ARM: shmobile: r8a7779 CCF DTS update

Update the r8a7779 CCF DTS with the following fixes:
- Use MSTP0 for SCIF clock control
- Use R8A7779_CLK_P as parent clock for SCIF (same as legacy code)
- Use "clock-indicies" instead of "renesas,clock-indices"

Signed-off-by Magnus Damm <[email protected]>

---

Written on top of renesas-devel-20141202-v3.18-rc7

 arch/arm/boot/dts/r8a7779.dtsi |   30 +++++++++++++++---------------
 1 file changed, 15 insertions(+), 15 deletions(-)

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 renamed from arch/arm/boot/dts/r8a7779.dtsi +0 -2

@@ -464,18 +464,18 @@
                                 <&cpg_clocks R8A7779_CLK_P>,
                                 <&cpg_clocks R8A7779_CLK_S>,
                                 <&cpg_clocks R8A7779_CLK_S>,
-                                <&cpg_clocks R8A7779_CLK_S1>,
-                                <&cpg_clocks R8A7779_CLK_S1>,

Hi Magnus,

Quoted 4 lines · Magnus Damm

Update the r8a7779 CCF DTS with the following fixes:
- Use MSTP0 for SCIF clock control
- Use R8A7779_CLK_P as parent clock for SCIF (same as legacy code)
- Use "clock-indicies" instead of "renesas,clock-indices"

We already have
"[PATCH v2 3/6] ARM: shmobile: r8a7779 dtsi: Change to using clock-indices"
(http://www.spinics.net/lists/linux-sh/msg37285.html) for the latter.

Quoted 8 lines

Signed-off-by Magnus Damm <[email protected]>

---

Written on top of renesas-devel-20141202-v3.18-rc7

 arch/arm/boot/dts/r8a7779.dtsi |   30 +++++++++++++++---------------
 1 file changed, 15 insertions(+), 15 deletions(-)

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 line 206

Quoted 11 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 renamed from arch/arm/boot/dts/r8a7779.dtsi +1 -1

@@ -200,7 +200,7 @@
                compatible = "renesas,scif-r8a7779", "renesas,scif";
                reg = <0xffe40000 0x100>;
                interrupts = <0 88 IRQ_TYPE_LEVEL_HIGH>;
-               clocks = <&cpg_clocks R8A7779_CLK_P>;
+               clocks = <&mstp0_clks R8A7779_CLK_SCIF0>;
                clock-names = "sci_ick";
                status = "disabled";
        };

According to the datasheet, the SCIF can use both S1 and "SCIF_CLK" (I assume that's the MSTP clock output for SCIF?) as clock input, selectable using the XIN bit of the Clock Select Register (CKS).
Do you have more information?

Quoted 6 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 renamed from arch/arm/boot/dts/r8a7779.dtsi +0 -2

@@ -464,18 +464,18 @@
                                 <&cpg_clocks R8A7779_CLK_P>,
                                 <&cpg_clocks R8A7779_CLK_S>,
                                 <&cpg_clocks R8A7779_CLK_S>,
-                                <&cpg_clocks R8A7779_CLK_S1>,
-                                <&cpg_clocks R8A7779_CLK_S1>,

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 line 472

Quoted 10 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 renamed from arch/arm/boot/dts/r8a7779.dtsi +6 -4

-                                <&cpg_clocks R8A7779_CLK_S1>,
-                                <&cpg_clocks R8A7779_CLK_S1>,
-                                <&cpg_clocks R8A7779_CLK_S1>,
-                                <&cpg_clocks R8A7779_CLK_S1>,
+                                <&cpg_clocks R8A7779_CLK_P>,
+                                <&cpg_clocks R8A7779_CLK_P>,
+                                <&cpg_clocks R8A7779_CLK_P>,
+                                <&cpg_clocks R8A7779_CLK_P>,
+                                <&cpg_clocks R8A7779_CLK_P>,
+                                <&cpg_clocks R8A7779_CLK_P>,

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

Quoted 6 lines · Geert Uytterhoeven

Quoted 4 lines · Magnus Damm

Update the r8a7779 CCF DTS with the following fixes:
- Use MSTP0 for SCIF clock control
- Use R8A7779_CLK_P as parent clock for SCIF (same as legacy code)
- Use "clock-indicies" instead of "renesas,clock-indices"

Quoted 3 lines

We already have
"[PATCH v2 3/6] ARM: shmobile: r8a7779 dtsi: Change to using clock-indices"
(http://www.spinics.net/lists/linux-sh/msg37285.html) for the latter.

Yes, I noticed that too late I'm afraid. =)

Are you aware of any outstanding issue for that series?

