[Bug 221065] ideapad_acpi: unexpected charge_types spam on Yoga Pro 7 14ASP9

[bugzilla-daemon@kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=221065

--- Comment #36 from Avraham Hollander (anhollander516@gmail.com) ---
(In reply to Rong Zhang from comment #34)
> I also found some time looking into ucsi_acpi and the AML code of the ACPI
> device (in SSDT17).
> 
> Two helper methods (\_SB.PCI0.LPC0.EC0.ECRD for reading,
> \_SB.PCI0.LPC0.EC0.ECWT for writing) are repeatedly called a total of 25 (!)
> times when handling a UCSI command. Their timeout for acquiring the ECMT
> mutex is 1s. Therefore, it is not surprising that some of these calls will
> time out.
> 
> Considering this, it seems that ucsi_acpi will cause significant pressure on
> the ECMT mutex, as each UCSI command acquires and releases it 25 times.
> Instead of blacklisting ideapad-laptop, could you try to blacklist
> ucsi_acpi? Doing so should reduce the pressure on the ECMT mutex a lot, so
> other ACPI methods may be less likely to time out. Again, this won't
> completely "fix" the issue, and a BIOS update is the only way to fix it.
> 
> Moreover, there is no way for us to determine if any call within them has
> failed, so we can't trust the UCSI ACPI device anymore -- a quirk will also
> be required to disable usci_acpi on 83DR.

I don't *seem* to be getting the mutex error with it blacklisted. However, I'm
still getting the original error from ideapad_acpi.

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

You are receiving this mail because:
You are watching the assignee of the bug.