/* * Copyright (c) 2018-2021, Andreas Kling * * SPDX-License-Identifier: BSD-2-Clause */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include namespace Kernel { static Singleton> s_list; SpinlockProtected& Thread::all_instances() { return *s_list; } ErrorOr> Thread::try_create(NonnullLockRefPtr process) { auto kernel_stack_region = TRY(MM.allocate_kernel_region(default_kernel_stack_size, {}, Memory::Region::Access::ReadWrite, AllocationStrategy::AllocateNow)); kernel_stack_region->set_stack(true); auto block_timer = TRY(try_make_lock_ref_counted()); auto name = TRY(KString::try_create(process->name())); return adopt_nonnull_lock_ref_or_enomem(new (nothrow) Thread(move(process), move(kernel_stack_region), move(block_timer), move(name))); } Thread::Thread(NonnullLockRefPtr process, NonnullOwnPtr kernel_stack_region, NonnullLockRefPtr block_timer, NonnullOwnPtr name) : m_process(move(process)) , m_kernel_stack_region(move(kernel_stack_region)) , m_name(move(name)) , m_block_timer(move(block_timer)) { bool is_first_thread = m_process->add_thread(*this); if (is_first_thread) { // First thread gets TID == PID m_tid = m_process->pid().value(); } else { m_tid = Process::allocate_pid().value(); } // FIXME: Handle KString allocation failure. m_kernel_stack_region->set_name(MUST(KString::formatted("Kernel stack (thread {})", m_tid.value()))); Thread::all_instances().with([&](auto& list) { list.append(*this); }); if constexpr (THREAD_DEBUG) dbgln("Created new thread {}({}:{})", m_process->name(), m_process->pid().value(), m_tid.value()); reset_fpu_state(); // Only IF is set when a process boots. m_regs.set_flags(0x0202); #if ARCH(I386) if (m_process->is_kernel_process()) { m_regs.cs = GDT_SELECTOR_CODE0; m_regs.ds = GDT_SELECTOR_DATA0; m_regs.es = GDT_SELECTOR_DATA0; m_regs.fs = 0; m_regs.ss = GDT_SELECTOR_DATA0; m_regs.gs = GDT_SELECTOR_PROC; } else { m_regs.cs = GDT_SELECTOR_CODE3 | 3; m_regs.ds = GDT_SELECTOR_DATA3 | 3; m_regs.es = GDT_SELECTOR_DATA3 | 3; m_regs.fs = GDT_SELECTOR_DATA3 | 3; m_regs.ss = GDT_SELECTOR_DATA3 | 3; m_regs.gs = GDT_SELECTOR_TLS | 3; } #elif ARCH(X86_64) if (m_process->is_kernel_process()) m_regs.cs = GDT_SELECTOR_CODE0; else m_regs.cs = GDT_SELECTOR_CODE3 | 3; #elif ARCH(AARCH64) TODO_AARCH64(); #else # error Unknown architecture #endif m_regs.cr3 = m_process->address_space().with([](auto& space) { return space->page_directory().cr3(); }); m_kernel_stack_base = m_kernel_stack_region->vaddr().get(); m_kernel_stack_top = m_kernel_stack_region->vaddr().offset(default_kernel_stack_size).get() & ~(FlatPtr)0x7u; if (m_process->is_kernel_process()) { m_regs.set_sp(m_kernel_stack_top); m_regs.set_sp0(m_kernel_stack_top); } else { // Ring 3 processes get a separate stack for ring 0. // The ring 3 stack will be assigned by exec(). #if ARCH(I386) m_regs.ss0 = GDT_SELECTOR_DATA0; #endif m_regs.set_sp0(m_kernel_stack_top); } // We need to add another reference if we could successfully create // all the resources needed for this thread. The reason for this is that // we don't want to delete this thread after dropping the reference, // it may still be running or scheduled to be run. // The finalizer is responsible for dropping this reference once this // thread is ready to be cleaned up. ref(); } Thread::~Thread() { VERIFY(!m_process_thread_list_node.is_in_list()); // We shouldn't be queued VERIFY(m_runnable_priority < 0); } Thread::BlockResult Thread::block_impl(BlockTimeout const& timeout, Blocker& blocker) { VERIFY(!Processor::current_in_irq()); VERIFY(this == Thread::current()); ScopedCritical critical; SpinlockLocker scheduler_lock(g_scheduler_lock); SpinlockLocker block_lock(m_block_lock); // We need to hold m_block_lock so that nobody can unblock a blocker as soon // as it is constructed and registered elsewhere ScopeGuard finalize_guard([&] { blocker.finalize(); }); if (!blocker.setup_blocker()) { blocker.will_unblock_immediately_without_blocking(Blocker::UnblockImmediatelyReason::UnblockConditionAlreadyMet); return BlockResult::NotBlocked; } // Relaxed semantics are fine for timeout_unblocked because we // synchronize on the spin locks already. Atomic timeout_unblocked(false); bool timer_was_added = false; switch (state()) { case Thread::State::Stopped: // It's possible that we were requested to be stopped! break; case Thread::State::Running: VERIFY(m_blocker == nullptr); break; default: VERIFY_NOT_REACHED(); } m_blocker = &blocker; if (auto& block_timeout = blocker.override_timeout(timeout); !block_timeout.is_infinite()) { // Process::kill_all_threads may be called at any time, which will mark all // threads to die. In that case timer_was_added = TimerQueue::the().add_timer_without_id(*m_block_timer, block_timeout.clock_id(), block_timeout.absolute_time(), [&]() { VERIFY(!Processor::current_in_irq()); VERIFY(!g_scheduler_lock.is_locked_by_current_processor()); VERIFY(!m_block_lock.is_locked_by_current_processor()); // NOTE: this may execute on the same or any other processor! SpinlockLocker scheduler_lock(g_scheduler_lock); SpinlockLocker block_lock(m_block_lock); if (m_blocker && !timeout_unblocked.exchange(true)) unblock(); }); if (!timer_was_added) { // Timeout is already in the past blocker.will_unblock_immediately_without_blocking(Blocker::UnblockImmediatelyReason::TimeoutInThePast); m_blocker = nullptr; return BlockResult::InterruptedByTimeout; } } blocker.begin_blocking({}); set_state(Thread::State::Blocked); block_lock.unlock(); scheduler_lock.unlock(); dbgln_if(THREAD_DEBUG, "Thread {} blocking on {} ({}) -->", *this, &blocker, blocker.state_string()); bool did_timeout = false; u32 lock_count_to_restore = 0; auto previous_locked = unlock_process_if_locked(lock_count_to_restore); for (;;) { // Yield to the scheduler, and wait for