#define __HAVE_ARCH_PGD_FREE
#include <asm-generic/pgalloc.h>
-/* Allocate the top level pgd (page directory)
- *
- * Here (for 64 bit kernels) we implement a Hybrid L2/L3 scheme: we
- * allocate the first pmd adjacent to the pgd. This means that we can
- * subtract a constant offset to get to it. The pmd and pgd sizes are
- * arranged so that a single pmd covers 4GB (giving a full 64-bit
- * process access to 8TB) so our lookups are effectively L2 for the
- * first 4GB of the kernel (i.e. for all ILP32 processes and all the
- * kernel for machines with under 4GB of memory) */
+/* Allocate the top level pgd (page directory) */
static inline pgd_t *pgd_alloc(struct mm_struct *mm)
{
- pgd_t *pgd = (pgd_t *)__get_free_pages(GFP_KERNEL,
- PGD_ALLOC_ORDER);
- pgd_t *actual_pgd = pgd;
+ pgd_t *pgd;
- if (likely(pgd != NULL)) {
- memset(pgd, 0, PAGE_SIZE<<PGD_ALLOC_ORDER);
-#if CONFIG_PGTABLE_LEVELS == 3
- actual_pgd += PTRS_PER_PGD;
- /* Populate first pmd with allocated memory. We mark it
- * with PxD_FLAG_ATTACHED as a signal to the system that this
- * pmd entry may not be cleared. */
- set_pgd(actual_pgd, __pgd((PxD_FLAG_PRESENT |
- PxD_FLAG_VALID |
- PxD_FLAG_ATTACHED)
- + (__u32)(__pa((unsigned long)pgd) >> PxD_VALUE_SHIFT)));
- /* The first pmd entry also is marked with PxD_FLAG_ATTACHED as
- * a signal that this pmd may not be freed */
- set_pgd(pgd, __pgd(PxD_FLAG_ATTACHED));
-#endif
- }
- spin_lock_init(pgd_spinlock(actual_pgd));
- return actual_pgd;
+ pgd = (pgd_t *) __get_free_pages(GFP_KERNEL, PGD_ORDER);
+ if (unlikely(pgd == NULL))
+ return NULL;
+
+ memset(pgd, 0, PAGE_SIZE << PGD_ORDER);
+
+ return pgd;
}
static inline void pgd_free(struct mm_struct *mm, pgd_t *pgd)
{
-#if CONFIG_PGTABLE_LEVELS == 3
- pgd -= PTRS_PER_PGD;
-#endif
- free_pages((unsigned long)pgd, PGD_ALLOC_ORDER);
+ free_pages((unsigned long)pgd, PGD_ORDER);
}
#if CONFIG_PGTABLE_LEVELS == 3
static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long address)
{
- return (pmd_t *)__get_free_pages(GFP_PGTABLE_KERNEL, PMD_ORDER);
+ pmd_t *pmd;
+
+ pmd = (pmd_t *)__get_free_pages(GFP_PGTABLE_KERNEL, PMD_ORDER);
+ if (likely(pmd))
+ memset ((void *)pmd, 0, PAGE_SIZE << PMD_ORDER);
+ return pmd;
}
static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd)
{
- if (pmd_flag(*pmd) & PxD_FLAG_ATTACHED) {
- /*
- * This is the permanent pmd attached to the pgd;
- * cannot free it.
- * Increment the counter to compensate for the decrement
- * done by generic mm code.
