diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2011-05-24 12:06:02 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2011-05-24 12:06:02 -0700 |
commit | 0d66cba1ac3ad38614077443d604d6a09cec99de (patch) | |
tree | 383981fd93c7e6d467fb239edd51f935f1a481dd /arch | |
parent | 4637f40f200063973553ce3c4c1ac6c247e4535c (diff) | |
parent | 5bd418784a2764a8d9de177a5462bfc008fd334a (diff) |
Merge branch 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6
* 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6: (29 commits)
[S390] cpu hotplug: fix external interrupt subclass mask handling
[S390] oprofile: dont access lowcore
[S390] oprofile: add missing irq stats counter
[S390] Ignore sendmmsg system call note wired up warning
[S390] s390,oprofile: fix compile error for !CONFIG_SMP
[S390] s390,oprofile: fix alert counter increment
[S390] Remove unused includes in process.c
[S390] get CPC image name
[S390] sclp: event buffer dissection
[S390] chsc: process channel-path-availability information
[S390] refactor page table functions for better pgste support
[S390] merge page_test_dirty and page_clear_dirty
[S390] qdio: prevent compile warning
[S390] sclp: remove unnecessary sendmask check
[S390] convert old cpumask API into new one
[S390] pfault: cleanup code
[S390] pfault: cpu hotplug vs missing completion interrupts
[S390] smp: add __noreturn attribute to cpu_die()
[S390] percpu: implement arch specific irqsafe_cpu_ops
[S390] vdso: disable gcov profiling
...
Diffstat (limited to 'arch')
35 files changed, 626 insertions, 668 deletions
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 4a7f14079e03..ff2d2371b2e9 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -230,17 +230,6 @@ config SYSVIPC_COMPAT config AUDIT_ARCH def_bool y -config S390_EXEC_PROTECT - def_bool y - prompt "Data execute protection" - help - This option allows to enable a buffer overflow protection for user - space programs and it also selects the addressing mode option above. - The kernel parameter noexec=on will enable this feature and also - switch the addressing modes, default is disabled. Enabling this (via - kernel parameter) on machines earlier than IBM System z9 this will - reduce system performance. - comment "Code generation options" choice diff --git a/arch/s390/appldata/appldata_base.c b/arch/s390/appldata/appldata_base.c index 5c91995b74e4..24bff4f1cc52 100644 --- a/arch/s390/appldata/appldata_base.c +++ b/arch/s390/appldata/appldata_base.c @@ -130,9 +130,7 @@ static void appldata_work_fn(struct work_struct *work) { struct list_head *lh; struct appldata_ops *ops; - int i; - i = 0; get_online_cpus(); mutex_lock(&appldata_ops_mutex); list_for_each(lh, &appldata_ops_list) { diff --git a/arch/s390/include/asm/cmpxchg.h b/arch/s390/include/asm/cmpxchg.h index 7488e52efa97..81d7908416cf 100644 --- a/arch/s390/include/asm/cmpxchg.h +++ b/arch/s390/include/asm/cmpxchg.h @@ -167,7 +167,6 @@ static inline unsigned long __cmpxchg(void *ptr, unsigned long old, #ifdef CONFIG_64BIT #define cmpxchg64(ptr, o, n) \ ({ \ - BUILD_BUG_ON(sizeof(*(ptr)) != 8); \ cmpxchg((ptr), (o), (n)); \ }) #else /* CONFIG_64BIT */ diff --git a/arch/s390/include/asm/elf.h b/arch/s390/include/asm/elf.h