linux.git/drivers/clk/ti, branch v6.5

clk: ti: clkctrl: check return value of kasprintf()

2023-06-17T04:52:50+00:00

kasprintf() returns a pointer to dynamically allocated memory.
Pointer could be NULL in case allocation fails. Check pointer validity.
Identified with coccinelle (kmerr.cocci script).

Fixes: 852049594b9a ("clk: ti: clkctrl: convert subclocks to use proper names also")
Fixes: 6c3090520554 ("clk: ti: clkctrl: Fix hidden dependency to node name")
Signed-off-by: Claudiu Beznea 
Link: https://lore.kernel.org/r/20230530093913.1656095-8-claudiu.beznea@microchip.com
Reviewed-by: Tony Lindgren 
Signed-off-by: Stephen Boyd

Merge branches 'clk-xilinx', 'clk-broadcom' and 'clk-platform' into clk-next

2023-04-25T18:50:36+00:00

 - BCM63268 timer clock and reset controller
 - Convert platform clk drivers to remove_new

* clk-xilinx:
  clocking-wizard: Support higher frequency accuracy
  clk: zynqmp: pll: Remove the limit

* clk-broadcom:
  clk: bcm: Add BCM63268 timer clock and reset driver
  dt-bindings: clock: Add BCM63268 timer binding
  dt-bindings: reset: add BCM63268 timer reset definitions
  dt-bindings: clk: add BCM63268 timer clock definitions

* clk-platform: (25 commits)
  clk: xilinx: Convert to platform remove callback returning void
  clk: x86: Convert to platform remove callback returning void
  clk: uniphier: Convert to platform remove callback returning void
  clk: ti: Convert to platform remove callback returning void
  clk: tegra: Convert to platform remove callback returning void
  clk: stm32: Convert to platform remove callback returning void
  clk: mvebu: Convert to platform remove callback returning void
  clk: mmp: Convert to platform remove callback returning void
  clk: keystone: Convert to platform remove callback returning void
  clk: hisilicon: Convert to platform remove callback returning void
  clk: stm32mp1: Convert to platform remove callback returning void
  clk: scpi: Convert to platform remove callback returning void
  clk: s2mps11: Convert to platform remove callback returning void
  clk: pwm: Convert to platform remove callback returning void
  clk: palmas: Convert to platform remove callback returning void
  clk: hsdk-pll: Convert to platform remove callback returning void
  clk: fixed-rate: Convert to platform remove callback returning void
  clk: fixed-mmio: Convert to platform remove callback returning void
  clk: fixed-factor: Convert to platform remove callback returning void
  clk: axm5516: Convert to platform remove callback returning void
  ...

clk: ti: Convert to platform remove callback returning void

2023-03-29T02:31:45+00:00

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is (mostly) ignored
and this typically results in resource leaks. To improve here there is a
quest to make the remove callback return void. In the first step of this
quest all drivers are converted to .remove_new() which already returns
void.

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König 
Link: https://lore.kernel.org/r/20230312161512.2715500-28-u.kleine-koenig@pengutronix.de
Signed-off-by: Stephen Boyd

clk: ti: Use of_address_to_resource()

2023-03-27T16:38:48+00:00

Replace of_get_address() and of_translate_address() calls with single
call to of_address_to_resource().

Signed-off-by: Rob Herring 
Link: https://lore.kernel.org/r/20230319163217.226144-1-robh@kernel.org
Reviewed-by: Tony Lindgren 
Signed-off-by: Stephen Boyd

clk: ti: fix typo in ti_clk_retry_init() code comment

2022-11-23T01:04:20+00:00

Replace "not" with "node".

Signed-off-by: Dario Binacchi 
Link: https://lore.kernel.org/r/20221106155625.3476564-1-dario.binacchi@amarulasolutions.com
Signed-off-by: Stephen Boyd

clk: ti: dra7-atl: don't allocate `parent_names' variable

2022-11-23T01:02:03+00:00

The `parent_names' variable was freed also in case of kzalloc() error.
Instead of modifying the code to perform a proper memory release, I
decided to fix the bug by not allocating memory.
Since only one parent name is referenced, it is not necessary to
allocate this variable at runtime and therefore you can avoid calling
the kzalloc() function. This simplifies the code (even calls to kfree
can be removed) and improves the performance of the routine.

