summaryrefslogtreecommitdiff
path: root/arch/arm/mm/fault-armv.c
diff options
context:
space:
mode:
Diffstat (limited to 'arch/arm/mm/fault-armv.c')
-rw-r--r--arch/arm/mm/fault-armv.c92
1 files changed, 51 insertions, 41 deletions
diff --git a/arch/arm/mm/fault-armv.c b/arch/arm/mm/fault-armv.c
index ae857f41f68d..39fd5df73317 100644
--- a/arch/arm/mm/fault-armv.c
+++ b/arch/arm/mm/fault-armv.c
@@ -17,7 +17,6 @@
#include <asm/bugs.h>
#include <asm/cacheflush.h>
#include <asm/cachetype.h>
-#include <asm/pgtable.h>
#include <asm/tlbflush.h>
#include "mm.h"
@@ -62,45 +61,27 @@ static int do_adjust_pte(struct vm_area_struct *vma, unsigned long address,
return ret;
}
-#if USE_SPLIT_PTE_PTLOCKS
-/*
- * If we are using split PTE locks, then we need to take the page
- * lock here. Otherwise we are using shared mm->page_table_lock
- * which is already locked, thus cannot take it.
- */
-static inline void do_pte_lock(spinlock_t *ptl)
-{
- /*
- * Use nested version here to indicate that we are already
- * holding one similar spinlock.
- */
- spin_lock_nested(ptl, SINGLE_DEPTH_NESTING);
-}
-
-static inline void do_pte_unlock(spinlock_t *ptl)
-{
- spin_unlock(ptl);
-}
-#else /* !USE_SPLIT_PTE_PTLOCKS */
-static inline void do_pte_lock(spinlock_t *ptl) {}
-static inline void do_pte_unlock(spinlock_t *ptl) {}
-#endif /* USE_SPLIT_PTE_PTLOCKS */
-
static int adjust_pte(struct vm_area_struct *vma, unsigned long address,
- unsigned long pfn)
+ unsigned long pfn, bool need_lock)
{
spinlock_t *ptl;
pgd_t *pgd;
+ p4d_t *p4d;
pud_t *pud;
pmd_t *pmd;
pte_t *pte;
+ pmd_t pmdval;
int ret;
pgd = pgd_offset(vma->vm_mm, address);
if (pgd_none_or_clear_bad(pgd))
return 0;
- pud = pud_offset(pgd, address);
+ p4d = p4d_offset(pgd, address);
+ if (p4d_none_or_clear_bad(p4d))
+ return 0;
+
+ pud = pud_offset(p4d, address);
if (pud_none_or_clear_bad(pud))
return 0;
@@ -108,18 +89,32 @@ static int adjust_pte(struct vm_area_struct *vma, unsigned long address,
if (pmd_none_or_clear_bad(pmd))
return 0;
+again:
/*
* This is called while another page table is mapped, so we
* must use the nested version. This also means we need to
* open-code the spin-locking.
*/
- ptl = pte_lockptr(vma->vm_mm, pmd);
- pte = pte_offset_map(pmd, address);
- do_pte_lock(ptl);
+ pte = pte_offset_map_rw_nolock(vma->vm_mm, pmd, address, &pmdval, &ptl);
+ if (!pte)
+ return 0;
+
+ if (need_lock) {
+ /*
+ * Use nested version here to indicate that we are already
+ * holding one similar spinlock.
+ */
+ spin_lock_nested(ptl, SINGLE_DEPTH_NESTING);
+ if (unlikely(!pmd_same(pmdval, pmdp_get_lockless(pmd)))) {
+ pte_unmap_unlock(pte, ptl);
+ goto again;
+ }
+ }
ret = do_adjust_pte(vma, address, pfn, pte);
- do_pte_unlock(ptl);
+ if (need_lock)
+ spin_unlock(ptl);
pte_unmap(pte);
return ret;
@@ -127,8 +122,10 @@ static int adjust_pte(struct vm_area_struct *vma, unsigned long address,
static void
make_coherent(struct address_space *mapping, struct vm_area_struct *vma,
- unsigned long addr, pte_t *ptep, unsigned long pfn)
+ unsigned long addr, pte_t *ptep, unsigned long pfn)
{
+ const unsigned long pmd_start_addr = ALIGN_DOWN(addr, PMD_SIZE);
+ const unsigned long pmd_end_addr = pmd_start_addr + PMD_SIZE;
struct mm_struct *mm = vma->vm_mm;
struct vm_area_struct *mpnt;
unsigned long offset;
@@ -145,6 +142,14 @@ make_coherent(struct address_space *mapping, struct vm_area_struct *vma,
flush_dcache_mmap_lock(mapping);
vma_interval_tree_foreach(mpnt, &mapping->i_mmap, pgoff, pgoff) {
/*
+ * If we are using split PTE locks, then we need to take the pte
+ * lock. Otherwise we are using shared mm->page_table_lock which
+ * is already locked, thus cannot take it.
+ */
+ bool need_lock = IS_ENABLED(CONFIG_SPLIT_PTE_PTLOCKS);
+ unsigned long mpnt_addr;
+
+ /*
* If this VMA is not in our MM, we can ignore it.
* Note that we intentionally mask out the VMA
* that we are fixing up.
@@ -154,7 +159,12 @@ make_coherent(struct address_space *mapping, struct vm_area_struct *vma,
if (!(mpnt->vm_flags & VM_MAYSHARE))
continue;
offset = (pgoff - mpnt->vm_pgoff) << PAGE_SHIFT;
- aliases += adjust_pte(mpnt, mpnt->vm_start + offset, pfn);
+ mpnt_addr = mpnt->vm_start + offset;
+
+ /* Avoid deadlocks by not grabbing the same PTE lock again. */
+ if (mpnt_addr >= pmd_start_addr && mpnt_addr < pmd_end_addr)
+ need_lock = false;
+ aliases += adjust_pte(mpnt, mpnt_addr, pfn, need_lock);
}
flush_dcache_mmap_unlock(mapping);
if (aliases)
@@ -174,12 +184,12 @@ make_coherent(struct address_space *mapping, struct vm_area_struct *vma,
*
* Note that the pte lock will be held.
*/
-void update_mmu_cache(struct vm_area_struct *vma, unsigned long addr,
- pte_t *ptep)
+void update_mmu_cache_range(struct vm_fault *vmf, struct vm_area_struct *vma,
+ unsigned long addr, pte_t *ptep, unsigned int nr)
{
unsigned long pfn = pte_pfn(*ptep);
struct address_space *mapping;
- struct page *page;
+ struct folio *folio;
if (!pfn_valid(pfn))
return;
@@ -188,13 +198,13 @@ void update_mmu_cache(struct vm_area_struct *vma, unsigned long addr,
* The zero page is never written to, so never has any dirty
* cache lines, and therefore never needs to be flushed.
*/
- page = pfn_to_page(pfn);
- if (page == ZERO_PAGE(0))
+ if (is_zero_pfn(pfn))
return;
- mapping = page_mapping_file(page);
- if (!test_and_set_bit(PG_dcache_clean, &page->flags))
- __flush_dcache_page(mapping, page);
+ folio = page_folio(pfn_to_page(pfn));
+ mapping = folio_flush_mapping(folio);
+ if (!test_and_set_bit(PG_dcache_clean, &folio->flags))
+ __flush_dcache_folio(mapping, folio);
if (mapping) {
if (cache_is_vivt())
make_coherent(mapping, vma, addr, ptep, pfn);