- */
- mm_inc_nr_pmds(mm);
- return;
- }
free_pages((unsigned long)pmd, PMD_ORDER);
}
-
#endif
static inline void
pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmd, pte_t *pte)
{
-#if CONFIG_PGTABLE_LEVELS == 3
- /* preserve the gateway marker if this is the beginning of
- * the permanent pmd */
- if(pmd_flag(*pmd) & PxD_FLAG_ATTACHED)
- set_pmd(pmd, __pmd((PxD_FLAG_PRESENT |
- PxD_FLAG_VALID |
- PxD_FLAG_ATTACHED)
- + (__u32)(__pa((unsigned long)pte) >> PxD_VALUE_SHIFT)));
- else
-#endif
- set_pmd(pmd, __pmd((PxD_FLAG_PRESENT | PxD_FLAG_VALID)
- + (__u32)(__pa((unsigned long)pte) >> PxD_VALUE_SHIFT)));
+ set_pmd(pmd, __pmd((PxD_FLAG_PRESENT | PxD_FLAG_VALID)
+ + (__u32)(__pa((unsigned long)pte) >> PxD_VALUE_SHIFT)));
}
#define pmd_populate(mm, pmd, pte_page) \
#include <asm/processor.h>
#include <asm/cache.h>
-static inline spinlock_t *pgd_spinlock(pgd_t *);
-
/*
* kern_addr_valid(ADDR) tests if ADDR is pointing to valid kernel
* memory. For the return value to be meaningful, ADDR must be >=
/* This is for the serialization of PxTLB broadcasts. At least on the N class
* systems, only one PxTLB inter processor broadcast can be active at any one
- * time on the Merced bus.
-
- * PTE updates are protected by locks in the PMD.
- */
+ * time on the Merced bus. */
extern spinlock_t pa_tlb_flush_lock;
-extern spinlock_t pa_swapper_pg_lock;
#if defined(CONFIG_64BIT) && defined(CONFIG_SMP)
extern int pa_serialize_tlb_flushes;
#else
* within a page table are directly modified. Thus, the following
* hook is made available.
*/
-#define set_pte(pteptr, pteval) \
- do{ \
- *(pteptr) = (pteval); \
- } while(0)
-
-#define set_pte_at(mm, addr, ptep, pteval) \
- do { \
- unsigned long flags; \
- spin_lock_irqsave(pgd_spinlock((mm)->pgd), flags);\
- set_pte(ptep, pteval); \
- purge_tlb_entries(mm, addr); \
- spin_unlock_irqrestore(pgd_spinlock((mm)->pgd), flags);\
+#define set_pte(pteptr, pteval) \
+ do { \
+ *(pteptr) = (pteval); \
+ barrier(); \
+ } while(0)
+
+#define set_pte_at(mm, addr, pteptr, pteval) \
+ do { \
+ *(pteptr) = (pteval); \
+ purge_tlb_entries(mm, addr); \
} while (0)
#endif /* !__ASSEMBLY__ */
#define KERNEL_INITIAL_SIZE (1 << KERNEL_INITIAL_ORDER)
#if CONFIG_PGTABLE_LEVELS == 3
-#define PGD_ORDER 1 /* Number of pages per pgd */
-#define PMD_ORDER 1 /* Number of pages per pmd */
-#define PGD_ALLOC_ORDER (2 + 1) /* first pgd contains pmd */
+#define PMD_ORDER 1
+#define PGD_ORDER 0
#else
-#define PGD_ORDER 1 /* Number of pages per pgd */
-#define PGD_ALLOC_ORDER (PGD_ORDER + 1)
+#define PGD_ORDER 1
#endif
/* Definitions for 3rd level (we use PLD here for Page Lower directory
* able to effectively address 40/42/44-bits of physical address space
* depending on 4k/16k/64k PAGE_SIZE */
#define _PxD_PRESENT_BIT 31
-#define _PxD_ATTACHED_BIT 30
-#define _PxD_VALID_BIT 29
+#define _PxD_VALID_BIT 30
#define PxD_FLAG_PRESENT (1 << xlate_pabit(_PxD_PRESENT_BIT))
-#define PxD_FLAG_ATTACHED (1 << xlate_pabit(_PxD_ATTACHED_BIT))
#define PxD_FLAG_VALID (1 << xlate_pabit(_PxD_VALID_BIT))
#define PxD_FLAG_MASK (0xf)
#define PxD_FLAG_SHIFT (4)
#define pgd_flag(x) (pgd_val(x) & PxD_FLAG_MASK)
#define pgd_address(x) ((unsigned long)(pgd_val(x) &~ PxD_FLAG_MASK) << PxD_VALUE_SHIFT)