Signed-off-by: Dario Binacchi 
Reported-by: kernel test robot 
Link: https://lore.kernel.org/r/20221113181147.1626585-2-dario.binacchi@amarulasolutions.com
Tested-by: Tony Lindgren 
Reviewed-by: Tony Lindgren 
Signed-off-by: Stephen Boyd

clk: ti: change ti_clk_register[_omap_hw]() API

2022-11-23T01:01:47+00:00

The ti_clk_register() and ti_clk_register_omap_hw() functions are always
called with the parameter of type "struct device" set to NULL, since the
functions from which they are called always have a parameter of type
"struct device_node". Replacing "struct device" type parameter with
"struct device_node" will allow you to register a TI clock to the common
clock framework by taking advantage of the facilities provided by the
"struct device_node" type. Further, adding the "of_" prefix to the name
of these functions explicitly binds them to the "struct device_node"
type.

The patch has been tested on a Beaglebone board.

Signed-off-by: Dario Binacchi 
Tested-by: Tony Lindgren 
Reviewed-by: Tony Lindgren 
Link: https://lore.kernel.org/r/20221113181147.1626585-1-dario.binacchi@amarulasolutions.com
Signed-off-by: Stephen Boyd

Merge tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux

2022-10-08T17:06:48+00:00

Pull clk updates from Stephen Boyd:
 "We have some late breaking reports that a patch series to rework clk
  rate range support broke boot on some devices, so I've left that
  branch out of this. Hopefully we can get to that next week, or punt on
  it and let it bake another cycle. That means we don't really have any
  changes to the core framework this time around besides a few typo
  fixes. Instead this is all clk driver updates and fixes.

  The usual suspects are here (again), with Qualcomm dominating the
  diffstat. We look to have gained support for quite a few new Qualcomm
  SoCs and Dmitry worked on updating many of the existing Qualcomm
  drivers to use clk_parent_data. After that we have MediaTek drivers
  getting some much needed updates, in particular to support GPU DVFS.
  There are also quite a few Samsung clk driver patches, but that's
  mostly because there was a maintainer change and so last release we
  missed some of those patches.

  Overall things look normal, but I'm slowly reviewing core framework
  code nowadays and that shows given the rate range patches had to be
  yanked last minute. Let's hope this situation changes soon.

  New Drivers:
   - Support for Renesas VersaClock7 clock generator family
   - Add Spreadtrum UMS512 SoC clk support
   - New clock drivers for MediaTek Helio X10 MT6795
   - Display clks for Qualcomm SM6115, SM8450
   - GPU clks for Qualcomm SC8280XP
   - Qualcomm MSM8909 and SM6375 global and SMD RPM clk drivers

