FS#38188 - [memtest86+] Bit fade test (#11) always fails

Attached to Project: Arch Linux
Opened by mpan (mpan) - Thursday, 19 December 2013, 03:47 GMT
Last edited by Toolybird (Toolybird) - Tuesday, 30 May 2023, 06:37 GMT
Task Type Bug Report
Category Upstream Bugs
Status Closed
Assigned To Levente Polyak (anthraxx)
Architecture x86_64
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 2
Private No

Details

Description:
In test #11 memtest reports error for every memory location. Tested on two physical machines. On both 4.20 reports no errors. The bug also consistently appears on a virtual machine.

Upstream has received the report (<http://forum.canardpc.com/threads/85380-Bug-5.01-bit-fade-test-%28-11%29-always-fails>).

Additional info:
* 5.01-1

Steps to reproduce:
1. Run memtest
2. Select test #11
3. Wait for a moment (with large memory this may actually take a few minutes or more)
   memtest-fail.png (6.3 KiB)
This task depends upon

Closed by  Toolybird (Toolybird)
Tuesday, 30 May 2023, 06:37 GMT
Reason for closing:  Fixed
Additional comments about closing:  See last comment. If not fixed, please report upstream.
Comment by mattia (nTia89) - Sunday, 16 November 2014, 15:34 GMT
I confirm the bug.
Comment by mattia (nTia89) - Sunday, 01 October 2017, 14:52 GMT
we can close the bug report

since it's 1)upstream and 2)three years old
currently I use memtest86-efi from AUR that it's ok (and magically there is no a #11 test...)
and likely 3)repo could switch to this version
Comment by mpan (mpan) - Monday, 02 October 2017, 04:12 GMT
aur/memtest86-efi is memtest86, not memtest86+ (note the “plus” sign). While growing from the same root, these are two different pieces of software now, with memtest86 being closed-source. Therefore it’s hardly replacement for anyone who wants to use FLOSS.

However, it’s hardly possible to not notice that memtest86+ is simply dead. This leaves three options: removal of memtest86+ from the repos or,
since some features of memtest86+ work, leaving it in the repos and either closing this bug or not. I believe that removal is pointless, because the package is still working and useful; also there are packages older than that (back to 2013). And if the package is kept, the report should stay up just to avoid a new being opened.
Comment by mattia (nTia89) - Monday, 02 October 2017, 17:49 GMT
ok!

IMHO we can keep package in our repo but we also need to close this bug report because it is not an Arch issue: our packagers can't fix it.
in fact should be reported upstream, in their forum, that is active (maybe them are able to help you).
Comment by oldherl (oldherl) - Sunday, 14 March 2021, 11:23 GMT
It seems that it's fixed in upstream version 5.31b last year, but not yet packaged in Arch.
Comment by oldherl (oldherl) - Sunday, 14 March 2021, 11:27 GMT
Oh, sorry. It's not fixed, just appears later in the test.
Comment by mattia (nTia89) - Saturday, 11 February 2023, 08:00 GMT
This report needs an update:

1) memtest86+ hits a new release in our repo (6.10); I cannot reproduce the issue anymore, and you?
2) current version does not ship anymore a #11 test; there is anyway a "fade" test #10, is it the same?
3) there is no a related bug report in the official bug tracker [https://github.com/memtest86plus/memtest86plus/issues/]; if you are still affected by this issue, open a new bug there (and report a link here, too)

Loading...