Announcement

Collapse
No announcement yet.

List of Motherboards with issues when running MemTest86 in multi-CPU selection modes

Collapse
This is a sticky topic.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #91
    Hi

    [this thread (although old) seems similar i.e. "Hang on MemTest86 startup during multiprocessor test" so hopefully ok to post here]

    When booting [USB] MemTest86 [v8.40 free] to test my new rig, it seems to hang during the startup multiprocessor test -

    Processor: AMD Ryzen 3700X [8 core]
    MB: ASRock X570 Phantom Gaming-ITX/TB3 i.e. AMD X570 chipset. UEFI v1.70 09/11/2019
    RAM: 32GB OLOy 3200MHz 16-20-20-38 [2 sticks, 64GB total]

    I tried adding this MB to the blacklist
    but now MemTest86 startup hangs showing 'Getting memory controller details...'

    Everything running stock i.e. no overclocking etc. Windows 10 installed with no issues - task manager shows correct CPU & cores, memory etc

    I have attached the MemTest86.log [I tried to cut out the entries for 2 failed boots - one before and other after I added MB to blacklist]

    TIA for any help/clues

    Cheers
    M
    Attached Files

    Comment


    • #92
      We suspect it might be related to this issue, which is related to setting the screen resolution.This occurs immediately after configuring the multiprocessor settings.
      https://www.passmark.com/forum/memte...et-screen-size

      Please give the following build a try at your convenience:

      https://www.passmark.com/temp/memtes...b-8.4.1003.zip

      If possible, please upload the MemTest86.log file as before.

      Comment


      • #93
        Hi
        Thanks for the quick response
        I rewrote my USB SD card with your new 1003 version [i.e. my MB will not now be in blacklist] and it hangs in the same spot i.e. last entry showing on screen is 'Testing multiprocessor support..."

        My monitor [ancient Dell 2407 - resolution 1920 x 1200] is attached to computer graphics card HDMI port [Radeon 5600XT]
        I don't know if this is relevant, but the 9 lines of info [Starting "Retrieving hardware info. Please wait..."] don't start at top left of screen - they start about a third way down screen and a third of the way across from left

        FYI: I also got the 404 downloading your link - seems there's an extra space at end. If copy link, shows as '%20' - if I stripped this off, then it worked
        i.e.
        https://www.passmark.com/temp/memtes...b-8.4.1003.zip%20

        HTH
        M
        Attached Files

        Comment


        • #94
          I'm also getting the hang on Testing multiprocessor support... I've tested with both the GA version and the version 8.4.1003. Please see attached log.

          Cheers
          Chris
          Attached Files

          Comment


          • #95
            Link with the space %20 has been fixed.
            We'll take a look at the logs and get back to you.

            Comment


            • #96
              Mine is hanging on Testing multiprocessor support... as well. I also just tried version 8.4.1003, but had the same problem. My log is attached.


              MemTest86.log

              Comment


              • #97
                Thanks for the logs.

                We added some more debug output, as well as disabled setting of the graphics resolutions. The screen may look off but hopefully it is a temporary workaround for the hang.

                https://www.passmark.com/temp/memtes...b-8.4.1004.zip

                Please attach the logs if possible.

                If the problem still occurs, can you also try the previous MemTest86 v8.3 release and see if the problem occurs as well:
                https://www.memtest86.com/downloads/...86-8.3-usb.zip
                Last edited by keith; Apr-17-2020, 01:18 AM.

                Comment


                • #98
                  Hi
                  With 8.4 v1004, screen message still hangs at same point "Testing multiprocessor support..." [I left for 10mins] but the log gets very sightly further
                  2020-04-17 14:10:27 - [CONFIG] This platform has 16 logical processors of which 16 are enabled. # Log for v1003 used to end here
                  2020-04-17 14:10:27 - [CONFIG] SelCPUNo = 0
                  2020-04-17 14:10:27 - [CONFIG] CurLang = 0
                  2020-04-17 14:10:28 - [CONFIG] ConsoleMode = 0 (MaxMode = 4)

                  With 8.3, it seems to work - it [auto] went into a test
                  The equiv log for above shows
                  2020-04-17 14:29:08 - [CONFIG] This platform has 16 logical processors of which 16 are enabled.
                  2020-04-17 14:29:08 - Console size = 80 x 25
                  2020-04-17 14:29:09 - Screen resolution is too high (1920 x 1200). Setting screen size to the maximum 1920 x 1080
                  2020-04-17 14:29:09 - Unable to set screen size
                  2020-04-17 14:29:09 - Screen size = 1920 x 1200
                  2020-04-17 14:29:09 - Char width=8 height=19

                  Maybe the 'too high' resolution 1920 x 1200 is not being handled correctly by 8.4?

