Skip to contentSkip to conversation

ACPI: OSL: Poweroff when encountering a fatal ACPI error

PATCH In review Armin Wolf 2 participants 9 messages Feb 4 – Mar 1

Subsystems (2)
lore
Share & tools
lore

Armin Wolf

author
#0 lore raw

PATCH v3 1/1 ACPI: OSL: Poweroff when encountering a fatal ACPI error

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

---
Changes since v2:
- poweroff instead of triggering a kernel panic

Changes since v1:
- use IS_ENABLED() for checking the presence of CONFIG_ACPI_PANIC_ON_FATAL

---
 .../admin-guide/kernel-parameters.txt         |  9 +++++++++
 drivers/acpi/Kconfig                          | 11 +++++++++++
 drivers/acpi/osl.c                            | 19 ++++++++++++++++++-
 3 files changed, 38 insertions(+), 1 deletion(-)
Documentation/admin-guide/kernel-parameters.txt +9 -0
index 1058f2a6d6a8..1f2eaa0ec424 100644
@@ -187,6 +187,15 @@ Kernel parameters
 			unusable.  The "log_buf_len" parameter may be useful
 			if you need to capture more output.
 
+	acpi.poweroff_on_fatal=	[ACPI]
+			{0 | 1}
+			Causes the system to poweroff when the ACPI bytecode signals
+			a fatal error. The default value of this setting can
+			be configured using CONFIG_ACPI_POWEROFF_ON_FATAL.
+			Overriding this value should only be done for diagnosing
+			ACPI firmware problems, as the system might behave erratically
+			after having encountered a fatal ACPI error.
+
 	acpi_enforce_resources=	[ACPI]
 			{ strict | lax | no }
 			Check for resource conflicts between native drivers
drivers/acpi/Kconfig +11 -0
index df0ff0764d0d..1610dd4c8278 100644
@@ -65,6 +65,17 @@ config ACPI_THERMAL_LIB
        depends on THERMAL
        bool
 
+config ACPI_POWEROFF_ON_FATAL
+	bool "Poweroff on fatal ACPI error"
+	default y
+	help
+	  The ACPI bytecode can signal that a fatal error has occurred using the Fatal()
+	  ASL operator, normaly causing the system to poweroff. Disabling this option causes
+	  such a condition to be treated like a ordinary ACPI error.
+
+	  This setting can also be overridden during boot using the acpi.poweroff_on_fatal
+	  kernel parameter.
+
 config ACPI_DEBUGGER
 	bool "AML debugger interface"
 	select ACPI_DEBUG
drivers/acpi/osl.c +18 -1
index 05393a7315fe..f2b45fa4a752 100644
@@ -11,8 +11,10 @@
 
 #define pr_fmt(fmt) "ACPI: OSL: " fmt
 
+#include <linux/kconfig.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
+#include <linux/reboot.h>
 #include <linux/slab.h>
 #include <linux/mm.h>
 #include <linux/highmem.h>
@@ -70,6 +72,10 @@ static bool acpi_os_initialized;
 unsigned int acpi_sci_irq = INVALID_ACPI_IRQ;
 bool acpi_permanent_mmap = false;
 
+static bool poweroff_on_fatal = IS_ENABLED(CONFIG_ACPI_POWEROFF_ON_FATAL);
+module_param(poweroff_on_fatal, bool, 0);
+MODULE_PARM_DESC(poweroff_on_fatal, "Poweroff when encountering a fatal ACPI error");
+
 /*
  * This list of permanent mappings is for memory that may be accessed from
  * interrupt context, where we can't do the ioremap().
@@ -1381,9 +1387,20 @@ acpi_status acpi_os_notify_command_complete(void)
 
 acpi_status acpi_os_signal(u32 function, void *info)
 {
+	struct acpi_signal_fatal_info *fatal_info;
+
 	switch (function) {
 	case ACPI_SIGNAL_FATAL:
-		pr_err("Fatal opcode executed\n");
+		fatal_info = info;
+		pr_emerg("Fatal error while evaluating ACPI control method\n");
+		pr_emerg("Type 0x%X Code 0x%X Argument 0x%X\n",
+			 fatal_info->type, fatal_info->code, fatal_info->argument);
+
+		if (poweroff_on_fatal)
+			orderly_poweroff(true);
+		else
+			add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
+
 		break;
 	case ACPI_SIGNAL_BREAKPOINT:
 		/*
Signature · 1 line
2.39.5
author
#1 lore raw

Am 04.02.26 um 22:29 schrieb Armin Wolf:

Quoted 6 lines

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

Any updates on this?

