Re: [BUG]i2c_hid_acpi broken with 4.17.2 on Framework Laptop 13 AMD

[Mario Limonciello <mario.limonciello@xxxxxxx>]

On 3/6/2024 15:31, Marek Marczykowski-Górecki wrote:
> On Wed, Mar 06, 2024 at 02:37:20PM -0600, Mario Limonciello wrote:
> > On 3/6/2024 14:34, Sébastien Chaumat wrote:
> > > Le mer. 6 mars 2024 à 19:51, Mario Limonciello
> > > <mario.limonciello@xxxxxxx <mailto:mario.limonciello@xxxxxxx>> a écrit :
> > >
> > >      On 3/6/2024 12:49, Sébastien Chaumat wrote:
> > >       >
> > >       >
> > >       > Le mer. 6 mars 2024 à 19:08, Mario Limonciello
> > >       > <mario.limonciello@xxxxxxx <mailto:mario.limonciello@xxxxxxx>
> > >      <mailto:mario.limonciello@xxxxxxx
> > >      <mailto:mario.limonciello@xxxxxxx>>> a écrit :
> > >       >
> > >       >     On 3/6/2024 12:05, Sébastien Chaumat wrote:
> > >       >      >
> > >       >      >
> > >       >      > Le mer. 6 mars 2024 à 18:33, Mario Limonciello
> > >       >      > <mario.limonciello@xxxxxxx
> > >       >      >     Also; I'd be really interested to hear what happens with
> > >       >     s2idle with
> > >       >      >     Xen
> > >       >      >     now (if it works).
> > >       >      >
> > >       >      >
> > >       >      > suspend to RAM now works :)
> > >       >
> > >       >     Do you see /sys/power/suspend_stats/last_hw_sleep increasing
> > >      with your
> > >       >     suspend cycle?
> > >       >
> > >       >
> > >       > No,  it does not increase (0).
> > >       >
> > >
> > >      OK, then in that case I suggest you run
> > >      https://gitlab.freedesktop.org/drm/amd/-/blob/master/scripts/amd_s2idle.py 
> > > <https://gitlab.freedesktop.org/drm/amd/-/blob/master/scripts/amd_s2idle.py>
> > >      in the hypervisor to find out what's wrong.
> > >
> > >
> > > It fails with xen (not bare metal) during the prerequisite tests :
> > The script might need some modifications for tests that don't make sense in
> > the context of Xen.
> >
> > > ✅ GPIO driver `pinctrl_amd` available
> > > Traceback (most recent call last):
> > >     File "/home/sch/Downloads/amd_s2idle.py", line 2447, in <module>
> > >       test = app.prerequisites()
> > >              ^^^^^^^^^^^^^^^^^^^
> > >     File "/home/sch/Downloads/amd_s2idle.py", line 1938, in prerequisites
> > >       if not check():
> > >              ^^^^^^^
> > >     File "/home/sch/Downloads/amd_s2idle.py", line 826, in check_msr
> > >       val = read_msr(reg, 0)
> > >             ^^^^^^^^^^^^^^^^
> > >     File "/home/sch/Downloads/amd_s2idle.py", line 813, in read_msr
> > >       val = struct.unpack("Q", os.read(f, 8))[0]
> > >                                ^^^^^^^^^^^^^
> > > OSError: [Errno 5] Input/output error
> > >
> > > Last line in the log is :
> > > 2024-03-06 21:29:33,146 DEBUG: Lockdown: [none] integrity confidentiality
> > Do you have userspace MSR support compiled in?  If not; that could explain
> > that problem.  But it's very unlikely you changed the MSRs.
> > For the purpose of finding where the suspend problem is, I'd comment out
> > that specific check for now.
> The s2idle isn't supported in Xen yet. Only S3 is. In fact, I have some
> work-in-progress patches for that, I may try testing them on this
> system.
OK; the "most important parts" I want to point out that are on bare 
metal and might not be obvious:
 * Infra from drivers/acpi/x86/s2idle.c
 * amd-pmc driver
 * USB4 driver (thunderbolt.ko)
 * NVME needs to use "simple suspend"

FYI - I'm going to be MIA for a while on personal leave starting this 
weekend, please keep me on CC on your progress and if you need help when 
I get back from my leave I'll comment on it.