us to resume unblocked. VERIFY(!g_scheduler_lock.is_locked_by_current_processor()); VERIFY(Processor::in_critical()); yield_without_releasing_big_lock(); VERIFY(Processor::in_critical()); SpinlockLocker block_lock2(m_block_lock); if (m_blocker && !m_blocker->can_be_interrupted() && !m_should_die) { block_lock2.unlock(); dbgln("Thread should not be unblocking, current state: {}", state_string()); set_state(Thread::State::Blocked); continue; } // Prevent the timeout from unblocking this thread if it happens to // be in the process of firing already did_timeout |= timeout_unblocked.exchange(true); if (m_blocker) { // Remove ourselves... VERIFY(m_blocker == &blocker); m_blocker = nullptr; } dbgln_if(THREAD_DEBUG, "<-- Thread {} unblocked from {} ({})", *this, &blocker, blocker.state_string()); break; } // Notify the blocker that we are no longer blocking. It may need // to clean up now while we're still holding m_lock auto result = blocker.end_blocking({}, did_timeout); // calls was_unblocked internally if (timer_was_added && !did_timeout) { // Cancel the timer while not holding any locks. This allows // the timer function to complete before we remove it // (e.g. if it's on another processor) TimerQueue::the().cancel_timer(*m_block_timer); } if (previous_locked != LockMode::Unlocked) { // NOTE: This may trigger another call to Thread::block(). relock_process(previous_locked, lock_count_to_restore); } return result; } void Thread::block(Kernel::Mutex& lock, SpinlockLocker& lock_lock, u32 lock_count) { VERIFY(!Processor::current_in_irq()); VERIFY(this == Thread::current()); ScopedCritical critical; SpinlockLocker scheduler_lock(g_scheduler_lock); SpinlockLocker block_lock(m_block_lock); switch (state()) { case Thread::State::Stopped: // It's possible that we were requested to be stopped! break; case Thread::State::Running: VERIFY(m_blocker == nullptr); break; default: dbgln("Error: Attempting to block with invalid thread state - {}", state_string()); VERIFY_NOT_REACHED(); } // If we're blocking on the big-lock we may actually be in the process // of unblocking from another lock. If that's the case m_blocking_mutex // is already set auto& big_lock = process().big_lock(); VERIFY((&lock == &big_lock && m_blocking_mutex != &big_lock) || !m_blocking_mutex); auto* previous_blocking_mutex = m_blocking_mutex; m_blocking_mutex = &lock; m_lock_requested_count = lock_count; set_state(Thread::State::Blocked); block_lock.unlock(); scheduler_lock.unlock(); lock_lock.unlock(); dbgln_if(THREAD_DEBUG, "Thread {} blocking on Mutex {}", *this, &lock); for (;;) { // Yield to the scheduler, and wait for us to resume unblocked. VERIFY(!g_scheduler_lock.is_locked_by_current_processor()); VERIFY(Processor::in_critical()); if (&lock != &big_lock && big_lock.is_exclusively_locked_by_current_thread()) { // We're locking another lock and already hold the big lock... // We need to release the big lock yield_and_release_relock_big_lock(); } else { // By the time we've reached this another thread might have // marked us as holding the big lock, so this call must not // verify that we're not holding it. yield_without_releasing_big_lock(VerifyLockNotHeld::No); } VERIFY(Processor::in_critical()); SpinlockLocker block_lock2(m_block_lock); VERIFY(!m_blocking_mutex); m_blocking_mutex = previous_blocking_mutex; break; } lock_lock.lock(); } u32 Thread::unblock_from_mutex(Kernel::Mutex& mutex) { SpinlockLocker scheduler_lock(g_scheduler_lock); SpinlockLocker block_lock(m_block_lock); VERIFY(!Processor::current_in_irq()); VERIFY(m_blocking_mutex == &mutex); dbgln_if(THREAD_DEBUG, "Thread {} unblocked from Mutex {}", *this, &mutex); auto requested_count = m_lock_requested_count; m_blocking_mutex = nullptr; if (Thread::current() == this) { set_state(Thread::State::Running); return requested_count; } VERIFY(m_state != Thread::State::Runnable && m_state != Thread::State::Running); set_state(Thread::State::Runnable); return requested_count; } void Thread::unblock_from_blocker(Blocker& blocker) { auto do_unblock = [&]() { SpinlockLocker scheduler_lock(g_scheduler_lock); SpinlockLocker block_lock(m_block_lock); if (m_blocker != &blocker) return; if (!should_be_stopped() && !is_stopped()) unblock(); }; if (Processor::current_in_irq() != 0) { Processor::deferred_call_queue([do_unblock = move(do_unblock), self = try_make_weak_ptr().release_value_but_fixme_should_propagate_errors()]() { if (auto this_thread = self.strong_ref()) do_unblock(); }); } else { do_unblock(); } } void Thread::unblock(u8 signal) { VERIFY(!Processor::current_in_irq()); VERIFY(g_scheduler_lock.is_locked_by_current_processor()); VERIFY(m_block_lock.is_locked_by_current_processor()); if (m_state != Thread::State::Blocked) return; if (m_blocking_mutex) return; VERIFY(m_blocker); if (signal != 0) { if (is_handling_page_fault()) { // Don't let signals unblock threads that are blocked inside a page fault handler. // This prevents threads from EINTR'ing the inode read in an inode page fault. // FIXME: There's probably a better way to solve this. return; } if (!m_blocker->can_be_interrupted() && !m_should_die) return; m_blocker->set_interrupted_by_signal(signal); } m_blocker = nullptr; if (Thread::current() == this) { set_state(Thread::State::Running); return; } VERIFY(m_state != Thread::State::Runnable && m_state != Thread::State::Running); set_state(Thread::State::Runnable); } void Thread::set_should_die() { if (m_should_die) { dbgln("{} Should already die", *this); return; } ScopedCritical critical; // Remember that we should die instead of returning to // the userspace. SpinlockLocker lock(g_scheduler_lock); m_should_die = true; // NOTE: Even the current thread can technically be in "Stopped" // state! This is the case when another thread sent a