sched_clock(), used by printk(), calls __USE_RTC() to know
whether to use realtime clock or timebase.
__USE_RTC() uses cpu_has_feature() which is initialised by
machine_init(). Before machine_init(), __USE_RTC() returns true,
leading to a program check exception on CPUs not having realtime
clock.
In order to be able to use printk() earlier, use feature fixup.
Feature fixups are applies in early_init(), enabling the use of
printk() earlier.
Signed-off-by: Christophe Leroy <redacted>
---
arch/powerpc/include/asm/time.h | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
@@ -42,7 +42,14 @@ struct div_result {/* Accessor functions for the timebase (RTC on 601) registers. *//* If one day CONFIG_POWER is added just define __USE_RTC as 1 */#ifdef CONFIG_PPC_BOOK3S_32-#define __USE_RTC() (cpu_has_feature(CPU_FTR_USE_RTC))+staticinlinebool__USE_RTC(void)+{+asm_volatile_goto(ASM_FTR_IFCLR("nop;","b %1;",%0)::+"i"(CPU_FTR_USE_RTC)::l_use_rtc);+returnfalse;+l_use_rtc:+returntrue;+}#else#define __USE_RTC() 0#endif
From: Michael Ellerman <mpe@ellerman.id.au> Date: 2019-08-26 11:41:47
Christophe Leroy [off-list ref] writes:
sched_clock(), used by printk(), calls __USE_RTC() to know
whether to use realtime clock or timebase.
__USE_RTC() uses cpu_has_feature() which is initialised by
machine_init(). Before machine_init(), __USE_RTC() returns true,
leading to a program check exception on CPUs not having realtime
clock.
In order to be able to use printk() earlier, use feature fixup.
Feature fixups are applies in early_init(), enabling the use of
printk() earlier.
Signed-off-by: Christophe Leroy <redacted>
---
arch/powerpc/include/asm/time.h | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
The other option would be just to make this a compile time decision, eg.
add CONFIG_PPC_601 and use that to gate whether we use RTC.
Given how many 601 users there are, maybe 1?, I think that would be a
simpler option and avoids complicating the code / binary for everyone
else.
cheers
@@ -42,7 +42,14 @@ struct div_result {/* Accessor functions for the timebase (RTC on 601) registers. *//* If one day CONFIG_POWER is added just define __USE_RTC as 1 */#ifdef CONFIG_PPC_BOOK3S_32-#define __USE_RTC() (cpu_has_feature(CPU_FTR_USE_RTC))+staticinlinebool__USE_RTC(void)+{+asm_volatile_goto(ASM_FTR_IFCLR("nop;","b %1;",%0)::+"i"(CPU_FTR_USE_RTC)::l_use_rtc);+returnfalse;+l_use_rtc:+returntrue;+}#else#define __USE_RTC() 0#endif
sched_clock(), used by printk(), calls __USE_RTC() to know
whether to use realtime clock or timebase.
__USE_RTC() uses cpu_has_feature() which is initialised by
machine_init(). Before machine_init(), __USE_RTC() returns true,
leading to a program check exception on CPUs not having realtime
clock.
In order to be able to use printk() earlier, use feature fixup.
Feature fixups are applies in early_init(), enabling the use of
printk() earlier.
Signed-off-by: Christophe Leroy <redacted>
---
arch/powerpc/include/asm/time.h | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
The other option would be just to make this a compile time decision, eg.
add CONFIG_PPC_601 and use that to gate whether we use RTC.
Euh ... yes OK, why not. And that would help simplify many places in the
code. I can propose something in that direction, but it will be a bigger
change.
Given how many 601 users there are, maybe 1?, I think that would be a
simpler option and avoids complicating the code / binary for everyone
else.
However this patch doesn't complicate things more than it was with
cpu_has_feature() which does exactly the same but using static keys,
does it ?
Christophe
@@ -42,7 +42,14 @@ struct div_result {/* Accessor functions for the timebase (RTC on 601) registers. *//* If one day CONFIG_POWER is added just define __USE_RTC as 1 */#ifdef CONFIG_PPC_BOOK3S_32-#define __USE_RTC() (cpu_has_feature(CPU_FTR_USE_RTC))+staticinlinebool__USE_RTC(void)+{+asm_volatile_goto(ASM_FTR_IFCLR("nop;","b %1;",%0)::+"i"(CPU_FTR_USE_RTC)::l_use_rtc);+returnfalse;+l_use_rtc:+returntrue;+}#else#define __USE_RTC() 0#endif
From: Benjamin Herrenschmidt <benh@kernel.crashing.org> Date: 2019-08-26 13:25:47
On Mon, 2019-08-26 at 21:41 +1000, Michael Ellerman wrote:
Christophe Leroy [off-list ref] writes:
quoted
sched_clock(), used by printk(), calls __USE_RTC() to know
whether to use realtime clock or timebase.
__USE_RTC() uses cpu_has_feature() which is initialised by
machine_init(). Before machine_init(), __USE_RTC() returns true,
leading to a program check exception on CPUs not having realtime
clock.
In order to be able to use printk() earlier, use feature fixup.
Feature fixups are applies in early_init(), enabling the use of
printk() earlier.
Signed-off-by: Christophe Leroy <redacted>
---
arch/powerpc/include/asm/time.h | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
The other option would be just to make this a compile time decision, eg.
add CONFIG_PPC_601 and use that to gate whether we use RTC.
Given how many 601 users there are, maybe 1?, I think that would be a
simpler option and avoids complicating the code / binary for everyone
else.
Didn't we ditch 601 support years ago anyway ? We had workaround we
threw out I think...
Cheers,
Ben.
@@ -42,7 +42,14 @@ struct div_result {/* Accessor functions for the timebase (RTC on 601) registers. *//* If one day CONFIG_POWER is added just define __USE_RTC as 1 */#ifdef CONFIG_PPC_BOOK3S_32-#define __USE_RTC() (cpu_has_feature(CPU_FTR_USE_RTC))+staticinlinebool__USE_RTC(void)+{+asm_volatile_goto(ASM_FTR_IFCLR("nop;","b %1;",%0)::+"i"(CPU_FTR_USE_RTC)::l_use_rtc);+returnfalse;+l_use_rtc:+returntrue;+}#else#define __USE_RTC() 0#endif
Le 26/08/2019 à 15:25, Benjamin Herrenschmidt a écrit :
On Mon, 2019-08-26 at 21:41 +1000, Michael Ellerman wrote:
quoted
Christophe Leroy [off-list ref] writes:
quoted
sched_clock(), used by printk(), calls __USE_RTC() to know
whether to use realtime clock or timebase.
__USE_RTC() uses cpu_has_feature() which is initialised by
machine_init(). Before machine_init(), __USE_RTC() returns true,
leading to a program check exception on CPUs not having realtime
clock.
In order to be able to use printk() earlier, use feature fixup.
Feature fixups are applies in early_init(), enabling the use of
printk() earlier.
Signed-off-by: Christophe Leroy <redacted>
---
arch/powerpc/include/asm/time.h | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
The other option would be just to make this a compile time decision, eg.
add CONFIG_PPC_601 and use that to gate whether we use RTC.
Given how many 601 users there are, maybe 1?, I think that would be a
simpler option and avoids complicating the code / binary for everyone
else.
Didn't we ditch 601 support years ago anyway ? We had workaround we
threw out I think...
There are still workarounds for 601 in the kernel, for exemple (in
arch/powerpc/include/asm/ppc_asm.h)
/* various errata or part fixups */
#ifdef CONFIG_PPC601_SYNC_FIX
#define SYNC \
BEGIN_FTR_SECTION \
sync; \
isync; \
END_FTR_SECTION_IFSET(CPU_FTR_601)
#define SYNC_601 \
BEGIN_FTR_SECTION \
sync; \
END_FTR_SECTION_IFSET(CPU_FTR_601)
#define ISYNC_601 \
BEGIN_FTR_SECTION \
isync; \
END_FTR_SECTION_IFSET(CPU_FTR_601)
#else
#define SYNC
#define SYNC_601
#define ISYNC_601
#endif
But if you think we can get rid of 601 completely, I'm happy with that.
Christophe
@@ -42,7 +42,14 @@ struct div_result {/* Accessor functions for the timebase (RTC on 601) registers. *//* If one day CONFIG_POWER is added just define __USE_RTC as 1 */#ifdef CONFIG_PPC_BOOK3S_32-#define __USE_RTC() (cpu_has_feature(CPU_FTR_USE_RTC))+staticinlinebool__USE_RTC(void)+{+asm_volatile_goto(ASM_FTR_IFCLR("nop;","b %1;",%0)::+"i"(CPU_FTR_USE_RTC)::l_use_rtc);+returnfalse;+l_use_rtc:+returntrue;+}#else#define __USE_RTC() 0#endif
From: Michael Ellerman <mpe@ellerman.id.au> Date: 2019-08-27 02:12:25
Benjamin Herrenschmidt [off-list ref] writes:
On Mon, 2019-08-26 at 21:41 +1000, Michael Ellerman wrote:
quoted
Christophe Leroy [off-list ref] writes:
quoted
sched_clock(), used by printk(), calls __USE_RTC() to know
whether to use realtime clock or timebase.
__USE_RTC() uses cpu_has_feature() which is initialised by
machine_init(). Before machine_init(), __USE_RTC() returns true,
leading to a program check exception on CPUs not having realtime
clock.
In order to be able to use printk() earlier, use feature fixup.
Feature fixups are applies in early_init(), enabling the use of
printk() earlier.
Signed-off-by: Christophe Leroy <redacted>
---
arch/powerpc/include/asm/time.h | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
The other option would be just to make this a compile time decision, eg.
add CONFIG_PPC_601 and use that to gate whether we use RTC.
Given how many 601 users there are, maybe 1?, I think that would be a
simpler option and avoids complicating the code / binary for everyone
else.
Didn't we ditch 601 support years ago anyway ? We had workaround we
threw out I think...
From: Michael Ellerman <mpe@ellerman.id.au> Date: 2019-08-27 02:13:55
Christophe Leroy [off-list ref] writes:
Le 26/08/2019 à 13:41, Michael Ellerman a écrit :
quoted
Christophe Leroy [off-list ref] writes:
quoted
sched_clock(), used by printk(), calls __USE_RTC() to know
whether to use realtime clock or timebase.
__USE_RTC() uses cpu_has_feature() which is initialised by
machine_init(). Before machine_init(), __USE_RTC() returns true,
leading to a program check exception on CPUs not having realtime
clock.
In order to be able to use printk() earlier, use feature fixup.
Feature fixups are applies in early_init(), enabling the use of
printk() earlier.
Signed-off-by: Christophe Leroy <redacted>
---
arch/powerpc/include/asm/time.h | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
The other option would be just to make this a compile time decision, eg.
add CONFIG_PPC_601 and use that to gate whether we use RTC.
Euh ... yes OK, why not. And that would help simplify many places in the
code. I can propose something in that direction, but it will be a bigger
change.
Thanks.
quoted
Given how many 601 users there are, maybe 1?, I think that would be a
simpler option and avoids complicating the code / binary for everyone
else.
However this patch doesn't complicate things more than it was with
cpu_has_feature() which does exactly the same but using static keys,
does it ?
It's more complicated in that it's not using cpu_has_feature() it's
doing some custom thing that is not used anywhere else. But yeah I guess
it's not much extra complication.
cheers