Skip to contentSkip to conversation

arm: kirkwood: 88f6281: Detect CONFIG_SYS_TCLK from SAR register

PATCH In review Pali Rohár 3 participants 10 messages Aug 16, 2022 – Aug 17, 2022

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 arm: kirkwood: 88f6281: Detect CONFIG_SYS_TCLK from SAR register

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz. This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] - https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] - https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] - https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

---
 arch/arm/mach-kirkwood/include/mach/kw88f6281.h | 3 ++-
 arch/arm/mach-kirkwood/include/mach/soc.h       | 2 ++
 2 files changed, 4 insertions(+), 1 deletion(-)
arch/arm/mach-kirkwood/include/mach/kw88f6281.h +2 -1
index 87406081cf54..f86cd0bb6013 100644
@@ -15,6 +15,7 @@
 #define KW_REGS_PHY_BASE		KW88F6281_REGS_PHYS_BASE
 
 /* TCLK Core Clock definition */
-#define CONFIG_SYS_TCLK	200000000 /* 200MHz */
+#define CONFIG_SYS_TCLK			((readl(CONFIG_SAR_REG) & BIT(21)) ? \
+					166666667 : 200000000)
 
 #endif /* _ASM_ARCH_KW88F6281_H */
arch/arm/mach-kirkwood/include/mach/soc.h +2 -0
index 1d7f2828cd38..5f545c6f4349 100644
@@ -62,6 +62,8 @@
 #define MVCPU_WIN_ENABLE	KWCPU_WIN_ENABLE
 #define MVCPU_WIN_DISABLE	KWCPU_WIN_DISABLE
 
+#define CONFIG_SAR_REG		(KW_MPP_BASE + 0x0030)
+
 #if defined (CONFIG_KW88F6281)
 #include <asm/arch/kw88f6281.h>
 #elif defined (CONFIG_KW88F6192)
Signature · 1 line
2.20.1
#1 lore raw

Am 2022-08-16 22:00, schrieb Pali Rohár:

Quoted 19 lines

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz. This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] -
https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] -
https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Quoted 4 lines
Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

Thanks, but this doesn't compile:

In file included from lib/time.c:19:
lib/time.c: In function ‘timer_get_boot_us’:
./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:50:5: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   50 | #if CONFIG_SYS_TIMER_RATE == 1000000
      |     ^~~~~~~~~~~~~~~~~~~~~

./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:52:7: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   52 | #elif CONFIG_SYS_TIMER_RATE > 1000000
      |       ^~~~~~~~~~~~~~~~~~~~~

make[1]: *** [scripts/Makefile.build:257: lib/time.o] Fehler 1

-michael

#2 lore raw
Quoted 55 lines · On Tuesday 16 August 2022 23:34:13 Michael Walle wrote:

Am 2022-08-16 22:00, schrieb Pali Rohár:

Quoted 20 lines

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz.
This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] -
https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] -
https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Quoted 4 lines
Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

Thanks, but this doesn't compile:

In file included from lib/time.c:19:
lib/time.c: In function ‘timer_get_boot_us’:
./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:50:5: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   50 | #if CONFIG_SYS_TIMER_RATE == 1000000
      |     ^~~~~~~~~~~~~~~~~~~~~

./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

Quoted 8 lines
   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:52:7: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   52 | #elif CONFIG_SYS_TIMER_RATE > 1000000
      |       ^~~~~~~~~~~~~~~~~~~~~

make[1]: *** [scripts/Makefile.build:257: lib/time.o] Fehler 1

-michael

Ah :-( And why it works on other Armada mvebu boards? Because of this:

$ git grep CONFIG_SYS_TIMER_RATE arch/arm/mach-mvebu
arch/arm/mach-mvebu/include/mach/config.h:#define CONFIG_SYS_TIMER_RATE         25000000

I have looked how is this option used and it is unit for CONFIG_SYS_TIMER_COUNTER option. On all kirkwood and mvebu platforms is that register set to SoC Global Timer 0 which on Armada 38x uses either fixed 25MHz clock or NBCLK clock, together with SoC Global Timer 0 Ratio. When NBCLK then the frequency of that timer register is:
NBCLK / 2 / (2 ^ ratio). Default ratio is 0 (but can be configured up to 7). By default fixed 25MHz clock is used.

Hence for mvebu it is OK as U-Boot does not change default values of SoC Global Timer 0.

For kirkwood it looks like that those timers are based on TCLK (page 283 in Functional Specifications).