Thanks,
Armin Wolf

Quoted 104 lines

---
Changes since v2:
- poweroff instead of triggering a kernel panic

Changes since v1:
- use IS_ENABLED() for checking the presence of CONFIG_ACPI_PANIC_ON_FATAL

---
  .../admin-guide/kernel-parameters.txt         |  9 +++++++++
  drivers/acpi/Kconfig                          | 11 +++++++++++
  drivers/acpi/osl.c                            | 19 ++++++++++++++++++-
  3 files changed, 38 insertions(+), 1 deletion(-)
Documentation/admin-guide/kernel-parameters.txt +9 -0
index 1058f2a6d6a8..1f2eaa0ec424 100644
@@ -187,6 +187,15 @@ Kernel parameters
  			unusable.  The "log_buf_len" parameter may be useful
  			if you need to capture more output.
 
+	acpi.poweroff_on_fatal=	[ACPI]
+			{0 | 1}
+			Causes the system to poweroff when the ACPI bytecode signals
+			a fatal error. The default value of this setting can
+			be configured using CONFIG_ACPI_POWEROFF_ON_FATAL.
+			Overriding this value should only be done for diagnosing
+			ACPI firmware problems, as the system might behave erratically
+			after having encountered a fatal ACPI error.
+
  	acpi_enforce_resources=	[ACPI]
  			{ strict | lax | no }
  			Check for resource conflicts between native drivers
drivers/acpi/Kconfig +11 -0
index df0ff0764d0d..1610dd4c8278 100644
@@ -65,6 +65,17 @@ config ACPI_THERMAL_LIB
         depends on THERMAL
         bool
 
+config ACPI_POWEROFF_ON_FATAL
+	bool "Poweroff on fatal ACPI error"
+	default y
+	help
+	  The ACPI bytecode can signal that a fatal error has occurred using the Fatal()
+	  ASL operator, normaly causing the system to poweroff. Disabling this option causes
+	  such a condition to be treated like a ordinary ACPI error.
+
+	  This setting can also be overridden during boot using the acpi.poweroff_on_fatal
+	  kernel parameter.
+
  config ACPI_DEBUGGER
  	bool "AML debugger interface"
  	select ACPI_DEBUG
drivers/acpi/osl.c +18 -1
index 05393a7315fe..f2b45fa4a752 100644
@@ -11,8 +11,10 @@
 
  #define pr_fmt(fmt) "ACPI: OSL: " fmt
 
+#include <linux/kconfig.h>
  #include <linux/module.h>
  #include <linux/kernel.h>
+#include <linux/reboot.h>
  #include <linux/slab.h>
  #include <linux/mm.h>
  #include <linux/highmem.h>
@@ -70,6 +72,10 @@ static bool acpi_os_initialized;
  unsigned int acpi_sci_irq = INVALID_ACPI_IRQ;
  bool acpi_permanent_mmap = false;
 
