Réseau CPC BIENDEBUTER.NET Crunchez vos adresses URL
|
Calculez la conso électrique de votre PC
|
Hébergez vos photos
+ Reply to Thread
Page 1 of 12 1 2 3 4 5 6 7 8 9 11 ... LastLast
Results 1 to 30 of 353
  1. #1
    Memtest86+ V5.00b6 changelog
    ----------------------------

    - Added support for up to 2 TB of RAM on X64 CPUs
    - Added experimental SMT support up to 32 cores
    - Added complete detection for memory controllers.
    - Added Motherboard Manufacturer & Model reporting
    - Added CPU temperature reporting
    - Added enhanced Fail Safe Mode (Press F1 at startup)
    - Added support for Intel "Sandy Bridge-E" CPUs
    - Added support for Intel "Ivy Bridge" CPUs
    - Added preliminary support for Intel "Haswell" CPUs
    - Added preliminary support for Intel "Haswell-ULT" CPUs
    - Added support for AMD "Kabini" (K16) CPUs
    - Added support for AMD "Bulldozer" CPUs
    - Added support for AMD "Trinity" CPUs
    - Added support for AMD E-/C-/G-/Z- "Bobcat" CPUs
    - Added support for Intel Atom "Pineview" CPUs
    - Added support for Intel Atom "Cedar Trail" CPUs
    - Added SPD detection on most AMD Chipsets
    - Enforced Coreboot support
    - Optimized run time for faster memory error detection
    - Rewriten lots of memory timings detection code
    - Corrected bugs, bugs and more bugs

    Known bugs
    ----------


    ************************************************** ***********

    ************************************************** ***********

    Memtest86+ 5.00 Beta 1

    Bin : http://www.memtest.org/download/beta/500b1/mt500b1.bin
    ISO : http://www.memtest.org/download/beta/500b1/mt500b1.iso
    USB Installer : http://www.memtest.org/download/beta...t500b1.usb.exe

    ---------------
    Memtest86+ 5.00 Beta 3 - [04/03/2013]

    Bin : http://www.memtest.org/download/beta/500b3/mt500b3.bin
    ISO : http://www.memtest.org/download/beta/500b3/mt500b3.iso
    USB Installer : http://www.memtest.org/download/beta...t500b3.usb.exe

    ---------------
    Memtest86+ 5.00 Beta 4 - [06/03/2013]

    Bin : http://www.memtest.org/download/beta/500b4/mt500b4.bin
    ISO : http://www.memtest.org/download/beta/500b4/mt500b4.iso
    USB Installer : http://www.memtest.org/download/beta...t500b4.usb.exe

    Beta 4 : Solve the "hang" bug on test #7

    ---------------
    Memtest86+ 5.00 Beta 6 - [14/03/2013]

    Bin : http://www.memtest.org/download/beta/500b6/mt500b6.bin
    ISO : http://www.memtest.org/download/beta/500b6/mt500b6.iso
    USB Installer : http://www.memtest.org/download/beta...t500b6.usb.exe

    Beta 6 : Add support for Intel HSW/HSW-ULT & AMD 16h (Kabini+)

    ---------------

    Memtest86+ 5.00 RC1 - [29/03/2013]

    Bin : http://www.memtest.org/download/beta...1/mt500rc1.bin
    ISO : http://www.memtest.org/download/beta...1/mt500rc1.iso
    USB Installer : http://www.memtest.org/download/beta...500rc1.usb.exe

    ---------------
    Please run this Beta on as many platform as possible and post a quick report (with a screenshot if possible), if you find any error or if all goes well.

    Some tips about failsafe modes at startup :

    - F1 : Enter global fail safe, with basic features only
    - F2 : Toggle SMP. Will try to activate SMP on systems where SMP is disabled by default (Xeon, Opteron & Engineering Sample CPU) or will disable SMP on all other platforms.
    - F3 : Hidden mode. Will use old MP Table instead of modern ACPI MADT Table for SMP initialization.
    Last edited by Doc TB; 29/03/2013 at 19h49.

  2. #2
    Memtest 5.00 beta (and alpha) shows only 1 core on all Gigabytes Boards I have tested so far.

  3. #3
    Please try to press F3 at startup.

  4. #4
    Hi,

    I've tried the beta on a MBP8,2 (early 2011) with 16GB of RAM. It seems to "hang". It only gets as far as the detection of the memory modules and then just sits there. Cannot exit or enter config.
    I've tried pressing F1-F3 -- no improvement. Please see attached screenshot.

    Thanks.

    Also, where can we get the source code from?

