Main page
One of the main things still missing in FreeBSD for it to be usable on a modern laptop such as the AMD Framework laptops and the newer Intel ones is the ability to go to sleep. In the past, this was done using something called ACPI S3.
S3 is one of the sleep states that ACPI defines (other examples include S0 when in regular operation and S5 when the computer is fully turned off).
When you tell your machine to go into the S3 sleep state, the acpi_EnterSleepState function is called, which eventually tells your ACPI firmware to put your machine to sleep.
With S3, your firmware is thus responsible for turning off the CPU and other devices when explicitly told to do so.
However, modern laptops have started ditching S3 and using something called S0ix instead. With S0ix, the operating system is the one responsible for figuring out what devices need to be put to sleep before the CPU can be turned off, and the firmware will automatically turn off the CPU only once all the requirements are met.
A fair warning: this article delves into the sombre depths and tedium of ACPI, so it's not the most exciting read.
Also I'm still currently figuring this out, so some of the information here might be incomplete or flat wrong. These are really mostly personal notes.
On FreeBSD, you can query the sleep states your machine supports by reading the hw.acpi.supported_sleep_state sysctl (hw.acpi.suspend_state gives you the sleep state used for suspend).
If you don't see S3 in the list, your machine probably only supports S0ix.
To be sure that your machine indeed does support S0ix, you need to check the FADT flags, specifically AcpiGbl_FADT.Flags & ACPI_FADT_LOW_POWER_S0.
Ben Widawsky from Intel started work on this in 2018 with two patches, D17675 and D17676 (Emulated S3 with s0ix). This work was never finished, however.
_LPI objectThe LPIT (Low Power Idle Table, defined in this Intel spec) is a table that describes the low-power idle states that the CPU supports. These table entries also contain residency counters, which just tell you how long a CPU has spent in a particular low-power state (useful for debugging).
It would seem as though LPIT has gone out of favour since ACPI 6.0. It says as much for ARM. I haven't found any further information on this for AMD, but it does seem like newer Intel devices still have the LPIT table (e.g. the Dell XPS 15 9570).
The ACPI spec now makes no mention of LPIT, but it does look like LPI is its replacement. Unfortunately, they made residency counters for each of these states optional, and it so happens that they are missing on my AMD Framework laptop, which is a little annoying for debugging.
I don't yet know if there are machines which have the LPIT but no _LPI object, but if there are then both will have to be implemented.
For now I'm just looking at reading the _LPI object.
In ACPI-speak, a _DSM (Device Specific Method) is just a command you can send to an ACPI device (with acpi_EvaluateDSMTyped).
Sets of these DSM's are identified by a UUID (which is queried with acpi_DSMQuery).
It seems like, in practice, the original Intel spec linked above is not actually used (with UUID c4eb40a0-6cd2-11e2-bcfd-0800200c9a66), at least on modern Intel or AMD platforms.
Instead, there's DSM UUID 11e00d56-ce64-47ce-837b-1f898f9aa461, which is documented here, and thankfully seems quite similar to the original DSM's, except with a couple extra "Modern Standby" functions:
| Index | Description |
|---|---|
| 0 | Enumerate functions. |
| 1 | Get device constraints (only in the Intel spec). |
| 2 | Get crash dump device (only in the Intel spec). |
| 3 | Display off notification. |
| 4 | Display on notification. |
| 5 | Entry notification. |
| 6 | Exit notification. |
| 7 | "Modern Standby" entry notification. |
| 8 | "Modern Standby" exit notification. |
I will detail what each of these DSM's are for in a bit.
AMD seems to have their own DSM set with UUID e3f32452-febc-43ce-9039-932122d37721, to which the only reference I've been able to find is this post on the Freedesktop mailing list.
This is what they look like:
| Index | Description |
|---|---|
| 0 | Enumerate functions. |
| 1 | Get device constraints. |
| 2 | Entry notification. |
| 3 | Exit notification. |
| 4 | Display off notification. |
| 5 | Display on notification. |
TODO I'm unsure of if I can just ignore this and use the Microsoft DSM's, as my machine reports both UUID's.
I've created a new acpi_lps0 driver, which looks for the PNP0D80 ACPI ID ("Windows-compatible System Power Management Controller"), which is the device on which the above DSM's are defined.
TODO Should I rename this to acpi_spmc like Ben's patches? I only named it acpi_lps0 because I didn't know what SPMC stood for at the time.
Here are the detailed steps for going to sleep with S0ix.
The first step when trying to enter a low-power state is to get all the device constraints required for entering that state in the first place. The "Get device constraints" DSM is used for this, and the output is parsed (for some reason, the AMD DSM's result package follows a different format for which I couldn't find a spec anywhere 🙃).
The output contains a list of all the ACPI devices on the system along with the minimum D-states required for the CPU to go to sleep. Higher D-state correspond to deeper sleep states, D0 means "fully on", and D3 means "fully off".
An acpi_pwr_switch_consumer function already exists to set the D-state of a device, but FreeBSD doesn't currently have an equivalent acpi_pwr_get_consumer function, which is necessary to know which are the constraining devices exactly.
ACPI defines multiple ways of getting the D-state of a device.
The first and simplest is through the _PSC (power state current, ACPI 7.3.6) control method, which simply spits out the device's D-state when evaluated.
_PSC isn't implemented for all devices, however:
This control method is not required if the device state can be inferred by the Power Resource settings. This would be the case when the device does not require a _PS0, _PS1, _PS2, or _PS3 control method.
The "Power Resource settings" the spec mentions are the _PR0, _PR1, _PR2, and _PR3 objects (ACPI 7.3.8 - 7.3.11), which evaluate to the power requirements (i.e. a list of power resource objects) for the device to enter the respective D-state.
From this, the process of inferring the D-state of the device is as follows:
_PRx object for that D-state._STA object (ACPI 7.2.4), which lets us know if it is on (0) or off (1).I've implemented this through the acpi_pwr_infer_state function, which is called as a fallback to _PSC in acpi_pwr_get_consumer.
Each device which is constraining the CPU from going to sleep (i.e. their D-state is lower than the minimum D-state required) needs to be put to sleep.
I think this is done by calling the acpi_pwr_switch_consumer function.
TODO Actually implement this. I'm going to need to work on a bit more infrastructure and understand the sleep process in FreeBSD before I can do this.
Now, the system should be ready to go to sleep.
TODO This is easy.
CPU power states are known as C-states. To put it to sleep, we need to set it to its maximum C-state and, if we did the previous setup right, the hardware will hopefully take care of the rest.
MWAIT can do this for us. It's an x86 instruction that's usually used in conjunction with MONITOR to enter an "implementation-dependent optimized state" and wait until a specific memory range is written to.
If CPUID.05H:ECX[bit 0] is set, it can also be used for power management.
Specifically, eax can be set to contain hints to MWAIT and ecx (extension) can be set to contain the C-state that the processor should enter.
For our purposes, we can set the lowest bit of eax to 1 to allow for interrupts to break out of MWAIT (i.e. wake the CPU up).
Thanks to this, we can forgo the need to set up a memory range to monitor.
Bits 7 to 4 of ecx are used to specify the target C-state to enter (we can ignore the lowest 4 bits which are for "sub C-states").
All in all, this is what that looks like on FreeBSD:
#include <machine/cpufunc.h>
cpu_mwait(MWAIT_INTRBREAK, MWAIT_C3); // TODO What's the maximum C-state?
TODO
S4 is really just S3 with the extra step of writing the contents of memory to disk before fully powering off. This essentially gives you the power savings of S5 while still having a relatively small wake latency.
TODO Modern standby.