[PATCH V3 2/4] arm_pmu: acpi: Add a representative platform device for TRBE

[Anshuman Khandual]

On 8/3/23 14:44, Yicong Yang wrote:
> On 2023/8/3 13:56, Anshuman Khandual wrote:
>> ACPI TRBE does not have a HID for identification which could create and add
>> a platform device into the platform bus. Also without a platform device, it
>> cannot be probed and bound to a platform driver.
>>
>> This creates a dummy platform device for TRBE after ascertaining that ACPI
>> provides required interrupts uniformly across all cpus on the system. This
>> device gets created inside drivers/perf/arm_pmu_acpi.c to accommodate TRBE
>> being built as a module.
>>
>> Cc: Catalin Marinas <catalin.marinas at arm.com>
>> Cc: Will Deacon <will at kernel.org>
>> Cc: Mark Rutland <mark.rutland at arm.com>
>> Cc: linux-arm-kernel at lists.infradead.org
>> Cc: linux-kernel at vger.kernel.org
>> Signed-off-by: Anshuman Khandual <anshuman.khandual at arm.com>
>> ---
>>  arch/arm64/include/asm/acpi.h |  3 +++
>>  drivers/perf/arm_pmu_acpi.c   | 37 ++++++++++++++++++++++++++++++++++-
>>  include/linux/perf/arm_pmu.h  |  1 +
>>  3 files changed, 40 insertions(+), 1 deletion(-)
>>
>> diff --git a/arch/arm64/include/asm/acpi.h b/arch/arm64/include/asm/acpi.h
>> index bd68e1b7f29f..4d537d56eb84 100644
>> --- a/arch/arm64/include/asm/acpi.h
>> +++ b/arch/arm64/include/asm/acpi.h
>> @@ -42,6 +42,9 @@
>>  #define ACPI_MADT_GICC_SPE  (offsetof(struct acpi_madt_generic_interrupt, \
>>  	spe_interrupt) + sizeof(u16))
>>  
>> +#define ACPI_MADT_GICC_TRBE  (offsetof(struct acpi_madt_generic_interrupt, \
>> +	trbe_interrupt) + sizeof(u16))
>> +
>>  /* Basic configuration for ACPI */
>>  #ifdef	CONFIG_ACPI
>>  pgprot_t __acpi_get_mem_attribute(phys_addr_t addr);
>> diff --git a/drivers/perf/arm_pmu_acpi.c b/drivers/perf/arm_pmu_acpi.c
>> index 235c14766a36..79feea548e6e 100644
>> --- a/drivers/perf/arm_pmu_acpi.c
>> +++ b/drivers/perf/arm_pmu_acpi.c
>> @@ -69,7 +69,7 @@ static void arm_pmu_acpi_unregister_irq(int cpu)
>>  		acpi_unregister_gsi(gsi);
>>  }
>>  
>> -#if IS_ENABLED(CONFIG_ARM_SPE_PMU)
>> +#if IS_ENABLED(CONFIG_ARM_SPE_PMU) || IS_ENABLED(CONFIG_CORESIGHT_TRBE)
>>  static int
>>  arm_acpi_register_pmu_device(struct platform_device *pdev, u8 len,
>>  			     u16 (*parse_gsi)(struct acpi_madt_generic_interrupt *))
>> @@ -166,6 +166,40 @@ static inline void arm_spe_acpi_register_device(void)
>>  }
>>  #endif /* CONFIG_ARM_SPE_PMU */
>>  
>> +#ifdef CONFIG_CORESIGHT_TRBE
> Maybe we should use #if IS_ENABLED(CONFIG_CORESIGHT_TRBE) here and other places?
> 
> As trbe can be configured as a module, when CONFIG_CORESIGHT_TRBE=m this block
> won't be compiled. Referred to
> 
> https://github.com/torvalds/linux/blob/c7c90e121e992eefdf07945e5a6e9cf097b29463/include/linux/kconfig.h#L68

You are right, just making CONFIG_CORESIGHT_TRBE=m make it fall back on the
stub arm_trbe_acpi_register_device() definition, which does not create the
required dummy platform device thus preventing TRBE probe on the platform.

Will change the above #ifdef as IS_ENABLED() for CONFIG_CORESIGHT_TRBE and
revert back using IS_ENABLED() in the previous patch for CONFIG_ARM_SPE_PMU
as well. Thanks for noticing this problem.