On 7/15/20 1:11 PM, Chris Murphy wrote:
Hi, While bad RAM is uncommon, it comes up with some regularity to cause folks a lot of grief. I'm wondering if there's a way to make it easier to get bad news :-\ In particular there are cases where RAM defects just don't show up with a few hours of memtest86+, it can take days of contiguous testing, which is so inconvenient the test itself seems worse. Here's what I've got so far: 1. Fedora includes /boot/memtest86+-5.01 on every installation. But this is a legacy/BIOS program. [..] 2. The kernel has a built-in memory tester. [..] 3. "memory interface test" used at Google,[..] 4. "multiple concurrent kernel compiles"
There's also 5. memtester, http://pyropus.ca/software/memtester/ , packaged in Fedora. It runs in userland, so obvously isn't as thorough but I'd think it has got to better than gcc and other userland utilities, because it tries the 'known-tricky' access patterns rather than hoping to see them in compiler output.
I remember the beginnings of memtest86+: its predecessor was written at the SGI hardware division by folks who understood how e.g specific access paterns appear on the physical traces on the memory modules, and can lead to crosstalk, but only if you simulataneously and repeatedly toggle very specific bits. This type of knowledge was essential to writing good tests, and probably also to discovering vulnerabilites like rowhammer.
Have you looked at memtester? What do you think of it? _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx
