_all':
> >>>> /root/data-profile-tools/src/reg.c:191: undefined reference to `doupdate'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_win_destroy':
> >>>> /root/data-profile-tools/src/reg.c:200: undefined reference to `delwin'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_line_write':
> >>>> /root/data-profile-tools/src/reg.c:226: undefined reference to `mvwprintw'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:230: undefined
> >>>> reference to `wattr_off'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:217: undefined
> >>>> reference to `wattr_on'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_highlight_write':
> >>>> /root/data-profile-tools/src/reg.c:245: undefined reference to `wattr_on'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:255: undefined
> >>>> reference to `wattr_off'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:252: undefined
> >>>> reference to `mvwprintw'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:255: undefined
> >>>> reference to `wattr_off'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_curses_fini':
> >>>> /root/data-profile-tools/src/reg.c:367: undefined reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:367: undefined
> >>>> reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:367: undefined
> >>>> reference to `wclear'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:368: undefined
> >>>> reference to `wrefresh'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:369: undefined
> >>>> reference to `endwin'
> >>>> /usr/bin/ld: ./.libs/libdatop.a(reg.o): in function `reg_curses_init':
> >>>> /root/data-profile-tools/src/reg.c:382: undefined reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:381: undefined
> >>>> reference to `initscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:382: undefined
> >>>> reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:382: undefined
> >>>> reference to `wrefresh'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:383: undefined
> >>>> reference to `use_default_colors'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:384: undefined
> >>>> reference to `start_color'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:385: undefined
> >>>> reference to `keypad'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:386: undefined
> >>>> reference to `nonl'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:387: undefined
> >>>> reference to `cbreak'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:388: undefined
> >>>> reference to `noecho'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:389: undefined
> >>>> reference to `curs_set'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:401: undefined
> >>>> reference to `stdscr'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:401: undefined
> >>>> reference to `mvwprintw'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:403: undefined
> >>>> reference to `mvwprintw'
> >>>> /usr/bin/ld: /root/data-profile-tools/src/reg.c:405: undefined
> >>>> reference to `wrefresh'
> >>>> collect2: error: ld returned 1 exit status
> >>>> make[1]: *** [Makefile:592: datop] Error 1
> >>>> make[1]: Leaving directory '/root/data-profile-tools'
> >>>> make: *** [Makefile:438: all] Error 2
> >>> Hi, Barry
> >>>
> >>> Now, the question made me realize that the compatibility of this tool is
> >>> very poor. I built a ubuntu environment at yesterday, and fixed above
> >>> errors by:
> >>>
> >>> diff --git a/configure.ac b/configure.ac
> >>> index 7922f27..1ed823c 100644
> >>> --- a/configure.ac
> >>> +++ b/configure.ac
> >>> @@ -21,13 +21,9 @@ AC_PROG_INSTALL
> >>>    AC_CHECK_LIB([numa], [numa_free])
> >>>    AC_CHECK_LIB([pthread], [pthread_create])
> >>>
> >>> -PKG_CHECK_MODULES([CHECK], [check])
> >>> -
> >>> -PKG_CHECK_MODULES([NCURSES], [ncursesw ncurses], [LIBS="$LIBS
> >>> $ncurses_LIBS"], [
> >>> -       AC_SEARCH_LIBS([delwin], [ncursesw ncurses], [], [
> >>> -               AC_MSG_ERROR([ncurses is required but was not found])
> >>> -       ], [])
> >>> -])
> >>> +AC_SEARCH_LIBS([stdscr], [ncurses ncursesw], [], [
> >>> +       AC_MSG_ERROR([required library libncurses or ncurses not found])
> >>> +       ])
> >>>
> >>
> >> I can confirm the patch fixed the issue I reported yesterday, thanks!
> >>
> >>> It works. But I found an another thing will hinder you using this tool.
> >>> We had developed other patches about DAMON base on upstream. This tool
> >>> only works well in ourselves kernel(anolis kernel, already open source).
> >>> Of course, I think it's unnecessary for you to change kernel, just let
> >>> you know this tool still has this problem.
> >>>
> >>
> >> Although I can't use this tool directly as I am not a NUMA right now,
> >> ~/data-profile-tools # ./datop --help
> >> Not support NUMA fault stat (DAMON)!
> >>
> >
> > I wonder if you can extend it to non-numa by setting "remote" to 0%
> > and local to "100%" always for non-numa machines rather than death.
> Hi Barry
>
> That's a great suggestion. Actually, I have removed 'numa_stat' check in
> datop. Maybe you can found. It does not enable numa stat when
> 'numa_stat' sysfs not found in the current system.

