There are a few manually-calculated variable-length struct allocations
left, this converts them to use struct_size. Found with the following
git grep command
git grep -A1 'kzalloc.*sizeof[^_].*+'
Signed-off-by: Stephen Kitt <steve@sk2.org>
Link: https://lkml.kernel.org/r/20190927185110.29897-1-steve@sk2.org
Acked-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
[sboyd@kernel.org: Add grep command]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Clock drivers for Armada AP and Armada CP use the same function to
generate unique clock name. A third drivers is coming with the same
need, so it's time to move this function in a common file.
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Link: https://lkml.kernel.org/r/20190710134346.30239-3-gregory.clement@bootlin.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
There are a few spelling mistakes in comments and a pr_err
error message. Fix these.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
These > comparisons should be >= to prevent reading beyond the end of
of the clk_data->hws[] buffer.
The clk_data->hws[] array is allocated in cp110_syscon_common_probe()
when we do:
cp110_clk_data = devm_kzalloc(dev, sizeof(*cp110_clk_data) +
sizeof(struct clk_hw *) * CP110_CLK_NUM,
GFP_KERNEL);
As you can see, it has CP110_CLK_NUM elements which is equivalent to
CP110_MAX_CORE_CLOCKS + CP110_MAX_GATABLE_CLOCKS.
Fixes: d3da3eaef7 ("clk: mvebu: new driver for Armada CP110 system controller")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Thanks to new documentation, we have a better view of the clock tree.
There were few mistakes in the first version of this driver, the main one
being the parental link between the clocks. Actually the tree is more
flat that we though. Most of the IP blocks require two clocks: one for
the IP itself and one for accessing the registers, and unlike what we
wrote there is no link between these two clocks.
The other mistakes were about the name of the clocks: the root clock is
not the Audio PLL but the PLL0, and what we called the EIP clock is named
the x2 Core clock and is used by other IP block than the EIP ones.
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Mark an array of strings static const and remove the dereference
of a function pointer when assigning to the platform driver probe
struct member.
drivers/clk/mvebu/cp110-system-controller.c:89:12:
warning: symbol 'gate_base_names' was not declared. Should it be static?
drivers/clk/mvebu/cp110-system-controller.c:447:18:
error: cannot dereference this type
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
This commit updates the CP110 system controller driver to add the
definition for a missing clock.
The SDIO clock is dedicated driving the SDHCI interface and its frequency
is 400MHz (2/5 of PLL source clock).
The SDIO interface should be bound to this clock and not the core clock
as in the older code.
Using the wrong clock lead to a maximum SDHCI frequency of 250 Mhz, while
the HW really supports up to 400 Mhz.
This patch also fixes the NAND clock relationship documentation.
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
[gregory.clement@free-electrons.com:
- use sdio instead of emmc to name the clock]
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
The initial intent when the binding of the cp110 system controller was to
have one flat node. The idea being that what is currently a clock-only
driver in drivers would become a MFD driver, exposing the clock, GPIO and
pinctrl functionality. However, after taking a step back, this would lead
to a messy binding. Indeed, a single node would be a GPIO controller,
clock controller, pinmux controller, and more.
This patch adopts a more classical solution of a top-level syscon node
with sub-nodes for the individual devices. The main benefit will be to
have each functional block associated to its own sub-node where we can
put its own properties.
The introduction of the Armada 7K/8K is still in the early stage so the
plan is to remove the old binding. However, we don't want to break the
device tree compatibility for the few devices already in the field. For
this we still keep the support of the legacy compatible string with a big
warning in the kernel about updating the device tree.
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
Using the *-clock-output-names property was a convenient way to have a
unique name for each clock even when there are multiple cp110 blocks
as we can find on Armada 8K.
However it has some drawbacks: the main one being a stronger link than
necessary between the driver and the device tree. For example the clock
name can't be changed, removed or moved. It is still the early stage of
introduction of the Armada 7K/8K and the hardware is still not totally
documented, especially for the clock part. By removing the use of
*-clock-output-names it will be easier to add new clocks without breaking
the compatibility.
The name of each clock is now created by using its physical address as a
prefix (as it was done for the platform device names). Thanks to this we
have an automatic way to compute a unique name.
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
In preparation to the addition of a new clock, rename the goto labels
used to handle the failure cases using a name related to the failure
cause. This will allow to insert additional failing cases without
renaming all the labels.
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
This commit:
- makes the GOP_DP (bit 9) gatable clock a child clock of the
SD_MMC_GOP (bit 18) clock, as it should have been. The clock for bit
18 was just named SD_MMC, but since it also covers the GOP block, it
is renamed SD_MMC_GOP.
- makes the MG (bit 5) gatable clock a child clock of the MG_CORE
clock (bit 6)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
The Kconfig currently controlling compilation of this code is:
drivers/clk/mvebu/Kconfig:config ARMADA_CP110_SYSCON
drivers/clk/mvebu/Kconfig: bool
...meaning that it currently is not being built as a module by anyone.
Lets remove the modular code that is essentially orphaned, so that
when reading the driver there is no doubt it is builtin-only.
We explicitly disallow a driver unbind, since that doesn't have a
sensible use case anyway, and it allows us to drop the ".remove"
code for non-modular drivers.
Since module_platform_driver() uses the same init level priority as
builtin_platform_driver() the init ordering remains unchanged with
this commit.
Also note that MODULE_DEVICE_TABLE is a no-op for non-modular code.
We also delete the MODULE_LICENSE tag etc. since all that information
is already contained at the top of the file in the comments.
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@codeaurora.org>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: linux-clk@vger.kernel.org
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Now that we have clk_hw based provider APIs to register clks, we
can get rid of struct clk pointers while registering clks in Armada
CP110 system controller driver. This commit introduces new
API and registration for all clocks in CP110 HW blocks.
Signed-off-by: Marcin Wojtas <mw@semihalf.com>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Original commit, which added support for Armada CP110 system controller
used global variables for storing all clock information. It worked
fine for Armada 7k SoC, with single CP110 block. After dual-CP110 Armada 8k
was introduced, the data got overwritten and corrupted.
This patch fixes the issue by allocating resources dynamically in the
driver probe and storing it as platform drvdata.
Fixes: d3da3eaef7 ("clk: mvebu: new driver for Armada CP110 system ...")
Signed-off-by: Marcin Wojtas <mw@semihalf.com>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
CC: <stable@vger.kernel.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Armada CP110 system controller comprises its own routine responsble
for registering gate clocks. Among others 'flags' field in
struct clk_init_data was not set, using a random values, which
may cause an unpredicted behavior.
This patch fixes the problem by resetting all fields of clk_init_data
before assigning values for all gated clocks of Armada 7k/8k SoCs family.
Fixes: d3da3eaef7 ("clk: mvebu: new driver for Armada CP110 system ...")
Signed-off-by: Marcin Wojtas <mw@semihalf.com>
CC: <stable@vger.kernel.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
The Armada CP110 system controller provides, amongst other things, a
number of clocks for the platform: a small number of core clocks, and
then a number of gatable clocks, derived from some of the core
clocks. Those clocks are configured via registers of the CP110 System
Controller.
The CP110 is the other core HW block (next to the AP806) used in the
Marvel Armada 7K and 8K SoCs.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[sboyd@codeaurora.org: Silence some checkpatch noise]
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>