[Bug 217931] amd-pstate lacks crucial features: CPU frequency and boost control

[bugzilla-daemon@kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=217931

--- Comment #68 from Mario Limonciello (AMD) (mario.limonciello@amd.com) ---
This bug should have been closed a while ago (boost is landed a while now)
rather than cargo culted for the new issue.  In any case let's work through
your new issue and then we can close it once and for all.

> AMD Ryzen 9 6900HS, ROG Zephyrus G14 GA402RK
> After S3 sleeping the machine

S3 resume?  I don't believe that's a POR configuration for Ryzen 6000.  Are you
sure you don't mean s0i3/s2idle?  Or have you done some hacking on your BIOS to
try to get S3 back?

It would be really illustrative to me if you can please clone this repo and run
amd_s2idle.py and share the report.

https://git.kernel.org/pub/scm/linux/kernel/git/superm1/amd-debug-tools.git/

> it goes haywire, even boosting CPUs (ignoring boost flag) and going above the
> set limits.

I'm wondering if there is a case that the CPPC registers are in the wrong state
after resume.  In addition to running amd_s2idle.py, can you please run
amd_pstate.py both at bootup and then again after suspend?  Save two separate
reports and attach them here.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.