Add a set of accessors to pack, unpack and modify page table entries (at all levels). This allows a paravirt implementation to control the contents of pgd/pmd/pte entries. For example, Xen uses this to convert the (pseudo-)physical address into a machine address when populating a pagetable entry, and converting back to pphys address when an entry is read. Signed-off-by: Jeremy Fitzhardinge -- arch/i386/kernel/paravirt.c | 113 ++++++++++++++++++++++++++++++++---- arch/i386/kernel/vmlinux.lds.S | 3 include/asm-i386/page.h | 18 ++++- include/asm-i386/paravirt.h | 68 +++++++++++++++++++++ include/asm-i386/pgtable-2level.h | 5 - include/asm-i386/pgtable-3level.h | 27 ++++---- 6 files changed, 199 insertions(+), 35 deletions(-) =================================================================== --- a/arch/i386/kernel/paravirt.c +++ b/arch/i386/kernel/paravirt.c @@ -34,7 +34,7 @@ #include /* nop stub */ -static void native_nop(void) +void native_nop(void) { } @@ -400,38 +400,76 @@ static void native_flush_tlb_single(u32 } #ifndef CONFIG_X86_PAE -static void native_set_pte(pte_t *ptep, pte_t pteval) +void native_set_pte(pte_t *ptep, pte_t pteval) { *ptep = pteval; } -static void native_set_pte_at(struct mm_struct *mm, u32 addr, pte_t *ptep, pte_t pteval) +void native_set_pte_at(struct mm_struct *mm, u32 addr, + pte_t *ptep, pte_t pteval) { *ptep = pteval; } -static void native_set_pmd(pmd_t *pmdp, pmd_t pmdval) +void native_set_pmd(pmd_t *pmdp, pmd_t pmdval) { *pmdp = pmdval; } +unsigned long native_pte_val(pte_t pte) +{ + return pte.pte_low; +} + +unsigned long native_pmd_val(pmd_t pmd) +{ + BUG(); + return 0; +} + +unsigned long native_pgd_val(pgd_t pgd) +{ + return pgd.pgd; +} + +pte_t native_make_pte(unsigned long pte) +{ + return (pte_t){ pte }; +} + +pmd_t native_make_pmd(unsigned long pmd) +{ + BUG(); +} + +pgd_t native_make_pgd(unsigned long pgd) +{ + return (pgd_t){ pgd }; +} + +pte_t native_ptep_get_and_clear(pte_t *ptep) +{ + return __pte(xchg(&(ptep)->pte_low, 0)); +} + #else /* CONFIG_X86_PAE */ -static void native_set_pte(pte_t *ptep, pte_t pte) +void native_set_pte(pte_t *ptep, pte_t pte) { ptep->pte_high = pte.pte_high; smp_wmb(); ptep->pte_low = pte.pte_low; } -static void native_set_pte_at(struct mm_struct *mm, u32 addr, pte_t *ptep, pte_t pte) +void native_set_pte_at(struct mm_struct *mm, u32 addr, pte_t *ptep, pte_t pte) { ptep->pte_high = pte.pte_high; smp_wmb(); ptep->pte_low = pte.pte_low; } -static void native_set_pte_present(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pte) +void native_set_pte_present(struct mm_struct *mm, u32 addr, + pte_t *ptep, pte_t pte) { ptep->pte_low = 0; smp_wmb(); @@ -440,35 +478,78 @@ static void native_set_pte_present(struc ptep->pte_low = pte.pte_low; } -static void native_set_pte_atomic(pte_t *ptep, pte_t pteval) +void native_set_pte_atomic(pte_t *ptep, pte_t pteval) { set_64bit((unsigned long long *)ptep,pte_val(pteval)); } -static void native_set_pmd(pmd_t *pmdp, pmd_t pmdval) +void native_set_pmd(pmd_t *pmdp, pmd_t pmdval) { set_64bit((unsigned long long *)pmdp,pmd_val(pmdval)); } -static void native_set_pud(pud_t *pudp, pud_t pudval) +void native_set_pud(pud_t *pudp, pud_t pudval) { *pudp = pudval; } -static void native_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) +void native_pte_clear(struct mm_struct *mm, u32 addr, pte_t *ptep) { ptep->pte_low = 0; smp_wmb(); ptep->pte_high = 0; } -static void native_pmd_clear(pmd_t *pmd) +void native_pmd_clear(pmd_t *pmd) { u32 *tmp = (u32 *)pmd; *tmp = 0; smp_wmb(); *(tmp + 1) = 0; } + +unsigned long long native_pte_val(pte_t pte) +{ + return pte.pte_low | ((unsigned long long)pte.pte_high << 32); +} + +unsigned long long native_pmd_val(pmd_t pmd) +{ + return pmd.pmd; +} + +unsigned long long native_pgd_val(pgd_t pgd) +{ + return pgd.pgd; +} + +pte_t native_make_pte(unsigned long long pte) +{ + return (pte_t){ pte }; +} + +pmd_t native_make_pmd(unsigned long long pmd) +{ + return (pmd_t){ pmd }; +} + +pgd_t native_make_pgd(unsigned long long pgd) +{ + return (pgd_t){ pgd }; +} + +pte_t native_ptep_get_and_clear(pte_t *ptep) +{ + pte_t res; + + /* xchg acts as a barrier before the setting of the high bits */ + res.pte_low = xchg(&ptep->pte_low, 0); + res.pte_high = ptep->pte_high; + ptep->pte_high = 0; + + return res; +} + #endif /* CONFIG_X86_PAE */ /* These are in entry.S */ @@ -561,6 +642,9 @@ struct paravirt_ops paravirt_ops = { .set_pmd = native_set_pmd, .pte_update = (void *)native_nop, .pte_update_defer = (void *)native_nop, + + .ptep_get_and_clear = native_ptep_get_and_clear, + #ifdef CONFIG_X86_PAE .set_pte_atomic = native_set_pte_atomic, .set_pte_present = native_set_pte_present, @@ -569,6 +653,14 @@ struct paravirt_ops paravirt_ops = { .pmd_clear = native_pmd_clear, #endif + .pte_val = native_pte_val, + .pmd_val = native_pmd_val, + .pgd_val = native_pgd_val, + + .make_pte = native_make_pte, + .make_pmd = native_make_pmd, + .make_pgd = native_make_pgd, + .irq_enable_sysexit = native_irq_enable_sysexit, .iret = native_iret, =================================================================== --- a/arch/i386/kernel/vmlinux.lds.S +++ b/arch/i386/kernel/vmlinux.lds.S @@ -21,6 +21,9 @@ #include #include #include + +#undef ENTRY +#undef ALIGN OUTPUT_FORMAT("elf32-i386", "elf32-i386", "elf32-i386") OUTPUT_ARCH(i386) =================================================================== --- a/include/asm-i386/page.h +++ b/include/asm-i386/page.h @@ -11,7 +11,6 @@ #ifdef __KERNEL__ #ifndef __ASSEMBLY__ - #ifdef CONFIG_X86_USE_3DNOW @@ -48,9 +47,11 @@ typedef struct { unsigned long long pmd; typedef struct { unsigned long long pmd; } pmd_t; typedef struct { unsigned long long pgd; } pgd_t; typedef struct { unsigned long long pgprot; } pgprot_t; +#ifndef CONFIG_PARAVIRT #define pmd_val(x) ((x).pmd) #define pte_val(x) ((x).pte_low | ((unsigned long long)(x).pte_high << 32)) #define __pmd(x) ((pmd_t) { (x) } ) +#endif /* CONFIG_PARAVIRT */ #define HPAGE_SHIFT 21 #include #else @@ -58,7 +59,9 @@ typedef struct { unsigned long pgd; } pg typedef struct { unsigned long pgd; } pgd_t; typedef struct { unsigned long pgprot; } pgprot_t; #define boot_pte_t pte_t /* or would you rather have a typedef */ +#ifndef CONFIG_PARAVIRT #define pte_val(x) ((x).pte_low) +#endif #define HPAGE_SHIFT 22 #include #endif @@ -71,12 +74,14 @@ typedef struct { unsigned long pgprot; } #define HAVE_ARCH_HUGETLB_UNMAPPED_AREA #endif +#define pgprot_val(x) ((x).pgprot) +#define __pgprot(x) ((pgprot_t) { (x) } ) + +#ifndef CONFIG_PARAVIRT #define pgd_val(x) ((x).pgd) -#define pgprot_val(x) ((x).pgprot) - #define __pte(x) ((pte_t) { (x) } ) #define __pgd(x) ((pgd_t) { (x) } ) -#define __pgprot(x) ((pgprot_t) { (x) } ) +#endif #endif /* !