Register a GPIO driver to access the GPIOs provided by the chip.
The GPIOs of the SoC should always start at 0 and the other GPIOs could
start at a random position. There is just one SoC in a system and when
they start at 0 the number is predictable.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Patchwork: http://patchwork.linux-mips.org/patch/4591Acked-by: NFlorian Fainelli <florian@openwrt.org>

The GPIOs are access through some registers in the chip common core or
over extif. We need locking around these GPIO accesses, all GPIOs are
accessed through the same registers and parallel writes will cause
problems.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Patchwork: http://patchwork.linux-mips.org/patch/4590Acked-by: NFlorian Fainelli <florian@openwrt.org>

Add functions to access the GPIO registers for pullup and pulldown.
These are needed for handling gpio registration.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Patchwork: http://patchwork.linux-mips.org/patch/4589Acked-by: NFlorian Fainelli <florian@openwrt.org>

Register a GPIO driver to access the GPIOs provided by the chip.
The GPIOs of the SoC should always start at 0 and the other GPIOs could
start at a random position. There is just one SoC in a system and when
they start at 0 the number is predictable.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Patchwork: http://patchwork.linux-mips.org/patch/4587Acked-by: NFlorian Fainelli <florian@openwrt.org>

Add description to the function.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Patchwork: http://patchwork.linux-mips.org/patch/4588Acked-by: NFlorian Fainelli <florian@openwrt.org>

Add functions to access the GPIO registers for pullup and pulldown.
These are needed for handling gpio registration.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Patchwork: http://patchwork.linux-mips.org/patch/4586Acked-by: NFlorian Fainelli <florian@openwrt.org>

The GPIOs are access through some registers in the chip common core.
We need locking around these GPIO accesses, all GPIOs are accessed
through the same registers and parallel writes will cause problems.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Patchwork: http://patchwork.linux-mips.org/patch/4585Acked-by: NFlorian Fainelli <florian@openwrt.org>

Signed-off-by: NGabor Juhos <juhosg@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4223Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Use the nlm_set_pic_extra_ack() call to setup the extra interrupt
ACK needed by XLR PCI and XLS PCIe. Simplify the code by adding
nlm_pci_link_to_irq().
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4561Signed-off-by: NJohn Crispin <blogic@openwrt.org>

BCM6345 has an intermediate 16-bits wide test control register between the
peripheral identifier register, and its clock control register is only 16-bits
wide contrary to other platforms where it is 32-bits wide. By shifting all
clocks bits by 16-bits to the left we ensure they get written to the proper
clock control register, without adding specific BCM6345 handling in the clock
code.
Signed-off-by: NFlorian Fainelli <florian@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4555/Signed-off-by: NJohn Crispin <blogic@openwrt.org>

The internal GPHYs need a firmware blob to function properly. This patch adds
the code needed to request the blob and load it to the PHY.
Signed-off-by: NJohn Crispin <blogic@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4523

The XRX200 family of SoCs has embedded gigabit PHYs. This patch adds code to
boot them up.
Signed-off-by: NJohn Crispin <blogic@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4522

Signed-off-by: NJohn Crispin <blogic@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4521

Print the hardware revision and port/channel info when starting the dma core.
Signed-off-by: NJohn Crispin <blogic@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4520

The XRX200 SoC family has a different register layout for reading the boot
selection bits.
Signed-off-by: NJohn Crispin <blogic@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4519

The bootmem was incorrectly freed resulting in lots of dangling pointers.
Additionally we should use of_platform_populate() as the Documentaion tells us
to do so.
Signed-off-by: NJohn Crispin <blogic@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4518

Fix a typo in the code.
Signed-off-by: NKelvin Cheung <keguang.zhang@gmail.com>
Patchwork: http://patchwork.linux-mips.org/patch/4434Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Update stmmac_mdio_bus_data accordingly due to the upstream change.
Signed-off-by: NKelvin Cheung <keguang.zhang@gmail.com>
Patchwork: http://patchwork.linux-mips.org/patch/4433Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Improve ls1x_serial_setup().
Signed-off-by: NKelvin Cheung <keguang.zhang@gmail.com>
Patchwork: http://patchwork.linux-mips.org/patch/4432Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Use common clock infrastructure instead of private APIs.
1. Enable COMMON_CLK in the Kconfig.
2. Remove private clock APIs, which are replaced by the code in
   drivers/clk/clk-ls1x.c.
3. Modify header file for drivers/clk/clk-ls1x.c.
Signed-off-by: NKelvin Cheung <keguang.zhang@gmail.com>
Patchwork: http://patchwork.linux-mips.org/patch/4431Signed-off-by: NJohn Crispin <blogic@openwrt.org>

On XLR/XLS, the cpu cores communicate with fast on-chip devices
(e.g. network accelerator, security engine etc.) using the Fast
Messaging Network(FMN). The FMN queues and credits needs to be
configured and intialized before it can be used.

The co-processor 2 on XLR/XLS CPU cores has registers for FMN access,
and the XLR/XLS has custom instructions for sending and loading
messages.  The FMN can deliver also per-cpu interrupts when messages
are available at the CPU.