  Deleted Drivers:
   - Remove DaVinci DM644x and DM646x clk driver support

  Updates:
   - Convert Baikal-T1 CCU driver to platform driver
   - Split reset support out of primary Baikal-T1 CCU driver
   - Add some missing clks required for RPiVid Video Decoder on
     RaspberryPi
   - Mark PLLC critical on bcm2835
   - More devm helpers for fixed rate registration
   - Various PXA168 clk driver fixes
   - Add resets for MediaTek MT8195 PCIe and USB
   - Miscellaneous of_node_put() fixes
   - Nuke dt-bindings/clk path (again) by moving headers to
     dt-bindings/clock
   - Convert gpio-clk-gate binding to YAML
   - Various fixes to AMD/Xilinx Zynqmp clk driver
   - Graduate AMD/Xilinx "clocking wizard" driver from staging
   - Add missing DPI1_HDMI clock in MT8195 VDOSYS1
   - Clock driver changes to support GPU DVFS on MT8183, MT8192, MT8195
   - Fix GPU clock topology on MT8195
   - Propogate rate changes from GPU clock gate up the tree
   - Clock mux notifiers for GPU-related PLLs
   - Conversion of more "simple" drivers to mtk_clk_simple_probe()
   - Hook up mtk_clk_simple_remove() for "simple" MT8192 clock drivers
   - Fixes to previous |struct clk| to |struct clk_hw| conversion on
     MediaTek
   - Shrink MT8192 clock driver by deduplicating clock parent lists
   - Change order between 'sim_enet_root_clk' and 'enet_qos_root_clk'
     clocks for i.MX8MP
   - Drop unnecessary newline in i.MX8MM dt-bindings
   - Add more MU1 and SAI clocks dt-bindings Ids
   - Introduce slice busy bit check for i.MX93 composite clock
   - Introduce white list bit check for i.MX93 composite clock
   - Add new i.MX93 clock gate
   - Add MU1 and MU2 clocks to i.MX93 clock provider
   - Add SAI IPG clocks to i.MX93 clock provider
   - add generic clocks for U(S)ART available on SAMA5D2 SoCs
   - reset controller support for Polarfire clocks
   - .round_rate and .set rate support for clk-mpfs
   - code cleanup for clk-mpfs
   - PLL support for PolarFire SoC's Clock Conditioning Circuitry
   - Add watchdog, I2C, pin control/GPIO, and Ethernet clocks on R-Car
     V4H
   - Add SDHI, Timer (CMT/TMU), and SPI (MSIOF) clocks on R-Car S4-8
   - Add I2C clocks and resets on RZ/V2M
   - Document clock support for the RZ/Five SoC
   - mux-variant clock using the table variant to select parents
   - clock controller for the rv1126 soc
   - conversion of rk3128 to yaml and relicensing of the yaml bindings
     to gpl2+MIT (following dt-binding guildelines)
   - Exynos7885: add FSYS, TREX and MFC clock controllers
   - Exynos850: add IS and AUD (audio) clock controllers with bindings
   - ExynosAutov9: add FSYS clock controllers with bindings
   - ExynosAutov9: correct clock IDs in bindings of Peric 0 and 1 clock
     controllers, due to duplicated entries. This is an acceptable ABI
     break: recently developed/added platform so without legacies, acked
     by known users/developers
   - ExynosAutov9: add few missing Peric 0/1 gates
   - ExynosAutov9: correct register offsets of few Peric 0/1 clocks
   - Minor code improvements (use of_device_get_match_data() helper,
     code style)
   - Add Krzysztof Kozlowski as co-maintainer of Samsung SoC clocks, as
     he already maintainers that architecture/platform
   - Keep Qualcomm GDSCs enabled when PWRSTS_RET flag is there, solving
     retention issues during suspend of USB on Qualcomm sc7180/sc7280
     and SC8280XP
   - Qualcomm SM6115 and QCM2260 are moved to reuse PLL configuration
   - Qualcomm SDM660 SDCC1 moved to floor clk ops
   - Support for the APCS PLLs for Qualcomm IPQ8064, IPQ8074 and IPQ6018
     was added/fixed
   - The Qualcomm MSM8996 CPU clocks are updated with support for ACD
   - Support for Qualcomm SDM670 GCC and RPMh clks was added
   - Transition to parent_data, parent_hws and use of ARRAY_SIZE() for
     num_parents was done for many Qualcomm SoCs
   - Support for per-reset defined delay on Qualcomm was introduced"

* tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux: (283 commits)
  clk: qcom: gcc-sm6375: Ensure unsigned long type
  clk: qcom: gcc-sm6375: Remove unused variables
  clk: qcom: kpss-xcc: convert to parent data API
  clk: introduce (devm_)hw_register_mux_parent_data_table API
  clk: allow building lan966x as a module
  clk: clk-xgene: simplify if-if to if-else
  clk: ast2600: BCLK comes from EPLL
  clk: clocking-wizard: Depend on HAS_IOMEM
  clk: clocking-wizard: Use dev_err_probe() helper
  clk: nxp: fix typo in comment
  clk: pxa: add a check for the return value of kzalloc()
  clk: vc5: Add support for IDT/Renesas VersaClock 5P49V6975
  dt-bindings: clock: vc5: Add 5P49V6975
  clk: mvebu: armada-37xx-tbg: Remove the unneeded result variable
  clk: ti: dra7-atl: Fix reference leak in of_dra7_atl_clk_probe
  clk: Renesas versaclock7 ccf device driver
  dt-bindings: Renesas versaclock7 device tree bindings
  clk: ti: Balance of_node_get() calls for of_find_node_by_name()
  clk: imx: scu: fix memleak on platform_device_add() fails
  clk: vc5: Use regmap_{set,clear}_bits() where appropriate
  ...

