RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-10-31 11:12:45 +00:00
Code Issues Activity Actions
serenity/Kernel/Syscalls/execve.cpp
Max Wipfli ee342f5ec3 Kernel: Replace usage of LexicalPath with KLexicalPath 
This replaces all uses of LexicalPath in the Kernel with the functions
from KLexicalPath. This also allows the Kernel to stop including
AK::LexicalPath.
2021-07-07 15:32:17 +02:00

976 lines
38 KiB
C++
Raw Blame History

/*
* Copyright (c) 2018-2021, Andreas Kling <kling@serenityos.org>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
#include <AK/ScopeGuard.h>
#include <AK/TemporaryChange.h>
#include <AK/WeakPtr.h>
#include <Kernel/Debug.h>
#include <Kernel/FileSystem/Custody.h>
#include <Kernel/FileSystem/FileDescription.h>
#include <Kernel/Panic.h>
#include <Kernel/PerformanceManager.h>
#include <Kernel/Process.h>
#include <Kernel/Random.h>
#include <Kernel/Time/TimeManagement.h>
#include <Kernel/VM/AllocationStrategy.h>
#include <Kernel/VM/MemoryManager.h>
#include <Kernel/VM/PageDirectory.h>
#include <Kernel/VM/Region.h>
#include <Kernel/VM/SharedInodeVMObject.h>
#include <LibC/limits.h>
#include <LibELF/AuxiliaryVector.h>
#include <LibELF/Image.h>
#include <LibELF/Validation.h>
namespace Kernel {
extern Region* g_signal_trampoline_region;
struct LoadResult {
OwnPtr<Space> space;
FlatPtr load_base { 0 };
FlatPtr entry_eip { 0 };
size_t size { 0 };
WeakPtr<Region> tls_region;
size_t tls_size { 0 };
size_t tls_alignment { 0 };
WeakPtr<Region> stack_region;
};
static Vector<ELF::AuxiliaryValue> generate_auxiliary_vector(FlatPtr load_base, FlatPtr entry_eip, uid_t uid, uid_t euid, gid_t gid, gid_t egid, String executable_path, int main_program_fd);
static bool validate_stack_size(const Vector<String>& arguments, const Vector<String>& environment)
{
size_t total_arguments_size = 0;
size_t total_environment_size = 0;
for (auto& a : arguments)
total_arguments_size += a.length() + 1;
for (auto& e : environment)
total_environment_size += e.length() + 1;
total_arguments_size += sizeof(char*) * (arguments.size() + 1);
total_environment_size += sizeof(char*) * (environment.size() + 1);
static constexpr size_t max_arguments_size = Thread::default_userspace_stack_size / 8;
static constexpr size_t max_environment_size = Thread::default_userspace_stack_size / 8;
if (total_arguments_size > max_arguments_size)
return false;
if (total_environment_size > max_environment_size)
return false;
// FIXME: This doesn't account for the size of the auxiliary vector
return true;
}
static KResultOr<FlatPtr> make_userspace_context_for_main_thread([[maybe_unused]] ThreadRegisters& regs, Region& region, Vector<String> arguments,
Vector<String> environment, Vector<ELF::AuxiliaryValue> auxiliary_values)
{
FlatPtr new_sp = region.range().end().get();
// Add some bits of randomness to the user stack pointer.
new_sp -= round_up_to_power_of_two(get_fast_random<u32>() % 4096, 16);
auto push_on_new_stack = [&new_sp](FlatPtr value) {
new_sp -= sizeof(FlatPtr);
Userspace<FlatPtr*> stack_ptr = new_sp;
return copy_to_user(stack_ptr, &value);
};
auto push_aux_value_on_new_stack = [&new_sp](auxv_t value) {
new_sp -= sizeof(auxv_t);
Userspace<auxv_t*> stack_ptr = new_sp;
return copy_to_user(stack_ptr, &value);
};
auto push_string_on_new_stack = [&new_sp](const String& string) {
new_sp -= round_up_to_power_of_two(string.length() + 1, sizeof(FlatPtr));
Userspace<FlatPtr*> stack_ptr = new_sp;
return copy_to_user(stack_ptr, string.characters(), string.length() + 1);
};
Vector<FlatPtr> argv_entries;
for (auto& argument : arguments) {
push_string_on_new_stack(argument);
if (!argv_entries.try_append(new_sp))
return ENOMEM;
}
Vector<FlatPtr> env_entries;
for (auto& variable : environment) {
push_string_on_new_stack(variable);
if (!env_entries.try_append(new_sp))
return ENOMEM;
}
for (auto& value : auxiliary_values) {
if (!value.optional_string.is_empty()) {
push_string_on_new_stack(value.optional_string);
value.auxv.a_un.a_ptr = (void*)new_sp;
}
}
for (ssize_t i = auxiliary_values.size() - 1; i >= 0; --i) {
auto& value = auxiliary_values[i];
push_aux_value_on_new_stack(value.auxv);
}
push_on_new_stack(0);
for (ssize_t i = env_entries.size() - 1; i >= 0; --i)
push_on_new_stack(env_entries[i]);
FlatPtr envp = new_sp;
push_on_new_stack(0);
for (ssize_t i = argv_entries.size() - 1; i >= 0; --i)
push_on_new_stack(argv_entries[i]);
FlatPtr argv = new_sp;
// NOTE: The stack needs to be 16-byte aligned.
