merged this merge window. The Amlogic driver was missing some flags
causing rates to be rounded improperly or clk_set_rate() to fail. The
Samsung driver wasn't freeing everything on error paths and improperly
saving/restoring PLL state across suspend/resume. The at91 driver was
calling msleep() too early when scheduling hadn't started, so we put in
place a quick solution until we can handle this sort of problem in the
core framework. There were also problems with the Allwinner driver and
operator precedence being incorrect causing subtle bugs. Finally, the TI
driver was duplicating aliases and not delaying long enough leading to
some unexpected timeouts.
-----BEGIN PGP SIGNATURE-----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=0xes
-----END PGP SIGNATURE-----
Merge tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk fixes from Stephen Boyd:
"Fixes for various clk driver issues that happened because of code we
merged this merge window.
The Amlogic driver was missing some flags causing rates to be rounded
improperly or clk_set_rate() to fail. The Samsung driver wasn't
freeing everything on error paths and improperly saving/restoring PLL
state across suspend/resume. The at91 driver was calling msleep() too
early when scheduling hadn't started, so we put in place a quick
solution until we can handle this sort of problem in the core
framework.
There were also problems with the Allwinner driver and operator
precedence being incorrect causing subtle bugs. Finally, the TI driver
was duplicating aliases and not delaying long enough leading to some
unexpected timeouts"
* tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
clk: ti: clkctrl: Fix failed to enable error with double udelay timeout
clk: ti: dra7-atl-clock: Remove ti_clk_add_alias call
clk: sunxi-ng: a80: fix the zero'ing of bits 16 and 18
clk: sunxi: Fix operator precedence in sunxi_divs_clk_setup
clk: ast2600: Fix enabling of clocks
clk: at91: avoid sleeping early
clk: imx8m: Use SYS_PLL1_800M as intermediate parent of CLK_ARM
clk: samsung: exynos5420: Preserve PLL configuration during suspend/resume
clk: samsung: exynos542x: Move G3D subsystem clocks to its sub-CMU
clk: samsung: exynos5433: Fix error paths
clk: at91: sam9x60: fix programmable clock
clk: meson: g12a: set CLK_MUX_ROUND_CLOSEST on the cpu clock muxes
clk: meson: g12a: fix cpu clock rate setting
clk: meson: gxbb: let sar_adc_clk_div set the parent clock rate
Some RCGs (the gfx_3d_src_clk in msm8998 for example) are basically just
some constant ratio from the input across the entire frequency range. It
would be great if we could specify the frequency table as a single entry
constant ratio instead of a long list, ie:
{ .src = P_GPUPLL0_OUT_EVEN, .pre_div = 3 },
{ }
So, lets support that.
We need to fix a corner case in qcom_find_freq() where if the freq table
is non-null, but has no frequencies, we end up returning an "entry" before
the table array, which is bad. Then, we need ignore the freq from the
table, and instead base everything on the requested freq.
Suggested-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Jeffrey Hugo <jeffrey.l.hugo@gmail.com>
Link: https://lkml.kernel.org/r/20191031185715.15504-1-jeffrey.l.hugo@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
When MSM8998 support was added, and analysis was done to determine what
clocks would be consumed. That analysis had a flaw, which caused the
pnoc to be skipped. The pnoc clock needs to be on to access the uart
for the console. The clock is on from boot, but has no consumer votes
in the RPM. When we attempt to boot the modem, it causes the RPM to
turn off pnoc, which kills our access to the console and causes CPU hangs.
We need pnoc to be defined, so that clk_smd_rpm_handoff() will put in
an implicit vote for linux and prevent issues when booting modem.
Hopefully pnoc can be consumed by the interconnect framework in future
so that Linux can rely on explicit votes.
Fixes: 6131dc81211c ("clk: qcom: smd: Add support for MSM8998 rpm clocks")
Signed-off-by: Jeffrey Hugo <jeffrey.l.hugo@gmail.com>
Link: https://lkml.kernel.org/r/20191107190615.5656-1-jeffrey.l.hugo@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
We need to control five additional clocks and a reset inorder to boot the
modem on msm8998. If we can boot the modem, we have a place to run the
wlan firmware and get wifi up and running.
