RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-07-24 14:37:43 +00:00
Code Issues Activity Actions

Revert "Kernel: Send SIGBUS to threads that use after valid Inode mmaped range"

Browse source 
This reverts commit 0c675192c9.
This commit is contained in:
Liav A 2022-09-24 14:31:32 +03:00 • committed by Andreas Kling
parent 4d7e4e5da8
commit 6e26e9fb29
4 changed files with 33 additions and 57 deletions
Download patch file Download diff file Expand all files Collapse all files

13
Kernel/Arch/x86/common/Interrupts.cpp
View file

@ -317,7 +317,7 @@ void page_fault_handler(TrapFrame* trap)
PageFault fault { regs.exception_code, VirtualAddress { fault_address } };
auto response = MM.handle_page_fault(fault);
if (response == PageFaultResponse::ShouldCrash || response == PageFaultResponse::OutOfMemory || response == PageFaultResponse::BusError) {
if (response == PageFaultResponse::ShouldCrash || response == PageFaultResponse::OutOfMemory) {
if (faulted_in_kernel && handle_safe_access_fault(regs, fault_address)) {
// If this would be a ring0 (kernel) fault and the fault was triggered by
// safe_memcpy, safe_strnlen, or safe_memset then we resume execution at
@ -325,11 +325,6 @@ void page_fault_handler(TrapFrame* trap)
return;
}
if (response == PageFaultResponse::BusError && current_thread->has_signal_handler(SIGBUS)) {
current_thread->send_urgent_signal_to_self(SIGBUS);
return;
}
if (response != PageFaultResponse::OutOfMemory && current_thread) {
if (current_thread->has_signal_handler(SIGSEGV)) {
current_thread->send_urgent_signal_to_self(SIGSEGV);
@ -346,9 +341,6 @@ void page_fault_handler(TrapFrame* trap)
constexpr FlatPtr free_scrub_pattern = explode_byte(FREE_SCRUB_BYTE);
constexpr FlatPtr kmalloc_scrub_pattern = explode_byte(KMALLOC_SCRUB_BYTE);
constexpr FlatPtr kfree_scrub_pattern = explode_byte(KFREE_SCRUB_BYTE);
if (response == PageFaultResponse::BusError) {
dbgln("Note: Address {} is an access to undefined memory range of a Inode-backed VMObject", VirtualAddress(fault_address));
} else {
if ((fault_address & 0xffff0000) == (malloc_scrub_pattern & 0xffff0000)) {
dbgln("Note: Address {} looks like it may be uninitialized malloc() memory", VirtualAddress(fault_address));
} else if ((fault_address & 0xffff0000) == (free_scrub_pattern & 0xffff0000)) {
@ -381,7 +373,6 @@ void page_fault_handler(TrapFrame* trap)
dbgln("Note: Address {} looks like it may be a recently destroyed NonnullLockRefPtr", VirtualAddress(fault_address));
}
}
}
if (current_thread) {
auto& current_process = current_thread->process();
@ -399,8 +390,6 @@ void page_fault_handler(TrapFrame* trap)
}
}
if (response == PageFaultResponse::BusError)
return handle_crash(regs, "Page Fault (Bus Error)", SIGBUS, false);
return handle_crash(regs, "Page Fault", SIGSEGV, response == PageFaultResponse::OutOfMemory);
} else if (response == PageFaultResponse::Continue) {
dbgln_if(PAGE_FAULT_DEBUG, "Continuing after resolved page fault");

1
Kernel/Memory/PageFaultResponse.h
View file

@ -10,7 +10,6 @@ namespace Kernel {
enum class PageFaultResponse {
ShouldCrash,
BusError,
OutOfMemory,
Continue,
};

11
Kernel/Memory/Region.cpp
View file

@ -362,7 +362,7 @@ PageFaultResponse Region::handle_fault(PageFault const& fault)
}
if (vmobject().is_inode()) {
dbgln_if(PAGE_FAULT_DEBUG, "NP(inode) fault in Region({})[{}]", this, page_index_in_region);
return handle_inode_fault(page_index_in_region, offset_in_page_from_address(fault.vaddr()));
return handle_inode_fault(page_index_in_region);
}
SpinlockLocker vmobject_locker(vmobject().m_lock);
@ -462,7 +462,7 @@ PageFaultResponse Region::handle_cow_fault(size_t page_index_in_region)
return response;
}
PageFaultResponse Region::handle_inode_fault(size_t page_index_in_region, size_t offset_in_page_in_region)
PageFaultResponse Region::handle_inode_fault(size_t page_index_in_region)
{
VERIFY(vmobject().is_inode());
VERIFY(!g_scheduler_lock.is_locked_by_current_processor());
@ -475,13 +475,6 @@ PageFaultResponse Region::handle_inode_fault(size_t page_index_in_region, size_t
{
// NOTE: The VMObject lock is required when manipulating the VMObject's physical page slot.
SpinlockLocker locker(inode_vmobject.m_lock);
if (inode_vmobject.inode().size() == 0)
return PageFaultResponse::BusError;
auto fault_vaddr = vaddr_from_page_index(page_index_in_vmobject).offset(offset_in_page_in_region);
auto inode_last_valid_address = vaddr().offset(inode_vmobject.inode().size());
if (inode_last_valid_address < fault_vaddr)
return PageFaultResponse::BusError;
if (!vmobject_physical_page_slot.is_null()) {
dbgln_if(PAGE_FAULT_DEBUG, "handle_inode_fault: Page faulted in by someone else before reading, remapping.");
if (!remap_vmobject_page(page_index_in_vmobject, *vmobject_physical_page_slot))

7
Kernel/Memory/Region.h
View file

@ -122,11 +122,6 @@ public:
return (vaddr - m_range.base()).get() / PAGE_SIZE;
}
[[nodiscard]] unsigned offset_in_page_from_address(VirtualAddress vaddr) const
{
return (vaddr - m_range.base()).get() % PAGE_SIZE;
}
[[nodiscard]] VirtualAddress vaddr_from_page_index(size_t page_index) const
{
return vaddr().offset(page_index * PAGE_SIZE);
@ -224,7 +219,7 @@ private:
}
[[nodiscard]] PageFaultResponse handle_cow_fault(size_t page_index);
[[nodiscard]] PageFaultResponse handle_inode_fault(size_t page_index, size_t offset_in_page_in_region);
[[nodiscard]] PageFaultResponse handle_inode_fault(size_t page_index);
[[nodiscard]] PageFaultResponse handle_zero_fault(size_t page_index, PhysicalPage& page_in_slot_at_time_of_fault);
[[nodiscard]] bool map_individual_page_impl(size_t page_index);