Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.user > #199254 > unrolled thread

question about memtest86+

Back to article view | Back to linux.debian.user

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  question about memtest86+ Long Wind <longwind2@yahoo.com> - 2018-08-23 00:10 +0200
    Re: question about memtest86+ David Christensen <dpchrist@holgerdanske.com> - 2018-08-23 03:00 +0200
    Re: question about memtest86+ Ben Caradoc-Davies <ben@transient.nz> - 2018-08-23 03:30 +0200
    Re: question about memtest86+ Felix Miata <mrmazda@earthlink.net> - 2018-08-23 03:50 +0200
    Re: question about memtest86+ Jimmy Johnson <field.engineer@gmail.com> - 2018-08-23 04:20 +0200
    Re: question about memtest86+ Felix Miata <mrmazda@earthlink.net> - 2018-08-23 10:40 +0200
    Re: question about memtest86+ Long Wind <longwind2@yahoo.com> - 2018-08-23 10:40 +0200
      Re: question about memtest86+ Ben Caradoc-Davies <ben@transient.nz> - 2018-08-24 00:00 +0200
        Re: question about memtest86+ Long Wind <longwind2@yahoo.com> - 2018-08-24 09:20 +0200
          Re: question about memtest86+ Jimmy Johnson <field.engineer@gmail.com> - 2018-08-24 11:20 +0200
            Re: question about memtest86+ Long Wind <longwind2@yahoo.com> - 2018-08-24 13:00 +0200
          Re: question about memtest86+ Ben Caradoc-Davies <ben@transient.nz> - 2018-08-25 02:40 +0200
          Re: question about memtest86+ David Christensen <dpchrist@holgerdanske.com> - 2018-08-25 05:50 +0200
            Re: question about memtest86+ Long Wind <longwind2@yahoo.com> - 2018-08-25 13:30 +0200
              (solved) Re: question about memtest86+ Long Wind <longwind2@yahoo.com> - 2018-08-26 12:00 +0200
      Re: question about memtest86+ David Christensen <dpchrist@holgerdanske.com> - 2018-08-24 03:20 +0200
    Re: question about memtest86+ Joe Pfeiffer <pfeiffer@cs.nmsu.edu> - 2018-08-23 17:20 +0200

#199254 — question about memtest86+

[Multipart message — attachments visible in raw view] — view raw

i install memtest86+ of stretch to test memory
i don't see any error msg, but after a few minutes, it shutdown my PC

does that mean my memory is bad?
i read manual of memtest86+, can't find explaination

[toc] | [next] | [standalone]

#199257

On 08/22/2018 02:58 PM, Long Wind wrote:
> i install memtest86+ of stretch to test memory

Which Memtest86+ did you install on Debian Stretch?  Please provide the URL.


How did you install Memtest86+?


> i don't see any error msg, but after a few minutes, it shutdown my PC
> 
> does that mean my memory is bad?

How did you run Memtest86+?


> i read manual of memtest86+, can't find explaination


David

[toc] | [prev] | [next] | [standalone]

#199260

On 23/08/2018 09:58, Long Wind wrote:
> i install memtest86+ of stretch to test memory
> i don't see any error msg, but after a few minutes, it shutdown my PC
> does that mean my memory is bad?
> i read manual of memtest86+, can't find explaination

Uncommanded shutdowns may be caused by overheating. What is your CPU 
temperature under load? Your BIOS may shut your computer down if a 
temperature limit is exceeded. Manufacturer? Laptop? Model? CPU? BIOS 
version?

I use and recommend memtester, which can be run after a normal boot 
("userspace"). I found that multiple concurrent instances of memtester 
detected memory problems not found by memtest86+. Run memtester as root 
so it can memlock.  Because memtester runs after a normal boot, you can 
use a desktop temperature monitoring tool. I use the Xfce sensors applet 
polling at one second intervals, and this script:

#!/bin/bash
tmax=0
while true; do
     t=$((`cat /sys/class/hwmon/hwmon2/temp1_input`/1000))
     tmax=$((t>tmax?t:tmax))
     echo "`date '+%T'`: $t / $tmax"
     sleep 1
done

I also use thermald to limit CPU temperature (but my problem was not 
temperature). I use a custom service file to reduce the polling interval 
to one second.

