diff mbox series

[03/13] mm: Add generic p?d_large() macros

Message ID 20190215170235.23360-4-steven.price@arm.com
State New, archived
Headers show
Series Convert x86 & arm64 to use generic page walk | expand

Commit Message

Steven Price Feb. 15, 2019, 5:02 p.m. UTC
From: James Morse <james.morse@arm.com>

Exposing the pud/pgd levels of the page tables to walk_page_range() means
we may come across the exotic large mappings that come with large areas
of contiguous memory (such as the kernel's linear map).

For architectures that don't provide p?d_large() macros, provided a
does nothing default.

Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Steven Price <steven.price@arm.com>
---
 include/asm-generic/pgtable.h | 10 ++++++++++
 1 file changed, 10 insertions(+)

Comments

Mark Rutland Feb. 18, 2019, 11:14 a.m. UTC | #1
On Fri, Feb 15, 2019 at 05:02:24PM +0000, Steven Price wrote:
> From: James Morse <james.morse@arm.com>
>
> Exposing the pud/pgd levels of the page tables to walk_page_range() means
> we may come across the exotic large mappings that come with large areas
> of contiguous memory (such as the kernel's linear map).
>
> For architectures that don't provide p?d_large() macros, provided a
> does nothing default.
>
> Signed-off-by: James Morse <james.morse@arm.com>
> Signed-off-by: Steven Price <steven.price@arm.com>
> ---
>  include/asm-generic/pgtable.h | 10 ++++++++++
>  1 file changed, 10 insertions(+)
>
> diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
> index 05e61e6c843f..7630d663cd51 100644
> --- a/include/asm-generic/pgtable.h
> +++ b/include/asm-generic/pgtable.h
> @@ -1186,4 +1186,14 @@ static inline bool arch_has_pfn_modify_check(void)
>  #define mm_pmd_folded(mm)__is_defined(__PAGETABLE_PMD_FOLDED)
>  #endif
>
> +#ifndef pgd_large
> +#define pgd_large(x)0
> +#endif
> +#ifndef pud_large
> +#define pud_large(x)0
> +#endif
> +#ifndef pmd_large
> +#define pmd_large(x)0
> +#endif

It might be worth a comment defining the semantics of these, e.g. how
they differ from p?d_huge() and p?d_trans_huge().

Thanks,
Mark.

> +
>  #endif /* _ASM_GENERIC_PGTABLE_H */
> --
> 2.20.1
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.
Peter Zijlstra Feb. 18, 2019, 11:31 a.m. UTC | #2
On Fri, Feb 15, 2019 at 05:02:24PM +0000, Steven Price wrote:
> From: James Morse <james.morse@arm.com>
> 
> Exposing the pud/pgd levels of the page tables to walk_page_range() means
> we may come across the exotic large mappings that come with large areas
> of contiguous memory (such as the kernel's linear map).
> 
> For architectures that don't provide p?d_large() macros, provided a
> does nothing default.

Kan was going to fix that for all archs I think..

See:

  http://lkml.kernel.org/r/20190204105409.GA17550@hirez.programming.kicks-ass.net

> Signed-off-by: James Morse <james.morse@arm.com>
> Signed-off-by: Steven Price <steven.price@arm.com>
> ---
>  include/asm-generic/pgtable.h | 10 ++++++++++
>  1 file changed, 10 insertions(+)
> 
> diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
> index 05e61e6c843f..7630d663cd51 100644
> --- a/include/asm-generic/pgtable.h
> +++ b/include/asm-generic/pgtable.h
> @@ -1186,4 +1186,14 @@ static inline bool arch_has_pfn_modify_check(void)
>  #define mm_pmd_folded(mm)	__is_defined(__PAGETABLE_PMD_FOLDED)
>  #endif
>  
> +#ifndef pgd_large
> +#define pgd_large(x)	0
> +#endif
> +#ifndef pud_large
> +#define pud_large(x)	0
> +#endif
> +#ifndef pmd_large
> +#define pmd_large(x)	0
> +#endif
> +
>  #endif /* _ASM_GENERIC_PGTABLE_H */
> -- 
> 2.20.1
>
Peter Zijlstra Feb. 18, 2019, 12:53 p.m. UTC | #3
On Mon, Feb 18, 2019 at 11:14:23AM +0000, Mark Rutland wrote:
> > +#ifndef pgd_large
> > +#define pgd_large(x)0
> > +#endif
> > +#ifndef pud_large
> > +#define pud_large(x)0
> > +#endif
> > +#ifndef pmd_large
> > +#define pmd_large(x)0
> > +#endif
> 
> It might be worth a comment defining the semantics of these, e.g. how
> they differ from p?d_huge() and p?d_trans_huge().

