RE: pmap bomb on 4.0-STABLE

2001-04-05 Thread Koster, K.J.

Dear Mahlon,

 
 Can anyone tell me what this means - and even better, a fix?  It's
 my understanding that pmap concerns shared memory, is it possible
 I have a bad stick of ram floating around?
 
Bad memory sticks are easy to find. Just rip out half the RAM and let the
box run for a few days, then let it run off the other half for a while.

Check case cooling too.

Kees Jan

PS. Please notice that crossposting is generally frowned upon. People tend
to ignore you if you do this.


 You are only young once,
   but you can stay immature all your life.

To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message



Re: pmap bomb on 4.0-STABLE

2001-04-05 Thread Mahlon Smith

  Can anyone tell me what this means - and even better, a fix?  It's
  my understanding that pmap concerns shared memory, is it possible
  I have a bad stick of ram floating around?
  
 Bad memory sticks are easy to find. Just rip out half the RAM and let the
 box run for a few days, then let it run off the other half for a while.


Of course, this is pretty far from scientific troubleshooting, especially
when it crashes at random times.  It's also highly undesirable to cripple
the machine, considering it's a production box.

I just need to know if pmap_entry really does have anything to do with physical
ram, before I go off on a ram swapping goose chase, just to find out a
month down the road the problem isn't fixed.

--
Mahlon Smith
InternetCDS
http://www.internetcds.com

To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message



Re: pmap bomb on 4.0-STABLE

2001-04-05 Thread Joseph Gleason

A friend of mine swears by this memory testing utility:

http://reality.sgi.com/cbrady_denver/memtest86/

Apparently it tries a bunch of diffrent test patters that are likely to find
memory problems that a simple test wouldn't find.  It is cool because you
just just write the image to a 1.44mb floppy and boot from that to do the
test.

A major downside is how long it takes.  It takes around 8 hours on my laptop
to do the full suite of tests.  Not very useful for a production server..but
something that probably be done on every system you create before you move
it to producton.

Joe Gleason

- Original Message -
From: "Mahlon Smith" [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Sent: Thursday, April 05, 2001 12:10
Subject: Re: pmap bomb on 4.0-STABLE


   Can anyone tell me what this means - and even better, a fix?  It's
   my understanding that pmap concerns shared memory, is it possible
   I have a bad stick of ram floating around?
  
  Bad memory sticks are easy to find. Just rip out half the RAM and let
the
  box run for a few days, then let it run off the other half for a while.


 Of course, this is pretty far from scientific troubleshooting, especially
 when it crashes at random times.  It's also highly undesirable to cripple
 the machine, considering it's a production box.

 I just need to know if pmap_entry really does have anything to do with
physical
 ram, before I go off on a ram swapping goose chase, just to find out a
 month down the road the problem isn't fixed.

 --
 Mahlon Smith
 InternetCDS
 http://www.internetcds.com

 To Unsubscribe: send mail to [EMAIL PROTECTED]
 with "unsubscribe freebsd-hackers" in the body of the message



To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message



Re: pmap bomb on 4.0-STABLE

2001-04-05 Thread David Scheidt

On Thu, 5 Apr 2001, Joseph Gleason wrote:

:A friend of mine swears by this memory testing utility:
:
:http://reality.sgi.com/cbrady_denver/memtest86/
:
:Apparently it tries a bunch of diffrent test patters that are likely to find
:memory problems that a simple test wouldn't find.  It is cool because you
:just just write the image to a 1.44mb floppy and boot from that to do the
:test.
:
:A major downside is how long it takes.  It takes around 8 hours on my laptop

The major downside is that software memory testing isn't conclusive.  If it
detects a problem, you've probably got one.  If it doesn't, you may still.  
The only way to be sure is to use a hardware tester.  If you don't have one,
and most of us don't, you have to resort to swapping memory.  
:

-- 
[EMAIL PROTECTED]
Bipedalism is only a fad.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message