SIGSTOP to // it while it was running and it calls e.g. exit() before // the scheduler gets involved again. if (is_stopped()) { // If we were stopped, we need to briefly resume so that // the kernel stacks can clean up. We won't ever return back // to user mode, though VERIFY(!process().is_stopped()); resume_from_stopped(); } if (is_blocked()) { SpinlockLocker block_lock(m_block_lock); if (m_blocker) { // We're blocked in the kernel. m_blocker->set_interrupted_by_death(); unblock(); } } } void Thread::die_if_needed() { VERIFY(Thread::current() == this); if (!m_should_die) return; u32 unlock_count; [[maybe_unused]] auto rc = unlock_process_if_locked(unlock_count); dbgln_if(THREAD_DEBUG, "Thread {} is dying", *this); { SpinlockLocker lock(g_scheduler_lock); // It's possible that we don't reach the code after this block if the // scheduler is invoked and FinalizerTask cleans up this thread, however // that doesn't matter because we're trying to invoke the scheduler anyway set_state(Thread::State::Dying); } ScopedCritical critical; // Flag a context switch. Because we're in a critical section, // Scheduler::yield will actually only mark a pending context switch // Simply leaving the critical section would not necessarily trigger // a switch. Scheduler::yield(); // Now leave the critical section so that we can also trigger the // actual context switch Processor::clear_critical(); dbgln("die_if_needed returned from clear_critical!!! in irq: {}", Processor::current_in_irq()); // We should never get here, but the scoped scheduler lock // will be released by Scheduler::context_switch again VERIFY_NOT_REACHED(); } void Thread::exit(void* exit_value) { VERIFY(Thread::current() == this); m_join_blocker_set.thread_did_exit(exit_value); set_should_die(); u32 unlock_count; [[maybe_unused]] auto rc = unlock_process_if_locked(unlock_count); if (m_thread_specific_range.has_value()) { process().address_space().with([&](auto& space) { auto* region = space->find_region_from_range(m_thread_specific_range.value()); space->deallocate_region(*region); }); } #ifdef ENABLE_KERNEL_COVERAGE_COLLECTION KCOVDevice::free_thread(); #endif die_if_needed(); } void Thread::yield_without_releasing_big_lock(VerifyLockNotHeld verify_lock_not_held) { VERIFY(!g_scheduler_lock.is_locked_by_current_processor()); VERIFY(verify_lock_not_held == VerifyLockNotHeld::No || !process().big_lock().is_exclusively_locked_by_current_thread()); // Disable interrupts here. This ensures we don't accidentally switch contexts twice InterruptDisabler disable; Scheduler::yield(); // flag a switch u32 prev_critical = Processor::clear_critical(); // NOTE: We may be on a different CPU now! Processor::restore_critical(prev_critical); } void Thread::yield_and_release_relock_big_lock() { VERIFY(!g_scheduler_lock.is_locked_by_current_processor()); // Disable interrupts here. This ensures we don't accidentally switch contexts twice InterruptDisabler disable; Scheduler::yield(); // flag a switch u32 lock_count_to_restore = 0; auto previous_locked = unlock_process_if_locked(lock_count_to_restore); // NOTE: Even though we call Scheduler::yield here, unless we happen // to be outside of a critical section, the yield will be postponed // until leaving it in relock_process. relock_process(previous_locked, lock_count_to_restore); } LockMode Thread::unlock_process_if_locked(u32& lock_count_to_restore) { return process().big_lock().force_unlock_exclusive_if_locked(lock_count_to_restore); } void Thread::relock_process(LockMode previous_locked, u32 lock_count_to_restore) { // Clearing the critical section may trigger the context switch // flagged by calling Scheduler::yield above. // We have to do it this way because we intentionally // leave the critical section here to be able to switch contexts. u32 prev_critical = Processor::clear_critical(); // CONTEXT SWITCH HAPPENS HERE! // NOTE: We may be on a different CPU now! Processor::restore_critical(prev_critical); if (previous_locked != LockMode::Unlocked) { // We've unblocked, relock the process if needed and carry on. process().big_lock().restore_exclusive_lock(lock_count_to_restore); } } // NOLINTNEXTLINE(readability-make-member-function-const) False positive; We call block which is not const auto Thread::sleep(clockid_t clock_id, Time const& duration, Time* remaining_time) -> BlockResult { VERIFY(state() == Thread::State::Running); return Thread::current()->block({}, Thread::BlockTimeout(false, &duration, nullptr, clock_id), remaining_time); } // NOLINTNEXTLINE(readability-make-member-function-const) False positive; We call block which is not const auto Thread::sleep_until(clockid_t clock_id, Time const& deadline) -> BlockResult { VERIFY(state() == Thread::State::Running); return Thread::current()->block({}, Thread::BlockTimeout(true, &deadline, nullptr, clock_id)); } StringView Thread::state_string() const { switch (state()) { case Thread::State::Invalid: return "Invalid"sv; case Thread::State::Runnable: return "Runnable"sv; case Thread::State::Running: return "Running"sv; case Thread::State::Dying: return "Dying"sv; case Thread::State::Dead: return "Dead"sv; case Thread::State::Stopped: return "Stopped"sv; case Thread::State::Blocked: { SpinlockLocker block_lock(m_block_lock); if (m_blocking_mutex) return "Mutex"sv; if (m_blocker) return m_blocker->state_string(); VERIFY_NOT_REACHED(); } } PANIC("Thread::state_string(): Invalid state: {}", (int)state()); } void Thread::finalize() { VERIFY(Thread::current() == g_finalizer); VERIFY(Thread::current() != this); #if LOCK_DEBUG VERIFY(!m_lock.is_locked_by_current_processor()); if (lock_count() > 0) { dbgln("Thread {} leaking {} Locks!", *this, lock_count()); SpinlockLocker list_lock(m_holding_locks_lock); for (auto& info : m_holding_locks_list) { auto