new_sp -= new_sp % 16;
#if ARCH(I386)
// GCC assumes that the return address has been pushed to the stack when it enters the function,
// so we need to reserve an extra pointer's worth of bytes below this to make GCC's stack alignment
// calculations work
new_sp -= sizeof(void*);
push_on_new_stack(envp);
push_on_new_stack(argv);
push_on_new_stack(argv_entries.size());
#else
regs.rdi = argv_entries.size();
regs.rsi = argv;
regs.rdx = envp;
#endif
push_on_new_stack(0); // return address
VERIFY((new_sp + sizeof(void*)) % 16 == 0);
return new_sp;
}
struct RequiredLoadRange {
FlatPtr start { 0 };
FlatPtr end { 0 };
};
static KResultOr<RequiredLoadRange> get_required_load_range(FileDescription& program_description)
{
auto& inode = *(program_description.inode());
auto vmobject = SharedInodeVMObject::create_with_inode(inode);
size_t executable_size = inode.size();
auto region = MM.allocate_kernel_region_with_vmobject(*vmobject, page_round_up(executable_size), "ELF memory range calculation", Region::Access::Read);
if (!region) {
dbgln("Could not allocate memory for ELF");
return ENOMEM;
}
auto elf_image = ELF::Image(region->vaddr().as_ptr(), executable_size);
if (!elf_image.is_valid()) {
return EINVAL;
}
RequiredLoadRange range {};
elf_image.for_each_program_header([&range](const auto& pheader) {
if (pheader.type() != PT_LOAD)
return;
auto region_start = (FlatPtr)pheader.vaddr().as_ptr();
auto region_end = region_start + pheader.size_in_memory();
if (range.start == 0 || region_start < range.start)
range.start = region_start;
if (range.end == 0 || region_end > range.end)
range.end = region_end;
});
VERIFY(range.end > range.start);
return range;
};
static KResultOr<FlatPtr> get_load_offset(const ElfW(Ehdr) & main_program_header, FileDescription& main_program_description, FileDescription* interpreter_description)
{
constexpr FlatPtr load_range_start = 0x08000000;
constexpr FlatPtr load_range_size = 65536 * PAGE_SIZE; // 2**16 * PAGE_SIZE = 256MB
constexpr FlatPtr minimum_load_offset_randomization_size = 10 * MiB;
auto random_load_offset_in_range([](auto start, auto size) {
return page_round_down(start + get_good_random<FlatPtr>() % size);
});
if (main_program_header.e_type == ET_DYN) {
return random_load_offset_in_range(load_range_start, load_range_size);
}
if (main_program_header.e_type != ET_EXEC)
return EINVAL;
auto main_program_load_range_result = get_required_load_range(main_program_description);
if (main_program_load_range_result.is_error())
return main_program_load_range_result.error();
auto main_program_load_range = main_program_load_range_result.value();
RequiredLoadRange selected_range {};
if (interpreter_description) {
auto interpreter_load_range_result = get_required_load_range(*interpreter_description);
if (interpreter_load_range_result.is_error())
return interpreter_load_range_result.error();
auto interpreter_size_in_memory = interpreter_load_range_result.value().end - interpreter_load_range_result.value().start;
auto interpreter_load_range_end = load_range_start + load_range_size - interpreter_size_in_memory;
// No intersection
if (main_program_load_range.end < load_range_start || main_program_load_range.start > interpreter_load_range_end)
return random_load_offset_in_range(load_range_start, load_range_size);
RequiredLoadRange first_available_part = { load_range_start, main_program_load_range.start };
RequiredLoadRange second_available_part = { main_program_load_range.end, interpreter_load_range_end };
// Select larger part
if (first_available_part.end - first_available_part.start > second_available_part.end - second_available_part.start)
selected_range = first_available_part;
else
selected_range = second_available_part;
} else
selected_range = main_program_load_range;
// If main program is too big and leaves us without enough space for adequate loader randomization
if (selected_range.end - selected_range.start < minimum_load_offset_randomization_size)
return E2BIG;
return random_load_offset_in_range(selected_range.start, selected_range.end - selected_range.start);
}