__ASSEMBLY__ */ @@ -143,6 +148,11 @@ extern int page_is_ram(unsigned long pag #include #include +#ifdef CONFIG_PARAVIRT +/* After pte_t, etc, have been defined */ +#include +#endif + #ifndef CONFIG_COMPAT_VDSO #define __HAVE_ARCH_GATE_AREA 1 #endif =================================================================== --- a/include/asm-i386/paravirt.h +++ b/include/asm-i386/paravirt.h @@ -25,6 +25,8 @@ #define CLBR_ANY 0x7 #ifndef __ASSEMBLY__ +#include + struct thread_struct; struct Xgt_desc_struct; struct tss_struct; @@ -140,12 +142,31 @@ struct paravirt_ops void (*set_pmd)(pmd_t *pmdp, pmd_t pmdval); void (*pte_update)(struct mm_struct *mm, u32 addr, pte_t *ptep); void (*pte_update_defer)(struct mm_struct *mm, u32 addr, pte_t *ptep); + + pte_t (*ptep_get_and_clear)(pte_t *ptep); + #ifdef CONFIG_X86_PAE void (*set_pte_atomic)(pte_t *ptep, pte_t pteval); - void (*set_pte_present)(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pte); + void (*set_pte_present)(struct mm_struct *mm, u32 addr, pte_t *ptep, pte_t pte); void (*set_pud)(pud_t *pudp, pud_t pudval); - void (*pte_clear)(struct mm_struct *mm, unsigned long addr, pte_t *ptep); + void (*pte_clear)(struct mm_struct *mm, u32 addr, pte_t *ptep); void (*pmd_clear)(pmd_t *pmdp); + + unsigned long long (*pte_val)(pte_t); + unsigned long long (*pmd_val)(pmd_t); + unsigned long long (*pgd_val)(pgd_t); + + pte_t (*make_pte)(unsigned long long pte); + pmd_t (*make_pmd)(unsigned long long pmd); + pgd_t (*make_pgd)(unsigned long long pgd); +#else + unsigned long (*pte_val)(pte_t); + unsigned long (*pmd_val)(pmd_t); + unsigned long (*pgd_val)(pgd_t); + + pte_t (*make_pte)(unsigned long pte); + pmd_t (*make_pmd)(unsigned long pmd); + pgd_t (*make_pgd)(unsigned long pgd); #endif void (*set_lazy_mode)(int mode); @@ -163,6 +184,24 @@ struct paravirt_ops __attribute__((__section__(".paravirtprobe"))) = fn extern struct paravirt_ops paravirt_ops; + +#ifdef CONFIG_X86_PAE +unsigned long long native_pte_val(pte_t); +unsigned long long native_pmd_val(pmd_t); +unsigned long long native_pgd_val(pgd_t); + +pte_t native_make_pte(unsigned long long pte); +pmd_t native_make_pmd(unsigned long long pmd); +pgd_t native_make_pgd(unsigned long long pgd); +#else +unsigned long native_pte_val(pte_t); +unsigned long native_pmd_val(pmd_t); +unsigned long native_pgd_val(pgd_t); + +pte_t native_make_pte(unsigned long pte); +pmd_t native_make_pmd(unsigned long pmd); +pgd_t native_make_pgd(unsigned long pgd); +#endif #define paravirt_enabled() (paravirt_ops.paravirt_enabled) @@ -215,6 +254,8 @@ static inline void __cpuid(unsigned int #define read_cr4() paravirt_ops.read_cr4() #define read_cr4_safe(x) paravirt_ops.read_cr4_safe() #define write_cr4(x) paravirt_ops.write_cr4(x) + +#define raw_ptep_get_and_clear(xp) (paravirt_ops.ptep_get_and_clear(xp)) static inline void raw_safe_halt(void) { @@ -297,6 +338,17 @@ static inline void halt(void) (paravirt_ops.write_idt_entry((dt), (entry), (low), (high))) #define set_iopl_mask(mask) (paravirt_ops.set_iopl_mask(mask)) +#define __pte(x) paravirt_ops.make_pte(x) +#define __pgd(x) paravirt_ops.make_pgd(x) + +#define pte_val(x) paravirt_ops.pte_val(x) +#define pgd_val(x) paravirt_ops.pgd_val(x) + +#ifdef CONFIG_X86_PAE +#define __pmd(x) paravirt_ops.make_pmd(x) +#define pmd_val(x) paravirt_ops.pmd_val(x) +#endif + /* The paravirtualized I/O functions */ static inline void slow_down_io(void) { paravirt_ops.io_delay(); @@ -336,6 +388,18 @@ static inline void setup_secondary_clock paravirt_ops.setup_secondary_clock(); } #endif + + +void native_set_pte(pte_t *ptep, pte_t pteval); +void native_set_pte_at(struct mm_struct *mm, u32 addr, pte_t *ptep, pte_t pteval); +void native_set_pmd(pmd_t *pmdp, pmd_t pmdval); +void native_set_pte_present(struct mm_struct *mm, u32 addr, pte_t *ptep, pte_t pte); +void native_set_pte_atomic(pte_t *ptep, pte_t pteval); +void native_set_pud(pud_t *pudp, pud_t pudval); +void native_pte_clear(struct mm_struct *mm, u32 addr, pte_t *ptep); +void native_pmd_clear(pmd_t *pmd); +pte_t native_ptep_get_and_clear(pte_t *ptep); +void native_nop(void); #ifdef CONFIG_SMP static inline void startup_ipi_hook(int phys_apicid, unsigned long start_eip, =================================================================== --- a/include/asm-i386/pgtable-2level.h +++ b/include/asm-i386/pgtable-2level.h @@ -15,6 +15,7 @@ #define set_pte(pteptr, pteval) (*(pteptr) = pteval) #define set_pte_at(mm,addr,ptep,pteval) set_pte(ptep,pteval) #define set_pmd(pmdptr, pmdval) (*(pmdptr) = (pmdval)) +#define raw_ptep_get_and_clear(xp) __pte(xchg(&(xp)->pte_low, 0)) #endif #define set_pte_atomic(pteptr, pteval) set_pte(pteptr,pteval) @@ -23,11 +24,9 @@ #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 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) -#define pte_pfn(x) ((unsigned long)(((x).pte_low >> PAGE_SHIFT))) +#define pte_pfn(x) (pte_val(x) >> PAGE_SHIFT) #define pfn_pte(pfn, prot) __pte(((pfn) << PAGE_SHIFT) | pgprot_val(prot)) #define pfn_pmd(pfn, prot) __pmd(((pfn) << PAGE_SHIFT) | pgprot_val(prot)) =================================================================== --- a/include/asm-i386/pgtable-3level.h +++ b/include/asm-i386/pgtable-3level.h @@ -98,6 +98,18 @@ static inline void pmd_clear(pmd_t *pmd) smp_wmb(); *(tmp + 1) = 0; } + +static inline pte_t raw_ptep_get_and_clear(pte_t *ptep) +{ + pte_t res; + + /* xchg acts as a barrier before the setting of the high bits */ + res.pte_low = xchg(&ptep->pte_low, 0); + res.pte_high = ptep->pte_high; + ptep->pte_high = 0; + + return res; +} #endif /* @@ -119,18 +131,6 @@ static inline void pud_clear (pud_t * pu #define pmd_offset(pud, address) ((pmd_t *) pud_page(*(pud)) + \ pmd_index(address)) -static inline pte_t raw_ptep_get_and_clear(pte_t *ptep) -{ - pte_t res; - - /* xchg acts as a barrier before the setting of the high bits */ - res.pte_low = xchg(&ptep->pte_low, 0); - res.pte_high = ptep->pte_high; - ptep->pte_high = 0; - - return res; -} - #define __HAVE_ARCH_PTE_SAME static inline int pte_same(pte_t a, pte_t b) { @@ -146,8 +146,7 @@ static inline int pte_none(pte_t pte) static inline unsigned long pte_pfn(pte_t pte) { - return (pte.pte_low >> PAGE_SHIFT) | - (pte.pte_high << (32 - PAGE_SHIFT)); + return pte_val(pte) >> PAGE_SHIFT; } extern unsigned long long __supported_pte_mask; -- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/