Another useful test tool is mprime <https://www.mersenne.org/download/> 
(AKA Prime95 on Windows). Be sure to disable AVX with "CpuSupportsAVX=0" 
in local.txt because synthetic AVX tests can be used to generate 
excessive and unrealistic amounts of heat.

Kind regards,

-- 
Ben Caradoc-Davies <ben@transient.nz>
Director
Transient Software Limited <https://transient.nz/>
New Zealand

[toc] | [prev] | [next] | [standalone]

#199263

Long Wind composed on 2018-08-22 21:58 (UTC):

> i install memtest86+ of stretch to test memory
> i don't see any error msg, but after a few minutes, it shutdown my PC

> does that mean my memory is bad?
> i read manual of memtest86+, can't find explaination

On recent hardware I have much better success with memtest86:
https://www.memtest86.com/download.htm
-- 
"Wisdom is supreme; therefore get wisdom. Whatever else you
get, get wisdom." Proverbs 4:7 (New Living Translation)

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

[toc] | [prev] | [next] | [standalone]

#199265

On 08/22/2018 02:58 PM, Long Wind wrote:
> i install memtest86+ of stretch to test memory
> i don't see any error msg, but after a few minutes, it shutdown my PC
> 
> does that mean my memory is bad?
> i read manual of memtest86+, can't find explaination


Maybe system shutdown due to heat..
-- 
Jimmy Johnson

Slackware64 14.2 - Just Say No! To SystemD, Plasma5 & Drugs!
KDE 4.14.32 - AMD A8-7600 - EXT4 at sda9
Registered Linux User #380263

[toc] | [prev] | [next] | [standalone]

#199272

Long Wind composed on 2018-08-23 08:26 (UTC):

> to Felix Miata:
> memtest86 seems same as  memtest86+

It's not. 86 gets updated. 86+ does not. Last I checked, 86+ is v5.01. 86 is
7.5. 7.4 works on UEFI. 5.01 doesn't work on UEFI. 86+ has an ISO to burn to DVD
or stick, as well as install to HD or SDD. 86+ doesn't log. 86 can log, and it's
configurable.
-- 
"Wisdom is supreme; therefore get wisdom. Whatever else you
get, get wisdom." Proverbs 4:7 (New Living Translation)

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

[toc] | [prev] | [next] | [standalone]

#199273

[Multipart message — attachments visible in raw view] — view raw

to All:i've just return home, Thanks to all who reply!

to David:i use apt-get to install memtest86+ on debian 9.4 for i386
it reconfig grub and i reboot to enter memtest86+ from grub menu

to Ben Caradoc-Davies:
my PC is HP dx5150, i use it with that memory for more than 2 years, it's OK, no overheating. maybe memory checking cause overheating? i've installed memtester, i'll try it.

to Felix Miata:
memtest86 seems same as  memtest86+

to Jimmy Johnson:
i didn't see any msg about overheating
   On Thursday, August 23, 2018, 10:50:58 AM GMT+8, Timothy M Butterworth <timothy.m.butterworth@gmail.com> wrote:  
 
 On 8/22/18, Long Wind <longwind2@yahoo.com> wrote:
> i install memtest86+ of stretch to test memory
> i don't see any error msg, but after a few minutes, it shutdown my PC
>
> does that mean my memory is bad?
> i read manual of memtest86+, can't find explaination
>

You can't run memtest or fsck on a running system they can cause file
loss and corruption. You need to use a live CD. I use openSUSE Live
Rescue works well for me. I do not recall Debian's Live CD's having a
MemTest option in grub.

Tim
  

[toc] | [prev] | [next] | [standalone]

#199298

On 23/08/2018 20:26, Long Wind wrote:
> to Ben Caradoc-Davies:
> my PC is HP dx5150, i use it with that memory for more than 2 years, it's OK, no overheating. maybe memory checking cause overheating? i've installed memtester, i'll try it.

Any dx5150 Front Panel LEDs lit?
https://support.hp.com/au-en/document/c00868914

Does the CPU fan spin? When was it last cleaned? Do you have a cat?