How to fix it? I do not know right now. I have just quirk & dirty solution by moving code from preprocessor to compiler:

lib/time.c +9 -7
index 96074b84af69..f2ffc0498d39 100644
@@ -46,13 +46,15 @@ unsigned long notrace timer_read_counter(void)
 ulong timer_get_boot_us(void)
 {
 	ulong count = timer_read_counter();
-
-#if CONFIG_SYS_TIMER_RATE == 1000000
-	return count;
-#elif CONFIG_SYS_TIMER_RATE > 1000000
-	return lldiv(count, CONFIG_SYS_TIMER_RATE / 1000000);
-#elif defined(CONFIG_SYS_TIMER_RATE)
-	return (unsigned long long)count * 1000000 / CONFIG_SYS_TIMER_RATE;
+#ifdef CONFIG_SYS_TIMER_RATE
+	const ulong timer_rate = CONFIG_SYS_TIMER_RATE;
+
+	if (timer_rate)
+		return count;
+	else if (timer_rate > 1000000)
+		return lldiv(count, timer_rate / 1000000);
+	else
+		return (unsigned long long)count * 1000000 / timer_rate;
 #else
 	/* Assume the counter is in microseconds */
 	return count;

At least it could be used for verifying if CONFIG_SYS_TCLK is working.

#3 lore raw
Quoted 96 lines · On Wednesday 17 August 2022 00:14:20 Pali Rohár wrote:
Quoted 55 lines · On Tuesday 16 August 2022 23:34:13 Michael Walle wrote:

Am 2022-08-16 22:00, schrieb Pali Rohár:

Quoted 20 lines

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz.
This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] -
https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] -
https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Quoted 4 lines
Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

Thanks, but this doesn't compile:

In file included from lib/time.c:19:
lib/time.c: In function ‘timer_get_boot_us’:
./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:50:5: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   50 | #if CONFIG_SYS_TIMER_RATE == 1000000
      |     ^~~~~~~~~~~~~~~~~~~~~

./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

Quoted 8 lines
   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:52:7: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   52 | #elif CONFIG_SYS_TIMER_RATE > 1000000
      |       ^~~~~~~~~~~~~~~~~~~~~

make[1]: *** [scripts/Makefile.build:257: lib/time.o] Fehler 1

-michael

Ah :-( And why it works on other Armada mvebu boards? Because of this:

$ git grep CONFIG_SYS_TIMER_RATE arch/arm/mach-mvebu
arch/arm/mach-mvebu/include/mach/config.h:#define CONFIG_SYS_TIMER_RATE         25000000

I have looked how is this option used and it is unit for CONFIG_SYS_TIMER_COUNTER option. On all kirkwood and mvebu platforms is that register set to SoC Global Timer 0 which on Armada 38x uses either fixed 25MHz clock or NBCLK clock, together with SoC Global Timer 0 Ratio. When NBCLK then the frequency of that timer register is:
NBCLK / 2 / (2 ^ ratio). Default ratio is 0 (but can be configured up to 7). By default fixed 25MHz clock is used.

Hence for mvebu it is OK as U-Boot does not change default values of SoC Global Timer 0.

For kirkwood it looks like that those timers are based on TCLK (page 283 in Functional Specifications).

How to fix it? I do not know right now. I have just quirk & dirty solution by moving code from preprocessor to compiler:

lib/time.c +0 -1
index 96074b84af69..f2ffc0498d39 100644
@@ -46,13 +46,15 @@ unsigned long notrace timer_read_counter(void)
 ulong timer_get_boot_us(void)
 {
 	ulong count = timer_read_counter();
-

lib/time.c unsigned long notrace timer_read_counter(void) · line 52

Quoted 10 lines
lib/time.c +4 -6
-#if CONFIG_SYS_TIMER_RATE == 1000000
-	return count;
-#elif CONFIG_SYS_TIMER_RATE > 1000000
-	return lldiv(count, CONFIG_SYS_TIMER_RATE / 1000000);
-#elif defined(CONFIG_SYS_TIMER_RATE)
-	return (unsigned long long)count * 1000000 / CONFIG_SYS_TIMER_RATE;
+#ifdef CONFIG_SYS_TIMER_RATE
+	const ulong timer_rate = CONFIG_SYS_TIMER_RATE;
+
+	if (timer_rate)
            ~~~~~~~~~~

Here needs to be if (timer_rate == 1000000)

Quoted 10 lines
lib/time.c +5 -0
+		return count;
+	else if (timer_rate > 1000000)
+		return lldiv(count, timer_rate / 1000000);
+	else
+		return (unsigned long long)count * 1000000 / timer_rate;
 #else
 	/* Assume the counter is in microseconds */
 	return count;

At least it could be used for verifying if CONFIG_SYS_TCLK is working.

Anyway, it looks like that this timer code is deprecated and boards should migrate to use DM.

So maybe the proper fix for this would be to add kirkwood dm driver?

#4 lore raw

Am 2022-08-17 00:39, schrieb Pali Rohár:

Quoted 119 lines · On Wednesday 17 August 2022 00:14:20 Pali Rohár wrote:
Quoted 96 lines · On Tuesday 16 August 2022 23:34:13 Michael Walle wrote:
Quoted 55 lines

Am 2022-08-16 22:00, schrieb Pali Rohár:

Quoted 20 lines

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz.
This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] -
https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] -
https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Quoted 4 lines
Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

Thanks, but this doesn't compile:

In file included from lib/time.c:19:
lib/time.c: In function ‘timer_get_boot_us’:
./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:50:5: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   50 | #if CONFIG_SYS_TIMER_RATE == 1000000
      |     ^~~~~~~~~~~~~~~~~~~~~

./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

Quoted 8 lines
   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:52:7: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   52 | #elif CONFIG_SYS_TIMER_RATE > 1000000
      |       ^~~~~~~~~~~~~~~~~~~~~

make[1]: *** [scripts/Makefile.build:257: lib/time.o] Fehler 1

-michael

Ah :-( And why it works on other Armada mvebu boards? Because of this:

$ git grep CONFIG_SYS_TIMER_RATE arch/arm/mach-mvebu
arch/arm/mach-mvebu/include/mach/config.h:#define CONFIG_SYS_TIMER_RATE         25000000

I have looked how is this option used and it is unit for CONFIG_SYS_TIMER_COUNTER option. On all kirkwood and mvebu platforms is that register set to SoC Global Timer 0 which on Armada 38x uses either fixed 25MHz clock or NBCLK clock, together with SoC Global Timer 0 Ratio. When NBCLK then the frequency of that timer register is:
NBCLK / 2 / (2 ^ ratio). Default ratio is 0 (but can be configured up to 7). By default fixed 25MHz clock is used.

Hence for mvebu it is OK as U-Boot does not change default values of SoC Global Timer 0.

For kirkwood it looks like that those timers are based on TCLK (page 283 in Functional Specifications).

How to fix it? I do not know right now. I have just quirk & dirty solution by moving code from preprocessor to compiler:

lib/time.c +0 -1
index 96074b84af69..f2ffc0498d39 100644
@@ -46,13 +46,15 @@ unsigned long notrace timer_read_counter(void)
 ulong timer_get_boot_us(void)
 {
 	ulong count = timer_read_counter();
-

lib/time.c unsigned long notrace timer_read_counter(void) · line 52

Quoted 10 lines
lib/time.c +4 -6
-#if CONFIG_SYS_TIMER_RATE == 1000000
-	return count;
-#elif CONFIG_SYS_TIMER_RATE > 1000000
-	return lldiv(count, CONFIG_SYS_TIMER_RATE / 1000000);
-#elif defined(CONFIG_SYS_TIMER_RATE)
-	return (unsigned long long)count * 1000000 / CONFIG_SYS_TIMER_RATE;
+#ifdef CONFIG_SYS_TIMER_RATE
+	const ulong timer_rate = CONFIG_SYS_TIMER_RATE;
+
+	if (timer_rate)
            ~~~~~~~~~~

Here needs to be if (timer_rate == 1000000)

Quoted 9 lines
lib/time.c +5 -0
+		return count;
+	else if (timer_rate > 1000000)
+		return lldiv(count, timer_rate / 1000000);
+	else
+		return (unsigned long long)count * 1000000 / timer_rate;
 #else
 	/* Assume the counter is in microseconds */
 	return count;
Quoted 1 line
Quoted 1 line

At least it could be used for verifying if CONFIG_SYS_TCLK is working.

FWIW, this is working.

Quoted 4 lines

Anyway, it looks like that this timer code is deprecated and boards should migrate to use DM.

So maybe the proper fix for this would be to add kirkwood dm driver?

Eventually sure, but for now I'd like to have a working board without spending too much time on fixing this. I'd propose to revert the broken commit for now and then, we can work on a proper timer driver.

But I can already say that the u-boot binary is almost at its size limit. I've looked into CONFIG_TIMER and it seems that we'd also need the whole clock infrastructure (and DM_CLK). No?

-michael

#5 lore raw
Quoted 135 lines · On Wednesday 17 August 2022 01:23:34 Michael Walle wrote:

Am 2022-08-17 00:39, schrieb Pali Rohár:

Quoted 125 lines · On Wednesday 17 August 2022 00:14:20 Pali Rohár wrote:
Quoted 102 lines · On Tuesday 16 August 2022 23:34:13 Michael Walle wrote:
Quoted 55 lines

Am 2022-08-16 22:00, schrieb Pali Rohár:

Quoted 20 lines

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz.
This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] -
https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] -
https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Quoted 4 lines
Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

Thanks, but this doesn't compile:

In file included from lib/time.c:19:
lib/time.c: In function ‘timer_get_boot_us’:
./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:50:5: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   50 | #if CONFIG_SYS_TIMER_RATE == 1000000
      |     ^~~~~~~~~~~~~~~~~~~~~

./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

Quoted 8 lines
   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:52:7: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   52 | #elif CONFIG_SYS_TIMER_RATE > 1000000
      |       ^~~~~~~~~~~~~~~~~~~~~

make[1]: *** [scripts/Makefile.build:257: lib/time.o] Fehler 1

-michael

Ah :-( And why it works on other Armada mvebu boards? Because of this:

$ git grep CONFIG_SYS_TIMER_RATE arch/arm/mach-mvebu
arch/arm/mach-mvebu/include/mach/config.h:#define
CONFIG_SYS_TIMER_RATE         25000000

I have looked how is this option used and it is unit for CONFIG_SYS_TIMER_COUNTER option. On all kirkwood and mvebu platforms is
that register set to SoC Global Timer 0 which on Armada 38x uses either
fixed 25MHz clock or NBCLK clock, together with SoC Global Timer 0 Ratio. When NBCLK then the frequency of that timer register is:
NBCLK / 2 / (2 ^ ratio). Default ratio is 0 (but can be configured up to
7). By default fixed 25MHz clock is used.

Hence for mvebu it is OK as U-Boot does not change default values of SoC
Global Timer 0.

For kirkwood it looks like that those timers are based on TCLK (page 283
in Functional Specifications).

How to fix it? I do not know right now. I have just quirk & dirty solution by moving code from preprocessor to compiler:

lib/time.c +0 -1
index 96074b84af69..f2ffc0498d39 100644
@@ -46,13 +46,15 @@ unsigned long notrace timer_read_counter(void)
 ulong timer_get_boot_us(void)
 {
 	ulong count = timer_read_counter();
-

lib/time.c unsigned long notrace timer_read_counter(void) · line 52

Quoted 10 lines
lib/time.c +4 -6
-#if CONFIG_SYS_TIMER_RATE == 1000000
-	return count;
-#elif CONFIG_SYS_TIMER_RATE > 1000000
-	return lldiv(count, CONFIG_SYS_TIMER_RATE / 1000000);
-#elif defined(CONFIG_SYS_TIMER_RATE)
-	return (unsigned long long)count * 1000000 / CONFIG_SYS_TIMER_RATE;
+#ifdef CONFIG_SYS_TIMER_RATE
+	const ulong timer_rate = CONFIG_SYS_TIMER_RATE;
+
+	if (timer_rate)
            ~~~~~~~~~~

Here needs to be if (timer_rate == 1000000)

Quoted 9 lines
lib/time.c +5 -0
+		return count;
+	else if (timer_rate > 1000000)
+		return lldiv(count, timer_rate / 1000000);
+	else
+		return (unsigned long long)count * 1000000 / timer_rate;
 #else
 	/* Assume the counter is in microseconds */
 	return count;
Quoted 1 line
Quoted 1 line

At least it could be used for verifying if CONFIG_SYS_TCLK is working.

FWIW, this is working.

Quoted 4 lines

Anyway, it looks like that this timer code is deprecated and boards should migrate to use DM.

So maybe the proper fix for this would be to add kirkwood dm driver?

Quoted 7 lines

Eventually sure, but for now I'd like to have a working board without spending too much time on fixing this. I'd propose to revert the broken commit for now and then, we can work on a proper timer driver.

But I can already say that the u-boot binary is almost at its size limit. I've looked into CONFIG_TIMER and it seems that we'd also need the whole clock infrastructure (and DM_CLK). No?

So, for now I see only two options:

1) Either apply that change which moves preprocessor macros to C code and apply this one fix.

2) Or revert that my kirkwood commit 8ac303d49f89 and do not apply this one fix.

Stefan, what do you prefer?

#6 lore raw

Am 2022-08-17 10:36, schrieb Pali Rohár:

Quoted 144 lines · On Wednesday 17 August 2022 01:23:34 Michael Walle wrote:
Quoted 135 lines

Am 2022-08-17 00:39, schrieb Pali Rohár:

Quoted 125 lines · On Wednesday 17 August 2022 00:14:20 Pali Rohár wrote:
Quoted 102 lines · On Tuesday 16 August 2022 23:34:13 Michael Walle wrote:
Quoted 55 lines

Am 2022-08-16 22:00, schrieb Pali Rohár:

Quoted 20 lines

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz.
This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] -
https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] -
https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Quoted 4 lines
Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

Thanks, but this doesn't compile:

In file included from lib/time.c:19:
lib/time.c: In function ‘timer_get_boot_us’:
./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:50:5: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   50 | #if CONFIG_SYS_TIMER_RATE == 1000000
      |     ^~~~~~~~~~~~~~~~~~~~~

./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’
18 | #define CONFIG_SYS_TCLK ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
| ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro ‘CONFIG_SYS_TCLK’

Quoted 8 lines
   56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
      |                                ^~~~~~~~~~~~~~~
lib/time.c:52:7: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
   52 | #elif CONFIG_SYS_TIMER_RATE > 1000000
      |       ^~~~~~~~~~~~~~~~~~~~~

make[1]: *** [scripts/Makefile.build:257: lib/time.o] Fehler 1

-michael

Ah :-( And why it works on other Armada mvebu boards? Because of this:

$ git grep CONFIG_SYS_TIMER_RATE arch/arm/mach-mvebu
arch/arm/mach-mvebu/include/mach/config.h:#define
CONFIG_SYS_TIMER_RATE         25000000

I have looked how is this option used and it is unit for CONFIG_SYS_TIMER_COUNTER option. On all kirkwood and mvebu platforms is
that register set to SoC Global Timer 0 which on Armada 38x uses either
fixed 25MHz clock or NBCLK clock, together with SoC Global Timer 0 Ratio. When NBCLK then the frequency of that timer register is:
NBCLK / 2 / (2 ^ ratio). Default ratio is 0 (but can be configured up to
7). By default fixed 25MHz clock is used.

Hence for mvebu it is OK as U-Boot does not change default values of SoC
Global Timer 0.

For kirkwood it looks like that those timers are based on TCLK (page 283
in Functional Specifications).

How to fix it? I do not know right now. I have just quirk & dirty solution by moving code from preprocessor to compiler:

lib/time.c +0 -1
index 96074b84af69..f2ffc0498d39 100644
@@ -46,13 +46,15 @@ unsigned long notrace timer_read_counter(void)
 ulong timer_get_boot_us(void)
 {
 	ulong count = timer_read_counter();
-

lib/time.c unsigned long notrace timer_read_counter(void) · line 52

Quoted 10 lines
lib/time.c +4 -6
-#if CONFIG_SYS_TIMER_RATE == 1000000
-	return count;
-#elif CONFIG_SYS_TIMER_RATE > 1000000
-	return lldiv(count, CONFIG_SYS_TIMER_RATE / 1000000);
-#elif defined(CONFIG_SYS_TIMER_RATE)
-	return (unsigned long long)count * 1000000 / CONFIG_SYS_TIMER_RATE;
+#ifdef CONFIG_SYS_TIMER_RATE
+	const ulong timer_rate = CONFIG_SYS_TIMER_RATE;
+
+	if (timer_rate)
            ~~~~~~~~~~

Here needs to be if (timer_rate == 1000000)

Quoted 9 lines
lib/time.c +5 -0
+		return count;
+	else if (timer_rate > 1000000)
+		return lldiv(count, timer_rate / 1000000);
+	else
+		return (unsigned long long)count * 1000000 / timer_rate;
 #else
 	/* Assume the counter is in microseconds */
 	return count;
Quoted 1 line
Quoted 1 line

At least it could be used for verifying if CONFIG_SYS_TCLK is working.

FWIW, this is working.

Quoted 4 lines

Anyway, it looks like that this timer code is deprecated and boards should migrate to use DM.

So maybe the proper fix for this would be to add kirkwood dm driver?

Quoted 7 lines

Eventually sure, but for now I'd like to have a working board without spending too much time on fixing this. I'd propose to revert the broken commit for now and then, we can work on a proper timer driver.

But I can already say that the u-boot binary is almost at its size limit. I've looked into CONFIG_TIMER and it seems that we'd also need the whole clock infrastructure (and DM_CLK). No?

Quoted 8 lines

So, for now I see only two options:

1) Either apply that change which moves preprocessor macros to C code and apply this one fix.

2) Or revert that my kirkwood commit 8ac303d49f89 and do not apply this one fix.

Stefan, what do you prefer?

3) There is a chance that the timer code is actually pretty straight-forward and we don't need DM_CLK.

FWIW I'm working on 3) at the moment. So if that works out, we can use CONFIG_TIMER on kirkwood and then apply your initial patch. As there are no bugfix releases for u-boot anyway, we don't have to care for a backport fix.

-michael

#7 lore raw

Am 2022-08-17 10:43, schrieb Michael Walle:

Quoted 7 lines

3) There is a chance that the timer code is actually pretty straight-forward and we don't need DM_CLK.

FWIW I'm working on 3) at the moment. So if that works out, we can use CONFIG_TIMER on kirkwood and then apply your initial patch. As there are no bugfix releases for u-boot anyway, we don't have to care for a backport fix.

Ah damn, we cannot do this depending on wether CONFIG_TIMER is set or not. E.g. on a per-device basis. Or we do some ugly stuff as

#ifdef CONFIG_TIMER

#define CONFIG_SYS_TCLK                        ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
                                       166666667 : 200000000)

#else
#define CONFIG_SYS_TCLK 200000000
#endif

So it seems we are back to 1) or 2) :(

-michael

#8 lore raw
In reply to Pali Rohár

Hi Pali,

Quoted 144 lines · Pali Rohár
Quoted 135 lines · On Wednesday 17 August 2022 01:23:34 Michael Walle wrote:

Am 2022-08-17 00:39, schrieb Pali Rohár:

Quoted 125 lines · On Wednesday 17 August 2022 00:14:20 Pali Rohár wrote:
Quoted 102 lines · On Tuesday 16 August 2022 23:34:13 Michael Walle wrote:
Quoted 60 lines

Am 2022-08-16 22:00, schrieb Pali Rohár:

Quoted 20 lines

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz.
This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] -
https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] -
https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Quoted 4 lines
Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

Thanks, but this doesn't compile:

In file included from lib/time.c:19:
lib/time.c: In function ‘timer_get_boot_us’:
./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’

    18 | #define CONFIG_SYS_TCLK   ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
       |                             ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro
‘CONFIG_SYS_TCLK’
    56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
       |                                ^~~~~~~~~~~~~~~
lib/time.c:50:5: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
    50 | #if CONFIG_SYS_TIMER_RATE == 1000000
       |     ^~~~~~~~~~~~~~~~~~~~~

./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’

    18 | #define CONFIG_SYS_TCLK   ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
       |                             ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro
‘CONFIG_SYS_TCLK’
    56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
       |                                ^~~~~~~~~~~~~~~
lib/time.c:52:7: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
    52 | #elif CONFIG_SYS_TIMER_RATE > 1000000
       |       ^~~~~~~~~~~~~~~~~~~~~
Quoted 3 lines

make[1]: *** [scripts/Makefile.build:257: lib/time.o] Fehler 1

-michael

Ah :-( And why it works on other Armada mvebu boards? Because of this:

$ git grep CONFIG_SYS_TIMER_RATE arch/arm/mach-mvebu
arch/arm/mach-mvebu/include/mach/config.h:#define
CONFIG_SYS_TIMER_RATE         25000000

I have looked how is this option used and it is unit for CONFIG_SYS_TIMER_COUNTER option. On all kirkwood and mvebu platforms is
that register set to SoC Global Timer 0 which on Armada 38x uses either
fixed 25MHz clock or NBCLK clock, together with SoC Global Timer 0 Ratio. When NBCLK then the frequency of that timer register is:
NBCLK / 2 / (2 ^ ratio). Default ratio is 0 (but can be configured up to
7). By default fixed 25MHz clock is used.

Hence for mvebu it is OK as U-Boot does not change default values of SoC
Global Timer 0.

For kirkwood it looks like that those timers are based on TCLK (page 283
in Functional Specifications).

How to fix it? I do not know right now. I have just quirk & dirty solution by moving code from preprocessor to compiler:

lib/time.c +0 -1
index 96074b84af69..f2ffc0498d39 100644
@@ -46,13 +46,15 @@ unsigned long notrace timer_read_counter(void)
  ulong timer_get_boot_us(void)
  {
  	ulong count = timer_read_counter();
-

lib/time.c unsigned long notrace timer_read_counter(void) · line 52

Quoted 10 lines
lib/time.c +4 -6
-#if CONFIG_SYS_TIMER_RATE == 1000000
-	return count;
-#elif CONFIG_SYS_TIMER_RATE > 1000000
-	return lldiv(count, CONFIG_SYS_TIMER_RATE / 1000000);
-#elif defined(CONFIG_SYS_TIMER_RATE)
-	return (unsigned long long)count * 1000000 / CONFIG_SYS_TIMER_RATE;
+#ifdef CONFIG_SYS_TIMER_RATE
+	const ulong timer_rate = CONFIG_SYS_TIMER_RATE;
+
+	if (timer_rate)
             ~~~~~~~~~~

Here needs to be if (timer_rate == 1000000)

Quoted 9 lines
lib/time.c +5 -0
+		return count;
+	else if (timer_rate > 1000000)
+		return lldiv(count, timer_rate / 1000000);
+	else
+		return (unsigned long long)count * 1000000 / timer_rate;
  #else
  	/* Assume the counter is in microseconds */
  	return count;
Quoted 1 line
Quoted 1 line

At least it could be used for verifying if CONFIG_SYS_TCLK is working.

FWIW, this is working.

Quoted 4 lines

Anyway, it looks like that this timer code is deprecated and boards should migrate to use DM.

So maybe the proper fix for this would be to add kirkwood dm driver?

Quoted 7 lines

Eventually sure, but for now I'd like to have a working board without spending too much time on fixing this. I'd propose to revert the broken commit for now and then, we can work on a proper timer driver.

But I can already say that the u-boot binary is almost at its size limit. I've looked into CONFIG_TIMER and it seems that we'd also need the whole clock infrastructure (and DM_CLK). No?

Quoted 7 lines

So, for now I see only two options:

1) Either apply that change which moves preprocessor macros to C code and apply this one fix.

2) Or revert that my kirkwood commit 8ac303d49f89 and do not apply this one fix.

Stefan, what do you prefer?

I prefer option 1. Could you please work on a new patch for this and include some reasoning for this change in the commit message?

Thanks,
Stefan

#9 lore raw
Quoted 149 lines · On Wednesday 17 August 2022 12:42:57 Stefan Roese wrote:

Hi Pali,

Quoted 146 lines · Pali Rohár
Quoted 135 lines · On Wednesday 17 August 2022 01:23:34 Michael Walle wrote:

Am 2022-08-17 00:39, schrieb Pali Rohár:

Quoted 125 lines · On Wednesday 17 August 2022 00:14:20 Pali Rohár wrote:
Quoted 102 lines · On Tuesday 16 August 2022 23:34:13 Michael Walle wrote:
Quoted 60 lines

Am 2022-08-16 22:00, schrieb Pali Rohár:

Quoted 20 lines

Bit 21 in SAR register specifies if TCLK is running at 166 MHz or 200 MHz.
This information is undocumented in public Marvell Kirkwood Functional Specifications [2], but is available in Linux v3.15 kirkwood code [1].

Commit 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK") broke support for Marvell 88F6281 SoCs because it was expected that all those SoCs have TCLK running at 200 MHz as specified in Marvell 88F6281 Hardware Specifications [3].

Fix broken support for 88F6281 by detecting CONFIG_SYS_TCLK from SAR register, like it was doing Linux v3.15.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/mach-kirkwood/common.c?h=v3.15#n542
[2] -
https://web.archive.org/web/20130730091033/http://www.marvell.com/embedded-processors/kirkwood/assets/FS_88F6180_9x_6281_OpenSource.pdf
[3] -
https://web.archive.org/web/20120620073511/http://www.marvell.com/embedded-processors/kirkwood/assets/HW_88F6281_OpenSource.pdf

Quoted 4 lines
Fixes 8ac303d49f89 ("arm: kirkwood: Do not overwrite CONFIG_SYS_TCLK")

---
Michael, please test this patch, if it fixes your boards.

Thanks, but this doesn't compile:

In file included from lib/time.c:19:
lib/time.c: In function ‘timer_get_boot_us’:
./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’

    18 | #define CONFIG_SYS_TCLK   ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
       |                             ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro
‘CONFIG_SYS_TCLK’
    56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
       |                                ^~~~~~~~~~~~~~~
lib/time.c:50:5: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
    50 | #if CONFIG_SYS_TIMER_RATE == 1000000
       |     ^~~~~~~~~~~~~~~~~~~~~

./arch/arm/include/asm/io.h:108:19: error: token "{" is not valid in preprocessor expressions
108 | #define readl(c) ({ u32 __v = __arch_getl(c); __iormb(); __v; })
| ^
./arch/arm/include/asm/arch/kw88f6281.h:18:29: note: in expansion of macro ‘readl’

    18 | #define CONFIG_SYS_TCLK   ((readl(CONFIG_SAR_REG) & BIT(21)) ? \
       |                             ^~~~~
./arch/arm/include/asm/arch/config.h:56:32: note: in expansion of macro
‘CONFIG_SYS_TCLK’
    56 | #define CONFIG_SYS_TIMER_RATE  CONFIG_SYS_TCLK
       |                                ^~~~~~~~~~~~~~~
lib/time.c:52:7: note: in expansion of macro ‘CONFIG_SYS_TIMER_RATE’
    52 | #elif CONFIG_SYS_TIMER_RATE > 1000000
       |       ^~~~~~~~~~~~~~~~~~~~~
Quoted 3 lines

make[1]: *** [scripts/Makefile.build:257: lib/time.o] Fehler 1

-michael

Ah :-( And why it works on other Armada mvebu boards? Because of this:

$ git grep CONFIG_SYS_TIMER_RATE arch/arm/mach-mvebu
arch/arm/mach-mvebu/include/mach/config.h:#define
CONFIG_SYS_TIMER_RATE         25000000

I have looked how is this option used and it is unit for CONFIG_SYS_TIMER_COUNTER option. On all kirkwood and mvebu platforms is
that register set to SoC Global Timer 0 which on Armada 38x uses either
fixed 25MHz clock or NBCLK clock, together with SoC Global Timer 0 Ratio. When NBCLK then the frequency of that timer register is:
NBCLK / 2 / (2 ^ ratio). Default ratio is 0 (but can be configured up to
7). By default fixed 25MHz clock is used.

Hence for mvebu it is OK as U-Boot does not change default values of SoC
Global Timer 0.

For kirkwood it looks like that those timers are based on TCLK (page 283
in Functional Specifications).

How to fix it? I do not know right now. I have just quirk & dirty solution by moving code from preprocessor to compiler:

lib/time.c +0 -1
index 96074b84af69..f2ffc0498d39 100644
@@ -46,13 +46,15 @@ unsigned long notrace timer_read_counter(void)
  ulong timer_get_boot_us(void)
  {
  	ulong count = timer_read_counter();
-

lib/time.c unsigned long notrace timer_read_counter(void) · line 52

Quoted 10 lines
lib/time.c +4 -6
-#if CONFIG_SYS_TIMER_RATE == 1000000
-	return count;
-#elif CONFIG_SYS_TIMER_RATE > 1000000
-	return lldiv(count, CONFIG_SYS_TIMER_RATE / 1000000);
-#elif defined(CONFIG_SYS_TIMER_RATE)
-	return (unsigned long long)count * 1000000 / CONFIG_SYS_TIMER_RATE;
+#ifdef CONFIG_SYS_TIMER_RATE
+	const ulong timer_rate = CONFIG_SYS_TIMER_RATE;
+
+	if (timer_rate)
             ~~~~~~~~~~

Here needs to be if (timer_rate == 1000000)

Quoted 9 lines
lib/time.c +5 -0
+		return count;
+	else if (timer_rate > 1000000)
+		return lldiv(count, timer_rate / 1000000);
+	else
+		return (unsigned long long)count * 1000000 / timer_rate;
  #else
  	/* Assume the counter is in microseconds */
  	return count;
Quoted 1 line
Quoted 1 line

At least it could be used for verifying if CONFIG_SYS_TCLK is working.

FWIW, this is working.

Quoted 4 lines

Anyway, it looks like that this timer code is deprecated and boards should migrate to use DM.

So maybe the proper fix for this would be to add kirkwood dm driver?

Quoted 7 lines

Eventually sure, but for now I'd like to have a working board without spending too much time on fixing this. I'd propose to revert the broken commit for now and then, we can work on a proper timer driver.

But I can already say that the u-boot binary is almost at its size limit. I've looked into CONFIG_TIMER and it seems that we'd also need the whole clock infrastructure (and DM_CLK). No?

So, for now I see only two options:

Quoted 8 lines
Quoted 5 lines

1) Either apply that change which moves preprocessor macros to C code and apply this one fix.

2) Or revert that my kirkwood commit 8ac303d49f89 and do not apply this one fix.

Stefan, what do you prefer?

I prefer option 1. Could you please work on a new patch for this and include some reasoning for this change in the commit message?

Michael, will you prepare and test it? I think you have now everything required.