factor/vm/os-unix.cpp

506 lines
14 KiB
C++
Raw Normal View History

2009-05-02 05:04:19 -04:00
#include "master.hpp"
2013-05-11 22:24:31 -04:00
namespace factor {
2009-05-04 02:46:13 -04:00
2013-05-11 22:24:31 -04:00
THREADHANDLE start_thread(void* (*start_routine)(void*), void* args) {
pthread_attr_t attr;
pthread_t thread;
if (pthread_attr_init(&attr) != 0)
fatal_error("pthread_attr_init() failed", 0);
if (pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_JOINABLE) != 0)
fatal_error("pthread_attr_setdetachstate() failed", 0);
if (pthread_create(&thread, &attr, start_routine, args) != 0)
fatal_error("pthread_create() failed", 0);
pthread_attr_destroy(&attr);
return thread;
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
static void* null_dll;
2009-05-02 05:04:19 -04:00
void sleep_nanos(uint64_t nsec) {
2013-05-11 22:24:31 -04:00
timespec ts;
timespec ts_rem;
int ret;
ts.tv_sec = nsec / 1000000000;
ts.tv_nsec = nsec % 1000000000;
ret = nanosleep(&ts, &ts_rem);
while (ret == -1 && errno == EINTR) {
memcpy(&ts, &ts_rem, sizeof(ts));
ret = nanosleep(&ts, &ts_rem);
}
2013-05-11 22:24:31 -04:00
if (ret == -1)
fatal_error("nanosleep failed", 0);
}
2013-05-11 22:24:31 -04:00
void factor_vm::init_ffi() { null_dll = dlopen(NULL, RTLD_LAZY); }
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
void factor_vm::ffi_dlopen(dll* dll) {
2014-02-08 10:31:59 -05:00
dll->handle = dlopen(alien_offset(dll->path), RTLD_LAZY | RTLD_GLOBAL);
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
void* factor_vm::ffi_dlsym_raw(dll* dll, symbol_char* symbol) {
return dlsym(dll ? dll->handle : null_dll, symbol);
2011-05-20 18:11:50 -04:00
}
2013-05-11 22:24:31 -04:00
void* factor_vm::ffi_dlsym(dll* dll, symbol_char* symbol) {
return FUNCTION_CODE_POINTER(ffi_dlsym_raw(dll, symbol));
2011-05-20 18:11:50 -04:00
}
#ifdef FACTOR_PPC
2013-05-11 22:24:31 -04:00
void* factor_vm::ffi_dlsym_toc(dll* dll, symbol_char* symbol) {
return FUNCTION_TOC_POINTER(ffi_dlsym_raw(dll, symbol));
2009-05-02 05:04:19 -04:00
}
2011-05-20 18:11:50 -04:00
#endif
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
void factor_vm::ffi_dlclose(dll* dll) {
if (dlclose(dll->handle))
general_error(ERROR_FFI, false_object, false_object);
dll->handle = NULL;
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
void factor_vm::primitive_existsp() {
struct stat sb;
char* path = (char*)(untag_check<byte_array>(ctx->pop()) + 1);
ctx->push(tag_boolean(stat(path, &sb) >= 0));
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
void factor_vm::move_file(const vm_char* path1, const vm_char* path2) {
int ret = 0;
do {
ret = rename((path1), (path2));
} while (ret < 0 && errno == EINTR);
2010-03-28 12:52:16 -04:00
2013-05-11 22:24:31 -04:00
if (ret < 0)
general_error(ERROR_IO, tag_fixnum(errno), false_object);
}
2013-05-11 22:24:31 -04:00
segment::segment(cell size_, bool executable_p) {
size = size_;
2013-05-11 22:24:31 -04:00
int pagesize = getpagesize();
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
int prot;
if (executable_p)
prot = (PROT_READ | PROT_WRITE | PROT_EXEC);
