| Message ID | 20240502180736.7330-1-a.fertl@t-online.de |
|---|---|
| State | Changes Requested |
| Headers | show |
| Series | [v3,1/1] drivers/rtc: rtc-sun6i: AutoCal Internal OSC Clock | expand |
Hi Alois, On Fri, 3 May 2024, at 6:07 AM, Alois Fertl wrote: > I have a M98-8K PLUS Magcubic TV-Box based on the Allwinner H618 SOC. > On board is a Sp6330 wifi/bt module that requires a 32kHz clock to > operate correctly. Without this change the clock from the SOC is > ~29kHz and BT module does not start up. The patch enables the Internal > OSC Clock Auto Calibration of the H616/H618 which than provides the > necessary 32kHz and the BT module initializes successfully. > Add a flag and set it for H6 AND H616. The H618 is the same as H616 > regarding rtc. > > Signed-off-by: Alois Fertl <a.fertl@t-online.de> > --- > > v1->v2 > - add flag and activate for H6 AND H616 > > v2->v3 > - correct findings from review > > I was hoping to get some feedback regarding the situation on H6, > where an external 32k crystal can be present. > From what I understand from the H6 manual there should be no > conflict as one can select which souce will drive the output. > Should certainly be tested but I don`t have H6 hardware. > > drivers/rtc/rtc-sun6i.c | 17 ++++++++++++++++- > > 1 file changed, 16 insertions(+), 1 deletion(-) > > diff --git a/drivers/rtc/rtc-sun6i.c b/drivers/rtc/rtc-sun6i.c > index e0b85a0d5645..20e81ccdef29 100644 > --- a/drivers/rtc/rtc-sun6i.c > +++ b/drivers/rtc/rtc-sun6i.c > @@ -42,6 +42,11 @@ > > #define SUN6I_LOSC_CLK_PRESCAL 0x0008 > > +#define SUN6I_LOSC_CLK_AUTO_CAL 0x000c > +#define SUN6I_LOSC_CLK_AUTO_CAL_16MS BIT(2) > +#define SUN6I_LOSC_CLK_AUTO_CAL_ENABLE BIT(1) > +#define SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL BIT(0) > + > /* RTC */ > #define SUN6I_RTC_YMD 0x0010 > #define SUN6I_RTC_HMS 0x0014 > @@ -126,7 +131,6 @@ > * registers (R40, H6) > * - SYS power domain controls (R40) > * - DCXO controls (H6) > - * - RC oscillator calibration (H6) > * > * These functions are not covered by this driver. > */ > @@ -138,6 +142,7 @@ struct sun6i_rtc_clk_data { > unsigned int has_losc_en : 1; > unsigned int has_auto_swt : 1; > unsigned int no_ext_losc : 1; > + unsigned int has_auto_cal : 1; > }; > > #define RTC_LINEAR_DAY BIT(0) > @@ -268,6 +273,14 @@ static void __init sun6i_rtc_clk_init(struct > device_node *node, > } > writel(reg, rtc->base + SUN6I_LOSC_CTRL); > > + if (rtc->data->has_auto_cal) { > + /* Enable internal OSC clock auto calibration */ > + reg = SUN6I_LOSC_CLK_AUTO_CAL_16MS | > + SUN6I_LOSC_CLK_AUTO_CAL_ENABLE | > + SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL; > + writel(reg, rtc->base + SUN6I_LOSC_CLK_AUTO_CAL); > + } > + > /* Yes, I know, this is ugly. */ > sun6i_rtc = rtc; > > @@ -380,6 +393,7 @@ static const struct sun6i_rtc_clk_data > sun50i_h6_rtc_data = { > .has_out_clk = 1, > .has_losc_en = 1, > .has_auto_swt = 1, > + .has_auto_cal = 1, > }; > > static void __init sun50i_h6_rtc_clk_init(struct device_node *node) I tried to test this with my H700 board but this struct is not actually in any mainline kernel and looks like it is from an earlier version of the H616 RTC changes which was never actually merged? Talked to Andre on IRC and he thinks the H6 and H616 RTC clocks *should* be equivalent. > @@ -395,6 +409,7 @@ static const struct sun6i_rtc_clk_data > sun50i_h616_rtc_data = { > .has_prescaler = 1, > .has_out_clk = 1, > .no_ext_losc = 1, > + .has_auto_cal = 1, > }; > > static void __init sun50i_h616_rtc_clk_init(struct device_node *node) > -- > 2.39.2 Regards, Ryan
