Skip to contentSkip to conversation

sanitize power management config menus, take two

TRIVIAL Tomas Szepe 4 participants 14 messages Aug 5, 2003 – Aug 6, 2003

lore
Share & tools
lore

Tomas Szepe

author
#0 lore raw

Resend: patch against test2-bk4.
Sanitize power management config menus, take two.

Signature · 184 lines
Tomas Szepe <[email protected]>


diff -urN a/arch/i386/Kconfig b/arch/i386/Kconfig
--- a/arch/i386/Kconfig	2003-07-14 09:38:20.000000000 +0200
+++ b/arch/i386/Kconfig	2003-07-27 13:57:09.000000000 +0200
@@ -797,11 +797,11 @@
 endmenu
 
 
-menu "Power management options (ACPI, APM)"
+menu "Power management support"
 	depends on !X86_VOYAGER
 
 config PM
-	bool "Power Management support"
+	bool "Power management support"
 	---help---
 	  "Power Management" means that parts of your computer are shut
 	  off or put into a power conserving "sleep" mode if they are not
@@ -820,34 +820,7 @@
 	  will issue the hlt instruction if nothing is to be done, thereby
 	  sending the processor to sleep and saving power.
 
-config SOFTWARE_SUSPEND
-	bool "Software Suspend (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && PM && SWAP
-	---help---
-	  Enable the possibilty of suspendig machine. It doesn't need APM.
-	  You may suspend your machine by 'swsusp' or 'shutdown -z <time>' 
-	  (patch for sysvinit needed). 
-
-	  It creates an image which is saved in your active swaps. By the next
-	  booting the, pass 'resume=/path/to/your/swap/file' and kernel will 
-	  detect the saved image, restore the memory from
-	  it and then it continues to run as before you've suspended.
-	  If you don't want the previous state to continue use the 'noresume'
-	  kernel option. However note that your partitions will be fsck'd and
-	  you must re-mkswap your swap partitions/files.
-
-	  Right now you may boot without resuming and then later resume but
-	  in meantime you cannot use those swap partitions/files which were
-	  involved in suspending. Also in this case there is a risk that buffers
-	  on disk won't match with saved ones.
-
-	  SMP is supported ``as-is''. There's a code for it but doesn't work.
-	  There have been problems reported relating SCSI.
-
-	  This option is about getting stable. However there is still some
-	  absence of features.
-
-	  For more information take a look at Documentation/swsusp.txt.
+source "kernel/Kconfig.swsusp"
 
 source "drivers/acpi/Kconfig"
 
diff -urN a/arch/i386/kernel/cpu/cpufreq/Kconfig b/arch/i386/kernel/cpu/cpufreq/Kconfig
--- a/arch/i386/kernel/cpu/cpufreq/Kconfig	2003-07-10 23:30:33.000000000 +0200
+++ b/arch/i386/kernel/cpu/cpufreq/Kconfig	2003-07-27 13:50:30.000000000 +0200
@@ -2,10 +2,9 @@
 # CPU Frequency scaling
 #
 
-menu "CPU Frequency scaling"
-
 config CPU_FREQ
 	bool "CPU Frequency scaling"
+	depends on PM
 	help
 	  Clock scaling allows you to change the clock speed of CPUs on the
 	  fly. This is a nice method to save battery power on notebooks,
@@ -16,6 +15,8 @@
 
 	  If in doubt, say N.
 
+if CPU_FREQ
+
 source "drivers/cpufreq/Kconfig"
 
 config CPU_FREQ_TABLE
@@ -162,4 +163,4 @@
 
 	  If in doubt, say N.
 
-endmenu
+endif
diff -urN a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig
--- a/arch/x86_64/Kconfig	2003-07-14 09:38:20.000000000 +0200
+++ b/arch/x86_64/Kconfig	2003-07-27 13:57:24.000000000 +0200
@@ -296,34 +296,7 @@
 	  sending the processor to limited sleep and saving power. However
 	  using ACPI will likely save more power.
 
