The <linux/module.h> pretty much brings in the kitchen sink along
with it, so it should be avoided wherever reasonably possible in
terms of being included from other commonly used <linux/something.h>
files, as it results in a measureable increase on compile times.

The worst culprit was probably device.h since it is used everywhere.
This file also had an implicit dependency/usage of mutex.h which was
masked by module.h, and is also fixed here at the same time.

There are over a dozen other headers that simply declare the
struct instead of pulling in the whole file, so follow their lead
and simply make it a few more.

Most of the implicit dependencies on module.h being present by
these headers pulling it in have been now weeded out, so we can
finally make this change with hopefully minimal breakage.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

The original implementations reference THIS_MODULE in an inline.
We could include <linux/export.h>, but it is better to avoid chaining.

Fortunately someone else already thought of this, and made a similar
inline into a #define in <linux/device.h> for device_schedule_callback(),
[see commit 523ded71] so follow that precedent here.

Also bubble up any __must_check that were used on the prev. wrapper inline
functions up one to the real __register functions, to preserve any prev.
sanity checks that were used in those instances.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

The <linux/crypto.h> (which is in turn in common headers
like tcp.h) wants to use module_name() in an inline fcn.
But having all of <linux/module.h> along for the ride is
overkill and slows down compiles by a measureable amount,
since it in turn includes lots of headers.

Since the inline is never used anywhere in the kernel[1],
we can just remove it, and then also remove the module.h
include as well.

In all the many crypto modules, there were some relying on
crypto.h including module.h -- for them we now explicitly
call out module.h for inclusion.

[1] git grep shows some staging drivers also define the same
static inline, but they also never ever use it.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Once we clean up the implicit presence of module.h (and all its
sub-includes), we'll see an implicit dependency on page.h for
the PAGE_SIZE define.  So fix it in advance.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This file was getting notifier.h via device.h --> module.h but
the module.h inclusion is going away, so add notifier.h directly.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

The implicit presence of module.h and all its sub-includes was
masking these implicit header usages:

include/linux/dmaengine.h:684: warning: 'struct page' declared inside parameter list
include/linux/dmaengine.h:684: warning: its scope is only this definition or declaration, which is probably not what you want
include/linux/dmaengine.h:687: warning: 'struct page' declared inside parameter list
include/linux/dmaengine.h:736:2: error: implicit declaration of function 'bitmap_zero'

With input from Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

By removing the implicit presence of module.h from this file, we
will see things like:

In file included from fs/dlm/user.c:9:
include/linux/miscdevice.h:50: error: field ‘list’ has incomplete type
include/linux/miscdevice.h:54: error: expected specifier-qualifier-list before ‘mode_t’

Call out lists.h and types.h for inclusion to fix each of the
above respectively.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This will show up on MIPS when we fix all the implicit header presences
that are because of module.h being everywhere.

In file included from kernel/trace/ftrace.c:16:
include/linux/stop_machine.h: In function 'stop_one_cpu':
include/linux/stop_machine.h:50: error: implicit declaration of function 'smp_processor_id'
include/linux/stop_machine.h: In function 'stop_cpus':
include/linux/stop_machine.h:80: error: implicit declaration of function 'raw_smp_processor_id'
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

It shows up as a build failure on MIPS, as it is used in
three of_property function stubs.

include/linux/of.h:275: error: 'ENOSYS' undeclared (first use in this function)
include/linux/of.h:282: error: 'ENOSYS' undeclared (first use in this function)
include/linux/of.h:295: error: 'ENOSYS' undeclared (first use in this function)
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

There is nothing modular in this file, and no reason to drag
in all the 357 headers that module.h brings with it, since
it just slows down compiles.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This file had an include of module.h which was probably added
in relation to this line:

  #define ACPI_EXPORT_SYMBOL(symbol)  EXPORT_SYMBOL(symbol);

However, we really expect symbol exporters to grab export.h
themselves, and since this is only a define, we can remove
the module.h include without aclinux.h itself causing any
compile issues.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This file has a define MODULE_ALIAS_MISCDEV which in turn will
use the MODULE_ALIAS define, but only if the former is explicitly
used by modular device driver code (and such code should be
already including module.h).

