Commit 2b05f6ae1e ("ARM: ux500: remove PMU IRQ bouncer")
deleted some code to bounce and work around the weird PMU
IRQs in the DB8500 ASIC, but did a semantic mistake:
since the auxdata was now unused, the call to
of_platform_populate() was removed, but this does not
work: the default platform population will only kick in
if .init_machine() is assigned NULL, and since the U8540
was still using the callback that was not the case.
Fix this by reinstating the call to of_platform_populate(),
but pass NULL as auxdata.
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Fixes: 2b05f6ae1e ("ARM: ux500: remove PMU IRQ bouncer")
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
The ux500 PMU IRQ bouncer is getting in the way of some fundametnal
changes to the ARM PMU driver, and it's the only special case that
exists today. Let's remove it.
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
The commit afece3ab9a ("PM / Domains: Add time accounting to various
genpd states") causes a boot regression for ux500.
The problem occurs when the ux500 machine code calls pm_genpd_init(), which
since the above change triggers a call to ktime_get(). More precisely,
because ux500 initializes PM domains in the init_IRQ() phase of the boot,
timekeeping has not yet been initialized.
Fix the problem by moving the initialization of the PM domains to after
timekeeping has been initialized.
Fixes: afece3ab9a ("PM / Domains: Add time accounting to various genpd..")
Cc: Thara Gopinath <thara.gopinath@linaro.org>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Since the PMU register interface is banked per CPU, CPU PMU interrrupts
cannot be handled by a CPU other than the one with the PMU asserting the
interrupt. This means that migrating PMU SPIs, as we do during a CPU
hotplug operation doesn't make any sense and can lead to the IRQ being
disabled entirely if we route a spurious IRQ to the new affinity target.
This has been observed in practice on AMD Seattle, where CPUs on the
non-boot cluster appear to take a spurious PMU IRQ when coming online,
which is routed to CPU0 where it cannot be handled.
This patch passes IRQF_PERCPU for PMU SPIs and forcefully sets their
affinity prior to requesting them, ensuring that they cannot
be migrated during hotplug events. This interacts badly with the DB8500
erratum workaround that ping-pongs the interrupt affinity from the handler,
so we avoid passing IRQF_PERCPU in that case by allowing the IRQ flags
to be overridden in the platdata.
Fixes: 3cf7ee98b8 ("drivers/perf: arm_pmu: move irq request/free into probe")
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
The pen hold/release scheme was copied over to Ux500 from the ARM
reference designs like most of these at the time. It is not needed
at all, and was mostly removed in commit c00def71ef
"ARM: ux500: simplify secondary CPU boot".
However on the suspend/resume path and hot plug/unplug of CPUs,
the .cpu_die() callback was still waiting for the pen to be
released which made it spin forever and the second core never come
back online after suspend/resume.
Fix this by simply replacing the strange custom .cpu_die() with
a oneline wfi() just like e.g. the qcom platform does. This fixes
the issue and makes the second core come up properly after
suspend/resume.
As a side effect, this rids us of the completely surplus local
setup.h and hotplug.c files, and we just compile this into platsmp.c
with everything else SMP.
Cc: stable@vger.kernel.org
Fixes: c00def71ef ("ARM: ux500: simplify secondary CPU boot")
Cc: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
This platform data is revoked: the drivers are getting the DMA
configuration from the device tree, it has been done like that
since the DMA support was merged and this data has not been used
since. The remaining auxdata is also unused.
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
The cpu.c and cache-l2x0.c files hold only two or three simple
functions each, and they are all called from the machine
descriptors, so we can just move them all into the same file
for simplicity and consistency.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
As the ux500 id code is basically a standalone driver, we can move it
out of the arch code into drivers/soc/ux500.
This is a user-visible change, as it moves all the devices in sysfs
from /sys/devices/soc0/ to /sys/devices/ and leaves the soc0 node as a
separate device.
Originally the idea was to put all on-chip devices under the soc node,
and ux500 was the first platform to have this device, but later platforms
almost all didn't follow that pattern, so this makes the platform do
the same thing as everyone else.
Since the platform is really obsolete now, I am optimistic that nothing
will break after moving the devices around.
As the SoC driver no longer has access to the private header files,
I'm changing the code to instead look up the address of the backupram
from devicetree, which is a good idea anyway.
Finally, having a separate Kconfig symbol means the driver is now
optional and could even be a loadable module rather than always being
built-in if we allowed that for soc_device.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
[Fixup mising Makefile, fixup BB_UID_BASE to fc0]
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
ux500_setup_id is currently called from u8500_map_io(), which is
really early, but nothing relies on the ID any more, other than
a printk message that is not really all that important to
have early during boot.
