diff options
Diffstat (limited to 'arch/x86/xen/mmu_pv.c')
-rw-r--r-- | arch/x86/xen/mmu_pv.c | 356 |
1 files changed, 39 insertions, 317 deletions
diff --git a/arch/x86/xen/mmu_pv.c b/arch/x86/xen/mmu_pv.c index a58d9c69807a..ff99e20a84e3 100644 --- a/arch/x86/xen/mmu_pv.c +++ b/arch/x86/xen/mmu_pv.c @@ -86,19 +86,8 @@ #include "mmu.h" #include "debugfs.h" -#ifdef CONFIG_X86_32 -/* - * Identity map, in addition to plain kernel map. This needs to be - * large enough to allocate page table pages to allocate the rest. - * Each page can map 2MB. - */ -#define LEVEL1_IDENT_ENTRIES (PTRS_PER_PTE * 4) -static RESERVE_BRK_ARRAY(pte_t, level1_ident_pgt, LEVEL1_IDENT_ENTRIES); -#endif -#ifdef CONFIG_X86_64 /* l3 pud for userspace vsyscall mapping */ static pud_t level3_user_vsyscall[PTRS_PER_PUD] __page_aligned_bss; -#endif /* CONFIG_X86_64 */ /* * Protects atomic reservation decrease/increase against concurrent increases. @@ -280,10 +269,7 @@ static inline void __xen_set_pte(pte_t *ptep, pte_t pteval) if (!xen_batched_set_pte(ptep, pteval)) { /* * Could call native_set_pte() here and trap and - * emulate the PTE write but with 32-bit guests this - * needs two traps (one for each of the two 32-bit - * words in the PTE) so do one hypercall directly - * instead. + * emulate the PTE write, but a hypercall is much cheaper. */ struct mmu_update u; @@ -439,26 +425,6 @@ static void xen_set_pud(pud_t *ptr, pud_t val) xen_set_pud_hyper(ptr, val); } -#ifdef CONFIG_X86_PAE -static void xen_set_pte_atomic(pte_t *ptep, pte_t pte) -{ - trace_xen_mmu_set_pte_atomic(ptep, pte); - __xen_set_pte(ptep, pte); -} - -static void xen_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) -{ - trace_xen_mmu_pte_clear(mm, addr, ptep); - __xen_set_pte(ptep, native_make_pte(0)); -} - -static void xen_pmd_clear(pmd_t *pmdp) -{ - trace_xen_mmu_pmd_clear(pmdp); - set_pmd(pmdp, __pmd(0)); -} -#endif /* CONFIG_X86_PAE */ - __visible pmd_t xen_make_pmd(pmdval_t pmd) { pmd = pte_pfn_to_mfn(pmd); @@ -466,7 +432,6 @@ __visible pmd_t xen_make_pmd(pmdval_t pmd) } PV_CALLEE_SAVE_REGS_THUNK(xen_make_pmd); -#ifdef CONFIG_X86_64 __visible pudval_t xen_pud_val(pud_t pud) { return pte_mfn_to_pfn(pud.pud); @@ -571,7 +536,6 @@ __visible p4d_t xen_make_p4d(p4dval_t p4d) } PV_CALLEE_SAVE_REGS_THUNK(xen_make_p4d); #endif /* CONFIG_PGTABLE_LEVELS >= 5 */ -#endif /* CONFIG_X86_64 */ static int xen_pmd_walk(struct mm_struct *mm, pmd_t *pmd, int (*func)(struct mm_struct *mm, struct page *, enum pt_level), @@ -636,11 +600,8 @@ static int xen_p4d_walk(struct mm_struct *mm, p4d_t *p4d, * will be STACK_TOP_MAX, but at boot we need to pin up to * FIXADDR_TOP. * - * For 32-bit the important bit is that we don't pin beyond there, - * because then we start getting into Xen's ptes. - * - * For 64-bit, we must skip the Xen hole in the middle of the address - * space, just after the big x86-64 virtual hole. + * We must skip the Xen hole in the middle of the address