[PATCH] driver-core: fix DEVICE_INT_ATTR to use correct show/store functions
DEVICE_INT_ATTR() should use device_show_int() and device_store_int()
not device_show_ulong() and device_store_ulong()
Signed-off-by: Michael Davidson <[email protected]>
---
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -508,7 +508,7 @@ ssize_t device_store_int(struct device *
{ __ATTR(_name, _mode, device_show_ulong, device_store_ulong), &(_var) }
#define DEVICE_INT_ATTR(_name, _mode, _var) \
struct dev_ext_attribute dev_attr_##_name = \
- { __ATTR(_name, _mode, device_show_ulong, device_store_ulong), &(_var) }
+ { __ATTR(_name, _mode, device_show_int, device_store_int), &(_var) }
extern int device_create_file(struct device *device,
const struct device_attribute *entry);
On Thu, May 03, 2012 at 04:19:02PM -0700, Michael Davidson wrote:
> [PATCH] driver-core: fix DEVICE_INT_ATTR to use correct show/store functions
This line isn't needed again in the patch body like this :(
> DEVICE_INT_ATTR() should use device_show_int() and device_store_int()
> not device_show_ulong() and device_store_ulong()
>
> Signed-off-by: Michael Davidson <[email protected]>
Is this something that fixes a problem you have seen, so it needs to go
in now, and backported to stable kernel releases, or can it wait for 3.5
and you found it just by looking at the code?
thanks,
greg k-h
I found the problem while porting some code to 3.3.
It looks as if the mce driver uses int attributes for a few things so I assume
that is broken but I have not actually verified it.
thanks,
md
On Thu, May 3, 2012 at 4:32 PM, Greg KH <[email protected]> wrote:
> On Thu, May 03, 2012 at 04:19:02PM -0700, Michael Davidson wrote:
>> [PATCH] driver-core: fix DEVICE_INT_ATTR to use correct show/store functions
>
> This line isn't needed again in the patch body like this :(
>
>> DEVICE_INT_ATTR() should use device_show_int() and device_store_int()
>> not device_show_ulong() and device_store_ulong()
>>
>> Signed-off-by: Michael Davidson <[email protected]>
>
> Is this something that fixes a problem you have seen, so it needs to go
> in now, and backported to stable kernel releases, or can it wait for 3.5
> and you found it just by looking at the code?
>
> thanks,
>
> greg k-h