linux-stable.git/drivers/clk/versatile, branch linux-3.14.y

clk: versatile: sp810: support reentrance

2016-05-11T09:20:59+00:00

commit ec7957a6aa0aaf981fb8356dc47a2cdd01cde03c upstream.

Despite care take to allocate clocks state containers the
SP810 driver actually just supports creating one instance:
all clocks registered for every instance will end up with the
exact same name and __clk_init() will fail.

Rename the timclken<0> .. timclken to sp810__
so every clock on every instance gets a unique name.

This is necessary for the RealView PBA8 which has two SP810
blocks: the second block will not register its clocks unless
every clock on every instance is unique and results in boot
logs like this:

------------[ cut here ]------------
WARNING: CPU: 0 PID: 0 at ../drivers/clk/versatile/clk-sp810.c:137
  clk_sp810_of_setup+0x110/0x154()
Modules linked in:
CPU: 0 PID: 0 Comm: swapper/0 Not tainted
4.5.0-rc2-00030-g352718fc39f6-dirty #225
Hardware name: ARM RealView Machine (Device Tree Support)
[] (unwind_backtrace) from []
             (show_stack+0x10/0x14)
[] (show_stack) from []
             (dump_stack+0x84/0x9c)
[] (dump_stack) from []
             (warn_slowpath_common+0x74/0xb0)
[] (warn_slowpath_common) from []
             (warn_slowpath_null+0x1c/0x24)
[] (warn_slowpath_null) from []
             (clk_sp810_of_setup+0x110/0x154)
[] (clk_sp810_of_setup) from []
             (of_clk_init+0x12c/0x1c8)
[] (of_clk_init) from []
             (time_init+0x20/0x2c)
[] (time_init) from []
             (start_kernel+0x244/0x3c4)
[] (start_kernel) from [<7000807c>] (0x7000807c)
---[ end trace cb88537fdc8fa200 ]---

Cc: Michael Turquette 
Cc: Pawel Moll 
Fixes: 6e973d2c4385 "clk: vexpress: Add separate SP810 driver"
Signed-off-by: Linus Walleij 
Signed-off-by: Stephen Boyd 
Signed-off-by: Greg Kroah-Hartman

clk: versatile: off by one in clk_sp810_timerclken_of_get()

2015-09-21T17:02:17+00:00

commit 3294bee87091be5f179474f6c39d1d87769635e2 upstream.

The ">" should be ">=" or we end up reading beyond the end of the array.

Fixes: 6e973d2c4385 ('clk: vexpress: Add separate SP810 driver')
Signed-off-by: Dan Carpenter 
Acked-by: Pawel Moll 
Signed-off-by: Stephen Boyd 
Signed-off-by: Greg Kroah-Hartman

clk: vexpress: NULL dereference on error path

2014-06-07T17:28:19+00:00

commit 6b4ed8b00e93bd31f24a25f59ed8d1b808d0cc00 upstream.

If the allocation fails then we dereference the NULL in the error path.
Just return directly.

Fixes: ed27ff1db869 ('clk: Versatile Express clock generators ("osc") driver')
Signed-off-by: Dan Carpenter 
Signed-off-by: Pawel Moll 
Signed-off-by: Greg Kroah-Hartman

Merge tag 'clk-for-linus-3.14-part1' of git://git.linaro.org/people/mike.turquette/linux

2014-01-24T02:56:08+00:00

Pull clk framework changes from Mike Turquette:
 "The first half of the clk framework pull request is made up almost
  entirely of new platform/driver support.  There are some conversions
  of existing drivers to the common-clock Device Tree binding, and a few
  non-critical fixes to the framework.

  Due to an entirely unnecessary cyclical dependency with the arm-soc
  tree this pull request is broken into two pieces.  The second piece
  will be sent out after arm-soc sends you the pull request that merged
  in core support for the HiSilicon 3620 platform.  That same pull
  request from arm-soc depends on this pull request to merge in those
  HiSilicon bits without causing build failures"

[ Just did the ARM SoC merges, so getting ready for the second clk tree
  pull request   - Linus ]

