Re: [PATCH v4 07/11] xen/arm: bitops: Implement a ffsll function

[Rahul Singh <Rahul.Singh@arm.com>]

Hello Julien,

> On 15 Jan 2021, at 12:16 pm, Julien Grall <julien@xen.org> wrote:
> 
> Hi Rahul,
> 
> On 08/01/2021 14:46, Rahul Singh wrote:
>> Implement the ffsll based on built-in function "__builtin_ffsll()"
>> ffsll will return one plus the index of the least significant 1-bit in
>> doublewords or if doublewords is zero, returns zero.
>> Signed-off-by: Rahul Singh <rahul.singh@arm.com>
>> ---
>> Changes in V4:
>>  - This patch is introduce in this verison.
>> ---
>>  xen/include/asm-arm/bitops.h | 12 ++++++++++++
>>  1 file changed, 12 insertions(+)
>> diff --git a/xen/include/asm-arm/bitops.h b/xen/include/asm-arm/bitops.h
>> index 71ae14cab3..7f83ee1828 100644
>> --- a/xen/include/asm-arm/bitops.h
>> +++ b/xen/include/asm-arm/bitops.h
>> @@ -170,6 +170,18 @@ static inline unsigned int find_first_set_bit(unsigned long word)
>>          return ffsl(word) - 1;
>>  }
>>  +/**
>> + * ffsll - find the first least significant set bit
>> + * @doubleword: double word to search
>> + *
>> + * Returns one plus the index of the least significant 1-bit in @doubleword
>> + * or if doubleword is zero, returns zero.
>> + */
>> +static inline int ffsll(long long doubleword)
> 
> If I am not mistaken, we already have an helper doing exactly the same (see ffs64() in xen/bitops.h).
> 
> Can you check if it is effectively the case and use it?

I checked we can use the ffs64() for SMMUv3 driver as it is doing exactly the same as ffsll().
I will modify the code to use the ffs64() in place of ffsll().

Regards,
Rahul
> 
> Cheers,
> 
> -- 
> Julien Grall