space, just after + * the big x86-64 virtual hole. */ static int __xen_pgd_walk(struct mm_struct *mm, pgd_t *pgd, int (*func)(struct mm_struct *mm, struct page *, @@ -654,14 +615,12 @@ static int __xen_pgd_walk(struct mm_struct *mm, pgd_t *pgd, limit--; BUG_ON(limit >= FIXADDR_TOP); -#ifdef CONFIG_X86_64 /* * 64-bit has a great big hole in the middle of the address * space, which contains the Xen mappings. */ hole_low = pgd_index(GUARD_HOLE_BASE_ADDR); hole_high = pgd_index(GUARD_HOLE_END_ADDR); -#endif nr = pgd_index(limit) + 1; for (i = 0; i < nr; i++) { @@ -787,6 +746,8 @@ static int xen_pin_page(struct mm_struct *mm, struct page *page, read-only, and can be pinned. */ static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd) { + pgd_t *user_pgd = xen_get_user_pgd(pgd); + trace_xen_mmu_pgd_pin(mm, pgd); xen_mc_batch(); @@ -800,26 +761,14 @@ static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd) xen_mc_batch(); } -#ifdef CONFIG_X86_64 - { - pgd_t *user_pgd = xen_get_user_pgd(pgd); + xen_do_pin(MMUEXT_PIN_L4_TABLE, PFN_DOWN(__pa(pgd))); - xen_do_pin(MMUEXT_PIN_L4_TABLE, PFN_DOWN(__pa(pgd))); - - if (user_pgd) { - xen_pin_page(mm, virt_to_page(user_pgd), PT_PGD); - xen_do_pin(MMUEXT_PIN_L4_TABLE, - PFN_DOWN(__pa(user_pgd))); - } + if (user_pgd) { + xen_pin_page(mm, virt_to_page(user_pgd), PT_PGD); + xen_do_pin(MMUEXT_PIN_L4_TABLE, + PFN_DOWN(__pa(user_pgd))); } -#else /* CONFIG_X86_32 */ -#ifdef CONFIG_X86_PAE - /* Need to make sure unshared kernel PMD is pinnable */ - xen_pin_page(mm, pgd_page(pgd[pgd_index(TASK_SIZE)]), - PT_PMD); -#endif - xen_do_pin(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(pgd))); -#endif /* CONFIG_X86_64 */ + xen_mc_issue(0); } @@ -870,9 +819,7 @@ static int __init xen_mark_pinned(struct mm_struct *mm, struct page *page, static void __init xen_after_bootmem(void) { static_branch_enable(&xen_struct_pages_ready); -#ifdef CONFIG_X86_64 SetPagePinned(virt_to_page(level3_user_vsyscall)); -#endif xen_pgd_walk(&init_mm, xen_mark_pinned, FIXADDR_TOP); } @@ -919,29 +866,19 @@ static int xen_unpin_page(struct mm_struct *mm, struct page *page, /* Release a pagetables pages back as normal RW */ static void __xen_pgd_unpin(struct mm_struct *mm, pgd_t *pgd) { + pgd_t *user_pgd = xen_get_user_pgd(pgd); + trace_xen_mmu_pgd_unpin(mm, pgd); xen_mc_batch(); xen_do_pin(MMUEXT_UNPIN_TABLE, PFN_DOWN(__pa(pgd))); -#ifdef CONFIG_X86_64 - { - pgd_t *user_pgd = xen_get_user_pgd(pgd); - - if (user_pgd) { - xen_do_pin(MMUEXT_UNPIN_TABLE, - PFN_DOWN(__pa(user_pgd))); - xen_unpin_page(mm, virt_to_page(user_pgd), PT_PGD); - } + if (user_pgd) { + xen_do_pin(MMUEXT_UNPIN_TABLE, + PFN_DOWN(__pa(user_pgd))); + xen_unpin_page(mm, virt_to_page(user_pgd), PT_PGD); } -#endif - -#ifdef CONFIG_X86_PAE - /* Need to make sure unshared kernel PMD is unpinned */ - xen_unpin_page(mm, pgd_page(pgd[pgd_index(TASK_SIZE)]), - PT_PMD); -#endif __xen_pgd_walk(mm, pgd, xen_unpin_page, USER_LIMIT); @@ -1089,7 +1026,6 @@ static void __init pin_pagetable_pfn(unsigned cmd, unsigned long pfn) BUG(); } -#ifdef CONFIG_X86_64 static void __init xen_cleanhighmap(unsigned long vaddr, unsigned long vaddr_end) { @@ -1273,17 +1209,15 @@ static void __init xen_pagetable_cleanhighmap(void) xen_cleanhighmap(addr, roundup(addr + size, PMD_SIZE * 2)); xen_start_info->pt_base = (unsigned long)__va(__pa(xen_start_info->pt_base)); } -#endif static void __init xen_pagetable_p2m_setup(void) { xen_vmalloc_p2m_tree(); -#ifdef CONFIG_X86_64 xen_pagetable_p2m_free(); xen_pagetable_cleanhighmap(); -#endif + /* And revector! Bye bye old array */ xen_start_info->mfn_list = (unsigned long)xen_p2m_addr; } @@ -1420,6 +1354,8 @@ static void __xen_write_cr3(bool kernel, unsigned long cr3) } static void xen_write_cr3(unsigned long cr3) { + pgd_t *user_pgd = xen_get_user_pgd(__va(cr3)); + BUG_ON(preemptible()); xen_mc_batch(); /* disables interrupts */ @@ -1430,20 +1366,14 @@ static void xen_write_cr3(unsigned long cr3) __xen_write_cr3(true, cr3); -#ifdef CONFIG_X86_64 - { - pgd_t *user_pgd = xen_get_user_pgd(__va(cr3)); - if (user_pgd) - __xen_write_cr3(false, __pa(user_pgd)); - else - __xen_write_cr3(false, 0); - } -#endif + if (user_pgd) + __xen_write_cr3(false, __pa(user_pgd)); + else + __xen_write_cr3(false, 0); xen_mc_issue(PARAVIRT_LAZY_CPU); /* interrupts restored */ } -#ifdef CONFIG_X86_64 /* * At the start of the day - when Xen launches a guest, it has already * built pagetables for the guest. We diligently look over them @@ -1478,49 +1408,39 @@ static void __init xen_write_cr3_init(unsigned long cr3) xen_mc_issue(PARAVIRT_LAZY_CPU); /* interrupts restored */ } -#endif static int xen_pgd_alloc(struct mm_struct *mm) { pgd_t *pgd = mm->pgd; - int ret = 0; + struct page *page = virt_to_page(pgd); + pgd_t *user_pgd; + int ret = -ENOMEM; BUG_ON(PagePinned(virt_to_page(pgd))); + BUG_ON(page->private != 0); -#ifdef CONFIG_X86_64 - { - struct page *page = virt_to_page(pgd); - pgd_t *user_pgd; - - BUG_ON(page->private != 0); - - ret = -ENOMEM; - - user_pgd = (pgd_t *)__get_free_page(GFP_KERNEL | __GFP_ZERO); - page->private = (unsigned long)user_pgd; + user_pgd = (pgd_t *)__get_free_page(GFP_KERNEL | __GFP_ZERO); + page->private = (unsigned long)user_pgd; - if (user_pgd != NULL) { + if (user_pgd != NULL) { #ifdef CONFIG_X86_VSYSCALL_EMULATION - user_pgd[pgd_index(VSYSCALL_ADDR)] = - __pgd(__pa(level3_user_vsyscall) | _PAGE_TABLE); + user_pgd[pgd_index(VSYSCALL_ADDR)] = + __pgd(__pa(level3_user_vsyscall) | _PAGE_TABLE); #endif - ret = 0; - } - - BUG_ON(PagePinned(virt_to_page(xen_get_user_pgd(pgd)))); + ret = 0; } -#endif + + BUG_ON(PagePinned(virt_to_page(xen_get_user_pgd(pgd)))); + return ret; } static void xen_pgd_free(struct mm_struct *mm, pgd_t *pgd) { -#ifdef CONFIG_X86_64 pgd_t *user_pgd = xen_get_user_pgd(pgd); if (user_pgd) free_page((unsigned long)user_pgd); -#endif } /* @@ -1539,7 +1459,6 @@ static void