We currently have a hidden dependency to the device tree node name for
the clkctrl clocks. Instead of using standard node name like "clock", we
must use "l4-per-clkctrl" type naming so the clock driver can find the
associated clock domain. Further, if "clk" is specified for a clock node
name, the driver sets TI_CLK_CLKCTRL_COMPAT flag that uses different
logic for the clock name based on the parent node name for the all the
clkctrl clocks for the SoC.
If the clock node naming dependency is not understood, the related
clockdomain is not found, or a wrong one can get used if a clock manager
has multiple clock domains.
As each clkctrl instance represents a single clock domain, let's allow
using domain specific compatible names to specify the clock domain.
This simplifies things and removes the hidden dependency to the node
name. And then later on, after the node names have been standardized,
we can drop the related code for parsing the node names.
Let's also update the binding to use standard "clock" node naming
instead of "clk" and add the missing description for reg.
Cc: devicetree@vger.kernel.org
Cc: Rob Herring <robh+dt@kernel.org>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Acked-by: Rob Herring <robh@kernel.org>
Acked-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Tero Kristo <t-kristo@ti.com>
then a bunch of driver updates and a handful of new drivers. In terms of
diffstat the Qualcomm and Amlogic drivers are high up there because of all the
clk data introcued by new drivers. The Nvidia Tegra driver had a lot of work
done this cycle too to support suspend/resume and memory controllers. And the
OMAP clk driver got proper clk and reset handling in place.
Rounding out the patches are various updates to remove unused data, mark things
static, correct incorrect data in drivers, etc. All the little things that
improve drivers and maintain code health. I will point out that there's a patch
in here for the GPIO clk driver, that almost nobody uses, which changes
behavior and causes clk_set_rate() to try to change the GPIO gate clk's parent.
Other than that things are fairly well SoC specific here.
Core:
- Add a clk provider API to get current parent index
- Plug a memory leak in clk_unregister() path
New Drivers:
- CGU in Ingenix X1000
- Bitmain BM1880 clks
- Qualcomm MSM8998 GPU clk controllers
- Qualcomm SC7180 GCC and RPMH clk controllers
- Qualcomm QCS404 Q6SSTOP clk controllers
- Add support for the Renesas R-Car M3-W+ (r8a77961) SoC
- Add support for the Renesas RZ/G2N (r8a774b1) SoC
- Add Tegra20/30 External Memory Clock (EMC) support
Updates:
- Make gpio gate clks propagate rate setting up to parent
- Prepare Armada 3700 for suspend to RAM by moving PCIe suspend/resume priority
- Drop unused variables, enums, etc. in various clk drivers
- Convert various drivers to use devm_platform_ioremap_resource()
- Use struct_size() some more in various clk drivers
- Improve Rockchip px30 clk tree
- Add suspend/resume support to Tegra210 clk driver
- Reimplement SOR clks on earlier Tegra SoCs, helping HDMI and DP
- Allwinner DT exports and H6 clk tree fixes
- Proper clk and reset handling for OMAP SoCs
- Revamped TI divider clk to clamp max divider
- Make 1443X/1416X PLL clock structure common for reusing among i.MX8 SoCs
- Drop IMX7ULP_CLK_MIPI_PLL clock, it shouldn't be used
- Add VIDEO2_PLL clock for imx8mq
- Add missing gate clock for pll1/2 fixed dividers on i.MX8 SoCs
- Add sm1 support in the Amlogic audio clock controller
- Switch some clocks on R-Car Gen2/3 to .determine_rate()
- Remove Renesas R-Car Gen2 legacy DT clock support
- Improve arithmetic divisions on Renesas R-Car Gen2 and Gen3
- Improve Renesas R-Car Gen3 SD clock handling
- Add rate table for Samsung exynos542x GPU and VPLL clks
- Fix potential CPU performance degradation after system suspend/resume cycle
on exynos542x SoCs
-----BEGIN PGP SIGNATURE-----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=qJNg
-----END PGP SIGNATURE-----
Merge tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk updates from Stephen Boyd:
"This merge window we have one small clk provider API in the core
framework and then a bunch of driver updates and a handful of new
drivers. In terms of diffstat the Qualcomm and Amlogic drivers are
high up there because of all the clk data introcued by new drivers.