const& location = info.lock_location; dbgln(" - Mutex: \"{}\" @ {} locked in function \"{}\" at \"{}:{}\" with a count of: {}", info.lock->name(), info.lock, location.function_name(), location.filename(), location.line_number(), info.count); } VERIFY_NOT_REACHED(); } #endif { SpinlockLocker lock(g_scheduler_lock); dbgln_if(THREAD_DEBUG, "Finalizing thread {}", *this); set_state(Thread::State::Dead); m_join_blocker_set.thread_finalizing(); } if (m_dump_backtrace_on_finalization) { auto trace_or_error = backtrace(); if (!trace_or_error.is_error()) { auto trace = trace_or_error.release_value(); dbgln("Backtrace:"); kernelputstr(trace->characters(), trace->length()); } } drop_thread_count(); } void Thread::drop_thread_count() { bool is_last = process().remove_thread(*this); if (is_last) process().finalize(); } void Thread::finalize_dying_threads() { VERIFY(Thread::current() == g_finalizer); Vector dying_threads; { SpinlockLocker lock(g_scheduler_lock); for_each_in_state(Thread::State::Dying, [&](Thread& thread) { if (!thread.is_finalizable()) return; auto result = dying_threads.try_append(&thread); // We ignore allocation failures above the first 32 guaranteed thread slots, and // just flag our future-selves to finalize these threads at a later point if (result.is_error()) g_finalizer_has_work.store(true, AK::MemoryOrder::memory_order_release); }); } for (auto* thread : dying_threads) { LockRefPtr process = thread->process(); dbgln_if(PROCESS_DEBUG, "Before finalization, {} has {} refs and its process has {}", *thread, thread->ref_count(), thread->process().ref_count()); thread->finalize(); dbgln_if(PROCESS_DEBUG, "After finalization, {} has {} refs and its process has {}", *thread, thread->ref_count(), thread->process().ref_count()); // This thread will never execute again, drop the running reference // NOTE: This may not necessarily drop the last reference if anything // else is still holding onto this thread! thread->unref(); } } void Thread::update_time_scheduled(u64 current_scheduler_time, bool is_kernel, bool no_longer_running) { if (m_last_time_scheduled.has_value()) { u64 delta; if (current_scheduler_time >= m_last_time_scheduled.value()) delta = current_scheduler_time - m_last_time_scheduled.value(); else delta = m_last_time_scheduled.value() - current_scheduler_time; // the unlikely event that the clock wrapped if (delta != 0) { // Add it to the global total *before* updating the thread's value! Scheduler::add_time_scheduled(delta, is_kernel); auto& total_time = is_kernel ? m_total_time_scheduled_kernel : m_total_time_scheduled_user; total_time.fetch_add(delta, AK::memory_order_relaxed); } } if (no_longer_running) m_last_time_scheduled = {}; else m_last_time_scheduled = current_scheduler_time; } bool Thread::tick() { if (previous_mode() == PreviousMode::KernelMode) { ++m_process->m_ticks_in_kernel; ++m_ticks_in_kernel; } else { ++m_process->m_ticks_in_user; ++m_ticks_in_user; } --m_ticks_left; return m_ticks_left != 0; } void Thread::check_dispatch_pending_signal() { auto result = DispatchSignalResult::Continue; { SpinlockLocker scheduler_lock(g_scheduler_lock); if (pending_signals_for_state() != 0) { result = dispatch_one_pending_signal(); } } if (result == DispatchSignalResult::Yield) { yield_without_releasing_big_lock(); } } u32 Thread::pending_signals() const { SpinlockLocker lock(g_scheduler_lock); return pending_signals_for_state(); } u32 Thread::pending_signals_for_state() const { VERIFY(g_scheduler_lock.is_locked_by_current_processor()); constexpr u32 stopped_signal_mask = (1 << (SIGCONT - 1)) | (1 << (SIGKILL - 1)) | (1 << (SIGTRAP - 1)); if (is_handling_page_fault()) return 0; return m_state != State::Stopped ? m_pending_signals : m_pending_signals & stopped_signal_mask; } void Thread::send_signal(u8 signal, [[maybe_unused]] Process* sender) { VERIFY(signal < NSIG); VERIFY(process().is_user_process()); SpinlockLocker scheduler_lock(g_scheduler_lock); // FIXME: Figure out what to do for masked signals. Should we also ignore them here? if (should_ignore_signal(signal)) { dbgln_if(SIGNAL_DEBUG, "Signal {} was ignored by {}", signal, process()); return; } if constexpr (SIGNAL_DEBUG) { if (sender) dbgln("Signal: {} sent {} to {}", *sender, signal, process()); else dbgln("Signal: Kernel send {} to {}", signal, process()); } m_pending_signals |= 1 << (signal - 1); m_signal_senders[signal] = sender ? sender->pid() : pid(); m_have_any_unmasked_pending_signals.store((pending_signals_for_state() & ~m_signal_mask) != 0, AK::memory_order_release); m_signal_blocker_set.unblock_all_blockers_whose_conditions_are_met(); if (!has_unmasked_pending_signals()) return; if (m_state == Thread::State::Stopped) { if (pending_signals_for_state() != 0) { dbgln_if(SIGNAL_DEBUG, "Signal: Resuming stopped {} to deliver signal {}", *this, signal); resume_from_stopped(); } } else { SpinlockLocker block_lock(m_block_lock); dbgln_if(SIGNAL_DEBUG, "Signal: Unblocking {} to deliver signal {}", *this, signal); unblock(signal); } } u32 Thread::update_signal_mask(u32 signal_mask) { SpinlockLocker lock(g_scheduler_lock); auto previous_signal_mask = m_signal_mask; m_signal_mask = signal_mask; m_have_any_unmasked_pending_signals.store((pending_signals_for_state() & ~m_signal_mask) != 0, AK::memory_order_release); return previous_signal_mask; } u32 Thread::signal_mask() const { SpinlockLocker lock(g_scheduler_lock); return m_signal_mask; } u32 Thread::signal_mask_block(sigset_t signal_set, bool block) { SpinlockLocker lock(g_scheduler_lock); auto previous_signal_mask = m_signal_mask; if (block) m_signal_mask |= signal_set; else m_signal_mask &= ~signal_set; m_have_any_unmasked_pending_signals.store((pending_signals_for_state() & ~m_signal_mask) != 0, AK::memory_order_release); return previous_signal_mask; } void Thread::reset_signals_for_exec() { SpinlockLocker lock(g_scheduler_lock); // The signal mask is preserved across execve(2). // The pending signal set is preserved across an execve(2). m_have_any_unmasked_pending_signals.store(false, AK::memory_order_release); m_signal_action_masks.fill({}); // A successful call to execve(2) removes any existing alternate signal stack m_alternative_signal_stack = 0; m_alternative_signal_stack_size = 0; } // Certain exceptions, such as SIGSEGV and SIGILL, put a // thread into a state where the signal handler must be // invoked immediately, otherwise it will continue to fault. // This function should be used in an exception handler to // ensure that when the thread resumes, it's executing in // the appropriate signal handler. void Thread::send_urgent_signal_to_self(u8 signal) { VERIFY(Thread::current() == this); DispatchSignalResult result; { SpinlockLocker lock(g_scheduler_lock); result = dispatch_signal(signal); } if (result == DispatchSignalResult::Terminate) { Thread::current()->die_if_needed(); VERIFY_NOT_REACHED(); // dispatch_signal will request termination of the thread, so the above call should never return } if (result == DispatchSignalResult::Yield) yield_and_release_relock_big_lock(); } DispatchSignalResult Thread::dispatch_one_pending_signal() { VERIFY(g_scheduler_lock.is_locked_by_current_processor()); u32 signal_candidates = pending_signals_for_state() & ~m_signal_mask; if (signal_candidates == 0) return DispatchSignalResult::Continue; u8 signal = 1; for (; signal < NSIG; ++signal) { if ((signal_candidates & (1 << (signal - 1))) != 0) { break; } } return dispatch_signal(signal); } DispatchSignalResult Thread::try_dispatch_one_pending_signal(u8 signal) { VERIFY(signal != 0); SpinlockLocker scheduler_lock(g_scheduler_lock); u32 signal_candidates = pending_signals_for_state() & ~m_signal_mask; if ((signal_candidates & (1 << (signal - 1))) == 0) return DispatchSignalResult::Continue; return dispatch_signal(signal); } enum class DefaultSignalAction { Terminate, Ignore, DumpCore, Stop, Continue, }; static DefaultSignalAction default_signal_action(u8 signal) { VERIFY(signal && signal < NSIG); switch (signal) { case SIGHUP: case SIGINT: case SIGKILL: case SIGPIPE: case SIGALRM: case SIGUSR1: case SIGUSR2: case SIGVTALRM: case SIGSTKFLT: case SIGIO: case SIGPROF: case SIGTERM: case SIGCANCEL: return DefaultSignalAction::Terminate; case SIGCHLD: case SIGURG: case SIGWINCH: case SIGINFO: return DefaultSignalAction::Ignore; case SIGQUIT: case SIGILL: case SIGTRAP: case SIGABRT: case SIGBUS: case SIGFPE: case SIGSEGV: case SIGXCPU: case SIGXFSZ: case SIGSYS: return DefaultSignalAction::DumpCore; case SIGCONT: return DefaultSignalAction::Continue; case SIGSTOP: case SIGTSTP: case SIGTTIN: case SIGTTOU: return DefaultSignalAction::Stop; default: VERIFY_NOT_REACHED(); } } bool Thread::should_ignore_signal(u8 signal) const { VERIFY(signal < NSIG); auto const& action = m_process->m_signal_action_data[signal]; if (action.handler_or_sigaction.is_null()) return default_signal_action(signal) == DefaultSignalAction::Ignore; return ((sighandler_t)action.handler_or_sigaction.get() == SIG_IGN); } bool Thread::has_signal_handler(u8 signal) const { VERIFY(signal < NSIG); auto const& action = m_process->m_signal_action_data[signal]; return !action.handler_or_sigaction.is_null(); } bool Thread::is_signal_masked(u8 signal) const { VERIFY(signal < NSIG); return (1 << (signal - 1)) & m_signal_mask; } bool Thread::has_alternative_signal_stack() const { return m_alternative_signal_stack_size != 0; } bool Thread::is_in_alternative_signal_stack() const { auto sp = get_register_dump_from_stack().userspace_sp(); return sp >= m_alternative_signal_stack && sp < m_alternative_signal_stack + m_alternative_signal_stack_size; } static ErrorOr push_value_on_user_stack(FlatPtr& stack, FlatPtr data) { stack -= sizeof(FlatPtr); return copy_to_user((FlatPtr*)stack, &data); } template static ErrorOr copy_value_on_user_stack(FlatPtr& stack, T const& data) { stack -= sizeof(data); return copy_to_user((RemoveCVReference*)stack, &data); } void Thread::resume_from_stopped() { VERIFY(is_stopped()); VERIFY(m_stop_state != State::Invalid); VERIFY(g_scheduler_lock.is_locked_by_current_processor()); if (m_stop_state == Thread::State::Blocked) { SpinlockLocker block_lock(m_block_lock); if (m_blocker || m_blocking_mutex) { // Hasn't been unblocked yet set_state(Thread::State::Blocked, 0); } else { // Was unblocked while stopped set_state(Thread::State::Runnable); } } else { set_state(m_stop_state, 0); } } DispatchSignalResult Thread::dispatch_signal(u8 signal) { VERIFY_INTERRUPTS_DISABLED(); VERIFY(g_scheduler_lock.is_locked_by_current_processor()); VERIFY(signal > 0 && signal <= NSIG); VERIFY(process().is_user_process()); VERIFY(this == Thread::current()); dbgln_if(SIGNAL_DEBUG, "Dispatch signal {} to {}, state: {}", signal, *this, state_string()); if (m_state == Thread::State::Invalid || !is_initialized()) { // Thread has barely been created, we need to wait until it is // at least in Runnable state and is_initialized() returns true, // which indicates that it is fully set up an we actually have // a register state on the stack that we can modify return DispatchSignalResult::Deferred; } auto& action = m_process->m_signal_action_data[signal]; auto sender_pid = m_signal_senders[signal]; auto sender = Process::from_pid(sender_pid); if (!current_trap() && !action.handler_or_sigaction.is_null()) { // We're trying dispatch a handled signal to a user process that was scheduled // after a yielding/blocking kernel thread, we don't have a register capture of // the thread, so just defer processing the signal to later. return DispatchSignalResult::Deferred; } // Mark