diff options
author | Alex Chiang <achiang@hp.com> | 2010-02-25 16:09:52 -0500 |
---|---|---|
committer | Tony Luck <tony.luck@intel.com> | 2010-02-25 18:15:13 -0500 |
commit | d868080d2a1c95526cb01e3d0c14096721cbb87a (patch) | |
tree | d87e4a8c2807beb582b20e296ff6dd694172ed18 /arch/ia64/kernel/palinfo.c | |
parent | 60b341b778cc2929df16c0a504c91621b3c6a4ad (diff) |
[IA64] Only build arch/ia64/kernel/acpi.o when CONFIG_ACPI
The following commit broke the ia64 sim_defconfig build:
3b2b84c0b81108a9a869a88bf2beeb5a95d81dd1
ACPI: processor: driver doesn't need to evaluate _PDC
This is because it added:
+#include <acpi/processor.h>
To arch/ia64/kernel/acpi.c. Unfortunately, the ia64_simdefconfig does
not turn on CONFIG_ACPI, and we get build errors.
The fix described in $subject seems to be the most sensible way to
untangle the mess.
The other issue is that acpi_get_sysname() is required for all configs,
most of which define CONFIG_ACPI, but are not CONFIG_IA64_GENERIC. Turn
it into an inline to cover the "non generic" ia64 configs; to prevent
a duplicate definition build error, we need to wrap the definition in
acpi.o inside an #ifdef.
Finally, move the pm_idle and pm_power_off exports into process.c (which
is always built), similar to other architectures, and allow the sim
defconfig to link.
Signed-off-by: Alex Chiang <achiang@hp.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Diffstat (limited to 'arch/ia64/kernel/palinfo.c')
0 files changed, 0 insertions, 0 deletions