Re: [PATCH 3/4] powerpc/kuap: Refactor static branch for disabling kuap

[Christophe Leroy]

Le 06/06/2023 à 11:16, Nicholas Piggin a écrit :
> On Mon Jun 5, 2023 at 9:04 PM AEST, Christophe Leroy wrote:
>> All but book3s/64 use a static branch key for disabling kuap.
>> book3s/64 uses a memory feature.
>>
>> Refactor all targets except book3s/64.
>>
>> Signed-off-by: Christophe Leroy <christophe.leroy@xxxxxxxxxx>
>> ---
>>   arch/powerpc/include/asm/book3s/32/kup.h     |  7 -------
>>   arch/powerpc/include/asm/book3s/64/kup.h     |  1 +
>>   arch/powerpc/include/asm/kup.h               | 15 +++++++++++++++
>>   arch/powerpc/include/asm/nohash/32/kup-8xx.h |  7 -------
>>   arch/powerpc/include/asm/nohash/kup-booke.h  |  7 -------
>>   arch/powerpc/mm/book3s32/kuap.c              |  3 ---
>>   arch/powerpc/mm/init-common.c                |  3 +++
>>   arch/powerpc/mm/nohash/kup.c                 |  3 ---
>>   8 files changed, 19 insertions(+), 27 deletions(-)
>>
>> diff --git a/arch/powerpc/include/asm/book3s/32/kup.h b/arch/powerpc/include/asm/book3s/32/kup.h
>> index 466a19cfb4df..8da9997a67ba 100644
>> --- a/arch/powerpc/include/asm/book3s/32/kup.h
>> +++ b/arch/powerpc/include/asm/book3s/32/kup.h
>> @@ -11,8 +11,6 @@
>>   
>>   #include <linux/jump_label.h>
>>   
>> -extern struct static_key_false disable_kuap_key;
>> -
>>   static __always_inline bool kuep_is_disabled(void)
>>   {
>>   	return !IS_ENABLED(CONFIG_PPC_KUEP);
>> @@ -25,11 +23,6 @@ static __always_inline bool kuep_is_disabled(void)
>>   #define KUAP_NONE	(~0UL)
>>   #define KUAP_ALL	(~1UL)
>>   
>> -static __always_inline bool kuap_is_disabled(void)
>> -{
>> -	return static_branch_unlikely(&disable_kuap_key);
>> -}
>> -
>>   static inline void kuap_lock_one(unsigned long addr)
>>   {
>>   	mtsr(mfsr(addr) | SR_KS, addr);
>> diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
>> index 1b0215ff3710..f8b8e93c488c 100644
>> --- a/arch/powerpc/include/asm/book3s/64/kup.h
>> +++ b/arch/powerpc/include/asm/book3s/64/kup.h
>> @@ -233,6 +233,7 @@ static __always_inline bool kuap_is_disabled(void)
>>   {
>>   	return !mmu_has_feature(MMU_FTR_BOOK3S_KUAP);
>>   }
>> +#define kuap_is_disabled kuap_is_disabled
> 
> Is there any point to doing this pattern since the code is in places
> that have ifdef PPC6 S etc?

I'm not sure what you have in mind.

There is a default kuap_is_disabled() in arch/powerpc/include/asm/kup.h
For 64s we want the version from 
arch/powerpc/include/asm/book3s/64/kup.h instead of the default one.

And kuap_is_disabled() is used in some common code in 
arch/powerpc/include/asm/kup.h, not only in the code encloded by #ifndef 64s

> 
>> diff --git a/arch/powerpc/mm/init-common.c b/arch/powerpc/mm/init-common.c
>> index 119ef491f797..74e140b1efef 100644
>> --- a/arch/powerpc/mm/init-common.c
>> +++ b/arch/powerpc/mm/init-common.c
>> @@ -32,6 +32,9 @@ EXPORT_SYMBOL_GPL(kernstart_virt_addr);
>>   bool disable_kuep = !IS_ENABLED(CONFIG_PPC_KUEP);
>>   bool disable_kuap = !IS_ENABLED(CONFIG_PPC_KUAP);
>>   
>> +struct static_key_false disable_kuap_key;
>> +EXPORT_SYMBOL(disable_kuap_key);
>> +
> 
> That's going to define it on 64s?

Is that a problem at all ?

By the way I was thinking about also using that key on 64s, will look at 
it as a second step.

Thanks for the review
Christophe

> 
> Nice refactoring though.
> 
> Thanks,
> Nick