If we move the call to ux500_setup_id() into ux500_soc_device_init(),
that file becomes usuable almost entirely standalone, and we can kill
off the u8500_map_io() callback as it just does the default
debug_ll_io_init() now.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Nothing else uses the global dbx500_asic_id structure, so
we can merge the two small files that reference it into one.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The generic IRQ init function also enables the l2 cache
implicitly when the machine descriptor sets an .l2c_aux_mask.
Let's use that on ux500 and remove the ux500_l2x0_init()
along with the cpu_is_u8500_family checks.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Except for the constant DB8500_PRCMU_FW_VERSION_OFFSET number, nothing
is ever passed through the platform data and used in a driver, so we
can simply stop passing it around.
Acked-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
There is only one instance of ab8500_regulator_platform_data, and it's
safe to assume we won't ever merge another one, so it's rather pointless
to pass it through multiple levels of platform data pointers.
This moves the structure and everything referenced by it into the
driver that uses it.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The code was executing a return with a pointer before reaching
iounmap().
Reported-by: David Binderman <dcb314@hotmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
The ARM core kernel already calls clocksource_of_init() so why
go to all the trouble of locating and probing this node in the
machine. CLOCKSOURCE_OF_DECLARE() will take care of it in the
clocksource driver, and thus we can also get rid of the
dangling header file <linux/clksrc-dbx500-prcmu.h>
Suggested-by: Arnd Bergmann <arndb@linaro.org>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Pull ARM development updates from Russell King:
"Included in this update:
- moving PSCI code from ARM64/ARM to drivers/
- removal of some architecture internals from global kernel view
- addition of software based "privileged no access" support using the
old domains register to turn off the ability for kernel
loads/stores to access userspace. Only the proper accessors will
be usable.
- addition of early fixup support for early console
- re-addition (and reimplementation) of OMAP special interconnect
barrier
- removal of finish_arch_switch()
- only expose cpuX/online in sysfs if hotpluggable
- a number of code cleanups"
* 'for-linus' of git://ftp.arm.linux.org.uk/~rmk/linux-arm: (41 commits)
ARM: software-based priviledged-no-access support
ARM: entry: provide uaccess assembly macro hooks
ARM: entry: get rid of multiple macro definitions
ARM: 8421/1: smp: Collapse arch_cpu_idle_dead() into cpu_die()
ARM: uaccess: provide uaccess_save_and_enable() and uaccess_restore()
ARM: mm: improve do_ldrd_abort macro
ARM: entry: ensure that IRQs are enabled when calling syscall_trace_exit()
ARM: entry: efficiency cleanups
ARM: entry: get rid of asm_trace_hardirqs_on_cond
ARM: uaccess: simplify user access assembly
ARM: domains: remove DOMAIN_TABLE
ARM: domains: keep vectors in separate domain
ARM: domains: get rid of manager mode for user domain
ARM: domains: move initial domain setting value to asm/domains.h
ARM: domains: provide domain_mask()
ARM: domains: switch to keeping domain value in register
ARM: 8419/1: dma-mapping: harmonize definition of DMA_ERROR_CODE
ARM: 8417/1: refactor bitops functions with BIT_MASK() and BIT_WORD()
ARM: 8416/1: Feroceon: use of_iomap() to map register base
ARM: 8415/1: early fixmap support for earlycon
...
This removes a lot of ancient cruft from the Ux500 SMP boot.
Instead of the pen grab/release, just point the ROM to
secondary_boot() and start the second CPU there, then send
the IPI.
Use our own SMP enable method. This enables us to remove the
last static mapping and get both CPUs booting properly.
Tested this and it just works.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Olof Johansson <olof@lixom.net>
To enable sharing of the arm_pmu code with arm64, this patch factors it
out to drivers/perf/. A new drivers/perf directory is added for
performance monitor drivers to live under.
MAINTAINERS is updated accordingly. Files added previously without a
corresponsing MAINTAINERS update (perf_regs.c, perf_callchain.c, and
perf_event.h) are also added.
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Will Deacon <will.deacon@arm.com>
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
[will: augmented Kconfig help slightly]
Signed-off-by: Will Deacon <will.deacon@arm.com>
This function is not mapping any I/O other than for probing
the chip ID. Rename it to make the function graspable.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This removes the static map for the PRCMU in favor of looking
up that resource from the device tree.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The l2x0 level 2 cache initialization used a static map
to get at the l2x0 registers. Get rid of this by getting
the register range from the device tree and just remap it
for the short time we need it.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
After moving the SMP and PM code to use dynamic mappings from
the device tree, we can delete the static maps for SCU and
backup RAM.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The GIC distributor was mapped statically for the PM code
but the driver already get the base address from the DT.