    Last edited by qwerty123; 18/05/2012 at 06h23.

  5. #5
    Hello,

    Motherboard was recogniced correctly, but I got a Re-boot during test7.

    System:
    Gigabyte X79A-GD65 (8D) with a Intel i7 3930K, 64G RAM (BIOS 1.8 (beta)

    Hermann

  6. #6
    For MBP> Please post a screenshot on safe mode (F1)

    For GB MB> Could you try again with F1 ?

  7. #7

  8. #8
    I tried F1, F2, F3 and nothing worked.
    ASRock 970 Extreme3, AMD FX6100, 2x4GB Patriot RAM.
    4.20 boots fine.

  9. #9

  10. #10
    Quote Originally Posted by bjlockie View Post
    I tried F1, F2, F3 and nothing worked.
    ASRock 970 Extreme3, AMD FX6100, 2x4GB Patriot RAM.
    4.20 boots fine.
    Please, we need more information than "nothing worked". Do you have a screenshot ? Does it freeze ? Reboot ? When ?

    ---------- Post added at 13h38 ---------- Previous post was at 13h32 ----------

    Quote Originally Posted by qwerty123 View Post
    Voila

    Safe mode on MBP.
    http://oi47.tinypic.com/2u8ifpc.jpg
    Well, this is not safe mode, I don't understand how the code can enter this state...

  11. #11
    Quote Originally Posted by Doc TB View Post
    Well, this is not safe mode, I don't understand how the code can enter this state...
    That may well be the case, but I was able to get the same result 10/10 attempts.
    When running 4.20 it crashes with "unexpected interrupt" at 7%.

  12. #12
    It seems Apple doesn't follow the established standards, one more time. The memory map is not filled correctly and Memtest86+ just try to access reserved memory marked as usable. I wrote an email to Apple dev support some months ago and their answer was "Apple computers are not PC compatible". Built with PC compatible components, but especially modded for not being "PC Compatible". It's a shame, but I don't know how to solve this issue : Apple doesn't publish the required technical informations. :-/

  13. #13
    Quote Originally Posted by Doc TB View Post
    It seems Apple doesn't follow the established standards, one more time. The memory map is not filled correctly and Memtest86+ just try to access reserved memory marked as usable. I wrote an email to Apple dev support some months ago and their answer was "Apple computers are not PC compatible". Built with PC compatible components, but especially modded for not being "PC Compatible". It's a shame, but I don't know how to solve this issue : Apple doesn't publish the required technical informations. :-/
    Sad but true. Unfortunately apples BIOS implementation is pretty poor, though it's only there for legacy compatibility for things like bootcamp. Have you considered making memtest work from EFI? BIOS won't be around much longer. I've tried running memtest from grub-efi, but it doesn't work because it uses 16bit mode. Is it possible to make it use the 32bit boot protocol?

  14. #14
    The issue is not 16 or 32 bits boot protocol, the main issue is the numerous BIOS calls required to do the memory initialization. EFI is a nice idea, but the first time I heard "EFI will replace BIOS in the upcoming months !!!" was in 2002 at an Intel Developer's forum. 10 years after, UEFI replaced EFI but BIOS is still present in 99% of PC Motherboard. It's an hard task to build an EFI-readyMemtest86+, with massive code rewrite, and that version will not be compatible with legacy BIOS. I will not consider supporting two forks at the same time, so when Memtest86+ will switch to EFI, the BIOS version will be discontinued. When BIOS will be not be available in standard PC components, I'll start working on en EFI revision.

  15. #15
    Hello
    Re-run in Save mode

    3 hours with no Crash (nearly one single Pass for 64GB ).
    But: two single fails at 60GB (Test3) and 61GB (Test6). So I tried to isolate the corresponding DIMM.
    The expected failing DIMM was used in a slot where I expect to see 8GB - 16GB range. So I adjusted the Tested Range from 10GB to 14 GB (expceted error at 12GB and 13GB ).
    I somehow saw, that test3 was done VERY fast and I did not get any fail. So I wanted to go back to test all Memory. Hitting "0" to enable the new addressrange caused a direct reboot.
    I know, that other configuration settings are broken (e. g. test or core selection), but this is not mentioned for the Address range setting.

    Restarting in Safe mode and checking the behavior for Test3 showed the following:
    - Test 1 runs fast, but one sees some pattern changes.
    - Test 2, Test3, Test4 are just a single blink and next test executed is Test 5.
    - Afterwards the testflow seems fine.


    I would not expect that Tests 2,3,4 are that fast.
    So If Test3 would not be executed it would be strange to get the fail on test3 when al 8 DIMMs are plugged in. Not sure how to interprete this obervation.


