Amd_pmc: Last suspend didn't reach deepest state

Which Linux distro are you using?
Arch Linux

Which release version?
rolling

updated today

Which kernel are you using?
6.19.6-zen1-1-zen

Which BIOS version are you using?
03.04

Which Framework Desktop model are you using?
AI 395 MAX+ / Radeon 8060S (128G)

The Problem:

for some days now this nice little beast refuses to go into deepest sleep state. I could immediately tell because the fan doesn’t stop spinning (it’s set to 30% min always-on to avoid the PSU fan noise issue).

For me the problems started after an update of Arch (I’m guessing here because I don’t do reboots otherwise - except to switch Kernels) at 2026-02-28T10:42:20+01:00

...
2026-02-24T19:46:05+01:00 Linux version 6.18.9-zen1-2-zen (linux-zen@archlinux) (gcc (GCC) 15.2.1 20260209, GNU ld (GNU Binutils) 2.46) #1 ZEN SMP PREEMPT_D
YNAMIC Mon, 09 Feb 2026 17:45:06 +0000
2026-02-24T22:20:49+01:00 PM: suspend entry (s2idle)
2026-02-24T23:33:00+01:00 PM: suspend exit
2026-02-25T00:10:48+01:00 PM: suspend entry (s2idle)
2026-02-25T07:18:04+01:00 PM: suspend exit
...
2026-02-27T00:33:36+01:00 PM: suspend entry (s2idle)
2026-02-27T07:19:33+01:00 PM: suspend exit
2026-02-27T08:10:03+01:00 PM: suspend entry (s2idle)
2026-02-27T20:06:38+01:00 PM: suspend exit
2026-02-28T10:42:20+01:00 Linux version 6.18.13-zen1-1-zen (linux-zen@archlinux) (gcc (GCC) 15.2.1 20260209, GNU ld (GNU Binutils) 2.46) #1 ZEN SMP PREEMPT_
DYNAMIC Wed, 25 Feb 2026 23:12:22 +0000
2026-02-28T13:26:23+01:00 PM: suspend entry (s2idle)
2026-02-28T13:26:26+01:00 amd_pmc AMDI000B:00: Last suspend didn't reach deepest state
2026-02-28T13:26:26+01:00 PM: suspend exit
2026-02-28T13:26:36+01:00 PM: suspend entry (s2idle)
2026-02-28T18:57:04+01:00 amd_pmc AMDI000B:00: Last suspend didn't reach deepest state
2026-02-28T18:57:04+01:00 PM: suspend exit
...

I’ve had this issue occasionally before (e.g. with a 6.12 LTS kernel, 6.18.3-zen1-1-zen or 6.18.1-arch1-2) but never so consistently, only once every 20+ cycles.

Updating to the latest 6.19.6 does not solve the issue for me, as reported in a similar issue for the Laptop here:

Then tried different kernels but always the same result, the machine never (not a single time) makes it into deep sleep:

2026-03-08T11:38:22+01:00 Linux version 6.19.6-arch1-1 (linux@archlinux) (gcc (GCC) 15.2.1 20260209, GNU ld (GNU Binutils) 2.46) #1 SMP PREEMPT_DYNAMIC Wed, 04 Mar 2026 18:25:08 +0000
2026-03-08T11:39:12+01:00 PM: suspend entry (s2idle)
2026-03-08T11:39:32+01:00 amd_pmc AMDI000B:00: Last suspend didn't reach deepest state
2026-03-08T11:39:32+01:00 PM: suspend exit
2026-03-08T11:39:51+01:00 PM: suspend entry (s2idle)
2026-03-08T11:40:00+01:00 amd_pmc AMDI000B:00: Last suspend didn't reach deepest state
2026-03-08T11:40:00+01:00 PM: suspend exit
...
2026-03-08T11:45:54+01:00 Linux version 6.18.16-1-lts (linux-lts@archlinux) (gcc (GCC) 15.2.1 20260209, GNU ld (GNU Binutils) 2.46) #1 SMP PREEMPT_DYNAMIC Wed, 04 Mar 2026 18:09:01 +0000
2026-03-08T11:48:34+01:00 PM: suspend entry (s2idle)
2026-03-08T11:48:46+01:00 amd_pmc AMDI000B:00: Last suspend didn't reach deepest state
2026-03-08T11:48:46+01:00 PM: suspend exit
2026-03-08T11:49:16+01:00 PM: suspend entry (s2idle)
2026-03-08T11:49:27+01:00 amd_pmc AMDI000B:00: Last suspend didn't reach deepest state
2026-03-08T11:49:27+01:00 PM: suspend exit
...

Apparently my pacman update log doesn’t go back enough to tell what was included in the update, but before the boot where the problem started, also the linux-firmware based packages where updated, including e.g. the AMD Microcode:

$ pacman -Qi amd-ucode 
Name            : amd-ucode
Version         : 20260221-1
Description     : Microcode update image for AMD CPUs
Architecture    : any
...
Build Date      : 2026-02-21T21:18:41%:z
Install Date    : 2026-02-28T00:14:26%:z <<<
Install Reason  : Explicitly installed
Install Script  : No
Validated By    : Signature

So now I am thinking in that direction instead of the Kernel.

Anyone have a better clue?

Possibly related to this?

I would open an issue in that issue tracker, using the s2idle bug template.

1 Like

Yes agree

Update to latest Linux firmware (just tagged today).

1 Like

Wow, thanx for the link, verified that downgrading to a previous firmware worked and my fan stops spinning, no more warnings:

sudo pacman -U /var/cache/pacman/pkg/linux-firmware-other-20260110-1-any.pkg.tar.zst

Just upgraded and tested: :check_mark:

So version 20260309-1 is fine again (at least for me).

Same here. My framework 13 with AMD AI 300 series on cachyos, finally reaches proper sleep after upgradng linux-firmware and linux-firmware-other to 20260309-1:

Start Time Duration Hardware Sleep Battery Start Battery Delta Battery Ave Rate Wake Pin Wake Interrupt
0 2026-03-23 10:30:34 0:02:02 96.72% 40.93% -0.07% -0.07W ACPI SCI