On 8/1/24 12:47, Stephan Gerhold wrote:
On Wed, Jul 31, 2024 at 05:25:30PM +0200, Adam Skladowski wrote:
From: Vladimir Lypak vladimir.lypak@gmail.com
Introduce support for audio card on MSM8953 platform. Main difference between MSM8953 and MSM8916 is Q6AFE CLK API supported by firmware which influence way we set codec clocks. SoCs shipping on at least msm-3.18 should use v2 clocks.
Signed-off-by: Vladimir Lypak vladimir.lypak@gmail.com [Adam: rename functions, add msg] Co-developed-by: Adam Skladowski a39.skl@gmail.com Signed-off-by: Adam Skladowski a39.skl@gmail.com
sound/soc/qcom/apq8016_sbc.c | 43 +++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 41 insertions(+), 2 deletions(-)
diff --git a/sound/soc/qcom/apq8016_sbc.c b/sound/soc/qcom/apq8016_sbc.c index 5a29adbd3f82..3ed35beb671a 100644 --- a/sound/soc/qcom/apq8016_sbc.c +++ b/sound/soc/qcom/apq8016_sbc.c @@ -22,6 +22,11 @@
Is there a particular reason why you decided to hardcode the q6afe_clk_ver for the SoCs rather than finishing up the dynamic detection Otto proposed [1]?
This works for MSM8953 but there are a few SoCs like MSM8909 where both clock API versions exist (depending on the firmware versions). If we want to support them at some point, we will need the dynamic detection anyway. It would be nice to finish up that patch set.
Thanks, Stephan
This probably sound obvious but i don't understand takes Srinivas had on these patches. On top i don't feel good sending code i don't understand much. On a note i managed to slightly modify fallback commit and provide match table translating new clks into v1. In theory if we want we can drop snd_soc_component_set_sysclk() from soundcard driver and use devm_clk_get/clk_set_rate/clk_prepare_enable to manage clocks provided by q6afe_clocks. Biggest issue for me is lack of people who i can even discuss with about, even here i fail to see much feedback.