enum class ShouldAllocateTls {
No,
Yes,
};
enum class ShouldAllowSyscalls {
No,
Yes,
};
static KResultOr<LoadResult> load_elf_object(NonnullOwnPtr<Space> new_space, FileDescription& object_description,
FlatPtr load_offset, ShouldAllocateTls should_allocate_tls, ShouldAllowSyscalls should_allow_syscalls)
{
auto& inode = *(object_description.inode());
auto vmobject = SharedInodeVMObject::create_with_inode(inode);
if (vmobject->writable_mappings()) {
dbgln("Refusing to execute a write-mapped program");
return ETXTBSY;
}
size_t executable_size = inode.size();
auto executable_region = MM.allocate_kernel_region_with_vmobject(*vmobject, page_round_up(executable_size), "ELF loading", Region::Access::Read);
if (!executable_region) {
dbgln("Could not allocate memory for ELF loading");
return ENOMEM;
}
auto elf_image = ELF::Image(executable_region->vaddr().as_ptr(), executable_size);
if (!elf_image.is_valid())
return ENOEXEC;
Region* master_tls_region { nullptr };
size_t master_tls_size = 0;
size_t master_tls_alignment = 0;
FlatPtr load_base_address = 0;
String elf_name = object_description.absolute_path();
VERIFY(!Processor::current().in_critical());
MemoryManager::enter_space(*new_space);
KResult ph_load_result = KSuccess;
elf_image.for_each_program_header([&](const ELF::Image::ProgramHeader& program_header) {
if (program_header.type() == PT_TLS) {
VERIFY(should_allocate_tls == ShouldAllocateTls::Yes);
VERIFY(program_header.size_in_memory());
if (!elf_image.is_within_image(program_header.raw_data(), program_header.size_in_image())) {
dbgln("Shenanigans! ELF PT_TLS header sneaks outside of executable.");
ph_load_result = ENOEXEC;
return IterationDecision::Break;
}
auto range = new_space->allocate_range({}, program_header.size_in_memory());
if (!range.has_value()) {
ph_load_result = ENOMEM;
return IterationDecision::Break;
}
auto region_or_error = new_space->allocate_region(range.value(), String::formatted("{} (master-tls)", elf_name), PROT_READ | PROT_WRITE, AllocationStrategy::Reserve);
if (region_or_error.is_error()) {
ph_load_result = region_or_error.error();
return IterationDecision::Break;
}
master_tls_region = region_or_error.value();
master_tls_size = program_header.size_in_memory();
master_tls_alignment = program_header.alignment();
if (!copy_to_user(master_tls_region->vaddr().as_ptr(), program_header.raw_data(), program_header.size_in_image())) {
ph_load_result = EFAULT;
return IterationDecision::Break;
}
return IterationDecision::Continue;
}
if (program_header.type() != PT_LOAD)
return IterationDecision::Continue;
if (program_header.is_writable()) {
// Writable section: create a copy in memory.
VERIFY(program_header.size_in_memory());
VERIFY(program_header.alignment() == PAGE_SIZE);
if (!elf_image.is_within_image(program_header.raw_data(), program_header.size_in_image())) {
dbgln("Shenanigans! Writable ELF PT_LOAD header sneaks outside of executable.");
ph_load_result = ENOEXEC;
return IterationDecision::Break;
}
int prot = 0;
if (program_header.is_readable())
prot |= PROT_READ;
if (program_header.is_writable())
prot |= PROT_WRITE;
auto region_name = String::formatted("{} (data-{}{})", elf_name, program_header.is_readable() ? "r" : "", program_header.is_writable() ? "w" : "");
auto range_base = VirtualAddress { page_round_down(program_header.vaddr().offset(load_offset).get()) };
auto range_end = VirtualAddress { page_round_up(program_header.vaddr().offset(load_offset).offset(program_header.size_in_memory()).get()) };
auto range = new_space->allocate_range(range_base, range_end.get() - range_base.get());
if (!range.has_value()) {
ph_load_result = ENOMEM;
return IterationDecision::Break;
}
auto region_or_error = new_space->allocate_region(range.value(), region_name, prot, AllocationStrategy::Reserve);
if (region_or_error.is_error()) {
ph_load_result = region_or_error.error();
return IterationDecision::Break;
}
// It's not always the case with PIE executables (and very well shouldn't be) that the
// virtual address in the program header matches the one we end up giving the process.
