linux-kernel - Re: [PATCH 1/3] sched/wait: use freezable

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20190207132801.GA2284@behemoth.owl.eu.com.local>
Date:   Thu, 7 Feb 2019 14:28:01 +0100
From:   Hugo Lefeuvre <hle@....eu.com>
To:     Joel Fernandes <joel@...lfernandes.org>
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Greg Hartman <ghartman@...gle.com>,
        Alistair Strachan <astrachan@...gle.com>,
        Arve Hjønnevåg <arve@...roid.com>,
        Todd Kjos <tkjos@...roid.com>,
        Martijn Coenen <maco@...roid.com>,
        Christian Brauner <christian@...uner.io>,
        Ingo Molnar <mingo@...hat.com>,
        Peter Zijlstra <peterz@...radead.org>,
        devel@...verdev.osuosl.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/3] sched/wait: use freezable_schedule when possible

Hi,

> > The result is a potential performance gain during freeze, since less
> > tasks have to be awaken.
> 
> I'm curious did you try the freezing process and see if pointless wakeups are
> reduced?  That would be an added bonus if you did.

Test env: fresh Debian QEMU vm with 4.19 stable kernel.

Test process:

- Added two debug logs to freeze_task:

    bool freeze_task(struct task_struct *p)
    {
        unsigned long flags;
    [snip]
        pr_info("freezing a task");
    [snip]
        if (freezer_should_skip(p)) {
            pr_info("skeeping a task");
            return false;
        }
    [snip]
    }

- Triggered manual freeze:

# echo freezer > /sys/power/pm_test
# echo test_resume > /sys/power/disk
# echo disk > /sys/power/state

- grep -c to get the number of "freezing a task" and "skeeping a task"
lines in kern.log.

Results:

Without my patch: 448 calls freeze_task, 12 skipped.
With my patch: 448 calls, 32 skipped.

2.6x more tasks skipped.

Not sure this is the best way to test this patch, though. Any advice?

regards,
 Hugo

-- 
                Hugo Lefeuvre (hle)    |    www.owl.eu.com
RSA4096_ 360B 03B3 BF27 4F4D 7A3F D5E8 14AA 1EB8 A247 3DFD
ed25519_ 37B2 6D38 0B25 B8A2 6B9F 3A65 A36F 5357 5F2D DC4C

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)