Aren/haiku
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

kernel/vm: Avoid committing memory in vm_map_file for PRIVATE_MAP without PROT_WRITE.

Browse Source 
Instead, rely on commitment being done later, when the
protections are changed. set_area_protection() already
did just that, but set_memory_protection did not, so
it is implemented here.

Fixes #18733.

Change-Id: Ia58aee93faf1296fce69d723b12d0fa0a8440706
Reviewed-on: https://review.haiku-os.org/c/haiku/+/7339
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
This commit is contained in:
Augustin Cavalier 2024-01-16 19:38:59 -05:00 committed by waddlesplash
parent 33fb08b1a5
commit a2270c7035
1 changed files with 42 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

47
src/system/kernel/vm/vm.cpp
View File

@ -2159,14 +2159,17 @@ _vm_map_file(team_id team, const char* name, void** _address,
if (addressSpec != B_EXACT_ADDRESS)
unmapAddressRange = false;
uint32 mappingFlags = 0;
if (unmapAddressRange)
mappingFlags |= CREATE_AREA_UNMAP_ADDRESS_RANGE;
if (fd < 0) {
uint32 flags = unmapAddressRange ? CREATE_AREA_UNMAP_ADDRESS_RANGE : 0;
virtual_address_restrictions virtualRestrictions = {};
virtualRestrictions.address = *_address;
virtualRestrictions.address_specification = addressSpec;
physical_address_restrictions physicalRestrictions = {};
return vm_create_anonymous_area(team, name, size, B_NO_LOCK, protection,
flags, 0, &virtualRestrictions, &physicalRestrictions, kernel,
mappingFlags, 0, &virtualRestrictions, &physicalRestrictions, kernel,
_address);
}
@ -2187,11 +2190,16 @@ _vm_map_file(team_id team, const char* name, void** _address,
}
uint32 protectionMax = 0;
if (mapping != REGION_PRIVATE_MAP) {
if (mapping == REGION_NO_PRIVATE_MAP) {
if ((openMode & O_ACCMODE) == O_RDWR)
protectionMax = protection | B_USER_PROTECTION;
else
protectionMax = protection | (B_USER_PROTECTION & ~B_WRITE_AREA);
} else if (mapping == REGION_PRIVATE_MAP) {
// For privately mapped read-only regions, skip committing memory.
// (If protections are changed later on, memory will be committed then.)
if ((protection & B_WRITE_AREA) == 0)
mappingFlags |= CREATE_AREA_DONT_COMMIT_MEMORY;
}
// get the vnode for the object, this also grabs a ref to it
@ -2260,8 +2268,7 @@ _vm_map_file(team_id team, const char* name, void** _address,
addressRestrictions.address = *_address;
addressRestrictions.address_specification = addressSpec;
status = map_backing_store(locker.AddressSpace(), cache, offset, name, size,
0, protection, protectionMax, mapping,
unmapAddressRange ? CREATE_AREA_UNMAP_ADDRESS_RANGE : 0,
0, protection, protectionMax, mapping, mappingFlags,
&addressRestrictions, kernel, &area, _address);
if (status != B_OK || mapping == REGION_PRIVATE_MAP) {
@ -6858,6 +6865,7 @@ _user_set_memory_protection(void* _address, size_t size, uint32 protection)
if (area->protection == protection)
continue;
if (offset == 0 && rangeSize == area->Size()) {
// The whole area is covered: let set_area_protection handle it.
status_t status = vm_set_area_protection(area->address_space->ID(),
area->id, protection, false);
if (status != B_OK)
@ -6876,6 +6884,35 @@ _user_set_memory_protection(void* _address, size_t size, uint32 protection)
VMCacheChainLocker cacheChainLocker(topCache);
cacheChainLocker.LockAllSourceCaches();
// Adjust the committed size, if necessary.
if (topCache->source != NULL && topCache->temporary) {
const bool becomesWritable = (protection & B_WRITE_AREA) != 0;
ssize_t commitmentChange = 0;
for (addr_t pageAddress = area->Base() + offset;
pageAddress < currentAddress; pageAddress += B_PAGE_SIZE) {
if (topCache->LookupPage(pageAddress) != NULL) {
// This page should already be accounted for in the commitment.
continue;
}
const bool isWritable
= (get_area_page_protection(area, pageAddress) & B_WRITE_AREA) != 0;
if (becomesWritable && !isWritable)
commitmentChange += B_PAGE_SIZE;
else if (!becomesWritable && isWritable)
commitmentChange -= B_PAGE_SIZE;
}
if (commitmentChange != 0) {
const off_t newCommitment = topCache->committed_size + commitmentChange;
ASSERT(newCommitment <= (topCache->virtual_end - topCache->virtual_base));
status_t status = topCache->Commit(newCommitment, VM_PRIORITY_USER);
if (status != B_OK)
return status;
}
}
for (addr_t pageAddress = area->Base() + offset;
pageAddress < currentAddress; pageAddress += B_PAGE_SIZE) {
map->Lock();