Re: [lkp] [mm] 55834c5909: boot-slabinfo.num_objs +31.9% increasement
From: Dmitry Vyukov
Date: Mon May 30 2016 - 03:53:50 EST
On Mon, May 30, 2016 at 4:17 AM, kernel test robot
<xiaolong.ye@xxxxxxxxx> wrote:
>
> FYI, we noticed boot-slabinfo.num_objs +31.9% increasement due to commit:
>
> commit 55834c59098d0c5a97b0f3247e55832b67facdcf ("mm: kasan: initial memory quarantine implementation")
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
>
> in testcase: boot
> on test machine: vm-lkp-wsx03-2G: 2 threads qemu-system-x86_64 -enable-kvm -cpu host with 2G memory
> with following parameters: sleep=1
>
>
> Details are as below:
> -------------------------------------------------------------------------------------------------->
>
>
> =========================================================================================
> compiler/kconfig/rootfs/sleep/tbox_group/testcase:
> gcc-6/x86_64-randconfig-s4-05280557/debian-x86_64-2015-02-07.cgz/1/vm-lkp-wsx03-2G/boot
>
> commit:
> b8f1a75d61d8405a753380c6fb17ba84a5603cd4
> 55834c59098d0c5a97b0f3247e55832b67facdcf
>
> b8f1a75d61d8405a 55834c59098d0c5a97b0f3247e
> ---------------- --------------------------
> %stddev %change %stddev
> \ | \
> 165450 Â 0% +31.9% 218176 Â 1% boot-slabinfo.num_objs
> 19095 Â 5% +104.3% 39013 Â 3% boot-slabinfo.num_pages
> 107.27 Â 2% +14.1% 122.43 Â 13% boot-time.boot
> 97.13 Â 1% +4.3% 101.31 Â 1% boot-time.idle
> 1105076 Â 0% -7.2% 1025056 Â 0% boot-meminfo.MemFree
> 36269 Â 12% +215.7% 114517 Â 5% boot-meminfo.SUnreclaim
> 76469 Â 5% +104.5% 156343 Â 4% boot-meminfo.Slab
>
>
>
> boot-slabinfo.num_objs
>
> 230000 ++-----------------------------------------------------------------+
> | O O |
> 220000 ++O O O OO OO O O OO OO |
> OO O OOOO O OOOOOOO O O OO O |
> 210000 ++ |
> | |
> 200000 ++ |
> | |
> 190000 ++ |
> | |
> 180000 ++ |
> | |
> 170000 ++ |
> ********************************************************************
> 160000 ++-----------------------------------------------------------------+
>
>
> boot-slabinfo.num_pages
>
> 45000 ++-------------O----------------------------------------------------+
> | O O OO O O O O OOO |
> 40000 +O OOO OO OO O O |
> O OO O O O O O OO O O |
> | O O O |
> 35000 ++ |
> | |
> 30000 ++ |
> | |
> 25000 ++ |
> | |
> | * * *
> 20000 +* * * ** * * * * * * ** **** ** * * * * * * :: * *** *|
> * * ** ** * * ** *** * ** *** * * * * * ** *** * ** ** ** * * |
> 15000 ++------------------------------------------------------------------+
>
>
> [*] bisect-good sample
> [O] bisect-bad sample
Hi Xiaolong,
It is with CONFIG_KASAN=y? If yes, then increased memory usage is expected.