Power management/Suspend and hibernate

From ArchWiki
Jump to navigation Jump to search

Currently there are three methods of suspending available: suspend to RAM (usually called just suspend), suspend to disk (usually known as hibernate), and hybrid suspend (sometimes aptly called suspend to both):

  • Suspend to RAM method cuts power to most parts of the machine aside from the RAM, which is required to restore the machine's state. Because of the large power savings, it is advisable for laptops to automatically enter this mode when the computer is running on batteries and the lid is closed (or the user is inactive for some time).
  • Suspend to disk method saves the machine's state into swap space and completely powers off the machine. When the machine is powered on, the state is restored. Until then, there is zero power consumption.
  • Suspend to both method saves the machine's state into swap space, but does not power off the machine. Instead, it invokes usual suspend to RAM. Therefore, if the battery is not depleted, the system can resume from RAM. If the battery is depleted, the system can be resumed from disk, which is much slower than resuming from RAM, but the machine's state has not been lost.

There are multiple low level interfaces (backends) providing basic functionality, and some high level interfaces providing tweaks to handle problematic hardware drivers/kernel modules (e.g. video card re-initialization).

Low level interfaces

Though these interfaces can be used directly, it is advisable to use some of high level interfaces to suspend/hibernate. Using low level interfaces directly is significantly faster than using any high level interface, since running all the pre- and post-suspend hooks takes time, but hooks can properly set hardware clock, restore wireless etc.

kernel (swsusp)

The most straightforward approach is to directly inform the in-kernel software suspend code (swsusp) to enter a suspended state; the exact method and state depends on the level of hardware support. On modern kernels, writing appropriate strings to /sys/power/state is the primary mechanism to trigger this suspend.

See kernel documentation for details.


The uswsusp ('Userspace Software Suspend') is a wrapper around the kernel's suspend-to-RAM mechanism, which performs some graphics adapter manipulations from userspace before suspending and after resuming.

See main article Uswsusp.

High level interfaces

The end goal of these packages is to provide binaries/scripts that can be invoked to perform suspend/hibernate. Actually hooking them up to power buttons or menu clicks or laptop lid events is usually left to other tools. To automatically suspend/hibernate on certain power events, such as laptop lid close or battery depletion percentage, you may want to look into running Acpid.


systemd provides native commands for suspend, hibernate and a hybrid suspend, see Power management#Power management with systemd for details. This is the default interface used in Arch Linux.

See Power management#Sleep hooks for additional information on configuring suspend/hibernate hooks. Also see systemctl(1), systemd-sleep(8), and systemd.special(7).


In order to use hibernation, you need to create a swap partition or file. You will need to point the kernel to your swap using the resume= kernel parameter, which is configured via the boot loader. You will also need to configure the initramfs. This tells the kernel to attempt resuming from the specified swap in early userspace. These three steps are described in detail below.


About swap partition/file size

Even if your swap partition is smaller than RAM, you still have a big chance of hibernating successfully. According to kernel documentation:

/sys/power/image_size controls the size of the image created by the suspend-to-disk mechanism. It can be written a string representing a non-negative integer that will be used as an upper limit of the image size, in bytes. The suspend-to-disk mechanism will do its best to ensure the image size will not exceed that number. However, if this turns out to be impossible, it will try to suspend anyway using the smallest image possible. In particular, if "0" is written to this file, the suspend image will be as small as possible. Reading from this file will display the current image size limit, which is set to 2/5 of available RAM by default.

You may either decrease the value of /sys/power/image_size to make the suspend image as small as possible (for small swap partitions), or increase it to possibly speed up the hibernation process. For systems with a large amount of RAM, smaller values may drastically increase the speed of resuming a hibernating system. See systemd#systemd-tmpfiles - temporary files to make this change persistent.

The suspend image cannot span multiple swap partitions and/or swap files. It must fully fit in one swap partition or one swap file.[1]

Required kernel parameters

The kernel parameter resume=swap_device must be used. Any of the persistent block device naming methods can be used as swap_device. For example:

  • resume=UUID=4209c845-f495-4c43-8a03-5363dd433153
  • resume="PARTLABEL=Swap partition"
  • resume=/dev/archVolumeGroup/archLogicalVolume -- if swap is on a LVM logical volume

The kernel parameters will only take effect after rebooting. To be able to hibernate right away, obtain the volume's major and minor device numbers from lsblk and echo them in format major:minor to /sys/power/resume. If using a swap file, additionally echo the resume offset to /sys/power/resume_offset.[2]

For example, if the swap device is 8:3:

# echo 8:3 > /sys/power/resume

Or when hibernating to a swap file, if the swap file is on volume 8:2 and has the offset 38912:

# echo 8:2 > /sys/power/resume
# echo 38912 > /sys/power/resume_offset

Hibernation into swap file

Warning: Btrfs on Linux kernel before version 5.0 does not support swap files. Failure to heed this warning may result in file system corruption. While a swap file may be used on Btrfs when mounted through a loop device, this will result in severely degraded swap performance.