* tag 'clk-for-linus-3.14-part1' of git://git.linaro.org/people/mike.turquette/linux: (97 commits)
  devicetree: bindings: Document qcom,mmcc
  devicetree: bindings: Document qcom,gcc
  clk: qcom: Add support for MSM8660's global clock controller (GCC)
  clk: qcom: Add support for MSM8974's multimedia clock controller (MMCC)
  clk: qcom: Add support for MSM8974's global clock controller (GCC)
  clk: qcom: Add support for MSM8960's multimedia clock controller (MMCC)
  clk: qcom: Add support for MSM8960's global clock controller (GCC)
  clk: qcom: Add reset controller support
  clk: qcom: Add support for branches/gate clocks
  clk: qcom: Add support for root clock generators (RCGs)
  clk: qcom: Add support for phase locked loops (PLLs)
  clk: qcom: Add a regmap type clock struct
  clk: Add set_rate_and_parent() op
  reset: Silence warning in reset-controller.h
  clk: sirf: re-arch to make the codes support both prima2 and atlas6
  clk: composite: pass mux_hw into determine_rate
  clk: shmobile: Fix MSTP clock array initialization
  clk: shmobile: Fix MSTP clock index
  ARM: dts: Add clock provider specific properties to max77686 node
  clk: max77686: Register OF clock provider
  ...

clk: versatile: fixup IM-PD1 clock implementation

2014-01-03T21:27:54+00:00

Register both VCO clocks, give per-logical module unique names
to the clocks so we can have several IM-PD1's connected (in
theory). Implement all the fixed-factor clocks as children of
VCO2.

Tested by using the UARTs and the PL181 MMC block on the IM-PD1,
works flawlessly.

Acked-by: Mike Turquette 
Signed-off-by: Linus Walleij

clk: versatile: pass a name to ICST clock provider

2014-01-03T21:27:38+00:00

When we have more than one of these clocks in a system (such as
on the IM-PD1) we need a mechanism to pass a name for the clock.
Refactor to add this as an argument.

Acked-by: Mike Turquette 
Signed-off-by: Linus Walleij

clk: versatile: Staticize clk_sp810_timerclken_of_get

2013-12-20T01:47:32+00:00

clk_sp810_timerclken_of_get is used only in this file. Make it static.

Signed-off-by: Sachin Kamat 
Signed-off-by: Mike Turquette

clk: fixup argument order when setting VCO parameters

2013-10-08T06:01:07+00:00

The order of arguments in the call to vco_set() for the ICST clocks appears to
have been switched in error, which results in the VCO not being initialised
correctly. This in turn stops the integrated LCD on things like Integrator/CP
from working correctly.

This patch fixes the order and restores the expected functionality.

Reviewed-by: Linus Walleij 
Signed-off-by: Jonathan Austin 
Signed-off-by: Mike Turquette 
Cc: stable@vger.kernel.org

clk: add CLK_SET_RATE_NO_REPARENT flag

2013-08-19T19:27:17+00:00

Add a CLK_SET_RATE_NO_REPARENT clock flag, which will prevent muxes
being reparented during clk_set_rate.

To avoid breaking existing platforms, all callers of clk_register_mux()
are adjusted to pass the new flag. Platform maintainers are encouraged
to remove the flag if they wish to allow mux reparenting on set_rate.

Signed-off-by: James Hogan 
Reviewed-by: Stephen Boyd 
Cc: Mike Turquette 
Cc: Russell King 
Cc: Sascha Hauer 
Cc: Stephen Warren 
Cc: Viresh Kumar 
Cc: Kukjin Kim 
Cc: Haojian Zhuang 
Cc: Chao Xie 
Cc: Arnd Bergmann 
Cc: "Emilio López" 
Cc: Gregory CLEMENT 
Cc: Maxime Ripard 
Cc: Prashant Gaikwad 
Cc: Thierry Reding 
Cc: Peter De Schrijver 
Cc: Pawel Moll 
Cc: Catalin Marinas 
Cc: Andrew Chew 
Cc: Doug Anderson 
Cc: Heiko Stuebner 
Cc: Paul Walmsley 
Cc: Sylwester Nawrocki 
Cc: Thomas Abraham 
Cc: Tomasz Figa 
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-samsung-soc@vger.kernel.org
Cc: spear-devel@list.st.com
Cc: linux-tegra@vger.kernel.org
Tested-by: Haojian Zhuang 
Acked-by: Stephen Warren  [tegra]
Acked-by: Maxime Ripard  [sunxi]
Acked-by: Sören Brinkmann  [Zynq]
Signed-off-by: Mike Turquette

clk: vexpress: Use full node name to identify individual clocks

2013-06-20T07:02:18+00:00

Previously all the clocks were reported as "osc". Now it will be
something like "/dcc/osc@0".

Signed-off-by: Pawel Moll 
Signed-off-by: Mike Turquette