For dt-enabled machines we want to use a big irq_domain over all controllers
and therefore need to access not only the main controllers but the
sub-controller as well.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Enables post-init setting of the desired typehandler for the interrupt.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

The list in used was from the s3c2450, a close cousin of the s3c2416.
As it's not possible to distinguish between the s3c2416 and s3c2450
the additional interrupts of the s3c2450 will only be available thru
devicetree later.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Might be confusing for people to read the code without having the
datasheet nearby.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

This move is necessary to make use of the irqchip infrastructure
for the following devicetree support for s3c24xx architectures.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

This removes the dependency on static irq mappings for basic irq handling
and makes the s3c24xx entry-macro.S obsolete.

Also the interrupts of the second full interrupt controller on the s3c2416
are really handled now, which was forgotten when adding them.

The handling itself does the same as the previous assembler-code in that
it tries to get the interrupt offset from the offset register first and
if that produces wrong results manually searches for the interrupt bit
in the pending register value. It also saves the historic comment which
explains the reason behind this.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

It's not used anywhere else.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

The s3c24xx_init_irq function that was the base for all irq inits
is now only used to initialize the real s3c2410 irqs.

Therefore rename it and also move its declaration from plat/cpu.h
to common.h

The eint declaration is used by the vast majority of the SoCs and
gets therefore placed outside any ifdefs.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

The current parent_irq check checks for a value != 0. This does of course
not work when the parent irq sits in the bit 0 of the parent register.
This only affects the eint0 interrupt of the s3c2412.

To fix this behaviour, check for the presence of a parent_intc in the
structure. In an s3c24xx interrupt controller either all interrupts have
parent interrupts or none have, so if a parent controller is available
the parent_irq value always points to a parent_irq.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

The check during the parent handling itself was wrong, as it should have
checked for parent_irq_data.

The interrupt controller structs always contain an irq_data array with 32
entries and the only possible error could be a parent_irq assignment of >31.

As this would point to outside the irq_data array this could contain
anything including non-NULL values. Therefore correct this to check
the parent_irq value to be in the right range.

With the same explanation of a valid interrupt controller always having a
full irq_data array, the topmost irq_data check in s3c24xx_irq_map
can also go away.

Finally the mapping function is only called thru the irq_domain ops, in
which case the intc struct is already successfully created, so there is
no need to check for it again.
Reported-by: Julia Lawall <julia.lawall@lip6.fr>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

commit 7b45ed96 ("ARM: S3C24XX: handle s3c2412 eints using new
infrastructure") introduced build warning and this patch fixes
that:

drivers/gpio/gpio-samsung.c: In function 's3c24xx_gpiolib_fbank_to_irq':
drivers/gpio/gpio-samsung.c:1126:5: warning: suggest explicit braces to avoid ambiguous 'else' [-Wparentheses]
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Cc: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

The s3c2412 handles the eints 0 to 3 different than all the other SoCs
of the 24xx range. These eints must be acked and masked in the regular
bits as well as the bits 0 to 3 of the eint registers, which are unused
on the other SoCs.

This of course can be realized using the new infrastructure with the
eint bits in the main register being the parent interrupts of the
same bits in the eint register.

The s3c2412 therefore gets its own IRQ_EINT0 to 4 constants that
reside in the newly created gap before IRQ_EINT4. gpio-samsung, as the
only user of these is modified to return the correct values when
handling gpio_to_irq requests on s3c2412 based machines.

Due to lack of hardware this is compile tested only, but should
hopefully work as intended.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Makes it possible to identify s3c2412 SoCs.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

This patch moves the irq numbers starting with EINT4 4 points down to
enable the inclusion of the first 4 bits of EINT register into the mapping
removing the need for special offset handling.

For most S3C24XX architectures this will simply create 4 additional unused
interrupts, but enables the S3C2412 to make use of the new infrastructure
to realize its special handling of the EINT0 to EINT3 interrupts.

All affected parts of the Samsung code (arch + drivers) seem to use the
real interrupt defines (IRQ_something) and not any form of S3C2410_IRQ(x)
whose numbering is changed here starting from S3C2410_IRQ(32).

This patch was runtime-tested on a s3c2416 based board.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Contains only the new mapping structure. The special handling
of the eint0 to eint3 interrupts still needs to be solved.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Combines the two independent init steps for the irqs into one.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Base for further cleanups
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Originally the s3c2412 used a overidden irq chip to set the rtc wakeup flag.
But with using the samsung_sync_wakemask function the same can be achieved
without introducing soc specific irq functions.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

As always a mapping structure is everything needed.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Simply declare a correct mapping structure to use the common irq code.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Now all the arch_initcalls for s3c244x irqs are gone.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Will be integrated in the following patch.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

s3c2440 and s3c2442 need separate init functions, as the s3c2440 contains
even more differing irqs that will be moved in the following patch.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Base for further modifications.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Signed-off-by: Naour Romain <romain.naour@openwide.fr>
Reviewed-by: Tomasz Figa <tomasz.figa@gmail.com>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Signed-off-by: Naour Romain <romain.naour@openwide.fr>
Reviewed-by: Tomasz Figa <tomasz.figa@gmail.com>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Signed-off-by: Naour Romain <romain.naour@openwide.fr>
[tomasz.figa@gmail.com: tested on a Tiny6410 board]
Reviewed-and-Tested-by: Tomasz Figa <tomasz.figa@gmail.com>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Signed-off-by: Naour Romain <romain.naour@openwide.fr>
Reviewed-by: Tomasz Figa <tomasz.figa@gmail.com>
[heiko@sntech.de: tested on a s3c2416 based machine]
Reviewed-and-Tested-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Signed-off-by: Naour Romain <romain.naour@openwide.fr>
Reviewed-by: Tomasz Figa <tomasz.figa@gmail.com>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Only the _init, _init_clocks, _init_uarts and _map_io functions need
NULL defines, as they are used in the cpu map.

Further integrate the two restart functions already in common.h in
their respective soc part and compact the numerous empty lines.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

The different soc functions are now only used in the mach-s3c24xx
directory, so it's not necessary anymore to keep the globally
visible.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

plat-samsung/irq.h did only contain functions for handling the spread out
subirqs on s3c24xx arches, which are not needed anymore.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
[kgene.kim@samsung.com: fixed build error on bast-irq.c]
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

The header file plat/common-smdk.h is used only in mach-s3c24xx/,
so this patch moves it into mach-s3c24xx directory.
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Since mach/regs-sdi.h is used only for s3cmci.c, so this moves the
header file into the driver file, drivers/mmc/host/s3cmci.c file.

Cc: Chris Ball <cjb@laptop.org>
Tested-by: Sylwester Nawrocki <sylvester.nawrocki@gmail.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>

Pull fbdev UAPI disintegration from David Howells:
 "You'll be glad to here that the end is nigh for the UAPI patches.
  Only the fbdev/framebuffer piece remains now that the SCSI stuff has
  gone in.

  Here are the UAPI disintegration bits for the fbdev drivers.  It
  appears that Florian hasn't had time to deal with my patch, but back
  in December he did say he didn't mind if I pushed it forward."

Yay.  No more uapi movement.  And hopefully no more big header file
cleanups coming up either, it just tends to be very painful.

* tag 'disintegrate-fbdev-20121220' of git://git.infradead.org/users/dhowells/linux-headers:
  UAPI: (Scripted) Disintegrate include/video

Merge tag 'stable/for-linus-3.9-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen

Pull Xen bug-fixes from Konrad Rzeszutek Wilk:
 - Update the Xen ACPI memory and CPU hotplug locking mechanism.
 - Fix PAT issues wherein various applications would not start
 - Fix handling of multiple MSI as AHCI now does it.
 - Fix ARM compile failures.

* tag 'stable/for-linus-3.9-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
  xenbus: fix compile failure on ARM with Xen enabled
  xen/pci: We don't do multiple MSI's.
  xen/pat: Disable PAT using pat_enabled value.
  xen/acpi: xen cpu hotplug minor updates
  xen/acpi: xen memory hotplug minor updates

Pull  more VFS bits from Al Viro:
 "Unfortunately, it looks like xattr series will have to wait until the
  next cycle ;-/

  This pile contains 9p cleanups and fixes (races in v9fs_fid_add()
  etc), fixup for nommu breakage in shmem.c, several cleanups and a bit
  more file_inode() work"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  constify path_get/path_put and fs_struct.c stuff
  fix nommu breakage in shmem.c
  cache the value of file_inode() in struct file
  9p: if v9fs_fid_lookup() gets to asking server, it'd better have hashed dentry
  9p: make sure ->lookup() adds fid to the right dentry
  9p: untangle ->lookup() a bit
  9p: double iput() in ->lookup() if d_materialise_unique() fails
  9p: v9fs_fid_add() can't fail now
  v9fs: get rid of v9fs_dentry
  9p: turn fid->dlist into hlist
  9p: don't bother with private lock in ->d_fsdata; dentry->d_lock will do just fine
  more file_inode() open-coded instances
  selinux: opened file can't have NULL or negative ->f_path.dentry

(In the meantime, the hlist traversal macros have changed, so this
required a semantic conflict fixup for the newly hlistified fid->dlist)

Pull btrfs fixup from Chris Mason:
 "Geert and James both sent this one in, sorry guys"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  btrfs/raid56: Add missing #include <linux/vmalloc.h>

Pull second set of s390 patches from Martin Schwidefsky:
 "The main part of this merge are Heikos uaccess patches.  Together with
  commit 09884964 ("mm: do not grow the stack vma just because of an
  overrun on preceding vma") the user string access is hopefully fixed
  for good.

  In addition some bug fixes and two cleanup patches."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
  s390/module: fix compile warning
  qdio: remove unused parameters
  s390/uaccess: fix kernel ds access for page table walk
  s390/uaccess: fix strncpy_from_user string length check
  input: disable i8042 PC Keyboard controller for s390
  s390/dis: Fix invalid array size
  s390/uaccess: remove pointless access_ok() checks
  s390/uaccess: fix strncpy_from_user/strnlen_user zero maxlen case
  s390/uaccess: shorten strncpy_from_user/strnlen_user
  s390/dasd: fix unresponsive device after all channel paths were lost
  s390/mm: ignore change bit for vmemmap
  s390/page table dumper: add support for change-recording override bit