linux-kernel - Re: taskstats root only breaking iotop

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: <CAA879TPzmVh4yR5n4Mfd8D6ERpvAFipeMXr1QBr3fwQMdvnRzw@mail.gmail.com>
Date:	Sun, 2 Oct 2011 00:41:59 +0200
From:	Guillaume Chazarain <guichaz@...il.com>
To:	Linus Torvalds <torvalds@...ux-foundation.org>
Cc:	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
	Balbir Singh <bsingharora@...il.com>,
	Vasiliy Kulikov <segoon@...nwall.com>
Subject: Re: taskstats root only breaking iotop

On Sat, Oct 1, 2011 at 11:59 PM, Linus Torvalds
<torvalds@...ux-foundation.org> wrote:
> I think the question should be: why is it *ever* a good idea to let
> *anybody* read how many bytes anybody has read.

I use iotop to get the throughput of a cp or tar because I don't
always think of piping to 'pv'. People also use it as a progress
indicator for their less verbose applications (total bytes
read/written instead of rate).

I see on forums people use it to find which process is hammering their
disk, though in this case the time spent on I/Os is more useful than
the throughput and this is not broken by these changes assuming they
have root.

> If you want that kind of detail, do "strace". Don't do that abortion
> that is taskstats.

strace is not always appropriate: it's invasive as it slows down the
traced process and it's not aggregating anything.

> Right now, TASKSTATS is a total and utter disaster.

I'm not attached to taskstats, I'd be happy to migrate to what comes next.

> and it's a crazy idea to
> begin with.

But it's not clear to me if you dislike just the taskstats
implementation or the general idea of precise per task I/O stats.

Thanks for taking the time to explain your reasoning, I'm not trying
to resurrect taskstats, just checking if I should expect a brighter
future for iotop.

-- 
Guillaume
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/