linux-kernel - Re: [PATCH 14/24] HWPOISON: return 0 if page is assured to be isolated

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: <20091202131554.GB13277@localhost>
Date:	Wed, 2 Dec 2009 21:15:54 +0800
From:	Wu Fengguang <fengguang.wu@...el.com>
To:	Andi Kleen <andi@...stfloor.org>
Cc:	Andrew Morton <akpm@...ux-foundation.org>,
	Nick Piggin <npiggin@...e.de>,
	"linux-mm@...ck.org" <linux-mm@...ck.org>,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 14/24] HWPOISON: return 0 if page is assured to be
	isolated

On Wed, Dec 02, 2009 at 08:47:30PM +0800, Andi Kleen wrote:
> On Wed, Dec 02, 2009 at 11:12:45AM +0800, Wu Fengguang wrote:
> > Introduce hpc.page_isolated to record if page is assured to be
> > isolated, ie. it won't be accessed in normal kernel code paths
> > and therefore won't trigger another MCE event.
> > 
> > __memory_failure() will now return 0 to indicate that page is
> > really isolated.  Note that the original used action result
> > RECOVERED is not a reliable criterion.
> > 
> > Note that we now don't bother to risk returning 0 for the
> > rare unpoison/truncated cases.
> 
> That's the only user of the new hwpoison_control structure right?
> I think I prefer for that single bit to extend the return values
> and keep the arguments around. structures are not nice to read.

Easier to read but harder to extend.  I saw Haicheng add some debug
bits to hwpoison_control to collect debug info ;)

> I'll change the code.

I originally introduce "struct hwpoison_control" to collect more info
(like data_recoverable) and to dump them via uevent. Then we decide to
drop them unless there comes explicit user demands..

In its current form, it does seem more clean to do without hwpoison_control.

Thanks,
Fengguang

--
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/