The Nvidia Tegra driver had a lot of work done this cycle too to
support suspend/resume and memory controllers. And the OMAP clk driver
got proper clk and reset handling in place.
Rounding out the patches are various updates to remove unused data,
mark things static, correct incorrect data in drivers, etc. All the
little things that improve drivers and maintain code health. I will
point out that there's a patch in here for the GPIO clk driver, that
almost nobody uses, which changes behavior and causes clk_set_rate()
to try to change the GPIO gate clk's parent. Other than that things
are fairly well SoC specific here.
Core:
- Add a clk provider API to get current parent index
- Plug a memory leak in clk_unregister() path
New Drivers:
- CGU in Ingenix X1000
- Bitmain BM1880 clks
- Qualcomm MSM8998 GPU clk controllers
- Qualcomm SC7180 GCC and RPMH clk controllers
- Qualcomm QCS404 Q6SSTOP clk controllers
- Add support for the Renesas R-Car M3-W+ (r8a77961) SoC
- Add support for the Renesas RZ/G2N (r8a774b1) SoC
- Add Tegra20/30 External Memory Clock (EMC) support
Updates:
- Make gpio gate clks propagate rate setting up to parent
- Prepare Armada 3700 for suspend to RAM by moving PCIe
suspend/resume priority
- Drop unused variables, enums, etc. in various clk drivers
- Convert various drivers to use devm_platform_ioremap_resource()
- Use struct_size() some more in various clk drivers
- Improve Rockchip px30 clk tree
- Add suspend/resume support to Tegra210 clk driver
- Reimplement SOR clks on earlier Tegra SoCs, helping HDMI and DP
- Allwinner DT exports and H6 clk tree fixes
- Proper clk and reset handling for OMAP SoCs
- Revamped TI divider clk to clamp max divider
- Make 1443X/1416X PLL clock structure common for reusing among i.MX8
SoCs
- Drop IMX7ULP_CLK_MIPI_PLL clock, it shouldn't be used
- Add VIDEO2_PLL clock for imx8mq
- Add missing gate clock for pll1/2 fixed dividers on i.MX8 SoCs
- Add sm1 support in the Amlogic audio clock controller
- Switch some clocks on R-Car Gen2/3 to .determine_rate()
- Remove Renesas R-Car Gen2 legacy DT clock support
- Improve arithmetic divisions on Renesas R-Car Gen2 and Gen3
- Improve Renesas R-Car Gen3 SD clock handling
- Add rate table for Samsung exynos542x GPU and VPLL clks
- Fix potential CPU performance degradation after system
suspend/resume cycle on exynos542x SoCs"
* tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux: (160 commits)
clk: aspeed: Add RMII RCLK gates for both AST2500 MACs
MAINTAINERS: Add entry for BM1880 SoC clock driver
clk: Add common clock driver for BM1880 SoC
dt-bindings: clock: Add devicetree binding for BM1880 SoC
clk: Add clk_hw_unregister_composite helper function definition
clk: Zero init clk_init_data in helpers
clk: ingenic: Allow drivers to be built with COMPILE_TEST
MAINTAINERS: Update section for Ux500 clock drivers
clk: mark clk_disable_unused() as __init
clk: Fix memory leak in clk_unregister()
clk: Ingenic: Add CGU driver for X1000.
dt-bindings: clock: Add X1000 bindings.
clk: tegra: Use match_string() helper to simplify the code
clk: pxa: fix one of the pxa RTC clocks
clk: sprd: Use IS_ERR() to validate the return value of syscon_regmap_lookup_by_phandle()
clk: armada-xp: remove unused code
clk: tegra: Fix build error without CONFIG_PM_SLEEP
clk: tegra: Add missing stubs for the case of !CONFIG_PM_SLEEP
clk: tegra: Optimize PLLX restore on Tegra20/30
clk: tegra: Add suspend and resume support on Tegra210
...
Commit 3d8598fb9c ("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: 3d8598fb9c ("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>
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>
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>
If we do a clk_get() for a clock that does not exists, we have
_ti_omap4_clkctrl_xlate() return uninitialized data if no match
is found. This can be seen in some cases with SLAB_DEBUG enabled:
Unable to handle kernel paging request at virtual address 5a5a5a5a
...
clk_hw_create_clk.part.33
sysc_notifier_call
notifier_call_chain
blocking_notifier_call_chain
device_add
Let's fix this by setting a found flag only when we find a match.