-config SOFTWARE_SUSPEND
-	bool "Software Suspend (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && PM && SWAP
-	---help---
-	  Enable the possibilty of suspending the machine. It doesn't need APM.
-	  You may suspend your machine by 'swsusp' or 'shutdown -z <time>' 
-	  (patch for sysvinit needed). 
-
-	  It creates an image which is saved in your active swaps. On the next
-	  boot, pass the 'resume=/path/to/your/swap/file' option and the kernel
-	  will detect the saved image, restore the memory from
-	  it, and then continue to run as before you suspended.
-	  If you don't want the previous state to continue, use the 'noresume'
-	  kernel option. However, note that your partitions will be fsck'd and
-	  you must re-mkswap your swap partitions/files.
-
-	  Right now you may boot without resuming and then later resume but
-	  in the meantime you cannot use those swap partitions/files which were
-	  involved in suspending. Also in this case there is a risk that buffers
-	  on disk won't match with saved ones.
-
-	  SMP is supported ``as-is''. There's code for it but doesn't work.
-	  There have been problems reported relating to SCSI.
-
-	  This option is close to getting stable. However there is still some
-	  absence of features.
-
-	  For more information take a look at Documentation/swsusp.txt.
+source "kernel/Kconfig.swsusp"
 
 source "drivers/acpi/Kconfig"
 
diff -urN a/drivers/acpi/Kconfig b/drivers/acpi/Kconfig
--- a/drivers/acpi/Kconfig	2003-06-22 22:27:39.000000000 +0200
+++ b/drivers/acpi/Kconfig	2003-07-27 13:50:30.000000000 +0200
@@ -2,7 +2,7 @@
 # ACPI Configuration
 #
 
-menu "ACPI Support"
+if PM
 
 config ACPI
 	bool "ACPI Support" if X86
@@ -36,6 +36,8 @@
 	  available at:
 	  <http://www.acpi.info>
 
+if ACPI
+
 config ACPI_HT_ONLY
 	bool "CPU Enumeration Only"
 	depends on X86 && ACPI && X86_LOCAL_APIC
@@ -236,5 +238,6 @@
 	depends on IA64 && (!IA64_HP_SIM || IA64_SGI_SN)
 	default y
 
-endmenu
+endif
 
+endif
diff -urN a/kernel/Kconfig.swsusp b/kernel/Kconfig.swsusp
--- a/kernel/Kconfig.swsusp	1970-01-01 01:00:00.000000000 +0100
+++ b/kernel/Kconfig.swsusp	2003-07-27 14:13:26.000000000 +0200
@@ -0,0 +1,26 @@
+config SOFTWARE_SUSPEND
+	bool "Software Suspend (EXPERIMENTAL)"
+	depends on EXPERIMENTAL && PM && SWAP
+	---help---
+	  Enable the possibility of suspending the machine.  This feature
+	  doesn't require APM.  You may suspend your machine by 'swsusp'
+	  or 'shutdown -z <time>' (with the appropriate patch for sysvinit).
+
+	  The suspend procedure will create an image that is saved in one
+	  of your active swap partitions (swap files are not supported).
+
+	  On the next boot, passing the kernel boot option
+	  'resume=/path/to/your/swap/partition' will cause the kernel
+	  to detect the saved image, restore system memory from it and
+	  then continue to run as before you suspended.  If you don't
+	  want the previous state restored, use the 'noresume' kernel
+	  boot option.  However, note that in this case your partitions
+	  will appear to be damaged and you will need to re-mkswap.
+
+	  You are not advised to boot without resuming and then resume
+	  later.  There is a high probability that buffers on disk won't
+	  match with the saved ones.
+
+	  SMP is NOT supported at this time.
+
+	  For more information take a look at Documentation/swsusp.txt.
#1 lore raw
Quoted 7 lines · Tomas Szepe

Resend: patch against test2-bk4.
Sanitize power management config menus, take two.

