Public bug reported: Description: After manually configuring my Ubuntu Server 25.04 system for ultra-low power consumption (idle ~3.2W), I’ve encountered an issue where the CPU no longer enters deep C-states (C8–C10) after a SATA SSD is woken up from standby (e.g., by accessing a file on the drive).
What works (before disk wakeup): CPU reaches deep C-states C8, C9, C10 reliably SATA SSD (crusial CT1000MX) is set to standby via hdparm -y System stays in ultra-low idle power (~3.2W) What breaks (after wakeup): After any disk access wakes up the SATA SSD from standby, CPU only goes as deep as C2 Even after the disk goes back to standby, CPU remains stuck in shallow C-states (C2 max) Requires system reboot to recover full C-state entry What I tried: Using TLP and powertop to re-apply tunables Manually re-enabling autosuspend and PCIe ASPM Checking /sys/devices/system/cpu/cpu*/cpuidle/state*/usage — confirms C8–C10 stop being used When I switched to Debian Trixie with kernel 6.12.22 there were no such problems and the system remained idle on C10 all the time despite frequent disk operations. System Info: Ubuntu Server 25.04 Kernel: 6.14.0-15-generic TLP: 1.8.0 with aggressive power-saving settings CPU: Intel J6412 (Elkhart Lake), ASPM and C10 enabled in BIOS SSD: Crusial CT1000MX 1TB (SATA) How to reproduce: Boot system and ensure CPU enters deep C-states (via powertop or /sys stats) Send SATA SSD to standby: hdparm -y /dev/sdX Wait until system is idle and in deep sleep Wake disk via file read (e.g., cat /mnt/ssd/file) Check CPU C-states — C8–C10 no longer used ** Affects: linux (Ubuntu) Importance: Undecided Status: New ** Attachment added: "apport.linux-image-6.14.0-15-generic.pjqudybo.zip" https://bugs.launchpad.net/bugs/2109897/+attachment/5875606/+files/apport.linux-image-6.14.0-15-generic.pjqudybo.zip -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/2109897 Title: CPU does not return to deep C-states (C8–C10) after SATA SSD wakeup from standby Status in linux package in Ubuntu: New Bug description: Description: After manually configuring my Ubuntu Server 25.04 system for ultra-low power consumption (idle ~3.2W), I’ve encountered an issue where the CPU no longer enters deep C-states (C8–C10) after a SATA SSD is woken up from standby (e.g., by accessing a file on the drive). What works (before disk wakeup): CPU reaches deep C-states C8, C9, C10 reliably SATA SSD (crusial CT1000MX) is set to standby via hdparm -y System stays in ultra-low idle power (~3.2W) What breaks (after wakeup): After any disk access wakes up the SATA SSD from standby, CPU only goes as deep as C2 Even after the disk goes back to standby, CPU remains stuck in shallow C-states (C2 max) Requires system reboot to recover full C-state entry What I tried: Using TLP and powertop to re-apply tunables Manually re-enabling autosuspend and PCIe ASPM Checking /sys/devices/system/cpu/cpu*/cpuidle/state*/usage — confirms C8–C10 stop being used When I switched to Debian Trixie with kernel 6.12.22 there were no such problems and the system remained idle on C10 all the time despite frequent disk operations. System Info: Ubuntu Server 25.04 Kernel: 6.14.0-15-generic TLP: 1.8.0 with aggressive power-saving settings CPU: Intel J6412 (Elkhart Lake), ASPM and C10 enabled in BIOS SSD: Crusial CT1000MX 1TB (SATA) How to reproduce: Boot system and ensure CPU enters deep C-states (via powertop or /sys stats) Send SATA SSD to standby: hdparm -y /dev/sdX Wait until system is idle and in deep sleep Wake disk via file read (e.g., cat /mnt/ssd/file) Check CPU C-states — C8–C10 no longer used To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2109897/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp