search for: ptep_get_and_clear

In shadow mode hypervisors, ptep_get_and_clear achieves the desired purpose of keeping the shadows in sync by issuing a native_get_and_clear, followed by a call to pte_update, which indicates the PTE has been modified. Direct mode hypervisors (Xen) have no need for this anyway, and will trap the update using writable pagetables. This means no...

In shadow mode hypervisors, ptep_get_and_clear achieves the desired purpose of keeping the shadows in sync by issuing a native_get_and_clear, followed by a call to pte_update, which indicates the PTE has been modified. Direct mode hypervisors (Xen) have no need for this anyway, and will trap the update using writable pagetables. This means no...

Hi all, This little series adds a new transaction-like abstraction for doing RMW updates to a pte, hooks it into paravirt_ops, and then makes use of it in Xen. The basic problem is that mprotect is very slow under Xen (up to 50x slower than native), primarily because of the ptent = ptep_get_and_clear(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ set_pte_at(mm, addr, pte, ptent); sequence in mm/mprotect.c:change_pte_range(). This is bad for Xen for two reasons: 1: ptep_get_and_clear() ends up being a xchg on the pte. Since the pte page is read-only (as it must be, b...

Hi all, This little series adds a new transaction-like abstraction for doing RMW updates to a pte, hooks it into paravirt_ops, and then makes use of it in Xen. The basic problem is that mprotect is very slow under Xen (up to 50x slower than native), primarily because of the ptent = ptep_get_and_clear(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ set_pte_at(mm, addr, pte, ptent); sequence in mm/mprotect.c:change_pte_range(). This is bad for Xen for two reasons: 1: ptep_get_and_clear() ends up being a xchg on the pte. Since the pte page is read-only (as it must be, b...

Hi all, This little series adds a new transaction-like abstraction for doing RMW updates to a pte, hooks it into paravirt_ops, and then makes use of it in Xen. The basic problem is that mprotect is very slow under Xen (up to 50x slower than native), primarily because of the ptent = ptep_get_and_clear(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ set_pte_at(mm, addr, pte, ptent); sequence in mm/mprotect.c:change_pte_range(). This is bad for Xen for two reasons: 1: ptep_get_and_clear() ends up being a xchg on the pte. Since the pte page is read-only (as it must be, b...

...ication interface. ] This little series adds a new transaction-like abstraction for doing RMW updates to a pte, hooks it into paravirt_ops, and then makes use of it in Xen. The basic problem is that mprotect is very slow under Xen (up to 50x slower than native), primarily because of the ptent = ptep_get_and_clear(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ set_pte_at(mm, addr, pte, ptent); sequence in mm/mprotect.c:change_pte_range(). This is bad for Xen for two reasons: 1: ptep_get_and_clear() ends up being a xchg on the pte. Since the pte page is read-only (as it must be, b...

...ication interface. ] This little series adds a new transaction-like abstraction for doing RMW updates to a pte, hooks it into paravirt_ops, and then makes use of it in Xen. The basic problem is that mprotect is very slow under Xen (up to 50x slower than native), primarily because of the ptent = ptep_get_and_clear(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ set_pte_at(mm, addr, pte, ptent); sequence in mm/mprotect.c:change_pte_range(). This is bad for Xen for two reasons: 1: ptep_get_and_clear() ends up being a xchg on the pte. Since the pte page is read-only (as it must be, b...

...ication interface. ] This little series adds a new transaction-like abstraction for doing RMW updates to a pte, hooks it into paravirt_ops, and then makes use of it in Xen. The basic problem is that mprotect is very slow under Xen (up to 50x slower than native), primarily because of the ptent = ptep_get_and_clear(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ set_pte_at(mm, addr, pte, ptent); sequence in mm/mprotect.c:change_pte_range(). This is bad for Xen for two reasons: 1: ptep_get_and_clear() ends up being a xchg on the pte. Since the pte page is read-only (as it must be, b...