Using a swap file requires also setting the resume=swap_device and additionally a resume_offset=swap_file_offset kernel parameters. See the kernel documentation.

swap_device is the the volume where the swap file resides and it follows the same format as for the root parameter. The value of swap_file_offset can be obtained by running filefrag -v swap_file, the output is in a table format and the required value is located in the first row of the physical_offset column. For example:

# filefrag -v /swapfile
Filesystem type is: ef53
File size of /swapfile is 4294967296 (1048576 blocks of 4096 bytes)
 ext:     logical_offset:        physical_offset: length:   expected: flags:
   0:        0..       0:      38912..     38912:      1:            
   1:        1..   22527:      38913..     61439:  22527:             unwritten
   2:    22528..   53247:     899072..    929791:  30720:      61440: unwritten

In the example the value of swap_file_offset is the first 38912 with the two periods.

  • The following command may be used to identify swap_device: findmnt -no UUID -T /swapfile
  • The following command may be used to identify swap_file_offset: filefrag -v /swapfile | awk '{ if($1=="0:"){print $4} }'
  • The value of swap_file_offset can also be obtained by running swap-offset swap_file. The swap-offset binary is provided within the set of tools uswsusp. If using this method, then these two parameters have to be provided in /etc/suspend.conf via the keys resume device and resume offset. No reboot is required in this case.
  • For a stacked block device such as an encrypted container (LUKS), RAID or LVM, the resume parameter must point to the unlocked/mapped device that contains the file system with the swap file.
  • If the swap file is in /home/, systemd-logind will not be able to determine its size and thus will prevent hibernation. See systemd issue 15354 for a workaround.
Tip: You might want to decrease the swappiness for your swap file if the only purpose is to be able to hibernate and not expand RAM.

Hibernation into swap file on Btrfs

Hibernation onto a swap file is supported by recent versions of systemd [3].

The resume_offset number can be computed using the tool btrfs_map_physical.c. Do not try to use the filefrag tool, on Btrfs the "physical" offset you get from filefrag is not the real physical offset on disk; there is a virtual disk address space in order to support multiple devices. [4]

Download or copy the tool btrfs_map_physical.c into a file named btrfs_map_physical.c, then compile it,

$ gcc -O2 -o btrfs_map_physical btrfs_map_physical.c

and run it. An example output is shown below.

# ./btrfs_map_physical /path/to/swapfile
0            regular      4096          2927632384      268435456      1      4009762816
4096         prealloc     268431360     2927636480      268431360      1      4009766912
268435456    prealloc     268435456     3251634176      268435456      1      4333764608
536870912    prealloc     268435456     3520069632      268435456      1      4602200064
805306368    prealloc     268435456     3788505088      268435456      1      4870635520
1073741824   prealloc     268435456     4056940544      268435456      1      5139070976
1342177280   prealloc     268435456     4325376000      268435456      1      5407506432
1610612736   prealloc     268435456     4593811456      268435456      1      5675941888

Note the the first physical offset returned by this tool. In this example, we use 4009762816. Also note the pagesize that can be found with getconf PAGESIZE.

To compute the resume_offset value, divide the physical offset by the pagesize. In this example, it is 4009762816 / 4096 = 978946.

Hibernation into a thinly-provisioned LVM volume

Hibernation into a thinly-provisioned LVM volume is possible, but you have to make sure that the volume is fully allocated. Otherwise resuming from it will fail.

You can fully allocate the LVM volume by simply filling it with zeros. E.g.:

# dd if=/dev/zero of=/dev/vg0/swap bs=1M status=progress

To verify the volume is fully allocated, you can use:

# lvs
 LV                   VG  Attr       LSize   Pool Origin    Data%  Meta%  Move Log Cpy%Sync Convert
 swap                 vg0 Vwi-aot--- 10.00g  pool           100

A fully allocated volume will show up as having 100% data usage.

Warning: Do not to use TRIM on thinly-provisioned swap volumes that are used for hibernation, i.e. do not use discard in /etc/fstab and the -d/--discard option of swapon. Otherwise the used space will be deallocated.

Configure the initramfs

  • When an initramfs with the base hook is used, which is the default, the resume hook is required in /etc/mkinitcpio.conf. Whether by label or by UUID, the swap partition is referred to with a udev device node, so the resume hook must go after the udev hook. This example was made starting from the default hook configuration:
HOOKS=(base udev autodetect keyboard modconf block filesystems resume fsck)
Remember to regenerate the initramfs for these changes to take effect.
Note: LVM users should add the resume hook after lvm2.
  • When an initramfs with the systemd hook is used, a resume mechanism is already provided, and no further hooks need to be added.