this signal as handled. m_pending_signals &= ~(1 << (signal - 1)); m_have_any_unmasked_pending_signals.store((m_pending_signals & ~m_signal_mask) != 0, AK::memory_order_release); auto& process = this->process(); auto* tracer = process.tracer(); if (signal == SIGSTOP || (tracer && default_signal_action(signal) == DefaultSignalAction::DumpCore)) { dbgln_if(SIGNAL_DEBUG, "Signal {} stopping this thread", signal); set_state(Thread::State::Stopped, signal); return DispatchSignalResult::Yield; } if (signal == SIGCONT) { dbgln("signal: SIGCONT resuming {}", *this); } else { if (tracer) { // when a thread is traced, it should be stopped whenever it receives a signal // the tracer is notified of this by using waitpid() // only "pending signals" from the tracer are sent to the tracee if (!tracer->has_pending_signal(signal)) { dbgln("signal: {} stopping {} for tracer", signal, *this); set_state(Thread::State::Stopped, signal); return DispatchSignalResult::Yield; } tracer->unset_signal(signal); } } auto handler_vaddr = action.handler_or_sigaction; if (handler_vaddr.is_null()) { switch (default_signal_action(signal)) { case DefaultSignalAction::Stop: set_state(Thread::State::Stopped, signal); return DispatchSignalResult::Yield; case DefaultSignalAction::DumpCore: process.set_should_generate_coredump(true); process.for_each_thread([](auto& thread) { thread.set_dump_backtrace_on_finalization(); }); [[fallthrough]]; case DefaultSignalAction::Terminate: m_process->terminate_due_to_signal(signal); return DispatchSignalResult::Terminate; case DefaultSignalAction::Ignore: VERIFY_NOT_REACHED(); case DefaultSignalAction::Continue: return DispatchSignalResult::Continue; } VERIFY_NOT_REACHED(); } if ((sighandler_t)handler_vaddr.as_ptr() == SIG_IGN) { dbgln_if(SIGNAL_DEBUG, "Ignored signal {}", signal); return DispatchSignalResult::Continue; } ScopedAddressSpaceSwitcher switcher(m_process); m_currently_handled_signal = signal; u32 old_signal_mask = m_signal_mask; u32 new_signal_mask = m_signal_action_masks[signal].value_or(action.mask); if ((action.flags & SA_NODEFER) == SA_NODEFER) new_signal_mask &= ~(1 << (signal - 1)); else new_signal_mask |= 1 << (signal - 1); m_signal_mask |= new_signal_mask; m_have_any_unmasked_pending_signals.store((m_pending_signals & ~m_signal_mask) != 0, AK::memory_order_release); bool use_alternative_stack = ((action.flags & SA_ONSTACK) != 0) && has_alternative_signal_stack() && !is_in_alternative_signal_stack(); auto setup_stack = [&](RegisterState& state) -> ErrorOr { FlatPtr stack; if (use_alternative_stack) stack = m_alternative_signal_stack + m_alternative_signal_stack_size; else stack = state.userspace_sp(); dbgln_if(SIGNAL_DEBUG, "Setting up user stack to return to IP {:p}, SP {:p}", state.ip(), state.userspace_sp()); __ucontext ucontext { .uc_link = nullptr, .uc_sigmask = old_signal_mask, .uc_stack = { .ss_sp = bit_cast(stack), .ss_flags = action.flags & SA_ONSTACK, .ss_size = use_alternative_stack ? m_alternative_signal_stack_size : 0, }, .uc_mcontext = {}, }; copy_kernel_registers_into_ptrace_registers(static_cast(ucontext.uc_mcontext), state); auto fill_signal_info_for_signal = [&](siginfo& signal_info) { if (signal == SIGCHLD) { if (!sender) { signal_info.si_code = CLD_EXITED; return; } auto const* thread = sender->thread_list().with([](auto& list) { return list.is_empty() ? nullptr : list.first(); }); if (!thread) { signal_info.si_code = CLD_EXITED; return; } switch (thread->m_state) { case State::Dead: if (sender->should_generate_coredump() && sender->is_dumpable()) { signal_info.si_code = CLD_DUMPED; signal_info.si_status = sender->termination_signal(); return; } [[fallthrough]]; case State::Dying: if (sender->termination_signal() == 0) { signal_info.si_code = CLD_EXITED; signal_info.si_status = sender->termination_status(); return; } signal_info.si_code = CLD_KILLED; signal_info.si_status = sender->termination_signal(); return; case State::Runnable: case State::Running: case State::Blocked: signal_info.si_code = CLD_CONTINUED; return; case State::Stopped: signal_info.si_code = CLD_STOPPED; return; case State::Invalid: // Something is wrong, but we're just an observer. break; } } signal_info.si_code = SI_NOINFO; }; siginfo signal_info { .si_signo = signal, // Filled in below by fill_signal_info_for_signal. .si_code = 0, // Set for SI_TIMER, we don't have the data here. .si_errno = 0, .si_pid = sender_pid.value(), .si_uid = sender ? sender->credentials()->uid().value() : 0, // Set for SIGILL, SIGFPE, SIGSEGV and SIGBUS // FIXME: We don't generate these signals in a way that can be handled. .si_addr = 0, // Set for SIGCHLD. .si_status = 0, // Set for SIGPOLL, we don't have SIGPOLL. .si_band = 0, // Set for SI_QUEUE, SI_TIMER, SI_ASYNCIO and SI_MESGQ // We do not generate any of these. .si_value = { .sival_int = 0, }, }; if (action.flags & SA_SIGINFO) fill_signal_info_for_signal(signal_info); #if ARCH(I386) constexpr static FlatPtr thread_red_zone_size = 0; #elif ARCH(X86_64) constexpr static FlatPtr thread_red_zone_size = 128; #elif ARCH(AARCH64) constexpr static FlatPtr thread_red_zone_size = 0; // FIXME TODO_AARCH64(); #else # error Unknown architecture in dispatch_signal #endif // Align the stack to 16 bytes. // Note that we push some elements on to the stack before the return address, // so we need to account for this here. constexpr static FlatPtr elements_pushed_on_stack_before_handler_address = 1; // one slot for a saved register FlatPtr const extra_bytes_pushed_on_stack_before_handler_address = sizeof(ucontext) + sizeof(signal_info); FlatPtr stack_alignment = (stack - elements_pushed_on_stack_before_handler_address * sizeof(FlatPtr) + extra_bytes_pushed_on_stack_before_handler_address) % 16; // Also note that we have