    Best regards

    Hermann




    System:
    MSI X79A-GD65 (8D) with a Intel i7 3930K, 64G RAM (BIOS 1.8 (beta)) (Sorry typed Gigabyte in the old message, but it is a MSI)



    Hermann

  16. #16
    - Some tests are bypassed in SMP mode
    - There is a bug with test numbers displayed in beta 1
    - "All memory" in Address mode is broken. To be more precise, every option that require a test restart is broken and will reboot the PC.

    About single fails, I would be really interested to see a screenshot.

    PS : I think the final revision of Mt86+ 5.00 will have SMP disabled by default in all case.

  17. #17
    Quote Originally Posted by Doc TB View Post
    The issue is not 16 or 32 bits boot protocol, the main issue is the numerous BIOS calls required to do the memory initialization. EFI is a nice idea, but the first time I heard "EFI will replace BIOS in the upcoming months !!!" was in 2002 at an Intel Developer's forum. 10 years after, UEFI replaced EFI but BIOS is still present in 99% of PC Motherboard. It's an hard task to build an EFI-readyMemtest86+, with massive code rewrite, and that version will not be compatible with legacy BIOS. I will not consider supporting two forks at the same time, so when Memtest86+ will switch to EFI, the BIOS version will be discontinued. When BIOS will be not be available in standard PC components, I'll start working on en EFI revision.
    I understand that. Please see here: https://www.gnu.org/software/grub/ma...e/linux16.html
    Seeing as Apple's BIOS implementations is so buggy, I was hoping that by using 32bit boot protocol I would at least be able to call it from grub-efi. That might give it a better chance of getting correct addresses for access.
    Can we expect a code drop soon? Perhaps someone else would like to work on the EFI port...

  18. #18
    Quote Originally Posted by Doc TB View Post
    Please, we need more information than "nothing worked". Do you have a screenshot ? Does it freeze ? Reboot ? When ?[COLOR="Silver"]
    It shows a black screen and just reboots .
    The ASRock 970 Extreme3 is UEFI BIOS but 4.20 works.

  19. #19
    USB Key or Iso ? No blue screen ? No F1 Prompt ?

    I have that board here. Will try again.

  20. #20
    Quote Originally Posted by Doc TB View Post
    USB Key or Iso ? No blue screen ? No F1 Prompt ?

    I have that board here. Will try again.
    I was using the .bin from grub.
    I used unetbin to put the ISO on a USB but all I get is the unetbin (grub?) screen that says "default" but that doesn't run.
    The time just resets so it probably isn't a bootable image.
    I will try a cd later.
    Last edited by bjlockie; 20/05/2012 at 01h37.

  21. #21
    Hi,

    I've tested the USB installer for beta-1, and the bug I've reported earlier is still in: Memtest86+ hangs when monitor is switched on.

    It is, for me, very easy to reproduce: start Memtest86+, switch monitor off, wait ten seconds, then switch monitor on. Tested both in normal and fail-safe mode.

    Details can be found in the topic I mentioned.

  22. #22
    So I've tested this new beta and it's having issues detecting all of the cores on my computer. I have AMD Phenom II X3 720 CPU that has been unlocked to a quad-core AMD Phenom II X4 20. I have a GIGABYTE GA-770TA-UD3 motherboard, and 2 sticks of 2GB OCZ RAM.

    The screenshots are all located in this album and have titles and descriptions detailing what each one is. I tried F1, F2, and F3 modes as well as no options at all.

    EDIT: I forgot to mention I used the ISO version, and burned to a rewritable CD that was verified.

  23. #23
    Quote Originally Posted by vanaalten View Post
    Hi,

    I've tested the USB installer for beta-1, and the bug I've reported earlier is still in: Memtest86+ hangs when monitor is switched on.

    It is, for me, very easy to reproduce: start Memtest86+, switch monitor off, wait ten seconds, then switch monitor on. Tested both in normal and fail-safe mode.

    Details can be found in the topic I mentioned.
    Please unplug the USB hub on your screen before launching Memtest86+ then try again.

    ---------- Post added at 22h06 ---------- Previous post was at 22h01 ----------

    Quote Originally Posted by foldor View Post
    So I've tested this new beta and it's having issues detecting all of the cores on my computer. I have AMD Phenom II X3 720 CPU that has been unlocked to a quad-core AMD Phenom II X4 20. I have a GIGABYTE GA-770TA-UD3 motherboard, and 2 sticks of 2GB OCZ RAM.

