powerpc/8xx: use pmd_off() to access a PMD entry in pte_update()
Subsystems (2)
PATCH 1/1 powerpc/8xx: use pmd_off() to access a PMD entry in pte_update()
From: Mike Rapoport <[email protected]>
The pte_update() implementation for PPC_8xx unfolds page table from the PGD level to access a PMD entry. Since 8xx has only 2-level page table this can be simplified with pmd_off() shortcut.
Replace explicit unfolding with pmd_off() and drop defines of pgd_index() and pgd_offset() that are no longer needed.
---
I think it's powerpc material, but I won't mind if Andrew picks it up :)
arch/powerpc/include/asm/nohash/32/pgtable.h | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-)
arch/powerpc/include/asm/nohash/32/pgtable.h +3 -5
@@ -205,10 +205,6 @@ static inline void pmd_clear(pmd_t *pmdp) *pmdp =__pmd (0 ); } -/* to find an entry in a page-table-directory */ -#define pgd_index (address) ((address) >> PGDIR_SHIFT) -#define pgd_offset (mm, address) ((mm)->pgd + pgd_index(address)) -/* * PTE updates. This function is called whenever an existing * valid PTE is updated. This does -not- include set_pte_at()
@@ -230,6 +226,8 @@ static inline void pmd_clear(pmd_t *pmdp)* For other page sizes, we have a single entry in the table. */ #ifdef CONFIG_PPC_8xx +static pmd_t *pmd_off (struct mm_struct *mm,unsigned long addr); +static inline pte_basic_t pte_update (struct mm_struct *mm,unsigned long addr,pte_t *p,unsigned long clr,unsigned long set,int huge) {
@@ -237,7 +235,7 @@ static inline pte_basic_t pte_update(struct mm_struct *mm, unsigned long addr, ppte_basic_t old =pte_val (*p);pte_basic_t new = (old & ~(pte_basic_t )clr) | set;int num, i; -pmd_t *pmd =pmd_offset (pud_offset (p4d_offset (pgd_offset (mm, addr), addr), addr), addr); +pmd_t *pmd =pmd_off (mm, addr);if (!huge) num =PAGE_SIZE /SZ_4K ;
Signature · 1 line
2.26.2
Le 15/06/2020 à 11:22, Mike Rapoport a écrit :
Quoted 10 lines
From: Mike Rapoport <[email protected]>
The
pte_update()implementation forPPC_8xxunfolds page table from the PGD level to access a PMD entry. Since 8xx has only 2-level page table this can be simplified withpmd_off()shortcut.Replace explicit unfolding with
pmd_off()and drop defines ofpgd_index()andpgd_offset()that are no longer needed.Signed-off-by Mike Rapoport <[email protected]>
Nice. When I implemented this function, I was not able to use pmd_ptr() due to some circular header inclusion (See comment in commit log). Looks like since the replacement of pmd_ptr() by pmd_off(), it works.
Christophe
Quoted 42 lines
---
I think it's powerpc material, but I won't mind if Andrew picks it up :)
arch/powerpc/include/asm/nohash/32/pgtable.h | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-)arch/powerpc/include/asm/nohash/32/pgtable.h +3
-5@@ -205,10 +205,6 @@ static inline void pmd_clear(pmd_t *pmdp) *pmdp =__pmd (0 ); } -/* to find an entry in a page-table-directory */ -#define pgd_index (address) ((address) >> PGDIR_SHIFT) -#define pgd_offset (mm, address) ((mm)->pgd + pgd_index(address)) -/* * PTE updates. This function is called whenever an existing * valid PTE is updated. This does -not- include set_pte_at() @@ -230,6 +226,8 @@ static inline void pmd_clear(pmd_t *pmdp)* For other page sizes, we have a single entry in the table. */ #ifdef CONFIG_PPC_8xx +static pmd_t *pmd_off (struct mm_struct *mm,unsigned long addr); +static inline pte_basic_t pte_update (struct mm_struct *mm,unsigned long addr,pte_t *p,unsigned long clr,unsigned long set,int huge) {@@ -237,7 +235,7 @@ static inline pte_basic_t pte_update(struct mm_struct *mm, unsigned long addr, ppte_basic_t old =pte_val (*p);pte_basic_t new = (old & ~(pte_basic_t )clr) | set;int num, i; -pmd_t *pmd =pmd_offset (pud_offset (p4d_offset (pgd_offset (mm, addr), addr), addr), addr); +pmd_t *pmd =pmd_off (mm, addr);if (!huge) num =PAGE_SIZE /SZ_4K ;
Andrew Morton
maintainerQuoted 6 lines · On Mon, 15 Jun 2020 12:22:29 +0300 Mike Rapoport <[email protected]> wrote:
From: Mike Rapoport <[email protected]>
The
pte_update()implementation forPPC_8xxunfolds page table from the PGD level to access a PMD entry. Since 8xx has only 2-level page table this can be simplified withpmd_off()shortcut.
