Attention is currently required from: Michael Niewöhner.
2 comments:
Commit Message:
tested? if you can, would be great if you test windoze and linux
Hmm. Not sure what I want to see, but it's probably not "CPU did not enter SLP_S0!!! (S0ix cnt=0)." The system does reach PC10 (from /sys/kernel/debug/pmc_core/package_cstate_show).
While I can't see which PCH IPs could be safely disabled, the LAN and WLAN devices remain powered on (based on reading, before and after, /sys/kernel/debug/pmc_core/mphy_core_lanes_power_gating_status and matching with HSIO lane ownership bits in FIA PCRs).
While the board has the SLP_S0# pad disconnected, this is not strictly required for S0ix, right?
Patchset:
Patch Set 3:
Patch Set 3:
Patch Set 3: -Code-Review
Patch Set 3: Code-Review-1
Patch Set 3: Code-Review+2
wait... did I really forget SKL? 😮
It was using PMC, not LPIT. Is that a problem?
No, vendor firmware for my Skylake board uses both devices as well. So this seems to be fine?
No, I reworked that. PMC/LPID/LPIT was actually confusing. PEP(D) is the "right" (or more correct) device name for this, so I cleaned this up but obviously forgot to add it to SKL as well 😄
LPIT is related but a (very) different table, though.
Öh oh well, actually that PMC that you mean (src/soc/intel/skylake/acpi/pmc.asl) is the "real" PMC. The "pmc.asl" (iirc the name correctly) was implementing LPID/PEPD. That was what I cleaned up.
Right, I meant the 'real' PMC. Since other platforms weren't moving from (PCIe device) PMC to PEPD, but from LPID to PEPD, it seems fine that Skylake has both?
To view, visit change 49140. To unsubscribe, or for help writing mail filters, visit settings.