Merge tag 'mvebu-drivers-5.9-1' of git://git./linux/kernel/git/gclement/mvebu into arm/drivers

mvebu drivers for 5.9 (part 1)

For firmware on the Turris MOX (Armada 3720 based board), add support
ECDSA signatures via debugfs.

* tag 'mvebu-drivers-5.9-1' of git://git.kernel.org/pub/scm/linux/kernel/git/gclement/mvebu:
  firmware: turris-mox-rwtm: add debugfs documentation
  firmware: turris-mox-rwtm: support ECDSA signatures via debugfs

Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Merge tag 'qcom-drivers-for-5.9' of git://git./linux/kernel/git/qcom/linux into arm/drivers

Qualcomm driver updates for v5.9

For RPMh this fixes an issue where ktime was used during suspend, allows
the driver to be used on ARM targets and some minor cleanups.

It adds support for the latest format version in the socinfo driver and
adds identifiers for SM8250 and SDM630.

SMD-RPM gains compatibles for MSM8994 and MSM8936 and the Qualcomm SCM
gains compatibles MSM8994 and IPQ8074.

The GENI core code gains interconnect path voting and performance level
support, with subsequent patches integrating this with the SPI, I2C,
UART and QSPI drivers.

Following this the KGDB support for the GENI serial driver is improved,
the performance related to chip-select is improved for SPI and QSPI.

* tag 'qcom-drivers-for-5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux: (35 commits)
  soc: qcom: geni: Fix NULL pointer dereference
  tty: serial: qcom-geni-serial: Drop the icc bw votes in suspend for console
  serial: qcom_geni_serial: Always use 4 bytes per TX FIFO word
  serial: qcom_geni_serial: Make kgdb work even if UART isn't console
  spi: spi-geni-qcom: Get rid of most overhead in prepare_message()
  spi: spi-geni-qcom: Set the clock properly at runtime resume
  spi: spi-geni-qcom: Avoid clock setting if not needed
  spi: spi-qcom-qspi: Set an autosuspend delay of 250 ms
  spi: spi-qcom-qspi: Avoid clock setting if not needed
  spi: spi-qcom-qspi: Use OPP API to set clk/perf state
  firmware: qcom_scm: Add msm8994 compatible
  firmware: qcom_scm: Fix legacy convention SCM accessors
  <linux/of.h>: add stub for of_get_next_parent() to fix qcom build error
  dt-bindings: firmware: qcom: Add compatible for IPQ8074 SoC
  spi: spi-geni-qcom: Use OPP API to set clk/perf state
  tty: serial: qcom_geni_serial: Use OPP API to set clk/perf state
  spi: spi-qcom-qspi: Add interconnect support
  spi: spi-geni-qcom: Add interconnect support
  spi: spi-geni-qcom: Combine the clock setting code
  tty: serial: qcom_geni_serial: Add interconnect support
  ...

Link: https://lore.kernel.org/r/20200721044812.3429652-1-bjorn.andersson@linaro.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Merge tag 'imx-drivers-5.9' of git://git./linux/kernel/git/shawnguo/linux into arm/drivers

i.MX drivers change for 5.9:

- Update SCU irq code to call pm_system_wakeup() in general MU IRQ
  handler, so that system can be waked up when MU IRQ arrives.
- Move i.MX SCU soc driver into imx firmware folder to get it
  initialized from i.MX SCU firmware driver.
- Clean up soc-imx-scu driver a bit by using devm_kasprintf().
- Correct postfix setting for cm40 power domain in scu-pd driver.
- Add resource management support for IMX_SCU firmware driver.
- Add more cm4 resources to i.MX SCU power domain driver.
- Select ARM_GIC_V3 from SOC_IMX8M for being able to use GICv3 driver
  in AARCH32 mode Linux on AARCH64 hardware.

* tag 'imx-drivers-5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
  soc: imx: select ARM_GIC_V3 for i.MX8M
  firmware: imx: Move i.MX SCU soc driver into imx firmware folder
  firmware: imx: scu-pd: add more cm4 resources
  firmware: imx: add resource management api
  firmware: imx: scu-pd: fix cm40 power domain
  soc: imx: scu: use devm_kasprintf
  firmware: imx: make sure MU irq can wake up system from suspend mode

Link: https://lore.kernel.org/r/20200720085536.24138-1-shawnguo@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Merge tag 'reset-for-v5.9' of git://git.pengutronix.de/pza/linux into arm/drivers

Reset controller updates for v5.9

This tag moves the reset-simple header out of drivers/reset for use by
drivers outside of drivers/reset, adds a .reset() callback to
reset-simple, converts i.MX reset bindings to json-schema, fixes a
compile warning in the reset-intel-gw driver, and replaces some HTTP
links with HTTPS ones in comments.

* tag 'reset-for-v5.9' of git://git.pengutronix.de/pza/linux:
  reset: Replace HTTP links with HTTPS ones
  reset: intel: fix a compile warning about REG_OFFSET redefined
  dt-bindings: reset: Convert i.MX7 reset to json-schema
  dt-bindings: reset: Convert i.MX reset to json-schema
  reset: simple: Add reset callback
  reset: Move reset-simple header out of drivers/reset

Link: https://lore.kernel.org/r/b718f052e38abbaac599d80645376b75e54aa5bd.camel@pengutronix.de
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

soc: samsung: exynos-regulator-coupler: Add simple voltage coupler for Exynos5800

Add a simple custom voltage regulator coupler for Exynos5800 SoCs, which
require coupling between "vdd_arm" and "vdd_int" regulators. This coupler
ensures that the voltage values don't go below the bootloader-selected
operation point during the boot process until the clients set their
constraints. It is achieved by assuming minimal voltage value equal to
the current value if no constraints are set. This also ensures proper
voltage balancing if any of the client driver is missing.

The balancing code comes from the regulator/core.c with the additional
logic for handling regulators without client constraints applied added.

Link: https://lore.kernel.org/r/20200721180900.13844-5-krzk@kernel.org
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

soc: qcom: geni: Fix NULL pointer dereference

pdev struct doesn't exits for the devices whose status are disabled
from DT node, in such cases NULL is returned from 'of_find_device_by_node'
Later when we try to get drvdata from pdev struct NULL pointer dereference
is triggered.

Add a NULL check for return values to fix the issue.

We were hitting this issue when one of QUP is disabled.

Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Fixes: 048eb908a1f2 ("soc: qcom-geni-se: Add interconnect support to fix earlycon crash")
Reported-by: Sai Prakash Ranjan <saipraka@codeaurora.org>
Signed-off-by: Akash Asthana <akashast@codeaurora.org>
Link: https://lore.kernel.org/r/1594996342-26964-1-git-send-email-akashast@codeaurora.org
[bjorn: s/wrapper_pdev/pdev/]
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

reset: Replace HTTP links with HTTPS ones

Rationale:
Reduces attack surface on kernel devs opening the links for MITM
as HTTPS traffic is much harder to manipulate.

Deterministic algorithm:
For each file:
  If not .svg:
    For each line:
      If doesn't contain `\bxmlns\b`:
        For each link, `\bhttp://[^# \t\r\n]*(?:\w|/)`:
  If neither `\bgnu\.org/license`, nor `\bmozilla\.org/MPL\b`:
            If both the HTTP and HTTPS versions
            return 200 OK and serve the same content:
              Replace HTTP with HTTPS.

Signed-off-by: Alexander A. Klimov <grandmaster@al2klimov.de>
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>

reset: intel: fix a compile warning about REG_OFFSET redefined

kernel test robot reports a compile warning about REG_OFFSET redefined
in the reset-intel-gw.c after merging commit e44ab4e14d6f4 ("regmap:
Simplify implementation of the regmap_read_poll_timeout() macro"). the
warning is like that:

drivers/reset/reset-intel-gw.c:18:0: warning: "REG_OFFSET" redefined
 #define REG_OFFSET GENMASK(31, 16)

In file included from ./arch/arm/mach-ixp4xx/include/mach/hardware.h:30:0,
                 from ./arch/arm/mach-ixp4xx/include/mach/io.h:15,
                 from ./arch/arm/include/asm/io.h:198,
                 from ./include/linux/io.h:13,
                 from ./include/linux/iopoll.h:14,
                 from ./include/linux/regmap.h:20,
                 from drivers/reset/reset-intel-gw.c:12:
./arch/arm/mach-ixp4xx/include/mach/platform.h:25:0: note: this is the location of the previous definition
 #define REG_OFFSET 3

Reported-by: kernel test robot <lkp@intel.com>
Fixes: c9aef213e38cde ("reset: intel: Add system reset controller driver")
Signed-off-by: Dejin Zheng <zhengdejin5@gmail.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>

firmware: turris-mox-rwtm: add debugfs documentation

Add debugfs ABI documentation for the ECDSA signatures.

Signed-off-by: Marek Behún <marek.behun@nic.cz>
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>

firmware: turris-mox-rwtm: support ECDSA signatures via debugfs

The firmware on Turris MOX secure processor offers signing messages
with ECDSA private key stored in protected OTP memory.

The optimal solution would be to register an akcipher provider via
kernel's crypto API, but crypto API does not yet support accessing
akcipher API from userspace (and probably won't for some time, see
https://www.spinics.net/lists/linux-crypto/msg38388.html).

At first I tried to put this via standard sysfs API, but the way I
designed it is not compatible with sysfs's standard "one file per
attribute".

This patch therefore adds support for accessing this signature
generation mechanism via debugfs. Since CZ.NIC's Turris MOX is the only
user of this module, the potential future change to akcipher API should
not cause problems, since we can just change our userspace software then.

Signed-off-by: Marek Behún <marek.behun@nic.cz>
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>

Merge tag 'tegra-for-5.9-soc' of git://git./linux/kernel/git/tegra/linux into arm/drivers

soc/tegra: Changes for v5.9-rc1

This adds missing SoC IDs for Tegra186 and Tegra194 and fixes a typo in
a warning message.

* tag 'tegra-for-5.9-soc' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux:
  soc/tegra: fuse: Fix typo in APB MISC warning
  soc/tegra: fuse: Add Tegra186 and Tegra194 SoC IDs

Link: https://lore.kernel.org/r/20200717161300.1661002-4-thierry.reding@gmail.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Merge tag 'tegra-for-5.9-memory' of git://git./linux/kernel/git/tegra/linux into arm/drivers

memory: tegra: Changes for v5.9-rc1

This contains the Tegra210 EMC frequency scaling support that didn't
make it into v5.8. In addition there are a couple of cleanups and minor
fixes.

* tag 'tegra-for-5.9-memory' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux:
  memory: tegra: Add Tegra132 compatible string match
  memory: tegra: Fix KCONFIG variables for Tegra186 and Tegra194
  memory: tegra: Delete some dead code
  memory: tegra: Avoid unused function warnings
  memory: tegra: Drop <linux/clk-provider.h>
  memory: tegra: Fix an error handling path in tegra186_emc_probe()
  memory: tegra30-emc: Poll EMC-CaR handshake instead of waiting for interrupt
  memory: tegra20-emc: Poll EMC-CaR handshake instead of waiting for interrupt
  memory: tegra: Support derated timings on Tegra210
  memory: tegra: Add EMC scaling sequence code for Tegra210
  memory: tegra: Add EMC scaling support code for Tegra210
  memory: tegra: Make debugfs permissions human-readable

Link: https://lore.kernel.org/r/20200717161300.1661002-3-thierry.reding@gmail.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Merge tag 'tegra-for-5.9-firmware' of git://git./linux/kernel/git/tegra/linux into arm/drivers

firmware: tegra: Changes for v5.9-rc1

This has a few cleanups and the addition of a new mechanism to query
debug information from the BPMP.

* tag 'tegra-for-5.9-firmware' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux:
  firmware: tegra: Update BPMP ABI
  firmware: tegra: Add support for in-band debug
  firmware: tegra: Prepare for supporting in-band debugfs
  firmware: tegra: Use consistent return variable name
  firmware: tegra: Add return code checks and increase debugfs size

Link: https://lore.kernel.org/r/20200717161300.1661002-2-thierry.reding@gmail.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

firmware: arm_scmi: Use NULL instead of integer 0 for rate pointer

Kbuild test robot reports the following sparse warning:

drivers/firmware/arm_scmi/clock.c:142:21:
sparse: Using plain integer as NULL pointer

Use NULL pointer instead of integer 0 for rate pointer and fix the
warning.

Link: https://lore.kernel.org/r/20200717140405.17905-1-sudeep.holla@arm.com
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Merge tag 'renesas-drivers-for-v5.9-tag1' of git://git./linux/kernel/git/geert/renesas-devel into arm/drivers

Renesas driver updates for v5.9

  - Add core support for the new RZ/G2H (R8A774E1) SoC, including System
    Controller (SYSC) and Reset (RST) support.

* tag 'renesas-drivers-for-v5.9-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-devel:
  soc: renesas: rcar-rst: Add support for RZ/G2H
  soc: renesas: Identify RZ/G2H
  soc: renesas: Add Renesas R8A774E1 config option
  soc: renesas: rcar-sysc: Add r8a774e1 support
  clk: renesas: Add r8a774e1 CPG Core Clock Definitions
  dt-bindings: power: Add r8a774e1 SYSC power domain definitions

Link: https://lore.kernel.org/r/20200717112427.26032-3-geert+renesas@glider.be
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

soc/tegra: fuse: Fix typo in APB MISC warning

The hardware block is called APB MISC, not ABP MISC, so fix the warning
to use the correct name.

Signed-off-by: Thierry Reding <treding@nvidia.com>

soc/tegra: fuse: Add Tegra186 and Tegra194 SoC IDs

SoC IDs for these generations had never been defined. Do so now.

Signed-off-by: Sandipan Patra <spatra@nvidia.com>
Reviewed-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

Merge tag 'v5.8-next-soc' of git://git./linux/kernel/git/matthias.bgg/linux into arm/drivers

add new functions to cmdq helper functions
- assign value to register
- export finalize function and don't call explicitely from flush async
- set specific event

* tag 'v5.8-next-soc' of git://git.kernel.org/pub/scm/linux/kernel/git/matthias.bgg/linux:
  soc: mediatek: cmdq: add set event function
  soc: mediatek: cmdq: export finalize function
  soc: mediatek: cmdq: add assign function

Link: https://lore.kernel.org/r/01399fb4-b2d0-e41b-dfd9-f2deba0ef651@gmail.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

memory: tegra: Add Tegra132 compatible string match

Ensure that the driver will bind against the Tegra132 instantiation of
the external memory controller. While the two are roughly the same from
a capability perspective, they do require some incompatible changes to
the programming sequences and therefore need separate compatible
strings.