I still suspect overheating. Sustained CPU load from a memory tester may 
be enough to put it over the edge. Sensors may help you gather evidence. 
What is your CPU model? The dx5150 has several variants and the safe 
operating temperatures may vary.

I note that this PC is about 13 years old. This exceeds the service life 
of many CPU fans. Capacitors are the next suspect (motherboard or power 
supply), but the symptoms suggest heat. Dust will exacerbate heat problems.

Kind regards,

-- 
Ben Caradoc-Davies <ben@transient.nz>
Director
Transient Software Limited <https://transient.nz/>
New Zealand

[toc] | [prev] | [next] | [standalone]

#199310

[Multipart message — attachments visible in raw view] — view raw

Sorry, Ben, i made mistake, it isn't hp dx5150, it's lenovo
motherboard is 865GV-M8, it uses Dual DDR 400 memory
i've used it for about 5 years, it's OK except sometimes it's noisy
noise must come from fans in case
since it's OK, why should i bother with memory or overheat problem?

it has 512M x 2, now i want to change one of them to 1G, so that total is 1.5G
i want to use memtester to test it, but it doesn't print progress, it seems i have to wait forever.

   On Friday, August 24, 2018, 5:51:04 AM GMT+8, Ben Caradoc-Davies <ben@transient.nz> wrote:  
 
 On 23/08/2018 20:26, Long Wind wrote:
> to Ben Caradoc-Davies:
> my PC is HP dx5150, i use it with that memory for more than 2 years, it's OK, no overheating. maybe memory checking cause overheating? i've installed memtester, i'll try it.

Any dx5150 Front Panel LEDs lit?
https://support.hp.com/au-en/document/c00868914

Does the CPU fan spin? When was it last cleaned? Do you have a cat?

I still suspect overheating. Sustained CPU load from a memory tester may 
be enough to put it over the edge. Sensors may help you gather evidence. 
What is your CPU model? The dx5150 has several variants and the safe 
operating temperatures may vary.

I note that this PC is about 13 years old. This exceeds the service life 
of many CPU fans. Capacitors are the next suspect (motherboard or power 
supply), but the symptoms suggest heat. Dust will exacerbate heat problems.

Kind regards,

-- 
Ben Caradoc-Davies <ben@transient.nz>
Director
Transient Software Limited <https://transient.nz/>
New Zealand

  

[toc] | [prev] | [next] | [standalone]

#199311

On 08/24/2018 12:05 AM, Long Wind wrote:
> 
> Sorry, Ben, i made mistake, it isn't hp dx5150, it's lenovo
> motherboard is 865GV-M8, it uses Dual DDR 400 memory
> i've used it for about 5 years, it's OK except sometimes it's noisy
> noise must come from fans in case
> since it's OK, why should i bother with memory or overheat problem?
> 
> it has 512M x 2, now i want to change one of them to 1G, so that total is 1.5G
> i want to use memtester to test it, but it doesn't print progress, it seems i have to wait forever.


You don't need to run memtest, be careful, install the new memory and 
start the computer, it does a self test, check the BIOS is the new 
memory there?

Memtest is known to cause heat to the max.
Memtest is known to cause budget chips to error.
-- 
Jimmy Johnson

Slackware64 14.2 - Just Say No! To SystemD, Plasma5 & Drugs!
KDE 4.14.32 - AMD A8-7600 - EXT4 at sda9
Registered Linux User #380263

[toc] | [prev] | [next] | [standalone]

#199313

[Multipart message — attachments visible in raw view] — view raw

Thank Jimmy Johnson for your advice!

i use xosview to monitor memory usage
maybe i shall use big program (memory hog) to test memory
but it sounds too unprofessional

   On Friday, August 24, 2018, 5:15:53 PM GMT+8, Jimmy Johnson <field.engineer@gmail.com> wrote:  
 
 On 08/24/2018 12:05 AM, Long Wind wrote:
> 
> Sorry, Ben, i made mistake, it isn't hp dx5150, it's lenovo
> motherboard is 865GV-M8, it uses Dual DDR 400 memory
> i've used it for about 5 years, it's OK except sometimes it's noisy
> noise must come from fans in case
> since it's OK, why should i bother with memory or overheat problem?
> 
> it has 512M x 2, now i want to change one of them to 1G, so that total is 1.5G
> i want to use memtester to test it, but it doesn't print progress, it seems i have to wait forever.