else
prot = (PROT_READ | PROT_WRITE);
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
char* array = (char*)mmap(NULL, pagesize + size + pagesize, prot,
MAP_ANON | MAP_PRIVATE, -1, 0);
if (array == (char*)- 1)
out_of_memory();
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
if (mprotect(array, pagesize, PROT_NONE) == -1)
fatal_error("Cannot protect low guard page", (cell)array);
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
if (mprotect(array + pagesize + size, pagesize, PROT_NONE) == -1)
fatal_error("Cannot protect high guard page", (cell)array);
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
start = (cell)(array + pagesize);
end = start + size;
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
segment::~segment() {
int pagesize = getpagesize();
int retval = munmap((void*)(start - pagesize), pagesize + size + pagesize);
if (retval)
fatal_error("Segment deallocation failed", 0);
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
void code_heap::guard_safepoint() {
if (mprotect(safepoint_page, getpagesize(), PROT_NONE) == -1)
fatal_error("Cannot protect safepoint guard page", (cell)safepoint_page);
}
2013-05-11 22:24:31 -04:00
void code_heap::unguard_safepoint() {
if (mprotect(safepoint_page, getpagesize(), PROT_WRITE) == -1)
fatal_error("Cannot unprotect safepoint guard page", (cell)safepoint_page);
}
2013-05-11 22:24:31 -04:00
void factor_vm::dispatch_signal(void* uap, void(handler)()) {
dispatch_signal_handler((cell*)&UAP_STACK_POINTER(uap),
(cell*)&UAP_PROGRAM_COUNTER(uap),
(cell)FUNCTION_CODE_POINTER(handler));
UAP_SET_TOC_POINTER(uap, (cell)FUNCTION_TOC_POINTER(handler));
2009-09-04 14:23:20 -04:00
}
2013-05-11 22:24:31 -04:00
void factor_vm::start_sampling_profiler_timer() {
struct itimerval timer;
memset((void*)&timer, 0, sizeof(struct itimerval));
timer.it_value.tv_usec = 1000000 / samples_per_second;
timer.it_interval.tv_usec = 1000000 / samples_per_second;
setitimer(ITIMER_REAL, &timer, NULL);
}
2013-05-11 22:24:31 -04:00
void factor_vm::end_sampling_profiler_timer() {
struct itimerval timer;
memset((void*)&timer, 0, sizeof(struct itimerval));
setitimer(ITIMER_REAL, &timer, NULL);
}
2013-05-11 22:24:31 -04:00
void memory_signal_handler(int signal, siginfo_t* siginfo, void* uap) {
factor_vm* vm = current_vm();
vm->verify_memory_protection_error((cell)siginfo->si_addr);
vm->signal_fault_addr = (cell)siginfo->si_addr;
vm->signal_fault_pc = (cell)UAP_PROGRAM_COUNTER(uap);
2013-05-11 22:24:31 -04:00
vm->dispatch_signal(uap, factor::memory_signal_handler_impl);
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
void synchronous_signal_handler(int signal, siginfo_t* siginfo, void* uap) {
if (factor_vm::fatal_erroring_p)
return;
2013-05-11 22:24:31 -04:00
factor_vm* vm = current_vm_p();
if (vm) {
vm->signal_number = signal;
vm->dispatch_signal(uap, factor::synchronous_signal_handler_impl);
} else
fatal_error("Foreign thread received signal", signal);
}
2013-05-11 22:24:31 -04:00
void safe_write_nonblock(int fd, void* data, ssize_t size);
2013-05-11 22:24:31 -04:00
static void enqueue_signal(factor_vm* vm, int signal) {
if (vm->signal_pipe_output != 0)
safe_write_nonblock(vm->signal_pipe_output, &signal, sizeof(int));
}
2013-05-11 22:24:31 -04:00
void enqueue_signal_handler(int signal, siginfo_t* siginfo, void* uap) {
if (factor_vm::fatal_erroring_p)
return;
2013-05-11 22:24:31 -04:00
factor_vm* vm = current_vm_p();
if (vm)
enqueue_signal(vm, signal);
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
void fep_signal_handler(int signal, siginfo_t* siginfo, void* uap) {
if (factor_vm::fatal_erroring_p)
return;
2013-05-11 22:24:31 -04:00
factor_vm* vm = current_vm_p();
if (vm) {
vm->safepoint.enqueue_fep(vm);
enqueue_signal(vm, signal);
} else
fatal_error("Foreign thread received signal", signal);
}
2013-05-11 22:24:31 -04:00
void sample_signal_handler(int signal, siginfo_t* siginfo, void* uap) {
factor_vm* vm = current_vm_p();
bool foreign_thread = false;
if (vm == NULL) {
foreign_thread = true;
vm = thread_vms.begin()->second;
}
if (atomic::load(&vm->sampling_profiler_p))
vm->safepoint.enqueue_samples(vm, 1, (cell)UAP_PROGRAM_COUNTER(uap),
2013-05-11 22:24:31 -04:00
foreign_thread);
else if (!foreign_thread)
enqueue_signal(vm, signal);
}
2013-05-11 22:24:31 -04:00
void ignore_signal_handler(int signal, siginfo_t* siginfo, void* uap) {}
2013-05-11 22:24:31 -04:00
void fpe_signal_handler(int signal, siginfo_t* siginfo, void* uap) {
factor_vm* vm = current_vm();
vm->signal_number = signal;
vm->signal_fpu_status = fpu_status(uap_fpu_status(uap));
uap_clear_fpu_status(uap);
2009-09-13 16:50:20 -04:00
2013-05-11 22:24:31 -04:00
vm->dispatch_signal(
uap, (siginfo->si_code == FPE_INTDIV || siginfo->si_code == FPE_INTOVF)
? factor::synchronous_signal_handler_impl
: factor::fp_signal_handler_impl);
}
2013-05-11 22:24:31 -04:00
static void sigaction_safe(int signum, const struct sigaction* act,
struct sigaction* oldact) {
int ret;
do {
ret = sigaction(signum, act, oldact);
} while (ret == -1 && errno == EINTR);
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
if (ret == -1)
fatal_error("sigaction failed", errno);
2009-05-02 05:04:19 -04:00
}
2013-05-11 22:24:31 -04:00
static void init_sigaction_with_handler(struct sigaction* act,
void (*handler)(int, siginfo_t*,
void*)) {
memset(act, 0, sizeof(struct sigaction));
sigemptyset(&act->sa_mask);
act->sa_sigaction = handler;
act->sa_flags = SA_SIGINFO | SA_ONSTACK;
}
2013-05-11 22:24:31 -04:00
static void safe_pipe(int* in, int* out) {
int filedes[2];
2013-05-11 22:24:31 -04:00
if (pipe(filedes) < 0)
fatal_error("Error opening pipe", errno);
2013-05-11 22:24:31 -04:00
*in = filedes[0];
*out = filedes[1];
2013-05-11 22:24:31 -04:00
if (fcntl(*in, F_SETFD, FD_CLOEXEC) < 0)
fatal_error("Error with fcntl", errno);
2013-05-11 22:24:31 -04:00
if (fcntl(*out, F_SETFD, FD_CLOEXEC) < 0)
fatal_error("Error with fcntl", errno);
}
2013-05-11 22:24:31 -04:00
static void init_signal_pipe(factor_vm* vm) {
safe_pipe(&vm->signal_pipe_input, &vm->signal_pipe_output);
2013-05-11 22:24:31 -04:00
if (fcntl(vm->signal_pipe_output, F_SETFL, O_NONBLOCK) < 0)