You might want to tweak your DSDT table to make it work. See DSDT article

Suspend/hibernate does not work, or does not work consistently

There have been many reports about the screen going black without easily viewable errors or the ability to do anything when going into and coming back from suspend and/or hibernate. These problems have been seen on both laptops and desktops. This is not an official solution, but switching to an older kernel, especially the LTS-kernel, will probably fix this.

Also problem may arise when using hardware watchdog timer (disabled by default, see RuntimeWatchdogSec= in systemd-system.conf(5) § OPTIONS). Bugged watchdog timer may reset the computer before the system finished creating the hibernation image.

Sometimes the screen goes black due to device initialization from within the initramfs. Removing any modules you might have in Mkinitcpio#MODULES and rebuilding the initramfs, can possibly solve this issue, specially graphics drivers for early KMS. Initializing such devices before resuming can cause inconsistencies that prevents the system resuming from hibernation. This does not affect resuming from RAM. Also, check the blog article best practices to debug suspend issues.

Moving from the radeon video driver to the newer AMDGPU driver could also help to make the hibernation and awakening process successful.

For Intel graphics drivers, enabling early KMS may help to solve the blank screen issue. Refer to Kernel mode setting#Early KMS start for details.

After upgrading to kernel 4.15.3, resume may fail with a static (non-blinking) cursor on a black screen. Blacklisting the module nvidiafb might help. [5]

Laptops with Intel CPU that load intel_lpss_pci module for touchpad, may face kernel panic on resume (blinking caps lock) [6]. The module needs to be added to initramfs as:

MODULES=(... intel_lpss_pci ...)

Then regenerate the initramfs.


If Wake-on-LAN is active, the network interface card will consume power even if the computer is hibernated.

Instantaneous wakeups from suspend

For some Intel Haswell systems with the LynxPoint and LynxPoint-LP chipset, instantaneous wakeups after suspend are reported. They are linked to erroneous BIOS ACPI implementations and how the xhci_hcd module interprets it during boot. As a work-around reported affected systems are added to a blacklist (named XHCI_SPURIOUS_WAKEUP) by the kernel case-by-case.[7]

Instantaneous resume may happen, for example, if a USB device is plugged during suspend and ACPI wakeup triggers are enabled. A viable work-around for such a system, if it is not on the blacklist yet, is to disable the wakeup triggers. An example to disable wakeup through USB is described as follows.[8]

To view the current configuration:

$ cat /proc/acpi/wakeup
Device  S-state   Status   Sysfs node
EHC1      S3    *enabled  pci:0000:00:1d.0
EHC2      S3    *enabled  pci:0000:00:1a.0
XHC       S3    *enabled  pci:0000:00:14.0

The relevant devices are EHC1, EHC2 and XHC (for USB 3.0). To toggle their state you have to echo the device name to the file as root.

# echo EHC1 > /proc/acpi/wakeup
# echo EHC2 > /proc/acpi/wakeup
# echo XHC > /proc/acpi/wakeup

This should result in suspension working again. However, this settings are only temporary and would have to be set at every reboot. To automate this take a look at systemd#systemd-tmpfiles - temporary files or see BBS thread for a possible solution and more information.

Example solution with disabling PTXH and XHC0 at the same time. For some reason, two lines with PTXH and XHC0 one per line or in different files does not work.

#  cat /etc/tmpfiles.d/100-disable-usb-wake.conf
#    Path                  Mode UID  GID  Age Argument
w    /proc/acpi/wakeup     -    -    -    -   PTXHXHC0

If you use nouveau driver, the reason of instantaneous wakeup may be a bug in that driver, which sometimes prevents graphics card from suspension. One possible workaround is unloading nouveau kernel module right before going to sleep and loading it back after wakeup. To do this, create the following script:


case $1/$2 in
    # echo "Going to $2..."
    /usr/bin/echo "0" > /sys/class/vtconsole/vtcon1/bind
    /usr/bin/rmmod nouveau
    # echo "Waking up from $2..."
    /usr/bin/modprobe nouveau
    /usr/bin/echo "1" > /sys/class/vtconsole/vtcon1/bind

The first echo line unbinds nouveaufb from the framebuffer console driver (fbcon). Usually it is vtcon1 as in this example, but it may also be another vtcon*. See /sys/class/vtconsole/vtcon*/name which one of them is a "frame buffer device" [9].

System does not power off when hibernating

When you hibernate your system, the system should power off (after saving the state on the disk). Sometimes, you might see the power LED is still glowing. If that happens, it might be instructive to set the HibernateMode to shutdown in sleep.conf.d(5):


With the above configuration, if everything else is set up correctly, on invocation of a systemctl hibernate the machine will shutdown saving state to disk as it does so.