linux.git/include/linux/device.h, branch v2.6.27-rc4

driver core: remove DEVICE_ID_SIZE define

2008-07-22T04:54:53+00:00

There is no such thing as a "device id size" in the driver core, so
remove the define and fix up any users of this odd define in the rest of
the kernel.

Cc: Kay Sievers 
Signed-off-by: Greg Kroah-Hartman

driver core: remove DEVICE_NAME_SIZE define

2008-07-22T04:54:53+00:00

There is no such thing as a "device name size" in the driver core, so
remove the define and fix up any users of this odd define in the rest of
the kernel.

Signed-off-by: Kay Sievers 
Signed-off-by: Greg Kroah-Hartman

driver core: remove KOBJ_NAME_LEN define

2008-07-22T04:54:52+00:00

Kobjects do not have a limit in name size since a while, so stop
pretending that they do.

Signed-off-by: Kay Sievers 
Signed-off-by: Greg Kroah-Hartman

class: add lockdep infrastructure

2008-07-22T04:54:52+00:00

This adds the infrastructure to properly handle lockdep issues when the
internal class semaphore is changed to a mutex.

Matthew wrote the original patch, and Greg fixed it up to work properly
with the class_create() function.

From: Matthew Wilcox 
Cc: Kay Sievers 
Cc: Dave Young 
Cc: Andrew Morton 
Cc: James Bottomley 
Cc: Peter Zijlstra 
Cc: Ingo Molnar 
Signed-off-by: Greg Kroah-Hartman

class: move driver core specific parts to a private structure

2008-07-22T04:54:51+00:00

This moves the portions of struct class that are dynamic (kobject and
lock and lists) out of the main structure and into a dynamic, private,
structure.


Cc: Kay Sievers 
Signed-off-by: Greg Kroah-Hartman

Driver Core: add ability for class_find_device to start in middle of list

2008-07-22T04:54:47+00:00

This mirrors the functionality that driver_find_device has as well.

We add a start variable, and all callers of the function are fixed up at
the same time.

The block layer will be using this new functionality in a follow-on
patch.


Cc: Kay Sievers 
Signed-off-by: Greg Kroah-Hartman

Driver Core: add ability for class_for_each_device to start in middle of list

2008-07-22T04:54:47+00:00

This mirrors the functionality that driver_for_each_device has as well.

We add a start variable, and all callers of the function are fixed up at
the same time.

The block layer will be using this new functionality in a follow-on
patch.


Cc: Kay Sievers 
Signed-off-by: Greg Kroah-Hartman

device create: convert device_create_drvdata to device_create

2008-07-22T04:54:47+00:00

Now that device_create() has been audited, rename things back to the
original call to be sane.

Keep the device_create_drvdata macro around to make merges easier.

Signed-off-by: Greg Kroah-Hartman

driver core: remove device_create()

2008-07-22T04:54:47+00:00

There are no more users of this, and it is racy.  Use
device_create_drvdata() or device_create_vargs() instead.

Signed-off-by: Greg Kroah-Hartman

sysfs: add /sys/dev/{char,block} to lookup sysfs path by major:minor

2008-07-22T04:54:40+00:00

Why?:
There are occasions where userspace would like to access sysfs
attributes for a device but it may not know how sysfs has named the
device or the path.  For example what is the sysfs path for
/dev/disk/by-id/ata-ST3160827AS_5MT004CK?  With this change a call to
stat(2) returns the major:minor then userspace can see that
/sys/dev/block/8:32 links to /sys/block/sdc.

What are the alternatives?:
1/ Add an ioctl to return the path: Doable, but sysfs is meant to reduce
   the need to proliferate ioctl interfaces into the kernel, so this
   seems counter productive.

2/ Use udev to create these symlinks: Also doable, but it adds a
   udev dependency to utilities that might be running in a limited
   environment like an initramfs.

3/ Do a full-tree search of sysfs.

[kay.sievers@vrfy.org: fix duplicate registrations]
[kay.sievers@vrfy.org: cleanup suggestions]

Cc: Neil Brown 
Cc: Tejun Heo 
Acked-by: Kay Sievers 
Reviewed-by: SL Baur 
Acked-by: Kay Sievers 
Acked-by: Mark Lord 
Acked-by: H. Peter Anvin 
Signed-off-by: Dan Williams 
Signed-off-by: Greg Kroah-Hartman