According to "KeyStone Architecture Inter-IC Control Bus User Guide", fixed
additive part of frequency divisors (referred as "d" in the code and datasheet)
always equals to 6, independent of module clock prescaler.

                         module clock frequency
master clock frequency = ----------------------
                         (ICCL + 6) + (ICCH + 6)

It was not the case with original Davinci IP. Introduce new compatible property
"ti,keystone-i2c", which triggers special handling in the driver.

Without this change Keystone-based systems (having 204.8MHz input clock) choose
prescaler 29 (PSC=28). Using d=5 in this case leads to bus bitrate ~353kHz
instead of requested 400kHz. After correction, assuming d=6 bus rate is ~392kHz.
This gives ~11% transfer rate increase.
Signed-off-by: NAlexander Sverdlin <alexander.sverdlin@nokia.com>
Reviewed-by: NGrygorii Strashko <grygorii.strashko@ti.com>
Tested-by: NHemanth Guruva Reddy <hemanth.guruva_reddy@nokia.com>
Tested-by: NLukasz Gemborowski <lukasz.gemborowski@nokia.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Issue the warning in all error paths when unable to register MSI or its
handler.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Propagate actual return code when requesting interrupt fails.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

struct pci_dev already has a flag to track if MSI is enabled or not. Use it
directly.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

There is no need to repeat the work that is already done in the PCI driver
core. Remove suspend and resume callbacks.

Note that there is no more calls performed to enable or disable a PCI
device during suspend-resume cycle. Nowadays they seems to be
superfluous. Someone can read more in [1].

[1] https://www.kernel.org/doc/ols/2009/ols2009-pages-319-330.pdfSigned-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

pcim_release() will release any requested region. There is no need to duplicate
this effort in the driver.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

This fixes the code to suppress sparse warnings like:

drivers/i2c/busses/i2c-ismt.c:725:36: warning: incorrect type in argument 2 (different address spaces)
drivers/i2c/busses/i2c-ismt.c:725:36:    expected void volatile [noderef] <asn:2>*addr
drivers/i2c/busses/i2c-ismt.c:725:36:    got void *
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

The way we currently scan I2C devices behind an I2C host controller does not
work in cases where the I2C device in question is not declared directly below
the host controller ACPI node.

This is perfectly legal according the ACPI 6.0 specification and some existing
systems are doing this.

To be able to enumerate all devices which are connected to a certain I2C host
controller we need to rework the current I2C scanning routine a bit. Instead of
scanning directly below the host controller we scan the whole ACPI namespace
for present devices with valid I2cSerialBus() connection pointing to the host
controller in question.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Tested-by: NDustin Byford <dustin@cumulusnetworks.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

There is some code duplication in i2c-designware-platdrv and
i2c-designware-pcidrv probe functions. What is even worse that duplication
requires i2c_dw_xfer(), i2c_dw_func() and i2c_dw_isr() i2c-designware-core
functions to be exported.

Therefore move common code into new i2c_dw_probe() and make functions above
local to i2c-designware-core.

While merging the code patch does following functional changes:

- I2C Adapter name will be "Synopsys DesignWare I2C adapter". Previously it
  was used for platform and ACPI devices but PCI device used
  "i2c-designware-pci".
- Using device name for interrupt name. Previous it was platform device name,
  ACPI device name or "i2c-designware-pci".
- Error code from devm_request_irq() and i2c_add_numbered_adapter() will be
  printed in case of error.
Signed-off-by: NJarkko Nikula <jarkko.nikula@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Make it easier to distinguish between i2c-designware-platdrv and
i2c-designware-core functions and to be consistent with
i2c-designware-pcidrv.
Signed-off-by: NJarkko Nikula <jarkko.nikula@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

dw_readl() and dw_writel() are not used outside of i2c-designware-core and
they are not exported so make them static and remove their forward
declaration.
Signed-off-by: NJarkko Nikula <jarkko.nikula@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

i2c_dw_is_enabled() became unused by the commit be58eda7
("i2c: designware-pci: Cleanup driver power management") and
i2c_dw_enable() by the commit 3a48d1c0 ("i2c: prevent spurious
interrupt on Designware controllers").
Signed-off-by: NJarkko Nikula <jarkko.nikula@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Device must not generate interrupts before registering the interrupt
handler so move i2c_dw_disable_int() before requesting it.

There are no known issues with this. The code has been here since commit
fe20ff5c ("i2c-designware: Add support for Designware core behind PCI
devices.").
Signed-off-by: NJarkko Nikula <jarkko.nikula@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

There is no need to clear interrupts in i2c_dw_pci_probe() since only place
where interrupts are unmasked is i2c_dw_xfer_init() and there interrupts
are always cleared after commit 2a2d95e9 ("i2c: designware: always
clear interrupts before enabling them").

This allows to cleanup the code and replace i2c_dw_clear_int() in
i2c_dw_xfer_init() by direct register read as there are no other callers.
Signed-off-by: NJarkko Nikula <jarkko.nikula@linux.intel.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

The requested bit rate can be outside the range supported by the driver.
The maximum bit rate this driver supports at the moment is 400Khz.

If the requested bit rate is larger than the maximum supported by the
driver, set the bitrate to the maximum supported before bitrate_khz is
calculated.

Maximum speed supported by the driver can be increased to 1Mhz by
adding support for "fast plus mode" in the future.

Fixes: commit 27bce457 ("i2c: img-scb: Add Imagination Technologies I2C SCB driver")
Signed-off-by: NSifan Naeem <sifan.naeem@imgtec.com>
Acked-by: NJames Hogan <james.hogan@imgtec.com>
Reviewed-by: NJames Hartley <james.hartley@imgtec.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Clear line status and all generated interrupts from the interrupt
status register before starting a transfer, as we may have
unserviced interrupts from previous transfers that might be
handled in the context of the new transfer.

Fixes: commit 27bce457 ("i2c: img-scb: Add Imagination Technologies I2C SCB driver")
Signed-off-by: NSifan Naeem <sifan.naeem@imgtec.com>
Acked-by: NJames Hogan <james.hogan@imgtec.com>
Reviewed-by: NJames Hartley <james.hartley@imgtec.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

i2c->line_status accumulates the line status bits that have been seen
with each interrupt. As we're only interested in that bit from the
current interrupt, refer to line_status (the argument to img_i2c_auto)
instead of i2c->line_status.
Signed-off-by: NSifan Naeem <sifan.naeem@imgtec.com>
Reviewed-by: NJames Hartley <james.hartley@imgtec.com>
Acked-by: NJames Hogan <james.hogan@imgtec.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Currently, after determining the minimum value for the High period
(TCKH) the remainder of the internal clock pulses is set as the Low
period (TCKL). This causes the i2c clock duty cycle to be much less
than 50%.

Modify the starting position to TCKH and TCKL at 50% of the internal
clock, and adjusts the TCKH and TCKL values from there should the
minimum value for TCKL not be met. This results in duty cycles closer
to 50%.

Fixes: commit 27bce457 ("i2c: img-scb: Add Imagination Technologies I2C SCB driver")
Signed-off-by: NSifan Naeem <sifan.naeem@imgtec.com>
Acked-by: NJames Hogan <james.hogan@imgtec.com>
Reviewed-by: NJames Hartley <james.hartley@imgtec.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Using % can be slow depending on the architecture.

Using DIV_ROUND_UP is nicer and more efficient way to do it.

Fixes: commit 27bce457 ("i2c: img-scb: Add Imagination Technologies I2C SCB driver")
Signed-off-by: NSifan Naeem <sifan.naeem@imgtec.com>
Acked-by: NJames Hogan <james.hogan@imgtec.com>
Reviewed-by: NJames Hartley <james.hartley@imgtec.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Move scb_wr_rd_fence to before reading from fifo and writing to
fifo to make sure the the first read/write is done after the required
number of cycles.

Fixes: commit 27bce457 ("i2c: img-scb: Add Imagination Technologies I2C SCB driver")
Signed-off-by: NSifan Naeem <sifan.naeem@imgtec.com>
Acked-by: NJames Hogan <james.hogan@imgtec.com>
Reviewed-by: NJames Hartley <james.hartley@imgtec.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

The code to read from the master read fifo, and write to the master
write fifo, checks a bit in an SCB register before every byte to
ensure that the fifo is not full (write fifo) or empty (read fifo).
Due to clock domain crossing inside the SCB block the updated value
of this bit is only visible after 2 cycles.

The scb_wr_rd_fence() function does 2 dummy writes (to the read-only
revision register), and it's called before reading from or writing to the
fifos to ensure that subsequent reads of the fifo status bits do not read
stale values.

As the 2 dummy writes are required in all versions of the ip, the version
check is dropped.

Fixes: commit 27bce457 ("i2c: img-scb: Add Imagination Technologies I2C SCB driver")
Signed-off-by: NSifan Naeem <sifan.naeem@imgtec.com>
Acked-by: NJames Hogan <james.hogan@imgtec.com>
Reviewed-by: NJames Hartley <james.hartley@imgtec.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Update the comments to match current behaviour. Shorten some comments.
Update copyrights.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

If we don't clear START generation as soon as possible, it may cause
another message to be generated. To keep the race window as small as
possible, we clear it right at the beginning of the interrupt. We don't
need checking since we always want to stop START and STOP generation on
the next occasion after we started it.