Signed-off-by: Thierry Reding <treding@nvidia.com>

firmware: tegra: Update BPMP ABI

Update the BPMP ABI to align with the the latest version.

Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

firmware: tegra: Add support for in-band debug

Add support for retrieving BPMP debug information via in-band messaging
as opposed to using shared-memory which older BPMP firmware used. Note
that it is possible to detect at runtime whether the BPMP firmware being
used supports the in-band messaging for retrieving the debug
informaation. Therefore, if the BPMP firmware supports the in-band
messaging for debug use this and otherwise fall-back to using shared
memory.

Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

firmware: tegra: Prepare for supporting in-band debugfs

Currently, BPMP debug information is accessible via the Linux debugfs
file-system using a shared-memory scheme. More recent BPMP firmware now
supports accessing the debug information by in-band messaging which does
not require shared-memory. To prepare for adding in-band debugfs support
for the BPMP, move the shared-memory specific initialisation from the
tegra_bpmp_init_debugfs() into a sub-function.

Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

firmware: tegra: Use consistent return variable name

Most functions in the BPMP driver use 'err' as the return variable
name but there are a few places that use 'ret'. Let's use 'err' to
be consistent.

Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

firmware: tegra: Add return code checks and increase debugfs size

Add checking of the BPMP-FW return code values for MRQ_DEBUGFS calls.

Also, development versions of the firmware may have debugfs with a
directory structure larger than 256 KiB. Hence increase the size of the
memory buffer to accommodate those firmware revisions.

And finally, ensure that no access outside of allocated memory buffer
happens in case BPMP-FW returns an invalid response size (nbytes) from
mrq_debugfs_dumpdir() call.

Signed-off-by: Timo Alho <talho@nvidia.com>
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

memory: tegra: Fix KCONFIG variables for Tegra186 and Tegra194

Commit a127e690b051 ("memory: tegra: Add support for the Tegra194 memory
controller") and commit 4e04b88633ae ("memory: tegra: Only include
support for enabled SoCs") incorrectly added the KCONFIG variables
CONFIG_ARCH_TEGRA186_SOC and CONFIG_ARCH_TEGRA194_SOC to the Tegra EMC
driver. These KCONFIG variables do not exist and prevent the EMC driver
from being probed on Tegra186 and Tegra194. These KCONFIG variable
names are simply missing one underscore and so fix this by adding the
necessary underscore to the variable names.

Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

Merge tag 'scmi-updates-5.9' of git://git./linux/kernel/git/sudeep.holla/linux into arm/drivers

ARM SCMI/SCPI updates for v5.9

The main addition for this time is the support for platform notifications.
SCMI protocol specification allows the platform to signal events to the
interested agents via notification messages. We are adding support for
the dispatch and delivery of such notifications to the interested users
inside the kernel.

Other than that, there are minor changes like checking and using the
fast_switch capability quering the firmware instead of doing it
unconditionally(using polling mode transfer), cosmetic trace update,
use of HAVE_ARM_SMCCC_DISCOVERY instead of ARM_PSCI_FW and a fix in
scmi clock registration logic for all the clocks with discrete rates.

* tag 'scmi-updates-5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux:
  firmware: arm_scmi: Remove fixed size fields from reports/scmi_event_header
  firmware: arm_scmi: Remove unneeded __packed attribute
  firmware: arm_scmi: Remove zero-length array in SCMI notifications
  firmware: arm_scmi: Provide a missing function param description
  clk: scmi: Fix min and max rate when registering clocks with discrete rates
  firmware: arm_scmi: Keep the discrete clock rates sorted
  firmware: arm_scmi: Add base notifications support
  firmware: arm_scmi: Add reset notifications support
  firmware: arm_scmi: Add sensor notifications support
  firmware: arm_scmi: Add perf notifications support
  firmware: arm_scmi: Add power notifications support
  firmware: arm_scmi: Enable notification core
  firmware: arm_scmi: Add notification dispatch and delivery
  firmware: arm_scmi: Add notification callbacks-registration
  firmware: arm_scmi: Add notification protocol-registration
  firmware: arm_scmi: Fix SCMI genpd domain probing
  firmware: arm_scmi: Use HAVE_ARM_SMCCC_DISCOVERY instead of ARM_PSCI_FW
  cpufreq: arm_scmi: Set fast_switch_possible conditionally
  firmware: arm_scmi: Add fast_switch_possible() interface
  firmware: arm_scmi: Use signed integer to report transfer status

Link: https://lore.kernel.org/r/20200713161410.12324-1-sudeep.holla@arm.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

tty: serial: qcom-geni-serial: Drop the icc bw votes in suspend for console

When using the geni-serial as console, its important to be
able to hit the lowest possible power state in suspend,
even with no_console_suspend.
The only thing that prevents it today on platforms like the sc7180
is the interconnect BW votes, which we certainly don't need when
the system is in suspend. So in the suspend handler mark them as
ACTIVE_ONLY (0x3) and on resume switch them back to the ALWAYS tag (0x7)

Signed-off-by: Rajendra Nayak <rnayak@codeaurora.org>
Reviewed-by: Akash Asthana <akashast@codeaurora.org>
Tested-by: Matthias Kaehlcke <mka@chromium.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/1594704709-26072-1-git-send-email-rnayak@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

serial: qcom_geni_serial: Always use 4 bytes per TX FIFO word

The geni serial driver had a rule that we'd only use 1 byte per FIFO
word for the TX FIFO if we were being used for the serial console.
This is ugly and a bit of a pain.  It's not too hard to fix, so fix
it.

Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Evan Green <evgreen@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20200626125844.2.Iabd56347670b9e4e916422773aba5b27943d19ee@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

serial: qcom_geni_serial: Make kgdb work even if UART isn't console

The geni serial driver had the rather sketchy hack in it where it
would adjust the number of bytes per RX FIFO word from 4 down to 1 if
it detected that CONFIG_CONSOLE_POLL was enabled (for kgdb) and this
was a console port (defined by the kernel directing output to this
port via the "console=" command line argument).

The problem with that sketchy hack is that it's possible to run kgdb
over a serial port even if it isn't used for console.

Let's avoid the hack by simply handling the 4-bytes-per-FIFO word case
for kdb.  We'll have to have a (very small) cache but that should be
fine.

A nice side effect of this patch is that an agetty (or similar)
running on this port is less likely to drop characters.  We'll
have roughly 4 times the RX FIFO depth than we used to now.

NOTE: the character cache here isn't shared between the polling API
and the non-polling API.  That means that, technically, the polling
API could eat a few extra bytes.  This doesn't seem to pose a huge
problem in reality because we'll only get several characters per FIFO
word if those characters are all received at nearly the same time and
we don't really expect non-kgdb characters to be sent to the same port
as kgdb at the exact same time we're exiting kgdb.

ALSO NOTE: we still have the sketchy hack for setting the number of
bytes per TX FIFO word in place, but that one is less bad.  kgdb
doesn't have any problem with this because it always just sends 1 byte
at a time and waits for it to finish.  The TX FIFO hack is only really
needed for console output.  In any case, a future patch will remove
that hack, too.

Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Evan Green <evgreen@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20200626125844.1.I8546ecb6c5beb054f70c5302d1a7293484212cd1@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-geni-qcom: Get rid of most overhead in prepare_message()

There's a bunch of overhead in spi-geni-qcom's prepare_message.  Get
rid of it.  Before this change spi_geni_prepare_message() took around
14.5 us.  After this change, spi_geni_prepare_message() takes about
1.75 us (as measured by ftrace).

What's here:
* We're always in FIFO mode, so no need to call it for every transfer.
  This avoids a whole ton of readl/writel calls.
* We don't need to write a whole pile of config registers if the mode
  isn't changing.  Cache the last mode and only do the work if needed.
* For several registers we were trying to do read/modify/write, but
  there was no reason.  The registers only have one thing in them, so
  just write them.

Acked-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Akash Asthana <akashast@codeaurora.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20200701174506.3.I2b3d7aeb1ea622335482cce60c58d2f8381e61dd@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-geni-qcom: Set the clock properly at runtime resume

In the patch ("spi: spi-geni-qcom: Avoid clock setting if not needed")
we avoid a whole pile of clock code.  As part of that, we should have
restored the clock at runtime resume.  Do that.

It turns out that, at least with today's configurations, this doesn't
actually matter.  That's because none of the current device trees have
an OPP table for geni SPI yet.  That makes dev_pm_opp_set_rate(dev, 0)
a no-op.  This is why it wasn't noticed in the testing of the original
patch.  It's still a good idea to fix, though.

Reviewed-by: Rajendra Nayak <rnayak@codeaurora.org>
Reviewed-by: Akash Asthana <akashast@codeaurora.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Acked-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20200709074037.v2.1.I0b701fc23eca911a5bde4ae4fa7f97543d7f960e@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-geni-qcom: Avoid clock setting if not needed

Every SPI transfer could have a different clock rate.  The
spi-geni-qcom controller code to deal with this was never very well
optimized and has always had a lot of code plus some calls into the
clk framework which, at the very least, would grab a mutex.  However,
until recently, the overhead wasn't _too_ much.  That changed with
commit 0e3b8a81f5df ("spi: spi-geni-qcom: Add interconnect support")
we're now calling geni_icc_set_bw(), which leads to a bunch of math
plus:
  geni_icc_set_bw()
    icc_set_bw()
      apply_constraints()
        qcom_icc_set()
          qcom_icc_bcm_voter_commit()
            rpmh_invalidate()
            rpmh_write_batch()
...and those rpmh commands can be a bit beefy if you call them too
often.

We already know what speed we were running at before, so if we see
that nothing has changed let's avoid the whole pile of code.

On my hardware, this made spi_geni_prepare_message() drop down from
~145 us down to ~14 us.

NOTE: Potentially it might also make sense to add some code into the
interconnect framework to avoid executing so much code when bandwidth
isn't changing, but even if we did that we still want to short circuit
here to save the extra math / clock calls.

Acked-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Akash Asthana<akashast@codeaurora.org>
Fixes: 0e3b8a81f5df ("spi: spi-geni-qcom: Add interconnect support")
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20200701174506.1.Icfdcee14649fc0a6c38e87477b28523d4e60bab3@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-qcom-qspi: Set an autosuspend delay of 250 ms

In commit cff80645d6d3 ("spi: spi-qcom-qspi: Add interconnect support")
the spi_geni_runtime_suspend() and spi_geni_runtime_resume()
became a bit slower.  Measuring on my hardware I see numbers in the
hundreds of microseconds now.

Let's use autosuspend to help avoid some of the overhead.  Now if
we're doing a bunch of transfers we won't need to be constantly
chruning.

The number 250 ms for the autosuspend delay was picked a bit
arbitrarily, so if someone has measurements showing a better value we
could easily change this.

Fixes: cff80645d6d3 ("spi: spi-qcom-qspi: Add interconnect support")
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Acked-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Rajendra Nayak <rnayak@codeaurora.org>
Tested-by: Rajendra Nayak <rnayak@codeaurora.org>
Reviewed-by: Mukesh Kumar Savaliya <msavaliy@codeaurora.org>
Reviewed-by: Akash Asthana <akashast@codeaurora.org>
Link: https://lore.kernel.org/r/20200709075113.v2.2.I3c56d655737c89bd9b766567a04b0854db1a4152@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-qcom-qspi: Avoid clock setting if not needed

As per recent changes to the spi-qcom-qspi, now when we set the clock
we'll call into the interconnect framework and also call the OPP API.
Those are expensive operations.  Let's avoid calling them if possible.
This has a big impact on getting transfer rates back up to where they
were (or maybe slightly better) before those patches landed.

Fixes: cff80645d6d3 ("spi: spi-qcom-qspi: Add interconnect support")
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Acked-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Rajendra Nayak <rnayak@codeaurora.org>
Tested-by: Rajendra Nayak <rnayak@codeaurora.org>
Reviewed-by: Mukesh Kumar Savaliya <msavaliy@codeaurora.org>
Reviewed-by: Akash Asthana <akashast@codeaurora.org>
Link: https://lore.kernel.org/r/20200709075113.v2.1.Ia7cb4f41ce93d37d0a764b47c8a453ce9e9c70ef@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-qcom-qspi: Use OPP API to set clk/perf state

QSPI needs to vote on a performance state of a power domain depending on
the clock rate. Add support for it by specifying the perf state/clock rate
as an OPP table in device tree.

Signed-off-by: Rajendra Nayak <rnayak@codeaurora.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Acked-by: Mark Brown <broonie@kernel.org>
Cc: Alok Chauhan <alokc@codeaurora.org>
Cc: Akash Asthana <akashast@codeaurora.org>
Cc: linux-spi@vger.kernel.org
Link: https://lore.kernel.org/r/1593769293-6354-2-git-send-email-rnayak@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

firmware: qcom_scm: Add msm8994 compatible

This change adds a compatible for msm8994,
which requires no additional clocks for
scm to probe correctly.

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Konrad Dybcio <konradybcio@gmail.com>
Link: https://lore.kernel.org/r/20200624150107.76234-2-konradybcio@gmail.com
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

firmware: qcom_scm: Fix legacy convention SCM accessors

The move to a combined driver for the QCOM SCM hardware changed the
io_writel and io_readl helpers to use non-atomic calls, despite the
commit message saying that atomic was a better option. This breaks these
helpers on hardware that uses the old legacy convention (access fails
with a -95 return code). Switch back to using the atomic calls.

Observed as a failure routing GPIO interrupts to the Apps processor on
an IPQ8064; fix is confirmed as correctly allowing the interrupts to be
routed and observed.

Reviewed-by: Elliot Berman <eberman@codeaurora.org>
Fixes: 57d3b816718c ("firmware: qcom_scm: Remove thin wrappers")
Cc: stable@vger.kernel.org
Signed-off-by: Jonathan McDowell <noodles@earth.li>
Link: https://lore.kernel.org/r/20200704172334.GA759@earth.li
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

Merge tag 'optee-bus-for-v5.9' of git://git.linaro.org/people/jens.wiklander/linux-tee into arm/drivers

Enable multi-stage OP-TEE bus enumeration

Probes drivers on the OP-TEE bus in two steps. First for drivers which
do not depend on tee-supplicant. After tee-supplicant has been started
probe the devices which do depend on tee-supplicant.

Also introduces driver which uses an OP-TEE based fTPM Trusted
Application depends on tee-supplicant NV RAM implementation based on
RPMB secure storage.

* tag 'optee-bus-for-v5.9' of git://git.linaro.org/people/jens.wiklander/linux-tee:
  tpm_ftpm_tee: register driver on TEE bus
  optee: enable support for multi-stage bus enumeration
  optee: use uuid for sysfs driver entry

Link: https://lore.kernel.org/r/20200710085230.GA1312913@jade
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

soc: imx: select ARM_GIC_V3 for i.MX8M

Select ARM_GIC_V3, then it is able to use gic v3 driver in aarch32
mode linux on aarch64 hardware. For aarch64 mode, it not hurts
to select ARM_GIC_V3.

Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>

firmware: arm_scmi: Remove fixed size fields from reports/scmi_event_header

Event reports are used to convey information describing events to the
registered user-callbacks: they are necessarily derived from the underlying
raw SCMI events' messages but they are not meant to expose or directly
mirror any of those messages data layout, which belong to the protocol
layer.

Using fixed size types for report fields, mirroring messages structure,
is at odd with this: get rid of them using more generic, equivalent,
typing.

Substitute scmi_event_header fixed size fields with generic types too and
shuffle around fields definitions to minimize implicit padding while
adapting involved functions.

Link: https://lore.kernel.org/r/20200710133919.39792-3-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Remove unneeded __packed attribute

Remove __packed attribute from struct scmi_event_header.

Link: https://lore.kernel.org/r/20200710133919.39792-2-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Remove zero-length array in SCMI notifications

Substitute zero-length array defined in scmi_base_error_report with
a flexible length array definition.

Link: https://lore.kernel.org/r/20200710133919.39792-1-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Provide a missing function param description

gcc as well as clang now produce warnings for missing kerneldoc function
parameter.

Fix the following W=1 kernel build warning:

drivers/firmware/arm_scmi/smc.c:32:
 warning: Function parameter or member 'shmem_lock' not described in 'scmi_smc'

Link: https://lore.kernel.org/r/20200709153155.22573-1-sudeep.holla@arm.com
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

clk: scmi: Fix min and max rate when registering clocks with discrete rates

Currently we are not initializing the scmi clock with discrete rates
correctly. We fetch the min_rate and max_rate value only for clocks with
ranges and ignore the ones with discrete rates. This will lead to wrong
initialization of rate range when clock supports discrete rate.

Fix this by using the first and the last rate in the sorted list of the
discrete clock rates while registering the clock.

Link: https://lore.kernel.org/r/20200709081705.46084-2-sudeep.holla@arm.com
Fixes: 6d6a1d82eaef7 ("clk: add support for clocks provided by SCMI")
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Reported-and-tested-by: Dien Pham <dien.pham.ry@renesas.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

soc: renesas: rcar-rst: Add support for RZ/G2H

Add support for RZ/G2H (R8A774E1) to the R-Car RST driver.

Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Link: https://lore.kernel.org/r/1594138692-16816-8-git-send-email-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

soc: renesas: Identify RZ/G2H

This patch adds support for identifying the RZ/G2H (r8a774e1) SoC.

Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Link: https://lore.kernel.org/r/1594138692-16816-2-git-send-email-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

soc: renesas: Add Renesas R8A774E1 config option

Add configuration option for the RZ/G2H (R8A774E1) SoC.

Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Link: https://lore.kernel.org/r/1594138692-16816-3-git-send-email-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

soc: renesas: rcar-sysc: Add r8a774e1 support

Add support for RZ/G2H (R8A774E1) SoC power areas to the R-Car SYSC
driver.

Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Link: https://lore.kernel.org/r/1594138692-16816-6-git-send-email-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

Merge tag 'renesas-r8a774e1-dt-binding-defs-tag' into renesas-drivers-for-v5.9

Renesas RZ/G2H DT Binding Definitions

Clock and Power Domain definitions for the Renesas RZ/G2H (R8A774E1)
SoC, shared by driver and DT source files.

clk: renesas: Add r8a774e1 CPG Core Clock Definitions

Add all RZ/G2H Clock Pulse Generator Core Clock Outputs, as listed in
Table 11.2 ("List of Clocks [RZ/G2H]") of the RZ/G2H Hardware User's
Manual.

Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Link: https://lore.kernel.org/r/1594138692-16816-10-git-send-email-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

dt-bindings: power: Add r8a774e1 SYSC power domain definitions

This patch adds power domain indices for the RZ/G2H (r8a774e1) SoC.

Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Link: https://lore.kernel.org/r/1594138692-16816-5-git-send-email-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

firmware: imx: Move i.MX SCU soc driver into imx firmware folder

The i.MX SCU soc driver depends on SCU firmware driver, so it has to
use platform driver model for proper defer probe operation, since
it has no device binding in DT file, a simple platform device is
created together inside the platform driver. To make it more clean,
we can just move the entire SCU soc driver into imx firmware folder
and initialized by i.MX SCU firmware driver.

Signed-off-by: Anson Huang <Anson.Huang@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>

<linux/of.h>: add stub for of_get_next_parent() to fix qcom build error

Fix a (COMPILE_TEST) build error when CONFIG_OF is not set/enabled
by adding a stub for of_get_next_parent().

../drivers/soc/qcom/qcom-geni-se.c:819:11: error: implicit declaration of function 'of_get_next_parent'; did you mean 'of_get_parent'? [-Werror=implicit-function-declaration]
../drivers/soc/qcom/qcom-geni-se.c:819:9: warning: assignment makes pointer from integer without a cast [-Wint-conversion]

Fixes: 048eb908a1f2 ("soc: qcom-geni-se: Add interconnect support to fix earlycon crash")
Acked-by: Rob Herring <robh@kernel.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Frank Rowand <frowand.list@gmail.com>
Cc: devicetree@vger.kernel.org
Cc: Andy Gross <agross@kernel.org>
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: linux-arm-msm@vger.kernel.org
Link: https://lore.kernel.org/r/ce0d7561-ff93-d267-b57a-6505014c728c@infradead.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

firmware: arm_scmi: Keep the discrete clock rates sorted

Instead of relying on the firmware to keep the clock rates sorted, let
us sort the list. This is not essential for clock layer but it helps
to find the min and max rates easily from the list.

Link: https://lore.kernel.org/r/20200709081705.46084-1-sudeep.holla@arm.com
Fixes: 5f6c6430e904 ("firmware: arm_scmi: add initial support for clock protocol")
Reported-and-tested-by: Dien Pham <dien.pham.ry@renesas.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

tpm_ftpm_tee: register driver on TEE bus

OP-TEE based fTPM Trusted Application depends on tee-supplicant to
provide NV RAM implementation based on RPMB secure storage. So this
dependency can be resolved via TEE bus where we only invoke fTPM
driver probe once fTPM device is registered on the bus which is only
true after the tee-supplicant is up and running. Additionally, TEE bus
provides auto device enumeration.

Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Suggested-by: Sumit Garg <sumit.garg@linaro.org>
Suggested-by: Arnd Bergmann <arnd@linaro.org>
Reviewed-by: Sumit Garg <sumit.garg@linaro.org>
Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Tested-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

optee: enable support for multi-stage bus enumeration

Some drivers (like ftpm) can operate only after tee-supplicant
runs because of tee-supplicant provides things like storage
services (rpmb, shm).  This patch splits probe of non tee-supplicant
dependable drivers to the early stage, and after tee-supplicant run
probe other drivers.

Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Suggested-by: Sumit Garg <sumit.garg@linaro.org>
Suggested-by: Arnd Bergmann <arnd@linaro.org>
Reviewed-by: Sumit Garg <sumit.garg@linaro.org>
Acked-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Tested-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

optee: use uuid for sysfs driver entry

With the evolving use-cases for TEE bus, now it's required to support
multi-stage enumeration process. But using a simple index doesn't
suffice this requirement and instead leads to duplicate sysfs entries.
So instead switch to use more informative device UUID for sysfs entry
like:
/sys/bus/tee/devices/optee-ta-<uuid>

Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Reviewed-by: Sumit Garg <sumit.garg@linaro.org>
Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Tested-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

Merge tag 'soc-attr-updates-5.9' of git://git./linux/kernel/git/sudeep.holla/linux into arm/drivers

SoC attributes update for v5.9

1. Addition of ARM SMCCC ARCH_SOC_ID support
2. Usage of the custom soc attribute groups already supported in the
   infrastucture instead of device_create_file which eliminates the need
   for any cleanup when soc is unregistered
3. Minor clean up switching to use standard DEVICE_ATTR_RO() instead of
   direct __ATTR

* tag 'soc-attr-updates-5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux:
  firmware: smccc: Add ARCH_SOC_ID support
  ARM: OMAP2: Use custom soc attribute group instead of device_create_file
  ARM: OMAP2: Switch to use DEVICE_ATTR_RO()
  soc: ux500: Use custom soc attribute group instead of device_create_file
  soc: ux500: Switch to use DEVICE_ATTR_RO()
  soc: integrator: Use custom soc attribute group instead of device_create_file
  soc: integrator: Switch to use DEVICE_ATTR_RO()
  soc: realview: Use custom soc attribute group instead of device_create_file
  soc: realview: Switch to use DEVICE_ATTR_RO()

Link: https://lore.kernel.org/r/20200706165312.40697-1-sudeep.holla@arm.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

firmware: smccc: Add ARCH_SOC_ID support

SMCCC v1.2 adds a new optional function SMCCC_ARCH_SOC_ID to obtain a
SiP defined SoC identification value. Add support for the same.

Also using the SoC bus infrastructure, let us expose the platform
specific SoC atrributes under sysfs.

There are various ways in which it can be represented in shortened form
for efficiency and ease of parsing for userspace. The chosen form is
described in the ABI document.

Link: https://lore.kernel.org/r/20200625095939.50861-1-sudeep.holla@arm.com
Cc: Etienne Carriere <etienne.carriere@st.com>
Reviewed-by: Steven Price <steven.price@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

ARM: OMAP2: Use custom soc attribute group instead of device_create_file

Commit c31e73121f4c ("base: soc: Handle custom soc information sysfs
entries") introduced custom soc attribute group in soc_device_attribute
structure but there are no users treewide. While trying to understand
the motivation and tried to use it, it was found lot of existing custom
attributes can moved to use it instead of device_create_file.

Though most of these never remove/cleanup the custom attribute as they
never call soc_device_unregister, using these custom attribute group
eliminate the need for any cleanup as the driver infrastructure will
take care of that.

Let us remove device_create_file and start using the custom attribute
group in soc_device_attribute.

Link: https://lore.kernel.org/r/20200523170859.50003-9-sudeep.holla@arm.com
Cc: Tony Lindgren <tony@atomide.com>
Cc: linux-omap@vger.kernel.org
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

ARM: OMAP2: Switch to use DEVICE_ATTR_RO()

Move device attributes to DEVICE_ATTR_RO() as that would make things
a lot more "obvious" what is happening over the existing __ATTR usage.

Link: https://lore.kernel.org/r/20200523170859.50003-8-sudeep.holla@arm.com
Cc: Tony Lindgren <tony@atomide.com>
Cc: linux-omap@vger.kernel.org
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

soc: ux500: Use custom soc attribute group instead of device_create_file

Commit c31e73121f4c ("base: soc: Handle custom soc information sysfs
entries") introduced custom soc attribute group in soc_device_attribute
structure but there are no users treewide. While trying to understand
the motivation and tried to use it, it was found lot of existing custom
attributes can moved to use it instead of device_create_file.

Though most of these never remove/cleanup the custom attribute as they
never call soc_device_unregister, using these custom attribute group
eliminate the need for any cleanup as the driver infrastructure will
take care of that.

Let us remove device_create_file and start using the custom attribute
group in soc_device_attribute.

Link: https://lore.kernel.org/r/20200523170859.50003-7-sudeep.holla@arm.com
Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

soc: ux500: Switch to use DEVICE_ATTR_RO()

Move device attributes to DEVICE_ATTR_RO() as that would make things
a lot more "obvious" what is happening over the existing __ATTR usage.

Link: https://lore.kernel.org/r/20200523170859.50003-6-sudeep.holla@arm.com
Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

soc: integrator: Use custom soc attribute group instead of device_create_file

Commit c31e73121f4c ("base: soc: Handle custom soc information sysfs
entries") introduced custom soc attribute group in soc_device_attribute
structure but there are no users treewide. While trying to understand
the motivation and tried to use it, it was found lot of existing custom
attributes can moved to use it instead of device_create_file.

Though most of these never remove/cleanup the custom attribute as they
never call soc_device_unregister, using these custom attribute group
eliminate the need for any cleanup as the driver infrastructure will
take care of that.

Let us remove device_create_file and start using the custom attribute
group in soc_device_attribute.

Link: https://lore.kernel.org/r/20200523170859.50003-5-sudeep.holla@arm.com
Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

soc: integrator: Switch to use DEVICE_ATTR_RO()

Move device attributes to DEVICE_ATTR_RO() as that would make things
a lot more "obvious" what is happening over the existing __ATTR usage.

Link: https://lore.kernel.org/r/20200523170859.50003-4-sudeep.holla@arm.com
Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

soc: realview: Use custom soc attribute group instead of device_create_file

Commit c31e73121f4c ("base: soc: Handle custom soc information sysfs
entries") introduced custom soc attribute group in soc_device_attribute
structure but there are no users treewide. While trying to understand
the motivation and tried to use it, it was found lot of existing custom
attributes can moved to use it instead of device_create_file.

Though most of these never remove/cleanup the custom attribute as they
never call soc_device_unregister, using these custom attribute group
eliminate the need for any cleanup as the driver infrastructure will
take care of that.

Let us remove device_create_file and start using the custom attribute
group in soc_device_attribute.

Link: https://lore.kernel.org/r/20200523170859.50003-3-sudeep.holla@arm.com
Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

soc: realview: Switch to use DEVICE_ATTR_RO()

Move device attributes to DEVICE_ATTR_RO() as that would make things
a lot more "obvious" what is happening over the existing __ATTR usage.

Link: https://lore.kernel.org/r/20200523170859.50003-2-sudeep.holla@arm.com
Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

dt-bindings: firmware: qcom: Add compatible for IPQ8074 SoC

Add compatible for IPQ8074 support.
This does not need clocks for scm calls.