Signature · 1 line
Tomas Szepe <[email protected]>

I don't understand this chunk:

Quoted 30 lines

diff -urN a/arch/i386/kernel/cpu/cpufreq/Kconfig b/arch/i386/kernel/cpu/cpufreq/Kconfig

arch/i386/kernel/cpu/cpufreq/Kconfig +4 -3
@@ -2,10 +2,9 @@
 # CPU Frequency scaling
 #
 
-menu "CPU Frequency scaling"
-
 config CPU_FREQ
 	bool "CPU Frequency scaling"
+	depends on PM
 	help
 	  Clock scaling allows you to change the clock speed of CPUs on the
 	  fly. This is a nice method to save battery power on notebooks,

@@ -16,6 +15,8 @@
 
 	  If in doubt, say N.
 
+if CPU_FREQ
+
 source "drivers/cpufreq/Kconfig"
 
 config CPU_FREQ_TABLE

@@ -162,4 +163,4 @@
 
 	  If in doubt, say N.
 
-endmenu
+endif
Signature · 4 lines
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.
author
#2 lore raw
Quoted 27 lines

[[email protected]]

I don't understand this chunk:

Quoted 23 lines

diff -urN a/arch/i386/kernel/cpu/cpufreq/Kconfig b/arch/i386/kernel/cpu/cpufreq/Kconfig

arch/i386/kernel/cpu/cpufreq/Kconfig +3 -2
@@ -2,10 +2,9 @@
 # CPU Frequency scaling
 #
 
-menu "CPU Frequency scaling"
-
 config CPU_FREQ
 	bool "CPU Frequency scaling"
+	depends on PM
 	help
 	  Clock scaling allows you to change the clock speed of CPUs on the
 	  fly. This is a nice method to save battery power on notebooks,

@@ -16,6 +15,8 @@
 
 	  If in doubt, say N.
 
+if CPU_FREQ
+
 source "drivers/cpufreq/Kconfig"
Quoted 7 lines

arch/i386/kernel/cpu/cpufreq/Kconfig line 166

Quoted 7 lines
arch/i386/kernel/cpu/cpufreq/Kconfig +1 -1
 config CPU_FREQ_TABLE

@@ -162,4 +163,4 @@
 
 	  If in doubt, say N.
 
-endmenu
+endif

o only enable cpufreq options if power management is selected
o don't put cpufreq options in a separate submenu

Signature · 1 line
Tomas Szepe <[email protected]>
#3 lore raw
Quoted 2 lines · Tomas Szepe

o only enable cpufreq options if power management is selected
o don't put cpufreq options in a separate submenu

Yes, but what I do not understand is why cpufreq need power management.

Signature · 4 lines
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.
author
#4 lore raw
Quoted 6 lines

[[email protected]]

Quoted 2 lines

o only enable cpufreq options if power management is selected
o don't put cpufreq options in a separate submenu

Yes, but what I do not understand is why cpufreq need power management.

Because it is a power management option. :)

CONFIG_PM is a dummy option, it does not link any code into the kernel by itself.

Signature · 1 line
Tomas Szepe <[email protected]>
author
#5 lore raw
Quoted 7 lines

[[email protected]]

Quoted 5 lines

[[email protected]]

Quoted 2 lines

o only enable cpufreq options if power management is selected
o don't put cpufreq options in a separate submenu

Quoted 6 lines
Quoted 1 line

Yes, but what I do not understand is why cpufreq need power management.

Because it is a power management option. :)

CONFIG_PM is a dummy option, it does not link any code into the kernel by itself.

Ooops, I ain't right. :)

There actually seems to be code that depends on CONFIG_PM, particularly so on arches other that x86, so the 'depends on PM' clause for cpufreq is indeed bogus.

Thanks for pointing this out, I'll post a fixed patch.

Signature · 1 line
Tomas Szepe <[email protected]>
#6 lore raw
In reply to Tomas Szepe
Quoted 10 lines
Quoted 5 lines
Quoted 3 lines

o only enable cpufreq options if power management is selected
o don't put cpufreq options in a separate submenu

Yes, but what I do not understand is why cpufreq need power management.