+static bool poweroff_on_fatal = IS_ENABLED(CONFIG_ACPI_POWEROFF_ON_FATAL);
+module_param(poweroff_on_fatal, bool, 0);
+MODULE_PARM_DESC(poweroff_on_fatal, "Poweroff when encountering a fatal ACPI error");
+
  /*
   * This list of permanent mappings is for memory that may be accessed from
   * interrupt context, where we can't do the ioremap().
@@ -1381,9 +1387,20 @@ acpi_status acpi_os_notify_command_complete(void)
 
  acpi_status acpi_os_signal(u32 function, void *info)
  {
+	struct acpi_signal_fatal_info *fatal_info;
+
  	switch (function) {
  	case ACPI_SIGNAL_FATAL:
-		pr_err("Fatal opcode executed\n");
+		fatal_info = info;
+		pr_emerg("Fatal error while evaluating ACPI control method\n");
+		pr_emerg("Type 0x%X Code 0x%X Argument 0x%X\n",
+			 fatal_info->type, fatal_info->code, fatal_info->argument);
+
+		if (poweroff_on_fatal)
+			orderly_poweroff(true);
+		else
+			add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
+
  		break;
  	case ACPI_SIGNAL_BREAKPOINT:
  		/*
#2 lore raw
Quoted 11 lines · Armin Wolf

Am 04.02.26 um 22:29 schrieb Armin Wolf:

Quoted 6 lines

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

Any updates on this?

I was about to apply it, but then I thought that I was not sure about the Kconfig option.

I don't see much value in it TBH. If you agree, I'll just apply the
patch without that part.

Thanks!

author
#3 lore raw

Am 25.02.26 um 22:28 schrieb Rafael J. Wysocki:

Quoted 9 lines · Armin Wolf
Quoted 8 lines

Am 04.02.26 um 22:29 schrieb Armin Wolf:

Quoted 6 lines

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

Quoted 8 lines
Quoted 1 line

Any updates on this?

I was about to apply it, but then I thought that I was not sure about the Kconfig option.

I don't see much value in it TBH. If you agree, I'll just apply the
patch without that part.

Thanks!

You can drop the Kconfig option if you want.

Thanks,
Armin Wolf

#4 lore raw
Quoted 16 lines · Armin Wolf

Am 25.02.26 um 22:28 schrieb Rafael J. Wysocki:

Quoted 13 lines · Armin Wolf
Quoted 9 lines

Am 04.02.26 um 22:29 schrieb Armin Wolf:

Quoted 6 lines

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

Any updates on this?

I was about to apply it, but then I thought that I was not sure about the Kconfig option.

Quoted 6 lines
Quoted 4 lines

I don't see much value in it TBH. If you agree, I'll just apply the
patch without that part.

Thanks!

You can drop the Kconfig option if you want.

OK, applied as 7.1 material.

Please double check
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/commit/?h=bleeding-edge&id=6f09a7009a1d7a132ddce3a8dd0c46aac66ad8e2

author
#5 lore raw

Am 26.02.26 um 19:46 schrieb Rafael J. Wysocki:

Quoted 19 lines · Armin Wolf
Quoted 18 lines

Am 25.02.26 um 22:28 schrieb Rafael J. Wysocki:

Quoted 16 lines · Armin Wolf
Quoted 9 lines

Am 04.02.26 um 22:29 schrieb Armin Wolf:

Quoted 6 lines

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

Any updates on this?

I was about to apply it, but then I thought that I was not sure about the Kconfig option.

I don't see much value in it TBH. If you agree, I'll just apply the
patch without that part.

Quoted 6 lines
Quoted 2 lines
Quoted 1 line

Thanks!

You can drop the Kconfig option if you want.

OK, applied as 7.1 material.

Please double check
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/commit/?h=bleeding-edge&id=6f09a7009a1d7a132ddce3a8dd0c46aac66ad8e2

Looks good to me, except that the include of kconfig.h should be dropped now.

Thanks,
Armin Wolf

author
#6 lore raw

Am 27.02.26 um 20:50 schrieb Armin Wolf:

Quoted 25 lines

Am 26.02.26 um 19:46 schrieb Rafael J. Wysocki:

Quoted 23 lines · Armin Wolf
Quoted 12 lines

Am 25.02.26 um 22:28 schrieb Rafael J. Wysocki:

Quoted 10 lines · Armin Wolf
Quoted 9 lines

Am 04.02.26 um 22:29 schrieb Armin Wolf:

Quoted 6 lines

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

Any updates on this?