...-2level.h Thu Oct 19 03:03:09 2006 -0700 +++ b/include/asm-i386/pgtable-2level.h Thu Oct 19 03:03:18 2006 -0700 @@ -22,8 +22,7 @@ #define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0) #define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0) -#define __HAVE_ARCH_PTEP_GET_AND_CLEAR -#define ptep_get_and_clear(mm,addr,xp) __pte(xchg(&(xp)->pte_low, 0)) +#define raw_ptep_get_and_clear(xp) __pte(xchg(&(xp)->pte_low, 0)) #define pte_page(x) pfn_to_page(pte_pfn(x)) #define pte_none(x) (!(x).pte_low) diff -r f1dd818c2f06 include/asm-i386/pgtable-3level.h --- a/i...

...-2level.h Thu Oct 19 03:03:09 2006 -0700 +++ b/include/asm-i386/pgtable-2level.h Thu Oct 19 03:03:18 2006 -0700 @@ -22,8 +22,7 @@ #define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0) #define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0) -#define __HAVE_ARCH_PTEP_GET_AND_CLEAR -#define ptep_get_and_clear(mm,addr,xp) __pte(xchg(&(xp)->pte_low, 0)) +#define raw_ptep_get_and_clear(xp) __pte(xchg(&(xp)->pte_low, 0)) #define pte_page(x) pfn_to_page(pte_pfn(x)) #define pte_none(x) (!(x).pte_low) diff -r f1dd818c2f06 include/asm-i386/pgtable-3level.h --- a/i...

...flush_tlb_single(addr); } - -#ifndef CONFIG_HIGHPTE -static void *native_kmap_atomic_pte(struct page *page, enum km_type type) -{ - return kmap_atomic(page, type); -} -#endif /* These are in entry.S */ extern void native_iret(void); @@ -324,7 +317,9 @@ struct paravirt_ops paravirt_ops = { .ptep_get_and_clear = native_ptep_get_and_clear, - .kmap_atomic_pte = native_kmap_atomic_pte, +#ifdef CONFIG_HIGHPTE + .kmap_atomic_pte = kmap_atomic, +#endif #ifdef CONFIG_X86_PAE .set_pte_atomic = native_set_pte_atomic, diff -r 5c03805411a6 include/asm-i386/highmem.h --- a/include/asm-i386/highmem.h Fri Apr 0...

...flush_tlb_single(addr); } - -#ifndef CONFIG_HIGHPTE -static void *native_kmap_atomic_pte(struct page *page, enum km_type type) -{ - return kmap_atomic(page, type); -} -#endif /* These are in entry.S */ extern void native_iret(void); @@ -324,7 +317,9 @@ struct paravirt_ops paravirt_ops = { .ptep_get_and_clear = native_ptep_get_and_clear, - .kmap_atomic_pte = native_kmap_atomic_pte, +#ifdef CONFIG_HIGHPTE + .kmap_atomic_pte = kmap_atomic, +#endif #ifdef CONFIG_X86_PAE .set_pte_atomic = native_set_pte_atomic, diff -r 5c03805411a6 include/asm-i386/highmem.h --- a/include/asm-i386/highmem.h Fri Apr 0...

...Thus, the following - * hook is made available. - */ -#define set_pte(pteptr, pteval) (*(pteptr) = pteval) -#define set_pte_at(mm,addr,ptep,pteval) set_pte(ptep,pteval) -#define set_pte_atomic(pteptr, pteval) set_pte(pteptr,pteval) -#define set_pmd(pmdptr, pmdval) (*(pmdptr) = (pmdval)) - -#define ptep_get_and_clear(mm,addr,xp) __pte(xchg(&(xp)->pte_low, 0)) #define pte_same(a, b) ((a).pte_low == (b).pte_low) #define pte_page(x) pfn_to_page(pte_pfn(x)) #define pte_none(x) (!(x).pte_low) @@ -61,4 +50,17 @@ static inline int pte_exec_kernel(pte_t #define __pte_to_swp_entry(pte) ((swp_entry_t) { (...

...Thus, the following - * hook is made available. - */ -#define set_pte(pteptr, pteval) (*(pteptr) = pteval) -#define set_pte_at(mm,addr,ptep,pteval) set_pte(ptep,pteval) -#define set_pte_atomic(pteptr, pteval) set_pte(pteptr,pteval) -#define set_pmd(pmdptr, pmdval) (*(pmdptr) = (pmdval)) - -#define ptep_get_and_clear(mm,addr,xp) __pte(xchg(&(xp)->pte_low, 0)) #define pte_same(a, b) ((a).pte_low == (b).pte_low) #define pte_page(x) pfn_to_page(pte_pfn(x)) #define pte_none(x) (!(x).pte_low) @@ -61,4 +50,17 @@ static inline int pte_exec_kernel(pte_t #define __pte_to_swp_entry(pte) ((swp_entry_t) { (...