// In order to copy the data image correctly into memory, we need to copy the data starting at
// the right initial page offset into the pages allocated for the elf_alloc-XX section.
// FIXME: There's an opportunity to munmap, or at least mprotect, the padding space between
// the .text and .data PT_LOAD sections of the executable.
// Accessing it would definitely be a bug.
auto page_offset = program_header.vaddr();
page_offset.mask(~PAGE_MASK);
if (!copy_to_user((u8*)region_or_error.value()->vaddr().as_ptr() + page_offset.get(), program_header.raw_data(), program_header.size_in_image())) {
ph_load_result = EFAULT;
return IterationDecision::Break;
}
return IterationDecision::Continue;
}
// Non-writable section: map the executable itself in memory.
VERIFY(program_header.size_in_memory());
VERIFY(program_header.alignment() == PAGE_SIZE);
int prot = 0;
if (program_header.is_readable())
prot |= PROT_READ;
if (program_header.is_writable())
prot |= PROT_WRITE;
if (program_header.is_executable())
prot |= PROT_EXEC;
auto range = new_space->allocate_range(program_header.vaddr().offset(load_offset), program_header.size_in_memory());
if (!range.has_value()) {
ph_load_result = ENOMEM;
return IterationDecision::Break;
}
auto region_or_error = new_space->allocate_region_with_vmobject(range.value(), *vmobject, program_header.offset(), elf_name, prot, true);
if (region_or_error.is_error()) {
ph_load_result = region_or_error.error();
return IterationDecision::Break;
}
if (should_allow_syscalls == ShouldAllowSyscalls::Yes)
region_or_error.value()->set_syscall_region(true);
if (program_header.offset() == 0)
load_base_address = (FlatPtr)region_or_error.value()->vaddr().as_ptr();
return IterationDecision::Continue;
});
if (ph_load_result.is_error()) {
dbgln("do_exec: Failure loading program ({})", ph_load_result.error());
return ph_load_result;
}
if (!elf_image.entry().offset(load_offset).get()) {
dbgln("do_exec: Failure loading program, entry pointer is invalid! {})", elf_image.entry().offset(load_offset));
return ENOEXEC;
}
auto stack_range = new_space->allocate_range({}, Thread::default_userspace_stack_size);
if (!stack_range.has_value()) {
dbgln("do_exec: Failed to allocate VM range for stack");
return ENOMEM;
}
auto stack_region_or_error = new_space->allocate_region(stack_range.value(), "Stack (Main thread)", PROT_READ | PROT_WRITE, AllocationStrategy::Reserve);
if (stack_region_or_error.is_error())
return stack_region_or_error.error();
auto& stack_region = *stack_region_or_error.value();
stack_region.set_stack(true);
return LoadResult {
move(new_space),
load_base_address,
elf_image.entry().offset(load_offset).get(),
executable_size,
AK::try_make_weak_ptr(master_tls_region),
master_tls_size,
master_tls_alignment,
stack_region.make_weak_ptr()
};
}
KResultOr<LoadResult> Process::load(NonnullRefPtr<FileDescription> main_program_description,
RefPtr<FileDescription> interpreter_description, const ElfW(Ehdr) & main_program_header)
{
auto new_space = Space::create(*this, nullptr);
if (!new_space)
return ENOMEM;
ScopeGuard space_guard([&]() {
MemoryManager::enter_process_paging_scope(*this);
});
auto load_offset = get_load_offset(main_program_header, main_program_description, interpreter_description);
if (load_offset.is_error()) {
return load_offset.error();
}
if (interpreter_description.is_null()) {
auto result = load_elf_object(new_space.release_nonnull(), main_program_description, load_offset.value(), ShouldAllocateTls::Yes, ShouldAllowSyscalls::No);
if (result.is_error())
return result.error();
m_master_tls_region = result.value().tls_region;
m_master_tls_size = result.value().tls_size;
m_master_tls_alignment = result.value().tls_alignment;
return result;
}
auto interpreter_load_result = load_elf_object(new_space.release_nonnull(), *interpreter_description, load_offset.value(), ShouldAllocateTls::No, ShouldAllowSyscalls::Yes);
if (interpreter_load_result.is_error())
return interpreter_load_result.error();
// TLS allocation will be done in userspace by the loader
VERIFY(!interpreter_load_result.value().tls_region);
VERIFY(!interpreter_load_result.value().tls_alignment);
VERIFY(!interpreter_load_result.value().tls_size);
return interpreter_load_result;
}
KResult Process::do_exec(NonnullRefPtr<FileDescription> main_program_description, Vector<String> arguments, Vector<String> environment,
RefPtr<FileDescription> interpreter_description, Thread*& new_main_thread, u32& prev_flags, const ElfW(Ehdr) & main_program_header)
{
VERIFY(is_user_process());
VERIFY(!Processor::current().in_critical());
auto path = main_program_description->absolute_path();
dbgln_if(EXEC_DEBUG, "do_exec: {}", path);
// FIXME: How much stack space does process startup need?
if (!validate_stack_size(arguments, environment))
return E2BIG;
auto parts = path.split('/');
if (parts.is_empty())
return ENOENT;
auto main_program_metadata = main_program_description->metadata();
auto load_result_or_error = load(main_program_description, interpreter_description, main_program_header);
if (load_result_or_error.is_error()) {
dbgln("do_exec: Failed to load main program or interpreter for {}", path);
return load_result_or_error.error();
}
auto signal_trampoline_range = load_result_or_error.value().space->allocate_range({}, PAGE_SIZE);
if (!signal_trampoline_range.has_value()) {
dbgln("do_exec: Failed to allocate VM for signal trampoline");
return ENOMEM;
}
// We commit to the new executable at this point. There is no turning back!
// Prevent other processes from attaching to us with ptrace while we're doing this.
Locker ptrace_locker(ptrace_lock());
// Disable profiling temporarily in case it's running on this process.
auto was_profiling = m_profiling;
TemporaryChange profiling_disabler(m_profiling, false);
kill_threads_except_self();
auto& load_result = load_result_or_error.value();
bool executable_is_setid = false;
if (!(main_program_description->custody()->mount_flags() & MS_NOSUID)) {
if (main_program_metadata.is_setuid()) {
executable_is_setid = true;
ProtectedDataMutationScope scope { *this };
m_euid = main_program_metadata.uid;
m_suid = main_program_metadata.uid;
}
if (main_program_metadata.is_setgid()) {
executable_is_setid = true;
ProtectedDataMutationScope scope { *this };
m_egid = main_program_metadata.gid;
m_sgid = main_program_metadata.gid;
}
}
set_dumpable(!executable_is_setid);
{
// We must disable global profiling (especially kfree tracing) here because
// we might otherwise end up walking the stack into the process' space that
// is about to be destroyed.
TemporaryChange global_profiling_disabler(g_profiling_all_threads, false);
m_space = load_result.space.release_nonnull();
}
MemoryManager::enter_space(*m_space);
auto signal_trampoline_region = m_space->allocate_region_with_vmobject(signal_trampoline_range.value(), g_signal_trampoline_region->vmobject(), 0, "Signal trampoline", PROT_READ | PROT_EXEC, true);
if (signal_trampoline_region.is_error()) {
VERIFY_NOT_REACHED();
}
signal_trampoline_region.value()->set_syscall_region(true);
m_executable = main_program_description->custody();
m_arguments = arguments;
m_environment = environment;
m_veil_state = VeilState::None;
m_unveiled_paths.clear();
m_unveiled_paths.set_metadata({ "/", UnveilAccess::None, false });
m_coredump_metadata.clear();
auto current_thread = Thread::current();
current_thread->clear_signals();
clear_futex_queues_on_exec();
fds().change_each([&](auto& file_description_metadata) {
if (file_description_metadata.is_valid() && file_description_metadata.flags() & FD_CLOEXEC)
file_description_metadata = {};
});
int main_program_fd = -1;
if (interpreter_description) {
main_program_fd = m_fds.allocate();
VERIFY(main_program_fd >= 0);
auto seek_result = main_program_description->seek(0, SEEK_SET);
VERIFY(!seek_result.is_error());
main_program_description->set_readable(true);
m_fds[main_program_fd].set(move(main_program_description), FD_CLOEXEC);
}
new_main_thread = nullptr;
if (&current_thread->process() == this) {
new_main_thread = current_thread;
} else {
for_each_thread([&](auto& thread) {
new_main_thread = &thread;
return IterationDecision::Break;
});
}
VERIFY(new_main_thread);
auto auxv = generate_auxiliary_vector(load_result.load_base, load_result.entry_eip, uid(), euid(), gid(), egid(), path, main_program_fd);
// NOTE: We create the new stack before disabling interrupts since it will zero-fault
// and we don't want to deal with faults after this point.
auto make_stack_result = make_userspace_context_for_main_thread(new_main_thread->regs(), *load_result.stack_region.unsafe_ptr(), move(arguments), move(environment), move(auxv));
if (make_stack_result.is_error())
return make_stack_result.error();
FlatPtr new_userspace_sp = make_stack_result.value();
if (wait_for_tracer_at_next_execve()) {
// Make sure we release the ptrace lock here or the tracer will block forever.
ptrace_locker.unlock();
Thread::current()->send_urgent_signal_to_self(SIGSTOP);
}
// We enter a critical section here because we don't want to get interrupted between do_exec()
// and Processor::assume_context() or the next context switch.
