[PATCH RFC u-boot-mvebu] arm: kirkwood: Move internal registers in arch_very_early_init() function

older
[PATCH u-boot-mvebu] doc/kwboot.1:...

Pali Rohár

11 Mar 2023 11 Mar '23

11:57 a.m.

Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org --- Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it. --- arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool + select ARCH_VERY_EARLY_INIT

config KW88F6281 bool + select ARCH_VERY_EARLY_INIT

config SHEEVA_88SV131 bool diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;

- /* Linux expects the internal registers to be at 0xf1000000 */ - writel(KW_REGS_PHY_BASE, KW_OFFSET_REG); - /* Enable and invalidate L2 cache in write through mode */ writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg); invalidate_l2_cache(); diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */ + +#include <config.h> +#include <linux/linkage.h> + +ENTRY(arch_very_early_init) + /* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */ + ldr r0, =KW_REGS_PHY_BASE + ldr r1, =KW_OFFSET_REG + str r0, [r1] + bx lr +ENDPROC(arch_very_early_init)

-- 2.20.1

Show replies by date

Tony Dinh

12 Mar 12 Mar

12:47 a.m.

Hi Pali,

On Sat, Mar 11, 2023 at 2:57 AM Pali Rohár pali@kernel.org wrote:

...

Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org

Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it.

I've run a regression test with this patch (Debug UART is not enabled). And everything was OK. No change in behavior.

However, when I turned on Debug UART on the nsa310s (88F6702) board, and ran with kwboot, it froze right away upon starting. Unrelated to this patch, I believe DEBUG_UART has been broken for Kirkwood lately, perhaps sometime late December to present, but I did not have time to track it down. Here is the last thread that I had Debug UART working on the Pogo V4 (88F6192): https://lists.denx.de/pipermail/u-boot/2022-December/502605.html

For reference, here is my local patch to configure the NSA310S, in addition to this patch.

diff --git a/configs/nsa310s_defconfig b/configs/nsa310s_defconfig index 76839e62dd..4bef35d576 100644 --- a/configs/nsa310s_defconfig +++ b/configs/nsa310s_defconfig @@ -15,8 +15,11 @@ CONFIG_ENV_SIZE=0x20000 CONFIG_ENV_OFFSET=0xE0000 CONFIG_DEFAULT_DEVICE_TREE="kirkwood-nsa310s" CONFIG_SYS_PROMPT="NSA310s> " +CONFIG_DEBUG_UART_BASE=0xf1012000 +CONFIG_DEBUG_UART_CLOCK=250000000 CONFIG_IDENT_STRING="\nZyXEL NSA310S/320S 1/2-Bay Power Media Server" CONFIG_SYS_LOAD_ADDR=0x800000 +CONFIG_DEBUG_UART=y CONFIG_DISTRO_DEFAULTS=y CONFIG_BOOTDELAY=3 CONFIG_USE_PREBOOT=y @@ -50,6 +53,7 @@ CONFIG_MTD_RAW_NAND=y CONFIG_PHY_MARVELL=y CONFIG_MVGBE=y CONFIG_MII=y +CONFIG_DEBUG_UART_ANNOUNCE=y CONFIG_USB=y CONFIG_USB_EHCI_HCD=y CONFIG_UBIFS_SILENCE_MSG=y

diff --git a/arch/arm/dts/kirkwood-nsa310s.dts b/arch/arm/dts/kirkwood-nsa310s.dts index 09ee76c2a2..2df45aa6da 100644 --- a/arch/arm/dts/kirkwood-nsa310s.dts +++ b/arch/arm/dts/kirkwood-nsa310s.dts @@ -22,6 +22,10 @@ reg = <0x00000000 0x10000000>; };

+ aliases { + serial0 = &uart0; + }; + chosen { bootargs = "console=ttyS0,115200"; stdout-path = &uart0; @@ -317,3 +321,8 @@ &pcie0 { status = "okay"; }; + +&uart0 { + status = "okay"; + u-boot,dm-pre-reloc; +};

If anybody is available to test this patch with Debug UART enabled, it would be great. I'll wait a few days and if there is no suggestion, I would do a bisect from Dec 19th.

Thanks, Tony

...

arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool
  select ARCH_VERY_EARLY_INIT
config KW88F6281 bool
  select ARCH_VERY_EARLY_INIT
config SHEEVA_88SV131 bool diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;
  /* Linux expects the internal registers to be at 0xf1000000 */
  writel(KW_REGS_PHY_BASE, KW_OFFSET_REG);
  /* Enable and invalidate L2 cache in write through mode */
  writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg);
  invalidate_l2_cache();
diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */

+#include <config.h> +#include <linux/linkage.h>

+ENTRY(arch_very_early_init)
  /* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */
  ldr     r0, =KW_REGS_PHY_BASE
  ldr     r1, =KW_OFFSET_REG
  str     r0, [r1]
  bx      lr
+ENDPROC(arch_very_early_init)

2.20.1

Pali Rohár

1:29 a.m.