to skip the thread red-zone (if needed), so do that here. stack -= thread_red_zone_size + stack_alignment; auto start_of_stack = stack; TRY(push_value_on_user_stack(stack, 0)); // syscall return value slot TRY(copy_value_on_user_stack(stack, ucontext)); auto pointer_to_ucontext = stack; TRY(copy_value_on_user_stack(stack, signal_info)); auto pointer_to_signal_info = stack; // Make sure we actually pushed as many elements as we claimed to have pushed. if (start_of_stack - stack != elements_pushed_on_stack_before_handler_address * sizeof(FlatPtr) + extra_bytes_pushed_on_stack_before_handler_address) { PANIC("Stack in invalid state after signal trampoline, expected {:x} but got {:x}", start_of_stack - elements_pushed_on_stack_before_handler_address * sizeof(FlatPtr) - extra_bytes_pushed_on_stack_before_handler_address, stack); } VERIFY(stack % 16 == 0); #if ARCH(I386) || ARCH(X86_64) // Save the FPU/SSE state TRY(copy_value_on_user_stack(stack, fpu_state())); #endif #if ARCH(I386) // Leave one empty slot to align the stack for a handler call. TRY(push_value_on_user_stack(stack, 0)); #endif TRY(push_value_on_user_stack(stack, pointer_to_ucontext)); TRY(push_value_on_user_stack(stack, pointer_to_signal_info)); TRY(push_value_on_user_stack(stack, signal)); #if ARCH(I386) VERIFY(stack % 16 == 0); #endif TRY(push_value_on_user_stack(stack, handler_vaddr.get())); // We write back the adjusted stack value into the register state. // We have to do this because we can't just pass around a reference to a packed field, as it's UB. state.set_userspace_sp(stack); return {}; }; // We now place the thread state on the userspace stack. // Note that we use a RegisterState. // Conversely, when the thread isn't blocking the RegisterState may not be // valid (fork, exec etc) but the tss will, so we use that instead. auto& regs = get_register_dump_from_stack(); auto result = setup_stack(regs); if (result.is_error()) { dbgln("Invalid stack pointer: {}", regs.userspace_sp()); process.set_should_generate_coredump(true); process.for_each_thread([](auto& thread) { thread.set_dump_backtrace_on_finalization(); }); m_process->terminate_due_to_signal(signal); return DispatchSignalResult::Terminate; } auto signal_trampoline_addr = process.signal_trampoline().get(); regs.set_ip(signal_trampoline_addr); dbgln_if(SIGNAL_DEBUG, "Thread in state '{}' has been primed with signal handler {:#04x}:{:p} to deliver {}", state_string(), m_regs.cs, m_regs.ip(), signal); return DispatchSignalResult::Continue; } RegisterState& Thread::get_register_dump_from_stack() { auto* trap = current_trap(); // We should *always* have a trap. If we don't we're probably a kernel // thread that hasn't been preempted. If we want to support this, we // need to capture the registers probably into m_regs and return it VERIFY(trap); while (trap) { if (!trap->next_trap) break; trap = trap->next_trap; } return *trap->regs; } ErrorOr> Thread::try_clone(Process& process) { auto clone = TRY(Thread::try_create(process)); m_signal_action_masks.span().copy_to(clone->m_signal_action_masks); clone->m_signal_mask = m_signal_mask; clone->m_fpu_state = m_fpu_state; clone->m_thread_specific_data = m_thread_specific_data; return clone; } void Thread::set_state(State new_state, u8 stop_signal) { State previous_state; VERIFY(g_scheduler_lock.is_locked_by_current_processor()); if (new_state == m_state) return; { previous_state = m_state; if (previous_state == Thread::State::Invalid) { // If we were *just* created, we may have already pending signals if (has_unmasked_pending_signals()) { dbgln_if(THREAD_DEBUG, "Dispatch pending signals to new thread {}", *this); dispatch_one_pending_signal(); } } m_state = new_state; dbgln_if(THREAD_DEBUG, "Set thread {} state to {}", *this, state_string()); } if (previous_state == Thread::State::Runnable) { Scheduler::dequeue_runnable_thread(*this); } else if (previous_state == Thread::State::Stopped) { m_stop_state = State::Invalid; auto& process = this->process(); if (process.set_stopped(false)) { process.for_each_thread([&](auto& thread) { if (&thread == this) return; if (!thread.is_stopped()) return; dbgln_if(THREAD_DEBUG, "Resuming peer thread {}", thread); thread.resume_from_stopped(); }); process.unblock_waiters(Thread::WaitBlocker::UnblockFlags::Continued); // Tell the parent process (if any) about this change. if (auto parent = Process::from_pid(process.ppid())) { [[maybe_unused]] auto result = parent->send_signal(SIGCHLD, &process); } } } if (m_state == Thread::State::Runnable) { Scheduler::enqueue_runnable_thread(*this); Processor::smp_wake_n_idle_processors(1); } else if (m_state == Thread::State::Stopped) { // We don't want to restore to Running state, only Runnable! m_stop_state = previous_state != Thread::State::Running ? previous_state : Thread::State::Runnable; auto& process = this->process(); if (!process.set_stopped(true)) { process.for_each_thread([&](auto& thread) { if (&thread == this) return; if (thread.is_stopped()) return; dbgln_if(THREAD_DEBUG, "Stopping peer thread {}", thread); thread.set_state(Thread::State::Stopped, stop_signal); }); process.unblock_waiters(Thread::WaitBlocker::UnblockFlags::Stopped, stop_signal); // Tell the parent process (if any) about this change. if (auto parent = Process::from_pid(process.ppid())) { [[maybe_unused]] auto result = parent->send_signal(SIGCHLD, &process); } } } else if (m_state == Thread::State::Dying) { VERIFY(previous_state != Thread::State::Blocked); if (this != Thread::current() && is_finalizable()) { // Some other thread set this thread to Dying, notify the // finalizer right away as it can be cleaned up now Scheduler::notify_finalizer(); } } } struct RecognizedSymbol { FlatPtr address; KernelSymbol const* symbol { nullptr }; }; static