Because it is a power management option. :)

CONFIG_PM is a dummy option, it does not link any code into the kernel by itself.

Actually, it does:

./arch/arm/kernel/Makefile:obj-$(CONFIG_PM) += pm.o
./arch/arm/mach-pxa/Makefile:obj-$(CONFIG_PM) += pm.o sleep.o
./arch/arm/mach-sa1100/Makefile:obj-$(CONFIG_PM) += pm.o sleep.o
./arch/i386/kernel/Makefile:obj-$(CONFIG_PM) += suspend.o
./drivers/pci/Makefile:obj-$(CONFIG_PM) += power.o ./kernel/Makefile:obj-$(CONFIG_PM) += pm.o power/

But, I agree with your change anyway.

-pat

author
#7 lore raw
Quoted 14 lines

[[email protected]]

Quoted 9 lines
Quoted 4 lines
Quoted 2 lines

o only enable cpufreq options if power management is selected
o don't put cpufreq options in a separate submenu

Yes, but what I do not understand is why cpufreq need power management.

Because it is a power management option. :)

CONFIG_PM is a dummy option, it does not link any code into the kernel by itself.

Actually, it does:

Quoted 8 lines

./arch/arm/kernel/Makefile:obj-$(CONFIG_PM) += pm.o
./arch/arm/mach-pxa/Makefile:obj-$(CONFIG_PM) += pm.o sleep.o
./arch/arm/mach-sa1100/Makefile:obj-$(CONFIG_PM) += pm.o sleep.o
./arch/i386/kernel/Makefile:obj-$(CONFIG_PM) += suspend.o
./drivers/pci/Makefile:obj-$(CONFIG_PM) += power.o ./kernel/Makefile:obj-$(CONFIG_PM) += pm.o power/

But, I agree with your change anyway.

Trouble is, the same goes for ACPI -- it doesn't require that CONFIG_PM code be present.

I think the correct x86 solution would be to introduce a real dummy option for the menus, and imply CONFIG_PM if APM or swsusp (the two options that seem to actually need CONFIG_PM code) is enabled.

Signature · 1 line
Tomas Szepe <[email protected]>
author
#8 lore raw
Quoted 10 lines

[[email protected]]

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

Yes, but what I do not understand is why cpufreq need power management.

CONFIG_PM is a dummy option, it does not link any code into the kernel by itself.

Actually, it does:
...

Quoted 8 lines
Quoted 1 line

But, I agree with your change anyway.

Trouble is, the same goes for ACPI -- it doesn't require that CONFIG_PM code be present.

I think the correct x86 solution would be to introduce a real dummy option for the menus, and imply CONFIG_PM if APM or swsusp (the two options that seem to actually need CONFIG_PM code) is enabled.

Patrick, Bruno, what is your opinion on the following replacement patch?

Signature · 196 lines
Tomas Szepe <[email protected]>


diff -urN a/arch/i386/Kconfig b/arch/i386/Kconfig
--- a/arch/i386/Kconfig	2003-08-05 18:39:25.000000000 +0200
+++ b/arch/i386/Kconfig	2003-08-05 18:56:48.000000000 +0200
@@ -797,13 +797,13 @@
 endmenu
 
 
-menu "Power management options (ACPI, APM)"
+menu "Power management support"
 	depends on !X86_VOYAGER
 
 config PM
-	bool "Power Management support"
+	bool "APM and software suspend options"
 	---help---
-	  "Power Management" means that parts of your computer are shut
+	  "Power management" provides for parts of your computer to be shut
 	  off or put into a power conserving "sleep" mode if they are not
 	  being used.  There are two competing standards for doing this: APM
 	  and ACPI.  If you want to use either one, say Y here and then also
@@ -820,37 +820,6 @@
 	  will issue the hlt instruction if nothing is to be done, thereby
 	  sending the processor to sleep and saving power.
 