You don't need to run memtest, be careful, install the new memory and 
start the computer, it does a self test, check the BIOS is the new 
memory there?

Memtest is known to cause heat to the max.
Memtest is known to cause budget chips to error.
-- 
Jimmy Johnson

Slackware64 14.2 - Just Say No! To SystemD, Plasma5 & Drugs!
KDE 4.14.32 - AMD A8-7600 - EXT4 at sda9
Registered Linux User #380263

  

[toc] | [prev] | [next] | [standalone]

#199338

On 24/08/2018 19:05, Long Wind wrote:
> Sorry, Ben, i made mistake, it isn't hp dx5150, it's lenovo
> motherboard is 865GV-M8, it uses Dual DDR 400 memory
> i've used it for about 5 years, it's OK except sometimes it's noisy
> noise must come from fans in case

Fans are a common source of case noise. Hard disk drives can be noisy. 
Electrical components such as coils or capacitors can also whine.

> since it's OK, why should i bother with memory or overheat problem?

Overheating can cause instability and shorten the life of your 
components. Cleaning the system and cleaning, replacing, or adding fans 
is an inexpensive and effective way of reducing overheating. Filters can 
also be clogged with dust or hair.

Defective memory will cause system instability and silent corruption. I 
would much rather have 1G of reliable memory than 1.5G of unreliable 
memory. Memory tests will tell you whether your 1.5G is reliable.

If your computer turns itself off, there is something wrong.

> it has 512M x 2, now i want to change one of them to 1G, so that total is 1.5G
> i want to use memtester to test it, but it doesn't print progress, it seems i have to wait forever.

If you do not specify a number of iterations, memtester will loop 
infinitely. When I run a single loop, I see progress reports, which 
change to "ok" for each successful test:

# memtester 1G 1
memtester version 4.3.0 (64-bit)
Copyright (C) 2001-2012 Charles Cazabon.
Licensed under the GNU General Public License version 2 (only).

pagesize is 4096
pagesizemask is 0xfffffffffffff000
want 1024MB (1073741824 bytes)
got  1024MB (1073741824 bytes), trying mlock ...locked.
Loop 1/1:
   Stuck Address       : ok
   Random Value        : ok
   Compare XOR         : ok
   Compare SUB         : ok
   Compare MUL         : ok
   Compare DIV         : ok
   Compare OR          : ok
   Compare AND         : ok
   Sequential Increment: ok
   Solid Bits          : ok
   Block Sequential    : ok
   Checkerboard        : ok
   Bit Spread          : ok
   Bit Flip            : ok
   Walking Ones        : ok
   Walking Zeroes      : ok
   8-bit Writes        : ok
   16-bit Writes       : ok

Done.

This test takes 3 minutes on my i7 7700 with dual-channel DDR4 RAM. Peak 
package temperature was 72 C. Ambient temperature is 19 C. I have a 
passively cooled Streacom mini-ITX case.

Kind regards,

-- 
Ben Caradoc-Davies <ben@transient.nz>
Director
Transient Software Limited <https://transient.nz/>
New Zealand

[toc] | [prev] | [next] | [standalone]

#199344

On 08/24/2018 12:05 AM, Long Wind wrote:
> Sorry, Ben, i made mistake, it isn't hp dx5150, it's lenovo
> motherboard is 865GV-M8, 

I'm having a hard time STFW for that motherboard.  There are several 
motherboards that use the 865GV chipset, both socket 478 and LGA 775. 
Please provide a URL to the exact motherboard.


My two oldest machines have Intel D865GBFLK motherboards with socket 478 
Pentium 4 processors (2.8E GHz and 3.4E GHz, respectively).  They were 
getting hot under load with the stock Intel CPU coolers, so I switched 
to Zalman high-speed ducted fan CPU coolers.  The processors run cooler 
now, but the fans are even noisier.  The computers have always generated 
a fair amount of heat.  I do not recall any thermal shutdown events. 
The place to look for such would be the BIOS event log.


