linux-kernel - Re: [PATCH] group xtime, xtime_lock, wall_to_monotonic, avenrun, calc_load

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: <20061211140034.fabb840f.akpm@osdl.org>
Date:	Mon, 11 Dec 2006 14:00:34 -0800
From:	Andrew Morton <akpm@...l.org>
To:	Eric Dumazet <dada1@...mosbay.com>
Cc:	Andi Kleen <ak@...e.de>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] group xtime, xtime_lock, wall_to_monotonic, avenrun,
 calc_load_count fields together in ktimed

On Mon, 11 Dec 2006 21:44:34 +0100
Eric Dumazet <dada1@...mosbay.com> wrote:

> Andrew Morton a __crit :
> > 
> > hm, the patch seems to transform a mess into a mess.  I guess it's a messy
> > problem.
> > 
> > I agree that aggregating all the time-related things into a struct like
> > this makes some sense.  As does aggregating them all into a similar-looking
> > namespace, but that'd probably be too intrusive - too late for that.
> 
> 
> Hi Andrew, thanks for your comments.
> 
> I sent two patches for the __attribute__((weak)) xtime_lock thing, and 
> calc_load() optimization, which dont depend on ktimed.

yup, thanks.

> Should I now send patches for aggregating things or is it considered too 
> intrusive ?

The previous version didn't look too intrusive.  But it would be nice to
have a plan to get rid of the macros:

#define xtime_lock	ktimed.xtime_lock

and just open-code this everywhere.

> (Sorry if I didnt understand your last sentence)

What I meant was: if we're not going to to aggregate all these globals like
this:

	ktimed.xtime_lock
	ktimed.wall_to_monotonic

then it would be nice if they were at least aggregated by naming convention:

	time_management_time_lock
	time_management_wall_to_monotonic
	etc

so the reader can see that these things are all part of the same subsystem.

But the proposed ktimed.xtime_lock achieves that, and has runtime benefits
too.

Can we please not call it ktimed?  That sounds like a kernel thread to me. 
time_data would be better.

> If yes, should I send separate patches to :
> 
> 1) define an empty ktimed (or with a placeholder for jiffies64, not yet used)
> 2) move xtime into ktimed
> 3) move xtime_lock into ktimed
> 4) move wall_to_monotonic into ktimed
> 5) move calc_load.count into ktimed
> 6) move avenrun into ktimed.

A single patch there would suffice, I suspect.

> 7) patches to use ktimed.jiffies64 on various arches (with the problem of 
> aliasing jiffies)

That might be a sprinkle of per-arch patches, but I'm not sure what is
entailed here.

-
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/