index 10c029cfcc7d..64b61bf72e93 100644 --- a/arch/s390/include/asm/elf.h +++ b/arch/s390/include/asm/elf.h @@ -196,18 +196,6 @@ do { \ } while (0) #endif /* __s390x__ */ -/* - * An executable for which elf_read_implies_exec() returns TRUE will - * have the READ_IMPLIES_EXEC personality flag set automatically. - */ -#define elf_read_implies_exec(ex, executable_stack) \ -({ \ - if (current->mm->context.noexec && \ - executable_stack != EXSTACK_DISABLE_X) \ - disable_noexec(current->mm, current); \ - current->mm->context.noexec == 0; \ -}) - #define STACK_RND_MASK 0x7ffUL #define ARCH_DLINFO \ diff --git a/arch/s390/include/asm/hugetlb.h b/arch/s390/include/asm/hugetlb.h index b56403c2df28..799ed0f1643d 100644 --- a/arch/s390/include/asm/hugetlb.h +++ b/arch/s390/include/asm/hugetlb.h @@ -111,21 +111,10 @@ static inline void huge_ptep_invalidate(struct mm_struct *mm, { pmd_t *pmdp = (pmd_t *) ptep; - if (!MACHINE_HAS_IDTE) { - __pmd_csp(pmdp); - if (mm->context.noexec) { - pmdp = get_shadow_table(pmdp); - __pmd_csp(pmdp); - } - return; - } - - __pmd_idte(address, pmdp); - if (mm->context.noexec) { - pmdp = get_shadow_table(pmdp); + if (MACHINE_HAS_IDTE) __pmd_idte(address, pmdp); - } - return; + else + __pmd_csp(pmdp); } #define huge_ptep_set_access_flags(__vma, __addr, __ptep, __entry, __dirty) \ diff --git a/arch/s390/include/asm/irq.h b/arch/s390/include/asm/irq.h index db14a311f1d2..1544b90bd6d6 100644 --- a/arch/s390/include/asm/irq.h +++ b/arch/s390/include/asm/irq.h @@ -15,6 +15,7 @@ enum interruption_class { EXTINT_VRT, EXTINT_SCP, EXTINT_IUC, + EXTINT_CPM, IOINT_QAI, IOINT_QDI, IOINT_DAS, diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h index 65e172f8209d..228cf0b295db 100644 --- a/arch/s390/include/asm/lowcore.h +++ b/arch/s390/include/asm/lowcore.h @@ -124,7 +124,7 @@ struct _lowcore { /* Address space pointer. */ __u32 kernel_asce; /* 0x02ac */ __u32 user_asce; /* 0x02b0 */ - __u32 user_exec_asce; /* 0x02b4 */ + __u32 current_pid; /* 0x02b4 */ /* SMP info area */ __u32 cpu_nr; /* 0x02b8 */ @@ -255,7 +255,7 @@ struct _lowcore { /* Address space pointer. */ __u64 kernel_asce; /* 0x0310 */ __u64 user_asce; /* 0x0318 */ - __u64 user_exec_asce; /* 0x0320 */ + __u64 current_pid; /* 0x0320 */ /* SMP info area */ __u32 cpu_nr; /* 0x0328 */ diff --git a/arch/s390/include/asm/mmu.h b/arch/s390/include/asm/mmu.h index 78522cdefdd4..82d0847896a0 100644 --- a/arch/s390/include/asm/mmu.h +++ b/arch/s390/include/asm/mmu.h @@ -5,19 +5,18 @@ typedef struct { atomic_t attach_count; unsigned int flush_mm; spinlock_t list_lock; - struct list_head crst_list; struct list_head pgtable_list; unsigned long asce_bits; unsigned long asce_limit; unsigned long vdso_base; - int noexec; - int has_pgste; /* The mmu context has extended page tables */ - int alloc_pgste; /* cloned contexts will have extended page tables */ + /* Cloned contexts will be created with extended page tables. */ + unsigned int alloc_pgste:1; + /* The mmu context has extended page tables. */ + unsigned int has_pgste:1; } mm_context_t; #define INIT_MM_CONTEXT(name) \ .context.list_lock = __SPIN_LOCK_UNLOCKED(name.context.list_lock), \ - .context.crst_list = LIST_HEAD_INIT(name.context.crst_list), \ .context.pgtable_list = LIST_HEAD_INIT(name.context.pgtable_list), #endif diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h index 8c277caa8d3a..5682f160ff82 100644 --- a/arch/s390/include/asm/mmu_context.h +++ b/arch/s390/include/asm/mmu_context.h @@ -35,11 +35,9 @@ static inline int init_new_context(struct task_struct *tsk, * and if has_pgste is set, it will create extended page * tables. */ - mm->context.noexec = 0; mm->context.has_pgste = 1; mm->context.alloc_pgste = 1; } else { - mm->context.noexec = (user_mode == SECONDARY_SPACE_MODE); mm->context.has_pgste = 0; mm->context.alloc_pgste = 0; } @@ -63,10 +61,8 @@ static inline void update_mm(struct mm_struct *mm, struct task_struct *tsk) S390_lowcore.user_asce = mm->context.asce_bits | __pa(pgd); if (user_mode != HOME_SPACE_MODE) { /* Load primary space page table origin. */ - pgd = mm->context.noexec ? get_shadow_table(pgd) : pgd; - S390_lowcore.user_exec_asce = mm->context.asce_bits | __pa(pgd); asm volatile(LCTL_OPCODE" 1,1,%0\n" - : : "m" (S390_lowcore.user_exec_asce) ); + : : "m" (S390_lowcore.user_asce) ); } else /* Load home space page table origin. */ asm volatile(LCTL_OPCODE" 13,13,%0" diff --git a/arch/s390/include/asm/page.h b/arch/s390/include/asm/page.h index 3c987e9ec8d6..accb372ddc7e 100644 --- a/arch/s390/include/asm/page.h +++ b/arch/s390/include/asm/page.h @@ -90,6 +90,7 @@ static inline void copy_page(void *to, void *from) */ typedef struct { unsigned long pgprot; } pgprot_t; +typedef struct { unsigned long pgste; } pgste_t; typedef struct { unsigned long pte; } pte_t; typedef struct { unsigned long pmd; } pmd_t; typedef struct { unsigned long pud; } pud_t; @@ -97,18 +98,21 @@ typedef struct { unsigned long pgd; } pgd_t; typedef pte_t *pgtable_t; #define pgprot_val(x) ((x).pgprot) +#define pgste_val(x) ((x).pgste) #define pte_val(x) ((x).pte) #define pmd_val(x) ((x).pmd) #define pud_val(x) ((x).pud) #define pgd_val(x) ((x).pgd) +#define __pgste(x) ((pgste_t) { (x) } ) #define __pte(x) ((pte_t) { (x) } ) #define __pmd(x) ((pmd_t) { (x) } ) +#define __pud(x) ((pud_t) { (x) } ) #define __pgd(x) ((pgd_t) { (x) } ) #define __pgprot(x) ((pgprot_t) { (x) } ) -static inline void -page_set_storage_key(unsigned long addr, unsigned int skey, int mapped) +static inline void page_set_storage_key(unsigned long addr, + unsigned char skey, int mapped) { if (!mapped) asm volatile(".insn rrf,0xb22b0000,%0,%1,8,0" @@ -117,15 +121,59 @@ page_set_storage_key(unsigned long addr, unsigned int skey, int mapped) asm volatile("sske %0,%1" : : "d" (skey), "a" (addr)); } -static inline unsigned int -page_get_storage_key(unsigned long addr) +static inline unsigned char page_get_storage_key(unsigned long addr) { - unsigned int skey; + unsigned char skey; - asm volatile("iske %0,%1" : "=d" (skey) : "a" (addr), "0" (0)); + asm volatile("iske %0,%1" : "=d" (skey) : "a" (addr)); return skey; } +static inline int page_reset_referenced(unsigned long addr) +{ + unsigned int ipm; + + asm volatile( + " rrbe 0,%1\n" + " ipm %0\n" + : "=d" (ipm) : "a" (addr) : "cc"); + return !!