Augment the PM code to get the base from the device tree
and remove the static map.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Delete the static maps for the Ux500 UARTs and rely
on debug_ll_io_init() to do the trick.
commit f87b95dd2e
"ARM: ux500: move debugmacro to debug includes"
defined the necessary addruart macro two years ago.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The static mappings of the GPIO blocks, clock/reset controller
and MTU timer are clearly superfluous. These are all probed
exclusively from the device tree. Tested on the U8500.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Symbols local to this file are made static.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Since all platform data has been transfered to DT, we don't need the
container for the mmci platform data anymore. Remove the file and the
corresponding references to it's data.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
The DBx500 and ABx500 should be getting their IRQs from the
device tree and nowhere else. Get rid of all the static assignments
everywhere, delete it from the driver, platform data and the
board files in one swift strike.
Lots of cross-dependencies in the MFD drivers for PRCMU and
AB8500 makes it necessary to strike everywhere at once to
eradicate IRQs passed as resources and platform data to the left
and right around the platform.
Cc: Mark Brown <broonie@kernel.org>
Cc: Samuel Ortiz <sameo@linux.intel.com>
Acked-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Supporting device tree on the Ux500 is not optional anymore,
so delete the config option and compile this in at all times.
Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This moves the set-up of the HREF500 with its AB8505 ASIC to
a device tree include. Since there is not yet any device tree
for this board the DTSI is currently unused. After this delete
the board file for pins for good and migration of pins to the
device tree is complete.
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This moves the AB8500 pin settings for the clock out pins over
to the device tree. We can delete the special setup calls for the
platforms only using the AB8500 and not AB8505.
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
- Delete some unused U8540 UART platform data.
- MMC/SD-fixes from Ulf Hansson so everything works with
the device tree again.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=vC9c
-----END PGP SIGNATURE-----
Merge tag 'ux500-devicetree-v3.14-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson into next/dt
From Linus Walleij:
Ux500 device tree patches for v3.14, take 2
- Delete some unused U8540 UART platform data.
- MMC/SD-fixes from Ulf Hansson so everything works with
the device tree again.
* tag 'ux500-devicetree-v3.14-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson:
ARM: ux500: regulators: Remove dead code for SD-card regulator
ARM: ux500: Configure regulator for I/O voltage for SD-card slot
ARM: ux500: Refactor common DT configs for sdi[n] devices
ARM: ux500: delete U8540 UART auxdata
Signed-off-by: Olof Johansson <olof@lixom.net>
The other Ux500's does not need this anymore, and the U8540
certainly is no different.
Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This reverts commit 49c129519a.
The special settings for the SD/MMC card parameters that are not
yet fully agreed upon how to encode into the device tree went missing
with this commit. We need to first put it into the device tree,
and then remove the platform data.
Cc: Lee Jones <lee.jones@linaro.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This reverts commit ce16feb851.
This commit stopped tying down the name of the MMC/SD devices, but these
names are used in the pin control table, so the MMC/SD cards stopped
working.
Cc: Lee Jones <lee.jones@linaro.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This deletes the AUXDATA hammering down the Nomadik pin
controller name to "pinctrl-db8500". We have removed all
dependencies on this name.
Acked-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Move the platform data from all these files into one, delete empty
files and remove all references to them.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
There are no more devices which require left to register which require
a parent. This is part of the ux500 conversion over to Device Tree only
booting.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
It's time to remove all ATAG support from ux500 and rely solely on
Device Tree booting. This patch is part of that endeavour.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
It's time to remove all ATAG support from ux500 and rely solely on
Device Tree booting. This patch is part of that endeavour.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
It's time to remove all ATAG support from ux500 and rely solely on
Device Tree booting. This patch is part of that endeavour.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
It's time to remove all ATAG support from ux500 and rely solely on
Device Tree booting. This patch is part of that endeavour.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
It's time to remove all ATAG support from ux500 and rely solely on
Device Tree booting. This patch is part of that endeavour.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
It's time to remove all ATAG support from ux500 and rely solely on
Device Tree booting. This patch is part of that endeavour.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
We still need to utilise the AUXDATA system for the PRCMU to pass
through platform data which can not be DT:ed i.e. regulator initialisation
values. All we're doing in this patch is changing the comment header to be
more accurate.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
