Re: [PATCH] platform/x86: hp-wmi: Add OMEN board 8A17 thermal profile support

[Ciro Pedrini]

Hey Krishna, sorry for my late response.

Unfortunately, I can't right now. I fixed a power usage issue on it
(the performance mode wasn't really using the expected and reported
wattage, perhaps you encountered the same thing?) and it died a few
weeks later.

Guess it was used to running on low energy and decided "enough is enough".

It's at a repair shop now, so if it comes back to life in a few weeks
and you still need the dump, I'll gladly send it over.

El mar, 6 oct 2026 a la(s) 1:52 a.m., Krishna Chomal
(krishna.chomal108@xxxxxxxxx) escribió:
>
> On Fri, Sep 25, 2026 at 05:52:26PM -0300, Ciro Pedrini wrote:
> >The HP OMEN 17-ck1xxx (board 8A17) isn't in omen_thermal_profile_boards,
> >so hp-wmi uses the generic thermal profile query. The firmware only
> >stores that value and never acts on it, so the EC stays in its default
> >mode and the CPU is held at 45 W whatever profile is selected.
> >
> >The board uses the OMEN interface (mode byte at EC 0x95, thermal policy
> >v1) and also needs the timed profile quirk: without it, the EC drops
> >back to the default mode about two minutes after each profile change.
> >
> >Add 8A17 to omen_thermal_profile_boards and
> >omen_timed_thermal_profile_boards.
> >
> >With this, all three profiles work and sustained package power under
> >load goes from 45 W to 56-79 W.
>
> Hi Ciro Pedrini,
>
> I was working a patch[1] involving omen_timed_thermal_profile_boards[]
> and needed some diagostic info. Would you mind attaching the acpidump
> for your board 8A17?
>
> Thanks!
>
> [1]: https://lore.kernel.org/platform-driver-x86/asPy5Q5wNYaMzTqJ@archlinux/#m922ff6cf9a68285d2b0f64e578fd6d2d6fa573f4
>
> >Cc: stable@xxxxxxxxxxxxxxx
> >Assisted-by: Claude:claude-opus-5-5
> >Signed-off-by: Ciro Pedrini <ciro.pedrini@xxxxxxxxx>
> >---
> >
> >Notes:
> >    Tested on 7.2.7, whose OMEN profile code matches for-next. Without the
> >    timed quirk, EC 0x63 counts down from about 120 after each profile
> >    write, and at zero the EC resets 0x95 to 0x30.
> >
> >    8A13, 8A14, 8A16 and 8A18 appear in the same DSDT. 8A18 is in the OMEN
> >    list but not the timed one, so it may have the same two-minute
> >    fallback. I can't test it.
> >
> >    I used an AI assistant (Claude) for the ACPI analysis and to draft this
> >    message. I tested the change on my machine and reviewed everything.
> >
> > drivers/platform/x86/hp/hp-wmi.c | 4 ++--
> > 1 file changed, 2 insertions(+), 2 deletions(-)
> >
> >diff --git a/drivers/platform/x86/hp/hp-wmi.c b/drivers/platform/x86/hp/hp-wmi.c
> >index 7ab81ce5f..76f489d5d 100644
> >--- a/drivers/platform/x86/hp/hp-wmi.c
> >+++ b/drivers/platform/x86/hp/hp-wmi.c
> >@@ -207,7 +207,7 @@ static const char * const omen_thermal_profile_boards[] = {
> >       "886B", "886C", "88C8", "88CB", "88D1", "88D2", "88F4", "88F5", "88F6",
> >       "88F7", "88FD", "88FE", "88FF",
> >       "8900", "8901", "8902", "8912", "8917", "8918", "8949", "894A", "89EB",
> >-      "8A15", "8A18", "8A42", "8A43",
> >+      "8A15", "8A17", "8A18", "8A42", "8A43",
> >       "8BAD",
> >       "8E41",
> > };
> >@@ -226,7 +226,7 @@ static const char * const omen_thermal_profile_force_v0_boards[] = {
> >  * "balanced" when reaching zero.
> >  */
> > static const char * const omen_timed_thermal_profile_boards[] = {
> >-      "8A15", "8A42",
> >+      "8A15", "8A17", "8A42",
> >       "8BAD",
> > };
> >
> >
> >base-commit: f475845eaf3d749114a63270bf2efea459e14dd2
> >--
> >2.55.0
> >