I'm setting up a new rig, with 4 X 4GB DDR2 modules.
I tried run memtest86 but it only detected 4GB of my memory.
In Linux and in XP-64 I see all 16GB of the RAM.
So how to test the RAM?
Anyone knows?
Please help. Thanks !
Duh......one stick at a time?
Sys_basher will test most of it. Memtest86 runs under DOS so it's a 32
bit program, sys_basher runs under Linux so it can see all of the RAM.
However sys_basher can't identify which DIMM is bad, only that you have a
bad DIMM. The problems is that Linux lacks a call which will translate a
logical address to a physical address, if anyone knows how to do this I'd
appreciate if they would post the instructions for how to do this.
Thanks for the info !
I will try it out. :)
Memtest86 is the old version. Try memtest86+.
The latest version is 2.11, and it can handle 64
bit addressing. It sees all of my 6GB.
Available here: www.memtest.org
General Schvantzkoph a écrit :
>
> Memtest86 runs under DOS so it's a 32 bit program
Nonsense. Memtest86 does not run under DOS.
If it ran under DOS it would be a 16-bit program.
> On Wed, 07 Jan 2009 23:36:35 -0800, pg wrote:
>
>> Hello.
>>
>> I'm setting up a new rig, with 4 X 4GB DDR2 modules.
>>
>> I tried run memtest86 but it only detected 4GB of my memory.
>>
>> In Linux and in XP-64 I see all 16GB of the RAM.
>>
>> So how to test the RAM?
>>
>> Anyone knows?
>>
>> Please help. Thanks !
>
> Sys_basher will test most of it. Memtest86 runs under DOS so it's a 32
> bit program,
memtest doesn't us an OS, it runs on the bare hardware. I believe
you're correct that it's only available as a 32 bit program (and
apparently doesn't know about PAE), however.
No it isn't.
There's the commercial version at memtest86.com and the open source
version at memtest86.org
> Try memtest86+.
> The latest version is 2.11, and it can handle 64
> bit addressing. It sees all of my 6GB.
>
> Available here: www.memtest.org
>
>
--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/
General Schvantzkoph <schvan...@yahoo.com> writes:
> On Wed, 07 Jan 2009 23:36:35 -0800, pg wrote:
>
>> Hello.
>>
>> I'm setting up a new rig, with 4 X 4GB DDR2 modules.
>>
>> I tried run memtest86 but it only detected 4GB of my memory.
>>
>> In Linux and in XP-64 I see all 16GB of the RAM.
>>
>> So how to test the RAM?
>>
>> Anyone knows?
>>
>> Please help. Thanks !
>
> Sys_basher will test most of it. Memtest86 runs under DOS so it's a 32
> bit program,
Note that memtest86 doesn't run under any OS, it runs on the bare
hardware. Someone else posted that memtest86+ will exercise all the
memory.
Easiest simplest way - 4gb at a time.
> Nonsense. Memtest86 does not run under DOS.
> If it ran under DOS it would be a 16-bit program.
And what *type* operating system (DOS is a type of OS, not a particular
one) do you think this *application* runs under? The program is NOT
written using machine or assembly code (but the boot sector loader was).
The bootstrap code in your BIOS is not going to load this application
directly into memory and transfer control to it.
"The initial versions of the source files bootsect.S, setup.S, head.S
and build.c are from the Linux 1.2.1 kernel and have been heavily
modified" (http://www.memtest86.com/about.html). The boot sector loader
(that your BIOS bootstrap code loads and to which it transfers control)
was written in assembly. That loads the C-written LinuxBIOS program
(http://en.wikipedia.org/wiki/LinuxBIOS) which loads the Linux kernel
which is the DOS that manages the C-written memtest application.
Memtest86 DOES run under an DOS (disk operating system). It's Linux.
> Pascal Hambourg wrote:
>
>> Nonsense. Memtest86 does not run under DOS.
>> If it ran under DOS it would be a 16-bit program.
>
> And what *type* operating system (DOS is a type of OS, not a particular
> one) do you think this *application* runs under? The program is NOT
> written using machine or assembly code (but the boot sector loader was).
> The bootstrap code in your BIOS is not going to load this application
> directly into memory and transfer control to it.
The rest of us long ago quit insisting that DOS was generic and
recognized that when somebody says DOS they mean MS-DOS or a clone.
> "The initial versions of the source files bootsect.S, setup.S, head.S
> and build.c are from the Linux 1.2.1 kernel and have been heavily
> modified" (http://www.memtest86.com/about.html). The boot sector loader
> (that your BIOS bootstrap code loads and to which it transfers control)
> was written in assembly. That loads the C-written LinuxBIOS program
> (http://en.wikipedia.org/wiki/LinuxBIOS) which loads the Linux kernel
> which is the DOS that manages the C-written memtest application.