This patch improves the situation but sadly does not completely fix it.
It is still to be researched if we can do better given this HW design.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Due to broken HW design, master IRQs are more timing critical, so give
them precedence over slave IRQ.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

The manual says (55.4.8.6) that HW does automatically send STOP after
NACK was received. My measuerments confirm that.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Setting up new messages was done in process context while handling a
message was in interrupt context. Because of the HW design, this IP core
is sensitive to timing, so the context switches were too expensive. Move
this setup to interrupt context as well.

In my test setup, this fixed the occasional 'data byte sent twice' issue
which a number of people have seen. It also fixes to send REP_START
after a read message which was wrongly send as a STOP + START sequence
before.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

We want to reuse this function later.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

We make sure to reinit the HW in the timeout case; then we know that
interrupts are always disabled in the sections protected by the
spinlock. Thus, we can simply remove it which is a preparation for
further refactoring. While here, rename the timeout variable to
time_left which is way more readable.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

We don't need to init HW before every transfer since we know the HW
state then. HW init at probe time is enough. While here, add setting the
clock register which belongs to init HW. Also, set MDBS bit since not
setting it is prohibited according to the manual.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Pull strscpy string copy function implementation from Chris Metcalf.

Chris sent this during the merge window, but I waffled back and forth on
the pull request, which is why it's going in only now.

The new "strscpy()" function is definitely easier to use and more secure
than either strncpy() or strlcpy(), both of which are horrible nasty
interfaces that have serious and irredeemable problems.

strncpy() has a useless return value, and doesn't NUL-terminate an
overlong result.  To make matters worse, it pads a short result with
zeroes, which is a performance disaster if you have big buffers.

strlcpy(), by contrast, is a mis-designed "fix" for strlcpy(), lacking
the insane NUL padding, but having a differently broken return value
which returns the original length of the source string.  Which means
that it will read characters past the count from the source buffer, and
you have to trust the source to be properly terminated.  It also makes
error handling fragile, since the test for overflow is unnecessarily
subtle.

strscpy() avoids both these problems, guaranteeing the NUL termination
(but not excessive padding) if the destination size wasn't zero, and
making the overflow condition very obvious by returning -E2BIG.  It also
doesn't read past the size of the source, and can thus be used for
untrusted source data too.

So why did I waffle about this for so long?

Every time we introduce a new-and-improved interface, people start doing
these interminable series of trivial conversion patches.

And every time that happens, somebody does some silly mistake, and the
conversion patch to the improved interface actually makes things worse.
Because the patch is mindnumbing and trivial, nobody has the attention
span to look at it carefully, and it's usually done over large swatches
of source code which means that not every conversion gets tested.

So I'm pulling the strscpy() support because it *is* a better interface.
But I will refuse to pull mindless conversion patches.  Use this in
places where it makes sense, but don't do trivial patches to fix things
that aren't actually known to be broken.

* 'strscpy' of git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile:
  tile: use global strscpy() rather than private copy
  string: provide strscpy()
  Make asm/word-at-a-time.h available on all architectures

Pull md fixes from Neil Brown:
 "Assorted fixes for md in 4.3-rc.

  Two tagged for -stable, and one is really a cleanup to match and
  improve kmemcache interface.

* tag 'md/4.3-fixes' of git://neil.brown.name/md:
  md/bitmap: don't pass -1 to bitmap_storage_alloc.
  md/raid1: Avoid raid1 resync getting stuck
  md: drop null test before destroy functions
  md: clear CHANGE_PENDING in readonly array
  md/raid0: apply base queue limits *before* disk_stack_limits
  md/raid5: don't index beyond end of array in need_this_block().
  raid5: update analysis state for failed stripe
  md: wait for pending superblock updates before switching to read-only

Pull MIPS updates from Ralf Baechle:
 "This week's round of MIPS fixes:
   - Fix JZ4740 build
   - Fix fallback to GFP_DMA
   - FP seccomp in case of ENOSYS
   - Fix bootmem panic
   - A number of FP and CPS fixes
   - Wire up new syscalls
   - Make sure BPF assembler objects can properly be disassembled
   - Fix BPF assembler code for MIPS I"

* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
  MIPS: scall: Always run the seccomp syscall filters
  MIPS: Octeon: Fix kernel panic on startup from memory corruption
  MIPS: Fix R2300 FP context switch handling
  MIPS: Fix octeon FP context switch handling
  MIPS: BPF: Fix load delay slots.
  MIPS: BPF: Do all exports of symbols with FEXPORT().
  MIPS: Fix the build on jz4740 after removing the custom gpio.h
  MIPS: CPS: #ifdef on CONFIG_MIPS_MT_SMP rather than CONFIG_MIPS_MT
  MIPS: CPS: Don't include MT code in non-MT kernels.
  MIPS: CPS: Stop dangling delay slot from has_mt.
  MIPS: dma-default: Fix 32-bit fall back to GFP_DMA
  MIPS: Wire up userfaultfd and membarrier syscalls.

