Message ID | 5e55012567da74870e1fb2edc2dc513b5821e523.1666801017.git.daniel@makrotopia.org (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | [v2] clk: mediatek: fix dependency of MT7986 ADC clocks | expand |
Quoting Daniel Golle (2022-10-26 09:18:07) > It seems like CLK_INFRA_ADC_FRC_CK always need to be enabled for > CLK_INFRA_ADC_26M_CK to work. Instead of adding this dependency to the > mtk-thermal and mt6577_auxadc drivers, add dependency to the clock > driver clk-mt7986-infracfg.c. Is this a cleanup patch? Or a pre-requisite for thermal and auxadc drivers? I don't understand the priority of this patch. Should I apply it to fix a regression?
On Thu, Oct 27, 2022 at 02:41:49PM -0700, Stephen Boyd wrote: > Quoting Daniel Golle (2022-10-26 09:18:07) > > It seems like CLK_INFRA_ADC_FRC_CK always need to be enabled for > > CLK_INFRA_ADC_26M_CK to work. Instead of adding this dependency to the > > mtk-thermal and mt6577_auxadc drivers, add dependency to the clock > > driver clk-mt7986-infracfg.c. > > Is this a cleanup patch? Or a pre-requisite for thermal and auxadc > drivers? I don't understand the priority of this patch. Should I apply > it to fix a regression? I'd say 'no', as AUXADC and thermal has not yet been added to mt7986a.dtsi, also the corresponding clocks are currently still unused. So while this commit does fix a previous commit, it doesn't have a direct impact and will only matter once thermal and auxadc units are added to mt7986.dtsi.
Quoting Daniel Golle (2022-10-27 16:24:35) > On Thu, Oct 27, 2022 at 02:41:49PM -0700, Stephen Boyd wrote: > > Quoting Daniel Golle (2022-10-26 09:18:07) > > > It seems like CLK_INFRA_ADC_FRC_CK always need to be enabled for > > > CLK_INFRA_ADC_26M_CK to work. Instead of adding this dependency to the > > > mtk-thermal and mt6577_auxadc drivers, add dependency to the clock > > > driver clk-mt7986-infracfg.c. > > > > Is this a cleanup patch? Or a pre-requisite for thermal and auxadc > > drivers? I don't understand the priority of this patch. Should I apply > > it to fix a regression? > > I'd say 'no', as AUXADC and thermal has not yet been added to > mt7986a.dtsi, also the corresponding clocks are currently still unused. > So while this commit does fix a previous commit, it doesn't have a direct > impact and will only matter once thermal and auxadc units are added to > mt7986.dtsi. Thanks! Please include these details next time.
On Wed, 26 Oct 2022 17:18:07 +0100, Daniel Golle wrote: > It seems like CLK_INFRA_ADC_FRC_CK always need to be enabled for > CLK_INFRA_ADC_26M_CK to work. Instead of adding this dependency to the > mtk-thermal and mt6577_auxadc drivers, add dependency to the clock > driver clk-mt7986-infracfg.c. > > Applied, thanks! [1/1] clk: mediatek: fix dependency of MT7986 ADC clocks commit: a46315295489933209e902638cd287aeb5f982ab Best regards,
diff --git a/drivers/clk/mediatek/clk-mt7986-infracfg.c b/drivers/clk/mediatek/clk-mt7986-infracfg.c index d90727a53283c7..49666047bf0ed5 100644 --- a/drivers/clk/mediatek/clk-mt7986-infracfg.c +++ b/drivers/clk/mediatek/clk-mt7986-infracfg.c @@ -153,7 +153,7 @@ static const struct mtk_gate infra_clks[] = { 18), GATE_INFRA1(CLK_INFRA_MSDC_66M_CK, "infra_msdc_66m", "infra_sysaxi_d2", 19), - GATE_INFRA1(CLK_INFRA_ADC_26M_CK, "infra_adc_26m", "csw_f26m_sel", 20), + GATE_INFRA1(CLK_INFRA_ADC_26M_CK, "infra_adc_26m", "infra_adc_frc", 20), GATE_INFRA1(CLK_INFRA_ADC_FRC_CK, "infra_adc_frc", "csw_f26m_sel", 21), GATE_INFRA1(CLK_INFRA_FBIST2FPC_CK, "infra_fbist2fpc", "nfi1x_sel", 23), /* INFRA2 */