>
> Memtest86 DOES run under an DOS (disk operating system). It's Linux.
The README file makes it pretty clear that a few files started out as
parts of a very old version of Linux; beyond that, it's a standalone
program.
Where did you find documentation claiming that it loads a Linux
kernel, or that memtest is an application running under it?
> VanguardLH <V...@nguard.LH> writes:
>
> > Memtest86 DOES run under an DOS (disk operating system). It's
> > Linux.
>
> The README file makes it pretty clear that a few files started out as
> parts of a very old version of Linux; beyond that, it's a standalone
> program.
>
> Where did you find documentation claiming that it loads a Linux
> kernel, or that memtest is an application running under it?
From README.build-process:
memtest.bin is the floppy/lilo bootable target.
memtest is the ELF bootable target.
And,
$ file memtest.bin
memtest.bin: Linux x86 kernel
which says memtest is a bootable target in ELF format. Not that it's
a linux kernel.
> And,
>
> $ file memtest.bin
> memtest.bin: Linux x86 kernel
which, of course, is simply from the fact that the startup code comes
from an antique version of linux's startup code.
I think I hit a brick wall at www.memtest86.org
All I got is a very generic domain squatter type of frontpage, which I
am listing below :
= = = = ==============================
Welcome to memtest86.org
For resources and information on Memory test and Memtest86 Download
Related Searches
Memory test
Memtest86 Download
Free Memory Test
Test My Ram
Memory diagnostic
How To Run Memtest86
Memory test software
Testing Computer Memory
Memory Testers Freeware
How Long Does Memtest86 Take
Memory Tester
PC diagnostic software
Check My Processor Speed
Cpu-z
Docmem
Sponsored Listings for MEMTEST86 DOWNLOAD
= = = = ==============================
>
> I think I hit a brick wall at www.memtest86.org
>
--
May's Law:
The quality of correlation is inversly proportional to the density
of control. (The fewer the data points, the smoother the curves.)
----------------------
Mandriva Linux release 2009.0 (Official) for x86_64
2.6.27.7-desktop-1mnb AMD Athlon(tm) 64 X2 Dual Core Processor 5000+
----------------------
http://www.memtest.org is the home page for Memtest86+...which I believe
the OP was after. :-)
Klick free download, voila!!!!
Memtest86+ v2.11 is more up to date than the product at
memtest86.com. Memtest86.com bases their CPU/chipset info
on memtest86+ v1.70, which I also have, but it doesn't correctly
detect the Core i7 CPU and X58 chipset. It says the i7 is a PIII.
It does detect the correct speed and my 6GB of DDR3. V2.11
corrects all that and detects the triple channel.
Memtest86.com may have put a few more bells and whistles in
their pay product.
www.memtest.org is not anything worth seeing. It is owned by some search
engine.
Try any of the links from that page.
--
I like being single. I'm always there when I need me.
-- Art Leo
> On Thu, 08 Jan 2009 18:16:45 -0800, propman wrote:
>
>> Jim Whitby wrote:
>>> On Thu, 08 Jan 2009 17:21:24 -0800, propman wrote:
>>>
>>>> pg wrote:
>>>>
>>>>
>>>>> I think I hit a brick wall at www.memtest86.org
>>>>>
>>>>>
>>>> http://www.memtest.org/
>>>
>>> try http://www.memtest.COM
>>>
>>>
>>>
>>>
>>>
>> http://www.memtest.org is the home page for Memtest86+...which I
>> believe the OP was after. :-)
>
> www.memtest.org is not anything worth seeing. It is owned by some search
> engine.
>
> Try any of the links from that page.
I stand corrected!!! memtest.org is not the same as memtest86.org.
Pardon the old man for being stupid and not being able to read.
propman <whose little grey cells ain't what they used to be>
> VanguardLH <V...@nguard.LH> writes:
>
>> Pascal Hambourg wrote:
>>
>>> Nonsense. Memtest86 does not run under DOS.
>>> If it ran under DOS it would be a 16-bit program.
>>
>> And what *type* operating system (DOS is a type of OS, not a particular
>> one) do you think this *application* runs under? The program is NOT
>> written using machine or assembly code (but the boot sector loader was).
>> The bootstrap code in your BIOS is not going to load this application
>> directly into memory and transfer control to it.
>
> The rest of us long ago quit insisting that DOS was generic and
> recognized that when somebody says DOS they mean MS-DOS or a clone.
MS-DOS
PC-DOS
FreeDOS
CalderaDOS
DOS Plus
DR-DOS
PTS-DOS
OpenDOS
AmigaDOS
AMSDOS
Apple DOS
Atari DOS
Commodore DOS
TRS-DOS
So it depends on how far you go back or at what point you take a
snapshot of the word "DOS" as representing whatever was in most use at
that snapshot's time frame.