                  P.S. It would be convenient if app displayed to screen the version of software being run e.g. the same string as put in log file?
                  MemTest86 V8.4 Free Build: 1004 (64-bit)

                  Thanks
                  M
                  Attached Files

                  Comment


                  • #99
                    Hi
                    Maybe I should add future comments to the other thread below as now reasonably sure my issue not due to Multi-CPU issues [I was just mislead by the screen messages]?
                    http://www.passmark.com/forum/memtes...et-screen-size

                    I thought I'd give my new computer a soak test with v8.3 anyway and it seems to run ok - 0 errors so far

                    A new issue is with my mouse [Logitech MX] - moving it 10cm moves the mouse pointer about 1mm on screen i.e. it does seem to move with mouse movements, just extremely low resolution.
                    This made the mouse pretty much useless so I had to use the keyboard [which is not the end of the world]
                    The mouse works OK when I used it when in the computer BIOS [or when boot up into windows], just memtest86 has issue

                    Regards
                    M

                    Comment


                    • Logitech MX mouse is too smart for its own good.
                      Seems it is non standard and needs Logitech device driver to work correctly. What happens if you go into the computer's BIOS screen? Does the mouse work there?

                      Comment


                      • Yes. I mentioned that mouse works ok in BIOS and Windows 10

                        Comment


                        • Also experiencing a hang during boot at "Testing multiprocessor support". Hang occurred in version 8.4, 8.4.1003, and 8.4.1.004; however, I was able to successfully boot with no issue on version 8.3 on the ASRock B450M PRO4.

                          I have two different systems that I have tried, the above was only tested on one system (ASRock B450M PRO4), but both systems experienced the same issues.

                          System One
                          =========
                          Motherboard: ASRock B450M PRO4 (BIOS VERSION 3.60 and 3.90)
                          CPU: AMD Ryzen 2700 (no overclock)
                          RAM: G.Skill 16GB DDR4 3200MHz F4-3200C16D-16GVKB (stock speeds, no XMP profile)


                          System Two
                          =========
                          Motherboard: ASRock B450M Steel Legend (unsure of BIOS version, will update later)
                          CPU: AMD Ryzen 3400G (no overclock)
                          RAM: Corsair 16GB DDR4 3200MHz CMK16GX4M2B3200C16 (XMP profile enabled)

                          Will upload logs for the second system when I get a change. I was also going to include the logs for the successful boot on version 8.3, but for some reason any USB created with the 8.3 image will not showing up in file explorer and therefore is inaccessible (diskpart shows the disk, but can open).
                          Attached Files

                          Comment


                          • Originally posted by JacobCoyle View Post
                            Also experiencing a hang during boot at "Testing multiprocessor support". Hang occurred in version 8.4, 8.4.1003, and 8.4.1.004; however, I was able to successfully boot with no issue on version 8.3 on the ASRock B450M PRO4.

                            I have two different systems that I have tried, the above was only tested on one system (ASRock B450M PRO4), but both systems experienced the same issues.
                            Thanks for the logs.

                            We found a memory related issue that may have corrupted the memory used by the UEFI firmware, which may have caused the hang in v8.4 but not in v8.3.

                            We have created a new thread to track the v8.4 builds and related issues.
                            https://www.passmark.com/forum/memte...-4-pre-release

                            Please give the build a try at your convenience.



                            Comment


                            • I am encountering hangs at around 1 hour and 30 minutes testing on Single CPU mode in version 8.4 Pro Build 1001. This issue continues to exist even after disabling SMT.

                              I installed Linux successfully with no issues and have not encountered hangs for days.

                              Motherboard: Supermicro H11SSL-NC R2.0 (BIOS version 2.1)
                              CPU: AMD EPYC 7302P 16C/32T@3.00GHz
                              RAM: Samsung M393A2K40DB3-CWE 16GB DDR4-25600 3200MHz

                              Attached log file and a screenshot of last time updated.
                              Attached Files

                              Comment


                              • EECS,

                                Others have reported similar problems with SuperMicro boards.
                                https://www.passmark.com/forum/memte...-for-all-tests
                                https://www.passmark.com/forum/memte...lp-to-diagnose
                                One suggestion was to change the CPU selection to CPU #1, instead #0.
                                We think their BIOS is buggy.

                                Comment

                                Working...
                                X