Quoted 8 lines

Quoted 8 lines

Signed-off-by Magnus Damm <[email protected]>

---

Written on top of renesas-devel-20141202-v3.18-rc7

 arch/arm/boot/dts/r8a7779.dtsi |   30 +++++++++++++++---------------
 1 file changed, 15 insertions(+), 15 deletions(-)

Quoted 16 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 line 206

Quoted 11 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 renamed from arch/arm/boot/dts/r8a7779.dtsi +1 -1

@@ -200,7 +200,7 @@
                compatible = "renesas,scif-r8a7779", "renesas,scif";
                reg = <0xffe40000 0x100>;
                interrupts = <0 88 IRQ_TYPE_LEVEL_HIGH>;
-               clocks = <&cpg_clocks R8A7779_CLK_P>;
+               clocks = <&mstp0_clks R8A7779_CLK_SCIF0>;
                clock-names = "sci_ick";
                status = "disabled";
        };

According to the datasheet, the SCIF can use both S1 and "SCIF_CLK" (I assume that's the MSTP clock output for SCIF?) as clock input, selectable using the XIN bit of the Clock Select Register (CKS).
Do you have more information?

I suspect that you may look at the left side of the r8a7779 SCIF overview page near the BRG where there is a "clks1" and the external SCIF_CLK. I'm looking at the right side and the "baud rate generator" where clkp is hooked up.

The SCIF driver today has relatively limited support for the BRG (which confusingly enough is a second baud rate generator, how many do one need?). With the "regular" baud rate generator and the BRG unit some of the SCIF variants can be configured to use many different clocks.

We currently lack code for the following:
A) Describe all the clocks hooked up to the SCIF
B) Select the best clock for any given baud rate

So with this patch we simply ignore the BRG.

Hi Magnus,

Quoted 11 lines · Magnus Damm

Quoted 9 lines · Geert Uytterhoeven

Quoted 4 lines · Magnus Damm

Update the r8a7779 CCF DTS with the following fixes:
- Use MSTP0 for SCIF clock control
- Use R8A7779_CLK_P as parent clock for SCIF (same as legacy code)
- Use "clock-indicies" instead of "renesas,clock-indices"

We already have
"[PATCH v2 3/6] ARM: shmobile: r8a7779 dtsi: Change to using clock-indices"
(http://www.spinics.net/lists/linux-sh/msg37285.html) for the latter.

Quoted 3 lines

Yes, I noticed that too late I'm afraid. =)

Are you aware of any outstanding issue for that series?

Not that I'm aware of. As Mike merged the bindings update, I had pinged Simon about the rest of the series, just before you sent your patch.

Quoted 25 lines

Quoted 24 lines

Quoted 11 lines

Signed-off-by Magnus Damm <[email protected]>

---

Written on top of renesas-devel-20141202-v3.18-rc7

 arch/arm/boot/dts/r8a7779.dtsi |   30 +++++++++++++++---------------
 1 file changed, 15 insertions(+), 15 deletions(-)

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 line 206

Quoted 8 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 renamed from arch/arm/boot/dts/r8a7779.dtsi +1 -1

@@ -200,7 +200,7 @@
                compatible = "renesas,scif-r8a7779", "renesas,scif";
                reg = <0xffe40000 0x100>;
                interrupts = <0 88 IRQ_TYPE_LEVEL_HIGH>;
-               clocks = <&cpg_clocks R8A7779_CLK_P>;
+               clocks = <&mstp0_clks R8A7779_CLK_SCIF0>;
                clock-names = "sci_ick";
                status = "disabled";
        };

According to the datasheet, the SCIF can use both S1 and "SCIF_CLK" (I assume that's the MSTP clock output for SCIF?) as clock input, selectable using the XIN bit of the Clock Select Register (CKS).
Do you have more information?

Quoted 4 lines

I suspect that you may look at the left side of the r8a7779 SCIF overview page near the BRG where there is a "clks1" and the external SCIF_CLK. I'm looking at the right side and the "baud rate generator" where clkp is hooked up.

Oh right, that "clkp" is really "clkp through MSTP bit X"?

Quoted 10 lines

The SCIF driver today has relatively limited support for the BRG (which confusingly enough is a second baud rate generator, how many do one need?). With the "regular" baud rate generator and the BRG unit some of the SCIF variants can be configured to use many different clocks.

We currently lack code for the following:
A) Describe all the clocks hooked up to the SCIF
B) Select the best clock for any given baud rate

Quoted 1 line

So with this patch we simply ignore the BRG.

OK. Fair enough.

