Re: Regression in workingset_refault latency on 5.15

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

To: Shakeel Butt <shakeelb@xxxxxxxxxx>
Subject: Re: Regression in workingset_refault latency on 5.15
From: Ivan Babrou <ivan@xxxxxxxxxxxxxx>
Date: Mon, 28 Feb 2022 15:34:25 -0800
Cc: Michal Koutný <mkoutny@xxxxxxxx>, Daniel Dao <dqminh@xxxxxxxxxxxxxx>, kernel-team <kernel-team@xxxxxxxxxxxxxx>, Linux MM <linux-mm@xxxxxxxxx>, Johannes Weiner <hannes@xxxxxxxxxxx>, Roman Gushchin <guro@xxxxxx>, Feng Tang <feng.tang@xxxxxxxxx>, Michal Hocko <mhocko@xxxxxxxxxx>, Hillf Danton <hdanton@xxxxxxxx>, Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>, Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
In-reply-to: <20220228230949.xrmy6j2glxsoffko@google.com>
References: <CABWYdi089+3wG-+6a-O8bnOU_U96i+c9AG3ofS+LocA5oXcE0A@mail.gmail.com> <CA+wXwBSeSzfMAzWzB-ocEDBDnyk3K2+wvuGSzrRUCF-DRuOG6Q@mail.gmail.com> <20220224165838.oir5clpkkqpstpx3@google.com> <CA+wXwBQ0jF0eKZ5x_TRUWmQ3fzV3J+SbBsLmxF0a-OmTVa25pA@mail.gmail.com> <CALvZod4Ln_Ny=OQW0DB7_arBnmjUaTLnYkn-H4BnSuw8XsMeFA@mail.gmail.com> <20220224185236.qgzm3jpoz2orjfcw@google.com> <CA+wXwBRqGwSR+FYGfvxcpJD+bbh4GA7aCz9J1z213cQ_5k181g@mail.gmail.com> <CABWYdi0=_c0VLSBj7RN+Hu9E2FwenqpF-PdLwzp1aWq0fwtmBw@mail.gmail.com> <20220225180345.GD12037@blackbody.suse.cz> <CABWYdi2V=Vf-3+bbsvepo=ha4G1hdizGmg_zVj2pQqx6dHzufQ@mail.gmail.com> <20220228230949.xrmy6j2glxsoffko@google.com>

On Mon, Feb 28, 2022 at 3:09 PM Shakeel Butt <shakeelb@xxxxxxxxxx> wrote:
> Hi Ivan & Daniel,
>
> Can you please try the following patch but without the async patch and
> let us know if there is any difference in regression you are observing?

Hey Shakeel,

I scheduled some reboots for overnight, one machine in each group:

1. Control on 5.15.25
2. Your previous patch with delayed flush on top of 5.15.25
3. You current patch with targeted flush on top of 5.15.25

Either Daniel or I should have some numbers tomorrow.

Follow-Ups:
- Re: Regression in workingset_refault latency on 5.15
  - From: Ivan Babrou
- Re: Regression in workingset_refault latency on 5.15
  - From: Shakeel Butt

References:
- Re: Regression in workingset_refault latency on 5.15
  - From: Ivan Babrou
- Re: Regression in workingset_refault latency on 5.15
  - From: Daniel Dao
- Re: Regression in workingset_refault latency on 5.15
  - From: Shakeel Butt
- Re: Regression in workingset_refault latency on 5.15
  - From: Daniel Dao
- Re: Regression in workingset_refault latency on 5.15
  - From: Shakeel Butt
- Re: Regression in workingset_refault latency on 5.15
  - From: Shakeel Butt
- Re: Regression in workingset_refault latency on 5.15
  - From: Daniel Dao
- Re: Regression in workingset_refault latency on 5.15
  - From: Ivan Babrou
- Re: Regression in workingset_refault latency on 5.15
  - From: Michal Koutný
- Re: Regression in workingset_refault latency on 5.15
  - From: Ivan Babrou
- Re: Regression in workingset_refault latency on 5.15
  - From: Shakeel Butt

Prev by Date: [linux-next:master 9168/9295] {standard input}:2505: Error: pcrel offset for branch to .LS0016 too far (0x3e)
Next by Date: Re: [PATCH 0/5] SLUB debugfs improvements based on stackdepot
Previous by thread: Re: Regression in workingset_refault latency on 5.15
Next by thread: Re: Regression in workingset_refault latency on 5.15
Index(es):
- Date
- Thread

[Index of Archives] [Linux ARM Kernel] [Linux ARM] [Linux Omap] [Fedora ARM] [IETF Annouce] [Bugtraq] [Linux OMAP] [Linux MIPS] [eCos] [Asterisk Internet PBX] [Linux API]