-#if CONFIG_PGTABLE_LEVELS == 3
-/* The first entry of the permanent pmd is not there if it contains
- * the gateway marker */
-#define pmd_none(x) (!pmd_val(x) || pmd_flag(x) == PxD_FLAG_ATTACHED)
-#else
#define pmd_none(x) (!pmd_val(x))
-#endif
#define pmd_bad(x) (!(pmd_flag(x) & PxD_FLAG_VALID))
#define pmd_present(x) (pmd_flag(x) & PxD_FLAG_PRESENT)
static inline void pmd_clear(pmd_t *pmd) {
-#if CONFIG_PGTABLE_LEVELS == 3
- if (pmd_flag(*pmd) & PxD_FLAG_ATTACHED)
- /* This is the entry pointing to the permanent pmd
- * attached to the pgd; cannot clear it */
- set_pmd(pmd, __pmd(PxD_FLAG_ATTACHED));
- else
-#endif
set_pmd(pmd, __pmd(0));
}
#define pud_bad(x) (!(pud_flag(x) & PxD_FLAG_VALID))
#define pud_present(x) (pud_flag(x) & PxD_FLAG_PRESENT)
static inline void pud_clear(pud_t *pud) {
-#if CONFIG_PGTABLE_LEVELS == 3
- if(pud_flag(*pud) & PxD_FLAG_ATTACHED)
- /* This is the permanent pmd attached to the pud; cannot
- * free it */
- return;
-#endif
set_pud(pud, __pud(0));
}
#endif
#define __pte_to_swp_entry(pte) ((swp_entry_t) { pte_val(pte) })
#define __swp_entry_to_pte(x) ((pte_t) { (x).val })
-
-static inline spinlock_t *pgd_spinlock(pgd_t *pgd)
-{
- if (unlikely(pgd == swapper_pg_dir))
- return &pa_swapper_pg_lock;
- return (spinlock_t *)((char *)pgd + (PAGE_SIZE << (PGD_ALLOC_ORDER - 1)));
-}
-
-
static inline int ptep_test_and_clear_young(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep)
{
pte_t pte;
- unsigned long flags;
if (!pte_young(*ptep))
return 0;
- spin_lock_irqsave(pgd_spinlock(vma->vm_mm->pgd), flags);
pte = *ptep;
if (!pte_young(pte)) {
- spin_unlock_irqrestore(pgd_spinlock(vma->vm_mm->pgd), flags);
return 0;
}
- set_pte(ptep, pte_mkold(pte));
- purge_tlb_entries(vma->vm_mm, addr);
- spin_unlock_irqrestore(pgd_spinlock(vma->vm_mm->pgd), flags);
+ set_pte_at(vma->vm_mm, addr, ptep, pte_mkold(pte));
return 1;
}
static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
{
pte_t old_pte;
- unsigned long flags;
- spin_lock_irqsave(pgd_spinlock(mm->pgd), flags);
old_pte = *ptep;
- set_pte(ptep, __pte(0));
- purge_tlb_entries(mm, addr);
- spin_unlock_irqrestore(pgd_spinlock(mm->pgd), flags);
+ set_pte_at(mm, addr, ptep, __pte(0));
return old_pte;
}
static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
{
- unsigned long flags;
- spin_lock_irqsave(pgd_spinlock(mm->pgd), flags);
- set_pte(ptep, pte_wrprotect(*ptep));
- purge_tlb_entries(mm, addr);
- spin_unlock_irqrestore(pgd_spinlock(mm->pgd), flags);
+ set_pte_at(mm, addr, ptep, pte_wrprotect(*ptep));
}
#define pte_same(A,B) (pte_val(A) == pte_val(B))
.level 2.0
#endif
- .import pa_tlb_lock,data
- .macro load_pa_tlb_lock reg
- mfctl %cr25,\reg
- addil L%(PAGE_SIZE << (PGD_ALLOC_ORDER - 1)),\reg
+ /* Get aligned page_table_lock address for this mm from cr28/tr4 */
+ .macro get_ptl reg
+ mfctl %cr28,\reg
.endm
/* space_to_prot macro creates a prot id from a space id */
# endif
#endif
dep %r0,31,PAGE_SHIFT,\pmd /* clear offset */
+#if CONFIG_PGTABLE_LEVELS < 3
copy %r0,\pte
+#endif
ldw,s \index(\pmd),\pmd
bb,>=,n \pmd,_PxD_PRESENT_BIT,\fault
dep %r0,31,PxD_FLAG_SHIFT,\pmd /* clear flags */
shladd \index,BITS_PER_PTE_ENTRY,\pmd,\pmd /* pmd is now pte */
.endm
- /* Look up PTE in a 3-Level scheme.