Yes; I took it to mean any large page mapping, so it would explicitly
include huge and thp.
Steven Price Feb. 18, 2019, 2:19 p.m. UTC | #4
On 18/02/2019 11:31, Peter Zijlstra wrote:
> On Fri, Feb 15, 2019 at 05:02:24PM +0000, Steven Price wrote:
>> From: James Morse <james.morse@arm.com>
>>
>> Exposing the pud/pgd levels of the page tables to walk_page_range() means
>> we may come across the exotic large mappings that come with large areas
>> of contiguous memory (such as the kernel's linear map).
>>
>> For architectures that don't provide p?d_large() macros, provided a
>> does nothing default.
> 
> Kan was going to fix that for all archs I think..

The latest series I can find from Kan is still x86 specific. I'm happy
to rebase onto something else if Kan has an implementation already
(please point me in the right direction). Otherwise Kan is obviously
free to base on these changes.

Steve

> See:
> 
>   http://lkml.kernel.org/r/20190204105409.GA17550@hirez.programming.kicks-ass.net
> 
>> Signed-off-by: James Morse <james.morse@arm.com>
>> Signed-off-by: Steven Price <steven.price@arm.com>
>> ---
>>  include/asm-generic/pgtable.h | 10 ++++++++++
>>  1 file changed, 10 insertions(+)
>>
>> diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
>> index 05e61e6c843f..7630d663cd51 100644
>> --- a/include/asm-generic/pgtable.h
>> +++ b/include/asm-generic/pgtable.h
>> @@ -1186,4 +1186,14 @@ static inline bool arch_has_pfn_modify_check(void)
>>  #define mm_pmd_folded(mm)	__is_defined(__PAGETABLE_PMD_FOLDED)
>>  #endif
>>  
>> +#ifndef pgd_large
>> +#define pgd_large(x)	0
>> +#endif
>> +#ifndef pud_large
>> +#define pud_large(x)	0
>> +#endif
>> +#ifndef pmd_large
>> +#define pmd_large(x)	0
>> +#endif
>> +
>>  #endif /* _ASM_GENERIC_PGTABLE_H */
>> -- 
>> 2.20.1
>>
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>
Steven Price Feb. 18, 2019, 2:20 p.m. UTC | #5
On 18/02/2019 12:53, Peter Zijlstra wrote:
> On Mon, Feb 18, 2019 at 11:14:23AM +0000, Mark Rutland wrote:
>>> +#ifndef pgd_large
>>> +#define pgd_large(x)0
>>> +#endif
>>> +#ifndef pud_large
>>> +#define pud_large(x)0
>>> +#endif
>>> +#ifndef pmd_large
>>> +#define pmd_large(x)0
>>> +#endif
>>
>> It might be worth a comment defining the semantics of these, e.g. how
>> they differ from p?d_huge() and p?d_trans_huge().
> 
> Yes; I took it to mean any large page mapping, so it would explicitly
> include huge and thp.

Yes your interpretation is correct. I'll add a comment explaining the
semantics.

Thanks,

Steve
Liang, Kan Feb. 19, 2019, 3:44 a.m. UTC | #6
On 2/18/2019 9:19 AM, Steven Price wrote:
> On 18/02/2019 11:31, Peter Zijlstra wrote:
>> On Fri, Feb 15, 2019 at 05:02:24PM +0000, Steven Price wrote:
>>> From: James Morse <james.morse@arm.com>
>>>
>>> Exposing the pud/pgd levels of the page tables to walk_page_range() means
>>> we may come across the exotic large mappings that come with large areas
>>> of contiguous memory (such as the kernel's linear map).
>>>
>>> For architectures that don't provide p?d_large() macros, provided a
>>> does nothing default.
>>
>> Kan was going to fix that for all archs I think..
>

Yes, I'm still working on a generic function to retrieve page size.
The generic p?d_large() issue has been fixed. However, I found that the 
pgd_page() is not generic either. I'm still working on it.
I will update you on the other thread when all issues are fixed.