fatal_error("Error with fcntl", errno);
2013-05-11 22:24:31 -04:00
vm->special_objects[OBJ_SIGNAL_PIPE] = tag_fixnum(vm->signal_pipe_input);
}
2013-05-11 22:24:31 -04:00
void factor_vm::unix_init_signals() {
init_signal_pipe(this);
2013-05-11 22:24:31 -04:00
signal_callstack_seg = new segment(callstack_size, false);
2013-05-11 22:24:31 -04:00
stack_t signal_callstack;
signal_callstack.ss_sp = (char*)signal_callstack_seg->start;
signal_callstack.ss_size = signal_callstack_seg->size;
signal_callstack.ss_flags = 0;
2013-05-11 22:24:31 -04:00
if (sigaltstack(&signal_callstack, (stack_t*)NULL) < 0)
fatal_error("sigaltstack() failed", 0);
2009-05-02 05:04:19 -04:00
{
struct sigaction memory_sigaction;
init_sigaction_with_handler(&memory_sigaction, memory_signal_handler);
sigaction_safe(SIGBUS, &memory_sigaction, NULL);
sigaction_safe(SIGSEGV, &memory_sigaction, NULL);
sigaction_safe(SIGTRAP, &memory_sigaction, NULL);
}
{
struct sigaction fpe_sigaction;
init_sigaction_with_handler(&fpe_sigaction, fpe_signal_handler);
sigaction_safe(SIGFPE, &fpe_sigaction, NULL);
}
{
struct sigaction synchronous_sigaction;
init_sigaction_with_handler(&synchronous_sigaction,
synchronous_signal_handler);
sigaction_safe(SIGILL, &synchronous_sigaction, NULL);
sigaction_safe(SIGABRT, &synchronous_sigaction, NULL);
}
{
struct sigaction enqueue_sigaction;
init_sigaction_with_handler(&enqueue_sigaction, enqueue_signal_handler);
sigaction_safe(SIGWINCH, &enqueue_sigaction, NULL);
sigaction_safe(SIGUSR1, &enqueue_sigaction, NULL);
sigaction_safe(SIGCONT, &enqueue_sigaction, NULL);
sigaction_safe(SIGURG, &enqueue_sigaction, NULL);
sigaction_safe(SIGIO, &enqueue_sigaction, NULL);
sigaction_safe(SIGPROF, &enqueue_sigaction, NULL);
sigaction_safe(SIGVTALRM, &enqueue_sigaction, NULL);
#ifdef SIGINFO
sigaction_safe(SIGINFO, &enqueue_sigaction, NULL);
#endif
}
2013-05-11 22:24:31 -04:00
handle_ctrl_c();
{
struct sigaction sample_sigaction;
init_sigaction_with_handler(&sample_sigaction, sample_signal_handler);
sigaction_safe(SIGALRM, &sample_sigaction, NULL);
}
2009-05-02 05:04:19 -04:00
2013-05-11 22:24:31 -04:00
/* We don't use SA_IGN here because then the ignore action is inherited
by subprocesses, which we don't want. There is a unit test in
io.launcher.unix for this. */
{
struct sigaction ignore_sigaction;
init_sigaction_with_handler(&ignore_sigaction, ignore_signal_handler);
sigaction_safe(SIGPIPE, &ignore_sigaction, NULL);
/* We send SIGUSR2 to the stdin_loop thread to interrupt it on FEP */
sigaction_safe(SIGUSR2, &ignore_sigaction, NULL);
}
2009-05-02 05:04:19 -04:00
}
/* On Unix, shared fds such as stdin cannot be set to non-blocking mode
2013-05-11 22:24:31 -04:00
(http://homepages.tesco.net/J.deBoynePollard/FGA/dont-set-shared-file-descriptors-to-non-blocking-mode.html)
so we kludge around this by spawning a thread, which waits on a control pipe
for a signal, upon receiving this signal it reads one block of data from
stdin and writes it to a data pipe. Upon completion, it writes a 4-byte
integer to the size pipe, indicating how much data was written to the data
pipe.
