Re: [PATCH RFC v8 04/24] arm64: Introduce por_elx_set_pkey_perms() helper

[Kevin Brodsky <kevin.brodsky@arm.com>]

On 22/06/2026 15:16, David Hildenbrand (Arm) wrote:
> On 5/26/26 13:15, Kevin Brodsky wrote:
>> Introduce a helper that sets the permissions of a given pkey
>> (POIndex) in the POR_ELx format, and make use of it in
>> arch_set_user_pkey_access().
>>
>> Signed-off-by: Kevin Brodsky <kevin.brodsky@arm.com>
>> ---
>>  arch/arm64/include/asm/por.h |  7 +++++++
>>  arch/arm64/mm/mmu.c          | 26 ++++++++++----------------
>>  2 files changed, 17 insertions(+), 16 deletions(-)
>>
>> diff --git a/arch/arm64/include/asm/por.h b/arch/arm64/include/asm/por.h
>> index d913d5b529e4..bffb4d2b1246 100644
>> --- a/arch/arm64/include/asm/por.h
>> +++ b/arch/arm64/include/asm/por.h
>> @@ -31,4 +31,11 @@ static inline bool por_elx_allows_exec(u64 por, u8 pkey)
>>  	return perm & POE_X;
>>  }
>>  
>> +static inline u64 por_elx_set_pkey_perms(u64 por, u8 pkey, u64 perms)
>> +{
>> +	u64 shift = POR_ELx_PERM_SHIFT(pkey);
>> +
>> +	return (por & ~(POE_MASK << shift)) | (perms << shift);
>> +}
>> +
>>  #endif /* _ASM_ARM64_POR_H */
>> diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c
>> index dd85e093ffdb..493310cf0486 100644
>> --- a/arch/arm64/mm/mmu.c
>> +++ b/arch/arm64/mm/mmu.c
>> @@ -2339,8 +2339,8 @@ void __cpu_replace_ttbr1(pgd_t *pgdp, bool cnp)
>>  #ifdef CONFIG_ARCH_HAS_PKEYS
>>  int arch_set_user_pkey_access(int pkey, unsigned long init_val)
>>  {
>> -	u64 new_por;
>> -	u64 old_por;
>> +	u64 new_perms;
> You should spell out the renaming of the variable.

Sure, will do.

> Given that perms is 4bit per key, should we use a u8 for it instead?

We could, in fact I'm just noticing that this all very inconsistent
right now: the POE_* constants are defined as UL, but por_elx_allows_*
and other functions use u8. I'll clean that up in a separate patch and
use u8 here.

>> +	u64 por;
>>  
>>  	if (!system_supports_poe())
>>  		return -ENOSPC;
>> @@ -2354,25 +2354,19 @@ int arch_set_user_pkey_access(int pkey, unsigned long init_val)
>>  		return -EINVAL;
>>  
>>  	/* Set the bits we need in POR:  */
>> -	new_por = POE_RWX;
>> +	new_perms = POE_RWX;
>>  	if (init_val & PKEY_DISABLE_WRITE)
>> -		new_por &= ~POE_W;
>> +		new_perms &= ~POE_W;
>>  	if (init_val & PKEY_DISABLE_ACCESS)
>> -		new_por &= ~POE_RW;
>> +		new_perms &= ~POE_RW;
>>  	if (init_val & PKEY_DISABLE_READ)
>> -		new_por &= ~POE_R;
>> +		new_perms &= ~POE_R;
>>  	if (init_val & PKEY_DISABLE_EXECUTE)
>> -		new_por &= ~POE_X;
>> +		new_perms &= ~POE_X;
>>  
>> -	/* Shift the bits in to the correct place in POR for pkey: */
>> -	new_por = POR_ELx_PERM_PREP(pkey, new_por);
>> -
>> -	/* Get old POR and mask off any old bits in place: */
>> -	old_por = read_sysreg_s(SYS_POR_EL0);
>> -	old_por &= ~(POE_MASK << POR_ELx_PERM_SHIFT(pkey));
>> -
>> -	/* Write old part along with new part: */
>> -	write_sysreg_s(old_por | new_por, SYS_POR_EL0);
>> +	por = read_sysreg_s(SYS_POR_EL0);
>> +	por = por_elx_set_pkey_perms(por, pkey, new_perms);
>> +	write_sysreg_s(por, SYS_POR_EL0);
> Was wondering whether to move reading+writing of the register into the same
> helper, but you cannot reuse this exactly the same way for SYS_POR_EL1 as it seems.

It wouldn't be useful because the new user of the helper,
por_set_kpkeys_context(), doesn't itself read or write the register.

> Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>

Thanks!

- Kevin