[PATCH 0/2][SRU][J/M/U] Fix ADL: System enabled AHCI can't get into s0ix when attached ODD

[Koba Ko]

lost Lunar, will send V2

On Mon, Nov 20, 2023 at 9:12 AM Stefan Bader <stefan.bader at canonical.com> wrote:
>
> On 19.11.23 00:54, Koba Ko wrote:
> > BugLink: https://bugs.launchpad.net/bugs/2037493
>
> Why all the other series and not Lunar?
>
> >
> > [Impact]
> > System can't enter suspend on AHCI mode,
> > the CPU FAN and System FAN is running, PSU LED is on.
> > When fail to enter suspend, remove ODD, the system can successfully enter suspend again.
> >
> > [Fix]
> > Enable LPM on Alder Lake-P AHCI.
> >
> > [Test Cases]
> > 1. enabled AHCI on target machine.
> > 2. boot with kernel applied fix.
> > 3. suspend then check slp_s0_residency_usec&package_cstate_show
> > ~~~
> > $ sudo cat /sys/kernel/debug/pmc_core/slp_s0_residency_usec;sudo cat /sys/kernel/debug/pmc_core/package_cstate_show
> > 4951200
> > Package C2 : 63571033
> > Package C3 : 6212
> > Package C6 : 5633477
> > Package C7 : 0
> > Package C8 : 0
> > Package C9 : 0
> > Package C10 : 5255561
> > $ sudo rtcwake -m mem -s 10
> > rtcwake: assuming RTC uses UTC ...
> > rtcwake: wakeup from "mem" using /dev/rtc0
> > $ sudo cat /sys/kernel/debug/pmc_core/slp_s0_residency_usec;sudo cat /sys/kernel/debug/pmc_core/package_cstate_show
> > 12965280
> > Package C2 : 75052691
> > Package C3 : 46989
> > Package C6 : 8108332
> > Package C7 : 0
> > Package C8 : 0
> > Package C9 : 0
> > Package C10 : 13628111
> > ~~~
> >
> > [where the issue could happen]
> > Medium, Intel promise there's no issue of ADL AHCI and
> > they will take responsibility after regression.
> > Intel also take LPM as POR so they will fix any issues.
> >
> > [Misc]
> > Because the board_ahci_mobile is still not modified to board_ahci_low_power on Jammy,
> > provide a single patch for Jammy.
> >
> > For generic J/M, passed the all arch compilations on cbd,
> > Jammy,
> > ~~~
> > $ git push j_cbd
> > Total 0 (delta 0), reused 0 (delta 0), pack-reused 0
> > remote: *** kernel-cbd *********************************************************
> > remote: * Queueing builds (your 'ubuntu_jammy_next'); ok to interrupt
> > remote: * For results: ssh cbd ls kobako-jammy-f5146c67dd2d-XWAJ
> > remote: * 0/1 workers busy, 0 builds queued
> > remote: 2023-11-17 19:50:25 kobako-jammy-f5146c67dd2d-XWAJ/amd64/BUILD-OK
> > remote: 2023-11-17 20:18:29 kobako-jammy-f5146c67dd2d-XWAJ/arm64/BUILD-OK
> > remote: 2023-11-17 20:01:41 kobako-jammy-f5146c67dd2d-XWAJ/armhf/BUILD-OK
> > remote: 2023-11-17 20:31:51 kobako-jammy-f5146c67dd2d-XWAJ/ppc64el/BUILD-OK
> > remote: 2023-11-17 20:37:25 kobako-jammy-f5146c67dd2d-XWAJ/s390x/BUILD-OK
> > remote: ************************************************************************
> > To cbd.kernel:jammy.git
> >   * [new branch] ubuntu_jammy_next -> ubuntu_jammy_next
> > ~~~
> > Mantic,
> > ~~~
> > $ git push m_cbd
> > Enumerating objects: 102, done.
> > Counting objects: 100% (102/102), done.
> > Delta compression using up to 8 threads
> > Compressing objects: 100% (17/17), done.
> > Writing objects: 100% (61/61), 252.10 KiB | 2.23 MiB/s, done.
> > Total 61 (delta 46), reused 54 (delta 41), pack-reused 0
> > remote: *** kernel-cbd *********************************************************
> > remote: * Queueing builds (your 'ubuntu_mantic_next'); ok to interrupt
> > remote: * For results: ssh cbd ls kobako-mantic-e489d3bb608c-9LiJ
> > remote: * 0/1 workers busy, 0 builds queued
> > remote: 2023-11-17 20:58:33 kobako-mantic-e489d3bb608c-9LiJ/amd64/BUILD-OK
> > remote: 2023-11-17 21:30:32 kobako-mantic-e489d3bb608c-9LiJ/arm64/BUILD-OK
> > remote: 2023-11-17 21:09:18 kobako-mantic-e489d3bb608c-9LiJ/armhf/BUILD-OK
> > remote: 2023-11-17 21:45:50 kobako-mantic-e489d3bb608c-9LiJ/ppc64el/BUILD-OK
> > remote: 2023-11-17 21:54:30 kobako-mantic-e489d3bb608c-9LiJ/s390x/BUILD-OK
> > remote: ************************************************************************
> > To cbd.kernel:mantic.git
> >   * [new branch] ubuntu_mantic_next -> ubuntu_mantic_next
> > ~~~
> >
> > For Unstable, successfully compiled with the Mantic configuration and without dkms.
> > ~~~
> > Total 0 (delta 0), reused 0 (delta 0), pack-reused 0
> > remote: *** kernel-cbd *********************************************************
> > remote: * Queueing builds (your 'ubuntu_unstable_next'); ok to interrupt
> > remote: * For results: ssh cbd ls kobako-mantic-dc0405f4bd4c-VZvy
> > remote: * 0/1 workers busy, 0 builds queued
> > remote: 2023-11-18 22:35:41 kobako-mantic-dc0405f4bd4c-VZvy/amd64/BUILD-OK
> > remote: 2023-11-18 23:03:15 kobako-mantic-dc0405f4bd4c-VZvy/arm64/BUILD-OK
> > remote: 2023-11-18 22:46:15 kobako-mantic-dc0405f4bd4c-VZvy/armhf/BUILD-OK
> > remote: 2023-11-18 23:16:02 kobako-mantic-dc0405f4bd4c-VZvy/ppc64el/BUILD-OK
> > remote: 2023-11-18 23:22:02 kobako-mantic-dc0405f4bd4c-VZvy/s390x/BUILD-OK
> > remote: ************************************************************************
> > To cbd.kernel:mantic.git
> >   * [new branch] ubuntu_unstable_next -> ubuntu_unstable_next
> > ~~~
> >
> > Mika Westerberg (1):
> >    ata: ahci: Add Intel Alder Lake-P AHCI controller to low power
> >      chipsets list
> >
> >   drivers/ata/ahci.c | 1 +
> >   1 file changed, 1 insertion(+)
> >
>
> --
> - Stefan
>