Announcement

Collapse
No announcement yet.

Ryzen-Test & Stress-Run Make It Easy To Cause Segmentation Faults On Zen CPUs

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • c2h5oh
    replied
    Originally posted by Michael View Post

    I will have more information out later today or tomorrow as well, running several hour long tests in different workload configurations... Now that I can reproduce super-easily via phoronix-test-suite stress-run, encourages me to run more tests whenever it's PTS automated, and being able to show off PTS stress-run capabilities since I don't often get to talk about it too much otherwise.
    One more for you to test: in bios set CPU voltage offset to +25mV. I've managed to crash my 1800X fairly consistently within 30 minutes when running large x264 encoding jobs, but this small voltage bump seems to have fixed it - It's been almost 40h of encoding and no issues. If you're running memory at speeds faster than 2666 make sure your SOC voltage is 1.1V (some bioses adjust that automatically, some don't)

    Leave a comment:


  • chuckula
    replied
    Originally posted by qsmcomp View Post
    Try disabling uop-cache from BIOS.
    Oh and turn off 7 of the cores!

    Leave a comment:


  • qsmcomp
    replied
    Try disabling uop-cache from BIOS.

    Leave a comment:


  • aufkrawall
    replied
    Mesa is compiled with Clang. But it also shows the issue, I randomly encountered it when I had a Ryzen R7 1700 and was compiling llvm several times.

    Leave a comment:


  • khnazile
    replied
    I've run into gcc random crashes on ryzen while compiling Mesa git master just few days ago. At first, I thought it was something wrong with recent code changes, but then I've remembered people talking about issues under load, run the compilation again and it finished just fine.

    Leave a comment:


  • chuckula
    replied
    Originally posted by Michael View Post

    I will have more information out later today or tomorrow as well, running several hour long tests in different workload configurations... Now that I can reproduce super-easily via phoronix-test-suite stress-run, encourages me to run more tests whenever it's PTS automated, and being able to show off PTS stress-run capabilities since I don't often get to talk about it too much otherwise.
    Thanks. Just for fun, try firing up the same test on your i9 7900X system and see if it encounters any issues.

    Leave a comment:


  • aufkrawall
    replied
    Good to see at least Phoronix does their job. Haven't seen it reported anywhere else in the media, even though the issue is going on for months.

    Leave a comment:


  • Michael
    replied
    Originally posted by chuckula View Post
    Thank you Michael for running these not-so-fun but important tests and spreading the word. It's great to see these issues go from a large number of forum posts to an automated tool that can help illustrate these very serious issues.

    I gave you a linked shout-out over on techreport to help spread the word about this and I appreciate that you even re-ran the tests with SMT disabled and lower-clocked RAM to help remove variables from the equation.
    I will have more information out later today or tomorrow as well, running several hour long tests in different workload configurations... Now that I can reproduce super-easily via phoronix-test-suite stress-run, encourages me to run more tests whenever it's PTS automated, and being able to show off PTS stress-run capabilities since I don't often get to talk about it too much otherwise.

    Leave a comment:


  • gQuigs
    replied
    Thanks from me too.. I had no idea it was this common. I used to donate my extra resources to BOINC but stopped because of instability when loading all the cores.

    Leave a comment:


  • chuckula
    replied
    Thank you Michael for running these not-so-fun but important tests and spreading the word. It's great to see these issues go from a large number of forum posts to an automated tool that can help illustrate these very serious issues.

    I gave you a linked shout-out over on techreport to help spread the word about this and I appreciate that you even re-ran the tests with SMT disabled and lower-clocked RAM to help remove variables from the equation.

    Leave a comment:

Working...
X