Delete the include, since module.h is such a giant thing that we
don't want it implicitly sneaking into compiles where it isn't
specifically required.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

There is nothing modular in this file, and no reason to drag
in all the extra headers that module.h brings with it, since
it just slows down compiles.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This file has modular references, but they are limited to
those which are covered by the simple "struct module;"
declaration used in dozens of other places.  In fact that
declaration is already there (just outside of the context
of this commit) so simply remove the include line.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

There is nothing module specific in this header, and removing
it doesn't seem to uncover any implicit dependencies either.
Must be simply a vestige of an ancient legacy.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

The <linux/module.h> pretty much brings in the kitchen sink along
with it, so it should be avoided wherever reasonably possible in
terms of being included from other commonly used <linux/something.h>
files, as it results in a measureable increase on compile times.

There doesn't appear to be any module specifics in this file.
The obvious people who were relying on the presence of
the vast amount of stuff module.h sucked in have been fixed.

If other files are implicitly relying on it, then lets see who
they are and fix them too.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This file consists of nothing other than things like:

  #ifdef CONFIG_FOO
  #define ....

There is no reason for it to require module.h
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

There are no modular calls here, so just the minimal header for
the EXPORT_SYMBOL macro will suffice.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This is a module and so needs module.h so that it will still build
once the implicit presence of module.h is removed.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

These macros are no longer in module.h and module.h is no longer
present everywhere.  Call out export.h for the users who are
making use of these macros.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Signed-off-by: NStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

As we will get this after the module.h usage cleanup.

drivers/pcmcia/db1xxx_ss.c:577: warning: type defaults to 'int' in declaration of 'MODULE_LICENSE'
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

These macros are no longer in module.h and module.h is no longer
present everywhere.  Call out export.h for the real users who
are making use of these macros, or else we'll get things like:

 CC      drivers/uwb/umc-drv.o
drivers/uwb/umc-dev.c:42: warning: data definition has no type or storage class
drivers/uwb/umc-dev.c:42: warning: type defaults to ‘int’ in declaration of ‘EXPORT_SYMBOL_GPL’
drivers/uwb/umc-dev.c:42: warning: parameter names (without types) in function declaration
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Lots of files were getting an implicit include of module.h
for years now.  But that will be fixed shortly.  So get the
real users to call out the include explicitly.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

The module.h (via device.h) was pulling handfuls of implicitly present
header files -- including basic stuff like stat.h.  Fix up these stat.h
users in advance so they don't break on the cleanup.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This uses both EXPORT_SYMBOL and THIS_MODULE, both which come from
the export.h file now.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

We are clipping down the presence of module.h, since it was
everywhere.  If you really need it, you better call it out,
as per this changeset.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This was implicitly appearing by way of module.h -- but when
we fix that, we'll get this:

In file included from drivers/regulator/dummy.c:21:
include/linux/regulator/driver.h:197: error: field 'notifier' has incomplete type
make[3]: *** [drivers/regulator/dummy.o] Error 1
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Another group of drivers that are taking advantage of the implicit
presence of module.h -- and will break when we pull the carpet out
from under them during a cleanup.  Fix 'em now.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

So that THIS_MODULE is still in scope once we clean up the presence
of module.h being everywhere.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Up to now, the module.h header was as hard to keep out as
sunlight.  But we are cleaning that up.  Fix the virtio users
who simply expect module.h to be there in every C file.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

We need this in advance of the module.h cleanup, or we'll
get compile errors like this:

  CC      drivers/lguest/lguest_device.o
drivers/lguest/lguest_device.c: In function ‘lguest_devices_init’:
drivers/lguest/lguest_device.c:490: error: ‘THIS_MODULE’ undeclared (first use in this function)
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This file has things like module_param_named() and MODULE_PARM_DESC()
so it needs the full module.h header present.  Without it, you'll get:

  CC      arch/x86/kvm/../../../virt/kvm/iommu.o
virt/kvm/iommu.c:37: error: expected ‘)’ before ‘bool’
virt/kvm/iommu.c:39: error: expected ‘)’ before string constant
make[3]: *** [arch/x86/kvm/../../../virt/kvm/iommu.o] Error 1
make[2]: *** [arch/x86/kvm] Error 2
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This was coming in via an implicit module.h (and its sub-includes)
before, but we'll be cleaning that up shortly.  Call out the stat.h
include requirement in advance.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Things like THIS_MODULE and EXPORT_SYMBOL were simply everywhere
because module.h was also everywhere.  But we are fixing the latter.
So we need to call out the real users in advance.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Previously these drivers just got module.h implicitly, but we
are cleaning that up and it will be no longer.  Call out the
real users of it.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

We are cleaning up the omnipresent module.h stuff, so people
who really use it need to call it out explicitly.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This 2nd batch of implicit module.h users only appeared when we
removed the unnecessary module.h from include/linux/miscdevice.h
[The 1st batch is already present in Greg's staging tree.]
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>