need help with memtest-86 v3.0

"Bob Davis" <iclick.DeleteThis@coxNIX.net> wrote in message news:<UweUb.9881$gl2.5325@lakeread05>...
 > "jerry" <jturba.DeleteThis@yahoo.com> wrote in message
 > news:af5c72f9.0402031841.3f6ae4f3@posting.google.com...
 >
  > > I ran memtest-86 for the hell of it to check my new setup and it freezes
  > > on the first pass, second test (address test, own test, no cache). It
  > > locks up tight, I have to reset or poweroff to get out. I have run it
  > > with only the video card and 1 stick of ram and the same problem. I tried
  > > the ram in each of the 4 slots same problem.
  > > My setup: Intel D865perlk, 3.0 GHZ cpu, Kingston KVR400x64c3ak2/1g
  > > (pc3200).
  > > I googled and did find one post with the same problem but no solution
  > > offered.
  > > I was able to load Win XP without problem and started to install Linux
  > > but have not finished. I have not really stressed the system yet.
  > > I don't have another motherboard laying around that will handle
  > > pc3200.Nor do I have access to another d865perlk/3.0 GHZ cpu combo.
  > >
  > > Any suggestions appreciated.
 >
 > Jerry, I have the same memory installed in a Gigabyte GA-8KNXP non-Ultra v1
 > mobo, and although I never had Memtest86 v3 lock up, it sporadically gave
 > errors at the very end (>97%) of Test 5. I never obtained a good
 > explanation for it, but the computer has been very stable since last June,
 > so I've decided Memtest may be producing false errors. Also, running M$'s
 > memory tester reveals no errors.
 > I first started testing with 1gb of RAM, then with my present configuration
 > of 2gb, with the same result.
 > The last time I looked, the Memtest website referred to "false errors" that
 > can be generated on "some systems," but the specifics were few.
 >
 > I wouldn't worry about it if the system is stable. Find the M$ tester and
 > try it. If it passes, I would think no more about it.

I have since run GoldMemory tester without any problems. I will check the
M$ website for their mem tester.I am beginning to think memtest may have
false errors.
Thanks
Jerry<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: need help with memtest-86 v3.0 

There is a newer revision of MemTest, called MemTest86+, which is based on
Chris Brady's original program, but with some bug fixes (taken from web
site):

a.. Lots of bug-fixes, works now with all AMD64.
a.. Added detection for Athlon 64 / Opteron
a.. Added detection for latest Athlon XP (including Barton)
a.. Added detection for latest Pentium 4 (including Prescott)
a.. Added detection for SiS735/740/745/748/755/645/655
a.. Added detection for ATi Radeon 9100 IGP (RS300)
a.. Added detection for nVidia nForce/nForce2(400/Ultra)/nForce3/nForce3-S
a.. Added detection for VIA KT266(A)/333/KT400(A)/KT600/PT800/PT880/K8T800
a.. Added detection for i865P/PE/G/i875P/E7500/E7501/E7205
a.. ECC initialization & ECC polling added for all AMD64 CPUs
a.. ECC initialization & ECC polling added for i875P and E7501 & E7205
a.. Added complete advanced settings detection for i875P, AMD64 and nForce2.
a.. Added partial advanced settings detection for KT266(A)/KT333.
a.. Corrected ECC Error not counted as a global error anymore.
a.. Added a "safe-mode" (press F1 when memtest86+ loading) in case of
freeze.
a.. Source code now compatible with gcc 3

You can find it at <a style='text-decoration: underline;' href="http://www.memtest.org/." target="_blank">http://www.memtest.org/.</a> Hope this helps.

Russell
<a style='text-decoration: underline;' href="http://tastycomputers.com" target="_blank">http://tastycomputers.com</a>

