Re: [PATCH] x86/mm/pti: in pti_clone_pgtable() don't increase addr by PUD_SIZE

[Song Liu]

> On Aug 20, 2019, at 6:55 AM, Rik van Riel <riel@xxxxxx> wrote:
> 
> On Tue, 2019-08-20 at 09:21 -0400, Song Liu wrote:
>>> On Aug 20, 2019, at 4:16 AM, Thomas Gleixner <tglx@xxxxxxxxxxxxx>
>>> wrote:
>>> 
>>> On Tue, 20 Aug 2019, Peter Zijlstra wrote:
>>>> What that code wants to do is skip to the end of the pud, a
>>>> pmd_size
>>>> increase will not do that. And right below this, there's a second
>>>> instance of this exact pattern.
>>>> 
>>>> Did I get the below right?
>>>> 
>>>> ---
>>>> diff --git a/arch/x86/mm/pti.c b/arch/x86/mm/pti.c
>>>> index b196524759ec..32b20b3cb227 100644
>>>> --- a/arch/x86/mm/pti.c
>>>> +++ b/arch/x86/mm/pti.c
>>>> @@ -330,12 +330,14 @@ pti_clone_pgtable(unsigned long start,
>>>> unsigned long end,
>>>> 
>>>> 		pud = pud_offset(p4d, addr);
>>>> 		if (pud_none(*pud)) {
>>>> +			addr &= PUD_MASK;
>>>> 			addr += PUD_SIZE;
>>> 
>>> 			round_up(addr, PUD_SIZE);
>> 
>> I guess we need "round_up(addr + PMD_SIZE, PUD_SIZE)". 
> 
> What does that do if start is less than PMD_SIZE
> away from the next PUD_SIZE boundary?

Great point!

> 
> How about:   round_up(addr + 1, PUD_SIZE)  ?

Yes. How about this?

=========================== 8< ============================