> it uses Dual DDR 400 memory
> i've used it for about 5 years, it's OK except sometimes it's noisy
> noise must come from fans in case

You probably have an LGA 775 model with a BIOS-controlled variable speed 
CPU fan that speeds up when the CPU is under load.


> since it's OK, 

Your computer is *not* OK if it shuts down during memory testing.


> why should i bother with memory or overheat problem?

Because if you request help from people and then refuse to act upon 
reasonable advice, you are unlikely to solve your problem and they are 
less likely to advise you in the future.


> it has 512M x 2, now i want to change one of them to 1G, so that total is 1.5G

That will defeat dual-channel mode.  You want a matched pair of the 
biggest, fastest, and lowest latency memory modules that are recommended 
for that computer/ motherboard by the memory manufacturer.


That said, I hesitate before sinking money into obsolete computers.  One 
of my old P4's has the original 2 @ 512 MB and 2 @ 256 MB modules from 
both computers; the other has 2 @ 1 GB upgrades from ~10 years ago.


> i want to use memtester to test it, but it doesn't print progress, it seems i have to wait forever.

RTFM memtester(8).  Tell it to use less memory and/or perform fewer 
iterations:

     https://linux.die.net/man/8/memtester


On 08/24/2018 03:51 AM, Long Wind wrote:
 > i use xosview to monitor memory usage
 > maybe i shall use big program (memory hog) to test memory
 > but it sounds too unprofessional

It is professional to use the right tool for the job.


All of the above said, the best way to avoid memory problems is to have 
computers with ECC memory.


David

[toc] | [prev] | [next] | [standalone]

#199358

[Multipart message — attachments visible in raw view] — view raw

Thank Ben and David!

fixing fan problem takes time and energy. as learner, i'm prone to make mistakes. i'm in poor health.  i'd rather buy new one if it break down

actually i have several old PCs, lenovo and hp dx5150 among other more dated ones. dx5150 has 2G memory and it pass memtest. if lenovo doesn't work, i can use hp

i bought lenovo and hp from 2nd-hand dealer, their price is about the same:200 yuan (one USD is about 6.9 yuan now) the price doesn't include monitor/keyboard/mouse. because they are cheap, i'll buy new one instead of fixing them if they don't work

from my point of view lenovo is OK in past five years, the only  problem is it doesn't pass memtest86+
my policy is don't fix them unless they don't work


   On Saturday, August 25, 2018, 11:43:51 AM GMT+8, David Christensen <dpchrist@holgerdanske.com> wrote:  
 
 On 08/24/2018 12:05 AM, Long Wind wrote:
> Sorry, Ben, i made mistake, it isn't hp dx5150, it's lenovo
> motherboard is 865GV-M8, 

I'm having a hard time STFW for that motherboard.  There are several 
motherboards that use the 865GV chipset, both socket 478 and LGA 775. 
Please provide a URL to the exact motherboard.


My two oldest machines have Intel D865GBFLK motherboards with socket 478 
Pentium 4 processors (2.8E GHz and 3.4E GHz, respectively).  They were 
getting hot under load with the stock Intel CPU coolers, so I switched 
to Zalman high-speed ducted fan CPU coolers.  The processors run cooler 
now, but the fans are even noisier.  The computers have always generated 
a fair amount of heat.  I do not recall any thermal shutdown events. 
The place to look for such would be the BIOS event log.


> it uses Dual DDR 400 memory
> i've used it for about 5 years, it's OK except sometimes it's noisy
> noise must come from fans in case

You probably have an LGA 775 model with a BIOS-controlled variable speed 
CPU fan that speeds up when the CPU is under load.


> since it's OK, 

Your computer is *not* OK if it shuts down during memory testing.


> why should i bother with memory or overheat problem?

Because if you request help from people and then refuse to act upon 
reasonable advice, you are unlikely to solve your problem and they are 
less likely to advise you in the future.


> it has 512M x 2, now i want to change one of them to 1G, so that total is 1.5G

That will defeat dual-channel mode.  You want a matched pair of the 
biggest, fastest, and lowest latency memory modules that are recommended 
for that computer/ motherboard by the memory manufacturer.


That said, I hesitate before sinking money into obsolete computers.  One 
of my old P4's has the original 2 @ 512 MB and 2 @ 256 MB modules from 
both computers; the other has 2 @ 1 GB upgrades from ~10 years ago.