Signed-off-by: Jeffrey Hugo <jeffrey.l.hugo@gmail.com>
Link: https://lkml.kernel.org/r/20191107192136.5880-1-jeffrey.l.hugo@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Add support for the global clock controller found on SC7180
based devices. This should allow most non-multimedia device
drivers to probe and control their clocks.
Signed-off-by: Taniya Das <tdas@codeaurora.org>
Link: https://lkml.kernel.org/r/20191014102308.27441-6-tdas@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Return NULL in the cases where the clk_hw is not registered with the
clock provider, but the clock consumer still requests for a clock id.
Signed-off-by: Taniya Das <tdas@codeaurora.org>
Link: https://lkml.kernel.org/r/20191014102308.27441-3-tdas@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Update the init data name for each of the dynamic frequency switch
controlled clock associated with the RCG clock name, so that it can be
generated as per the hardware plan. Thus update the macro accordingly.
Signed-off-by: Taniya Das <tdas@codeaurora.org>
Link: https://lkml.kernel.org/r/20191014102308.27441-2-tdas@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Add support for the Q6SSTOP clock control used on qcs404
based devices. This would allow wcss remoteproc driver to
control the required WCSS Q6SSTOP clock/reset controls to
bring the subsystem out of reset and shutdown the WCSS Q6DSP.
Signed-off-by: Govind Singh <govinds@codeaurora.org>
Link: https://lkml.kernel.org/r/20191011132928.9388-3-govinds@codeaurora.org
[sboyd@kernel.org: Sort makefile]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Initialize the clock drivers on sdm845 and qcs404 in core_initcall so we
can have earlier access to cpufreq during booting.
Signed-off-by: Amit Kucheria <amit.kucheria@linaro.org>
Acked-by: Stephen Boyd <sboyd@kernel.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Link: https://lore.kernel.org/r/75ae9c3a1c0e69b95818c6ffe7181fdeaaf2d70e.1571656015.git.amit.kucheria@linaro.org
Video-In and -Out interconnect clocks need to stay on all the
time for the peripheral to work and we do not model the actual
interconnect at this point. So mark them as critical for now.
Signed-off-by: Heiko Stuebner <heiko.stuebner@theobroma-systems.com>
Link: https://lore.kernel.org/r/20190917081903.25139-4-heiko@sntech.de
The clocks in the px30 critical clock section are from the regular cru not
the pmucru, so move them to the correct place.
Signed-off-by: Heiko Stuebner <heiko.stuebner@theobroma-systems.com>
Link: https://lore.kernel.org/r/20190917081903.25139-3-heiko@sntech.de
Some IPs, such as NAND, EMMC, SDIO and SDMMC need clock of 50% duty
cycle, divfree50 can generate clock of 50% duty cycle even in odd
value divisor.
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Link: https://lore.kernel.org/r/20190917081903.25139-2-heiko@sntech.de
MBUS clock will be referenced in MBUS controller node.
Export it.
Acked-by: Maxime Ripard <mripard@kernel.org>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Jernej Skrabec <jernej.skrabec@siol.net>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQRcEzekXsqa64kGDp7j7w1vZxhRxQUCXb2lYgAKCRDj7w1vZxhR
xX++AP931wTnnYugarLm6km0xXE/UZhYl+XBbRnUYqe+a3aN5AD9EoyuhZ3hpNzM
juyr/D9G17u6WBHFWaZdXwOp0ylIYgA=
=hj+3
-----END PGP SIGNATURE-----
Merge tag 'sunxi-clk-fixes-for-5.4-1' of https://git.kernel.org/pub/scm/linux/kernel/git/sunxi/linux into clk-fixes
Two patches that fix some operator precedence and zeroing of bits
* tag 'sunxi-clk-fixes-for-5.4-1' of https://git.kernel.org/pub/scm/linux/kernel/git/sunxi/linux:
clk: sunxi-ng: a80: fix the zero'ing of bits 16 and 18
clk: sunxi: Fix operator precedence in sunxi_divs_clk_setup
Commit 3d8598fb9c5a ("clk: ti: clkctrl: use fallback udelay approach if
timekeeping is suspended") added handling for cases when timekeeping is
suspended. But looks like we can still get occasional "failed to enable"
errors on the PM runtime resume path with udelay() returning faster than
expected.