(ipm & 0x20000000); +} + +/* Bits int the storage key */ +#define _PAGE_CHANGED 0x02 /* HW changed bit */ +#define _PAGE_REFERENCED 0x04 /* HW referenced bit */ +#define _PAGE_FP_BIT 0x08 /* HW fetch protection bit */ +#define _PAGE_ACC_BITS 0xf0 /* HW access control bits */ + +/* + * Test and clear dirty bit in storage key. + * We can't clear the changed bit atomically. This is a potential + * race against modification of the referenced bit. This function + * should therefore only be called if it is not mapped in any + * address space. + */ +#define __HAVE_ARCH_PAGE_TEST_AND_CLEAR_DIRTY +static inline int page_test_and_clear_dirty(unsigned long pfn, int mapped) +{ + unsigned char skey; + + skey = page_get_storage_key(pfn << PAGE_SHIFT); + if (!(skey & _PAGE_CHANGED)) + return 0; + page_set_storage_key(pfn << PAGE_SHIFT, skey & ~_PAGE_CHANGED, mapped); + return 1; +} + +/* + * Test and clear referenced bit in storage key. + */ +#define __HAVE_ARCH_PAGE_TEST_AND_CLEAR_YOUNG +static inline int page_test_and_clear_young(unsigned long pfn) +{ + return page_reset_referenced(pfn << PAGE_SHIFT); +} + struct page; void arch_free_page(struct page *page, int order); void arch_alloc_page(struct page *page, int order); diff --git a/arch/s390/include/asm/percpu.h b/arch/s390/include/asm/percpu.h index f7ad8719d02d..5325c89a5843 100644 --- a/arch/s390/include/asm/percpu.h +++ b/arch/s390/include/asm/percpu.h @@ -1,6 +1,9 @@ #ifndef __ARCH_S390_PERCPU__ #define __ARCH_S390_PERCPU__ +#include <linux/preempt.h> +#include <asm/cmpxchg.h> + /* * s390 uses its own implementation for per cpu data, the offset of * the cpu local data area is cached in the cpu's lowcore memory. @@ -16,6 +19,71 @@ #define ARCH_NEEDS_WEAK_PER_CPU #endif +#define arch_irqsafe_cpu_to_op(pcp, val, op) \ +do { \ + typedef typeof(pcp) pcp_op_T__; \ + pcp_op_T__ old__, new__, prev__; \ + pcp_op_T__ *ptr__; \ + preempt_disable(); \ + ptr__ = __this_cpu_ptr(&(pcp)); \ + prev__ = *ptr__; \ + do { \ + old__ = prev__; \ + new__ = old__ op (val); \ + switch (sizeof(*ptr__)) { \ + case 8: \ + prev__ = cmpxchg64(ptr__, old__, new__); \ + break; \ + default: \ + prev__ = cmpxchg(ptr__, old__, new__); \ + } \ + } while (prev__ != old__); \ + preempt_enable(); \ +} while (0) + +#define irqsafe_cpu_add_1(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, +) +#define irqsafe_cpu_add_2(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, +) +#define irqsafe_cpu_add_4(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, +) +#define irqsafe_cpu_add_8(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, +) + +#define irqsafe_cpu_and_1(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, &) +#define irqsafe_cpu_and_2(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, &) +#define irqsafe_cpu_and_4(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, &) +#define irqsafe_cpu_and_8(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, &) + +#define irqsafe_cpu_or_1(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, |) +#define irqsafe_cpu_or_2(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, |) +#define irqsafe_cpu_or_4(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, |) +#define irqsafe_cpu_or_8(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, |) + +#define