On Saturday 11 March 2023 15:47:29 Tony Dinh wrote:

...

Hi Pali,

On Sat, Mar 11, 2023 at 2:57 AM Pali Rohár pali@kernel.org wrote:

...
Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org

Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it.

I've run a regression test with this patch (Debug UART is not enabled). And everything was OK. No change in behavior.

Ok! Thanks for testing.

...

However, when I turned on Debug UART on the nsa310s (88F6702) board, and ran with kwboot, it froze right away upon starting. Unrelated to this patch, I believe DEBUG_UART has been broken for Kirkwood lately, perhaps sometime late December to present, but I did not have time to track it down. Here is the last thread that I had Debug UART working on the Pogo V4 (88F6192): https://lists.denx.de/pipermail/u-boot/2022-December/502605.html

For reference, here is my local patch to configure the NSA310S, in addition to this patch.

I do not know what could broke it, but I see there two things which you could change in your configuration.

1. Try to set UART shift register to 2. Not sure what is default but in kirkwood.dtsi file it is 2.

CONFIG_DEBUG_UART_SHIFT=2

2. Recheck UART clock. In kw88f6281.h is defined that TCLK is either 166666667 or 200000000. And it is configured by strapping pins. TCLK cannot be 250000000 on 6281 for sure. I do not know to which clock is connected UART base clock, but my guess is that it is TCLK.

...

diff --git a/configs/nsa310s_defconfig b/configs/nsa310s_defconfig index 76839e62dd..4bef35d576 100644 --- a/configs/nsa310s_defconfig +++ b/configs/nsa310s_defconfig @@ -15,8 +15,11 @@ CONFIG_ENV_SIZE=0x20000 CONFIG_ENV_OFFSET=0xE0000 CONFIG_DEFAULT_DEVICE_TREE="kirkwood-nsa310s" CONFIG_SYS_PROMPT="NSA310s> " +CONFIG_DEBUG_UART_BASE=0xf1012000 +CONFIG_DEBUG_UART_CLOCK=250000000 CONFIG_IDENT_STRING="\nZyXEL NSA310S/320S 1/2-Bay Power Media Server" CONFIG_SYS_LOAD_ADDR=0x800000 +CONFIG_DEBUG_UART=y CONFIG_DISTRO_DEFAULTS=y CONFIG_BOOTDELAY=3 CONFIG_USE_PREBOOT=y @@ -50,6 +53,7 @@ CONFIG_MTD_RAW_NAND=y CONFIG_PHY_MARVELL=y CONFIG_MVGBE=y CONFIG_MII=y +CONFIG_DEBUG_UART_ANNOUNCE=y CONFIG_USB=y CONFIG_USB_EHCI_HCD=y CONFIG_UBIFS_SILENCE_MSG=y

diff --git a/arch/arm/dts/kirkwood-nsa310s.dts b/arch/arm/dts/kirkwood-nsa310s.dts index 09ee76c2a2..2df45aa6da 100644 --- a/arch/arm/dts/kirkwood-nsa310s.dts +++ b/arch/arm/dts/kirkwood-nsa310s.dts @@ -22,6 +22,10 @@ reg = <0x00000000 0x10000000>; };
  aliases {
          serial0 = &uart0;
  };
  chosen {
          bootargs = "console=ttyS0,115200";
          stdout-path = &uart0;
@@ -317,3 +321,8 @@ &pcie0 { status = "okay"; };

+&uart0 {
   status = "okay";
   u-boot,dm-pre-reloc;
+};

If anybody is available to test this patch with Debug UART enabled, it would be great. I'll wait a few days and if there is no suggestion, I would do a bisect from Dec 19th.

Thanks, Tony

...
arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool
  select ARCH_VERY_EARLY_INIT
config KW88F6281 bool
  select ARCH_VERY_EARLY_INIT
config SHEEVA_88SV131 bool diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;
  /* Linux expects the internal registers to be at 0xf1000000 */
  writel(KW_REGS_PHY_BASE, KW_OFFSET_REG);
  /* Enable and invalidate L2 cache in write through mode */
  writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg);
  invalidate_l2_cache();
diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */

+#include <config.h> +#include <linux/linkage.h>

+ENTRY(arch_very_early_init)
  /* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */
  ldr     r0, =KW_REGS_PHY_BASE
  ldr     r1, =KW_OFFSET_REG
  str     r0, [r1]
  bx      lr
+ENDPROC(arch_very_early_init)

2.20.1

Tony Dinh

10:30 p.m.

Hi Pali,

On Sat, Mar 11, 2023 at 4:29 PM Pali Rohár pali@kernel.org wrote:

...

