On Fri, Jul 20, 2012 at 10:56:37PM +0100, David Howells wrote:

Here's the second installment of patches from step 1 of my plan below to clean
up the kernel header files and sort out the inclusion recursion problems.

Note that these patches will need regenerating if the header files they alter
change before they're applied.  However, the disintegration is scripted, so
that just takes a few minutes normally.


===================================
BACKGROUND ON THE RECURSION PROBLEM
===================================

I occasionally run into a problem where I can't write an inline function in a
header file because I need to access something from another header that
includes this one.  Due to this, I end up writing it as a #define instead.

The problems are mainly due to inline functions.  If we split some headers
(linux/sched.h being the biggest culprit) to separate the inline functions from
the data structs (e.g. task_struct) then we could reduce the problems.  Other
splits and rearrangements could help also.

Quite often it's a case of an inline function in header A wanting a struct[*]
from header B, but header B already has an inline function that wants a struct
from header A.

	[*] or constant or whatever.

In the past someone tried to add a kernel-offsets file (an analogue to
asm-offsets) to deal with the problems of dealing with both linux/rcupdate.h
and linux/sched.h - each header needed to be included before the other.

Indeed, and later attempts to work around this problem using per-CPU
variables did not go well either.  This separation should allow
__rcu_read_lock() to be inlined for PREEMPT=y kernels, which would be
quite nice.

Acked-by: Paul E. McKenney <redacted>