https://bugzilla.kernel.org/show_bug.cgi?id=220705

--- Comment #32 from Lukas Wunner ([email protected]) ---
Thank you! Just to double-check, this is with a stock kernel, right? I.e.
without any of the custom patches from this bugzilla or elsewhere?

Could you capture full dmesg output as well as full "sudo lspci -vvv" output
after booting and after a suspend/resume cycle and attach it to this bugzilla
so that I can double-check everything looks okay?

To answer your question, the PCH SRCCLKREQ# change in the coreboot devicetree
for the Eve only disables CLKREQ# at the single Root Port leading to the wifi
card. Everything else remains as before.

Normally what happens is, the Root Port disables the clock in L1.2 and when the
wifi card has data to transfer, it signals via CLKREQ# pin that the clock shall
be reinstated. The Root Port can't do that quickly enough. By disabling
SRCCLKREQ# at the Root Port, it keeps the clock enabled even in L1.2. The issue
was analyzed extensively inside Intel and this workaround was found to have the
best performance/energy tradeoff. Apparently leaving the clock enabled consumes
only little extra energy versus other approaches.

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

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

_______________________________________________
acpi-bugzilla mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-bugzilla

Reply via email to