With ti-sysc interconnect target module driver this leads into device
failure with PM runtime failing with "failed to enable" clkctrl error.
Let's fix the issue with a delay of two times the desired delay as in
often done for udelay() to account for the inaccuracy.
Fixes: 3d8598fb9c5a ("clk: ti: clkctrl: use fallback udelay approach if timekeeping is suspended")
Cc: Keerthy <j-keerthy@ti.com>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Link: https://lkml.kernel.org/r/20190930154001.46581-1-tony@atomide.com
Tested-by: Keerthy <j-keerthy@ti.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
It is not correct that sys3_pll_out use sys2_pll1_ref_sel as parent.
According to the current imx_clk_sccg_pll design, it uses both
bypass1/2, however set bypass2 as 1 is not correct, because it will
make sys[x]_pll_out use wrong parent and might access wrong registers.
So correct bypass2 to 0 and fix sys3_pll_out_sels.
Fixes: e9dda4af685f ("clk: imx: Refactor entire sccg pll clk")
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Add support for the R-Car M3-W+ (R8A77961) SoC to the Renesas Clock
Pulse Generator / Module Standby and Software Reset driver.
R-Car M3-W+ is very similar to R-Car M3-W (R8A77960), which allows for
both SoCs to share a driver. R-Car M3-W+ lacks a few modules, so their
clocks must be nullified.
Based on a patch in the BSP by Takeshi Kihara
<takeshi.kihara.df@renesas.com>.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Tested-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Link: https://lore.kernel.org/r/20191023122941.12342-5-geert+renesas@glider.be
Rename CONFIG_CLK_R8A7796 for R-Car M3-W (R8A77960) to
CONFIG_CLK_R8A77960, to avoid confusion with R-Car M3-W+ (R8A77961),
which will use CONFIG_CLK_R8A77961.
Extend the dependency of CONFIG_CLK_R8A77960 from CONFIG_ARCH_R8A7796 to
CONFIG_ARCH_R8A77960, to relax dependencies for a future rename of the
SoC configuration symbol.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Link: https://lore.kernel.org/r/20191023122941.12342-4-geert+renesas@glider.be
There is no need to terminate a function with a semicolon. Remove it.
Reported-by: Biju Das <biju.das@bp.renesas.com>
Fixes: 7ce36da900c0a2ff ("clk: renesas: cpg-mssr: Add support for R-Car M3-N")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20191016150711.30305-1-geert+renesas@glider.be
The existing width field used to check divider validity does not provide
enough protection against bad values. For example, if max divider value
is 4, the smallest all-1 bitmask that can hold this value is 7, which
allows values higher than 4 to be used. This typically causes
unpredictable results with hardware. So far this issue hasn't been
noticed as most of the dividers actually have maximum values which fit
the whole bitfield, but there are certain clocks for which this is a
problem, like dpll4_m4 divider on omap3 devices.
Thus, convert the whole validity logic to use min,max and mask values
for determining if a specific divider is valid or not. This prevents
the odd cases where bad value would otherwise be written to a divider
config register.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Adam Ford <aford173@gmail.com>
Cleanup the ti_clk_parse_divider_data to pass the divider data struct
directly instead of individual values of it. This makes it easier
to modify the implementation later on.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Adam Ford <aford173@gmail.com>
Cleanup couple of TI divider clock internal APIs. These currently pass
huge amount of parameters, which makes it difficult to track what is
going on. Abstract most of these under struct clk_omap_div which gets
passed over the APIs.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Adam Ford <aford173@gmail.com>
Due to the way ti sysc and hardreset line control is now implemented,
it is not possible to poll the clock status for gfx clock independent
of hardreset line control. Thus, add a flag to prevent handling this
status bit from clock driver. Correct sequencing of events is guaranteed
by ti-sysc bus driver.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Due to the way ti sysc and hardreset line control is now implemented,
it is not possible to poll the clock status for gfx clock independent
of hardreset line control. Thus, add a flag to prevent handling this
status bit from clock driver. Correct sequencing of events is guaranteed
by ti-sysc bus driver.
Reported-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Tero Kristo <t-kristo@ti.com>
The PRUSS module on AM33xx SoCs has a hardreset line and is controlled
by a PRCM reset line. Any clkctrl enable/disable operations cannot be
checked for module enabled/disabled status independent of the reset
operation, and this causes some unwanted timeouts in the kernel and
unbalanced states for the PRUSS clocks. These details should be handled
by the driver integration code itself.
