Jerome, On Tue, May 30, 2023 at 07:49:33PM +0300, Dmitry Rokosov wrote: > Jerome, > > On Tue, May 30, 2023 at 06:14:01PM +0200, Jerome Brunet wrote: > > On Wed 17 May 2023 at 16:33, Dmitry Rokosov <ddrokosov@xxxxxxxxxxxxxx> wrote: > > > > > A1 SoC has four clock controllers on the board: PLL, Peripherals, CPU, > > > and Audio. The audio clock controller is different from others, but the > > > rest are very similar from a functional and regmap point of view. > > > This patch series add support for Amlogic A1 PLL and Peripherals clock > > > drivers. > > > It blocks all A1 peripherals mainline support and a couple of patch series, > > > which were already reviewed and acked, but weren't merged due to pending > > > clock controller drivers series, e.g. > > > https://lore.kernel.org/all/20230418111612.19479-1-ddrokosov@xxxxxxxxxxxxxx/ > > > > > > TODO: CPU and Audio clock controllers are not included in this patch > > > series, it will be sent later. The following clks from these controllers > > > are not supported for now: > > > * Audio clks - vad, mclk_vad, mclk_d, resample_a, locker_in, mclk_b, > > > pdmdclk, pdmsysclk, eqdrc, spdifin, mclk_a, audio2_toaudiotop, > > > audio2_tovad, audio2_toddr_vad, audio2_tdmin_vad, audio2_pdm, > > > audio2_ddr_arb, audio_audiolocker, audio_eqdrc, audio_resamplea, > > > audio_spdifin, audio_toddrb, audio_toddra, audio_frddrb, audio_frddra, > > > audio_tdmoutb, audio_tdmouta, audio_loopbacka, audio_tdminlb, > > > audio_tdminb, audio_tdmina, audio_ddr_arb, mclk_c > > > > > > * CPU clks: cpu_fixed_source_sel0, cpu_fixed_source_div0, > > > cpu_fixed_source_sel1, cpu_fixed_source_div1, cpu_clk > > > > > > Validation: > > > * to double check all clk flags run below helper script: > > > pushd /sys/kernel/debug/clk > > > for f in *; do > > > if [[ -f "$f/clk_flags" ]]; then > > > flags="$(cat $f/clk_flags | awk '{$1=$1};1' | sed ':a;N;$!ba;s/\n/ | /g')" > > > echo -e "$f: $flags" > > > fi > > > done > > > popd > > > > > > * to trace current clks state use '/sys/kernel/debug/clk/clk_dump' node > > > with jq post-processing: > > > $ cat /sys/kernel/debug/clk/clk_dump | jq '.' > clk_dump.json > > > > > > * to debug clk rate propagation, compile kernel with the following > > > definition: > > > $ sed -i "s/undef CLOCK_ALLOW_WRITE_DEBUGFS/define CLOCK_ALLOW_WRITE_DEBUGFS/g" drivers/clk/clk.c > > > after that, clk_rate debug node for each clock will be available for > > > write operation > > > > > > > Applied, Thx > > Thank you very much! > > I noticed that you have applied version 15 of the patch. If I understand > correctly, this is because CLK_SET_RATE_NO_REPARENT for all RTC children > is preferred from your perspective. I agree with this decision. However, > version 16 includes a small refactoring patch that corrects a > misstyping by changing all sys_* related dividers and gates to > read-only operations, since they are set up from BootROM. Should I > submit this patch separately at a later time? > > Also could you please clarify the official process for handling these > patches? Currently, I don't see them in the linux-amlogic for-next > branches. Is there a separate repository for clock changes? I believe > the next repository is a fork of the clock framework, but unfortunately, > I cannot seem to locate it on git.kernel.org... Apologies for the > potentially simple questions... I came across your development tree on GitHub, and I noticed that you've applied the latest v16 version. Apologies for the noise! https://github.com/BayLibre/clk-meson/commit/84af914404dbc01f388c440cac72428784b8a161#diff-9e8258bc44803391f868de351058eef033cf3828d52801e1711f786501bc5911 -- Thank you, Dmitry : , , . , , , , , , , . , , . CONFIDENTIALITY NOTICE: This email and any files attached to it are confidential. If you are not the intended recipient you are notified that using, copying, distributing or taking any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error please notify the sender and delete this email.
