static void beat_lpar_hpte_invalidate(unsigned long slot, unsigned long va,
int psize, int ssize, int local)
{
unsigned long want_v;
unsigned long lpar_rc;
u64 dummy1, dummy2;
unsigned long flags;
DBG_LOW(" inval : slot=%lx, va=%016lx, psize: %d, local: %d\n",
slot, va, psize, local);
want_v = hpte_encode_v(va, psize, MMU_SEGSIZE_256M);
atomic_spin_lock_irqsave(&beat_htab_lock, flags);
dummy1 = beat_lpar_hpte_getword0(slot);
if ((dummy1 & ~0x7FUL) != (want_v & ~0x7FUL)) {
DBG_LOW("not found !\n");
atomic_spin_unlock_irqrestore(&beat_htab_lock, flags);
return;
}
lpar_rc = beat_write_htab_entry(0, slot, 0, 0, HPTE_V_VALID, 0,
&dummy1, &dummy2);
atomic_spin_unlock_irqrestore(&beat_htab_lock, flags);
BUG_ON(lpar_rc != 0);
}
/*
* NOTE: for updatepp ops we are fortunate that the linux "newpp" bits and
* the low 3 bits of flags happen to line up. So no transform is needed.
* We can probably optimize here and assume the high bits of newpp are
* already zero. For now I am paranoid.
*/
static long beat_lpar_hpte_updatepp_v3(unsigned long slot,
unsigned long newpp,
unsigned long va,
int psize, int ssize, int local)
{
unsigned long lpar_rc;
unsigned long want_v;
unsigned long pss;
want_v = hpte_encode_v(va, psize, MMU_SEGSIZE_256M);
pss = (psize == MMU_PAGE_4K) ? -1UL : mmu_psize_defs[psize].penc;
DBG_LOW(" update: "
"avpnv=%016lx, slot=%016lx, psize: %d, newpp %016lx ... ",
want_v & HPTE_V_AVPN, slot, psize, newpp);
lpar_rc = beat_update_htab_permission3(0, slot, want_v, pss, 7, newpp);
if (lpar_rc == 0xfffffff7) {
DBG_LOW("not found !\n");
return -1;
}
DBG_LOW("ok\n");
BUG_ON(lpar_rc != 0);
return 0;
}
static long beat_lpar_hpte_insert(unsigned long hpte_group,
unsigned long va, unsigned long pa,
unsigned long rflags, unsigned long vflags,
int psize, int ssize)
{
unsigned long lpar_rc;
unsigned long slot;
unsigned long hpte_v, hpte_r;
/* same as iseries */
if (vflags & HPTE_V_SECONDARY)
return -1;
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW("hpte_insert(group=%lx, va=%016lx, pa=%016lx, "
"rflags=%lx, vflags=%lx, psize=%d)\n",
hpte_group, va, pa, rflags, vflags, psize);
hpte_v = hpte_encode_v(va, psize, MMU_SEGSIZE_256M) |
vflags | HPTE_V_VALID;
hpte_r = hpte_encode_r(pa, psize) | rflags;
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW(" hpte_v=%016lx, hpte_r=%016lx\n", hpte_v, hpte_r);
if (rflags & (_PAGE_GUARDED|_PAGE_NO_CACHE))
hpte_r &= ~_PAGE_COHERENT;
spin_lock(&beat_htab_lock);
lpar_rc = beat_read_mask(hpte_group);
if (lpar_rc == 0) {
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW(" full\n");
spin_unlock(&beat_htab_lock);
return -1;
}
lpar_rc = beat_insert_htab_entry(0, hpte_group, lpar_rc << 48,
hpte_v, hpte_r, &slot);
spin_unlock(&beat_htab_lock);
/*
* Since we try and ioremap PHBs we don't own, the pte insert
* will fail. However we must catch the failure in hash_page
* or we will loop forever, so return -2 in this case.