> The latest series I can find from Kan is still x86 specific. I'm happy
> to rebase onto something else if Kan has an implementation already
> (please point me in the right direction). Otherwise Kan is obviously
> free to base on these changes.
>

My implementation is similar as yours. I'm happy to re-base on your changes.

Could you please also add a generic p4d_large()?

Thanks,
Kan

> Steve
> 
>> See:
>>
>>    http://lkml.kernel.org/r/20190204105409.GA17550@hirez.programming.kicks-ass.net
>>
>>> Signed-off-by: James Morse <james.morse@arm.com>
>>> Signed-off-by: Steven Price <steven.price@arm.com>
>>> ---
>>>   include/asm-generic/pgtable.h | 10 ++++++++++
>>>   1 file changed, 10 insertions(+)
>>>
>>> diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
>>> index 05e61e6c843f..7630d663cd51 100644
>>> --- a/include/asm-generic/pgtable.h
>>> +++ b/include/asm-generic/pgtable.h
>>> @@ -1186,4 +1186,14 @@ static inline bool arch_has_pfn_modify_check(void)
>>>   #define mm_pmd_folded(mm)	__is_defined(__PAGETABLE_PMD_FOLDED)
>>>   #endif
>>>   
>>> +#ifndef pgd_large
>>> +#define pgd_large(x)	0
>>> +#endif
>>> +#ifndef pud_large
>>> +#define pud_large(x)	0
>>> +#endif
>>> +#ifndef pmd_large
>>> +#define pmd_large(x)	0
>>> +#endif
>>> +
>>>   #endif /* _ASM_GENERIC_PGTABLE_H */
>>> -- 
>>> 2.20.1
>>>
>>
>> _______________________________________________
>> linux-arm-kernel mailing list
>> linux-arm-kernel@lists.infradead.org
>> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>>
>
Steven Price Feb. 20, 2019, 1:56 p.m. UTC | #7
On 19/02/2019 03:44, Liang, Kan wrote:
> 
> 
> On 2/18/2019 9:19 AM, Steven Price wrote:
>> On 18/02/2019 11:31, Peter Zijlstra wrote:
>>> On Fri, Feb 15, 2019 at 05:02:24PM +0000, Steven Price wrote:
>>>> From: James Morse <james.morse@arm.com>
>>>>
>>>> Exposing the pud/pgd levels of the page tables to walk_page_range()
>>>> means
>>>> we may come across the exotic large mappings that come with large areas
>>>> of contiguous memory (such as the kernel's linear map).
>>>>
>>>> For architectures that don't provide p?d_large() macros, provided a
>>>> does nothing default.
>>>
>>> Kan was going to fix that for all archs I think..
>>
> 
> Yes, I'm still working on a generic function to retrieve page size.
> The generic p?d_large() issue has been fixed. However, I found that the
> pgd_page() is not generic either. I'm still working on it.
> I will update you on the other thread when all issues are fixed.
> 
> 
> 
>> The latest series I can find from Kan is still x86 specific. I'm happy
>> to rebase onto something else if Kan has an implementation already
>> (please point me in the right direction). Otherwise Kan is obviously
>> free to base on these changes.
>>
> 
> My implementation is similar as yours. I'm happy to re-base on your
> changes.
> 
> Could you please also add a generic p4d_large()?

Sure, I'll include that in the next posting.

Thanks,

Steve
diff mbox series

Patch

diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
index 05e61e6c843f..7630d663cd51 100644
--- a/include/asm-generic/pgtable.h
+++ b/include/asm-generic/pgtable.h
@@ -1186,4 +1186,14 @@  static inline bool arch_has_pfn_modify_check(void)
 #define mm_pmd_folded(mm)	__is_defined(__PAGETABLE_PMD_FOLDED)
 #endif
 
+#ifndef pgd_large
+#define pgd_large(x)	0
+#endif
+#ifndef pud_large
+#define pud_large(x)	0
+#endif
+#ifndef pmd_large
+#define pmd_large(x)	0
+#endif
+
 #endif /* _ASM_GENERIC_PGTABLE_H */