Merge branches 'clk-ofnode', 'clk-bindings', 'clk-cleanup', 'clk-zynq' and 'clk-xilinx' into clk-next

2022-10-04T17:53:04+00:00

 - Miscellaneous of_node_put() fixes
 - Nuke dt-bindings/clk path (again) by moving headers to dt-bindings/clock
 - Convert gpio-clk-gate binding to YAML
 - Various fixes to AMD/Xilinx Zynqmp clk driver
 - Graduate AMD/Xilinx "clocking wizard" driver from staging

* clk-ofnode:
  clk: ti: Balance of_node_get() calls for of_find_node_by_name()
  clk: tegra20: Fix refcount leak in tegra20_clock_init
  clk: tegra: Fix refcount leak in tegra114_clock_init
  clk: tegra: Fix refcount leak in tegra210_clock_init
  clk: sprd: Hold reference returned by of_get_parent()
  clk: berlin: Add of_node_put() for of_get_parent()
  clk: at91: dt-compat: Hold reference returned by of_get_parent()
  clk: qoriq: Hold reference returned by of_get_parent()
  clk: oxnas: Hold reference returned by of_get_parent()
  clk: st: Hold reference returned by of_get_parent()
  clk: tegra: Add missing of_node_put()
  clk: meson: Hold reference returned by of_get_parent()
  clk: nomadik: Add missing of_node_put()

* clk-bindings:
  dt-bindings: clock: drop minItems equal to maxItems
  dt-bindings: clock: gpio-gate-clock: Convert to json-schema
  dt-bindings: clock: Move versaclock.h to dt-bindings/clock
  dt-bindings: clock: Move lochnagar.h to dt-bindings/clock

* clk-cleanup:
  clk: allow building lan966x as a module
  clk: clk-xgene: simplify if-if to if-else
  clk: nxp: fix typo in comment
  clk: mvebu: armada-37xx-tbg: Remove the unneeded result variable
  clk: ti: dra7-atl: Fix reference leak in of_dra7_atl_clk_probe
  clkdev: Simplify devm_clk_hw_register_clkdev() function
  clkdev: Remove never used devm_clk_release_clkdev()
  clk: Remove never used devm_of_clk_del_provider()
  clk: pistachio: Fix initconst confusion
  clk: clk-npcm7xx: Remove unused struct npcm7xx_clk_gate_data and npcm7xx_clk_div_fixed_data
  clk: do not initialize ret
  clk: remove extra empty line
  clk: Fix comment typo
  clk: move from strlcpy with unused retval to strscpy

* clk-zynq:
  clk: zynqmp: pll: rectify rate rounding in zynqmp_pll_round_rate
  clk: zynqmp: Check the return type zynqmp_pm_query_data
  clk: zynqmp: Add a check for NULL pointer
  clk: zynqmp: Replaced strncpy() with strscpy()
  clk: zynqmp: Fix stack-out-of-bounds in strncpy`
  clk: zynqmp: make bestdiv unsigned

* clk-xilinx:
  clk: clocking-wizard: Depend on HAS_IOMEM
  clk: clocking-wizard: Use dev_err_probe() helper
  clk: clocking-wizard: Update the compatible
  clk: clocking-wizard: Fix the reconfig for 5.2
  clk: clocking-wizard: Rename nr-outputs to xlnx,nr-outputs
  clk: clocking-wizard: Move clocking-wizard out
  dt-bindings: add documentation of xilinx clocking wizard

clk: ti: dra7-atl: Fix reference leak in of_dra7_atl_clk_probe

2022-10-01T00:39:46+00:00

pm_runtime_get_sync() will increment pm usage counter.
Forgetting to putting operation will result in reference leak.
Add missing pm_runtime_put_sync in some error paths.

Fixes: 9ac33b0ce81f ("CLK: TI: Driver for DRA7 ATL (Audio Tracking Logic)")
Signed-off-by: Miaoqian Lin 
Link: https://lore.kernel.org/r/20220602030838.52057-1-linmq006@gmail.com
Reviewed-by: Tony Lindgren 
Signed-off-by: Stephen Boyd