yep. i am able to run it on a non-numa machine, but datop immediately crashes
due to some memory corruption issues:

Monitoring 270 processes (interval: 5.5s)

   PID           PROC       TYPE           START             END
SIZE(KiB)    ACCESS      AGE
  1693    Binder:1693       ----               0               0
   0         0        0
   428        ueventd       ----               0               0
   0         0        0
 28654           adbd       ----               0               0
   0         0        0
   971    usb@...-ser       ----               0               0
   0         0        0
   619           logd       ----               0               0
   0         0        0
  4311    a...

<- Hotkey for sorting: 1(PID), 2(START), 3(SIZE), 4(ACCESS), 5(RMA) ->
CPU% = system CPU utilization

Q: Quit; H: Home; B: Back; R: Refresh; D: DAMON
double free or corruption (!prev)

         Aborted

if i move to monitor only one process, datop doesn't crash but it
doesn't show any
data either:

# pgrep youtube
4311
 # ./datop -p 4311

Monitoring 1 processes (interval: 5.0s)

   PID           PROC       TYPE           START             END
SIZE(KiB)   *ACCESS      AGE
  4311    youtube      ----               0               0          0
        0        0


>
> What's more, a new hot key 'f' will be introduced which can enable some
> features dynamically, such as numa stat. Others features can be used
> only in our internal version, likes 'f' in top, and will be open source
> when stable.
>
> > as your tools can map regions to .so, which seems to be quite useful.
> enen, I'm agree with you. But you know, one region maybe covers one or
> more VMAs, hard to map access count of regions to the related .so or
> anon. A lazy way used by me now. I still think it's valuable in the future.
>

it seems really an interesting topic worth more investigation. I wonder if
damon vaddr monitor should actually take vmas, or at least the types of
vmas into consideration while splitting.

Different vma types should be inherently different in hotness. for example,
if 1mb text and 1mb data are put in the same region, the monitored data
to reflect the hotness for the whole 2mb seems to be pointless at all.

Hi SeongJae,
what do you think about it?

> Anyway, any idea are welcome.
>
> Thanks,
> -wrw
>
> >
> >> I am still quite interested in your design and the purpose of this project.
> >> Unfortunately the project seems to be lacking some design doc.
> >>
> >> And would you like to send patches to lkml regarding what you
> >> have changed atop DAMON?
> >>
> >>> Anyway, the question that you reported was valuable, made me realize
> >>> what we need to improve next.
> >>>
> >>> Thanks,
> >>> Rongwei Wang
> >>>>
> >>>>>
> >>>>>>>
> >>>>>>> Typical characteristics of a large Android app is that it has
> >>>>>>> thousands of vma and very large virtual address spaces:
> >>>>>>> ~/damo # pmap 2550 | wc -l
> >>>>>>> 8522
> >>>>>>>
> >>>>>>> ~/damo # pmap 2550
> >>>>>>> ...
> >>>>>>> 0000007992bbe000      4K r----   [ anon ]
> >>>>>>> 0000007992bbf000     24K rw---   [ anon ]
> >>>>>>> 0000007fe8753000      4K -----   [ anon ]
> >>>>>>> 0000007fe8754000   8188K rw---   [ stack ]
> >>>>>>>     total         36742112K
> >>>>>>>
> >>>>>>> Because the whole vma list is too long, I have put the list here for
> >>>>>>> you to download:
> >>>>>>> wget http://www.linuxep.com/patches/android-app-vmas
> >>>>>>>
> >>>>>>> I can reproduce this problem on other Apps like youtube as well.
> >>>>>>> I suppose we need to boost the algorithm of splitting regions for this
> >>>>>>> kind of application.
> >>>>>>> Any thoughts?
> >>>>>>>
> >>>>>
> >

Thanks
Barry

Powered by blists - more mailing lists