>> >> static void __init sun50i_h616_rtc_clk_init(struct device_node *node) Sorry, this struct. The H6 struct is present. > > Ryan
On Thu, 2 May 2024 20:07:36 +0200 Alois Fertl <a.fertl@t-online.de> wrote: Hi Alois, many thanks for taking care and sending a patch! I think this problem is plaguing some people. > I have a M98-8K PLUS Magcubic TV-Box based on the Allwinner H618 SOC. > On board is a Sp6330 wifi/bt module that requires a 32kHz clock to > operate correctly. Without this change the clock from the SOC is > ~29kHz and BT module does not start up. The patch enables the Internal > OSC Clock Auto Calibration of the H616/H618 which than provides the > necessary 32kHz and the BT module initializes successfully. > Add a flag and set it for H6 AND H616. The H618 is the same as H616 > regarding rtc. (many thanks to Ryan for the head up on this) So what tree is this patch based on? It certainly is not mainline, is it? Mainline never supported the H616 RTC clocks via the RTC driver, this was only through the new driver in the clk directory, in drivers/clk/sunxi-ng/ccu-sun6i-rtc.c: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=8a93720329d4d2 https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d91612d7f01aca Please note that the H6 RTC clocks were intended to be handled via the new driver as well, but this part was reverted: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=60d9f050da63b Please only send patches based on the mainline tree. I doesn't look that hard to adjust the patch to mainline, though to cover the H6 is well this would require this code in both drivers. So when we want to address the H6 as well, it might make sense to solve the problem that triggered the revert first, to only have that change only in one file. > Signed-off-by: Alois Fertl <a.fertl@t-online.de> > --- > > v1->v2 > - add flag and activate for H6 AND H616 > > v2->v3 > - correct findings from review > > I was hoping to get some feedback regarding the situation on H6, > where an external 32k crystal can be present. > From what I understand from the H6 manual there should be no > conflict as one can select which souce will drive the output. > Should certainly be tested but I don`t have H6 hardware. I think I should have H6 boards both with and without an external crystal oscillator, so can check the situation there, but only next week. > drivers/rtc/rtc-sun6i.c | 17 ++++++++++++++++- > > 1 file changed, 16 insertions(+), 1 deletion(-) > > diff --git a/drivers/rtc/rtc-sun6i.c b/drivers/rtc/rtc-sun6i.c > index e0b85a0d5645..20e81ccdef29 100644 > --- a/drivers/rtc/rtc-sun6i.c > +++ b/drivers/rtc/rtc-sun6i.c > @@ -42,6 +42,11 @@ > > #define SUN6I_LOSC_CLK_PRESCAL 0x0008 > > +#define SUN6I_LOSC_CLK_AUTO_CAL 0x000c > +#define SUN6I_LOSC_CLK_AUTO_CAL_16MS BIT(2) > +#define SUN6I_LOSC_CLK_AUTO_CAL_ENABLE BIT(1) > +#define SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL BIT(0) > + > /* RTC */ > #define SUN6I_RTC_YMD 0x0010 > #define SUN6I_RTC_HMS 0x0014 > @@ -126,7 +131,6 @@ > * registers (R40, H6) > * - SYS power domain controls (R40) > * - DCXO controls (H6) > - * - RC oscillator calibration (H6) > * > * These functions are not covered by this