// If we used an InterruptDisabler that sti()'d on exit, we might timer tick'd too soon in exec().
Processor::current().enter_critical(prev_flags);
// NOTE: Be careful to not trigger any page faults below!
m_name = parts.take_last();
new_main_thread->set_name(m_name);
{
ProtectedDataMutationScope scope { *this };
m_promises = m_execpromises;
m_has_promises = m_has_execpromises;
m_execpromises = 0;
m_has_execpromises = false;
m_signal_trampoline = signal_trampoline_region.value()->vaddr();
// FIXME: PID/TID ISSUE
m_pid = new_main_thread->tid().value();
}
auto tsr_result = new_main_thread->make_thread_specific_region({});
if (tsr_result.is_error()) {
// FIXME: We cannot fail this late. Refactor this so the allocation happens before we commit to the new executable.
VERIFY_NOT_REACHED();
}
new_main_thread->reset_fpu_state();
auto& regs = new_main_thread->m_regs;
#if ARCH(I386)
regs.cs = GDT_SELECTOR_CODE3 | 3;
regs.ds = GDT_SELECTOR_DATA3 | 3;
regs.es = GDT_SELECTOR_DATA3 | 3;
regs.ss = GDT_SELECTOR_DATA3 | 3;
regs.fs = GDT_SELECTOR_DATA3 | 3;
regs.gs = GDT_SELECTOR_TLS | 3;
regs.eip = load_result.entry_eip;
regs.esp = new_userspace_sp;
#else
regs.rip = load_result.entry_eip;
regs.rsp = new_userspace_sp;
#endif
regs.cr3 = space().page_directory().cr3();
{
TemporaryChange profiling_disabler(m_profiling, was_profiling);
PerformanceManager::add_process_exec_event(*this);
}
{
ScopedSpinLock lock(g_scheduler_lock);
new_main_thread->set_state(Thread::State::Runnable);
}
u32 lock_count_to_restore;
[[maybe_unused]] auto rc = big_lock().force_unlock_if_locked(lock_count_to_restore);
VERIFY_INTERRUPTS_DISABLED();
VERIFY(Processor::current().in_critical());
return KSuccess;
}
static Vector<ELF::AuxiliaryValue> generate_auxiliary_vector(FlatPtr load_base, FlatPtr entry_eip, uid_t uid, uid_t euid, gid_t gid, gid_t egid, String executable_path, int main_program_fd)
{
Vector<ELF::AuxiliaryValue> auxv;
// PHDR/EXECFD
// PH*
auxv.append({ ELF::AuxiliaryValue::PageSize, PAGE_SIZE });
auxv.append({ ELF::AuxiliaryValue::BaseAddress, (void*)load_base });
auxv.append({ ELF::AuxiliaryValue::Entry, (void*)entry_eip });
// NOTELF
auxv.append({ ELF::AuxiliaryValue::Uid, (long)uid });
auxv.append({ ELF::AuxiliaryValue::EUid, (long)euid });
auxv.append({ ELF::AuxiliaryValue::Gid, (long)gid });
auxv.append({ ELF::AuxiliaryValue::EGid, (long)egid });
auxv.append({ ELF::AuxiliaryValue::Platform, Processor::current().platform_string() });
// FIXME: This is platform specific
auxv.append({ ELF::AuxiliaryValue::HwCap, (long)CPUID(1).edx() });
auxv.append({ ELF::AuxiliaryValue::ClockTick, (long)TimeManagement::the().ticks_per_second() });
// FIXME: Also take into account things like extended filesystem permissions? That's what linux does...