xen_pgd_free(struct mm_struct *mm, pgd_t *pgd) */ __visible pte_t xen_make_pte_init(pteval_t pte) { -#ifdef CONFIG_X86_64 unsigned long pfn; /* @@ -1553,7 +1472,7 @@ __visible pte_t xen_make_pte_init(pteval_t pte) pfn >= xen_start_info->first_p2m_pfn && pfn < xen_start_info->first_p2m_pfn + xen_start_info->nr_p2m_frames) pte &= ~_PAGE_RW; -#endif + pte = pte_pfn_to_mfn(pte); return native_make_pte(pte); } @@ -1561,13 +1480,6 @@ PV_CALLEE_SAVE_REGS_THUNK(xen_make_pte_init); static void __init xen_set_pte_init(pte_t *ptep, pte_t pte) { -#ifdef CONFIG_X86_32 - /* If there's an existing pte, then don't allow _PAGE_RW to be set */ - if (pte_mfn(pte) != INVALID_P2M_ENTRY - && pte_val_ma(*ptep) & _PAGE_PRESENT) - pte = __pte_ma(((pte_val_ma(*ptep) & _PAGE_RW) | ~_PAGE_RW) & - pte_val_ma(pte)); -#endif __xen_set_pte(ptep, pte); } @@ -1702,7 +1614,6 @@ static void xen_release_pmd(unsigned long pfn) xen_release_ptpage(pfn, PT_PMD); } -#ifdef CONFIG_X86_64 static void xen_alloc_pud(struct mm_struct *mm, unsigned long pfn) { xen_alloc_ptpage(mm, pfn, PT_PUD); @@ -1712,20 +1623,6 @@ static void xen_release_pud(unsigned long pfn) { xen_release_ptpage(pfn, PT_PUD); } -#endif - -void __init xen_reserve_top(void) -{ -#ifdef CONFIG_X86_32 - unsigned long top = HYPERVISOR_VIRT_START; - struct xen_platform_parameters pp; - - if (HYPERVISOR_xen_version(XENVER_platform_parameters, &pp) == 0) - top = pp.virt_start; - - reserve_top_address(-top); -#endif /* CONFIG_X86_32 */ -} /* * Like __va(), but returns address in the kernel mapping (which is @@ -1733,11 +1630,7 @@ void __init xen_reserve_top(void) */ static void * __init __ka(phys_addr_t paddr) { -#ifdef CONFIG_X86_64 return (void *)(paddr + __START_KERNEL_map); -#else - return __va(paddr); -#endif } /* Convert a machine address to physical address */ @@ -1771,56 +1664,7 @@ static void __init set_page_prot(void *addr, pgprot_t prot) { return set_page_prot_flags(addr, prot, UVMF_NONE); } -#ifdef CONFIG_X86_32 -static void __init xen_map_identity_early(pmd_t *pmd, unsigned long max_pfn) -{ - unsigned pmdidx, pteidx; - unsigned ident_pte; - unsigned long pfn; - - level1_ident_pgt = extend_brk(sizeof(pte_t) * LEVEL1_IDENT_ENTRIES, - PAGE_SIZE); - - ident_pte = 0; - pfn = 0; - for (pmdidx = 0; pmdidx < PTRS_PER_PMD && pfn < max_pfn; pmdidx++) { - pte_t *pte_page; - - /* Reuse or allocate a page of ptes */ - if (pmd_present(pmd[pmdidx])) - pte_page = m2v(pmd[pmdidx].pmd); - else { - /* Check for free pte pages */ - if (ident_pte == LEVEL1_IDENT_ENTRIES) - break; - - pte_page = &level1_ident_pgt[ident_pte]; - ident_pte += PTRS_PER_PTE; - - pmd[pmdidx] = __pmd(__pa(pte_page) | _PAGE_TABLE); - } - - /* Install mappings */ - for (pteidx = 0; pteidx < PTRS_PER_PTE; pteidx++, pfn++) { - pte_t pte; - - if (pfn > max_pfn_mapped) - max_pfn_mapped = pfn; - - if (!pte_none(pte_page[pteidx])) - continue; - - pte = pfn_pte(pfn, PAGE_KERNEL_EXEC); - pte_page[pteidx] = pte; - } - } - for (pteidx = 0; pteidx < ident_pte; pteidx += PTRS_PER_PTE) - set_page_prot(&level1_ident_pgt[pteidx], PAGE_KERNEL_RO); - - set_page_prot(pmd, PAGE_KERNEL_RO); -} -#endif void __init xen_setup_machphys_mapping(void) { struct xen_machphys_mapping mapping; @@ -1831,13 +1675,8 @@ void __init xen_setup_machphys_mapping(void) } else { machine_to_phys_nr = MACH2PHYS_NR_ENTRIES; } -#ifdef CONFIG_X86_32 - WARN_ON((machine_to_phys_mapping + (machine_to_phys_nr - 1)) - < machine_to_phys_mapping); -#endif } -#ifdef CONFIG_X86_64 static void __init convert_pfn_mfn(void *v) { pte_t *pte = v; @@ -2168,105 +2007,6 @@ void __init xen_relocate_p2m(void) xen_start_info->nr_p2m_frames = n_frames; } -#else /* !CONFIG_X86_64 */ -static RESERVE_BRK_ARRAY(pmd_t, initial_kernel_pmd, PTRS_PER_PMD); -static RESERVE_BRK_ARRAY(pmd_t, swapper_kernel_pmd, PTRS_PER_PMD); -RESERVE_BRK(fixup_kernel_pmd, PAGE_SIZE); -RESERVE_BRK(fixup_kernel_pte, PAGE_SIZE); - -static void __init xen_write_cr3_init(unsigned long cr3) -{ - unsigned long pfn = PFN_DOWN(__pa(swapper_pg_dir)); - - BUG_ON(read_cr3_pa() != __pa(initial_page_table)); - BUG_ON(cr3 != __pa(swapper_pg_dir)); - - /* - * We are switching to swapper_pg_dir for the first time (from - * initial_page_table) and therefore need to mark that page - * read-only and then pin it. - * - * Xen disallows sharing of kernel PMDs for PAE - * guests. Therefore we must copy the kernel PMD from - * initial_page_table into a new kernel PMD to be used in - * swapper_pg_dir. - */ - swapper_kernel_pmd = - extend_brk(sizeof(pmd_t) * PTRS_PER_PMD, PAGE_SIZE); - copy_page(swapper_kernel_pmd, initial_kernel_pmd); - swapper_pg_dir[KERNEL_PGD_BOUNDARY] = - __pgd(__pa(swapper_kernel_pmd) | _PAGE_PRESENT); - set_page_prot(swapper_kernel_pmd, PAGE_KERNEL_RO); - - set_page_prot(swapper_pg_dir, PAGE_KERNEL_RO); - xen_write_cr3(cr3); - pin_pagetable_pfn(MMUEXT_PIN_L3_TABLE, pfn); - - pin_pagetable_pfn(MMUEXT_UNPIN_TABLE, - PFN_DOWN(__pa(initial_page_table))); - set_page_prot(initial_page_table, PAGE_KERNEL); - set_page_prot(initial_kernel_pmd, PAGE_KERNEL); - - pv_ops.mmu.write_cr3 = &xen_write_cr3; -} - -/* - * For 32 bit domains xen_start_info->pt_base is the pgd address which might be - * not the first page table in the page table pool. - * Iterate through the initial page tables to find the real page table base. - */ -static phys_addr_t __init xen_find_pt_base(pmd_t *pmd) -{ - phys_addr_t pt_base, paddr; - unsigned pmdidx; - - pt_base = min(__pa(xen_start_info->pt_base), __pa(pmd)); - - for (pmdidx = 0; pmdidx < PTRS_PER_PMD; pmdidx++) - if (pmd_present(pmd[pmdidx]) && !pmd_large(pmd[pmdidx])) { - paddr = m2p(pmd[pmdidx].pmd); - pt_base = min(pt_base, paddr); - } - - return pt_base; -} - -void __init xen_setup_kernel_pagetable(pgd_t *pgd, unsigned long max_pfn) -{ - pmd_t *kernel_pmd; - - kernel_pmd = m2v(pgd[KERNEL_PGD_BOUNDARY].pgd); - - xen_pt_base = xen_find_pt_base(kernel_pmd); - xen_pt_size = xen_start_info->nr_pt_frames * PAGE_SIZE; - - initial_kernel_pmd = - extend_brk(sizeof(pmd_t) * PTRS_PER_PMD, PAGE_SIZE); - - max_pfn_mapped = PFN_DOWN(xen_pt_base + xen_pt_size + 512 * 1024); - - copy_page(initial_kernel_pmd, kernel_pmd); - - xen_map_identity_early(initial_kernel_pmd, max_pfn); - - copy_page(initial_page_table, pgd); - initial_page_table[KERNEL_PGD_BOUNDARY] = - __pgd(__pa(initial_kernel_pmd) | _PAGE_PRESENT); - - set_page_prot(initial_kernel_pmd, PAGE_KERNEL_RO); - set_page_prot(initial_page_table, PAGE_KERNEL_RO); - set_page_prot(empty_zero_page, PAGE_KERNEL_RO); - - pin_pagetable_pfn(MMUEXT_UNPIN_TABLE, PFN_DOWN(__pa(pgd))); - - pin_pagetable_pfn(MMUEXT_PIN_L3_TABLE, - PFN_DOWN(__pa(initial_page_table))); - xen_write_cr3(__pa(initial_page_table)); - - memblock_reserve(xen_pt_base, xen_pt_size); -} -#endif /* CONFIG_X86_64 */ - void __init xen_reserve_special_pages(void) { phys_addr_t paddr; @@ -2300,12 +2040,7 @@ static void xen_set_fixmap(unsigned idx, phys_addr_t phys, pgprot_t prot) switch (idx) { case FIX_BTMAP_END ... FIX_BTMAP_BEGIN: -#ifdef CONFIG_X86_32 - case FIX_WP_TEST: -# ifdef CONFIG_HIGHMEM - case FIX_KMAP_BEGIN ... FIX_KMAP_END: -# endif -#elif defined(CONFIG_X86_VSYSCALL_EMULATION) +#ifdef CONFIG_X86_VSYSCALL_EMULATION case VSYSCALL_PAGE: #endif /* All local page mappings */ @@ -2357,9 +2092,7 @@ static void __init xen_post_allocator_init(void) pv_ops.mmu.set_pte = xen_set_pte; pv_ops.mmu.set_pmd = xen_set_pmd; pv_ops.mmu.set_pud = xen_set_pud; -#ifdef CONFIG_X86_64 pv_ops.mmu.set_p4d = xen_set_p4d; -#endif /* This will work as long as patching hasn't happened yet (which it hasn't) */ @@ -2367,15 +2100,11 @@ static void __init xen_post_allocator_init(void) pv_ops.mmu.alloc_pmd = xen_alloc_pmd; pv_ops.mmu.release_pte = xen_release_pte; pv_ops.mmu.release_pmd = xen_release_pmd; -#ifdef CONFIG_X86_64 pv_ops.mmu.alloc_pud = xen_alloc_pud; pv_ops.mmu.release_pud = xen_release_pud; -#endif pv_ops.mmu.make_pte = PV_CALLEE_SAVE(xen_make_pte); -#ifdef CONFIG_X86_64 pv_ops.mmu.write_cr3 = &xen_write_cr3; -#endif } static void xen_leave_lazy_mmu(void) @@ -2420,17 +2149,11 @@ static const struct pv_mmu_ops xen_mmu_ops __initconst = { .make_pte = PV_CALLEE_SAVE(xen_make_pte_init), .make_pgd = PV_CALLEE_SAVE(xen_make_pgd), -#ifdef CONFIG_X86_PAE - .set_pte_atomic = xen_set_pte_atomic, - .pte_clear = xen_pte_clear, - .pmd_clear = xen_pmd_clear, -#endif /* CONFIG_X86_PAE */ .set_pud = xen_set_pud_hyper, .make_pmd = PV_CALLEE_SAVE(xen_make_pmd), .pmd_val = PV_CALLEE_SAVE(xen_pmd_val), -#ifdef CONFIG_X86_64 .pud_val = PV_CALLEE_SAVE(xen_pud_val), .make_pud = PV_CALLEE_SAVE(xen_make_pud), .set_p4d = xen_set_p4d_hyper, @@ -2442,7 +2165,6 @@ static const struct pv_mmu_ops xen_mmu_ops __initconst = { .p4d_val = PV_CALLEE_SAVE(xen_p4d_val), .make_p4d = PV_CALLEE_SAVE(xen_make_p4d), #endif -#endif /* CONFIG_X86_64 */ .activate_mm = xen_activate_mm, .dup_mmap = xen_dup_mmap, |