Signed-off-by: Gokul Sriram Palanisamy <gokulsri@codeaurora.org>
Signed-off-by: Sricharan R <sricharan@codeaurora.org>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/1589362265-22702-9-git-send-email-gokulsri@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

firmware: arm_scmi: Add base notifications support

Make SCMI base protocol register with the notification core.

Link: https://lore.kernel.org/r/20200701155348.52864-10-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Add reset notifications support

Make SCMI reset protocol register with the notification core.

Link: https://lore.kernel.org/r/20200701155348.52864-9-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Add sensor notifications support

Make SCMI sensor protocol register with the notification core.

Link: https://lore.kernel.org/r/20200701155348.52864-8-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Add perf notifications support

Make SCMI perf protocol register with the notification core.

Link: https://lore.kernel.org/r/20200701155348.52864-7-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Add power notifications support

Make SCMI power protocol register with the notification core.

Link: https://lore.kernel.org/r/20200701155348.52864-6-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Enable notification core

Initialize and enable SCMI notifications core support during bus/driver
probe phase, so that protocols can start registering their supported
events during their initialization.

Link: https://lore.kernel.org/r/20200701155348.52864-5-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Add notification dispatch and delivery

Add the core SCMI notifications dispatch and delivery support logic
which is able to dispatch well-known received events from the Rx
interrupt handler to the dedicated deferred worker. From there, it will
deliver the events to the registered users' callbacks.

Dispatch and delivery support is just added here, still not enabled.

Link: https://lore.kernel.org/r/20200701155348.52864-4-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Add notification callbacks-registration

Add the core SCMI notifications callbacks-registration support: allow
users to register their own callbacks against the desired events.

Whenever a registration request is issued against a still non existent
event, mark such request as pending for later processing, in order to
account for possible late initializations of SCMI Protocols associated
to loadable drivers.

Link: https://lore.kernel.org/r/20200701155348.52864-3-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Add notification protocol-registration

Add the core SCMI notifications protocol-registration support: allow
protocols to register their own set of supported events, during their
initialization phase. Notification core can track multiple platform
instances by their handles.

Link: https://lore.kernel.org/r/20200701155348.52864-2-cristian.marussi@arm.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Fix SCMI genpd domain probing

When, at probe time, an SCMI communication failure inhibits the capacity
to query power domains states, such domains should be skipped.

Registering partially initialized SCMI power domains with genpd will
causes kernel panic.

 arm-scmi timed out in resp(caller: scmi_power_state_get+0xa4/0xd0)
 scmi-power-domain scmi_dev.2: failed to get state for domain 9
 Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000
 Mem abort info:
   ESR = 0x96000006
   EC = 0x25: DABT (current EL), IL = 32 bits
   SET = 0, FnV = 0
   EA = 0, S1PTW = 0
 Data abort info:
   ISV = 0, ISS = 0x00000006
   CM = 0, WnR = 0
 user pgtable: 4k pages, 48-bit VAs, pgdp=00000009f3691000
 [0000000000000000] pgd=00000009f1ca0003, p4d=00000009f1ca0003, pud=00000009f35ea003, pmd=0000000000000000
 Internal error: Oops: 96000006 [#1] PREEMPT SMP
 CPU: 2 PID: 381 Comm: bash Not tainted 5.8.0-rc1-00011-gebd118c2cca8 #2
 Hardware name: ARM LTD ARM Juno Development Platform/ARM Juno Development Platform, BIOS EDK II Jan  3 2020
 Internal error: Oops: 96000006 [#1] PREEMPT SMP
 pstate: 80000005 (Nzcv daif -PAN -UAO BTYPE=--)
 pc : of_genpd_add_provider_onecell+0x98/0x1f8
 lr : of_genpd_add_provider_onecell+0x48/0x1f8
 Call trace:
  of_genpd_add_provider_onecell+0x98/0x1f8
  scmi_pm_domain_probe+0x174/0x1e8
  scmi_dev_probe+0x90/0xe0
  really_probe+0xe4/0x448
  driver_probe_device+0xfc/0x168
  device_driver_attach+0x7c/0x88
  bind_store+0xe8/0x128
  drv_attr_store+0x2c/0x40
  sysfs_kf_write+0x4c/0x60
  kernfs_fop_write+0x114/0x230
  __vfs_write+0x24/0x50
  vfs_write+0xbc/0x1e0
  ksys_write+0x70/0xf8
  __arm64_sys_write+0x24/0x30
  el0_svc_common.constprop.3+0x94/0x160
  do_el0_svc+0x2c/0x98
  el0_sync_handler+0x148/0x1a8
  el0_sync+0x158/0x180

Do not register any power domain that failed to be queried with genpd.

Fixes: 898216c97ed2 ("firmware: arm_scmi: add device power domain support using genpd")
Link: https://lore.kernel.org/r/20200619220330.12217-1-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Use HAVE_ARM_SMCCC_DISCOVERY instead of ARM_PSCI_FW

Commit e5bfb21d98b6 ("firmware: smccc: Add HAVE_ARM_SMCCC_DISCOVERY to
identify SMCCC v1.1 and above") introduced new config option to identify
the availability of SMCCC discoverability of version and features
transparently hiding the indirect dependency on ARM_PSCI_FW.

Commit 5a897e3ab429 ("firmware: arm_scmi: fix psci dependency") just
worked around the build dependency making SCMI SMC/HVC transport depend
on ARM_PSCI_FW at the time. Since it really just relies on SMCCC directly
and not on ARM_PSCI_FW, let us move to use CONFIG_HAVE_ARM_SMCCC_DISCOVERY
instead of CONFIG_ARM_PSCI_FW.

Link: https://lore.kernel.org/r/20200625101937.51939-1-sudeep.holla@arm.com
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

cpufreq: arm_scmi: Set fast_switch_possible conditionally

Currently the fast_switch_possible flag is set unconditionally to true.
Based on this, schedutil does not create a thread for frequency
switching and would always use the fast switch path.

However, if the platform does not support SCMI fast channel, we use
polling mode for SCMI message transfer. This may be possible only if
there is dedicated channel for DVFS and all operations are in polling
mode.

Update this by retrieving the fast_switch capability based on the
presence of fast channels in SCMI platform firmware.

Link: https://lore.kernel.org/r/20200617094332.8391-2-nicola.mazzucato@arm.com
Suggested-by: Lukasz Luba <lukasz.luba@arm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Nicola Mazzucato <nicola.mazzucato@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Add fast_switch_possible() interface

Add a new fast_switch_possible interface to the existing perf_ops to
export the information of whether or not fast_switch is possible for a
given device.

This can be used by the cpufreq driver and framework to choose proper
mechanism for frequency change.

Link: https://lore.kernel.org/r/20200617094332.8391-1-nicola.mazzucato@arm.com
Suggested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Nicola Mazzucato <nicola.mazzucato@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

firmware: arm_scmi: Use signed integer to report transfer status

Currently the trace event 'scmi_xfer_end' reports the status of the
transfer using the unsigned status field read from the firmware which
may not be easy to interpret. It may also miss to emit any timeouts
that happen in the driver resulting in emitting garbage in the status
field in those scenarios.