On Saturday 11 March 2023 15:47:29 Tony Dinh wrote:

...
Hi Pali,

On Sat, Mar 11, 2023 at 2:57 AM Pali Rohár pali@kernel.org wrote:

...
Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org

Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it.

I've run a regression test with this patch (Debug UART is not enabled). And everything was OK. No change in behavior.

Ok! Thanks for testing.

...
However, when I turned on Debug UART on the nsa310s (88F6702) board, and ran with kwboot, it froze right away upon starting. Unrelated to this patch, I believe DEBUG_UART has been broken for Kirkwood lately, perhaps sometime late December to present, but I did not have time to track it down. Here is the last thread that I had Debug UART working on the Pogo V4 (88F6192): https://lists.denx.de/pipermail/u-boot/2022-December/502605.html

For reference, here is my local patch to configure the NSA310S, in addition to this patch.

I do not know what could broke it, but I see there two things which you could change in your configuration.

Try to set UART shift register to 2. Not sure what is default but in kirkwood.dtsi file it is 2.

CONFIG_DEBUG_UART_SHIFT=2

Recheck UART clock. In kw88f6281.h is defined that TCLK is either 166666667 or 200000000. And it is configured by strapping pins. TCLK cannot be 250000000 on 6281 for sure. I do not know to which clock is connected UART base clock, but my guess is that it is TCLK.

Thanks for the suggestion. Indeed, the combination of CONFIG_DEBUG_UART_CLOCK=166666667 and CONFIG_DEBUG_UART_SHIFT=2 works as it should (88F6702/88F6192 TCLK=166666667).

Without setting CONFIG_DEBUG_UART_SHIFT in defconfig, it is default to 0.

CONFIG_DEBUG_UART_SHIFT=2 is the key here, because U-Boot does not hang with CONFIG_DEBUG_UART_CLOCK=250000000. The board runs OK, but the Debug UART announcement output is gibberish.

So I guess back on the Dec 19th build, perhaps I was just lucky that some internal values were used for Kirkwood, but not CONFIG_xxx.

Thanks, Tony

...

...
diff --git a/configs/nsa310s_defconfig b/configs/nsa310s_defconfig index 76839e62dd..4bef35d576 100644 --- a/configs/nsa310s_defconfig +++ b/configs/nsa310s_defconfig @@ -15,8 +15,11 @@ CONFIG_ENV_SIZE=0x20000 CONFIG_ENV_OFFSET=0xE0000 CONFIG_DEFAULT_DEVICE_TREE="kirkwood-nsa310s" CONFIG_SYS_PROMPT="NSA310s> " +CONFIG_DEBUG_UART_BASE=0xf1012000 +CONFIG_DEBUG_UART_CLOCK=250000000 CONFIG_IDENT_STRING="\nZyXEL NSA310S/320S 1/2-Bay Power Media Server" CONFIG_SYS_LOAD_ADDR=0x800000 +CONFIG_DEBUG_UART=y CONFIG_DISTRO_DEFAULTS=y CONFIG_BOOTDELAY=3 CONFIG_USE_PREBOOT=y @@ -50,6 +53,7 @@ CONFIG_MTD_RAW_NAND=y CONFIG_PHY_MARVELL=y CONFIG_MVGBE=y CONFIG_MII=y +CONFIG_DEBUG_UART_ANNOUNCE=y CONFIG_USB=y CONFIG_USB_EHCI_HCD=y CONFIG_UBIFS_SILENCE_MSG=y

diff --git a/arch/arm/dts/kirkwood-nsa310s.dts b/arch/arm/dts/kirkwood-nsa310s.dts index 09ee76c2a2..2df45aa6da 100644 --- a/arch/arm/dts/kirkwood-nsa310s.dts +++ b/arch/arm/dts/kirkwood-nsa310s.dts @@ -22,6 +22,10 @@ reg = <0x00000000 0x10000000>; };
  aliases {
          serial0 = &uart0;
  };
  chosen {
          bootargs = "console=ttyS0,115200";
          stdout-path = &uart0;
@@ -317,3 +321,8 @@ &pcie0 { status = "okay"; };

+&uart0 {
   status = "okay";
   u-boot,dm-pre-reloc;
+};

If anybody is available to test this patch with Debug UART enabled, it would be great. I'll wait a few days and if there is no suggestion, I would do a bisect from Dec 19th.

Thanks, Tony

...
arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool
  select ARCH_VERY_EARLY_INIT
config KW88F6281 bool
  select ARCH_VERY_EARLY_INIT
config SHEEVA_88SV131 bool diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;
  /* Linux expects the internal registers to be at 0xf1000000 */
  writel(KW_REGS_PHY_BASE, KW_OFFSET_REG);
  /* Enable and invalidate L2 cache in write through mode */
  writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg);
  invalidate_l2_cache();
diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */

+#include <config.h> +#include <linux/linkage.h>

+ENTRY(arch_very_early_init)
  /* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */
  ldr     r0, =KW_REGS_PHY_BASE
  ldr     r1, =KW_OFFSET_REG
  str     r0, [r1]
  bx      lr
+ENDPROC(arch_very_early_init)

2.20.1

Pali Rohár

10:38 p.m.