driver. > */ > @@ -138,6 +142,7 @@ struct sun6i_rtc_clk_data { > unsigned int has_losc_en : 1; > unsigned int has_auto_swt : 1; > unsigned int no_ext_losc : 1; > + unsigned int has_auto_cal : 1; > }; > > #define RTC_LINEAR_DAY BIT(0) > @@ -268,6 +273,14 @@ static void __init sun6i_rtc_clk_init(struct device_node *node, > } > writel(reg, rtc->base + SUN6I_LOSC_CTRL); > > + if (rtc->data->has_auto_cal) { > + /* Enable internal OSC clock auto calibration */ > + reg = SUN6I_LOSC_CLK_AUTO_CAL_16MS | > + SUN6I_LOSC_CLK_AUTO_CAL_ENABLE | > + SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL; > + writel(reg, rtc->base + SUN6I_LOSC_CLK_AUTO_CAL); > + } > + > /* Yes, I know, this is ugly. */ > sun6i_rtc = rtc; > > @@ -380,6 +393,7 @@ static const struct sun6i_rtc_clk_data sun50i_h6_rtc_data = { > .has_out_clk = 1, > .has_losc_en = 1, > .has_auto_swt = 1, > + .has_auto_cal = 1, > }; > > static void __init sun50i_h6_rtc_clk_init(struct device_node *node) > @@ -395,6 +409,7 @@ static const struct sun6i_rtc_clk_data sun50i_h616_rtc_data = { For the records: this whole struct does not exist in mainline. Cheers, Andre > .has_prescaler = 1, > .has_out_clk = 1, > .no_ext_losc = 1, > + .has_auto_cal = 1, > }; > > static void __init sun50i_h616_rtc_clk_init(struct device_node *node)
On Fri, 03 May 2024 23:02:13 +1200 "Ryan Walklin" <ryan@testtoast.com> wrote: Hi, > On Fri, 3 May 2024, at 6:07 AM, Alois Fertl wrote: > > I have a M98-8K PLUS Magcubic TV-Box based on the Allwinner H618 SOC. > > On board is a Sp6330 wifi/bt module that requires a 32kHz clock to > > operate correctly. Without this change the clock from the SOC is > > ~29kHz and BT module does not start up. The patch enables the Internal > > OSC Clock Auto Calibration of the H616/H618 which than provides the > > necessary 32kHz and the BT module initializes successfully. > > Add a flag and set it for H6 AND H616. The H618 is the same as H616 > > regarding rtc. > > > > Signed-off-by: Alois Fertl <a.fertl@t-online.de> > > --- > > > > v1->v2 > > - add flag and activate for H6 AND H616 > > > > v2->v3 > > - correct findings from review > > > > I was hoping to get some feedback regarding the situation on H6, > > where an external 32k crystal can be present. > > From what I understand from the H6 manual there should be no > > conflict as one can select which souce will drive the output. > > Should certainly be tested but I don`t have H6 hardware. > > > > drivers/rtc/rtc-sun6i.c | 17 ++++++++++++++++- > > > > 1 file changed, 16 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/rtc/rtc-sun6i.c b/drivers/rtc/rtc-sun6i.c > > index e0b85a0d5645..20e81ccdef29 100644 > > --- a/drivers/rtc/rtc-sun6i.c > > +++ b/drivers/rtc/rtc-sun6i.c > > @@ -42,6 +42,11 @@ > > > > #define SUN6I_LOSC_CLK_PRESCAL 0x0008 > > > > +#define SUN6I_LOSC_CLK_AUTO_CAL 0x000c > > +#define SUN6I_LOSC_CLK_AUTO_CAL_16MS BIT(2) > > +#define SUN6I_LOSC_CLK_AUTO_CAL_ENABLE BIT(1) > > +#define SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL BIT(0) > > + > > /* RTC */ > > #define SUN6I_RTC_YMD 0x0010 > > #define SUN6I_RTC_HMS 0x0014 > > @@ -126,7 +131,6 @@ > > * registers (R40, H6) > > * - SYS power domain controls (R40) > > * - DCXO controls (H6) > > - * - RC oscillator calibration (H6) > > * > > * These functions are not covered by this driver. > > */ > > @@ -138,6 +142,7 @@ struct sun6i_rtc_clk_data { > > unsigned