Linux Kernel Commit 025cee7f8fef02af09b03c8e1cd9843cb32adf9b

Linux Kernel Commit 025cee7f8fef02af09b03c8e1cd9843cb32adf9b

[John L. Males]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Dave,

With respect to Linux Kernel Commit
025cee7f8fef02af09b03c8e1cd9843cb32adf9b Change Log:

<http://www.kernel.org/pub/linux/kernel/v3.0/ChangeLog-3.4.34>

comment "It has probably been the cause of a number of subtle
bugs over the years, although the conditions to excite them
would have been hard to trigger."

Would the Linux Kernel under "unique" Virtual Memory Subsystem
stress excite some of these "subtle bugs" in Linux Kernels
prior to this change set for Commit
025cee7f8fef02af09b03c8e1cd9843cb32adf9b?  I am not asking for
you to identify the subtle bugs.  It is obvious it would be
difficult to determine from reported and unreported cases such
bugs. Just your sense or first hand knowledge if a Linux Kernel
under "unique" stress including Virtual Memory Subsystem stress
would be a key element of?

I am one of those who seems to often cause system kernels of
any OS stress due to my very high power user use of Operating
Systems and hence why my question.


Regards,

John L. Males
Toronto, Ontario
Canada
28 February 2013 17:11
<mailto:jlmales@gmail.com>


==============================================================
2013-02-28 16:57:24.110757989-0500-EST

28 Feb 16:57:24 ntpdate[27847]: ntpdate 4.2.6p2@1.2194-o Sun
Oct 17 13:35:14 UTC 2010 (1)

28 Feb 16:58:59 ntpdate[27852]: step time server 132.246.11.228
offset -0.000138 sec

Linux 3.4.24-kernel.org-jlm-010-amd64 #1 SMP PREEMPT Sun Dec 23
10:06:41 EST 2012

Modified Debian GNU/Linux 6.0.3 (squeeze)
(Evaluating alternatives to Debian)

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iEYEARECAAYFAlEv1h8ACgkQ+V/XUtB6aBDSzQCgiu+hskZzz2bMfLG5u+Ao9YzJ
hwIAn1IY5jnq0sJjIe0nxFnA+LKrGtAh
=1KDN
-----END PGP SIGNATURE-----

Re: Linux Kernel Commit 025cee7f8fef02af09b03c8e1cd9843cb32adf9b

[Dave Hansen]

On 02/28/2013 02:11 PM, John L. Males wrote:
> Would the Linux Kernel under "unique" Virtual Memory Subsystem
> stress excite some of these "subtle bugs" in Linux Kernels
> prior to this change set for Commit
> 025cee7f8fef02af09b03c8e1cd9843cb32adf9b?

I don't think stress would really do it.

It also, as far as I know, only applies to 32-bit NUMA kernels where
CONFIG_NUMA was turned on _and_ where the machine actually was NUMA.