On Sunday 12 March 2023 14:30:29 Tony Dinh wrote:

...

Hi Pali,

On Sat, Mar 11, 2023 at 4:29 PM Pali Rohár pali@kernel.org wrote:

...
On Saturday 11 March 2023 15:47:29 Tony Dinh wrote:

...
Hi Pali,

On Sat, Mar 11, 2023 at 2:57 AM Pali Rohár pali@kernel.org wrote:

...
Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org

Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it.

I've run a regression test with this patch (Debug UART is not enabled). And everything was OK. No change in behavior.

Ok! Thanks for testing.

...
However, when I turned on Debug UART on the nsa310s (88F6702) board, and ran with kwboot, it froze right away upon starting. Unrelated to this patch, I believe DEBUG_UART has been broken for Kirkwood lately, perhaps sometime late December to present, but I did not have time to track it down. Here is the last thread that I had Debug UART working on the Pogo V4 (88F6192): https://lists.denx.de/pipermail/u-boot/2022-December/502605.html

For reference, here is my local patch to configure the NSA310S, in addition to this patch.

I do not know what could broke it, but I see there two things which you could change in your configuration.

Try to set UART shift register to 2. Not sure what is default but in kirkwood.dtsi file it is 2.

CONFIG_DEBUG_UART_SHIFT=2

Recheck UART clock. In kw88f6281.h is defined that TCLK is either 166666667 or 200000000. And it is configured by strapping pins. TCLK cannot be 250000000 on 6281 for sure. I do not know to which clock is connected UART base clock, but my guess is that it is TCLK.

Thanks for the suggestion. Indeed, the combination of CONFIG_DEBUG_UART_CLOCK=166666667 and CONFIG_DEBUG_UART_SHIFT=2 works as it should (88F6702/88F6192 TCLK=166666667).

Without setting CONFIG_DEBUG_UART_SHIFT in defconfig, it is default to 0.

CONFIG_DEBUG_UART_SHIFT=2 is the key here, because U-Boot does not hang with CONFIG_DEBUG_UART_CLOCK=250000000. The board runs OK, but the Debug UART announcement output is gibberish.

Perfect! So ideally send a patch with adding those options into nsa310s_defconfig. I guess that CONFIG_DEBUG_UART_ANNOUNCE is not needed.

...

So I guess back on the Dec 19th build, perhaps I was just lucky that some internal values were used for Kirkwood, but not CONFIG_xxx.

Thanks, Tony

...
...
diff --git a/configs/nsa310s_defconfig b/configs/nsa310s_defconfig index 76839e62dd..4bef35d576 100644 --- a/configs/nsa310s_defconfig +++ b/configs/nsa310s_defconfig @@ -15,8 +15,11 @@ CONFIG_ENV_SIZE=0x20000 CONFIG_ENV_OFFSET=0xE0000 CONFIG_DEFAULT_DEVICE_TREE="kirkwood-nsa310s" CONFIG_SYS_PROMPT="NSA310s> " +CONFIG_DEBUG_UART_BASE=0xf1012000 +CONFIG_DEBUG_UART_CLOCK=250000000 CONFIG_IDENT_STRING="\nZyXEL NSA310S/320S 1/2-Bay Power Media Server" CONFIG_SYS_LOAD_ADDR=0x800000 +CONFIG_DEBUG_UART=y CONFIG_DISTRO_DEFAULTS=y CONFIG_BOOTDELAY=3 CONFIG_USE_PREBOOT=y @@ -50,6 +53,7 @@ CONFIG_MTD_RAW_NAND=y CONFIG_PHY_MARVELL=y CONFIG_MVGBE=y CONFIG_MII=y +CONFIG_DEBUG_UART_ANNOUNCE=y CONFIG_USB=y CONFIG_USB_EHCI_HCD=y CONFIG_UBIFS_SILENCE_MSG=y

diff --git a/arch/arm/dts/kirkwood-nsa310s.dts b/arch/arm/dts/kirkwood-nsa310s.dts index 09ee76c2a2..2df45aa6da 100644 --- a/arch/arm/dts/kirkwood-nsa310s.dts +++ b/arch/arm/dts/kirkwood-nsa310s.dts @@ -22,6 +22,10 @@ reg = <0x00000000 0x10000000>; };
  aliases {
          serial0 = &uart0;
  };
  chosen {
          bootargs = "console=ttyS0,115200";
          stdout-path = &uart0;
@@ -317,3 +321,8 @@ &pcie0 { status = "okay"; };

+&uart0 {
   status = "okay";
   u-boot,dm-pre-reloc;
+};

If anybody is available to test this patch with Debug UART enabled, it would be great. I'll wait a few days and if there is no suggestion, I would do a bisect from Dec 19th.