Let us use signed integer so that error values are emitted out after
they are mapped from firmware error formats to standard linux error
codes. While at this, also include any timeouts in the driver itself.

Link: https://lore.kernel.org/r/20200609134503.55860-1-sudeep.holla@arm.com
Cc: Jim Quinlan <james.quinlan@broadcom.com>
Cc: Lukasz Luba <lukasz.luba@arm.com>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>

spi: spi-geni-qcom: Use OPP API to set clk/perf state

geni spi needs to express a perforamnce state requirement on CX
depending on the frequency of the clock rates. Use OPP table from
DT to register with OPP framework and use dev_pm_opp_set_rate() to
set the clk/perf state.

Signed-off-by: Rajendra Nayak <rnayak@codeaurora.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Acked-by: Mark Brown <broonie@kernel.org>
Cc: Alok Chauhan <alokc@codeaurora.org>
Cc: Akash Asthana <akashast@codeaurora.org>
Cc: linux-spi@vger.kernel.org
Link: https://lore.kernel.org/r/1592222564-13556-3-git-send-email-rnayak@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

tty: serial: qcom_geni_serial: Use OPP API to set clk/perf state

geni serial needs to express a perforamnce state requirement on CX
powerdomain depending on the frequency of the clock rates.
Use OPP table from DT to register with OPP framework and use
dev_pm_opp_set_rate() to set the clk/perf state.

Signed-off-by: Rajendra Nayak <rnayak@codeaurora.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Akash Asthana <akashast@codeaurora.org>
Cc: linux-serial@vger.kernel.org
Link: https://lore.kernel.org/r/1592222564-13556-2-git-send-email-rnayak@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-qcom-qspi: Add interconnect support

Get the interconnect paths for QSPI device and vote according to the
current bus speed of the driver.

Acked-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Akash Asthana <akashast@codeaurora.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Link: https://lore.kernel.org/r/1592908737-7068-8-git-send-email-akashast@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-geni-qcom: Add interconnect support

Get the interconnect paths for SPI based Serial Engine device
and vote according to the current bus speed of the driver.

Acked-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Akash Asthana <akashast@codeaurora.org>
Link: https://lore.kernel.org/r/1592908737-7068-7-git-send-email-akashast@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

spi: spi-geni-qcom: Combine the clock setting code

There is code for adjusting the clock both in setup_fifo_params()
(called from prepare_message()) and in setup_fifo_xfer() (called from
transfer_one()).  The code is the same.  Abstract it out to a shared
function.

This is a no-op cleanup patch.  The only change is to the error string
if we fail to set the clock.  Since the two paths has marginally
different error messages I picked the clean one.

Acked-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Akash Asthana <akashast@codeaurora.org>
Link: https://lore.kernel.org/r/1592908737-7068-6-git-send-email-akashast@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

tty: serial: qcom_geni_serial: Add interconnect support

Get the interconnect paths for Uart based Serial Engine device
and vote according to the baud rate requirement of the driver.

Signed-off-by: Akash Asthana <akashast@codeaurora.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/1592908737-7068-5-git-send-email-akashast@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

i2c: i2c-qcom-geni: Add interconnect support

Get the interconnect paths for I2C based Serial Engine device
and vote according to the bus speed of the driver.

Signed-off-by: Akash Asthana <akashast@codeaurora.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Acked-by: Wolfram Sang <wsa@kernel.org>
Link: https://lore.kernel.org/r/1592908737-7068-4-git-send-email-akashast@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

soc: qcom-geni-se: Add interconnect support to fix earlycon crash

QUP core clock is shared among all the SE drivers present on particular
QUP wrapper, the system will reset(unclocked access) if earlycon used after
QUP core clock is put to 0 from other SE drivers before real console comes
up.

As earlycon can't vote for it's QUP core need, to fix this add ICC
support to common/QUP wrapper driver and put vote for QUP core from
probe on behalf of earlycon and remove vote during earlycon exit call.

Signed-off-by: Akash Asthana <akashast@codeaurora.org>
Reported-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Link: https://lore.kernel.org/r/1592908737-7068-3-git-send-email-akashast@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

soc: qcom: geni: Support for ICC voting

Add necessary macros and structure variables to support ICC BW
voting from individual SE drivers.

Signed-off-by: Akash Asthana <akashast@codeaurora.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Link: https://lore.kernel.org/r/1592908737-7068-2-git-send-email-akashast@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

soc: qcom: rpmh-rsc: Don't use ktime for timeout in write_tcs_reg_sync()

The write_tcs_reg_sync() may be called after timekeeping is suspended
so it's not OK to use ktime.  The readl_poll_timeout_atomic() macro
implicitly uses ktime.  This was causing a warning at suspend time.

Change to just loop 1000000 times with a delay of 1 us between loops.
This may give a timeout of more than 1 second but never less and is
safe even if timekeeping is suspended.

NOTE: I don't have any actual evidence that we need to loop here.
It's possibly that all we really need to do is just read the value
back to ensure that the pipes are cleaned and the looping/comparing is
totally not needed.  I never saw the loop being needed in my tests.
However, the loop shouldn't hurt.

Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Reviewed-by: Maulik Shah <mkshah@codeaurora.org>
Fixes: 91160150aba0 ("soc: qcom: rpmh-rsc: Timeout after 1 second in write_tcs_reg_sync()")
Reported-by: Maulik Shah <mkshah@codeaurora.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20200528074530.1.Ib86e5b406fe7d16575ae1bb276d650faa144b63c@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

soc: qcom: smd-rpm: Add msm8994 compatible

Add the compatible for the RPM in msm8994.

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Konrad Dybcio <konradybcio@gmail.com>
Link: https://lore.kernel.org/r/20200602200407.320908-1-konradybcio@gmail.com
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

soc: qcom: socinfo: add SM8250 entry to soc_id array

Add an entry for SM8250 SoC.

Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200525164817.2938638-4-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

soc: qcom: socinfo: add file with SoC info format version

To ease debugging socinfo driver for newer chips add debugfs file
returning SoC info format version.

Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200525164817.2938638-3-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

soc: qcom: socinfo: fix printing of pmic_model

Print sensible string instead of just "(null)" for unknown PMIC models.
Also as we are at it, do not let debugfs handler access past pmic_models
array.

Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200525164817.2938638-2-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

soc: qcom: socinfo: add support for newer socinfo data

Add support for newer Qualcomm SoC info structures (up to version 0.15).

Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200525164817.2938638-1-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>

firmware: imx: scu-pd: add more cm4 resources

Add more cm4 resources, then linux could use cm4's i2c/lpuart and
could kick cm4 core.

Reviewed-by: Dong Aisheng <aisheng.dong@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>

firmware: imx: add resource management api

Add resource management API, when we have multiple
partition running together, resources not owned to current
partition should not be used.

Reviewed-by: Leonard Crestez <leonard.crestez@nxp.com>
Reviewed-by: Dong Aisheng <aisheng.dong@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>

firmware: imx: scu-pd: fix cm40 power domain

The postfix needs to be false. Alought compiler use 0 for postfix now,
and take start_from as 0, it is better we add explicit false to postfix.

Fixes: 705dcca91d0a("firmware: imx: scu-pd: add power domain for I2C and INTMUX in CM40 SS")
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Dong Aisheng <aisheng.dong@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>