d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Only demote when needed in the arm64 pmap_change_props_locked

Browse Source 
When changing page table properties there is no need to demote a
level 1 or level 2 block if we are changing the entire memory range the
block is mapping. In this case just change the block directly.

Reported by:	alc, kib, markj
Sponsored by:	The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32339
This commit is contained in:
Andrew Turner 2021-10-06 17:38:22 +01:00
parent a85ce4ad72
commit 806a88e742
1 changed files with 49 additions and 35 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

84
sys/arm64/arm64/pmap.c
View File

@ -5982,7 +5982,8 @@ pmap_change_props_locked(vm_offset_t va, vm_size_t size, vm_prot_t prot,
int mode)
{
vm_offset_t base, offset, tmpva;
pt_entry_t l3, *pte, *newpte;
vm_size_t pte_size;
pt_entry_t pte, *ptep, *newpte;
pt_entry_t bits, mask;
int lvl, rv;
@ -6028,11 +6029,11 @@ pmap_change_props_locked(vm_offset_t va, vm_size_t size, vm_prot_t prot,
}
for (tmpva = base; tmpva < base + size; ) {
pte = pmap_pte(kernel_pmap, tmpva, &lvl);
if (pte == NULL)
ptep = pmap_pte(kernel_pmap, tmpva, &lvl);
if (ptep == NULL)
return (EINVAL);
if ((pmap_load(pte) & mask) == bits) {
if ((pmap_load(ptep) & mask) == bits) {
/*
* We already have the correct attribute,
* ignore this entry.
@ -6059,47 +6060,60 @@ pmap_change_props_locked(vm_offset_t va, vm_size_t size, vm_prot_t prot,
default:
panic("Invalid DMAP table level: %d\n", lvl);
case 1:
newpte = pmap_demote_l1(kernel_pmap, pte,
if ((tmpva & L1_OFFSET) == 0 &&
(base + size - tmpva) >= L1_SIZE) {
pte_size = L1_SIZE;
break;
}
newpte = pmap_demote_l1(kernel_pmap, ptep,
tmpva & ~L1_OFFSET);
if (newpte == NULL)
return (EINVAL);
pte = pmap_l1_to_l2(pte, tmpva);
ptep = pmap_l1_to_l2(ptep, tmpva);
/* FALLTHROUGH */
case 2:
newpte = pmap_demote_l2(kernel_pmap, pte,
if ((tmpva & L2_OFFSET) == 0 &&
(base + size - tmpva) >= L2_SIZE) {
pte_size = L2_SIZE;
break;
}
newpte = pmap_demote_l2(kernel_pmap, ptep,
tmpva);
if (newpte == NULL)
return (EINVAL);
pte = pmap_l2_to_l3(pte, tmpva);
ptep = pmap_l2_to_l3(ptep, tmpva);
/* FALLTHROUGH */
case 3:
/* Update the entry */
l3 = pmap_load(pte);
l3 &= ~mask;
l3 |= bits;
pmap_update_entry(kernel_pmap, pte, l3, tmpva,
PAGE_SIZE);
if (!VIRT_IN_DMAP(tmpva)) {
/*
* Keep the DMAP memory in sync.
*/
rv = pmap_change_props_locked(
PHYS_TO_DMAP(l3 & ~ATTR_MASK),
L3_SIZE, prot, mode);
if (rv != 0)
return (rv);
}
/*
* If moving to a non-cacheable entry flush
* the cache.
*/
if (mode == VM_MEMATTR_UNCACHEABLE)
cpu_dcache_wbinv_range(tmpva, L3_SIZE);
pte_size = PAGE_SIZE;
break;
}
tmpva += PAGE_SIZE;
/* Update the entry */
pte = pmap_load(ptep);
pte &= ~mask;
pte |= bits;
pmap_update_entry(kernel_pmap, ptep, pte, tmpva,
pte_size);
if (!VIRT_IN_DMAP(tmpva)) {
/*
* Keep the DMAP memory in sync.
*/
rv = pmap_change_props_locked(
PHYS_TO_DMAP(pte & ~ATTR_MASK), pte_size,
prot, mode);
if (rv != 0)
return (rv);
}
/*
* If moving to a non-cacheable entry flush
* the cache.
*/
if (mode == VM_MEMATTR_UNCACHEABLE)
cpu_dcache_wbinv_range(tmpva, pte_size);
tmpva += pte_size;
}
}