This patch adds FMN initialization, adds interrupt setup and handling,
and also provides support for sending and receiving FMN messages.
Signed-off-by: NGanesan Ramalingam <ganesanr@broadcom.com>
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4468Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Create struct nlm_pic_irq for interrupts handled by the PIC.
This simplifies IRQ handling for multi-SoC as well as
the single SoC cases. Also split the setup of percpu and PIC
interrupts so that we can configure the PIC interrupts for
every node.
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4467Signed-off-by: NJohn Crispin <blogic@openwrt.org>

There can be 1, 2 or 4 SoCs(nodes) in a multi-chip XLP board. Add an
option for multi-chip boards in case of XLP, and make the number of
nodes configurable.
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4470Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Upto 4 Netlogic XLP SoCs can be connected over ICI links to form a
coherent multi-node system.  Each SoC has its own set of on-chip
devices including PIC.  To support this, add a per SoC stucture and
use it for the PIC and SYS block addresses instead of using global
variables.
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4469Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Initial code to support more than 32 cpus. The platform CPU mask
is updated from 32-bit mask to cpumask_t. Convert places that use
cpu_/cpus_ functions to use cpumask_* functions.
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4464Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Remove unused and trivial PIC accesss functions, update nlm_pic_send_ipi()
and nlm_set_irt_to_cpu() to use similar logic, and use correct type for
reg in nlm_pic_disable_irt().
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4463Signed-off-by: NJohn Crispin <blogic@openwrt.org>

The cpuid was not passed into early_init_secondary even though the
comment indicated that it will be. Fix this.
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4458Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Used the hardware thread id passed in while writing to IRT in
nlm_pic_init_irt()
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4465Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Fix Kconfig for both XLR and XLP to select ZONE_DMA32 (instead of ZONE_DMA)
in case of 64-bit compilation. This can be used for devices that can only
do DMA to 32-bit address. ZONE_DMA is not useful on XLR or XLP.
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4466Signed-off-by: NJohn Crispin <blogic@openwrt.org>

At this point early printk is available, so debugging device tree
issues is easier.
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4460Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Enable Speculative Unmap Enable bit, which will enable speculative L2
cache requests for unmapped memory. This should give better performance
for kernel code/data which is in KSEG0
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4461Signed-off-by: NJohn Crispin <blogic@openwrt.org>

This allows us to use the r2 optimized code from kernel headers
while compilation.

Disable PGD_C0_CONTEXT option for XLP, which does not work.
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4456Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Add support for XLP performance counters register in perf. Update
mips/Kconfig so that perf events can be selected for XLP.
Signed-off-by: NZi Shen Lim <zlim@netlogicmicro.com>
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4457Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Add support for XLR and XLS processors in MIPS Oprofile code. These
processors are multi-threaded and have two counters per core. Each
counter can track either all the events in the core (global mode),
or events in just one thread.

We use the counters in the global mode, and use only the first thread
in each core to handle the configuration etc.
Signed-off-by: NMadhusudan Bhat <mbhat@netlogicmicro.com>
Signed-off-by: NJayachandran C <jchandra@broadcom.com>
Patchwork: http://patchwork.linux-mips.org/patch/4471Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Refactor nvram related functions into its own unit for easier expansion
and exposure of the values to other drivers.
Signed-off-by: NJonas Gorski <jonas.gorski@gmail.com>
Patchwork: http://patchwork.linux-mips.org/patch/4516Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Use the new reset helper where appropriate.
Signed-off-by: NJonas Gorski <jonas.gorski@gmail.com>
Patchwork: http://patchwork.linux-mips.org/patch/4453Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Add a reset helper for resetting the different cores.
Signed-off-by: NJonas Gorski <jonas.gorski@gmail.com>
Patchwork: http://patchwork.linux-mips.org/patch/4455Signed-off-by: NJohn Crispin <blogic@openwrt.org>

The softreset register description for BCM6358 was missing, so add it.
Signed-off-by: NJonas Gorski <jonas.gorski@gmail.com>
Patchwork: http://patchwork.linux-mips.org/patch/4454Signed-off-by: NJohn Crispin <blogic@openwrt.org>

Add a PCIe clock and use that instead of directly touching the clock
control register. While at it, fail if there is no such clock.
Signed-off-by: NJonas Gorski <jonas.gorski@gmail.com>
Acked-by: NFlorian Fainelli <florian@openwrt.org>
Patchwork: http://patchwork.linux-mips.org/patch/4452Signed-off-by: NJohn Crispin <blogic@openwrt.org>

There are bcma based devices like the Linksys E2000 out there, which do
have one ieee80211 core, but no PCIe core and they are using no
prefixes for the sprom. In addition some values like boardtype are
stored without a prefix for the main SoC chip also when they have an
additional PCIe wifi chip with an own boardtype var on some devices.

The Ethernet addresses are now also read out correctly without a prefix
so calling bcm47xx_fill_sprom_ethernet is not needed any more.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Patchwork: http://patchwork.linux-mips.org/patch/4364Signed-off-by: NJohn Crispin <blogic@openwrt.org>