auxv.append({ ELF::AuxiliaryValue::Secure, ((uid != euid) || (gid != egid)) ? 1 : 0 });
char random_bytes[16] {};
get_fast_random_bytes((u8*)random_bytes, sizeof(random_bytes));
auxv.append({ ELF::AuxiliaryValue::Random, String(random_bytes, sizeof(random_bytes)) });
auxv.append({ ELF::AuxiliaryValue::ExecFilename, executable_path });
auxv.append({ ELF::AuxiliaryValue::ExecFileDescriptor, main_program_fd });
auxv.append({ ELF::AuxiliaryValue::Null, 0L });
return auxv;
}
static KResultOr<Vector<String>> find_shebang_interpreter_for_executable(const char first_page[], int nread)
{
int word_start = 2;
int word_length = 0;
if (nread > 2 && first_page[0] == '#' && first_page[1] == '!') {
Vector<String> interpreter_words;
for (int i = 2; i < nread; ++i) {
if (first_page[i] == '\n') {
break;
}
if (first_page[i] != ' ') {
++word_length;
}
if (first_page[i] == ' ') {
if (word_length > 0) {
interpreter_words.append(String(&first_page[word_start], word_length));
}
word_length = 0;
word_start = i + 1;
}
}
if (word_length > 0)
interpreter_words.append(String(&first_page[word_start], word_length));
if (!interpreter_words.is_empty())
return interpreter_words;
}
return ENOEXEC;
}
KResultOr<RefPtr<FileDescription>> Process::find_elf_interpreter_for_executable(const String& path, const ElfW(Ehdr) & main_program_header, int nread, size_t file_size)
{
// Not using KResultOr here because we'll want to do the same thing in userspace in the RTLD
String interpreter_path;
if (!ELF::validate_program_headers(main_program_header, file_size, (const u8*)&main_program_header, nread, &interpreter_path)) {
dbgln("exec({}): File has invalid ELF Program headers", path);
return ENOEXEC;
}
if (!interpreter_path.is_empty()) {
dbgln_if(EXEC_DEBUG, "exec({}): Using program interpreter {}", path, interpreter_path);
auto interp_result = VFS::the().open(interpreter_path, O_EXEC, 0, current_directory());
if (interp_result.is_error()) {
dbgln("exec({}): Unable to open program interpreter {}", path, interpreter_path);
return interp_result.error();
}
auto interpreter_description = interp_result.value();
auto interp_metadata = interpreter_description->metadata();
VERIFY(interpreter_description->inode());
// Validate the program interpreter as a valid elf binary.
// If your program interpreter is a #! file or something, it's time to stop playing games :)
if (interp_metadata.size < (int)sizeof(ElfW(Ehdr)))
return ENOEXEC;
char first_page[PAGE_SIZE] = {};
auto first_page_buffer = UserOrKernelBuffer::for_kernel_buffer((u8*)&first_page);
auto nread_or_error = interpreter_description->read(first_page_buffer, sizeof(first_page));
if (nread_or_error.is_error())
return ENOEXEC;
nread = nread_or_error.value();
if (nread < (int)sizeof(ElfW(Ehdr)))
return ENOEXEC;
auto elf_header = (ElfW(Ehdr)*)first_page;
if (!ELF::validate_elf_header(*elf_header, interp_metadata.size)) {
dbgln("exec({}): Interpreter ({}) has invalid ELF header", path, interpreter_description->absolute_path());
return ENOEXEC;
}
// Not using KResultOr here because we'll want to do the same thing in userspace in the RTLD
String interpreter_interpreter_path;
if (!ELF::validate_program_headers(*elf_header, interp_metadata.size, (u8*)first_page, nread, &interpreter_interpreter_path)) {
dbgln("exec({}): Interpreter ({}) has invalid ELF Program headers", path, interpreter_description->absolute_path());
return ENOEXEC;
}
if (!interpreter_interpreter_path.is_empty()) {
dbgln("exec({}): Interpreter ({}) has its own interpreter ({})! No thank you!", path, interpreter_description->absolute_path(), interpreter_interpreter_path);
return ELOOP;
}
return interpreter_description;
}
if (main_program_header.e_type == ET_REL) {
// We can't exec an ET_REL, that's just an object file from the compiler
return ENOEXEC;
}
if (main_program_header.e_type == ET_DYN) {
// If it's ET_DYN with no PT_INTERP, then it's a dynamic executable responsible
// for its own relocation (i.e. it's /usr/lib/Loader.so)
if (path != "/usr/lib/Loader.so")
dbgln("exec({}): WARNING - Dynamic ELF executable without a PT_INTERP header, and isn't /usr/lib/Loader.so", path);
return nullptr;
}
// No interpreter, but, path refers to a valid elf image
return KResult(KSuccess);
}
KResult Process::exec(String path, Vector<String> arguments, Vector<String> environment, int recursion_depth)
{
if (recursion_depth > 2) {
dbgln("exec({}): SHENANIGANS! recursed too far trying to find #! interpreter", path);
return ELOOP;
}
// Open the file to check what kind of binary format it is
// Currently supported formats:
// - #! interpreted file
// - ELF32
// * ET_EXEC binary that just gets loaded