The read end of the size pipe can be set to non-blocking. */
2009-05-02 05:04:19 -04:00
extern "C" {
2013-05-11 22:24:31 -04:00
int stdin_read;
int stdin_write;
int control_read;
int control_write;
int size_read;
int size_write;
bool stdin_thread_initialized_p = false;
THREADHANDLE stdin_thread;
pthread_mutex_t stdin_mutex;
}
void safe_close(int fd) {
if (close(fd) < 0)
fatal_error("error closing fd", errno);
}
bool check_write(int fd, void* data, ssize_t size) {
if (write(fd, data, size) == size)
return true;
else {
if (errno == EINTR)
return check_write(fd, data, size);
else
return false;
}
}
void safe_write(int fd, void* data, ssize_t size) {
if (!check_write(fd, data, size))
fatal_error("error writing fd", errno);
}
void safe_write_nonblock(int fd, void* data, ssize_t size) {
if (!check_write(fd, data, size) && errno != EAGAIN)
fatal_error("error writing fd", errno);
}
bool safe_read(int fd, void* data, ssize_t size) {
ssize_t bytes = read(fd, data, size);
if (bytes < 0) {
if (errno == EINTR)
return safe_read(fd, data, size);
else {
fatal_error("error reading fd", errno);
return false;
}
} else
return (bytes == size);
}
void* stdin_loop(void* arg) {
unsigned char buf[4096];
bool loop_running = true;
sigset_t mask;
sigfillset(&mask);
sigdelset(&mask, SIGUSR2);
sigdelset(&mask, SIGTTIN);
sigdelset(&mask, SIGTERM);
sigdelset(&mask, SIGQUIT);
pthread_sigmask(SIG_SETMASK, &mask, NULL);
int unused;
pthread_setcancelstate(PTHREAD_CANCEL_ENABLE, &unused);
pthread_setcanceltype(PTHREAD_CANCEL_ASYNCHRONOUS, &unused);
while (loop_running) {
if (!safe_read(control_read, buf, 1))
break;
if (buf[0] != 'X')
fatal_error("stdin_loop: bad data on control fd", buf[0]);
for (;;) {
/* If we fep, the parent thread will grab stdin_mutex and send us
SIGUSR2 to interrupt the read() call. */
pthread_mutex_lock(&stdin_mutex);
pthread_mutex_unlock(&stdin_mutex);
ssize_t bytes = read(0, buf, sizeof(buf));
if (bytes < 0) {
if (errno == EINTR)
continue;
else {
loop_running = false;
break;
}
} else if (bytes >= 0) {
safe_write(size_write, &bytes, sizeof(bytes));
if (!check_write(stdin_write, buf, bytes))
loop_running = false;
break;
}
}
}
safe_close(stdin_write);
safe_close(control_read);
return NULL;
}
void factor_vm::open_console() {
FACTOR_ASSERT(!stdin_thread_initialized_p);
safe_pipe(&control_read, &control_write);
safe_pipe(&size_read, &size_write);
safe_pipe(&stdin_read, &stdin_write);
stdin_thread = start_thread(stdin_loop, NULL);
stdin_thread_initialized_p = true;
pthread_mutex_init(&stdin_mutex, NULL);
}
2011-11-13 01:39:28 -05:00
/* This method is used to kill the stdin_loop before exiting from factor.
2013-05-11 22:24:31 -04:00
A Nvidia driver bug on Linux is the reason this has to be done, see:
http://www.nvnews.net/vbulletin/showthread.php?t=164619 */
void factor_vm::close_console() {
if (stdin_thread_initialized_p) {
pthread_cancel(stdin_thread);
pthread_join(stdin_thread, 0);
}
}
void factor_vm::lock_console() {
FACTOR_ASSERT(stdin_thread_initialized_p);
/* Lock the stdin_mutex and send the stdin_loop thread a signal to interrupt
any read() it has in progress. When the stdin loop iterates again, it will
try to lock the same mutex and wait until unlock_console() is called. */
pthread_mutex_lock(&stdin_mutex);
pthread_kill(stdin_thread, SIGUSR2);
}
void factor_vm::unlock_console() {
FACTOR_ASSERT(stdin_thread_initialized_p);
pthread_mutex_unlock(&stdin_mutex);
}
void factor_vm::ignore_ctrl_c() {
sig_t ret;
do {
ret = signal(SIGINT, SIG_DFL);
} while (ret == SIG_ERR && errno == EINTR);
}
void factor_vm::handle_ctrl_c() {
struct sigaction fep_sigaction;
init_sigaction_with_handler(&fep_sigaction, fep_signal_handler);
sigaction_safe(SIGINT, &fep_sigaction, NULL);
}
void abort() {
sig_t ret;
do {
ret = signal(SIGABRT, SIG_DFL);
} while (ret == SIG_ERR && errno == EINTR);
factor_vm::close_console();
::abort();
}
2009-05-04 02:46:13 -04:00
}