*/
if (unlikely(lpar_rc != 0)) {
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW(" lpar err %lx\n", lpar_rc);
return -2;
}
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW(" -> slot: %lx\n", slot);
/* We have to pass down the secondary bucket bit here as well */
return (slot ^ hpte_group) & 15;
}
void flush_hash_page(unsigned long va, real_pte_t pte, int psize, int local)
{
unsigned long hash, index, shift, hidx, slot;
DBG_LOW("flush_hash_page(va=%016x)\n", va);
pte_iterate_hashed_subpages(pte, psize, va, index, shift) {
hash = hpt_hash(va, shift);
hidx = __rpte_to_hidx(pte, index);
if (hidx & _PTEIDX_SECONDARY)
hash = ~hash;
slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
slot += hidx & _PTEIDX_GROUP_IX;
DBG_LOW(" sub %d: hash=%x, hidx=%x\n", index, slot, hidx);
ppc_md.hpte_invalidate(slot, va, psize, local);
} pte_iterate_hashed_end();
static long beat_lpar_hpte_insert_v3(unsigned long hpte_group,
unsigned long vpn, unsigned long pa,
unsigned long rflags, unsigned long vflags,
int psize, int ssize)
{
unsigned long lpar_rc;
u64 hpte_v, hpte_r, slot;
if (vflags & HPTE_V_SECONDARY)
return -1;
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW("hpte_insert(group=%lx, vpn=%016lx, pa=%016lx, "
"rflags=%lx, vflags=%lx, psize=%d)\n",
hpte_group, vpn, pa, rflags, vflags, psize);
hpte_v = hpte_encode_v(vpn, psize, MMU_SEGSIZE_256M) |
vflags | HPTE_V_VALID;
hpte_r = hpte_encode_r(pa, psize) | rflags;
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW(" hpte_v=%016lx, hpte_r=%016lx\n", hpte_v, hpte_r);
if (rflags & _PAGE_NO_CACHE)
hpte_r &= ~_PAGE_COHERENT;
/* insert into not-volted entry */
lpar_rc = beat_insert_htab_entry3(0, hpte_group, hpte_v, hpte_r,
HPTE_V_BOLTED, 0, &slot);
/*
* Since we try and ioremap PHBs we don't own, the pte insert
* will fail. However we must catch the failure in hash_page
* or we will loop forever, so return -2 in this case.
*/
if (unlikely(lpar_rc != 0)) {
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW(" lpar err %lx\n", lpar_rc);
return -2;
}
if (!(vflags & HPTE_V_BOLTED))
DBG_LOW(" -> slot: %lx\n", slot);
/* We have to pass down the secondary bucket bit here as well */
return (slot ^ hpte_group) & 15;
}
void hash_preload(struct mm_struct *mm, unsigned long ea,
unsigned long access, unsigned long trap)
{
unsigned long vsid;
void *pgdir;
pte_t *ptep;
cpumask_t mask;
unsigned long flags;
int local = 0;
/* We don't want huge pages prefaulted for now
*/
if (unlikely(in_hugepage_area(mm->context, ea)))
return;
DBG_LOW("hash_preload(mm=%p, mm->pgdir=%p, ea=%016lx, access=%lx,"
" trap=%lx\n", mm, mm->pgd, ea, access, trap);
/* Get PTE, VSID, access mask */
pgdir = mm->pgd;
if (pgdir == NULL)
return;
ptep = find_linux_pte(pgdir, ea);
if (!ptep)
return;
vsid = get_vsid(mm->context.id, ea);
/* Hash it in */
local_irq_save(flags);
mask = cpumask_of_cpu(smp_processor_id());
if (cpus_equal(mm->cpu_vm_mask, mask))
local = 1;
#ifndef CONFIG_PPC_64K_PAGES
__hash_page_4K(ea, access, vsid, ptep, trap, local);
#else
if (mmu_ci_restrictions) {
/* If this PTE is non-cacheable, switch to 4k */
if (mm->context.user_psize == MMU_PAGE_64K &&
(pte_val(*ptep) & _PAGE_NO_CACHE)) {
mm->context.user_psize = MMU_PAGE_4K;
mm->context.sllp = SLB_VSID_USER |
mmu_psize_defs[MMU_PAGE_4K].sllp;
get_paca()->context = mm->context;
slb_flush_and_rebolt();
}
}
if (mm->context.user_psize == MMU_PAGE_64K)
__hash_page_64K(ea, access, vsid, ptep, trap, local);
else
__hash_page_4K(ea, access, vsid, ptep, trap, local);
#endif /* CONFIG_PPC_64K_PAGES */
local_irq_restore(flags);
}
/*
* NOTE: for updatepp ops we are fortunate that the linux "newpp" bits and
* the low 3 bits of flags happen to line up. So no transform is needed.