Acked-by Geert Uytterhoeven <[email protected]>

Hi Geert,

Quoted 16 lines · Geert Uytterhoeven

Hi Magnus,

Quoted 13 lines · Magnus Damm

Quoted 9 lines · Geert Uytterhoeven

Quoted 4 lines · Magnus Damm

Update the r8a7779 CCF DTS with the following fixes:
- Use MSTP0 for SCIF clock control
- Use R8A7779_CLK_P as parent clock for SCIF (same as legacy code)
- Use "clock-indicies" instead of "renesas,clock-indices"

We already have
"[PATCH v2 3/6] ARM: shmobile: r8a7779 dtsi: Change to using clock-indices"
(http://www.spinics.net/lists/linux-sh/msg37285.html) for the latter.

Yes, I noticed that too late I'm afraid. =)

Quoted 4 lines

Quoted 1 line

Are you aware of any outstanding issue for that series?

Not that I'm aware of. As Mike merged the bindings update, I had pinged Simon about the rest of the series, just before you sent your patch.

Thanks, sounds like I should make a V2 of this patch once Simon has merged patch 3/6 above.

Quoted 26 lines

Quoted 26 lines

Quoted 25 lines

Quoted 12 lines

Signed-off-by Magnus Damm <[email protected]>

---

Written on top of renesas-devel-20141202-v3.18-rc7

 arch/arm/boot/dts/r8a7779.dtsi |   30 +++++++++++++++---------------
 1 file changed, 15 insertions(+), 15 deletions(-)

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 line 206

Quoted 8 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 +0900 renamed from arch/arm/boot/dts/r8a7779.dtsi +1 -1

@@ -200,7 +200,7 @@
                compatible = "renesas,scif-r8a7779", "renesas,scif";
                reg = <0xffe40000 0x100>;
                interrupts = <0 88 IRQ_TYPE_LEVEL_HIGH>;
-               clocks = <&cpg_clocks R8A7779_CLK_P>;
+               clocks = <&mstp0_clks R8A7779_CLK_SCIF0>;
                clock-names = "sci_ick";
                status = "disabled";
        };

According to the datasheet, the SCIF can use both S1 and "SCIF_CLK" (I assume that's the MSTP clock output for SCIF?) as clock input, selectable using the XIN bit of the Clock Select Register (CKS).
Do you have more information?

Quoted 6 lines

Quoted 4 lines

I suspect that you may look at the left side of the r8a7779 SCIF overview page near the BRG where there is a "clks1" and the external SCIF_CLK. I'm looking at the right side and the "baud rate generator" where clkp is hooked up.

Oh right, that "clkp" is really "clkp through MSTP bit X"?

Yes, this is how we historically have hooked up the SCIF clocks. It is however not entirely clear from the documentation how the MSTP bit is connected, so the assumption may be wrong. And the SCIF driver has support for interface clocks as well. The "clkp through MSTP bit X" implementation seems better than the current code at least.

Quoted 10 lines

Quoted 10 lines

The SCIF driver today has relatively limited support for the BRG (which confusingly enough is a second baud rate generator, how many do one need?). With the "regular" baud rate generator and the BRG unit some of the SCIF variants can be configured to use many different clocks.

We currently lack code for the following:
A) Describe all the clocks hooked up to the SCIF
B) Select the best clock for any given baud rate

Quoted 5 lines

Quoted 1 line

So with this patch we simply ignore the BRG.

OK. Fair enough.

Acked-by Geert Uytterhoeven <[email protected]>

Thanks!

/ magnus

Quoted 24 lines · Magnus Damm

Hi Geert,

Quoted 20 lines · Geert Uytterhoeven

Hi Magnus,

Quoted 7 lines · Magnus Damm

Quoted 6 lines · Geert Uytterhoeven

Quoted 4 lines · Magnus Damm

Update the r8a7779 CCF DTS with the following fixes:
- Use MSTP0 for SCIF clock control
- Use R8A7779_CLK_P as parent clock for SCIF (same as legacy code)
- Use "clock-indicies" instead of "renesas,clock-indices"

Quoted 7 lines

Quoted 3 lines

We already have
"[PATCH v2 3/6] ARM: shmobile: r8a7779 dtsi: Change to using clock-indices"
(http://www.spinics.net/lists/linux-sh/msg37285.html) for the latter.

Yes, I noticed that too late I'm afraid. =)

Are you aware of any outstanding issue for that series?

Not that I'm aware of. As Mike merged the bindings update, I had pinged Simon about the rest of the series, just before you sent your patch.

Quoted 2 lines

Thanks, sounds like I should make a V2 of this patch once Simon has merged patch 3/6 above.

In general I prefer one fix per patch.

So unless there is a dependency between the remaining two fixes could you please make them separate patches when you drop the clock-indicies changes?

Hi Simon,

Quoted 28 lines · Simon Horman

Quoted 26 lines · Magnus Damm

Hi Geert,

Quoted 14 lines · Geert Uytterhoeven

Hi Magnus,

Quoted 11 lines · Magnus Damm

Quoted 9 lines · Geert Uytterhoeven

Quoted 4 lines · Magnus Damm

Update the r8a7779 CCF DTS with the following fixes:
- Use MSTP0 for SCIF clock control
- Use R8A7779_CLK_P as parent clock for SCIF (same as legacy code)
- Use "clock-indicies" instead of "renesas,clock-indices"

We already have
"[PATCH v2 3/6] ARM: shmobile: r8a7779 dtsi: Change to using clock-indices"
(http://www.spinics.net/lists/linux-sh/msg37285.html) for the latter.

Quoted 6 lines

Quoted 3 lines

Yes, I noticed that too late I'm afraid. =)

Are you aware of any outstanding issue for that series?

Not that I'm aware of. As Mike merged the bindings update, I had pinged Simon about the rest of the series, just before you sent your patch.

Thanks, sounds like I should make a V2 of this patch once Simon has merged patch 3/6 above.

Quoted 5 lines

In general I prefer one fix per patch.

So unless there is a dependency between the remaining two fixes could you please make them separate patches when you drop the clock-indicies changes?

I think both are needed to make it work, but I don't mind breaking them out into two separate commits. I agree that it is cleaner.

Quoted 13 lines · Geert Uytterhoeven

Quoted 12 lines · Magnus Damm

Quoted 10 lines · Geert Uytterhoeven

Quoted 4 lines · Magnus Damm

Update the r8a7779 CCF DTS with the following fixes:
- Use MSTP0 for SCIF clock control
- Use R8A7779_CLK_P as parent clock for SCIF (same as legacy code)
- Use "clock-indicies" instead of "renesas,clock-indices"

We already have
"[PATCH v2 3/6] ARM: shmobile: r8a7779 dtsi: Change to using clock-indices"
(http://www.spinics.net/lists/linux-sh/msg37285.html) for the latter.

Quoted 6 lines

Quoted 3 lines

Yes, I noticed that too late I'm afraid. =)

Are you aware of any outstanding issue for that series?

Not that I'm aware of. As Mike merged the bindings update, I had pinged Simon about the rest of the series, just before you sent your patch.

Thanks, sounds like I should make a V2 of this patch once Simon has merged patch 3/6 above.

Quoted 31 lines

Quoted 30 lines

Quoted 17 lines

Quoted 17 lines

Signed-off-by Magnus Damm <[email protected]>

---

Written on top of renesas-devel-20141202-v3.18-rc7

 arch/arm/boot/dts/r8a7779.dtsi |   30 +++++++++++++++---------------
 1 file changed, 15 insertions(+), 15 deletions(-)

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 renamed from arch/arm/boot/dts/r8a7779.dtsi +2 -1

+0900

@@ -200,7 +200,7 @@
                compatible = "renesas,scif-r8a7779", "renesas,scif";
                reg = <0xffe40000 0x100>;
                interrupts = <0 88 IRQ_TYPE_LEVEL_HIGH>;
-               clocks = <&cpg_clocks R8A7779_CLK_P>;
+               clocks = <&mstp0_clks R8A7779_CLK_SCIF0>;

Quoted 8 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 line 206

Quoted 3 lines

arch/arm/boot/dts/r8a7779.dtsi 2014-12-03 20:22:26.000000000 renamed from arch/arm/boot/dts/r8a7779.dtsi +0 -0

                clock-names = "sci_ick";
                status = "disabled";
        };

According to the datasheet, the SCIF can use both S1 and "SCIF_CLK" (I assume that's the MSTP clock output for SCIF?) as clock input, selectable using the XIN bit of the Clock Select Register (CKS). Do you have more information?

I suspect that you may look at the left side of the r8a7779 SCIF overview page near the BRG where there is a "clks1" and the external SCIF_CLK. I'm looking at the right side and the "baud rate generator" where clkp is hooked up.

Quoted 1 line

Oh right, that "clkp" is really "clkp through MSTP bit X"?

Yes, this is how we historically have hooked up the SCIF clocks. It is however not entirely clear from the documentation how the MSTP bit is connected, so the assumption may be wrong. And the SCIF driver has support for interface clocks as well.

The "clkp through MSTP bit X" implementation seems better than the current code at least.

Quoted 15 lines

Quoted 11 lines

The SCIF driver today has relatively limited support for the BRG (which confusingly enough is a second baud rate generator, how many do one need?). With the "regular" baud rate generator and the BRG unit some of the SCIF variants can be configured to use many different clocks.

We currently lack code for the following:
A) Describe all the clocks hooked up to the SCIF
B) Select the best clock for any given baud rate

So with this patch we simply ignore the BRG.

OK. Fair enough.

Acked-by Geert Uytterhoeven <[email protected]>