Make Octeon GPIO IRQ chip CPU hotplug-aware
Seems that irq_cpu_offline callbacks were forgotten in v1 and v2 CIU
GPIO chips. There is such a callback for octeon_irq_chip_ciu2_gpio,
covering CIU2 chips. Without this callback GPIO IRQs are not being migrated
during core offlining. Patch is tested on Octeon II.
Signed-off-by: Alexander Sverdlin <alexander.sverdlin@nsn.com>
Cc: David Daney <ddaney.cavm@gmail.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/8201/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
__get_cpu_var() is used for multiple purposes in the kernel source. One of
them is address calculation via the form &__get_cpu_var(x). This calculates
the address for the instance of the percpu variable of the current processor
based on an offset.
Other use cases are for storing and retrieving data from the current
processors percpu area. __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.
This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset. Thereby address calculations are avoided and less registers
are used when code is generated.
At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.
The patch set includes passes over all arches as well. Once these operations
are used throughout then specialized macros can be defined in non -x86
arches as well in order to optimize per cpu access by f.e. using a global
register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu
variable.
DEFINE_PER_CPU(int, y);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(&x, this_cpu_ptr(&y), sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
__this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
__this_cpu_inc(y)
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
In octeon_3xxx.dts file, there is a definiton for twsi/twsi2 interrupts.
But there is no code for initialization of this interrupts. This patch adds
code for initialization of twsi interrupts.
Signed-off-by: Eunbong Song <eunb.song@samsung.com>
Cc: linux-kernel@vger.kernel.org
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/6816/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The current implementation of irq_set_affinity() refuses rightfully to
route an interrupt to an offline cpu.
But there is a special case, where this is actually desired. Some of
the ARM SoCs have per cpu timers which require setting the affinity
during cpu startup where the cpu is not yet in the online mask.
If we can't do that, then the local timer interrupt for the about to
become online cpu is routed to some random online cpu.
The developers of the affected machines tried to work around that
issue, but that results in a massive mess in that timer code.
We have a yet unused argument in the set_affinity callbacks of the irq
chips, which I added back then for a similar reason. It was never
required so it got not used. But I'm happy that I never removed it.
That allows us to implement a sane handling of the above scenario. So
the affected SoC drivers can add the required force handling to their
interrupt chip, switch the timer code to irq_force_affinity() and
things just work.
This does not affect any existing user of irq_set_affinity().
Tagged for stable to allow a simple fix of the affected SoC clock
event drivers.
Reported-and-tested-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Cc: Tomasz Figa <t.figa@samsung.com>,
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>,
Cc: Kukjin Kim <kgene.kim@samsung.com>
Cc: linux-arm-kernel@lists.infradead.org,
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/20140416143315.717251504@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
commit 3747069b25e419f6b51395f48127e9812abc3596 upstream.
The __cpuinit type of throwaway sections might have made sense
some time ago when RAM was more constrained, but now the savings
do not offset the cost and complications. For example, the fix in
commit 5e427ec2d0 ("x86: Fix bit corruption at CPU resume time")
is a good example of the nasty type of bugs that can be created
with improper use of the various __init prefixes.
After a discussion on LKML[1] it was decided that cpuinit should go
the way of devinit and be phased out. Once all the users are gone,
we can then finally remove the macros themselves from linux/init.h.
Note that some harmless section mismatch warnings may result, since
notify_cpu_starting() and cpu_up() are arch independent (kernel/cpu.c)
and are flagged as __cpuinit -- so if we remove the __cpuinit from
the arch specific callers, we will also get section mismatch warnings.
As an intermediate step, we intend to turn the linux/init.h cpuinit
related content into no-ops as early as possible, since that will get
rid of these warnings. In any case, they are temporary and harmless.
Here, we remove all the MIPS __cpuinit from C code and __CPUINIT
from asm files. MIPS is interesting in this respect, because there
are also uasm users hiding behind their own renamed versions of the
__cpuinit macros.
[1] https://lkml.org/lkml/2013/5/20/589
[ralf@linux-mips.org: Folded in Paul's followup fix.]
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/5494/
Patchwork: https://patchwork.linux-mips.org/patch/5495/
Patchwork: https://patchwork.linux-mips.org/patch/5509/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Use irq_get_trigger_type() to get the IRQ trigger type flags
instead calling irqd_get_trigger_type(irq_desc_get_irq_data(irq))
Signed-off-by: Javier Martinez Canillas <javier.martinez@collabora.co.uk>
Acked-by: David Daney <david.daney@cavium.com>
Cc: Grant Likely <grant.likely@linaro.org>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Samuel Ortiz <sameo@linux.intel.com>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-mips@linux-mips.org
Link: http://lkml.kernel.org/r/1371228049-27080-7-git-send-email-javier.martinez@collabora.co.uk
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Current GPIO chip implementation in octeon-irq is still broken, even after upstream
commit 87161ccdc6 (MIPS: Octeon: Fix broken interrupt
controller code). It works for GPIO IRQs that have reset-default configuration, but
not for edge-triggered ones.