    The screenshots are all located in this album and have titles and descriptions detailing what each one is. I tried F1, F2, and F3 modes as well as no options at all.

    EDIT: I forgot to mention I used the ISO version, and burned to a rewritable CD that was verified.
    Buggued BIOS.

    Even the PSN is not defined correctly. The MP table is corrupted and the ACPI table only declare one core. You should try without core unlocker.

  24. #24
    Ran 5.00b1 on following machine (boot from USB Stick):

    Mobo: MSI X79A-GD65 (8D) (BIOS v1.6)
    CPU: Core i7-3820
    RAM: 8x8GB Corsair CMZ8GX3M1A1600C10 (DDR3-1600)

    Completed 2 passes successfully in default mode (No keys pressed, Multithreading enabled). All Info displayed on-screen appeared to be correct (MoBo type, SPD, Frequencies, CPU Temp).


    Just two little notes:

    Misspelling of "Sequential" during Test #1: displayed as "Seqential"

    Test percentage counter seems to be a little bugged during Test #1: It already reaches 100% when address test is in progress on CPU #5 (out of 8 SMT cores total) somewhere between 10G and 20G in address range.

    But these are just "appearances", I think...

  25. #25
    Hello,

    I have tried 5.00b1 on two different motherboards of the same model. Both of them will run without error for several passes, then experience an error in test 5. The error only seems to occur during test 5. Motherboard is an Asus M5A97, which has UEFI. Memory is Kingston ECC DDR3-1600. I am attaching 2 screenshots for your reference. Notice that the speed detection is inconsistent; one shows "741 MHz (DDR3-1482) - BCLK: 92" and the other shows "802 MHz (DDR3-1605) - BCLK: 100".

    Can you tell me if I'm hitting a bug in the beta, or if I probably do have bad memory? I was able to run Memtest86+ 4.20 many, many times (over 20 runs) without any errors.

    Thanks for all your hard work!




    ---------- Post added at 15h33 ---------- Previous post was at 14h31 ----------

    I should add that ECC is enabled in BIOS, which has been updated to the latest version. The error always occurs at a very low memory address, usually 0.1MB or 0.4MB.

  26. #26
    Quote Originally Posted by Doc TB View Post
    Please unplug the USB hub on your screen before launching Memtest86+ then try again.
    Good suggestion! Unplugging the USB connector on my screen solves the problem - switching monitor on/off and Memtest86+ will keep running.

    So apparently the screen signals 'something' to the PC via the USB connection - perhaps the PC/Memtest86+ should ignore this?

  27. #27
    Seems to work well for me. Asus P9X79 Deluxe. Detects 12 cores and uses 'em all. 32GB of RAM, tested fine 1 pass so far.

  28. #28
    Testing this version 5.0b1 on 2 Netbook with AMD Processors. E350 (Dual Core) and E450 (Dual Core).
    The Program does not find 2 core in these processor.

  29. #29
    Quote Originally Posted by brickfrog View Post
    Hello,

    I have tried 5.00b1 on two different motherboards of the same model. Both of them will run without error for several passes, then experience an error in test 5. The error only seems to occur during test 5. Motherboard is an Asus M5A97, which has UEFI. Memory is Kingston ECC DDR3-1600. I am attaching 2 screenshots for your reference. Notice that the speed detection is inconsistent; one shows "741 MHz (DDR3-1482) - BCLK: 92" and the other shows "802 MHz (DDR3-1605) - BCLK: 100".

    Can you tell me if I'm hitting a bug in the beta, or if I probably do have bad memory? I was able to run Memtest86+ 4.20 many, many times (over 20 runs) without any errors.

    Thanks for all your hard work!

    http://img18.imageshack.us/img18/2673/errorkd.jpg
    http://img40.imageshack.us/img40/6663/memoryj.jpg

    ---------- Post added at 15h33 ---------- Previous post was at 14h31 ----------

    I should add that ECC is enabled in BIOS, which has been updated to the latest version. The error always occurs at a very low memory address, usually 0.1MB or 0.4MB.
    The error in test#5 at 0x13CF0 & 0x13CF8 is a bug in the code. It will be solved in the next beta. Did you get an error at another address ? It will help to get a screenshot.

  30. #30
    Quote Originally Posted by Doc TB View Post
    The error in test#5 at 0x13CF0 & 0x13CF8 is a bug in the code. It will be solved in the next beta. Did you get an error at another address ? It will help to get a screenshot.
    I believe it was at 0x274f0 and 0x274f8, but I am not certain. I just started another run to see if I can duplicate the error. I will provide a screenshot as soon as possible. Thank you for looking into this.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts