As of commit 9a1091ef ("irqchip: gic: Support hierarchy irq
domain."), Marzen legacy hangs during boot with:

   Image Name:   'Linux-3.19.0-rc4'
   Image Type:   ARM Linux Kernel Image (uncompressed)
   Data Size:    3445880 Bytes = 3.3 MiB
   Load Address: 60008000
   Entry Point:  60008000
   Verifying Checksum ... OK
   Loading Kernel Image ... OK
OK

Starting kernel ...

Enabling DEBUG_LL does not seem to change the situation, however this
patch by itself fixes this issue and re-enables normal boot.

This issue happens because the IRQ numbers of the GIC are now virtual,
and no longer match the hardcoded hardware IRQ numbers in the platform
board code.

To fix this, instantiate the GIC from platform board code when compiling
a legacy kernel, like is done for the sh73a0, r8a7740 and r8a7778 legacy code.

Follows same style as the r8a7740 legacy GIC fix by Geert Uytterhoeven,
thanks to him for the initial work.
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Acked-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Add serial port debug macros for the SCIF(A) serial ports.
This includes all supported shmobile SoCs, except for EMEV2.

The configuration logic (both Kconfig and #ifdef) is more complicated than
one would expect, for several reasons:
  1. Not all SoCs have the same serial devices, and they're not always
     at the same addresses.
  2. There are two different types: SCIF and SCIFA. Fortunately they can
     easily be distinguished by physical address.
  3. Not all boards use the same serial port for the console.
     The defaults correspond to the boards that are supported in
     mainline. If you want to use a different serial port, just change
     the value of CONFIG_DEBUG_UART_PHYS, and the rest will auto-adapt.
  4. debug_ll_io_init() maps the SCIF(A) registers to a fixed virtual
     address. 0xfdxxxxxx was chosen, as it should lie below VMALLOC_END
     = 0xff000000, and must not conflict with the 2 MiB reserved region
     at PCI_IO_VIRT_BASE = 0xfee00000.
       - On SoCs not using the legacy machine_desc.map_io(),
	 debug_ll_io_init() is called by the ARM core code.
       - On SoCs using the legacy machine_desc.map_io(),
	 debug_ll_io_init() must be called explicitly. Calls are added
	 for r8a7740, r8a7779, sh7372, and sh73a0.

This was derived from the r8a7790 version by Laurent Pinchart.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: NArnd Bergmann <arnd@arndb.de>
Tested-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

The information is already included in the COPYING file in the kernel
sources root directory, we don't want to modify all source files when
the FSF will move to a new address, and I'm tired of seeing the related
checkpatch.pl warnings.
Signed-off-by: NLaurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Remove NR_IRQS_LEGACY from the r8a7779 generic machine
vector. The generic r8a7779 machine vector requires use
of Multiplatform, and in such case SPARSE_IRQ is enabled
by default. This in turns means that the default value
of .nr_irqs equals NR_IRQS and NR_IRQS_LEGACY. Because
of this we can simply remove NR_IRQS_LEGACY and move one
step closer to a cruft-free environment.
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Fix r8a7779 SoC machine vector to include shmobile_init_late()
so Suspend-to-RAM and CPUIdle are setup and platform devices
are omitted as expected.
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

No more device needs to be added from platform code when booting the
reference platform, remove the now empty
r8a7779_add_standard_devices_dt() function completely.
Signed-off-by: NLaurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Tested-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

A number of board files in arch/arm and arch/unicore32
explicitly reference platform_bus device as a parent
for new platform devices.

This is unnecessary, as platform device API guarantees
that devices with NULL parent are going to by adopted
by the mentioned "root" device.

This patch removes or replaces with NULL such references.
Signed-off-by: NPawel Moll <pawel.moll@arm.com>
Acked-by: NOlof Johansson <olof@lixom.net>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Change location of r8a7779.h so it can be included as "r8a7779.h"
instead of the old style <mach/r8a7779.h>
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Initialise SCIF devices using DT when booting marzen
using multiplatform.
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Remove the now unused r8a7779_init_delay() function.

Based on work for the r8a7740 SoC by Magnus Damm.

Cc: Magnus Damm <damm+renesas@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Convert the common C-code-less r8a7779 DT board support
to use shmobile_init_delay() to be able to migrate away
from per-SoC delay setup functions.

Based on work by Magnus dam for the r8a7740 SoC.

Cc: Magnus Damm <damm+renesas@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Move the clock initialisation and OF device population from
SoC to board code. This is in keeping with the pattern used by Lager.
And the clock portion is part of decoupling clock initialisation
from SoC code in preparation for moving to the common clock framework.
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

r8a7779_earlytimer_init() calls r8a7779_clock_init() and
r8a7779_clock_init() is defined in clock-r8a7779.c.

If both CONFIG_COMMON_CLK and CONFIG_ARCH_R8A7779 are enabled,
as will be the case when marzen-reference moves to use
the common clock framework, then setup-r8a7779.c is compiled
but clock-r8a7779.c is not.

As r8a7779_earlytimer_init() is not used by marzen-reference
simply move it to clock-r8a7779.c.

[horms+renesas@verge.net.au: rebase]
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Add and use helper to read mode pins.
This will be re-used when moving marzen-reference to
the common clock framework.
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Change location for common.h so it can be used as #include "common.h"
instead of the old style #include <mach/common.h>.
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Acked-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Acked-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Move most of irqs.h so it can be used as #include "irqs.h"
instead of the old style #include <mach/irqs.h>.

Legacy code in drivers/pinctrl needs more work to get rid
of the "mach" portion of the include path, so some part
is left in the original location.
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Acked-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Acked-by: NArnd Bergmann <arnd@arndb.de>
[horms+renesas@verge.net.au: Do not add trailing blank line to irqs.h]
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

The cache size should already be present in the L2 cache auxiliary
control register: it is part of the integration process to configure
the hardware IP.  Most platforms get this right, yet still many
cargo-cult program, and assume that they always need specifying to
the L2 cache code.  Remove them so we can find out which really need
this.
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Since we now automatically enable early BRESP in core L2C-310 code when
we detect a Cortex-A9, we don't need platforms/SoCs to set this bit
explicitly.  Instead, they should seek to preserve the value of bit 30
in the auxiliary control register.
Acked-by: NTony Lindgren <tony@atomide.com>
Acked-by: NStephen Warren <swarren@nvidia.com>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

The TMU (Timer Unit) driver implements a new style of platform data that
handles the timer as a single device with multiple channel. Switch from
the old-style platform data to the new-style platform data.
Signed-off-by: NLaurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

The field will be removed from the sh-sci driver. Don't set it and let
the driver handle baud rate calculation internally.
Signed-off-by: NLaurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Passing the register base address and IRQ through platform data is
deprecated. Use resources instead.
Signed-off-by: NLaurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Replace copy-n-paste SCIF platform data and device declaration with a
macro. This reduces the amount of code and improves readability.
Signed-off-by: NLaurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

sh_eth driver which needs platform data at the time of
registration is not used.
Now, ARM/shmobile aims to support DT,
and the C code base board support will be removed
if DT support is completed.
Current driver registration method which needs platform data
and which is not shared complicates codes.
This means legacy C code cleanup after DT supporting
will be more complicated
This patch removes r8a7779_add_ether_device()
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

VIN driver which needs platform data at the time of
registration is used from Marzen only.
Now, ARM/shmobile aims to support DT,
and the C code base board support will be removed
if DT support is completed.
Current driver registration method which needs platform data
and which is not shared complicates codes.
This means legacy C code cleanup after DT supporting
will be more complicated
This patch registers it on board code as cleanup C code
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Acked-by: NMagnus Damm <damm@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

r8a7779 INTC needs IRL pin mode settings to determine
behavior of IRQ0 - IRQ3, and r8a7779_init_irq_extpin()
is controlling it via irlm parameter.
But this function registers renesas_intc_irqpin driver
if irlm was set, and this value depends on platform.
This is not good for DT.
This patch splits r8a7779_init_irq_extpin() function
into "mode settings" and "funtion register" parts
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Add HPB-DMAC platform device on R8A7779 SoC along with its slave and channel
configurations (only for SDHI0 so far).
Signed-off-by: NMax Filippov <max.filippov@cogentembedded.com>
[Sergei: moved *enum* declaring HPB-DMAC slave IDs from now removed <mach/dma.h>
to <mach/r8a7779.h>, removed #include <mach/dma.h> from setup-r8a7779.c, removed
SSI-related *enum* values and SSI-related data from hpb_dmae_slaves[] and
hpb_dmae_channels[], added ASYNCMDR.ASBTMD{20|24|43} and ASYNCMDR.ASMD{20|24|43}
fields/values, fixed comments to ASYNCMDR.ASBTMD2[123] and ASYNCMDR.ASMD2[123]
fields/values, renamed all the bit/field/value #define's to include 'HBP_DMAE_'
prefix to match the driver, moved comments after the element initializers of
hpb_dmae_channels[].]
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

r8a7779 INTC needs IRL pin mode settings to determine
behavior of IRQ0 - IRQ3, and r8a7779_init_irq_extpin()
is controlling it via irlm parameter.
But this function registers renesas_intc_irqpin driver
if irlm was set, and this value depends on platform.
This is not good for DT.
This patch splits r8a7779_init_irq_extpin() function
into "mode settings" and "funtion register" parts
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Add HPB-DMAC platform device on R8A7779 SoC along with its slave and channel
configurations (only for SDHI0 so far).
Signed-off-by: NMax Filippov <max.filippov@cogentembedded.com>
[Sergei: moved *enum* declaring HPB-DMAC slave IDs from now removed <mach/dma.h>
to <mach/r8a7779.h>, removed #include <mach/dma.h> from setup-r8a7779.c, removed
SSI-related *enum* values and SSI-related data from hpb_dmae_slaves[] and
hpb_dmae_channels[], added ASYNCMDR.ASBTMD{20|24|43} and ASYNCMDR.ASMD{20|24|43}
fields/values, fixed comments to ASYNCMDR.ASBTMD2[123] and ASYNCMDR.ASMD2[123]
fields/values, renamed all the bit/field/value #define's to include 'HBP_DMAE_'
prefix to match the driver, moved comments after the element initializers of
hpb_dmae_channels[].]
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

a894fcc2 ("ARM: smp_twd: Divorce smp_twd
from local timer API") altered twd_local_timer_common_register() so that it
may make use of late_timer_init.

This is problematic on marzen with Magnus's recent patch "ARM: shmobile:
marzen: Switch to DT_MACHINE_START" which switches marzen around to enable
USE_OF and thus shmobile_timer_init(), which is registered as
late_time_init by shmobile_earlytimer_init() stops being a no-op.

As a work-around I have updated r8a7779_earlytimer_init() so that
shmobile_earlytimer_init() is called after r8a7779_register_twd().
Or in other words, the shmobile_earlytimer_init() setting of
late_time_init overwrites that of twd_local_timer_common_register().
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Acked-by: NMagnus Damm <damm@opensource.se>

Add VIN clocks and platform devices for R8A7779 SoC; add function to register
the VIN platform devices.
[Sergei: added 'id' parameter check to r8a7779_add_vin_device(), used '*pdata'
in *sizeof* operator there, renamed some variables, annotated vin[0-3]_resources
[] and 'vin[0-3]_info' as '__initdata'.]
Signed-off-by: NVladimir Barinov <vladimir.barinov@cogentembedded.com>
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Acked-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

This patch moves r8a7779_init_irq_xxx() to setup code,
and remove intc-r8a7779.

Now, r8a7779_init_irq_extpin() uses
platform_device_register_resndata() instead of
platform_device_register()
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

This patch adds missing __initdata to driver data/resource
which are used from platform_device_register_xxx()
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Leave ->init_time() set to NULL to use the default ARM behaviour.
Signed-off-by: NMagnus Damm <damm@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

usb phy driver which needs platform data at the time of
registration is used from Marzen only.
Now, ARM/shmobile aims to support DT,
and the C code base board support will be removed
if DT support is completed.
Current driver registration method which needs platform data
and which is not shared complicates codes.
This means legacy C code cleanup after DT supporting
will be more complicated
This patch registers it on board code as cleanup C code
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Now when the intc-irqpin driver is used for external
IRQ pins on r8a7779 it possible to get rid of unused
INTC headers mach/intc.h and linux/sh_intc.h.
Signed-off-by: NMagnus Damm <damm@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Remove unused <mach/hardware.h> on r8a7779.
Signed-off-by: NMagnus Damm <damm@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Pass NULL to of_platform_populate instead of passing
an empty list. This cleans up the code and reduces the
number of lines.
Signed-off-by: NMagnus Damm <damm@opensource.se>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

Since we're now going to setup the USBPCTRL0 register using the USB PHY device's
platform data, we now need a way to pass those platform data from the board file
to the device which is situated in setup-r8a7779.c -- and what I'm suggesting is
r8a7779_add_usb_phy_device() that will register USB PHY platform device with the
passed platform data using platform_device_register_resndata() call; creating
this function involves deletion of 'usb_phy_device' from r8a7779_devices_dt[],
so that it will no longer be registered for the generic R8A7779 machine (where
we can't provide the platform data anyway), hence EHCI/OHCI drivers will fail
to load as well.

For the Marzen board, this new function will be called from marzen_init() to
register the USB PHY device early enough.

Note that the board and the SoC code have to be in one patch to keep the code
bisectable...

The patch has been tested on the Marzen board.
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Acked-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
[horms+renesas@verge.net.au: manually applied]
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>

The memory region that is used by the driver overlaps EHCI and OHCI  register
regions for absolutely no reason now  -- fix it  by adding offset of 0x800 to
the base address, changing the register #define's accordingly. This has extra
positive effect that we now can use devm_ioremap_resource()...

Note that the driver and the SoC code have to be in one patch to keep the code
bisectable...

The patch has been tested on the Marzen board.
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Acked-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Acked-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>