Add the CLKF_NO_IDLEST flag to the PRUSS clkctrl clock so that these
module status checks are skipped.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
The PRUSS modules on AM43xx SoCs have a hardreset line and are controlled
by a PRCM reset line. Any clkctrl enable/disable operations cannot be
checked for module enabled/disabled status independent of the reset
operation, and this causes some unwanted timeouts in the kernel and
unbalanced states for the PRUSS clocks. These details should be handled
by the driver integration code itself.
Add the CLKF_NO_IDLEST flag to the PRUSS clkctrl clock so that these
module status checks are skipped.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
The IPU and DSP remote processor cores and their corresponding MMUs on
OMAP5 SoCs have hardreset lines associated with them and are controlled
by a PRCM reset line each. Any clkctrl enable/disable operations cannot
be checked for module enabled/disabled status independent of the reset
operation, and this causes some unwanted timeouts in the kernel and
unbalanced states for these clocks. These details should be handled by
the driver integration code itself.
Add the CLKF_NO_IDLEST flag to both the IPU and DSP clkctrl clocks so
that these module status checks are skipped.
Signed-off-by: Suman Anna <s-anna@ti.com>
Signed-off-by: Tero Kristo <t-kristo@ti.com>
The IPU and DSP remote processor cores and their corresponding MMUs on
OMAP4 SoCs have hardreset lines associated with them and are controlled
by a PRCM reset line each. Any clkctrl enable/disable operations cannot
be checked for module enabled/disabled status independent of the reset
operation, and this causes some unwanted timeouts in the kernel and
unbalanced states for these clocks. These details should be handled by
the driver integration code itself.
Add the CLKF_NO_IDLEST flag to both the IPU and DSP clkctrl clocks so
that these module status checks are skipped.
Signed-off-by: Suman Anna <s-anna@ti.com>
Signed-off-by: Tero Kristo <t-kristo@ti.com>
The IPU and DSP remote processor cores and their corresponding MMUs on
DRA7 SoCs have hardreset lines associated with them and are controlled
by a PRCM reset line each. Any clkctrl enable/disable operations cannot
be checked for module enabled/disabled status independent of the reset
operation, and this causes some unwanted timeouts in the kernel and
unbalanced states for these clocks. These details should be handled by
the driver integration code itself.
Add the CLKF_NO_IDLEST flag to both the IPU and DSP clkctrl clocks so
that these module status checks are skipped.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Standby status is provided for certain clkctrl clocks to see if the
given module has entered standby or not. This is mostly needed by
remoteproc code to see if the remoteproc has entered standby and the clock
can be turned off safely.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Apply the proper register function for clkctrl clocks, so they get
registered under the clk_hw_omap list also. This allows checking their
type runtime.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
The clk_half_divider_ops is not used outside or declared
outside of drivers/clk/rockchip/clk-half-divider.c so make
it static to avoid the following warning:
drivers/clk/rockchip/clk-half-divider.c:142:22: warning: symbol 'clk_half_divider_ops' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Link: https://lore.kernel.org/r/20191017105348.8061-1-ben.dooks@codethink.co.uk
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Add CLK_SET_RATE_PARENT flag to all clocks on the path from VPLL to G3D,
so the G3D MALI driver can simply adjust the rate of its clock by doing
a single clk_set_rate() call, without the need to know the whole clock
topology in Exynos542x SoCs.
Suggested-by: Marian Mihailescu <mihailescu2m@gmail.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Save and restore top PLL related configuration registers for big (APLL)
and LITTLE (KPLL) cores during suspend/resume cycle. So far, CPU clocks
were reset to default values after suspend/resume cycle and performance
after system resume was affected when performance governor has been selected.
Fixes: 773424326b51 ("clk: samsung: exynos5420: add more registers to restore list")
Signed-off-by: Marian Mihailescu <mihailescu2m@gmail.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Add new table rate for VPLL for Exynos 542x SoC required to support
Mali GPU clock frequencies.
Signed-off-by: Marian Mihailescu <mihailescu2m@gmail.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
The zero'ing of bits 16 and 18 is incorrect. Currently the code
is masking with the bitwise-and of BIT(16) & BIT(18) which is
0, so the updated value for val is always zero. Fix this by bitwise
and-ing value with the correct mask that will zero bits 16 and 18.