-config SOFTWARE_SUSPEND
-	bool "Software Suspend (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && PM && SWAP
-	---help---
-	  Enable the possibilty of suspendig machine. It doesn't need APM.
-	  You may suspend your machine by 'swsusp' or 'shutdown -z <time>' 
-	  (patch for sysvinit needed). 
-
-	  It creates an image which is saved in your active swaps. By the next
-	  booting the, pass 'resume=/path/to/your/swap/file' and kernel will 
-	  detect the saved image, restore the memory from
-	  it and then it continues to run as before you've suspended.
-	  If you don't want the previous state to continue use the 'noresume'
-	  kernel option. However note that your partitions will be fsck'd and
-	  you must re-mkswap your swap partitions/files.
-
-	  Right now you may boot without resuming and then later resume but
-	  in meantime you cannot use those swap partitions/files which were
-	  involved in suspending. Also in this case there is a risk that buffers
-	  on disk won't match with saved ones.
-
-	  SMP is supported ``as-is''. There's a code for it but doesn't work.
-	  There have been problems reported relating SCSI.
-
-	  This option is about getting stable. However there is still some
-	  absence of features.
-
-	  For more information take a look at Documentation/swsusp.txt.
-
-source "drivers/acpi/Kconfig"
-
 config APM
 	tristate "Advanced Power Management BIOS support"
 	depends on PM
@@ -996,6 +965,10 @@
 	  a work-around for a number of buggy BIOSes. Switch this option on if
 	  your computer crashes instead of powering off properly.
 
+source "kernel/Kconfig.swsusp"
+
+source "drivers/acpi/Kconfig"
+
 source "arch/i386/kernel/cpu/cpufreq/Kconfig"
 
 endmenu
diff -urN a/arch/i386/kernel/cpu/cpufreq/Kconfig b/arch/i386/kernel/cpu/cpufreq/Kconfig
--- a/arch/i386/kernel/cpu/cpufreq/Kconfig	2003-07-10 23:30:33.000000000 +0200
+++ b/arch/i386/kernel/cpu/cpufreq/Kconfig	2003-08-05 18:42:00.000000000 +0200
@@ -2,8 +2,6 @@
 # CPU Frequency scaling
 #
 
-menu "CPU Frequency scaling"
-
 config CPU_FREQ
 	bool "CPU Frequency scaling"
 	help
@@ -16,6 +14,8 @@
 
 	  If in doubt, say N.
 
+if CPU_FREQ
+
 source "drivers/cpufreq/Kconfig"
 
 config CPU_FREQ_TABLE
@@ -162,4 +162,4 @@
 
 	  If in doubt, say N.
 
-endmenu
+endif
diff -urN a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig
--- a/arch/x86_64/Kconfig	2003-08-05 18:39:26.000000000 +0200
+++ b/arch/x86_64/Kconfig	2003-08-05 18:41:45.000000000 +0200
@@ -296,34 +296,7 @@
 	  sending the processor to limited sleep and saving power. However
 	  using ACPI will likely save more power.
 
-config SOFTWARE_SUSPEND
-	bool "Software Suspend (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && PM && SWAP
-	---help---
-	  Enable the possibilty of suspending the machine. It doesn't need APM.
-	  You may suspend your machine by 'swsusp' or 'shutdown -z <time>' 
-	  (patch for sysvinit needed). 
-
-	  It creates an image which is saved in your active swaps. On the next
-	  boot, pass the 'resume=/path/to/your/swap/file' option and the kernel
-	  will detect the saved image, restore the memory from
-	  it, and then continue to run as before you suspended.
-	  If you don't want the previous state to continue, use the 'noresume'
-	  kernel option. However, note that your partitions will be fsck'd and
-	  you must re-mkswap your swap partitions/files.
-
-	  Right now you may boot without resuming and then later resume but
-	  in the meantime you cannot use those swap partitions/files which were
-	  involved in suspending. Also in this case there is a risk that buffers
-	  on disk won't match with saved ones.
-
-	  SMP is supported ``as-is''. There's code for it but doesn't work.
-	  There have been problems reported relating to SCSI.
-
-	  This option is close to getting stable. However there is still some
-	  absence of features.
-
-	  For more information take a look at Documentation/swsusp.txt.
+source "kernel/Kconfig.swsusp"
 