> i want to use memtester to test it, but it doesn't print progress, it seems i have to wait forever.

RTFM memtester(8).  Tell it to use less memory and/or perform fewer 
iterations:

    https://linux.die.net/man/8/memtester


On 08/24/2018 03:51 AM, Long Wind wrote:
 > i use xosview to monitor memory usage
 > maybe i shall use big program (memory hog) to test memory
 > but it sounds too unprofessional

It is professional to use the right tool for the job.


All of the above said, the best way to avoid memory problems is to have 
computers with ECC memory.


David

  

[toc] | [prev] | [next] | [standalone]

#199382 — (solved) Re: question about memtest86+

[Multipart message — attachments visible in raw view] — view raw

i run memtest86+ again, it pass, i mean it reach Pass 100%i'm not sure of reason, maybe room temperature is a little low
and i bought 1G memory at price 16.5 yuan, upgrade it to 1.5G, and test it, it reach  Pass 100%
memtest86+ shows that my memory is ddr2 399, i think it's ddr 400, maybe my hardware is too old
Thanks again to all who reply!below are some kernel msg that says cpu too hot:
Aug 26 16:59:16 debian kernel: [    0.096000] NMI watchdog: enabled on all CPUs, permanently consumes one hw-PMU counter.
Aug 26 16:59:16 debian kernel: [    0.096000] CPU 1 irqstacks, hard=f7550000 soft=f7552000
Aug 26 16:59:16 debian kernel: [    0.096000] x86: Booting SMP configuration:
Aug 26 16:59:16 debian kernel: [    0.096000] .... node  #0, CPUs:      #1
Aug 26 16:59:16 debian kernel: [    0.012000] Initializing CPU#1
Aug 26 16:59:16 debian kernel: [    0.012000] CPU1: Core temperature above threshold, cpu clock throttled (total events = 1)
Aug 26 16:59:16 debian kernel: [    0.012000] mce: [Hardware Error]: Machine check events logged
Aug 26 16:59:16 debian kernel: [    0.186100] x86: Booted up 1 node, 2 CPUs
Aug 26 16:59:16 debian kernel: [    0.186100] smpboot: Total of 2 processors activated (11182.99 BogoMIPS)
Aug 26 16:59:16 debian kernel: [    0.188862] devtmpfs: initialized
Aug 26 16:59:16 debian kernel: [    0.189099] PM: Registering ACPI NVS region [mem 0x5f7f0000-0x5f7f2fff] (12288 bytes)
Aug 26 16:59:16 debian kernel: [    0.189099] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
Aug 26 16:59:16 debian kernel: [    0.189099] futex hash table entries: 1024 (order: 4, 65536 bytes)
Aug 26 16:59:16 debian kernel: [    0.189099] CPU1: Core temperature/speed normal
Aug 26 16:59:16 debian kernel: [    0.189099] pinctrl core: initialized pinctrl subsystem
Aug 26 16:59:16 debian kernel: [    0.189099] mce: [Hardware Error]: Machine check events logged
Aug 26 16:59:16 debian kernel: [    0.189099] NET: Registered protocol family 16
Aug 26 16:59:16 debian kernel: [    0.200008] cpuidle: using governor ladder
Aug 26 16:59:16 debian kernel: [    0.212007] cpuidle: using governor menu
Aug 26 16:59:16 debian kernel: [    0.212070] ACPI: bus type PCI registered
Aug 26 16:59:16 debian kernel: [    0.212074] acpiphp: ACPI Hot Plug PCI Controller Driver version: 0.5
 

    On Saturday, August 25, 2018 7:24 PM, Long Wind <longwind2@yahoo.com> wrote:
 

 Thank Ben and David!

fixing fan problem takes time and energy. as learner, i'm prone to make mistakes. i'm in poor health.  i'd rather buy new one if it break down

actually i have several old PCs, lenovo and hp dx5150 among other more dated ones. dx5150 has 2G memory and it pass memtest. if lenovo doesn't work, i can use hp

i bought lenovo and hp from 2nd-hand dealer, their price is about the same:200 yuan (one USD is about 6.9 yuan now) the price doesn't include monitor/keyboard/mouse. because they are cheap, i'll buy new one instead of fixing them if they don't work

from my point of view lenovo is OK in past five years, the only  problem is it doesn't pass memtest86+
my policy is don't fix them unless they don't work