Addresses-Coverity: (" Suspicious &= or |= constant expression")
Fixes: b8eb71dcdd08 ("clk: sunxi-ng: Add A80 CCU")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Maxime Ripard <mripard@kernel.org>
r375326 in Clang exposes an issue with operator precedence in
sunxi_div_clk_setup:
drivers/clk/sunxi/clk-sunxi.c:1083:30: warning: operator '?:' has lower
precedence than '|'; '|' will be evaluated first
[-Wbitwise-conditional-parentheses]
data->div[i].critical ?
~~~~~~~~~~~~~~~~~~~~~ ^
drivers/clk/sunxi/clk-sunxi.c:1083:30: note: place parentheses around
the '|' expression to silence this warning
data->div[i].critical ?
^
)
drivers/clk/sunxi/clk-sunxi.c:1083:30: note: place parentheses around
the '?:' expression to evaluate it first
data->div[i].critical ?
^
(
1 warning generated.
It appears that the intention was for ?: to be evaluated first so that
CLK_IS_CRITICAL could be added to clkflags if the critical boolean was
set; right now, | is being evaluated first. Add parentheses around the
?: block to have it be evaluated first.
Fixes: 9919d44ff297 ("clk: sunxi: Use CLK_IS_CRITICAL flag for critical clks")
Link: https://github.com/ClangBuiltLinux/linux/issues/745
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Maxime Ripard <mripard@kernel.org>
The struct clk_ops enable callback for the aspeed gates mixes up the set
to clear and write to set registers.
Fixes: d3d04f6c330a ("clk: Add support for AST2600 SoC")
Reviewed-by: Andrew Jeffery <andrew@aj.id.au>
Signed-off-by: Joel Stanley <joel@jms.id.au>
Link: https://lkml.kernel.org/r/20191016131319.31318-1-joel@jms.id.au
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
As now we also need to probe in the reset driver as well.
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: Allison Randal <allison@lohutok.net>
Cc: Peter Griffin <peter.griffin@linaro.org>
Cc: linux-clk@vger.kernel.org
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Link: https://lkml.kernel.org/r/20191001182546.70090-1-john.stultz@linaro.org
[sboyd@kernel.org: Add comment about reset driver]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
It is not allowed to sleep to early in the boot process and this may lead
to kernel issues if the bootloader didn't prepare the slow clock and main
clock.
This results in the following error and dump stack on the AriettaG25:
bad: scheduling from the idle thread!
Ensure it is possible to sleep, else simply have a delay.
Reported-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lkml.kernel.org/r/20190920153906.20887-1-alexandre.belloni@bootlin.com
Fixes: 80eded6ce8bb ("clk: at91: add slow clks driver")
Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
The mipi pll clock comes from the MIPI PHY PLL output, so
it should not be a fixed clock.
MIPI PHY PLL is in the MIPI DSI space, and it is used as
the bit clock for transferring the pixel data out and its
output clock is configured according to the display mode.
So it should be used only for MIPI DSI and not be exported
out for other usages.
Signed-off-by: Fancy Fang <chen.fang@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
During cpu frequency switching the main "CLK_ARM" is reparented to an
intermediate "step" clock. On imx8mm and imx8mn the 24M oscillator is
used for this purpose but it is extremely slow, increasing wakeup
latencies to the point that i2c transactions can timeout and system
becomes unresponsive.
Fix by switching the "step" clk to SYS_PLL1_800M, matching the behavior
of imx8m cpufreq drivers in imx vendor tree.
This bug was not immediately apparent because upstream arm64 defconfig
uses the "performance" governor by default so no cpufreq transitions
happen.
Fixes: ba5625c3e272 ("clk: imx: Add clock driver support for imx8mm")
Fixes: 96d6392b54db ("clk: imx: Add support for i.MX8MN clock driver")
Cc: stable@vger.kernel.org
Signed-off-by: Leonard Crestez <leonard.crestez@nxp.com>
Link: https://lkml.kernel.org/r/f5d2b9c53f1ed5ccb1dd3c6624f56759d92e1689.1571771777.git.leonard.crestez@nxp.com
Acked-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
