linux/drivers/clk, branch v5.7

clk: qcom: gcc: Fix parent for gpll0_out_even

2020-05-27T00:15:06Z

Documentation says that gpll0 is parent of gpll0_out_even, somehow driver coded that as bi_tcxo, so fix it Fixes: 2a1d7eb854bb ("clk: qcom: gcc: Add global clock controller driver for SM8150") Reported-by: Jonathan Marek Signed-off-by: Vinod Koul Link: https://lkml.kernel.org/r/20200521052728.2141377-1-vkoul@kernel.org Reviewed-by: Bjorn Andersson Signed-off-by: Stephen Boyd

clk: qcom: sm8250 gcc depends on QCOM_GDSC

2020-05-26T23:27:43Z

The driver will always fail to probe without QCOM_GDSC, so select it. Signed-off-by: Jonathan Marek Link: https://lkml.kernel.org/r/20200523040947.31946-1-jonathan@marek.ca Reviewed-by: Bjorn Andersson Fixes: 3e5770921a88 ("clk: qcom: gcc: Add global clock controller driver for SM8250") Signed-off-by: Stephen Boyd

clk: ti: clkctrl: convert subclocks to use proper names also

2020-05-13T03:18:19Z

Addition of the new internal API to get the clkctrl names missed adding the same conversion in place for the subclocks. This leads into missed parent/child relationships (i.e. orphaned clocks) with mixed node name handling, for example with omap4/omap5 where the l4_per clocks are using new naming, but rest are using old. Fix by converting the subclock registration to pick correct names for the clocks also. Fixes: 6c3090520554 ("clk: ti: clkctrl: Fix hidden dependency to node name") Signed-off-by: Tero Kristo Link: https://lkml.kernel.org/r/20200430083451.8562-1-t-kristo@ti.com Acked-by: Tony Lindgren Signed-off-by: Stephen Boyd

clk: ti: am33xx: fix RTC clock parent

2020-05-13T03:17:46Z

Right now, trying to use RTC purely with the ti-sysc / clkctrl framework fails to enable the RTC module properly. Based on experimentation, this appears to be because RTC is sourced from the clkdiv32k optional clock. TRM is not very clear on this topic, but fix the RTC to use the proper source clock nevertheless. Reported-by: Tony Lindgren Signed-off-by: Tero Kristo Link: https://lkml.kernel.org/r/20200424152301.4018-1-t-kristo@ti.com Acked-by: Tony Lindgren Signed-off-by: Stephen Boyd

clk: ti: clkctrl: Fix Bad of_node_put within clkctrl_get_name

2020-05-13T03:16:10Z

clkctrl_get_name incorrectly calls of_node_put when it is not really doing of_node_get. This causes a boot time warning later on: [ 0.000000] OF: ERROR: Bad of_node_put() on /ocp/interconnect@4a000000/segmen t@0/target-module@5000/cm_core_aon@0/ipu-cm@500/ipu1-clkctrl@20 Fix by dropping the of_node_put from the function. Reported-by: Naresh Kamboju Fixes: 6c3090520554 ("clk: ti: clkctrl: Fix hidden dependency to node name") Signed-off-by: Tero Kristo Link: https://lkml.kernel.org/r/20200424124725.9895-1-t-kristo@ti.com Acked-by: Tony Lindgren Signed-off-by: Stephen Boyd

Merge tag 'v5.7-rockchip-clk-fixes1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip into clk-fixes

2020-05-12T23:36:13Z

Pull one Rockchip clk fix from Heiko Stuebner: - Fix for wrongly defines rk3228 aclk_gpu* * tag 'v5.7-rockchip-clk-fixes1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip: clk: rockchip: fix incorrect configuration of rk3228 aclk_gpu* clocks

clk: tegra: Fix initial rate for pll_a on Tegra124

2020-05-12T23:26:18Z

pll_a_out0 and the I2S clocks are already configured to default to rates corresponding to a 44.1 kHz sampling rate, but the pll_a configuration was set to a default that is not listed in the frequency table, which caused the PLL code to compute an invalid configuration. As a result of this invalid configuration, Jetson TK1 fails to resume from suspend. This used to get papered over because the ASoC driver would force audio clocks to a 44.1 kHz configuration on boot. However, that's not really necessary and was hence removed in commit ff5d18cb04f4 ("ASoC: tegra: Enable audio mclk during tegra_asoc_utils_init()"). Fix the initial rate for pll_a so that it matches the 44.1 kHz entry in the pll_a frequency table. Fixes: ff5d18cb04f4 ("ASoC: tegra: Enable audio mclk during tegra_asoc_utils_init()") Signed-off-by: Thierry Reding Link: https://lkml.kernel.org/r/20200505071655.644773-1-thierry.reding@gmail.com Signed-off-by: Stephen Boyd

clk: impd1: Look up clock-output-names

2020-05-05T19:13:12Z

The IM-PD1 still need to pass the clock output names. Signed-off-by: Linus Walleij Link: https://lkml.kernel.org/r/20200416081348.326833-1-linus.walleij@linaro.org Fixes: 84655b762a27 ("clk: versatile: Add device tree probing for IM-PD1 clocks") Signed-off-by: Stephen Boyd

clk: Unlink clock if failed to prepare or enable

2020-05-05T19:12:00Z

On failing to prepare or enable a clock, remove the core structure from the list it has been inserted as it is about to be freed. This otherwise leads to random crashes when subsequent clocks get registered, during which parsing of the clock tree becomes adventurous. Observed with QEMU's RPi-3 emulation. Fixes: 12ead77432f2 ("clk: Don't try to enable critical clocks if prepare failed") Signed-off-by: Marc Zyngier Cc: Guenter Roeck Cc: Stephen Boyd Cc: Michael Turquette Link: https://lkml.kernel.org/r/20200505140953.409430-1-maz@kernel.org Signed-off-by: Stephen Boyd

clk: sprd: don't gate uart console clock

2020-04-13T19:20:07Z

Don't gate uart1_eb which provides console clock, gating that clock would make serial stop working if serial driver didn't enable that explicitly. Fixes: 0e4b8a2349f3 ("clk: sprd: add clocks support for SC9863A") Signed-off-by: Chunyan Zhang Link: https://lkml.kernel.org/r/20200408020234.31764-1-zhang.lyra@gmail.com Signed-off-by: Stephen Boyd