[SOLVED] Error message during shutdown

[manmath]

I see the following error message during shutdown:
fw_domains_get_normal [i915] *error* render: timed out waiting for forcewa
Is there a way to fix it? Is it just a warning that I should not worry about?

[rab0171610]

Try sudo apt-get install lm-sensors -y
After it installs, open a terminal window and enter sudo sensors-detect and answer yes to the prompts.

[manmath]

Did. The error message still shows up after multiple reboots.

[rab0171610]

Good to know. Which kernel are you using? Also, which CPU do you have? Is this an older PC?

I know that this was an issue with the Intel CPU several years back. Reading a patch for the kernel at that time I see:

/* IVB configs may use multi-threaded forcewake */

+

+ /* A small trick here - if the bios hasn't configured

+ * MT forcewake, and if the device is in RC6, then

+ * force_wake_mt_get will not wake the device and the

+ * ECOBUS read will return zero. Which will be

+ * (correctly) interpreted by the test below as MT

+ * forcewake being disabled.

+ */

It refers to Multi Threading forced wake. RC6:

Intel RC6 power-savings can yield a significant reduction in power while at the same time it's able to boost the graphics core performance. Supporting RC6 has been a mess for the Intel OTC developers since for a small number of Sandy Bridge systems there have been stability issues and other problems when having the feature enabled. RC6 is a new feature to Sandy Bridge

I am wondering if there are any settings in the bios that refer to power saving, the cpu, SMP, multithreading, etc. that can be changed to prevent this error message. It seems it is referring to waking the CPU from a power saving state. Check the bios and see if there is a setting that might affect this and disable it. Just a suggestion to try.

Good Luck!

[manmath]

Good to know. Which kernel are you using? Also, which CPU do you have? Is this an older PC?

Mine is a decade-old system with Ivy Bridge (i5-3330 CPU) on Gigabyte H61 mobo. Here's neofetch report:

OS: Kubuntu 22.04 LTS x86_64
Host: H61M-DS2
Kernel: 5.17.3-zen1
Uptime: 5 mins
Packages: 1674 (dpkg)
Shell: bash 5.1.16
Resolution: 1920x1080
DE: Plasma 5.24.4
WM: KWin
Theme: [Plasma], Breeze [GTK2/3]
Icons: [Plasma], breeze [GTK2/3]
Terminal: konsole
CPU: Intel i5-3330 (4) @ 3.101GHz
GPU: Intel HD Graphics
Memory: 1085MiB / 7881MiB

The only change I made in the BIOS is switched to AHCI from IDE in the SATA section. Also I'm using Zen kernel instead of stock Ubuntu kernel.

[rab0171610]

Under Advanced CPU Core Feature:
Some settings for that motherboard you might want to check/toggle to help troubleshoot:

Internal CPU PLL Overvoltage

Enabled allows CPU PLL voltage to operate at a higher value.Disabled allows CPU PLL voltage to operate
at default value.Auto lets the BIOS automatically configure this setting. (Default: Auto)
Real-Time CPU Ratio Control In OS&(Note)
Enabled allows you to make real-time changes to the CPU clock ratio in your operating system. (Default:
Disabled)
Intel(R) Turbo Boost Technology&(Note)
Allows you to determine whether to enable the Intel CPU Turbo Boost technology.Auto lets the BIOS
automatically configure this setting. (Default: Auto)
Turbo Ratio (1-Core Active~4-Core Active)&(Note)
Allows you to set the CPU Turbo ratios for different number of active cores.Auto sets the CPU Turbo ratios
according to the CPU specifications. (Default: Auto)
Turbo Power Limit (Watts)&
Allows you to set a power limit for CPU Turbo mode. When the CPU power consumption exceeds the
specified power limit, the CPU will automatically reduce the core frequency in order to reduce the power.
Auto sets the power limit according to the CPU specifications. (Default: Auto)
Core Current Limit (Amps)&
Allows you to set a current limit for CPU Turbo mode. When the CPU current exceeds the specified current
limit, the CPU will automatically reduce the core frequency in order to reduce the current.Auto sets the
current limit according to the CPU specifications. (Default: Auto)
CPU Core Enabled&(Note)
Allows you to determine the number of CPU cores you want to enable.Auto lets the BIOS automatically
configure this setting. (Default: Auto)
Hyper-Threading Technology&(Note)
Allows you to determine whether to enable Hyper-Threading technology when using an Intel CPU that
supports this function. This feature only works for operating systems that support multi-processor mode.
Auto lets the BIOS automatically configure this setting. (Default: Auto)
CPU Enhanced Halt (C1E)&(Note)
Enables or disables Intel CPU Enhanced Halt (C1E) function, a CPU power-saving function in system
halt state. When enabled, the CPU core frequency and voltage will be reduced during system halt state to
decrease power consumption.Auto lets the BIOS automatically configure this setting. (Default: Auto)
C3/C6 State Support&(Note)
Allows you to determine whether to let the CPU enter C3/C6 mode in system halt state. When enabled, the
CPU core frequency and voltage will be reduced during system halt state to decrease power consumption.
The C3/C6 state is a more enhanced power-saving state than C1.Auto lets the BIOS automatically configure
this setting. (Default: Auto)
CPU Thermal Monitor&(Note)
Enables or disables Intel CPU Thermal Monitor function, a CPU overheating protection function. When
enabled, the CPU core frequency and voltage will be reduced when the CPU is overheated.Auto lets the
BIOS automatically configure this setting. (Default: Auto