int has_losc_en : 1; > > unsigned int has_auto_swt : 1; > > unsigned int no_ext_losc : 1; > > + unsigned int has_auto_cal : 1; > > }; > > > > #define RTC_LINEAR_DAY BIT(0) > > @@ -268,6 +273,14 @@ static void __init sun6i_rtc_clk_init(struct > > device_node *node, > > } > > writel(reg, rtc->base + SUN6I_LOSC_CTRL); > > > > + if (rtc->data->has_auto_cal) { > > + /* Enable internal OSC clock auto calibration */ > > + reg = SUN6I_LOSC_CLK_AUTO_CAL_16MS | > > + SUN6I_LOSC_CLK_AUTO_CAL_ENABLE | > > + SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL; > > + writel(reg, rtc->base + SUN6I_LOSC_CLK_AUTO_CAL); > > + } > > + > > /* Yes, I know, this is ugly. */ > > sun6i_rtc = rtc; > > > > @@ -380,6 +393,7 @@ static const struct sun6i_rtc_clk_data > > sun50i_h6_rtc_data = { > > .has_out_clk = 1, > > .has_losc_en = 1, > > .has_auto_swt = 1, > > + .has_auto_cal = 1, > > }; > > > > static void __init sun50i_h6_rtc_clk_init(struct device_node *node) > > I tried to test this with my H700 board but this struct is not actually in any mainline kernel and looks like it is from an earlier version of the H616 RTC changes which was never actually merged? Yes, thanks for the heads up, Ryan. Is this some Armbian branch, by any chance? > Talked to Andre on IRC and he thinks the H6 and H616 RTC clocks *should* > be equivalent. Yeah, that was my first naive idea, but it's actually not true, and the increasing complexity of the RTC *clocks* were the trigger for the new driver. That doesn't mean that the IOSC calibration routine cannot be the same between the H6 and H616, though, but as it stands that would need to live in two drivers (see my other email). Cheers, Andre > > > @@ -395,6 +409,7 @@ static const struct sun6i_rtc_clk_data > > sun50i_h616_rtc_data = { > > .has_prescaler = 1, > > .has_out_clk = 1, > > .no_ext_losc = 1, > > + .has_auto_cal = 1, > > }; > > > > static void __init sun50i_h616_rtc_clk_init(struct device_node *node) > > -- > > 2.39.2 > > Regards, > > Ryan >
On Fri, 2024-05-03 at 12:13 +0100, Andre Przywara wrote: Hi Andre and Ryan, thanks for responding, and sorry. Andre's reply was somehow lost and I just found it on the archive mirror. So hopefully my reply is correct. And it is my first journey to do a patch. > On Thu, 2 May 2024 20:07:36 +0200 > Alois Fertl <a.fertl@t-online.de> wrote: > > Hi Alois, > > many thanks for taking care and sending a patch! I think this problem > is > plaguing some people. > > > I have a M98-8K PLUS Magcubic TV-Box based on the Allwinner H618 > > SOC. > > On board is a Sp6330 wifi/bt module that requires a 32kHz clock to > > operate correctly. Without this change the clock from the SOC is > > ~29kHz and BT module does not start up. The patch enables the > > Internal > > OSC Clock Auto Calibration of the H616/H618 which than provides the > > necessary 32kHz and the BT module initializes successfully. > > Add a flag and set it for H6 AND H616. The H618 is the same as H616 > > regarding rtc. > > (many thanks to Ryan for the head up on this) > > So what tree is this patch based on? It certainly is not mainline, is > it? I'm running armbian based on linux-kernel-worktree/6.7__sunxi64__arm64 which has one other rtc/rtc-sun6i.c patch applied before. "[PATCH] drv:rtc:sun6i: support RTCs without external LOSCs" > Mainline never supported the H616 RTC clocks via the RTC driver, this > was > only through