 source "drivers/acpi/Kconfig"
 
diff -urN a/drivers/acpi/Kconfig b/drivers/acpi/Kconfig
--- a/drivers/acpi/Kconfig	2003-06-22 22:27:39.000000000 +0200
+++ b/drivers/acpi/Kconfig	2003-08-05 18:46:01.000000000 +0200
@@ -2,8 +2,6 @@
 # ACPI Configuration
 #
 
-menu "ACPI Support"
-
 config ACPI
 	bool "ACPI Support" if X86
 	depends on !X86_VISWS
@@ -36,6 +34,8 @@
 	  available at:
 	  <http://www.acpi.info>
 
+if ACPI
+
 config ACPI_HT_ONLY
 	bool "CPU Enumeration Only"
 	depends on X86 && ACPI && X86_LOCAL_APIC
@@ -236,5 +236,4 @@
 	depends on IA64 && (!IA64_HP_SIM || IA64_SGI_SN)
 	default y
 
-endmenu
-
+endif
diff -urN a/kernel/Kconfig.swsusp b/kernel/Kconfig.swsusp
--- a/kernel/Kconfig.swsusp	1970-01-01 01:00:00.000000000 +0100
+++ b/kernel/Kconfig.swsusp	2003-08-05 18:41:45.000000000 +0200
@@ -0,0 +1,26 @@
+config SOFTWARE_SUSPEND
+	bool "Software Suspend (EXPERIMENTAL)"
+	depends on EXPERIMENTAL && PM && SWAP
+	---help---
+	  Enable the possibility of suspending the machine.  This feature
+	  doesn't require APM.  You may suspend your machine by 'swsusp'
+	  or 'shutdown -z <time>' (with the appropriate patch for sysvinit).
+
+	  The suspend procedure will create an image that is saved in one
+	  of your active swap partitions (swap files are not supported).
+
+	  On the next boot, passing the kernel boot option
+	  'resume=/path/to/your/swap/partition' will cause the kernel
+	  to detect the saved image, restore system memory from it and
+	  then continue to run as before you suspended.  If you don't
+	  want the previous state restored, use the 'noresume' kernel
+	  boot option.  However, note that in this case your partitions
+	  will appear to be damaged and you will need to re-mkswap.
+
+	  You are not advised to boot without resuming and then resume
+	  later.  There is a high probability that buffers on disk won't
+	  match with the saved ones.
+
+	  SMP is NOT supported at this time.
+
+	  For more information take a look at Documentation/swsusp.txt.
#9 lore raw
In reply to Tomas Szepe
Quoted 3 lines

Trouble is, the same goes for ACPI -- it doesn't require that CONFIG_PM code be present.

I initially missed that part of your patch, and that is incorrect - Only part of ACPI (CONFIG_ACPI_SLEEP) should depend on CONFIG_PM.

Quoted 3 lines

I think the correct x86 solution would be to introduce a real dummy option for the menus, and imply CONFIG_PM if APM or swsusp (the two options that seem to actually need CONFIG_PM code) is enabled.

I can buy that. There are actually three levels of power management that we handle:

- System Power Management (swsusp, CONFIG_ACPI_SLEEP)
- Device Power Management (kernel/pm.c, future driver model support)
- CPU Power Management (cpufreq)

SPM implies that DPM will be enabled, but both DPM and CPM can exist without SPM, and independently of each other. All of them would essentially fall under CONFIG_PM..

Would you willing to whip up a patch for the Kconfig entries?

-pat

author
#10 lore raw
Quoted 5 lines

[[email protected]]

I can buy that. There are actually three levels of power management that we handle:

Quoted 7 lines

- System Power Management (swsusp, CONFIG_ACPI_SLEEP)
- Device Power Management (kernel/pm.c, future driver model support)
- CPU Power Management (cpufreq)

SPM implies that DPM will be enabled, but both DPM and CPM can exist without SPM, and independently of each other. All of them would essentially fall under CONFIG_PM..