- *
- * Here we implement a Hybrid L2/L3 scheme: we allocate the
- * first pmd adjacent to the pgd. This means that we can
- * subtract a constant offset to get to it. The pmd and pgd
- * sizes are arranged so that a single pmd covers 4GB (giving
- * a full LP64 process access to 8TB) so our lookups are
- * effectively L2 for the first 4GB of the kernel (i.e. for
- * all ILP32 processes and all the kernel for machines with
- * under 4GB of memory) */
+ /* Look up PTE in a 3-Level scheme. */
.macro L3_ptep pgd,pte,index,va,fault
-#if CONFIG_PGTABLE_LEVELS == 3 /* we might have a 2-Level scheme, e.g. with 16kb page size */
+#if CONFIG_PGTABLE_LEVELS == 3
+ copy %r0,\pte
extrd,u \va,63-ASM_PGDIR_SHIFT,ASM_BITS_PER_PGD,\index
- extrd,u,*= \va,63-ASM_PGDIR_SHIFT,64-ASM_PGDIR_SHIFT,%r0
ldw,s \index(\pgd),\pgd
- extrd,u,*= \va,63-ASM_PGDIR_SHIFT,64-ASM_PGDIR_SHIFT,%r0
bb,>=,n \pgd,_PxD_PRESENT_BIT,\fault
- extrd,u,*= \va,63-ASM_PGDIR_SHIFT,64-ASM_PGDIR_SHIFT,%r0
- shld \pgd,PxD_VALUE_SHIFT,\index
- extrd,u,*= \va,63-ASM_PGDIR_SHIFT,64-ASM_PGDIR_SHIFT,%r0
- copy \index,\pgd
- extrd,u,*<> \va,63-ASM_PGDIR_SHIFT,64-ASM_PGDIR_SHIFT,%r0
- ldo ASM_PGD_PMD_OFFSET(\pgd),\pgd
+ shld \pgd,PxD_VALUE_SHIFT,\pgd
#endif
L2_ptep \pgd,\pte,\index,\va,\fault
.endm
- /* Acquire pa_tlb_lock lock and check page is present. */
- .macro tlb_lock spc,ptp,pte,tmp,tmp1,fault
-#ifdef CONFIG_SMP
+ /* Acquire page_table_lock and check page is present. */
+ .macro ptl_lock spc,ptp,pte,tmp,tmp1,fault
+#ifdef CONFIG_TLB_PTLOCK
98: cmpib,COND(=),n 0,\spc,2f
- load_pa_tlb_lock \tmp
+ get_ptl \tmp
1: LDCW 0(\tmp),\tmp1
cmpib,COND(=) 0,\tmp1,1b
nop
3:
.endm
- /* Release pa_tlb_lock lock without reloading lock address.
+ /* Release page_table_lock without reloading lock address.
Note that the values in the register spc are limited to
NR_SPACE_IDS (262144). Thus, the stw instruction always
stores a nonzero value even when register spc is 64 bits.