the new driver in the clk > directory, in drivers/clk/sunxi-ng/ccu-sun6i-rtc.c: > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=8a93720329d4d2 > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d91612d7f01aca I have found some code in for calibration in the sunxi-ng directory that was never called and I doubt that it gives working calibration. Should a kernel then use exclusively use the one or the other driver? > > Please note that the H6 RTC clocks were intended to be handled via > the new > driver as well, but this part was reverted: > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=60d9f050da63b > > Please only send patches based on the mainline tree. > > I doesn't look that hard to adjust the patch to mainline, though to > cover > the H6 is well this would require this code in both drivers. So when > we > want to address the H6 as well, it might make sense to solve the > problem > that triggered the revert first, to only have that change only in one > file. Yes, the code would be useful for H6 when no external crystal is present. > > > Signed-off-by: Alois Fertl <a.fertl@t-online.de> > > --- > > > > v1->v2 > > - add flag and activate for H6 AND H616 > > > > v2->v3 > > - correct findings from review > > > > I was hoping to get some feedback regarding the situation on H6, > > where an external 32k crystal can be present. > > From what I understand from the H6 manual there should be no > > conflict as one can select which souce will drive the output. > > Should certainly be tested but I don`t have H6 hardware. > > I think I should have H6 boards both with and without an external > crystal > oscillator, so can check the situation there, but only next week. > > > drivers/rtc/rtc-sun6i.c | 17 ++++++++++++++++- > > > > 1 file changed, 16 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/rtc/rtc-sun6i.c b/drivers/rtc/rtc-sun6i.c > > index e0b85a0d5645..20e81ccdef29 100644 > > --- a/drivers/rtc/rtc-sun6i.c > > +++ b/drivers/rtc/rtc-sun6i.c > > @@ -42,6 +42,11 @@ > > > > #define SUN6I_LOSC_CLK_PRESCAL 0x0008 > > > > +#define SUN6I_LOSC_CLK_AUTO_CAL 0x000c > > +#define SUN6I_LOSC_CLK_AUTO_CAL_16MS BIT(2) > > +#define SUN6I_LOSC_CLK_AUTO_CAL_ENABLE BIT(1) > > +#define SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL BIT(0) > > + > > /* RTC */ > > #define SUN6I_RTC_YMD 0x0010 > > #define SUN6I_RTC_HMS 0x0014 > > @@ -126,7 +131,6 @@ > > * registers (R40, H6) > > * - SYS power domain controls (R40) > > * - DCXO controls (H6) > > - * - RC oscillator calibration (H6) > > * > > * These functions are not covered by this driver. > > */ > > @@ -138,6 +142,7 @@ struct sun6i_rtc_clk_data { > > unsigned int has_losc_en : 1; > > unsigned int has_auto_swt : 1; > > unsigned int no_ext_losc : 1; > > + unsigned int has_auto_cal : 1; > > }; > > > > #define RTC_LINEAR_DAY BIT(0) > > @@ -268,6 +273,14 @@ static void __init sun6i_rtc_clk_init(struct > > device_node *node, > > } > > writel(reg, rtc->base + SUN6I_LOSC_CTRL); > > > > + if (rtc->data->has_auto_cal) { > > + /* Enable internal OSC clock auto calibration */ > > + reg = SUN6I_LOSC_CLK_AUTO_CAL_16MS | > > + SUN6I_LOSC_CLK_AUTO_CAL_ENABLE | > > + SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL; > > + writel(reg, rtc->base + SUN6I_LOSC_CLK_AUTO_CAL); > > + } > > + > > /* Yes, I know, this is ugly. */ > > sun6i_rtc = rtc; > > > > @@ -380,6 +393,7 @@ static const