search for: paravirt_lazy_mmu

...gntdev device. Each m2p_add/remove_override call issues a MULTI_grant_table_op and a __flush_tlb_single if kmap_op != NULL. Batching all the calls together is a great performance benefit because it means issuing one hypercall total rather than two hypercall per page. If paravirt_lazy_mode is set PARAVIRT_LAZY_MMU, all these calls are going to be batched together, otherwise they are issued one at a time. Adding arch_enter_lazy_mmu_mode/arch_leave_lazy_mmu_mode around the m2p_add/remove_override calls forces paravirt_lazy_mode to PARAVIRT_LAZY_MMU, therefore makes sure that they are always batched. Changes...

...te_percpu(paravirt_lazy_mode, mode); +} + +void paravirt_leave_lazy(enum paravirt_lazy_mode mode) +{ + BUG_ON(x86_read_percpu(paravirt_lazy_mode) != mode); + BUG_ON(preemptible()); + + x86_write_percpu(paravirt_lazy_mode, PARAVIRT_LAZY_NONE); +} + +void paravirt_enter_lazy_mmu(void) +{ + enter_lazy(PARAVIRT_LAZY_MMU); +} + +void paravirt_leave_lazy_mmu(void) +{ + paravirt_leave_lazy(PARAVIRT_LAZY_MMU); +} + +void paravirt_enter_lazy_cpu(void) +{ + enter_lazy(PARAVIRT_LAZY_CPU); +} + +void paravirt_leave_lazy_cpu(void) +{ + paravirt_leave_lazy(PARAVIRT_LAZY_CPU); +} + +enum paravirt_lazy_mode paravirt_get_lazy_...

...te_percpu(paravirt_lazy_mode, mode); +} + +void paravirt_leave_lazy(enum paravirt_lazy_mode mode) +{ + BUG_ON(x86_read_percpu(paravirt_lazy_mode) != mode); + BUG_ON(preemptible()); + + x86_write_percpu(paravirt_lazy_mode, PARAVIRT_LAZY_NONE); +} + +void paravirt_enter_lazy_mmu(void) +{ + enter_lazy(PARAVIRT_LAZY_MMU); +} + +void paravirt_leave_lazy_mmu(void) +{ + paravirt_leave_lazy(PARAVIRT_LAZY_MMU); +} + +void paravirt_enter_lazy_cpu(void) +{ + enter_lazy(PARAVIRT_LAZY_CPU); +} + +void paravirt_leave_lazy_cpu(void) +{ + paravirt_leave_lazy(PARAVIRT_LAZY_CPU); +} + +enum paravirt_lazy_mode paravirt_get_lazy_...

...zy_mode, PARAVIRT_LAZY_NONE); + (*ops->leave)(); +} + +static void flush_lazy(enum paravirt_lazy_mode mode, struct pv_lazy_ops *ops) +{ + if (x86_read_percpu(paravirt_lazy_mode) == mode) { + (*ops->leave)(); + (*ops->enter)(); + } +} + +void arch_enter_lazy_mmu_mode(void) +{ + enter_lazy(PARAVIRT_LAZY_MMU, &pv_mmu_ops.lazy_mode); +} + +void arch_leave_lazy_mmu_mode(void) +{ + leave_lazy(PARAVIRT_LAZY_MMU, &pv_mmu_ops.lazy_mode); +} + +void arch_flush_lazy_mmu_mode(void) +{ + flush_lazy(PARAVIRT_LAZY_MMU, &pv_mmu_ops.lazy_mode); +} + +void arch_enter_lazy_cpu_mode(void) +{ + enter_lazy(PA...

...zy_mode, PARAVIRT_LAZY_NONE); + (*ops->leave)(); +} + +static void flush_lazy(enum paravirt_lazy_mode mode, struct pv_lazy_ops *ops) +{ + if (x86_read_percpu(paravirt_lazy_mode) == mode) { + (*ops->leave)(); + (*ops->enter)(); + } +} + +void arch_enter_lazy_mmu_mode(void) +{ + enter_lazy(PARAVIRT_LAZY_MMU, &pv_mmu_ops.lazy_mode); +} + +void arch_leave_lazy_mmu_mode(void) +{ + leave_lazy(PARAVIRT_LAZY_MMU, &pv_mmu_ops.lazy_mode); +} + +void arch_flush_lazy_mmu_mode(void) +{ + flush_lazy(PARAVIRT_LAZY_MMU, &pv_mmu_ops.lazy_mode); +} + +void arch_enter_lazy_cpu_mode(void) +{ + enter_lazy(PA...

...7 +236,14 @@ *u = *update; } -void xen_set_pmd(pmd_t *ptr, pmd_t val) +static bool page_pinned(void *ptr) +{ + struct page *page = virt_to_page(ptr); + + return PagePinned(page); +} + +void xen_set_pmd_hyper(pmd_t *ptr, pmd_t val) { struct mmu_update u; @@ -251,6 +258,18 @@ xen_mc_issue(PARAVIRT_LAZY_MMU); preempt_enable(); +} + +void xen_set_pmd(pmd_t *ptr, pmd_t val) +{ + /* If page is not pinned, we can just update the entry + directly */ + if (!page_pinned(ptr)) { + *ptr = val; + return; + } + + xen_set_pmd_hyper(ptr, val); } /* @@ -380,7 +399,7 @@ return ret; } -void xen_set_...

