[PATCH 0/1][SRU][P] System hang while connecting dock during s2idle on AMD system

[AceLan Kao]

From: "Chia-Lin Kao (AceLan)" <acelan.kao at canonical.com>

BugLink: https://bugs.launchpad.net/bugs/2106783

[Impact]
Some platforms are reported to have a failure that can occur when the
EC SCI is asserted but the system doesn't wake up (such as no Notify()
after processing EC events). What happens is the APU rapidly enters
and exits s0i3 multiple times. On affected systems this triggers the
over voltage protection (OVP) mechanism of the voltage regulator module
(VRM). The system will be forced off, and to be reset the battery will
need to be disconnected and reconnected.

[Fix]
AMD will submit the patches to the upstream soon.
https://lore.kernel.org/platform-driver-x86/70dfa642-4c97-4aaf-aa79-70127974f03e@amd.com/T/#m174cfe2f4ec5893e39cff6994a93ebd499ec29e7

Included in v6.15-rc5
9f5595d5f03fd platform/x86/amd: pmc: Require at least 2.5 seconds between HW sleep cycles

[Test]
>From the listed AMD machines, boot up and makes it enter suspended, and
then plug the dock. The system should work normally without hang.
DELL PRO 13 PLUS PB13255, DELL PRO 13 PLUS PB13255 360,
DELL PRO 14 PLUS PB14255, DELL PRO 16 PLUS PB16255,
DELL PRO 14 PC14255, DELL PRO 16 PC16255

[Where problems could occur]
Add sleep() in some cases should be safe and not possible to introduce regressions.

Mario Limonciello (1):
  platform/x86/amd: pmc: Require at least 2.5 seconds between HW sleep
    cycles

 drivers/platform/x86/amd/pmc/pmc.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

-- 
2.43.0