Thanks, Tony

...
arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool
  select ARCH_VERY_EARLY_INIT
config KW88F6281 bool
  select ARCH_VERY_EARLY_INIT
config SHEEVA_88SV131 bool diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;
  /* Linux expects the internal registers to be at 0xf1000000 */
  writel(KW_REGS_PHY_BASE, KW_OFFSET_REG);
  /* Enable and invalidate L2 cache in write through mode */
  writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg);
  invalidate_l2_cache();
diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */

+#include <config.h> +#include <linux/linkage.h>

+ENTRY(arch_very_early_init)
  /* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */
  ldr     r0, =KW_REGS_PHY_BASE
  ldr     r1, =KW_OFFSET_REG
  str     r0, [r1]
  bx      lr
+ENDPROC(arch_very_early_init)

2.20.1

Tony Dinh

13 Mar 13 Mar

1:12 a.m.

Hi Pali,

On Sun, Mar 12, 2023 at 2:38 PM Pali Rohár pali@kernel.org wrote:

...

On Sunday 12 March 2023 14:30:29 Tony Dinh wrote:

...
Hi Pali,

On Sat, Mar 11, 2023 at 4:29 PM Pali Rohár pali@kernel.org wrote:

...
On Saturday 11 March 2023 15:47:29 Tony Dinh wrote:

...
Hi Pali,

On Sat, Mar 11, 2023 at 2:57 AM Pali Rohár pali@kernel.org wrote:

...
Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org

Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it.

I've run a regression test with this patch (Debug UART is not enabled). And everything was OK. No change in behavior.

Ok! Thanks for testing.

...
However, when I turned on Debug UART on the nsa310s (88F6702) board, and ran with kwboot, it froze right away upon starting. Unrelated to this patch, I believe DEBUG_UART has been broken for Kirkwood lately, perhaps sometime late December to present, but I did not have time to track it down. Here is the last thread that I had Debug UART working on the Pogo V4 (88F6192): https://lists.denx.de/pipermail/u-boot/2022-December/502605.html

For reference, here is my local patch to configure the NSA310S, in addition to this patch.

I do not know what could broke it, but I see there two things which you could change in your configuration.

Try to set UART shift register to 2. Not sure what is default but in kirkwood.dtsi file it is 2.

CONFIG_DEBUG_UART_SHIFT=2

Recheck UART clock. In kw88f6281.h is defined that TCLK is either 166666667 or 200000000. And it is configured by strapping pins. TCLK cannot be 250000000 on 6281 for sure. I do not know to which clock is connected UART base clock, but my guess is that it is TCLK.

Thanks for the suggestion. Indeed, the combination of CONFIG_DEBUG_UART_CLOCK=166666667 and CONFIG_DEBUG_UART_SHIFT=2 works as it should (88F6702/88F6192 TCLK=166666667).

Without setting CONFIG_DEBUG_UART_SHIFT in defconfig, it is default to 0.

CONFIG_DEBUG_UART_SHIFT=2 is the key here, because U-Boot does not hang with CONFIG_DEBUG_UART_CLOCK=250000000. The board runs OK, but the Debug UART announcement output is gibberish.

Perfect! So ideally send a patch with adding those options into nsa310s_defconfig. I guess that CONFIG_DEBUG_UART_ANNOUNCE is not needed.

Sure, I will do that. In the meantime, I ran some more tests and here are some interesting results (thanks to Debug UART working now).

1. I've confirmed that, at the moment, we need your patch to see DEBUG_UART early output for Kirkwood. Without it, with DEBUG_UART configured correctly, I don't see any output on this board. It will run fine but behaving like there is no Debug UART enabled.

2. There is no need for the u-boot,dm-pre-reloc tag on Kirkwood boards. The serial-uclass function serial_check_stdout() binds it anyway ff the console is not marked to be bound before relocation: https://github.com/u-boot/u-boot/blob/master/drivers/serial/serial-uclass.c#...

3. The previous problem we saw with the Pogo V4 (Kirwood 6192) DM serial might be related to all this. Hopefully with your patch I can track it down further.

Thanks, Tony

...

...
So I guess back on the Dec 19th build, perhaps I was just lucky that some internal values were used for Kirkwood, but not CONFIG_xxx.

Thanks, Tony

...
...
diff --git a/configs/nsa310s_defconfig b/configs/nsa310s_defconfig index 76839e62dd..4bef35d576 100644 --- a/configs/nsa310s_defconfig +++ b/configs/nsa310s_defconfig @@ -15,8 +15,11 @@ CONFIG_ENV_SIZE=0x20000 CONFIG_ENV_OFFSET=0xE0000 CONFIG_DEFAULT_DEVICE_TREE="kirkwood-nsa310s" CONFIG_SYS_PROMPT="NSA310s> " +CONFIG_DEBUG_UART_BASE=0xf1012000 +CONFIG_DEBUG_UART_CLOCK=250000000 CONFIG_IDENT_STRING="\nZyXEL NSA310S/320S 1/2-Bay Power Media Server" CONFIG_SYS_LOAD_ADDR=0x800000 +CONFIG_DEBUG_UART=y CONFIG_DISTRO_DEFAULTS=y CONFIG_BOOTDELAY=3 CONFIG_USE_PREBOOT=y @@ -50,6 +53,7 @@ CONFIG_MTD_RAW_NAND=y CONFIG_PHY_MARVELL=y CONFIG_MVGBE=y CONFIG_MII=y +CONFIG_DEBUG_UART_ANNOUNCE=y CONFIG_USB=y CONFIG_USB_EHCI_HCD=y CONFIG_UBIFS_SILENCE_MSG=y

diff --git a/arch/arm/dts/kirkwood-nsa310s.dts b/arch/arm/dts/kirkwood-nsa310s.dts index 09ee76c2a2..2df45aa6da 100644 --- a/arch/arm/dts/kirkwood-nsa310s.dts +++ b/arch/arm/dts/kirkwood-nsa310s.dts @@ -22,6 +22,10 @@ reg = <0x00000000 0x10000000>; };
  aliases {
          serial0 = &uart0;
  };
  chosen {
          bootargs = "console=ttyS0,115200";
          stdout-path = &uart0;
@@ -317,3 +321,8 @@ &pcie0 { status = "okay"; };

+&uart0 {
   status = "okay";
   u-boot,dm-pre-reloc;
+};

If anybody is available to test this patch with Debug UART enabled, it would be great. I'll wait a few days and if there is no suggestion, I would do a bisect from Dec 19th.

Thanks, Tony

...
arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool
  select ARCH_VERY_EARLY_INIT
config KW88F6281 bool
  select ARCH_VERY_EARLY_INIT
config SHEEVA_88SV131 bool diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;
  /* Linux expects the internal registers to be at 0xf1000000 */
  writel(KW_REGS_PHY_BASE, KW_OFFSET_REG);
  /* Enable and invalidate L2 cache in write through mode */
  writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg);
  invalidate_l2_cache();
diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */

+#include <config.h> +#include <linux/linkage.h>

+ENTRY(arch_very_early_init)
  /* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */
  ldr     r0, =KW_REGS_PHY_BASE
  ldr     r1, =KW_OFFSET_REG
  str     r0, [r1]
  bx      lr
+ENDPROC(arch_very_early_init)

2.20.1

Tony Dinh

14 Mar 14 Mar

7:02 a.m.

On Sun, Mar 12, 2023 at 5:12 PM Tony Dinh mibodhi@gmail.com wrote:

...

Hi Pali,

On Sun, Mar 12, 2023 at 2:38 PM Pali Rohár pali@kernel.org wrote:

...
On Sunday 12 March 2023 14:30:29 Tony Dinh wrote:

...
Hi Pali,

On Sat, Mar 11, 2023 at 4:29 PM Pali Rohár pali@kernel.org wrote:

...
On Saturday 11 March 2023 15:47:29 Tony Dinh wrote:

...
Hi Pali,

On Sat, Mar 11, 2023 at 2:57 AM Pali Rohár pali@kernel.org wrote:

...
Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org

Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it.

I've run a regression test with this patch (Debug UART is not enabled). And everything was OK. No change in behavior.

Ok! Thanks for testing.

...
However, when I turned on Debug UART on the nsa310s (88F6702) board, and ran with kwboot, it froze right away upon starting. Unrelated to this patch, I believe DEBUG_UART has been broken for Kirkwood lately, perhaps sometime late December to present, but I did not have time to track it down. Here is the last thread that I had Debug UART working on the Pogo V4 (88F6192): https://lists.denx.de/pipermail/u-boot/2022-December/502605.html

For reference, here is my local patch to configure the NSA310S, in addition to this patch.

I do not know what could broke it, but I see there two things which you could change in your configuration.

Try to set UART shift register to 2. Not sure what is default but in kirkwood.dtsi file it is 2.

CONFIG_DEBUG_UART_SHIFT=2

Recheck UART clock. In kw88f6281.h is defined that TCLK is either 166666667 or 200000000. And it is configured by strapping pins. TCLK cannot be 250000000 on 6281 for sure. I do not know to which clock is connected UART base clock, but my guess is that it is TCLK.

Thanks for the suggestion. Indeed, the combination of CONFIG_DEBUG_UART_CLOCK=166666667 and CONFIG_DEBUG_UART_SHIFT=2 works as it should (88F6702/88F6192 TCLK=166666667).

Without setting CONFIG_DEBUG_UART_SHIFT in defconfig, it is default to 0.

CONFIG_DEBUG_UART_SHIFT=2 is the key here, because U-Boot does not hang with CONFIG_DEBUG_UART_CLOCK=250000000. The board runs OK, but the Debug UART announcement output is gibberish.

Perfect! So ideally send a patch with adding those options into nsa310s_defconfig. I guess that CONFIG_DEBUG_UART_ANNOUNCE is not needed.

Sure, I will do that. In the meantime, I ran some more tests and here are some interesting results (thanks to Debug UART working now).

I've confirmed that, at the moment, we need your patch to see

DEBUG_UART early output for Kirkwood. Without it, with DEBUG_UART configured correctly, I don't see any output on this board. It will run fine but behaving like there is no Debug UART enabled.

There is no need for the u-boot,dm-pre-reloc tag on Kirkwood

boards. The serial-uclass function serial_check_stdout() binds it anyway ff the console is not marked to be bound before relocation: https://github.com/u-boot/u-boot/blob/master/drivers/serial/serial-uclass.c#...

The previous problem we saw with the Pogo V4 (Kirwood 6192) DM

serial might be related to all this. Hopefully with your patch I can track it down further.

Thanks, Tony

...
...
So I guess back on the Dec 19th build, perhaps I was just lucky that some internal values were used for Kirkwood, but not CONFIG_xxx.

Thanks, Tony

...
...
diff --git a/configs/nsa310s_defconfig b/configs/nsa310s_defconfig index 76839e62dd..4bef35d576 100644 --- a/configs/nsa310s_defconfig +++ b/configs/nsa310s_defconfig @@ -15,8 +15,11 @@ CONFIG_ENV_SIZE=0x20000 CONFIG_ENV_OFFSET=0xE0000 CONFIG_DEFAULT_DEVICE_TREE="kirkwood-nsa310s" CONFIG_SYS_PROMPT="NSA310s> " +CONFIG_DEBUG_UART_BASE=0xf1012000 +CONFIG_DEBUG_UART_CLOCK=250000000 CONFIG_IDENT_STRING="\nZyXEL NSA310S/320S 1/2-Bay Power Media Server" CONFIG_SYS_LOAD_ADDR=0x800000 +CONFIG_DEBUG_UART=y CONFIG_DISTRO_DEFAULTS=y CONFIG_BOOTDELAY=3 CONFIG_USE_PREBOOT=y @@ -50,6 +53,7 @@ CONFIG_MTD_RAW_NAND=y CONFIG_PHY_MARVELL=y CONFIG_MVGBE=y CONFIG_MII=y +CONFIG_DEBUG_UART_ANNOUNCE=y CONFIG_USB=y CONFIG_USB_EHCI_HCD=y CONFIG_UBIFS_SILENCE_MSG=y

diff --git a/arch/arm/dts/kirkwood-nsa310s.dts b/arch/arm/dts/kirkwood-nsa310s.dts index 09ee76c2a2..2df45aa6da 100644 --- a/arch/arm/dts/kirkwood-nsa310s.dts +++ b/arch/arm/dts/kirkwood-nsa310s.dts @@ -22,6 +22,10 @@ reg = <0x00000000 0x10000000>; };
  aliases {
          serial0 = &uart0;
  };
  chosen {
          bootargs = "console=ttyS0,115200";
          stdout-path = &uart0;
@@ -317,3 +321,8 @@ &pcie0 { status = "okay"; };

+&uart0 {
   status = "okay";
   u-boot,dm-pre-reloc;
+};

If anybody is available to test this patch with Debug UART enabled, it would be great. I'll wait a few days and if there is no suggestion, I would do a bisect from Dec 19th.

Thanks, Tony

...
arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool
  select ARCH_VERY_EARLY_INIT
config KW88F6281 bool
  select ARCH_VERY_EARLY_INIT
config SHEEVA_88SV131 bool diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;
  /* Linux expects the internal registers to be at 0xf1000000 */
  writel(KW_REGS_PHY_BASE, KW_OFFSET_REG);
  /* Enable and invalidate L2 cache in write through mode */
  writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg);
  invalidate_l2_cache();
diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */

+#include <config.h> +#include <linux/linkage.h>

+ENTRY(arch_very_early_init)
  /* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */
  ldr     r0, =KW_REGS_PHY_BASE
  ldr     r1, =KW_OFFSET_REG
  str     r0, [r1]
  bx      lr
+ENDPROC(arch_very_early_init)

2.20.1

All checked out OK on these Kirkwood boards:

Zyxel NSA310S (88F6702) Cloudengine Pogo V4 (88F6192) Seagate Goflex Home (88F6281)

Tested-by: Tony Dinh mibodhi@gmail.com

Thanks, Tony

Stefan Roese

16 Mar 16 Mar

10:04 a.m.

On 3/11/23 11:57, Pali Rohár wrote:

...

Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org

Reviewed-by: Stefan Roese sr@denx.de

Thanks, Stefan

...

Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it.

arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool

select ARCH_VERY_EARLY_INIT

config KW88F6281 bool

select ARCH_VERY_EARLY_INIT

config SHEEVA_88SV131 bool

diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;

/* Linux expects the internal registers to be at 0xf1000000 */

writel(KW_REGS_PHY_BASE, KW_OFFSET_REG);

/* Enable and invalidate L2 cache in write through mode */ writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg); invalidate_l2_cache();

diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */

+#include <config.h> +#include <linux/linkage.h>

+ENTRY(arch_very_early_init)

/* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */

ldr r0, =KW_REGS_PHY_BASE

ldr r1, =KW_OFFSET_REG

str r0, [r1]

bx lr

+ENDPROC(arch_very_early_init)

Viele Grüße, Stefan Roese

-- DENX Software Engineering GmbH, Managing Director: Erika Unter HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-51 Fax: (+49)-8142-66989-80 Email: sr@denx.de

Stefan Roese

24 Mar 24 Mar

3:42 p.m.

On 3/11/23 11:57, Pali Rohár wrote:

...

Same change as was done for mvebu in commit 5bb2c550b11e ("arm: mvebu: Move internal registers in arch_very_early_init() function") but for kirkwood.

Signed-off-by: Pali Rohár pali@kernel.org

Applied to u-boot-marvell/next

Thanks, Stefan

...

Hello! Please test this RFC patch on more Kirkwood boards if there is any issue with it.

arch/arm/mach-kirkwood/Kconfig | 2 ++ arch/arm/mach-kirkwood/Makefile | 1 + arch/arm/mach-kirkwood/cpu.c | 3 --- arch/arm/mach-kirkwood/lowlevel.S | 12 ++++++++++++ 4 files changed, 15 insertions(+), 3 deletions(-) create mode 100644 arch/arm/mach-kirkwood/lowlevel.S

diff --git a/arch/arm/mach-kirkwood/Kconfig b/arch/arm/mach-kirkwood/Kconfig index c8a193dd4cdf..ba39e9ae416e 100644 --- a/arch/arm/mach-kirkwood/Kconfig +++ b/arch/arm/mach-kirkwood/Kconfig @@ -5,9 +5,11 @@ config FEROCEON_88FR131

config KW88F6192 bool

select ARCH_VERY_EARLY_INIT

config KW88F6281 bool

select ARCH_VERY_EARLY_INIT

config SHEEVA_88SV131 bool

diff --git a/arch/arm/mach-kirkwood/Makefile b/arch/arm/mach-kirkwood/Makefile index 3b2eef8d5419..0fb5a2326f5f 100644 --- a/arch/arm/mach-kirkwood/Makefile +++ b/arch/arm/mach-kirkwood/Makefile @@ -6,6 +6,7 @@

obj-y = cpu.o obj-y += cache.o +obj-y += lowlevel.o obj-y += mpp.o

# cpu.o and cache.o contain CP15 instructions which cannot be run in diff --git a/arch/arm/mach-kirkwood/cpu.c b/arch/arm/mach-kirkwood/cpu.c index df3e8f11782a..2b493b36c20d 100644 --- a/arch/arm/mach-kirkwood/cpu.c +++ b/arch/arm/mach-kirkwood/cpu.c @@ -189,9 +189,6 @@ int arch_cpu_init(void) struct kwcpu_registers *cpureg = (struct kwcpu_registers *)KW_CPU_REG_BASE;

/* Linux expects the internal registers to be at 0xf1000000 */

writel(KW_REGS_PHY_BASE, KW_OFFSET_REG);

/* Enable and invalidate L2 cache in write through mode */ writel(readl(&cpureg->l2_cfg) | 0x18, &cpureg->l2_cfg); invalidate_l2_cache();

diff --git a/arch/arm/mach-kirkwood/lowlevel.S b/arch/arm/mach-kirkwood/lowlevel.S new file mode 100644 index 000000000000..3b339f97f056 --- /dev/null +++ b/arch/arm/mach-kirkwood/lowlevel.S @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0+ */

+#include <config.h> +#include <linux/linkage.h>

+ENTRY(arch_very_early_init)

/* Move internal registers from KW_OFFSET_REG to KW_REGS_PHY_BASE */

ldr r0, =KW_REGS_PHY_BASE

ldr r1, =KW_OFFSET_REG

str r0, [r1]

bx lr

+ENDPROC(arch_very_early_init)

Viele Grüße, Stefan Roese

780

Age (days ago)

793

Last active (days ago)

List overview

Download

8 comments

3 participants

tags (0)

participants (3)

Pali Rohár
Stefan Roese
Tony Dinh