...; + /* These two are jmp to, not actually called. */ void (fastcall *irq_enable_sysexit)(void); void (fastcall *iret)(void); @@ -386,6 +388,19 @@ static inline void pmd_clear(pmd_t *pmdp } #endif +/* Lazy mode for batching updates / context switch */ +#define PARAVIRT_LAZY_NONE 0 +#define PARAVIRT_LAZY_MMU 1 +#define PARAVIRT_LAZY_CPU 2 + +#define __HAVE_ARCH_ENTER_LAZY_CPU_MODE +#define arch_enter_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_CPU) +#define arch_leave_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_NONE) + +#define __HAVE_ARCH_ENTER_LAZY_MMU_MODE +#define arch...

...; + /* These two are jmp to, not actually called. */ void (fastcall *irq_enable_sysexit)(void); void (fastcall *iret)(void); @@ -386,6 +388,19 @@ static inline void pmd_clear(pmd_t *pmdp } #endif +/* Lazy mode for batching updates / context switch */ +#define PARAVIRT_LAZY_NONE 0 +#define PARAVIRT_LAZY_MMU 1 +#define PARAVIRT_LAZY_CPU 2 + +#define __HAVE_ARCH_ENTER_LAZY_CPU_MODE +#define arch_enter_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_CPU) +#define arch_leave_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_NONE) + +#define __HAVE_ARCH_ENTER_LAZY_MMU_MODE +#define arch...

...; + /* These two are jmp to, not actually called. */ void (fastcall *irq_enable_sysexit)(void); void (fastcall *iret)(void); @@ -386,6 +388,19 @@ static inline void pmd_clear(pmd_t *pmdp } #endif +/* Lazy mode for batching updates / context switch */ +#define PARAVIRT_LAZY_NONE 0 +#define PARAVIRT_LAZY_MMU 1 +#define PARAVIRT_LAZY_CPU 2 + +#define __HAVE_ARCH_ENTER_LAZY_CPU_MODE +#define arch_enter_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_CPU) +#define arch_leave_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_NONE) + +#define __HAVE_ARCH_ENTER_LAZY_MMU_MODE +#define arch...

...; + /* These two are jmp to, not actually called. */ void (fastcall *irq_enable_sysexit)(void); void (fastcall *iret)(void); @@ -386,6 +388,19 @@ static inline void pmd_clear(pmd_t *pmdp } #endif +/* Lazy mode for batching updates / context switch */ +#define PARAVIRT_LAZY_NONE 0 +#define PARAVIRT_LAZY_MMU 1 +#define PARAVIRT_LAZY_CPU 2 + +#define __HAVE_ARCH_ENTER_LAZY_CPU_MODE +#define arch_enter_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_CPU) +#define arch_leave_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_NONE) + +#define __HAVE_ARCH_ENTER_LAZY_MMU_MODE +#define arch...

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,

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,

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,

Hi all, [ Change since last post: change name to ptep_modify_prot_, on the grounds that it isn't really a general pte-modification 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),

Hi all, [ Change since last post: change name to ptep_modify_prot_, on the grounds that it isn't really a general pte-modification 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),

Hi all, [ Change since last post: change name to ptep_modify_prot_, on the grounds that it isn't really a general pte-modification 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),

Hi Linus, Here's a set of patches to update paravirt_ops and Xen for 2.6.24 A quick overview of the patchset: paravirt_ops: Remove the monolithic paravirt_ops structure, and replace it with smaller structures of related functions. Also, clean up the handling of lazy mode to make it easier to implement. x86/mm/init.c: remove a chunk of dead code Xen: - remove duplicate includes -

Hi Linus, Here's a set of patches to update paravirt_ops and Xen for 2.6.24 A quick overview of the patchset: paravirt_ops: Remove the monolithic paravirt_ops structure, and replace it with smaller structures of related functions. Also, clean up the handling of lazy mode to make it easier to implement. x86/mm/init.c: remove a chunk of dead code Xen: - remove duplicate includes -

Hi Linus, Here's a set of patches to update paravirt_ops and Xen for 2.6.24 A quick overview of the patchset: paravirt_ops: Remove the monolithic paravirt_ops structure, and replace it with smaller structures of related functions. Also, clean up the handling of lazy mode to make it easier to implement. x86/mm/init.c: remove a chunk of dead code Xen: - remove duplicate includes -

Hi folks, With this series, the bulk of the work of pvops64 is done. Here, I integrate most of the paravirt.c and paravirt.h files, making them applicable to both architectures. CONFIG_PARAVIRT is _not_ present yet. Basically, this code is missing page table integration (patches currently being worked on by Jeremy). Enjoy