linux-kernel - Re: [PATCH 12/28] readahead: state based method

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] [day] [month] [year] [list]

Message-ID: <20061116133919.GA6645@mail.ustc.edu.cn>
Date:	Thu, 16 Nov 2006 21:39:19 +0800
From:	Wu Fengguang <wfg@...l.ustc.edu.cn>
To:	Christoph Lameter <clameter@....com>
Cc:	Andrew Morton <akpm@...l.org>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 12/28] readahead: state based method - aging accounting

On Wed, Nov 15, 2006 at 08:54:44AM -0800, Christoph Lameter wrote:
> On Wed, 15 Nov 2006, Wu Fengguang wrote:
> 
> > Collect info about the global available memory and its consumption speed.
> > The data are used by the stateful method to estimate the thrashing threshold.
> 
> Looks like you should use a ZVC counter for total scanned. See 
> include/linux/mmzone.h.

OK.

By using zone.total_scanned, I have chose an easy way :)

To do the general vm timing in something like zone.vm_stat[NR_SCAN_INACTIVE],
a set of new functions will be required:

        global_page_state_raw()
        zone_page_state_raw()
        node_page_state_raw()

They do not check overflows, so that we can do

        time_elapsed = new_raw_value - old_raw_value;

However, before introducing the ugly *_raw() functions, I'd like to know if

        #ifdef CONFIG_SMP
                if (x < 0)
                        x = 0;
        #endif  

really helps some big NUMA system. I suspect object counters like
NR_FILE_PAGES will _never_ overflow, and an accumulated counter like
NR_VMSCAN_WRITE is expected to overflow. In either case, it is ok to
return an unsigned long raw counter.

Regards,
Wu
-
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/