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: <4A311727.40403@lwfinger.net>
Date:	Thu, 11 Jun 2009 09:39:35 -0500
From:	Larry Finger <Larry.Finger@...inger.net>
To:	Pekka Enberg <penberg@...helsinki.fi>
CC:	Christoph Lameter <cl@...ux-foundation.org>,
	linux-kernel@...r.kernel.org, mel@....ul.ie
Subject: Re: [PATCH 2/2] SLUB: Disable debugging if it increases the minimum
 page order

Pekka Enberg wrote:
> Hi Christoph,
> 
> On Thu, 2009-06-11 at 09:43 -0400, Christoph Lameter wrote:
>> We have had that with SLAB. NO! This leads to the situation that some
>> slabs have debug on and some have not. You just do not know which.
> 
> I do see your point but surely we don't want to use order 1 allocations
> in the fall-back case for kmalloc-4096? Couldn't we just add a printk
> saying that debug was disabled for the cache? After all, my patch is
> much better than what SLAB does.
> 
> On Thu, 2009-06-11 at 09:43 -0400, Christoph Lameter wrote:
>> Note that CONFIG_SLUB_DEBUG only enables the code to debug a slab. It does
>> not enable debugging for each slab. CONFIG_SLUB_DEBUG_ON does that.
> 
> True. Larry, do you have CONFIG_SLUB_DEBUG_ON enabled or are you passing
> SLUB debugging options to the kernel?

It is enabled in the kernel. My SLUB options in .config are:

finger@...rylap:~> grep SLUB .config
CONFIG_SLUB_DEBUG=y
CONFIG_SLUB=y
CONFIG_SLUB_DEBUG_ON=y
# CONFIG_SLUB_STATS is not set

Would having STATS enabled help?

For a bug that hits infrequently, determining that it is fixed is
difficult. That said, my system has been up about 22.5 hours during
which I have tried to force the failure. I see the fragmentation of
memory vary widely as shown below:

finger@...rylap:~> date ; cat /proc/buddyinfo
Wed Jun 10 20:15:34 CDT 2009
Node 0, zone      DMA      4      3      5      2      4      1      2
     0      1      0      0
Node 0, zone    DMA32   5920  11678   2245    369    117     21      2
     1      0      0      0
finger@...rylap:~> date ; cat /proc/buddyinfo
Wed Jun 10 23:32:39 CDT 2009
Node 0, zone      DMA      4      3      5      2      4      1      2
     0      1      0      0
Node 0, zone    DMA32   2605   4140   3804      7      0      1      1
     1      0      0      0
finger@...rylap:~> date ; cat /proc/buddyinfo
Thu Jun 11 09:28:06 CDT 2009
Node 0, zone      DMA      4      3      5      2      4      1      2
     0      1      0      0
Node 0, zone    DMA32   2231    429   2726     54      1      0      1
     1      0      0      0
finger@...rylap:~> cat /proc/uptime
80678.11 78.54

The latest value of the number of O(1) fragments is about as low as I
have seen.

Larry
--
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/

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