[rab0171610]

Hopefully someone else will be able to make sense of this. This is from the kernel code for the driver:


linux/drivers/gpu/drm/i915/intel_uncore.c
https://github.com/torvalds/linux/bl...intel_uncore.c
The kernel driver for this CPU uses 'render' as one of the domain names variables.

static const char * const forcewake_domain_names[] = {

"render",

"gt",

"media",

Also the specific driver code that instructs to return this error that you are receiving:

static inline void

fw_domain_get(const struct intel_uncore_forcewake_domain *d)

{

fw_set(d, FORCEWAKE_KERNEL);

}

static inline void

fw_domain_wait_ack_set(const struct intel_uncore_forcewake_domain *d)

{

if (wait_ack_set(d, FORCEWAKE_KERNEL)) {

DRM_ERROR("%s: timed out waiting for forcewake ack request.\n",

intel_uncore_forcewake_domain_to_str(d->id));

add_taint_for_CI(d->uncore->i915, TAINT_WARN); /* CI now unreliable */

}

}

Also an interesting comment in the driver code:

/*

* There is a possibility of driver's wake request colliding

* with hardware's own wake requests and that can cause

* hardware to not deliver the driver's ack message.

*

* Use a fallback bit toggle to kick the gpu state machine

* in the hope that the original ack will be delivered along with

* the fallback ack.

*

* This workaround is described in HSDES #1604254524 and it's known as:

* WaRsForcewakeAddDelayForAck:skl,bxt,kbl,glk,cfl,cn l,icl

* although the name is a bit misleading.

*/

Hopefully there is a fix but I know this was an issue back in the day and they made many patches and code fixes to the driver. I am posting all this in hopes that someone with some experience with coding or this particular driver will know more than I do and maybe point you in the right direction. Hopefully it is just a setting in the bios related to the cpu power saving features but it could just be a buggy driver.

In the meantime, you could try setting all of the Advanced Cpu Features in the bios to the default settings above and toggling them one by one to see if the error goes away. Considering the comment in the code about problems when the hardware and driver can both send colliding wake acknowledgement requests, I would start by toggling the first one:

Real-Time CPU Ratio Control In OS&(Note)
Enabled allows you to make real-time changes to the CPU clock ratio in your operating system. (Default:
Disabled)

Since it is disabled by default, try enabling it to see if this allows the kernel driver to have some control.
Good Luck!

[manmath]

Thanks for all support. I did try them all, but still some how that error message persisted. As a last resort I switched to stock kernel, and the error message is gone. Might be something was wrong with Zen kernel on my hardware.

[rab0171610]

Well I am glad you found a solution. I learned something in the process as well! It is admirable that you are able to keep your older hardware still working and useful. Hopefully if anyone else has this problem they will see this post.

[manmath]

It is admirable that you are able to keep your older hardware still working and useful. Hopefully if anyone else has this problem they will see this post.

Definitely the post is helpful. I learnt some important points too. In some time, I will also look into the issue with Zen kernel on a separate drive. Thanks for the information.