Reported-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Fixes: 88a172526c ("clk: ti: add support for clkctrl clocks")
Signed-off-by: Tony Lindgren <tony@atomide.com>
Tested-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Tested-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
We need to always call clkdm_clk_enable() and clkdm_clk_disable() even
the clkctrl clock(s) enabled for the domain do not have any gate register
bits. Otherwise clockdomains may never get enabled except when devices get
probed with the legacy "ti,hwmods" devicetree property.
Fixes: 88a172526c ("clk: ti: add support for clkctrl clocks")
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
RNG and TIMER12 are reserved for secure side usage only on HS devices,
so disable their clkctrl clocks on HS SoCs also.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Certain clkctrl clocks (like the USB_OTG_SS4) do not exist on some
variants of the dra7x SoC. Append a flag for these clocks and skip
the registration in cases where the clocks do not exist.
Reported-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Commit a72d785021 ("clk: ti: Prepare for remove of OF node name")
changed the code to use kasprintf() for provider->clkdm_name but also
changed the offset used later on by three. We don't need to change the
offset as we already have the extra three characters in the format for
kasprintf with "%pOFnxxx".
This caused the clocks with TI_CLK_CLKCTRL_COMPAT to have NULL
clk->clkdm_name for omap4 and 5. And null clkdm_name can cause module
reset, enable, and idle to fail.
The issue can also be seen also when enabling DEBUG for clkctrl.c
and then we start seeing "clock: could not associate" messages for
omap4 and 5 as the generated name is something like "l4_wkclkdm" instead
of "l4_wkup_clkdm" that's needed.
Let's fix the issue with a partial revert of commit a72d785021 ("clk:
ti: Prepare for remove of OF node name").
ALso note that in general code should not depend on the dts node names.
And the node names should be generic types like clock-domain in this case.
This could be fixed later by using separate compatible properties for the
clockdomains, or by adding soc_device_match() table with reg offsets
to the driver. But let's fix the regression first.
Fixes: a72d785021 ("clk: ti: Prepare for remove of OF node name")
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Remove the usage of CLK_IS_BASIC flag completely from TI clock driver.
In most cases, the use is completely redundant, but in some cases
we need to use the new API to check if the clock is an OMAP clock or not.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Tested-by: Keerthy <j-keerthy@ti.com>
Convert string compares of DT node names to use of_node_name_eq helper
instead. This removes direct access to the node name pointer.
For instances using of_node_cmp, this has the side effect of now using
case sensitive comparisons. This should not matter for any FDT based
system which all of these are.
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: Tero Kristo <t-kristo@ti.com>
Cc: Ulf Hansson <ulf.hansson@linaro.org>
Cc: linux-renesas-soc@vger.kernel.org
Cc: linux-clk@vger.kernel.org
Cc: linux-omap@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Signed-off-by: Rob Herring <robh@kernel.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Geert Uytterhoeven <geert+renesas@glider.be> (clk-mstp)
Acked-by: Ulf Hansson <ulf.hansson@linaro.org> (ux500)
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
* clk-ti:
clk: ti: Prepare for remove of OF node name
clk: Clean up suspend/resume coding style
clk: ti: Add functions to save/restore clk context
clk: clk: Add clk_gate_restore_context function
clk: Add functions to save/restore clock context en-masse
clk: ti: dra7: add new clkctrl data
clk: ti: dra7xx: rename existing clkctrl data as compat data
clk: ti: am43xx: add new clkctrl data for am43xx
clk: ti: am43xx: rename existing clkctrl data as compat data
clk: ti: am33xx: add new clkctrl data for am33xx
clk: ti: am33xx: rename existing clkctrl data as compat data
clk: ti: clkctrl: replace dashes from clkdm name with underscore
clk: ti: clkctrl: support multiple clkctrl nodes under a cm node
dt-bindings: clock: dra7xx: add clkctrl indices for new data layout
dt-bindings: clock: am43xx: add clkctrl indices for new data layout
dt-bindings: clock: am33xx: add clkctrl indices for new data layout
Another patch is going to change this code to use %pOFn for DT node
names. Fix up the code to make this easy to pick this side of the merge
instead of fixing it up in a merge commit later.