ErrorOr symbolicate(RecognizedSymbol const& symbol, Process& process, StringBuilder& builder) { if (symbol.address == 0) return false; auto credentials = process.credentials(); bool mask_kernel_addresses = !credentials->is_superuser(); if (!symbol.symbol) { if (!Memory::is_user_address(VirtualAddress(symbol.address))) { TRY(builder.try_append("0xdeadc0de\n"sv)); } else { TRY(process.address_space().with([&](auto& space) -> ErrorOr { if (auto* region = space->find_region_containing({ VirtualAddress(symbol.address), sizeof(FlatPtr) })) { size_t offset = symbol.address - region->vaddr().get(); if (auto region_name = region->name(); !region_name.is_null() && !region_name.is_empty()) TRY(builder.try_appendff("{:p} {} + {:#x}\n", (void*)symbol.address, region_name, offset)); else TRY(builder.try_appendff("{:p} {:p} + {:#x}\n", (void*)symbol.address, region->vaddr().as_ptr(), offset)); } else { TRY(builder.try_appendff("{:p}\n", symbol.address)); } return {}; })); } return true; } unsigned offset = symbol.address - symbol.symbol->address; if (symbol.symbol->address == g_highest_kernel_symbol_address && offset > 4096) TRY(builder.try_appendff("{:p}\n", (void*)(mask_kernel_addresses ? 0xdeadc0de : symbol.address))); else TRY(builder.try_appendff("{:p} {} + {:#x}\n", (void*)(mask_kernel_addresses ? 0xdeadc0de : symbol.address), symbol.symbol->name, offset)); return true; } ErrorOr> Thread::backtrace() { Vector recognized_symbols; auto& process = const_cast(this->process()); auto stack_trace = TRY(Processor::capture_stack_trace(*this)); VERIFY(!g_scheduler_lock.is_locked_by_current_processor()); ScopedAddressSpaceSwitcher switcher(process); for (auto& frame : stack_trace) { if (Memory::is_user_range(VirtualAddress(frame), sizeof(FlatPtr) * 2)) { TRY(recognized_symbols.try_append({ frame })); } else { TRY(recognized_symbols.try_append({ frame, symbolicate_kernel_address(frame) })); } } StringBuilder builder; for (auto& symbol : recognized_symbols) { if (!TRY(symbolicate(symbol, process, builder))) break; } return KString::try_create(builder.string_view()); } size_t Thread::thread_specific_region_alignment() const { return max(process().m_master_tls_alignment, alignof(ThreadSpecificData)); } size_t Thread::thread_specific_region_size() const { return align_up_to(process().m_master_tls_size, thread_specific_region_alignment()) + sizeof(ThreadSpecificData); } ErrorOr Thread::make_thread_specific_region(Badge) { // The process may not require a TLS region, or allocate TLS later with sys$allocate_tls (which is what dynamically loaded programs do) if (!process().m_master_tls_region) return {}; return process().address_space().with([&](auto& space) -> ErrorOr { auto* region = TRY(space->allocate_region(Memory::RandomizeVirtualAddress::Yes, {}, thread_specific_region_size(), PAGE_SIZE, "Thread-specific"sv, PROT_READ | PROT_WRITE)); m_thread_specific_range = region->range(); SmapDisabler disabler; auto* thread_specific_data = (ThreadSpecificData*)region->vaddr().offset(align_up_to(process().m_master_tls_size, thread_specific_region_alignment())).as_ptr(); auto* thread_local_storage = (u8*)((u8*)thread_specific_data) - align_up_to(process().m_master_tls_size, process().m_master_tls_alignment); m_thread_specific_data = VirtualAddress(thread_specific_data); thread_specific_data->self = thread_specific_data; if (process().m_master_tls_size != 0) memcpy(thread_local_storage, process().m_master_tls_region.unsafe_ptr()->vaddr().as_ptr(), process().m_master_tls_size); return {}; }); } LockRefPtr Thread::from_tid(ThreadID tid) { return Thread::all_instances().with([&](auto& list) -> LockRefPtr { for (Thread& thread : list) { if (thread.tid() == tid) return thread; } return nullptr; }); } void Thread::reset_fpu_state() { memcpy(&m_fpu_state, &Processor::clean_fpu_state(), sizeof(FPUState)); } bool Thread::should_be_stopped() const { return process().is_stopped(); } void Thread::track_lock_acquire(LockRank rank) { // Nothing to do for locks without a rank. if (rank == LockRank::None) return; if (m_lock_rank_mask != LockRank::None) { // Verify we are only attempting to take a lock of a higher rank. VERIFY(m_lock_rank_mask > rank); } m_lock_rank_mask |= rank; } void Thread::track_lock_release(LockRank rank) { // Nothing to do for locks without a rank. if (rank == LockRank::None) return; // The rank value from the caller should only contain a single bit, otherwise // we are disabling the tracking for multiple locks at once which will corrupt // the lock tracking mask, and we will assert somewhere else. auto rank_is_a_single_bit = [](auto rank_enum) -> bool { auto rank = to_underlying(rank_enum); auto rank_without_least_significant_bit = rank - 1; return (rank & rank_without_least_significant_bit) == 0; }; // We can't release locks out of order, as that would violate the ranking. // This is validated by toggling the least significant bit of the mask, and // then bit wise or-ing the rank we are trying to release with the resulting // mask. If the rank we are releasing is truly the highest rank then the mask // we get back will be equal to the current mask stored on the thread. auto rank_is_in_order = [](auto mask_enum, auto rank_enum) -> bool { auto mask = to_underlying(mask_enum); auto rank = to_underlying(rank_enum); auto mask_without_least_significant_bit = mask - 1; return ((mask & mask_without_least_significant_bit) | rank) == mask; }; VERIFY(has_flag(m_lock_rank_mask, rank)); VERIFY(rank_is_a_single_bit(rank)); VERIFY(rank_is_in_order(m_lock_rank_mask, rank)); m_lock_rank_mask ^= rank; } } ErrorOr AK::Formatter::format(FormatBuilder& builder, Kernel::Thread const& value) { return AK::Formatter::format( builder, "{}({}:{})"sv, value.process().name(), value.pid().value(), value.tid().value()); }