* We can probably optimize here and assume the high bits of newpp are
* already zero. For now I am paranoid.
*/
static long beat_lpar_hpte_updatepp(unsigned long slot,
unsigned long newpp,
unsigned long vpn,
int psize, int apsize,
int ssize, unsigned long flags)
{
unsigned long lpar_rc;
u64 dummy0, dummy1;
unsigned long want_v;
want_v = hpte_encode_avpn(vpn, psize, MMU_SEGSIZE_256M);
DBG_LOW(" update: "
"avpnv=%016lx, slot=%016lx, psize: %d, newpp %016lx ... ",
want_v & HPTE_V_AVPN, slot, psize, newpp);
raw_spin_lock(&beat_htab_lock);
dummy0 = beat_lpar_hpte_getword0(slot);
if ((dummy0 & ~0x7FUL) != (want_v & ~0x7FUL)) {
DBG_LOW("not found !\n");
raw_spin_unlock(&beat_htab_lock);
return -1;
}
lpar_rc = beat_write_htab_entry(0, slot, 0, newpp, 0, 7, &dummy0,
&dummy1);
raw_spin_unlock(&beat_htab_lock);
if (lpar_rc != 0 || dummy0 == 0) {
DBG_LOW("not found !\n");
return -1;
}
DBG_LOW("ok %lx %lx\n", dummy0, dummy1);
BUG_ON(lpar_rc != 0);
return 0;
}
static void beat_lpar_hpte_invalidate_v3(unsigned long slot, unsigned long va,
int psize, int ssize, int local)
{
unsigned long want_v;
unsigned long lpar_rc;
unsigned long pss;
DBG_LOW(" inval : slot=%lx, va=%016lx, psize: %d, local: %d\n",
slot, va, psize, local);
want_v = hpte_encode_v(va, psize, MMU_SEGSIZE_256M);
pss = (psize == MMU_PAGE_4K) ? -1UL : mmu_psize_defs[psize].penc;
lpar_rc = beat_invalidate_htab_entry3(0, slot, want_v, pss);
/* E_busy can be valid output: page may be already replaced */
BUG_ON(lpar_rc != 0 && lpar_rc != 0xfffffff7);
}
static long beat_lpar_hpte_remove(unsigned long hpte_group)
{
DBG_LOW("hpte_remove(group=%lx)\n", hpte_group);
return -1;
}
/* Result code is:
* 0 - handled
* 1 - normal page fault
* -1 - critical hash insertion error
*/
int hash_page(unsigned long ea, unsigned long access, unsigned long trap)
{
void *pgdir;
unsigned long vsid;
struct mm_struct *mm;
pte_t *ptep;
cpumask_t tmp;
int rc, user_region = 0, local = 0;
int psize;
DBG_LOW("hash_page(ea=%016lx, access=%lx, trap=%lx\n",
ea, access, trap);
if ((ea & ~REGION_MASK) >= PGTABLE_RANGE) {
DBG_LOW(" out of pgtable range !\n");
return 1;
}
/* Get region & vsid */
switch (REGION_ID(ea)) {
case USER_REGION_ID:
user_region = 1;
mm = current->mm;
if (! mm) {
DBG_LOW(" user region with no mm !\n");
return 1;
}
vsid = get_vsid(mm->context.id, ea);
psize = mm->context.user_psize;
break;
case VMALLOC_REGION_ID:
mm = &init_mm;
vsid = get_kernel_vsid(ea);
if (ea < VMALLOC_END)
psize = mmu_vmalloc_psize;