Ok, that makes a lot of sense.

Quoted 1 line

Would you willing to whip up a patch for the Kconfig entries?

Sure, I'll try to put something together so we have a patch to start playing with.

Signature · 1 line
Tomas Szepe <[email protected]>
#11 lore raw
In reply to Patrick Mochel
Quoted 18 lines · Patrick Mochel
Quoted 3 lines

Trouble is, the same goes for ACPI -- it doesn't require that CONFIG_PM code be present.

I initially missed that part of your patch, and that is incorrect - Only part of ACPI (CONFIG_ACPI_SLEEP) should depend on CONFIG_PM.

Quoted 3 lines

I think the correct x86 solution would be to introduce a real dummy option for the menus, and imply CONFIG_PM if APM or swsusp (the two options that seem to actually need CONFIG_PM code) is enabled.

I can buy that. There are actually three levels of power management that we handle:

- System Power Management (swsusp, CONFIG_ACPI_SLEEP)
- Device Power Management (kernel/pm.c, future driver model support)
- CPU Power Management (cpufreq)

Quoted 5 lines

SPM implies that DPM will be enabled, but both DPM and CPM can exist without SPM, and independently of each other. All of them would essentially fall under CONFIG_PM..

Would you willing to whip up a patch for the Kconfig entries?

I mostly agree. The only trouble is then:

$ egrep -rl '#ifdef[:space:]+CONFIG_PM$' linux-2.6.0-test2/ | wc -l
    96

I think it make sense to change this by CONFIG_DPM.

Opinions?

Signature · 4 lines
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.
#12 lore raw
In reply to Patrick Mochel

Hi!

Quoted 11 lines
Quoted 3 lines

I think the correct x86 solution would be to introduce a real dummy option for the menus, and imply CONFIG_PM if APM or swsusp (the two options that seem to actually need CONFIG_PM code) is enabled.

I can buy that. There are actually three levels of power management that we handle:

- System Power Management (swsusp, CONFIG_ACPI_SLEEP)
- Device Power Management (kernel/pm.c, future driver model support)
- CPU Power Management (cpufreq)

Quoted 5 lines

SPM implies that DPM will be enabled, but both DPM and CPM can exist without SPM, and independently of each other. All of them would essentially fall under CONFIG_PM..

Would you willing to whip up a patch for the Kconfig entries?

We have enough trouble making sure *current* PM code runs with all possible config combinations; I do not think we want more PM options for now.

Signature · 2 lines
				Pavel
Written on sharp zaurus, because my Velo1 broke. If you have Velo you don't need...
#13 lore raw
In reply to Patrick Mochel
Quoted 21 lines · Patrick Mochel
Quoted 10 lines
Quoted 5 lines
Quoted 3 lines

o only enable cpufreq options if power management is selected
o don't put cpufreq options in a separate submenu

Yes, but what I do not understand is why cpufreq need power management.

Because it is a power management option. :)

CONFIG_PM is a dummy option, it does not link any code into the kernel by itself.

Actually, it does:

./arch/arm/kernel/Makefile:obj-$(CONFIG_PM) += pm.o
./arch/arm/mach-pxa/Makefile:obj-$(CONFIG_PM) += pm.o sleep.o
./arch/arm/mach-sa1100/Makefile:obj-$(CONFIG_PM) += pm.o sleep.o
./arch/i386/kernel/Makefile:obj-$(CONFIG_PM) += suspend.o
./drivers/pci/Makefile:obj-$(CONFIG_PM) += power.o ./kernel/Makefile:obj-$(CONFIG_PM) += pm.o power/

Quoted 1 line

But, I agree with your change anyway.

Then why? You may want to scale cpu frequency without having power managements in mind. Likewise for acpi btw (think HT only or irq routing via acpi for instance).

Signature · 4 lines
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.