irqsafe_cpu_xor_1(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, ^) +#define irqsafe_cpu_xor_2(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, ^) +#define irqsafe_cpu_xor_4(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, ^) +#define irqsafe_cpu_xor_8(pcp, val) arch_irqsafe_cpu_to_op(pcp, val, ^) + +#define arch_irqsafe_cpu_cmpxchg(pcp, oval, nval) \ +({ \ + typedef typeof(pcp) pcp_op_T__; \ + pcp_op_T__ ret__; \ + pcp_op_T__ *ptr__; \ + preempt_disable(); \ + ptr__ = __this_cpu_ptr(&(pcp)); \ + switch (sizeof(*ptr__)) { \ + case 8: \ + ret__ = cmpxchg64(ptr__, oval, nval); \ + break; \ + default: \ + ret__ = cmpxchg(ptr__, oval, nval); \ + } \ + preempt_enable(); \ + ret__; \ +}) + +#define irqsafe_cpu_cmpxchg_1(pcp, oval, nval) arch_irqsafe_cpu_cmpxchg(pcp, oval, nval) +#define irqsafe_cpu_cmpxchg_2(pcp, oval, nval) arch_irqsafe_cpu_cmpxchg(pcp, oval, nval) +#define irqsafe_cpu_cmpxchg_4(pcp, oval, nval) arch_irqsafe_cpu_cmpxchg(pcp, oval, nval) +#define irqsafe_cpu_cmpxchg_8(pcp, oval, nval) arch_irqsafe_cpu_cmpxchg(pcp, oval, nval) + #include <asm-generic/percpu.h> #endif /* __ARCH_S390_PERCPU__ */ diff --git a/arch/s390/include/asm/pgalloc.h b/arch/s390/include/asm/pgalloc.h index 082eb4e50e8b..f6314af3b354 100644 --- a/arch/s390/include/asm/pgalloc.h +++ b/arch/s390/include/asm/pgalloc.h @@ -19,14 +19,13 @@ #define check_pgt_cache() do {} while (0) -unsigned long *crst_table_alloc(struct mm_struct *, int); +unsigned long *crst_table_alloc(struct mm_struct *); void crst_table_free(struct mm_struct *, unsigned long *); void crst_table_free_rcu(struct mm_struct *, unsigned long *); unsigned long *page_table_alloc(struct mm_struct *); void page_table_free(struct mm_struct *, unsigned long *); void page_table_free_rcu(struct mm_struct *, unsigned long *); -void disable_noexec(struct mm_struct *, struct task_struct *); static inline void clear_table(unsigned long *s, unsigned long val, size_t n) { @@ -50,9 +49,6 @@ static inline void clear_table(unsigned long *s, unsigned long val, size_t n) static inline void crst_table_init(unsigned long *crst, unsigned long entry) { clear_table(crst, entry, sizeof(unsigned long)*2048); - crst = get_shadow_table(crst); - if (crst) - clear_table(crst, entry, sizeof(unsigned long)*2048); } #ifndef __s390x__ @@ -69,10 +65,7 @@ static inline unsigned long pgd_entry_type(struct mm_struct *mm) #define pmd_free(mm, x) do { } while (0) #define pgd_populate(mm, pgd, pud) BUG() -#define pgd_populate_kernel(mm, pgd, pud) BUG() - #define pud_populate(mm, pud, pmd) BUG() -#define pud_populate_kernel(mm, pud, pmd) BUG() #else /* __s390x__ */ @@ -90,7 +83,7 @@ void crst_table_downgrade(struct mm_struct *, unsigned long limit); static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long address) { - unsigned long *table = crst_table_alloc(mm, mm->context.noexec); + unsigned long *table = crst_table_alloc(mm); if (table) crst_table_init(table, _REGION3_ENTRY_EMPTY); return (pud_t *) table; @@ -99,43 +92,21 @@ static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long address) static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long vmaddr) { - unsigned long *table = crst_table_alloc(mm, mm->context.noexec); + unsigned long *table = crst_table_alloc(mm); if (table) crst_table_init(table, _SEGMENT_ENTRY_EMPTY); return (pmd_t *) table; } #define pmd_free(mm, pmd) crst_table_free(mm, (unsigned long *) pmd) -static inline void pgd_populate_kernel(struct mm_struct *mm, - pgd_t *pgd, pud_t *pud) -{ - pgd_val(*pgd) = _REGION2_ENTRY | __pa(pud); -} - static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pud_t *pud) { - pgd_populate_kernel(mm, pgd, pud); - if (mm->context.noexec) { - pgd = get_shadow_table(pgd); - pud = get_shadow_table(pud); - pgd_populate_kernel(mm, pgd, pud); - } -} - -static inline void pud_populate_kernel(struct mm_struct *mm, - pud_t *pud, pmd_t *pmd) -{ - pud_val(*pud) = _REGION3_ENTRY | __pa(pmd); + pgd_val(*pgd) = _REGION2_ENTRY | __pa(pud); } static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) { - pud_populate_kernel(mm, pud, pmd); - if (mm->context.noexec) { - pud = get_shadow_table(pud); - pmd = get_shadow_table(pmd); - pud_populate_kernel(mm, pud, pmd); - } + pud_val(*pud) = _REGION3_ENTRY | __pa(pmd); } #endif /* __s390x__ */ @@ -143,29 +114,19 @@ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) static inline pgd_t *pgd_alloc(struct mm_struct *mm) { spin_lock_init(&mm->context.list_lock); - INIT_LIST_HEAD(&mm->context.crst_list); INIT_LIST_HEAD(&mm->context.pgtable_list); - return (pgd_t *) - crst_table_alloc(mm, user_mode == SECONDARY_SPACE_MODE); + return (pgd_t *) crst_table_alloc(mm); } #define pgd_free(mm, pgd) crst_table_free(mm, (unsigned long *) pgd) -static inline void pmd_populate_kernel(struct mm_struct *mm, - pmd_t *pmd, pte_t *pte) -{ - pmd_val(*pmd) = _SEGMENT_ENTRY + __pa(pte); -} - static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd, pgtable_t pte) { - pmd_populate_kernel(mm, pmd, pte); - if (mm->context.noexec) { - pmd = get_shadow_table(pmd); - pmd_populate_kernel(mm, pmd, pte + PTRS_PER_PTE); - } + pmd_val(*pmd) = _SEGMENT_ENTRY + __pa(pte); } +#define pmd_populate_kernel(mm, pmd, pte) pmd_populate(mm, pmd, pte) + #define pmd_pgtable(pmd) \ (pgtable_t)(pmd_val(pmd) & -sizeof(pte_t)*PTRS_PER_PTE) diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 02ace3491c51..c4773a2ef3d3 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -31,9 +31,8 @@ #ifndef __ASSEMBLY__ #include <linux/sched.h> #include <linux/mm_types.h> -#include <asm/bitops.h> #include <asm/bug.h> -#include <asm/processor.h> +#include <asm/page.h> extern pgd_t swapper_pg_dir[] __attribute__ ((aligned (4096))); extern void paging_init(void); @@ -243,11 +242,13 @@ extern unsigned long VMALLOC_START; /* Software bits in the page table entry */ #define _PAGE_SWT 0x001 /* SW pte type bit t */ #define _PAGE_SWX 0x002 /* SW pte type bit x */ -#define _PAGE_SPECIAL 0x004 /* SW associated with special page */ +#define _PAGE_SWC 0x004 /* SW pte changed bit (for KVM) */ +#define _PAGE_SWR 0x008 /* SW pte referenced bit (for KVM) */ +#define _PAGE_SPECIAL 0x010 /* SW associated with special page */ #define __HAVE_ARCH_PTE_SPECIAL /* Set of bits not changed in pte_modify */ -#define _PAGE_CHG_MASK (PAGE_MASK | _PAGE_SPECIAL) +#define _PAGE_CHG_MASK (PAGE_MASK | _PAGE_SPECIAL | _PAGE_SWC | _PAGE_SWR) /* Six different types of pages. */ #define _PAGE_TYPE_EMPTY 