else
psize = mmu_io_psize;
break;
default:
/* Not a valid range
* Send the problem up to do_page_fault
*/
return 1;
}
DBG_LOW(" mm=%p, mm->pgdir=%p, vsid=%016lx\n", mm, mm->pgd, vsid);
/* Get pgdir */
pgdir = mm->pgd;
if (pgdir == NULL)
return 1;
/* Check CPU locality */
tmp = cpumask_of_cpu(smp_processor_id());
if (user_region && cpus_equal(mm->cpu_vm_mask, tmp))
local = 1;
/* Handle hugepage regions */
if (unlikely(in_hugepage_area(mm->context, ea))) {
DBG_LOW(" -> huge page !\n");
return hash_huge_page(mm, access, ea, vsid, local, trap);
}
/* Get PTE and page size from page tables */
ptep = find_linux_pte(pgdir, ea);
if (ptep == NULL || !pte_present(*ptep)) {
DBG_LOW(" no PTE !\n");
return 1;
}
#ifndef CONFIG_PPC_64K_PAGES
DBG_LOW(" i-pte: %016lx\n", pte_val(*ptep));
#else
DBG_LOW(" i-pte: %016lx %016lx\n", pte_val(*ptep),
pte_val(*(ptep + PTRS_PER_PTE)));
#endif
/* Pre-check access permissions (will be re-checked atomically
* in __hash_page_XX but this pre-check is a fast path
*/
if (access & ~pte_val(*ptep)) {
DBG_LOW(" no access !\n");
return 1;
}
/* Do actual hashing */
#ifndef CONFIG_PPC_64K_PAGES
rc = __hash_page_4K(ea, access, vsid, ptep, trap, local);
#else
if (mmu_ci_restrictions) {
/* If this PTE is non-cacheable, switch to 4k */
if (psize == MMU_PAGE_64K &&
(pte_val(*ptep) & _PAGE_NO_CACHE)) {
if (user_region) {
psize = MMU_PAGE_4K;
mm->context.user_psize = MMU_PAGE_4K;
mm->context.sllp = SLB_VSID_USER |
mmu_psize_defs[MMU_PAGE_4K].sllp;
} else if (ea < VMALLOC_END) {
/*
* some driver did a non-cacheable mapping
* in vmalloc space, so switch vmalloc
* to 4k pages
*/
printk(KERN_ALERT "Reducing vmalloc segment "
"to 4kB pages because of "
"non-cacheable mapping\n");
psize = mmu_vmalloc_psize = MMU_PAGE_4K;
}
}
if (user_region) {
if (psize != get_paca()->context.user_psize) {
get_paca()->context = mm->context;
slb_flush_and_rebolt();
}
} else if (get_paca()->vmalloc_sllp !=
mmu_psize_defs[mmu_vmalloc_psize].sllp) {
get_paca()->vmalloc_sllp =
mmu_psize_defs[mmu_vmalloc_psize].sllp;
slb_flush_and_rebolt();
}
}
if (psize == MMU_PAGE_64K)
rc = __hash_page_64K(ea, access, vsid, ptep, trap, local);
else
rc = __hash_page_4K(ea, access, vsid, ptep, trap, local);
#endif /* CONFIG_PPC_64K_PAGES */
#ifndef CONFIG_PPC_64K_PAGES
DBG_LOW(" o-pte: %016lx\n", pte_val(*ptep));
#else
DBG_LOW(" o-pte: %016lx %016lx\n", pte_val(*ptep),
pte_val(*(ptep + PTRS_PER_PTE)));
#endif
DBG_LOW(" -> rc=%d\n", rc);
return rc;
}