The problem is in octeon_irq_gpio_map_common(), which passes modified "hw" variable
(which has range of possible values 16..31) as "gpio_line" parameter to
octeon_irq_set_ciu_mapping(), which saves it in private data of the IRQ chip. Later,
neither octeon_irq_gpio_setup() is able to re-configure GPIOs (cvmx_write_csr() is
writing to non-existent CVMX_GPIO_BIT_CFGX), nor octeon_irq_ciu_gpio_ack() is able
to acknowledge such IRQ, because "mask" is incorrect.
Fix is trivial and has been tested on Cavium Octeon II -based board, including
both level-triggered and edge-triggered GPIO IRQs.
Signed-off-by: Alexander Sverdlin <alexander.sverdlin.ext@nsn.com>
Cc: David Daney <david.daney@cavium.com>
Acked-by: David Daney <david.daney@cavium.com>
Patchwork: http://patchwork.linux-mips.org/patch/4980/
Acked-by: John Crispin <blogic@openwrt.org>
Having received another series of whitespace patches I decided to do this
once and for all rather than dealing with this kind of patches trickling
in forever.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The patch needs to eliminate the definition of OCTEON_IRQ_BOOTDMA so
that the device tree code can map the interrupt, so in order to not
temporarily break things, we do a single patch to both the interrupt
registration code and the pata_octeon_cf driver.
Also rolled in is a conversion to use hrtimers and corrections to the
timing calculations.
Acked-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: David Daney <david.daney@cavium.com>
For CIUv1 controllers, we were relying on all calls to the irq_chip
functions to be done from the CPU that received the irq, and that they
would all be done from interrupt contest. These assumptions do not
hold for threaded handlers.
We make all the masking actually mask the irq source, and use real
raw_spin_locks instead of manually twiddling the Status[IE] bit.
Signed-off-by: David Daney <david.daney@cavium.com>
The cn68XX has a new interrupt controller named CIU2, add support for
this, and use it if cn68XX detected at runtime.
Signed-off-by: David Daney <david.daney@cavium.com>
Since 3.6.0-rc1, We are getting many messages like:
WARNING: at kernel/irq/irqdomain.c:444 irq_domain_associate_many+0x23c/0x260()
Modules linked in:
Call Trace:
[<ffffffff814cb698>] dump_stack+0x8/0x34
[<ffffffff81133d00>] warn_slowpath_common+0x78/0xa8
[<ffffffff81187e44>] irq_domain_associate_many+0x23c/0x260
[<ffffffff81187f38>] irq_create_mapping+0xd0/0x220
[<ffffffff81188104>] irq_create_of_mapping+0x7c/0x158
[<ffffffff813e5f08>] irq_of_parse_and_map+0x28/0x40
.
.
.
Both the CIU and GPIO interrupt domains were somewhat screwed up.
For the CIU domain, we need to call irq_domain_associate() for each of
the preassigned irq numbers. For the GPIO domain, we were applying
the register bit offset in octeon_irq_gpio_xlat, but it should be done
in octeon_irq_gpio_map instead.
Also: Reserve all 8 'core' irqs for the 'core' irq_chip so that they
don't get used by the other domains. Remove unused OCTEON_IRQ_*
symbols.
Signed-off-by: David Daney <david.daney@cavium.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/4190/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Switch to using the device tree to register serial ports.
Add all the ports with compatible = "cavium,octeon-3860-uart". Octeon serial
ports have their own device type, required port flags, and I/O
functions, so using of_serial.c is not indicated.
We need to do this as late_initcall, as the 8250 driver must be
initialized before we add any ports. 8250 initialization is done at
device_initcall time.
The OCTEON_IRQ_UART{0,1,2} symbols are removed as they are now unused
and interfere with irq_domain used by the device tree code.
Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: Grant Likely <grant.likely@secretlab.ca>
Cc: linux-mips@linux-mips.org
Cc: devicetree-discuss@lists.ozlabs.org
Cc: Rob Herring <rob.herring@calxeda.com>
Cc: linux-kernel@vger.kernel.org
Cc: David Daney <david.daney@cavium.com>
Patchwork: https://patchwork.linux-mips.org/patch/3942/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The device tree will supply the register bank base addresses, make
register addressing relative to those. PHY connection is now
described by the device tree.