Pull irq fixes from Thomas Gleixner:
 "This update contains:

   - Fix for a long standing race affecting /proc/irq/NNN

   - One line fix for ARM GICV3-ITS counting the wrong data

   - Warning silencing in ARM GICV3-ITS.  Another GCC trying to be
     overly clever issue"

* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  irqchip/gic-v3-its: Count additional LPIs for the aliased devices
  irqchip/gic-v3-its: Silence warning when its_lpi_alloc_chunks gets inlined
  genirq: Fix race in register_irq_proc()

The MIPS syscall handler code used to return -ENOSYS on invalid
syscalls. Whilst this is expected, it caused problems for seccomp
filters because the said filters never had the change to run since
the code returned -ENOSYS before triggering them. This caused
problems on the chromium testsuite for filters looking for invalid
syscalls. This has now changed and the seccomp filters are always
run even if the syscall is invalid. We return -ENOSYS once we
return from the seccomp filters. Moreover, similar codepaths have
been merged in the process which simplifies somewhat the overall
syscall code.
Signed-off-by: NMarkos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/11236/Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Pull x86 fixes from Ingo Molnar:
 "Fixes all around the map: W+X kernel mapping fix, WCHAN fixes, two
  build failure fixes for corner case configs, x32 header fix and a
  speling fix"

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/headers/uapi: Fix __BITS_PER_LONG value for x32 builds
  x86/mm: Set NX on gap between __ex_table and rodata
  x86/kexec: Fix kexec crash in syscall kexec_file_load()
  x86/process: Unify 32bit and 64bit implementations of get_wchan()
  x86/process: Add proper bound checks in 64bit get_wchan()
  x86, efi, kasan: Fix build failure on !KASAN && KMEMCHECK=y kernels
  x86/hyperv: Fix the build in the !CONFIG_KEXEC_CORE case
  x86/cpufeatures: Correct spelling of the HWP_NOTIFY flag

Pull timer fixes from Ingo Molnar:
 "An abs64() fix in the watchdog driver, and two clocksource driver
  NO_IRQ assumption fixes"

* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  clocksource: Fix abs() usage w/ 64bit values
  clocksource/drivers/keystone: Fix bad NO_IRQ usage
  clocksource/drivers/rockchip: Fix bad NO_IRQ usage

Pull EFI fixes from Ingo Molnar:
 "Two EFI fixes: one for x86, one for ARM, fixing a boot crash bug that
  can trigger under newer EFI firmware"

* 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  arm64/efi: Fix boot crash by not padding between EFI_MEMORY_RUNTIME regions
  x86/efi: Fix boot crash by mapping EFI memmap entries bottom-up at runtime, instead of top-down

Pull drm fixes from Dave Airlie:
 "Bunch of fixes all over the place, all pretty small: amdgpu, i915,
  exynos, one qxl and one vmwgfx.

  There is also a bunch of mst fixes, I left some cleanups in the series
  as I didn't think it was worth splitting up the tested series"

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (37 commits)
  drm/dp/mst: add some defines for logical/physical ports
  drm/dp/mst: drop cancel work sync in the mstb destroy path (v2)
  drm/dp/mst: split connector registration into two parts (v2)
  drm/dp/mst: update the link_address_sent before sending the link address (v3)
  drm/dp/mst: fixup handling hotplug on port removal.
  drm/dp/mst: don't pass port into the path builder function
  drm/radeon: drop radeon_fb_helper_set_par
  drm: handle cursor_set2 in restore_fbdev_mode
  drm/exynos: Staticize local function in exynos_drm_gem.c
  drm/exynos: fimd: actually disable dp clock
  drm/exynos: dp: remove suspend/resume functions
  drm/qxl: recreate the primary surface when the bo is not primary
  drm/amdgpu: only print meaningful VM faults
  drm/amdgpu/cgs: remove import_gpu_mem
  drm/i915: Call non-locking version of drm_kms_helper_poll_enable(), v2
  drm: Add a non-locking version of drm_kms_helper_poll_enable(), v2
  drm/vmwgfx: Fix a command submission hang regression
  drm/exynos: remove unused mode_fixup() code
  drm/exynos: remove decon_mode_fixup()
  drm/exynos: remove fimd_mode_fixup()
  ...