>
>> "The initial versions of the source files bootsect.S, setup.S, head.S
>> and build.c are from the Linux 1.2.1 kernel and have been heavily
>> modified" (http://www.memtest86.com/about.html). The boot sector loader
>> (that your BIOS bootstrap code loads and to which it transfers control)
>> was written in assembly. That loads the C-written LinuxBIOS program
>> (http://en.wikipedia.org/wiki/LinuxBIOS) which loads the Linux kernel
>> which is the DOS that manages the C-written memtest application.
>>
>> Memtest86 DOES run under an DOS (disk operating system). It's Linux.
>
> The README file makes it pretty clear that a few files started out as
> parts of a very old version of Linux; beyond that, it's a standalone
> program.
Yep, just like edlin.com, notepad.exe, and winword.exe are standalone
programs - that run atop a DOS.
> Where did you find documentation claiming that it loads a Linux
> kernel, or that memtest is an application running under it?
Try the links that I provided. The first one is from the memtest site
itself. Download the source code and enjoy looking at bootsec, head,
and linuxbios.c. The 2nd link I got after seeing in the code that
memtest uses coreboot (aka linuxbios).
Damn kids, get off my lawn!
Jon
> Ron May's Law:
> The quality of correlation is inversly proportional to the density
> of of my head. Da da do deee dumb
--
"If Milbank aka hummingHomo and Me.Whore stand by Bear Bottoms,
the idiot who outed himself as William "Bear" Bottoms, the baby killer,
then they are no better than baby killers themselves." Ari ~1Jan09
The Ubuntu 64-bit live-cd comes with a 64-bit version of Memtest. I'm
pretty sure most Linux distros' live-cd does this also.
Yousuf Khan
If it gives the failing address, can't you just divide by 4 GiB (in this
case) to get the bad stick?
--
-eben QebWe...@vTerYizUonI.nOetP
1101000 1110100 1110100 1110000 0111010 0101111 0101111 1110010 1101111
1111001 1100001 1101100 1110100 1111001 0101110 1101101 1101001 1101110
1100101 0101110 1101110 1110101 0111010 0111000 0110001 0101111
> So it depends on how far you go back or at what point you take a
> snapshot of the word "DOS" as representing whatever was in most use at
> that snapshot's time frame.
DOS the point, isn't it?
HAHAHA
I crack me up.
--
Social ecology encompasses the fabric of societal structures,
including peace and human rights, dignity and democracy, employment
and social integration, security and safety." ~ Walter Stahel
> On Thu, 8 Jan 2009 23:10:09 -0600, VanguardLH wrote:
>
>> So it depends on how far you go back or at what point you take a
>> snapshot of the word "DOS" as representing whatever was in most use at
>> that snapshot's time frame.
>
> DOS the point, isn't it?
>
> HAHAHA
>
> I crack me up.
--
Peter.
You don't understand Newton's Third Law of Motion?
It's not rocket science, you know.
Don't test it, run it.
--
<<//--------------------\\>>
Van Chocstraw
>>\\--------------------//<<
Then when it crashes, how do you know what to fix?
Of course, you can hope it works and run to failure.
The commercial model of PC production.
I prefer to have some justifiable confidence in my machines by
testing upfront. I get parts from all over the place. While they
are to-standards and good quality, no-one can say they'll interact
perfectly. Tolerences could line up just the wrong way.
So I'm a big advocate of testing, especially on homebuilts. Stability
and longevity are the rewards. My main machine is almost 10 years
old and I cannot remember the last crash apart from powerfails.
-- Robert author `cpuburn` http://pages.sbcglobal.net/redelm
pg wrote:
> I'm setting up a new rig, with 4 X 4GB DDR2 modules.
>
> I tried run memtest86 but it only detected 4GB of my memory.
>
> In Linux and in XP-64 I see all 16GB of the RAM.
>
> So how to test the RAM?
Memtest+, from www.memtest.org, will test it all, but so should the
newest ver. of Memtest86, ver. 3.4A. I've used the latter to test
systems equipped with 6GB and 8GB.
This only works if:
1) The problem occurs often / quickly enough; and
2) You are sure that memory is the cause.
Also, absence of proof isn't proof of absence.
Many reliability problems don't occur often, or at least don't
produce visible consequences frequently. Testing is almost alwyas
more intense than regular usage and monitors continually for errors.
The cause is similarly uncertain. When you have a crash, how
can you possibly know what caused it? If it reproduces well,
then it's probably software. But if not, there still are software
interactions, and all the different peices of equipment (HD, PSU,
...) that could cause it. Testing eliminates them one-by-one.
-- Robert
Van Chocstraw wrote:
> You can test it by removing one strip at a time and trying it if you are
> having problems.
I've had modules work fine when run alone but always failed when run
with another module, both in single-channel and dual-channel modes.
This includes two OCZ Gold matched pairs in a row.