On Wednesday, 22 of October 2008, Len Brown wrote:> > > > drivers/acpi/sleep/main.c:67: warning: ‘acpi_pm_disable_gpes’ defined but not used> > > > > > > Shows that this code has an identity crisis due to a maze of #ifdefs, in> > > the PM_SLEEP && !SUSPEND && !HIBERNATION case for example.> > > > This case is invalid, because PM_SLEEP == SUSPEND || HIBERNATION !> > I though so too, but 2.6.27 appears to have added XEN to the mix:> > config PM_SLEEP> bool> depends on SUSPEND || HIBERNATION || XEN_SAVE_RESTORE
Yeah, which breaks things. :-(
The patch below should fix it and BTW I think ACPI_SLEEP should not dependon XEN_SAVE_RESTORE anyway, so this is a bug fix really (please consider as.28 material, probably -stable too).
Thanks,Rafael
---From: Rafael J. Wysocki <rjw@xxxxxxx>
ACPI suspend: Fix CONFIG_ACPI_SLEEP dependence and some compilation warnings
Initially CONFIG_PM_SLEEP was defined asCONFIG_SUSPEND || CONFIG_HIBERNATION and some ACPI code, mostimportantly the code in drivers/acpi/main.c, was written with thisassumption. Currently, however, CONFIG_PM_SLEEP is also set whenCONFIG_XEN_SAVE_RESTORE is set.
This causes some compilation warnings to appear indrivers/acpi/main.c if both CONFIG_SUSPEND and CONFIG_HIBERNATIONare unset and CONFIG_PM_SLEEP is set (this was impossible before).To fix this problem, redefine CONFIG_ACPI_SLEEP do depend directlyon CONFIG_SUSPEND || CONFIG_HIBERNATION, as originally intended, anduse it instead of CONFIG_PM_SLEEP in drivers/acpi/main.c, whereverappropriate.
Additionally, move the acpi_target_sleep_state definition from underthe #ifdef to prevent compilation from failing in some cases.
Signed-off-by: Rafael J. Wysocki <rjw@xxxxxxx>--- drivers/acpi/Kconfig | 2 +- drivers/acpi/sleep/main.c | 7 +++---- 2 files changed, 4 insertions(+), 5 deletions(-)
Index: linux-2.6/drivers/acpi/Kconfig===================================================================--- linux-2.6.orig/drivers/acpi/Kconfig+++ linux-2.6/drivers/acpi/Kconfig@@ -42,7 +42,7 @@ if ACPI config ACPI_SLEEP bool- depends on PM_SLEEP+ depends on SUSPEND || HIBERNATION default y config ACPI_PROCFSIndex: linux-2.6/drivers/acpi/sleep/main.c===================================================================--- linux-2.6.orig/drivers/acpi/sleep/main.c+++ linux-2.6/drivers/acpi/sleep/main.c@@ -23,6 +23,7 @@ #include "sleep.h" u8 sleep_states[ACPI_S_STATE_COUNT];+static u32 acpi_target_sleep_state = ACPI_STATE_S0; static int acpi_sleep_prepare(u32 acpi_state) {@@ -45,9 +46,7 @@ static int acpi_sleep_prepare(u32 acpi_s return 0; } -#ifdef CONFIG_PM_SLEEP-static u32 acpi_target_sleep_state = ACPI_STATE_S0;-+#ifdef CONFIG_ACPI_SLEEP /* * ACPI 1.0 wants us to execute _PTS before suspending devices, so we allow the * user to request that behavior by using the 'acpi_old_suspend_ordering'@@ -132,7 +131,7 @@ static void acpi_pm_end(void) */ acpi_target_sleep_state = ACPI_STATE_S0; }-#endif /* CONFIG_PM_SLEEP */+#endif /* CONFIG_ACPI_SLEEP */ #ifdef CONFIG_SUSPEND extern void do_suspend_lowlevel(void);_______________________________________________linux-pm mailing listlinux-pm@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx://lists.linux-foundation.org/mailman/listinfo/linux-pm
