Re: [PATCH] x86/mm: fix return value of p[um]dp_set_access_flags

[Wei Yang]

On Fri, Sep 20, 2019 at 09:16:12AM -0700, Dave Hansen wrote:
>On 9/19/19 7:18 PM, Wei Yang wrote:
>> Last but not least, since update_mmu_cache_pmd is empty, even return
>> value is not correct, it doesn't break anything.
>
>In other words, this patch has no functional effect and does not provide
>a "fix".  What's the point of patching this stuff if it has no effect?

It correct the function semantics. The return value of these function doesn't
meet the requirement, which will be misleading and we still need to dig in the
history to find out the correct answer.

In case we would have a valid cache update mechanism, this would introduce a
problem.

So I suggest to fix the return value to reflect the true meaning.

-- 
Wei Yang
Help you, Help me