// * ET_DYN binary that requires a program interpreter
//
auto file_or_error = VFS::the().open(path, O_EXEC, 0, current_directory());
if (file_or_error.is_error())
return file_or_error.error();
auto description = file_or_error.release_value();
auto metadata = description->metadata();
if (!metadata.is_regular_file())
return EACCES;
// Always gonna need at least 3 bytes. these are for #!X
if (metadata.size < 3)
return ENOEXEC;
VERIFY(description->inode());
// Read the first page of the program into memory so we can validate the binfmt of it
char first_page[PAGE_SIZE];
auto first_page_buffer = UserOrKernelBuffer::for_kernel_buffer((u8*)&first_page);
auto nread_or_error = description->read(first_page_buffer, sizeof(first_page));
if (nread_or_error.is_error())
return ENOEXEC;
// 1) #! interpreted file
auto shebang_result = find_shebang_interpreter_for_executable(first_page, nread_or_error.value());
if (!shebang_result.is_error()) {
auto shebang_words = shebang_result.release_value();
auto shebang_path = shebang_words.first();
arguments[0] = move(path);
if (!arguments.try_prepend(move(shebang_words)))
return ENOMEM;
return exec(move(shebang_path), move(arguments), move(environment), ++recursion_depth);
}
// #2) ELF32 for i386
if (nread_or_error.value() < (int)sizeof(ElfW(Ehdr)))
return ENOEXEC;
auto main_program_header = (ElfW(Ehdr)*)first_page;
if (!ELF::validate_elf_header(*main_program_header, metadata.size)) {
dbgln("exec({}): File has invalid ELF header", path);
return ENOEXEC;
}
auto elf_result = find_elf_interpreter_for_executable(path, *main_program_header, nread_or_error.value(), metadata.size);
// Assume a static ELF executable by default
RefPtr<FileDescription> interpreter_description;
// We're getting either an interpreter, an error, or KSuccess (i.e. no interpreter but file checks out)
if (!elf_result.is_error()) {
// It's a dynamic ELF executable, with or without an interpreter. Do not allocate TLS
interpreter_description = elf_result.value();
} else if (elf_result.error().is_error())
return elf_result.error();
// The bulk of exec() is done by do_exec(), which ensures that all locals
// are cleaned up by the time we yield-teleport below.
Thread* new_main_thread = nullptr;
u32 prev_flags = 0;
auto result = do_exec(move(description), move(arguments), move(environment), move(interpreter_description), new_main_thread, prev_flags, *main_program_header);
if (result.is_error())
return result;
VERIFY_INTERRUPTS_DISABLED();
VERIFY(Processor::current().in_critical());
auto current_thread = Thread::current();
if (current_thread == new_main_thread) {
// We need to enter the scheduler lock before changing the state
// and it will be released after the context switch into that
// thread. We should also still be in our critical section
VERIFY(!g_scheduler_lock.own_lock());
VERIFY(Processor::current().in_critical() == 1);
g_scheduler_lock.lock();
current_thread->set_state(Thread::State::Running);
Processor::assume_context(*current_thread, prev_flags);
VERIFY_NOT_REACHED();
}
Processor::current().leave_critical(prev_flags);
return KSuccess;
}
KResultOr<FlatPtr> Process::sys$execve(Userspace<const Syscall::SC_execve_params*> user_params)
{
REQUIRE_PROMISE(exec);
// NOTE: Be extremely careful with allocating any kernel memory in exec().
// On success, the kernel stack will be lost.
Syscall::SC_execve_params params;
if (!copy_from_user(&params, user_params))
return EFAULT;
if (params.arguments.length > ARG_MAX || params.environment.length > ARG_MAX)
return E2BIG;
String path;
{
auto path_arg = get_syscall_path_argument(params.path);
if (path_arg.is_error())
return path_arg.error();
path = path_arg.value()->view();
}
auto copy_user_strings = [](const auto& list, auto& output) {
if (!list.length)
return true;
Checked<size_t> size = sizeof(*list.strings);
size *= list.length;
if (size.has_overflow())
return false;
Vector<Syscall::StringArgument, 32> strings;
if (!strings.try_resize(list.length))
return false;
if (!copy_from_user(strings.data(), list.strings, size.value()))
return false;
for (size_t i = 0; i < list.length; ++i) {
auto string = copy_string_from_user(strings[i]);
if (string.is_null())
return false;
if (!output.try_append(move(string)))
return false;
}
return true;
};
Vector<String> arguments;
if (!copy_user_strings(params.arguments, arguments))
return EFAULT;
Vector<String> environment;
if (!copy_user_strings(params.environment, environment))
return EFAULT;
auto result = exec(move(path), move(arguments), move(environment));
VERIFY(result.is_error()); // We should never continue after a successful exec!
return result.error();
}
}
Reference in a new issue View git blame Copy permalink