[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: memtest kernel option



On Sun, Dec 04, 2011 at 01:27:57PM +0100, Thierry GRAUSS wrote:
> Hello,
> 
> I have a laptop computer with half of the RAM soldered on the motherboard.
> This soldered RAM has a huge amount of bad RAM (too many errors to be able
> to use the badram option).
> 
> Fortunatly, there is a Linux kernel option (available in the mainstream
> kernel, no need to patch anything) which can run tests like memtest86 when
> Linux is booting so that it will automatically exclude these parts of the
> memory from being used.
> 
> This option, when enabled in the kernel, can be enabled and disabled with
> a kernel boot argument (option memtest in grub).
> 
> Is it possible to include this option in the default kernel, even if
> disabled by default, but having the option enabled in the installation CD
> (at least to be able to install the system without trouble)?
> I managed to install Debian Squeeze, but it was really a pain with this
> bad RAM.
> 
> Performance benchmarks show also no impact at all from the memtest option.
> 
> This option is not only useful for laptop computer  like mine, but also
> for the new "plug" servers based on ARM processors. Often they have only
> soldered RAM chip. As reliability is even more important for a server than
> for a laptop computer, I think that the activation of this option whould
> help a lot to prevent to trash these little computers when/if they have
> bad RAM modules.
> 
> - Origin of the module, before it was included in the mainstream branch of
> the kernel : http://rick.vanrein.org/linux/badram/
> - Benchmark of the memtest module :
> http://rick.vanrein.org/linux/badram/results.html
> - Example of use and explanation of the usefulness of the option when
> running a server :
> http://onlyjob.blogspot.com/2011/01/memtest-explained-linux-kernel.html
> 
> Thank you in advance

>From http://lists.debian.org/debian-kernel/2011/12/msg00333.html do I
understand that MEMTEST is now default in new kernels.


> Best regards
> Thierry GRAUSS


Groeten
Geert Stappers
-- 
> And is there a policy on top-posting vs. bottom-posting?
Yes.


Reply to: