linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] mips: smp-bmips: fix CPU mappings
@ 2021-02-23 12:48 Álvaro Fernández Rojas
  2021-02-24  3:45 ` Florian Fainelli
  2021-02-24  7:33 ` [PATCH v2] " Álvaro Fernández Rojas
  0 siblings, 2 replies; 7+ messages in thread
From: Álvaro Fernández Rojas @ 2021-02-23 12:48 UTC (permalink / raw)
  To: jonas.gorski, Florian Fainelli, Thomas Gleixner, Marc Zyngier,
	bcm-kernel-feedback-list, linux-mips, linux-kernel
  Cc: Álvaro Fernández Rojas

When booting bmips with SMP enabled on a BCM6358 running on CPU #1 instead of
CPU #0, the current CPU mapping code produces the following:
- smp_processor_id(): 0
- cpu_logical_map(): 1
- cpu_number_map(): 1

This is because SMP isn't supported on BCM6358 since it has a shared TLB, so
it is disabled and max_cpus is decreased from 2 to 1.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
---
 arch/mips/kernel/smp-bmips.c | 27 +++++++++++++++++----------
 1 file changed, 17 insertions(+), 10 deletions(-)

diff --git a/arch/mips/kernel/smp-bmips.c b/arch/mips/kernel/smp-bmips.c
index 359b176b665f..c4760cb48a67 100644
--- a/arch/mips/kernel/smp-bmips.c
+++ b/arch/mips/kernel/smp-bmips.c
@@ -134,17 +134,24 @@ static void __init bmips_smp_setup(void)
 	if (!board_ebase_setup)
 		board_ebase_setup = &bmips_ebase_setup;
 
-	__cpu_number_map[boot_cpu] = 0;
-	__cpu_logical_map[0] = boot_cpu;
-
-	for (i = 0; i < max_cpus; i++) {
-		if (i != boot_cpu) {
-			__cpu_number_map[i] = cpu;
-			__cpu_logical_map[cpu] = i;
-			cpu++;
+	if (max_cpus > 1) {
+		__cpu_number_map[boot_cpu] = 0;
+		__cpu_logical_map[0] = boot_cpu;
+
+		for (i = 0; i < max_cpus; i++) {
+			if (i != boot_cpu) {
+				__cpu_number_map[i] = cpu;
+				__cpu_logical_map[cpu] = i;
+				cpu++;
+			}
+			set_cpu_possible(i, 1);
+			set_cpu_present(i, 1);
 		}
-		set_cpu_possible(i, 1);
-		set_cpu_present(i, 1);
+	} else {
+		__cpu_number_map[0] = boot_cpu;
+		__cpu_logical_map[0] = 0;
+		set_cpu_possible(0, 1);
+		set_cpu_possible(0, 1);
 	}
 }
 
-- 
2.20.1


^ permalink raw reply related	[flat|nested] 7+ messages in thread

* Re: [PATCH] mips: smp-bmips: fix CPU mappings
  2021-02-23 12:48 [PATCH] mips: smp-bmips: fix CPU mappings Álvaro Fernández Rojas
@ 2021-02-24  3:45 ` Florian Fainelli
  2021-02-24  7:07   ` Álvaro Fernández Rojas
  2021-02-24  7:33 ` [PATCH v2] " Álvaro Fernández Rojas
  1 sibling, 1 reply; 7+ messages in thread
From: Florian Fainelli @ 2021-02-24  3:45 UTC (permalink / raw)
  To: Álvaro Fernández Rojas, jonas.gorski, Florian Fainelli,
	Thomas Gleixner, Marc Zyngier, bcm-kernel-feedback-list,
	linux-mips, linux-kernel



On 2/23/2021 4:48 AM, Álvaro Fernández Rojas wrote:
> When booting bmips with SMP enabled on a BCM6358 running on CPU #1 instead of
> CPU #0, the current CPU mapping code produces the following:
> - smp_processor_id(): 0
> - cpu_logical_map(): 1
> - cpu_number_map(): 1
> 
> This is because SMP isn't supported on BCM6358 since it has a shared TLB, so
> it is disabled and max_cpus is decreased from 2 to 1.
> 
> Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
> ---
>  arch/mips/kernel/smp-bmips.c | 27 +++++++++++++++++----------
>  1 file changed, 17 insertions(+), 10 deletions(-)
> 
> diff --git a/arch/mips/kernel/smp-bmips.c b/arch/mips/kernel/smp-bmips.c
> index 359b176b665f..c4760cb48a67 100644
> --- a/arch/mips/kernel/smp-bmips.c
> +++ b/arch/mips/kernel/smp-bmips.c
> @@ -134,17 +134,24 @@ static void __init bmips_smp_setup(void)
>  	if (!board_ebase_setup)
>  		board_ebase_setup = &bmips_ebase_setup;
>  
> -	__cpu_number_map[boot_cpu] = 0;
> -	__cpu_logical_map[0] = boot_cpu;
> -
> -	for (i = 0; i < max_cpus; i++) {
> -		if (i != boot_cpu) {
> -			__cpu_number_map[i] = cpu;
> -			__cpu_logical_map[cpu] = i;
> -			cpu++;
> +	if (max_cpus > 1) {
> +		__cpu_number_map[boot_cpu] = 0;
> +		__cpu_logical_map[0] = boot_cpu;
> +
> +		for (i = 0; i < max_cpus; i++) {
> +			if (i != boot_cpu) {
> +				__cpu_number_map[i] = cpu;
> +				__cpu_logical_map[cpu] = i;
> +				cpu++;
> +			}
> +			set_cpu_possible(i, 1);
> +			set_cpu_present(i, 1);
>  		}
> -		set_cpu_possible(i, 1);
> -		set_cpu_present(i, 1);
> +	} else {
> +		__cpu_number_map[0] = boot_cpu;
> +		__cpu_logical_map[0] = 0;
> +		set_cpu_possible(0, 1);
> +		set_cpu_possible(0, 1);

Duplicate line, with that fixed:

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [PATCH] mips: smp-bmips: fix CPU mappings
  2021-02-24  3:45 ` Florian Fainelli
@ 2021-02-24  7:07   ` Álvaro Fernández Rojas
  0 siblings, 0 replies; 7+ messages in thread
From: Álvaro Fernández Rojas @ 2021-02-24  7:07 UTC (permalink / raw)
  To: Florian Fainelli
  Cc: Jonas Gorski, Thomas Gleixner, Marc Zyngier,
	bcm-kernel-feedback-list, linux-mips, linux-kernel

Hi Florian,

> El 24 feb 2021, a las 4:45, Florian Fainelli <f.fainelli@gmail.com> escribió:
> 
> 
> 
> On 2/23/2021 4:48 AM, Álvaro Fernández Rojas wrote:
>> When booting bmips with SMP enabled on a BCM6358 running on CPU #1 instead of
>> CPU #0, the current CPU mapping code produces the following:
>> - smp_processor_id(): 0
>> - cpu_logical_map(): 1
>> - cpu_number_map(): 1
>> 
>> This is because SMP isn't supported on BCM6358 since it has a shared TLB, so
>> it is disabled and max_cpus is decreased from 2 to 1.
>> 
>> Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
>> ---
>> arch/mips/kernel/smp-bmips.c | 27 +++++++++++++++++----------
>> 1 file changed, 17 insertions(+), 10 deletions(-)
>> 
>> diff --git a/arch/mips/kernel/smp-bmips.c b/arch/mips/kernel/smp-bmips.c
>> index 359b176b665f..c4760cb48a67 100644
>> --- a/arch/mips/kernel/smp-bmips.c
>> +++ b/arch/mips/kernel/smp-bmips.c
>> @@ -134,17 +134,24 @@ static void __init bmips_smp_setup(void)
>> 	if (!board_ebase_setup)
>> 		board_ebase_setup = &bmips_ebase_setup;
>> 
>> -	__cpu_number_map[boot_cpu] = 0;
>> -	__cpu_logical_map[0] = boot_cpu;
>> -
>> -	for (i = 0; i < max_cpus; i++) {
>> -		if (i != boot_cpu) {
>> -			__cpu_number_map[i] = cpu;
>> -			__cpu_logical_map[cpu] = i;
>> -			cpu++;
>> +	if (max_cpus > 1) {
>> +		__cpu_number_map[boot_cpu] = 0;
>> +		__cpu_logical_map[0] = boot_cpu;
>> +
>> +		for (i = 0; i < max_cpus; i++) {
>> +			if (i != boot_cpu) {
>> +				__cpu_number_map[i] = cpu;
>> +				__cpu_logical_map[cpu] = i;
>> +				cpu++;
>> +			}
>> +			set_cpu_possible(i, 1);
>> +			set_cpu_present(i, 1);
>> 		}
>> -		set_cpu_possible(i, 1);
>> -		set_cpu_present(i, 1);
>> +	} else {
>> +		__cpu_number_map[0] = boot_cpu;
>> +		__cpu_logical_map[0] = 0;
>> +		set_cpu_possible(0, 1);
>> +		set_cpu_possible(0, 1);
> 
> Duplicate line, with that fixed:

Nice catch, it should be set_cpu_present().

> 
> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
> -- 
> Florian


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [PATCH v2] mips: smp-bmips: fix CPU mappings
  2021-02-23 12:48 [PATCH] mips: smp-bmips: fix CPU mappings Álvaro Fernández Rojas
  2021-02-24  3:45 ` Florian Fainelli
@ 2021-02-24  7:33 ` Álvaro Fernández Rojas
  2021-02-27  6:47   ` Álvaro Fernández Rojas
  2021-03-01 14:50   ` Thomas Bogendoerfer
  1 sibling, 2 replies; 7+ messages in thread
From: Álvaro Fernández Rojas @ 2021-02-24  7:33 UTC (permalink / raw)
  To: jonas.gorski, Florian Fainelli, Thomas Gleixner, Marc Zyngier,
	bcm-kernel-feedback-list, linux-mips, linux-kernel
  Cc: Álvaro Fernández Rojas

When booting bmips with SMP enabled on a BCM6358 running on CPU #1 instead of
CPU #0, the current CPU mapping code produces the following:
- smp_processor_id(): 0
- cpu_logical_map(0): 1
- cpu_number_map(0): 1

This is because SMP isn't supported on BCM6358 since it has a shared TLB, so
it is disabled and max_cpus is decreased from 2 to 1.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 v2: Fix duplicated line

 arch/mips/kernel/smp-bmips.c | 27 +++++++++++++++++----------
 1 file changed, 17 insertions(+), 10 deletions(-)

diff --git a/arch/mips/kernel/smp-bmips.c b/arch/mips/kernel/smp-bmips.c
index 359b176b665f..b6ef5f7312cf 100644
--- a/arch/mips/kernel/smp-bmips.c
+++ b/arch/mips/kernel/smp-bmips.c
@@ -134,17 +134,24 @@ static void __init bmips_smp_setup(void)
 	if (!board_ebase_setup)
 		board_ebase_setup = &bmips_ebase_setup;
 
-	__cpu_number_map[boot_cpu] = 0;
-	__cpu_logical_map[0] = boot_cpu;
-
-	for (i = 0; i < max_cpus; i++) {
-		if (i != boot_cpu) {
-			__cpu_number_map[i] = cpu;
-			__cpu_logical_map[cpu] = i;
-			cpu++;
+	if (max_cpus > 1) {
+		__cpu_number_map[boot_cpu] = 0;
+		__cpu_logical_map[0] = boot_cpu;
+
+		for (i = 0; i < max_cpus; i++) {
+			if (i != boot_cpu) {
+				__cpu_number_map[i] = cpu;
+				__cpu_logical_map[cpu] = i;
+				cpu++;
+			}
+			set_cpu_possible(i, 1);
+			set_cpu_present(i, 1);
 		}
-		set_cpu_possible(i, 1);
-		set_cpu_present(i, 1);
+	} else {
+		__cpu_number_map[0] = boot_cpu;
+		__cpu_logical_map[0] = 0;
+		set_cpu_possible(0, 1);
+		set_cpu_present(0, 1);
 	}
 }
 
-- 
2.20.1


^ permalink raw reply related	[flat|nested] 7+ messages in thread

* Re: [PATCH v2] mips: smp-bmips: fix CPU mappings
  2021-02-24  7:33 ` [PATCH v2] " Álvaro Fernández Rojas
@ 2021-02-27  6:47   ` Álvaro Fernández Rojas
  2021-03-01 12:33     ` Thomas Bogendoerfer
  2021-03-01 14:50   ` Thomas Bogendoerfer
  1 sibling, 1 reply; 7+ messages in thread
From: Álvaro Fernández Rojas @ 2021-02-27  6:47 UTC (permalink / raw)
  To: Jonas Gorski, Florian Fainelli, Thomas Gleixner, Marc Zyngier,
	bcm-kernel-feedback-list, linux-mips, linux-kernel

Hi all,

Apparently, this patch was flagged as "Not Applicable" without an explanation. Why?
https://patchwork.kernel.org/project/linux-mips/patch/20210224073336.32265-1-noltari@gmail.com/

Best regarss,
Álvaro.

> El 24 feb 2021, a las 8:33, Álvaro Fernández Rojas <noltari@gmail.com> escribió:
> 
> When booting bmips with SMP enabled on a BCM6358 running on CPU #1 instead of
> CPU #0, the current CPU mapping code produces the following:
> - smp_processor_id(): 0
> - cpu_logical_map(0): 1
> - cpu_number_map(0): 1
> 
> This is because SMP isn't supported on BCM6358 since it has a shared TLB, so
> it is disabled and max_cpus is decreased from 2 to 1.
> 
> Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
> ---
> v2: Fix duplicated line
> 
> arch/mips/kernel/smp-bmips.c | 27 +++++++++++++++++----------
> 1 file changed, 17 insertions(+), 10 deletions(-)
> 
> diff --git a/arch/mips/kernel/smp-bmips.c b/arch/mips/kernel/smp-bmips.c
> index 359b176b665f..b6ef5f7312cf 100644
> --- a/arch/mips/kernel/smp-bmips.c
> +++ b/arch/mips/kernel/smp-bmips.c
> @@ -134,17 +134,24 @@ static void __init bmips_smp_setup(void)
> 	if (!board_ebase_setup)
> 		board_ebase_setup = &bmips_ebase_setup;
> 
> -	__cpu_number_map[boot_cpu] = 0;
> -	__cpu_logical_map[0] = boot_cpu;
> -
> -	for (i = 0; i < max_cpus; i++) {
> -		if (i != boot_cpu) {
> -			__cpu_number_map[i] = cpu;
> -			__cpu_logical_map[cpu] = i;
> -			cpu++;
> +	if (max_cpus > 1) {
> +		__cpu_number_map[boot_cpu] = 0;
> +		__cpu_logical_map[0] = boot_cpu;
> +
> +		for (i = 0; i < max_cpus; i++) {
> +			if (i != boot_cpu) {
> +				__cpu_number_map[i] = cpu;
> +				__cpu_logical_map[cpu] = i;
> +				cpu++;
> +			}
> +			set_cpu_possible(i, 1);
> +			set_cpu_present(i, 1);
> 		}
> -		set_cpu_possible(i, 1);
> -		set_cpu_present(i, 1);
> +	} else {
> +		__cpu_number_map[0] = boot_cpu;
> +		__cpu_logical_map[0] = 0;
> +		set_cpu_possible(0, 1);
> +		set_cpu_present(0, 1);
> 	}
> }
> 
> -- 
> 2.20.1
> 


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [PATCH v2] mips: smp-bmips: fix CPU mappings
  2021-02-27  6:47   ` Álvaro Fernández Rojas
@ 2021-03-01 12:33     ` Thomas Bogendoerfer
  0 siblings, 0 replies; 7+ messages in thread
From: Thomas Bogendoerfer @ 2021-03-01 12:33 UTC (permalink / raw)
  To: Álvaro Fernández Rojas
  Cc: Jonas Gorski, Florian Fainelli, Thomas Gleixner, Marc Zyngier,
	bcm-kernel-feedback-list, linux-mips, linux-kernel

On Sat, Feb 27, 2021 at 07:47:42AM +0100, Álvaro Fernández Rojas wrote:
> Hi all,
> 
> Apparently, this patch was flagged as "Not Applicable" without an
> explanation. Why?

hmm, I probaly wanted to drop the first version, but changed v2. It's
back to new again and I'm gogin to apply it soon.

Sorry about that.

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea.                                                [ RFC1925, 2.3 ]

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [PATCH v2] mips: smp-bmips: fix CPU mappings
  2021-02-24  7:33 ` [PATCH v2] " Álvaro Fernández Rojas
  2021-02-27  6:47   ` Álvaro Fernández Rojas
@ 2021-03-01 14:50   ` Thomas Bogendoerfer
  1 sibling, 0 replies; 7+ messages in thread
From: Thomas Bogendoerfer @ 2021-03-01 14:50 UTC (permalink / raw)
  To: Álvaro Fernández Rojas
  Cc: jonas.gorski, Florian Fainelli, Thomas Gleixner, Marc Zyngier,
	bcm-kernel-feedback-list, linux-mips, linux-kernel

On Wed, Feb 24, 2021 at 08:33:36AM +0100, Álvaro Fernández Rojas wrote:
> When booting bmips with SMP enabled on a BCM6358 running on CPU #1 instead of
> CPU #0, the current CPU mapping code produces the following:
> - smp_processor_id(): 0
> - cpu_logical_map(0): 1
> - cpu_number_map(0): 1
> 
> This is because SMP isn't supported on BCM6358 since it has a shared TLB, so
> it is disabled and max_cpus is decreased from 2 to 1.
> 
> Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
> ---
>  v2: Fix duplicated line
> 
>  arch/mips/kernel/smp-bmips.c | 27 +++++++++++++++++----------
>  1 file changed, 17 insertions(+), 10 deletions(-)

applied to mips-next.

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea.                                                [ RFC1925, 2.3 ]

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2021-03-01 15:04 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-02-23 12:48 [PATCH] mips: smp-bmips: fix CPU mappings Álvaro Fernández Rojas
2021-02-24  3:45 ` Florian Fainelli
2021-02-24  7:07   ` Álvaro Fernández Rojas
2021-02-24  7:33 ` [PATCH v2] " Álvaro Fernández Rojas
2021-02-27  6:47   ` Álvaro Fernández Rojas
2021-03-01 12:33     ` Thomas Bogendoerfer
2021-03-01 14:50   ` Thomas Bogendoerfer

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).