The OCTEON_IRQ_MII{0,1} symbols are also removed as they are now
unused and interfere with the irq_domain used for device tree irq
mapping.
Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: linux-mips@linux-mips.org
Cc: devicetree-discuss@lists.ozlabs.org
Cc: Grant Likely <grant.likely@secretlab.ca>
Cc: Rob Herring <rob.herring@calxeda.com>
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/3941/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
There are three parts to this:
1) Remove the definitions of OCTEON_IRQ_TWSI and OCTEON_IRQ_TWSI2.
The interrupts are specified by the device tree and these hard
coded irq numbers block the used of the irq lines by the irq_domain
code.
2) Remove platform device setup code from octeon-platform.c, it is
now unused.
3) Convert i2c-octeon.c to use device tree. Part of this includes
using the devm_* functions instead of the raw counterparts, thus
simplifying error handling. No functionality is changed.
Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: Rob Herring <rob.herring@calxeda.com>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
Cc: linux-mips@linux-mips.org
Cc: devicetree-discuss@lists.ozlabs.org
Cc: Grant Likely <grant.likely@secretlab.ca>
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/3939/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
We can consolidate octeon_irq_chip_ciu_edge and octeon_irq_chip_ciu as
they only differ in the .irq_ack element, and that is unused by the
level handler. This gets rid of a bunch of duplicate definitions.
Signed-off-by: David Daney <david.daney@cavium.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/3931/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The follow-on patch to add irq_domain support will be the supported
method for using these irq lines, so get these defines out of the way
in preperation for that.
Signed-off-by: David Daney <david.daney@cavium.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/3930/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
This is needed for follow-on on patches for Octeon to use the Device
Tree to configure GPIO interrupts.
Signed-off-by: David Daney <david.daney@cavium.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/3932/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Make use of the IRQCHIP_ONOFFLINE_ENABLED flag and remove the
wrappers. Use irqd_irq_disabled() instead of desc->status, which will
go away.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: David Daney <ddaney@caviumnetworks.com>
This includes conversion to new style irq_chip functions, and
correctly enabling/disabling per-CPU interrupts.
The hardware interrupt bit to irq number mapping is now done with a
flexible map, instead of by bit twiddling the irq number.
[ tglx: Adjusted to new irq_cpu_on/offline callbacks and
__irq_set_affinity_lock ]
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Cc: linux-mips@linux-mips.org
Cc: ralf@linux-mips.org
LKML-Reference: <1301081931-11240-5-git-send-email-ddaney@caviumnetworks.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The original version went behind the back of everything, leaving
things in an inconsistent state.
Now we use the irq_set_affinity() to do the work for us. This has the
advantage that the IRQ core's view of the affinity stays consistent.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
To: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/1486/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The main change is to change most of the IRQs from handle_percpu_irq
to handle_fasteoi_irq. This necessitates extracting all the .ack code
to common functions that are not exposed to the irq core.
The affinity code now acts more sanely, by doing round-robin
distribution instead of broadcasting.
Because of the change to handle_fasteoi_irq and affinity, some of the
IRQs had to be split into separate groups with their own struct
irq_chip to prevent undefined operations on specific IRQ lines.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
To: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/1485/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Put all the MSI code in one place (msi-octeon.c). This simplifies
octeon-irq.c and gets rid of some ugly #ifdefs
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
To: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/1484/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
If an interrupt handler disables interrupts, the EOI function will
just reenable them. This will put us in an endless loop when the
upcoming Ethernet driver patches are applied.
Only reenable the interrupt on EOI if it is not IRQ_DISABLED. This
requires that the EOI function be separate from the ENABLE function.
We also rename the ACK functions to correspond with their function.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
To: linux-mips@linux-mips.org
To: gregkh@suse.de
Patchwork: http://patchwork.linux-mips.org/patch/840/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Some newer Octeon chips have registers that allow lockless operation of
the interrupt controller. Take advantage of them.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Since the locks are used from interrupt context we need the
irqsave/irqrestore versions of the locking functions.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Some of the were relying into smp.h being dragged in by another header
which of course is fragile. <asm/cpu-info.h> uses smp_processor_id()
only in macros and including smp.h there leads to an include loop, so
don't change cpu-info.h.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
This patch adds support for PCI and PCIe to the base Cavium OCTEON
processor support.
Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
according to Ingo, change set_affinity() in irq_chip should return int,
because that way we can handle failure cases in a much cleaner way, in
the genirq layer.
v2: fix two typos
[ Impact: extend API ]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: linux-arch@vger.kernel.org
LKML-Reference: <49F654E9.4070809@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup
Convert the last remaining users to struct irq_desc.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Alexander Sverdlin