We use an ordered store to ensure all prior accesses are
performed prior to releasing the lock. */
- .macro tlb_unlock0 spc,tmp
-#ifdef CONFIG_SMP
+ .macro ptl_unlock0 spc,tmp
+#ifdef CONFIG_TLB_PTLOCK
98: or,COND(=) %r0,\spc,%r0
stw,ma \spc,0(\tmp)
99: ALTERNATIVE(98b, 99b, ALT_COND_NO_SMP, INSN_NOP)
#endif
.endm
- /* Release pa_tlb_lock lock. */
- .macro tlb_unlock1 spc,tmp
-#ifdef CONFIG_SMP
-98: load_pa_tlb_lock \tmp
+ /* Release page_table_lock. */
+ .macro ptl_unlock1 spc,tmp
+#ifdef CONFIG_TLB_PTLOCK
+98: get_ptl \tmp
+ ptl_unlock0 \spc,\tmp
99: ALTERNATIVE(98b, 99b, ALT_COND_NO_SMP, INSN_NOP)
- tlb_unlock0 \spc,\tmp
#endif
.endm
L3_ptep ptp,pte,t0,va,dtlb_check_alias_20w
- tlb_lock spc,ptp,pte,t0,t1,dtlb_check_alias_20w
+ ptl_lock spc,ptp,pte,t0,t1,dtlb_check_alias_20w
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L3_ptep ptp,pte,t0,va,nadtlb_check_alias_20w
- tlb_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_20w
+ ptl_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_20w
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,dtlb_check_alias_11
- tlb_lock spc,ptp,pte,t0,t1,dtlb_check_alias_11
+ ptl_lock spc,ptp,pte,t0,t1,dtlb_check_alias_11
update_accessed ptp,pte,t0,t1
make_insert_tlb_11 spc,pte,prot
mtsp t1, %sr1 /* Restore sr1 */
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,nadtlb_check_alias_11
- tlb_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_11
+ ptl_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_11
update_accessed ptp,pte,t0,t1
make_insert_tlb_11 spc,pte,prot
mtsp t1, %sr1 /* Restore sr1 */
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,dtlb_check_alias_20
- tlb_lock spc,ptp,pte,t0,t1,dtlb_check_alias_20
+ ptl_lock spc,ptp,pte,t0,t1,dtlb_check_alias_20
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,nadtlb_check_alias_20
- tlb_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_20
+ ptl_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_20
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L3_ptep ptp,pte,t0,va,itlb_fault
- tlb_lock spc,ptp,pte,t0,t1,itlb_fault
+ ptl_lock spc,ptp,pte,t0,t1,itlb_fault
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot,t1
iitlbt pte,prot
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L3_ptep ptp,pte,t0,va,naitlb_check_alias_20w
- tlb_lock spc,ptp,pte,t0,t1,naitlb_check_alias_20w
+ ptl_lock spc,ptp,pte,t0,t1,naitlb_check_alias_20w
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot,t1
iitlbt pte,prot
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,itlb_fault
- tlb_lock spc,ptp,pte,t0,t1,itlb_fault
+ ptl_lock spc,ptp,pte,t0,t1,itlb_fault
update_accessed ptp,pte,t0,t1
make_insert_tlb_11 spc,pte,prot
mtsp t1, %sr1 /* Restore sr1 */
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,naitlb_check_alias_11
- tlb_lock spc,ptp,pte,t0,t1,naitlb_check_alias_11
+ ptl_lock spc,ptp,pte,t0,t1,naitlb_check_alias_11
update_accessed ptp,pte,t0,t1
make_insert_tlb_11 spc,pte,prot
mtsp t1, %sr1 /* Restore sr1 */
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,itlb_fault
- tlb_lock spc,ptp,pte,t0,t1,itlb_fault
+ ptl_lock spc,ptp,pte,t0,t1,itlb_fault
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot,t1
iitlbt pte,prot
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,naitlb_check_alias_20
- tlb_lock spc,ptp,pte,t0,t1,naitlb_check_alias_20
+ ptl_lock spc,ptp,pte,t0,t1,naitlb_check_alias_20
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot,t1
iitlbt pte,prot
- tlb_unlock1 spc,t0
+ ptl_unlock1 spc,t0
rfir
nop
L3_ptep ptp,pte,t0,va,dbit_fault
- tlb_lock spc,ptp,pte,t0,t1,dbit_fault
+ ptl_lock spc,ptp,pte,t0,t1,dbit_fault
update_dirty ptp,pte,t1
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
- tlb_unlock0 spc,t0
+ ptl_unlock0 spc,t0
rfir
nop
#else
L2_ptep ptp,pte,t0,va,dbit_fault
- tlb_lock spc,ptp,pte,t0,t1,dbit_fault
+ ptl_lock spc,ptp,pte,t0,t1,dbit_fault
update_dirty ptp,pte,t1
make_insert_tlb_11 spc,pte,prot
mtsp t1, %sr1 /* Restore sr1 */
- tlb_unlock0 spc,t0
+ ptl_unlock0 spc,t0
rfir
nop
L2_ptep ptp,pte,t0,va,dbit_fault
- tlb_lock spc,ptp,pte,t0,t1,dbit_fault
+ ptl_lock spc,ptp,pte,t0,t1,dbit_fault
update_dirty ptp,pte,t1
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
- tlb_unlock0 spc,t0
+ ptl_unlock0 spc,t0
rfir
nop
#endif