Quoted 7 lines
Replace explicit unfolding with
pmd_off()and drop defines ofpgd_index()andpgd_offset()that are no longer needed.Signed-off-by Mike Rapoport <[email protected]>---
I think it's powerpc material, but I won't mind if Andrew picks it up :)
Via the powerpc tree would be better, please.
Michael Ellerman
reviewerQuoted 11 lines · Andrew Morton
Quoted 9 lines · On Mon, 15 Jun 2020 12:22:29 +0300 Mike Rapoport <[email protected]> wrote:
From: Mike Rapoport <[email protected]>
The
pte_update()implementation forPPC_8xxunfolds page table from the PGD level to access a PMD entry. Since 8xx has only 2-level page table this can be simplified withpmd_off()shortcut.Replace explicit unfolding with
pmd_off()and drop defines ofpgd_index()andpgd_offset()that are no longer needed.
Quoted 6 lines
Quoted 4 lines
Signed-off-by Mike Rapoport <[email protected]>---
I think it's powerpc material, but I won't mind if Andrew picks it up :)
Via the powerpc tree would be better, please.
I'll take it into next for v5.9, unless there's a reason it needs to go into v5.8.
cheers
Quoted 15 lines · Michael Ellerman
Quoted 14 lines · Andrew Morton
Quoted 12 lines · On Mon, 15 Jun 2020 12:22:29 +0300 Mike Rapoport <[email protected]> wrote:
From: Mike Rapoport <[email protected]>
The
pte_update()implementation forPPC_8xxunfolds page table from the PGD level to access a PMD entry. Since 8xx has only 2-level page table this can be simplified withpmd_off()shortcut.Replace explicit unfolding with
pmd_off()and drop defines ofpgd_index()andpgd_offset()that are no longer needed.Signed-off-by Mike Rapoport <[email protected]>---
Quoted 6 lines
Quoted 3 lines
Quoted 1 line
I think it's powerpc material, but I won't mind if Andrew picks it up :)
Via the powerpc tree would be better, please.
I'll take it into next for v5.9, unless there's a reason it needs to go into v5.8.
I consider it a fixup for 5.8 merge window conflicts. Besides, mering it now may avoid new conflicts in 5.9 ;-)
Quoted 1 line
cheers
Signature · 2 lines
Sincerely yours, Mike.
Michael Ellerman
reviewerQuoted 18 lines · Mike Rapoport
Quoted 17 lines · Michael Ellerman
Quoted 16 lines · Andrew Morton
Quoted 13 lines · On Mon, 15 Jun 2020 12:22:29 +0300 Mike Rapoport <[email protected]> wrote:
From: Mike Rapoport <[email protected]>
The
pte_update()implementation forPPC_8xxunfolds page table from the PGD level to access a PMD entry. Since 8xx has only 2-level page table this can be simplified withpmd_off()shortcut.Replace explicit unfolding with
pmd_off()and drop defines ofpgd_index()andpgd_offset()that are no longer needed.Signed-off-by Mike Rapoport <[email protected]>---
I think it's powerpc material, but I won't mind if Andrew picks it up :)
Quoted 7 lines
Quoted 4 lines
Quoted 1 line
Via the powerpc tree would be better, please.
I'll take it into next for v5.9, unless there's a reason it needs to go into v5.8.
I consider it a fixup for 5.8 merge window conflicts. Besides, mering it now may avoid new conflicts in 5.9 ;-)
OK, I'll pick it up for v5.8.
cheers
Michael Ellerman
AppliedQuoted 6 lines · Mike Rapoport
The
pte_update()implementation forPPC_8xxunfolds page table from the PGD level to access a PMD entry. Since 8xx has only 2-level page table this can be simplified withpmd_off()shortcut.Replace explicit unfolding with
pmd_off()and drop defines ofpgd_index()andpgd_offset()that are no longer needed.
Applied to powerpc/fixes.
[1/1] powerpc/8xx: use pmd_off() to access a PMD entry in pte_update()
https://git.kernel.org/powerpc/c/687993ccf3b05070598b89fad97410b26d7bc9d2
cheers