Public bug reported: This machine successfully enters s2idle, but with the integrated Intel Ethernet controller enabled in BIOS the CPU package remains in C2 for the entire suspend and never enters package C10 or Intel S0ix. Disabling only Ethernet LAN in BIOS restores package C10 and S0ix for approximately 98% of the suspend interval.
Hardware and software: - Lenovo ThinkPad P14s Gen 5 Intel, machine type 21G2 - BIOS R2DET42W 1.27 (latest available; issue also occurred with 1.26) - Ubuntu 24.04.4 LTS - Kernel 7.0.0-28-generic - Intel Meteor Lake - Ethernet PCI ID 8086:550a rev 20 - Lenovo subsystem 17aa:50e9 - PCI address 0000:00:1f.6 - Driver e1000e - Firmware exposes only s2idle; legacy S3 is unavailable Reproduction conditions: - Laptop disconnected from AC, dock, monitors and all external USB devices - Suspend initiated with systemctl suspend - RTC wake scheduled after 120 seconds - Package-state and PMC counters captured by a systemd-sleep hook immediately before suspend and immediately after resume Result with Security -> I/O Port Access -> Ethernet LAN enabled: - Package C2: +117.38 seconds - Package C10: +0 - SLP_S0 residency: +0 - PMC0/PMC1 S0i2.0, S0i2.1 and S0i2.2: +0 - Suspend itself reports success - Longer 4004-second test also produced zero SLP_S0 and zero S0ix, with roughly 1.5-1.8 W estimated suspended draw Result after disabling only Ethernet LAN in BIOS and rebooting, under the same battery-only isolation conditions: - Ethernet function 0000:00:1f.6 is absent, as expected - Package C10: +118.15 seconds - S0i2.2: +118.03 seconds on both PMCs - SLP_S0 residency: +118.07 seconds - Approximately 98% of the 120-second interval was genuine deep sleep Additional isolation performed: - Setting /sys/bus/pci/devices/0000:00:1f.6/power/control to auto moved the NIC to runtime-suspended state but did not restore package C10 or S0ix during system suspend. - The power-saver platform profile made no difference. - Timestamped IRQ tracing showed no device interrupts during the machine_suspend interval, ruling out a wake/interrupt storm. - intel_idle C10 is enabled, max_cstate=9, and no process holds /dev/cpu_dma_latency. - intel_pmc_core ltr_ignore_all_suspend is enabled. - With Ethernet enabled, the PMC live state showed GBE_PGD0_PG_STS, GBE_D3_STS and GBE_VNN_REQ_STS unmet. - The e1000e log contains "EEE TX LPI TIMER: 00000011" at resume but no explicit "Failed to enable ULP" warning. - Updating BIOS from 1.26 to 1.27 did not change the failure. Expected result: The integrated Ethernet controller should remain available while allowing package C10/S0ix during s2idle, at least when there is no Ethernet link and Wake-on-LAN is not enabled on battery. Actual result: The machine must sacrifice its integrated Ethernet controller through a BIOS setting to obtain functional low-power suspend. Related reports and upstream change: - Fedora bug for the same 8086:550a/e1000e device preventing suspend: https://bugzilla.redhat.com/show_bug.cgi?id=2306163 - Kernel bug linked from Fedora: https://bugzilla.kernel.org/show_bug.cgi?id=218940 - Earlier e1000e suspend bug: https://bugzilla.kernel.org/show_bug.cgi?id=205015 - Commit making e1000e suspend errors non-fatal: https://github.com/torvalds/linux/commit/0a6ad4d9e1690c7faa3a53f762c877e477093657 - P14s Gen 5 owner reporting that disabling Ethernet in BIOS restored suspend: https://discussion.fedoraproject.org/t/suspend-failure-and-degraded-performance-after-failed-suspend-on-kernel-6-10-3/128299/7 Commit 0a6ad4d9e169 notes that ignoring the Ethernet hardware error may cause higher suspended power draw. This report may be the residual failure mode: suspend no longer aborts, but the Ethernet controller still prevents package C10/S0ix entirely. ProblemType: Bug DistroRelease: Ubuntu 24.04 Package: linux-image-7.0.0-28-generic 7.0.0-28.28~24.04.1 ProcVersionSignature: Ubuntu 7.0.0-28.28~24.04.1-generic 7.0.12 Uname: Linux 7.0.0-28-generic x86_64 ApportVersion: 2.28.1-0ubuntu3.8 Architecture: amd64 AudioDevicesInUse: USER PID ACCESS COMMAND /dev/snd/controlC0: jm 3525 F.... pipewire jm 3530 F.... wireplumber /dev/snd/seq: jm 3525 F.... pipewire CRDA: N/A CasperMD5CheckResult: pass CurrentDesktop: ubuntu:GNOME Date: Sat Jul 18 19:12:37 2026 HibernationDevice: RESUME=UUID=48863ccc-2f44-4081-bc90-9d0cca775a3a InstallationDate: Installed on 2025-02-24 (509 days ago) InstallationMedia: Ubuntu 24.04.1 LTS "Noble Numbat" - Release amd64 (20240827.1) MachineType: LENOVO 21G2CTO1WW ProcFB: 0 i915drmfb ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-7.0.0-28-generic root=UUID=48863ccc-2f44-4081-bc90-9d0cca775a3a ro quiet splash i915.enable_psr2_sel_fetch=0 resume=UUID=48863ccc-2f44-4081-bc90-9d0cca775a3a resume_offset=12544000 vt.handoff=7 RelatedPackageVersions: linux-restricted-modules-7.0.0-28-generic N/A linux-backports-modules-7.0.0-28-generic N/A linux-firmware 20240318.git3b128b60-0ubuntu2.27 SourcePackage: linux-hwe-7.0 UpgradeStatus: No upgrade log present (probably fresh install) dmi.bios.date: 04/24/2026 dmi.bios.release: 1.27 dmi.bios.vendor: LENOVO dmi.bios.version: R2DET42W (1.27 ) dmi.board.asset.tag: Not Available dmi.board.name: 21G2CTO1WW dmi.board.vendor: LENOVO dmi.board.version: Not Defined dmi.chassis.asset.tag: No Asset Information dmi.chassis.type: 10 dmi.chassis.vendor: LENOVO dmi.chassis.version: None dmi.ec.firmware.release: 1.18 dmi.modalias: dmi:bvnLENOVO:bvrR2DET42W(1.27):bd04/24/2026:br1.27:efr1.18:svnLENOVO:pn21G2CTO1WW:pvrThinkPadP14sGen5:rvnLENOVO:rn21G2CTO1WW:rvrNotDefined:cvnLENOVO:ct10:cvrNone:skuLENOVO_MT_21G2_BU_Think_FM_ThinkPadP14sGen5:pfaThinkPadP14sGen5: dmi.product.family: ThinkPad P14s Gen 5 dmi.product.name: 21G2CTO1WW dmi.product.sku: LENOVO_MT_21G2_BU_Think_FM_ThinkPad P14s Gen 5 dmi.product.version: ThinkPad P14s Gen 5 dmi.sys.vendor: LENOVO ** Affects: linux-hwe-7.0 (Ubuntu) Importance: Undecided Status: New ** Tags: amd64 apport-bug e1000e meteor-lake noble s2idle suspend -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/2161207 Title: ThinkPad P14s Gen 5: Intel 8086:550a prevents package C10/S0ix unless disabled in BIOS To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux-hwe-7.0/+bug/2161207/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