Quoted 8 lines
Quoted 7 lines

I was about to apply it, but then I thought that I was not sure about the Kconfig option.

   I don't see much value in it TBH.  If you agree, I'll just apply the patch without that part.

Thanks!

You can drop the Kconfig option if you want.

OK, applied as 7.1 material.

Quoted 9 lines
Quoted 3 lines

Please double check
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/commit/?h=bleeding-edge&id=6f09a7009a1d7a132ddce3a8dd0c46aac66ad8e2

Looks good to me, except that the include of kconfig.h should be dropped now.

Thanks,
Armin Wolf

I also just noticed that we still need to include panic.h for add_taint(). Should i send a new revision?

Thanks,
Armin Wolf

#7 lore raw
Quoted 33 lines · Armin Wolf

Am 27.02.26 um 20:50 schrieb Armin Wolf:

Quoted 30 lines

Am 26.02.26 um 19:46 schrieb Rafael J. Wysocki:

Quoted 27 lines · Armin Wolf
Quoted 15 lines

Am 25.02.26 um 22:28 schrieb Rafael J. Wysocki:

Quoted 13 lines · Armin Wolf
Quoted 9 lines

Am 04.02.26 um 22:29 schrieb Armin Wolf:

Quoted 6 lines

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

Any updates on this?

I was about to apply it, but then I thought that I was not sure about the Kconfig option.

Quoted 6 lines
Quoted 5 lines

I don't see much value in it TBH. If you agree, I'll just apply
the
patch without that part.

Thanks!

You can drop the Kconfig option if you want.

OK, applied as 7.1 material.

Please double check
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/commit/?h=bleeding-edge&id=6f09a7009a1d7a132ddce3a8dd0c46aac66ad8e2

Quoted 7 lines
Quoted 6 lines

Looks good to me, except that the include of kconfig.h should be dropped now.

Thanks,
Armin Wolf

I also just noticed that we still need to include panic.h for add_taint().

It complied regardless, likely because kernel.h was included, but yes.

Quoted 1 line

Should i send a new revision?

No need, I've just updated it by hand and pushed it out.

author
#8 lore raw

Am 27.02.26 um 21:39 schrieb Rafael J. Wysocki:

Quoted 35 lines · Armin Wolf
Quoted 34 lines

Am 27.02.26 um 20:50 schrieb Armin Wolf:

Quoted 32 lines

Am 26.02.26 um 19:46 schrieb Rafael J. Wysocki:

Quoted 20 lines · Armin Wolf
Quoted 19 lines

Am 25.02.26 um 22:28 schrieb Rafael J. Wysocki:

Quoted 17 lines · Armin Wolf
Quoted 9 lines

Am 04.02.26 um 22:29 schrieb Armin Wolf:

Quoted 6 lines

The ACPI spec states that the operating system should respond to a fatal ACPI error by "performing a controlled OS shutdown in a timely fashion". Comply with the ACPI specification by powering off the system when ACPICA signals a fatal ACPI error. Users can still disable this behavior by using the acpi.poweroff_on_fatal kernel option to work around firmware bugs.

Any updates on this?

I was about to apply it, but then I thought that I was not sure about the Kconfig option.

I don't see much value in it TBH. If you agree, I'll just apply
the
patch without that part.

Quoted 7 lines
Quoted 2 lines
Quoted 1 line

Thanks!

You can drop the Kconfig option if you want.

OK, applied as 7.1 material.

Please double check
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/commit/?h=bleeding-edge&id=6f09a7009a1d7a132ddce3a8dd0c46aac66ad8e2

Looks good to me, except that the include of kconfig.h should be dropped now.

Quoted 8 lines
Quoted 4 lines
Quoted 3 lines

Thanks,
Armin Wolf

I also just noticed that we still need to include panic.h for add_taint().

It complied regardless, likely because kernel.h was included, but yes.

Quoted 1 line

Should i send a new revision?

No need, I've just updated it by hand and pushed it out.

Thank you very much :)

Armin Wolf