   On Saturday, August 25, 2018, 11:43:51 AM GMT+8, David Christensen <dpchrist@holgerdanske.com> wrote:  
 
 On 08/24/2018 12:05 AM, Long Wind wrote:
> Sorry, Ben, i made mistake, it isn't hp dx5150, it's lenovo
> motherboard is 865GV-M8, 

I'm having a hard time STFW for that motherboard.  There are several 
motherboards that use the 865GV chipset, both socket 478 and LGA 775. 
Please provide a URL to the exact motherboard.


My two oldest machines have Intel D865GBFLK motherboards with socket 478 
Pentium 4 processors (2.8E GHz and 3.4E GHz, respectively).  They were 
getting hot under load with the stock Intel CPU coolers, so I switched 
to Zalman high-speed ducted fan CPU coolers.  The processors run cooler 
now, but the fans are even noisier.  The computers have always generated 
a fair amount of heat.  I do not recall any thermal shutdown events. 
The place to look for such would be the BIOS event log.


> it uses Dual DDR 400 memory
> i've used it for about 5 years, it's OK except sometimes it's noisy
> noise must come from fans in case

You probably have an LGA 775 model with a BIOS-controlled variable speed 
CPU fan that speeds up when the CPU is under load.


> since it's OK, 

Your computer is *not* OK if it shuts down during memory testing.


> why should i bother with memory or overheat problem?

Because if you request help from people and then refuse to act upon 
reasonable advice, you are unlikely to solve your problem and they are 
less likely to advise you in the future.


> it has 512M x 2, now i want to change one of them to 1G, so that total is 1.5G

That will defeat dual-channel mode.  You want a matched pair of the 
biggest, fastest, and lowest latency memory modules that are recommended 
for that computer/ motherboard by the memory manufacturer.


That said, I hesitate before sinking money into obsolete computers.  One 
of my old P4's has the original 2 @ 512 MB and 2 @ 256 MB modules from 
both computers; the other has 2 @ 1 GB upgrades from ~10 years ago.


> i want to use memtester to test it, but it doesn't print progress, it seems i have to wait forever.

RTFM memtester(8).  Tell it to use less memory and/or perform fewer 
iterations:

    https://linux.die.net/man/8/memtester


On 08/24/2018 03:51 AM, Long Wind wrote:
 > i use xosview to monitor memory usage
 > maybe i shall use big program (memory hog) to test memory
 > but it sounds too unprofessional

It is professional to use the right tool for the job.


All of the above said, the best way to avoid memory problems is to have 
computers with ECC memory.


David

  

   

[toc] | [prev] | [next] | [standalone]

#199303

On 08/23/2018 01:26 AM, Long Wind wrote:
> to All:i've just return home, Thanks to all who reply!
> 
> to David:i use apt-get to install memtest86+ on debian 9.4 for i386
> it reconfig grub and i reboot to enter memtest86+ from grub menu

Thank you for making me aware of memtest86+ integration into Debian 
(I've always used the bootable ISO CD image):

2018-08-23 17:32:15 dpchrist@vstretch ~
$ apt-cache search memtest
memtest86 - thorough real-mode memory tester
memtest86+ - thorough real-mode memory tester
memtester - Utility for testing the memory subsystem


David

[toc] | [prev] | [next] | [standalone]

#199283

Long Wind <longwind2@yahoo.com> writes:

> i install memtest86+ of stretch to test memory
> i don't see any error msg, but after a few minutes, it shutdown my PC
>
> does that mean my memory is bad?
> i read manual of memtest86+, can't find explaination

I see other people have mentioned possible temperature issues, so I'll
mention something that may or may not be relevant:  a BIOS update pushed
by Dell to my laptop enabled a CPU "turbo boost" mode that overclocked
the CPU.  Under any sort of load (and memtest86+ is certainly a load!)
it would go into thermal shutdown in a couple of minutes, without having
logged any sort of messages about heat.  Disabling the mode in the BIOS
gave me a stable system again.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.debian.user

csiph-web