struct sun6i_rtc_clk_data > > sun50i_h6_rtc_data = { > > .has_out_clk = 1, > > .has_losc_en = 1, > > .has_auto_swt = 1, > > + .has_auto_cal = 1, > > }; > > > > static void __init sun50i_h6_rtc_clk_init(struct device_node > > *node) > > @@ -395,6 +409,7 @@ static const struct sun6i_rtc_clk_data > > sun50i_h616_rtc_data = { > > For the records: this whole struct does not exist in mainline. > > Cheers, > Andre > > > .has_prescaler = 1, > > .has_out_clk = 1, > > .no_ext_losc = 1, > > + .has_auto_cal = 1, > > }; > > > > static void __init sun50i_h616_rtc_clk_init(struct device_node > > *node) > Cheers, Alois
diff --git a/drivers/rtc/rtc-sun6i.c b/drivers/rtc/rtc-sun6i.c index e0b85a0d5645..20e81ccdef29 100644 --- a/drivers/rtc/rtc-sun6i.c +++ b/drivers/rtc/rtc-sun6i.c @@ -42,6 +42,11 @@ #define SUN6I_LOSC_CLK_PRESCAL 0x0008 +#define SUN6I_LOSC_CLK_AUTO_CAL 0x000c +#define SUN6I_LOSC_CLK_AUTO_CAL_16MS BIT(2) +#define SUN6I_LOSC_CLK_AUTO_CAL_ENABLE BIT(1) +#define SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL BIT(0) + /* RTC */ #define SUN6I_RTC_YMD 0x0010 #define SUN6I_RTC_HMS 0x0014 @@ -126,7 +131,6 @@ * registers (R40, H6) * - SYS power domain controls (R40) * - DCXO controls (H6) - * - RC oscillator calibration (H6) * * These functions are not covered by this driver. */ @@ -138,6 +142,7 @@ struct sun6i_rtc_clk_data { unsigned int has_losc_en : 1; unsigned int has_auto_swt : 1; unsigned int no_ext_losc : 1; + unsigned int has_auto_cal : 1; }; #define RTC_LINEAR_DAY BIT(0) @@ -268,6 +273,14 @@ static void __init sun6i_rtc_clk_init(struct device_node *node, } writel(reg, rtc->base + SUN6I_LOSC_CTRL); + if (rtc->data->has_auto_cal) { + /* Enable internal OSC clock auto calibration */ + reg = SUN6I_LOSC_CLK_AUTO_CAL_16MS | + SUN6I_LOSC_CLK_AUTO_CAL_ENABLE | + SUN6I_LOSC_CLK_AUTO_CAL_SEL_CAL; + writel(reg, rtc->base + SUN6I_LOSC_CLK_AUTO_CAL); + } + /* Yes, I know, this is ugly. */ sun6i_rtc = rtc; @@ -380,6 +393,7 @@ static const struct sun6i_rtc_clk_data sun50i_h6_rtc_data = { .has_out_clk = 1, .has_losc_en = 1, .has_auto_swt = 1, + .has_auto_cal = 1, }; static void __init sun50i_h6_rtc_clk_init(struct device_node *node) @@ -395,6 +409,7 @@ static const struct sun6i_rtc_clk_data sun50i_h616_rtc_data = { .has_prescaler = 1, .has_out_clk = 1, .no_ext_losc = 1, + .has_auto_cal = 1, }; static void __init sun50i_h616_rtc_clk_init(struct device_node *node)
I have a M98-8K PLUS Magcubic TV-Box based on the Allwinner H618 SOC. On board is a Sp6330 wifi/bt module that requires a 32kHz clock to operate correctly. Without this change the clock from the SOC is ~29kHz and BT module does not start up. The patch enables the Internal OSC Clock Auto Calibration of the H616/H618 which than provides the necessary 32kHz and the BT module initializes successfully. Add a flag and set it for H6 AND H616. The H618 is the same as H616 regarding rtc. Signed-off-by: Alois Fertl <a.fertl@t-online.de> --- v1->v2 - add flag and activate for H6 AND H616 v2->v3 - correct findings from review I was hoping to get some feedback regarding the situation on H6, where an external 32k crystal can be present. From what I understand from the H6 manual there should be no conflict as one can select which souce will drive the output. Should certainly be tested but I don`t have H6 hardware. drivers/rtc/rtc-sun6i.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-)