Cc: Tero Kristo <t-kristo@ti.com>
Cc: Rob Herring <robh+dt@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
The new clkctrl data layout for dra7xx is split based on clockdomain
boundaries. Previously the split was based on CM boundaries. This patch
adds the new data as separate data entity, retaining the compatibility
data also for now. The compatibility data can be removed once no longer
needed.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Tony Lindgren <tony@atomide.com>
Rename the existing clkctrl data in preparation of upcoming clkdm
based split for it. Once the DT data has transitioned also, the
compat data can be removed.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Tony Lindgren <tony@atomide.com>
The new clkctrl data layout for am43xx is split based on clockdomain
boundaries. Previously the split was based on CM boundaries. This patch
adds the new data as separate data entity, retaining the compatibility
data also for now. The compatibility data can be removed once no longer
needed.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Tony Lindgren <tony@atomide.com>
Rename the existing clkctrl data in preparation of upcoming clkdm
based split for it. Once the DT data has transitioned also, the
compat data can be removed.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Tony Lindgren <tony@atomide.com>
The new clkctrl data layout for am33xx is split based on clockdomain
boundaries. Previously the split was based on CM boundaries. This patch
adds the new data as separate data entity, retaining the compatibility
data also for now. The compatibility data can be removed once no longer
needed.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Tony Lindgren <tony@atomide.com>
Rename the existing clkctrl data in preparation of upcoming clkdm
based split for it. Once the DT data has transitioned also, the
compat data can be removed.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Tony Lindgren <tony@atomide.com>
The change in the DTS data node naming prevents using underscore
within the node names and force usage of dash instead. On the other
hand, clockdomains use underscore instead of dash, so this must be
replaced within the driver code so that the mapping between the two
can be done properly.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Tony Lindgren <tony@atomide.com>
Currently, only one clkctrl node can be added under a specific CM node
due to limitation with the implementation. Modify the code to pick-up
clockdomain name from the clkctrl node instead of CM node if provided.
Also, add a new flag to the TI clock driver so that both modes can
be supported simultaneously.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Tested-by: Tony Lindgren <tony@atomide.com>
In preparation to remove the node name pointer from struct device_node,
convert printf users to use the %pOFn format specifier.
Cc: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-renesas-soc@vger.kernel.org
Cc: linux-omap@vger.kernel.org
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Certain clkctrl clocks, notably the display ones, use the
CLK_SET_RATE_PARENT feature extensively. Add support for this flag
to the clkctrl clocks.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Reported-by: Jyri Sarha <jsarha@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Tested-by: Jyri Sarha <jsarha@ti.com>
Add data for am43xx clkctrl clocks, and register it within the clkctrl
driver.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Acked-by: Stephen Boyd <sboyd@codeaurora.org>
In case the clkctrl node contains assigned-clock-* entries, registering
the provider can fail with -EPROBE_DEFER. In this case, add the
provider to the retry_init clock list so it will be cleaned up later.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Acked-by: Stephen Boyd <sboyd@codeaurora.org>
In certain cases it is possible that the timekeeping has been suspended
already when attempting to disable/enable a clkctrl clock. This will
happen at least on am43xx platform when attempting to enable / disable
the clockevent source itself, burping out a warning from timekeeping core.
The sequence of events leading to this:
-> timekeeping_suspend()
-> clockevents_suspend()
-> omap_clkevt_idle()
-> omap_hwmod_idle()
-> _omap4_clkctrl_clk_disable()
-> _omap4_is_timeout()
Avoid the issue by checking if the timekeeping is suspended and using
the fallback udelay approach for checking timeouts.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Acked-by: Stephen Boyd <sboyd@codeaurora.org>
Clkctrl clocks now support clockdomain init also. This will be needed
so that hwmod core can drop the support for clockdomain handling.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Current node name does not convey any information, as it is always "clk".
Instead, print out the full node path, which will tell us better where
something went wrong.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Add data for omap4 clkctrl clocks, and register it within the clkctrl
driver.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Previously, hwmod core has been used for controlling the hwmod level
clocks directly. This has certain drawbacks, like being unable to share
the clocks for multiple users, missing usecounting and generally being
totally incompatible with the common clock framework.
This patch adds support for clkctrl clocks for addressing the above
issues. These support the modulemode handling, which will replace the
direct hwmod clkctrl linkage. Any optional clocks are also supported,
gate, mux and divider.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Tony Lindgren