into home.transmeta.com:/home/torvalds/v2.5/linux

No more need to include buffer_head.h in fs.h.  Finally.

Make the 144 files in fs/ that need it include buffer_head.h directly.
Again some uses in the VFS files are layering violations and need to
be addressed later.  The new include statement gives a nice grep pattern
for that :)

Make kernel/ksyms.c include buffer_head.h directly.

Make the 5 headers in include/linux that need it include buffer_head.h
directly.

Make the 7 file that need it in mm/ include buffer_head.h directly.
Once again most files shouln't need it and want fixing.

Include buffer_head.h directly in the file in drivers/ that need it
(9 files).  Note that most of this uses are layering violations that I
will address later.

Remove two unused declarations in fs/jffs/intrep.h.

Remove unused bh_kmap/bh_kunmap inlines from highmem.h.

Declare buffer_init() extern in init/main.c like the other _init so that
it doesn't have to include buffer_head.h.  Remove buffer_init() there.

Now that fs.h grow due to the lock.h removal let's reduce it's overhead
again:  Instead of penalizing ever user of fs.h with the overhead of the
buffer head interface let it's users include it directly.

This also shows nicely which parts of the core kernel still depend on the
buffer head interface, and allows that to be cleaned up properly.

This is the first of ten patches and adds the includes needed by
buffer_head.h to it and fixes it's inclusion guard.

into home.transmeta.com:/home/torvalds/v2.5/linux

We now always allocate the section __ksymtab, which to modutils means
not to implicitly export any symbols.

That means that EXPORT_NO_SYMBOLS is now default, and thus can go away
in the sources.

into home.transmeta.com:/home/torvalds/v2.5/linux

o Standardize building math-emu.o for arch/alpha
o Remove old boilerplate code / redundant comments
o Remove unnecessary obj-m := $(sort $(obj-m)) in ALSA

o Use a shorter expression to transform 'foo.o' -> 'foo-objs'
o Define EXPORT_FLAGS (as -DEXPORT_SYMTAB) in the main Makefile
o Small rearrangement of code in Rules.make

The attached patch fixes the compilation problems Andrew Morton has been
seeing with the rw-semaphores in the kernel when using gcc 2.95.3 with
-O1.

into home.transmeta.com:/home/torvalds/v2.5/linux

into home.transmeta.com:/home/torvalds/v2.5/linux

into home.transmeta.com:/home/torvalds/v2.5/linux

(Christoph Hellwig)

The copy_to/from_user fixes left some variables unused, which
the compiler warned about.

into tp1.ruhr-uni-bochum.de:/home/kai/kernel/v2.5/linux-2.5.isdn

- Yes we really do need sched.h in central.c
- Include namei.h in solaris syscall emul layer
- Use generic_fls in sparc64 bitops.
- Add dummy asm-sparc64/suspend.h

- Include dcache.h/namei.h in fs/autofs/autofs_i.h not dirhash.c
- Include list.h and spinlock.h in dcache.h
- Include list.h in mount.h and namei.h

Removed lots of unneeded #includes from the tty style drivers.

addition of #include <linux/version.h> due to removal from <linux/usb.h>

fix compile time error, due to addition of jiffies.h

Group the variable-processing part together, which figures out what
needs to be done. (Nothing actually changes in this cset)

We have
o CFLAGS_KERNEL: Additional flags for files compiled into vmlinux
o AFLAGS_KERNEL: Additional flags for files assembled into vmlinux
o MODFLAGS: Additonal flags for files compiled as module

For consistency, rename MODFLAGS to CFLAGS_MODULE, introduce
AFLAGS_MODULE but leave it empty for now.

While we're at, removing "MAKING_MODULES=1", nobody is using it.