0x400 @@ -256,8 +257,6 @@ extern unsigned long VMALLOC_START; #define _PAGE_TYPE_FILE 0x601 /* bit 0x002 is used for offset !! */ #define _PAGE_TYPE_RO 0x200 #define _PAGE_TYPE_RW 0x000 -#define _PAGE_TYPE_EX_RO 0x202 -#define _PAGE_TYPE_EX_RW 0x002 /* * Only four types for huge pages, using the invalid bit and protection bit @@ -287,8 +286,6 @@ extern unsigned long VMALLOC_START; * _PAGE_TYPE_FILE 11?1 -> 11?1 * _PAGE_TYPE_RO 0100 -> 1100 * _PAGE_TYPE_RW 0000 -> 1000 - * _PAGE_TYPE_EX_RO 0110 -> 1110 - * _PAGE_TYPE_EX_RW 0010 -> 1010 * * pte_none is true for bits combinations 1000, 1010, 1100, 1110 * pte_present is true for bits combinations 0000, 0010, 0100, 0110, 1001 @@ -297,14 +294,17 @@ extern unsigned long VMALLOC_START; */ /* Page status table bits for virtualization */ -#define RCP_PCL_BIT 55 -#define RCP_HR_BIT 54 -#define RCP_HC_BIT 53 -#define RCP_GR_BIT 50 -#define RCP_GC_BIT 49 - -/* User dirty bit for KVM's migration feature */ -#define KVM_UD_BIT 47 +#define RCP_ACC_BITS 0xf000000000000000UL +#define RCP_FP_BIT 0x0800000000000000UL +#define RCP_PCL_BIT 0x0080000000000000UL +#define RCP_HR_BIT 0x0040000000000000UL +#define RCP_HC_BIT 0x0020000000000000UL +#define RCP_GR_BIT 0x0004000000000000UL +#define RCP_GC_BIT 0x0002000000000000UL + +/* User dirty / referenced bit for KVM's migration feature */ +#define KVM_UR_BIT 0x0000800000000000UL +#define KVM_UC_BIT 0x0000400000000000UL #ifndef __s390x__ @@ -377,85 +377,54 @@ extern unsigned long VMALLOC_START; #define _ASCE_USER_BITS (_ASCE_SPACE_SWITCH | _ASCE_PRIVATE_SPACE | \ _ASCE_ALT_EVENT) -/* Bits int the storage key */ -#define _PAGE_CHANGED 0x02 /* HW changed bit */ -#define _PAGE_REFERENCED 0x04 /* HW referenced bit */ - /* * Page protection definitions. */ #define PAGE_NONE __pgprot(_PAGE_TYPE_NONE) #define PAGE_RO __pgprot(_PAGE_TYPE_RO) #define PAGE_RW __pgprot(_PAGE_TYPE_RW) -#define PAGE_EX_RO __pgprot(_PAGE_TYPE_EX_RO) -#define PAGE_EX_RW __pgprot(_PAGE_TYPE_EX_RW) #define PAGE_KERNEL PAGE_RW #define PAGE_COPY PAGE_RO /* - * Dependent on the EXEC_PROTECT option s390 can do execute protection. - * Write permission always implies read permission. In theory with a - * primary/secondary page table execute only can be implemented but - * it would cost an additional bit in the pte to distinguish all the - * different pte types. To avoid that execute permission currently - * implies read permission as well. + * On s390 the page table entry has an invalid bit and a read-only bit. + * Read permission implies execute permission and write permission + * implies read permission. */ /*xwr*/ #define __P000 PAGE_NONE #define __P001 PAGE_RO #define __P010 PAGE_RO #define __P011 PAGE_RO -#define __P100 PAGE_EX_RO -#define __P101 PAGE_EX_RO -#define __P110 PAGE_EX_RO -#define __P111 PAGE_EX_RO +#define __P100 PAGE_RO +#define __P101 PAGE_RO +#define __P110 PAGE_RO +#define __P111 PAGE_RO #define __S000 PAGE_NONE #define __S001 PAGE_RO #define __S010 PAGE_RW #define __S011 PAGE_RW -#define __S100 PAGE_EX_RO -#define __S101 PAGE_EX_RO -#define __S110 PAGE_EX_RW -#define __S111 PAGE_EX_RW - -#ifndef __s390x__ -# define PxD_SHADOW_SHIFT 1 -#else /* __s390x__ */ -# define PxD_SHADOW_SHIFT 2 -#endif /* __s390x__ */ +#define __S100 PAGE_RO +#define __S101 PAGE_RO +#define __S110 PAGE_RW +#define __S111 PAGE_RW -static inline void *get_shadow_table(void *table) +static inline int mm_exclusive(struct mm_struct *mm) { - unsigned long addr, offset; - struct page *page; - - addr = (unsigned long) table; - offset = addr & ((PAGE_SIZE << PxD_SHADOW_SHIFT) - 1); - page = virt_to_page((void *)(addr ^ offset)); - return (void *)(addr_t)(page->index ? (page->index | offset) : 0UL); + return likely(mm == current->active_mm && + atomic_read(&mm->context.attach_count) <= 1); } -/* - * Certain architectures need to do special things when PTEs - * within a page table are directly modified. Thus, the following - * hook is made available. - */ -static inline void set_pte_at(struct mm_struct *mm, unsigned long addr, - pte_t *ptep, pte_t entry) +static inline int mm_has_pgste(struct mm_struct *mm) { - *ptep = entry; - if (mm->context.noexec) { - if (!(pte_val(entry) & _PAGE_INVALID) && - (pte_val(entry) & _PAGE_SWX)) - pte_val(entry) |= _PAGE_RO; - else - pte_val(entry) = _PAGE_TYPE_EMPTY; - ptep[PTRS_PER_PTE] = entry; - } +#ifdef CONFIG_PGSTE + if (unlikely(mm->context.has_pgste)) + return 1; +#endif + return 0; } - /* * pgd/pmd/pte query functions */ @@ -568,52 +537,127 @@ static inline int pte_special(pte_t pte) } #define __HAVE_ARCH_PTE_SAME -#define pte_same(a,b) (pte_val(a) == pte_val(b)) +static inline int pte_same(pte_t a, pte_t b) +{ + return pte_val(a) == pte_val(b); +} -static inline void rcp_lock(pte_t *ptep) +static inline pgste_t pgste_get_lock(pte_t *ptep) { + unsigned long new = 0; #ifdef CONFIG_PGSTE - unsigned long *pgste = (unsigned long *) (ptep + PTRS_PER_PTE); + unsigned long old; + preempt_disable(); - while (test_and_set_bit(RCP_PCL_BIT, pgste)) - ; + asm( + " lg %0,%2\n" + "0: lgr %1,%0\n" + " nihh %0,0xff7f\n" /* clear RCP_PCL_BIT in old */ + " oihh %1,0x0080\n" /* set RCP_PCL_BIT in new */ + " csg %0,%1,%2\n" + " jl 0b\n" + : "=&d" (old), "=&d" (new), "=Q" (ptep[PTRS_PER_PTE]) + : "Q" (ptep[PTRS_PER_PTE]) : "cc"); #endif + return __pgste(new); } -static inline void rcp_unlock(pte_t *ptep) +static inline void pgste_set_unlock(pte_t *ptep, pgste_t pgste) { #ifdef CONFIG_PGSTE - unsigned long *pgste = (unsigned long *) (ptep + PTRS_PER_PTE); - clear_bit(RCP_PCL_BIT, pgste); + asm( + " nihh %1,0xff7f\n" /* clear RCP_PCL_BIT */ + " stg %1,%0\n" + : "=Q" (ptep[PTRS_PER_PTE]) + : "d" (pgste_val(pgste)), "Q" (ptep[PTRS_PER_PTE]) : "cc"); preempt_enable(); #endif } -/* forward declaration for SetPageUptodate in page-flags.h*/ -static inline void page_clear_dirty(struct page *page, int mapped); -#include <linux/page-flags.h> - -static inline void ptep_rcp_copy(pte_t *ptep) +static inline pgste_t pgste_update_all(pte_t *ptep, pgste_t pgste) { #ifdef CONFIG_PGSTE - struct page *page = virt_to_page(pte_val(*ptep)); - unsigned int skey; - unsigned long *pgste = (unsigned long *) (ptep + PTRS_PER_PTE); - - skey = page_get_storage_key(page_to_phys(page)); - if (skey & _PAGE_CHANGED) { - set_bit_simple(RCP_GC_BIT, pgste); - set_bit_simple(KVM_UD_BIT, pgste); + unsigned long pfn, bits; + unsigned char skey; + + pfn = pte_val(*ptep) >> PAGE_SHIFT; |