...mpfn = migrate_pfn(pfn) | MIGRATE_PFN_MIGRATE; > - mpfn |= pte_write(pte) ? MIGRATE_PFN_WRITE : 0; > + if (pte_write(pte)) > + writable = true; > } > > /* FIXME support THP */ > @@ -2284,8 +2286,7 @@ static int migrate_vma_collect_pmd(pmd_t *pmdp, > ptep_get_and_clear(mm, addr, ptep); > > /* Setup special migration page table entry */ > - entry = make_migration_entry(page, mpfn & > - MIGRATE_PFN_WRITE); > + entry = make_migration_entry(page, writable); > swp_pte = swp_entry_to_pte(entry); > if (pte_soft_di...

...set_pmd +#define set_pud native_set_pud +#define set_pgd native_set_pgd +#define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0) +#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0) +#define pud_clear native_pud_clear +#define pgd_clear native_pgd_clear +#define ptep_get_and_clear(mm,addr,xp) __pte(xchg(&(xp)->pte, 0)) +#endif -static inline void set_pte(pte_t *dst, pte_t val) +static inline void native_set_pte(pte_t *dst, pte_t val) { - pte_val(*dst) = pte_val(val); + dst->pte = pte_val(val); } -#define set_pte_at(mm,addr,ptep,pteval) set_pte(ptep,pteval) -...

...set_pmd +#define set_pud native_set_pud +#define set_pgd native_set_pgd +#define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0) +#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0) +#define pud_clear native_pud_clear +#define pgd_clear native_pgd_clear +#define ptep_get_and_clear(mm,addr,xp) __pte(xchg(&(xp)->pte, 0)) +#endif -static inline void set_pte(pte_t *dst, pte_t val) +static inline void native_set_pte(pte_t *dst, pte_t val) { - pte_val(*dst) = pte_val(val); + dst->pte = pte_val(val); } -#define set_pte_at(mm,addr,ptep,pteval) set_pte(ptep,pteval) -...

...make_pte(unsigned long pte) +{ + return (pte_t){ pte }; +} + +pud_t native_make_pud(unsigned long pud) +{ + return (pud_t){ pud }; +} + +pmd_t native_make_pmd(unsigned long pmd) +{ + return (pmd_t){ pmd }; +} + +pgd_t native_make_pgd(unsigned long pgd) +{ + return (pgd_t){ pgd }; +} + +pte_t native_ptep_get_and_clear(struct mm_struct *mm, u64 addr, + pte_t *ptep) +{ + return __pte(xchg(&(ptep)->pte, 0)); +} + +void native_set_pte_at(struct mm_struct *mm, u64 addr, pte_t *ptep, + pte_t pteval) +{ + native_set_pte(ptep,pteval); +} + +void native_pte_clear(struct mm_struct *mm, u64 addr, pte_t *ptep...

...make_pte(unsigned long pte) +{ + return (pte_t){ pte }; +} + +pud_t native_make_pud(unsigned long pud) +{ + return (pud_t){ pud }; +} + +pmd_t native_make_pmd(unsigned long pmd) +{ + return (pmd_t){ pmd }; +} + +pgd_t native_make_pgd(unsigned long pgd) +{ + return (pgd_t){ pgd }; +} + +pte_t native_ptep_get_and_clear(struct mm_struct *mm, u64 addr, + pte_t *ptep) +{ + return __pte(xchg(&(ptep)->pte, 0)); +} + +void native_set_pte_at(struct mm_struct *mm, u64 addr, pte_t *ptep, + pte_t pteval) +{ + native_set_pte(ptep,pteval); +} + +void native_pte_clear(struct mm_struct *mm, u64 addr, pte_t *ptep...

Chris Wright wrote: >* Zachary Amsden (zach@vmware.com) wrote: > > >>Does Xen assume page aligned descriptor tables? I assume from this >> >> > >Yes. > > > >>patch and snippets I have gathered from others, that is a yes, and other >>things here imply that DT pages are not shadowed. If so, Xen itself >>must have live segments