"jerry" <jturba.TakeThisOut@yahoo.com> wrote in message
news:af5c72f9.0402050637.4eec597f@posting.google.com...
 > "Bob Davis" <iclick.TakeThisOut@coxNIX.net> wrote in message
news:<UweUb.9881$gl2.5325@lakeread05>...
  > > "jerry" <jturba.TakeThisOut@yahoo.com> wrote in message
  > > news:af5c72f9.0402031841.3f6ae4f3@posting.google.com...
  > >
   > > > I ran memtest-86 for the hell of it to check my new setup and it
freezes
   > > > on the first pass, second test (address test, own test, no cache). It
   > > > locks up tight, I have to reset or poweroff to get out. I have run it
   > > > with only the video card and 1 stick of ram and the same problem. I
tried
   > > > the ram in each of the 4 slots same problem.
   > > > My setup: Intel D865perlk, 3.0 GHZ cpu, Kingston KVR400x64c3ak2/1g
   > > > (pc3200).
   > > > I googled and did find one post with the same problem but no solution
   > > > offered.
   > > > I was able to load Win XP without problem and started to install Linux
   > > > but have not finished. I have not really stressed the system yet.
   > > > I don't have another motherboard laying around that will handle
   > > > pc3200.Nor do I have access to another d865perlk/3.0 GHZ cpu combo.
   > > >
   > > > Any suggestions appreciated.
  > >
  > > Jerry, I have the same memory installed in a Gigabyte GA-8KNXP non-Ultra
v1
  > > mobo, and although I never had Memtest86 v3 lock up, it sporadically
gave
  > > errors at the very end (>97%) of Test 5. I never obtained a good
  > > explanation for it, but the computer has been very stable since last
June,
  > > so I've decided Memtest may be producing false errors. Also, running
M$'s
  > > memory tester reveals no errors.
  > > I first started testing with 1gb of RAM, then with my present
configuration
  > > of 2gb, with the same result.
  > > The last time I looked, the Memtest website referred to "false errors"
that
  > > can be generated on "some systems," but the specifics were few.
  > >
  > > I wouldn't worry about it if the system is stable. Find the M$ tester
and
  > > try it. If it passes, I would think no more about it.
 >
 > I have since run GoldMemory tester without any problems. I will check the
 > M$ website for their mem tester.I am beginning to think memtest may have
 > false errors.
 > Thanks
 > Jerry<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: need help with memtest-86 v3.0 

"Russell" <rsullivan@tastycomputersdotcom_replacedotwith "."> wrote in message news:<RvtUb.185302$nt4.786018@attbi_s51>...
 > There is a newer revision of MemTest, called MemTest86+, which is based on
 > Chris Brady's original program, but with some bug fixes (taken from web
 > site):
 >
 > a.. Lots of bug-fixes, works now with all AMD64.
 > a.. Added detection for Athlon 64 / Opteron
 > a.. Added detection for latest Athlon XP (including Barton)
 > a.. Added detection for latest Pentium 4 (including Prescott)
 > a.. Added detection for SiS735/740/745/748/755/645/655
 > a.. Added detection for ATi Radeon 9100 IGP (RS300)
 > a.. Added detection for nVidia nForce/nForce2(400/Ultra)/nForce3/nForce3-S
 > a.. Added detection for VIA KT266(A)/333/KT400(A)/KT600/PT800/PT880/K8T800
 > a.. Added detection for i865P/PE/G/i875P/E7500/E7501/E7205
 > a.. ECC initialization & ECC polling added for all AMD64 CPUs
 > a.. ECC initialization & ECC polling added for i875P and E7501 & E7205
 > a.. Added complete advanced settings detection for i875P, AMD64 and nForce2.
 > a.. Added partial advanced settings detection for KT266(A)/KT333.
 > a.. Corrected ECC Error not counted as a global error anymore.
 > a.. Added a "safe-mode" (press F1 when memtest86+ loading) in case of
 > freeze.
 > a.. Source code now compatible with gcc 3
 >
 > You can find it at <a style='text-decoration: underline;' href="http://www.memtest.org/." target="_blank">http://www.memtest.org/.</a> Hope this helps.
 >
 > Russell
<font color=purple> > <a style='text-decoration: underline;' href="http://tastycomputers.com</font" target="_blank">http://tastycomputers.com</font</a>>
 >

I downloaded memtest+, and ran it. It locks up in exactly the same
place. I did read on the updated website that there is a reported bug
with the i865 motherboards (among others).
In the interest of accuracy memtest-86 and memtest+ both appear to
crash at the start of test #2 when run from the default configuration
; but in actuality they crash at the very end of test #1. If I use the
configuration menu to run each test independently(there are 12 tests
0-11) test 2 loops and runs perfectly. Test 1 will crash at the very
end and lock up.
I have run GoldMemory, and a Microsoft memory tester (mtinst.exe)
without any errors, and I have not had any symptoms (yet!), so I will
wait to see if memtest+ has a bug patch.
Thanks for the suggestion Russell.
Jerry<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: need help with memtest-86 v3.0