event 0.4.4 → 0.8.0

data/ext/event/{backend → selector}/epoll.h

@@ -20,6 +20,6 @@
 
 #pragma once
 
-#define EVENT_BACKEND_EPOLL
+#define EVENT_SELECTOR_EPOLL
 
-void Init_Event_Backend_EPoll(VALUE Event_Backend);
+void Init_Event_Selector_EPoll(VALUE Event_Selector);

data/ext/event/selector/kqueue.c

@@ -0,0 +1,637 @@
+// Copyright, 2021, by Samuel G. D. Williams. <http://www.codeotaku.com>
+// 
+// Permission is hereby granted, free of charge, to any person obtaining a copy
+// of this software and associated documentation files (the "Software"), to deal
+// in the Software without restriction, including without limitation the rights
+// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+// copies of the Software, and to permit persons to whom the Software is
+// furnished to do so, subject to the following conditions:
+// 
+// The above copyright notice and this permission notice shall be included in
+// all copies or substantial portions of the Software.
+// 
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+// THE SOFTWARE.
+
+#include "kqueue.h"
+#include "selector.h"
+
+#include <sys/event.h>
+#include <sys/ioctl.h>
+#include <time.h>
+#include <errno.h>
+
+static VALUE Event_Selector_KQueue = Qnil;
+
+enum {KQUEUE_MAX_EVENTS = 64};
+
+struct Event_Selector_KQueue {
+	struct Event_Selector backend;
+	int descriptor;
+};
+
+void Event_Selector_KQueue_Type_mark(void *_data)
+{
+	struct Event_Selector_KQueue *data = _data;
+	Event_Selector_mark(&data->backend);
+}
+
+static
+void close_internal(struct Event_Selector_KQueue *data) {
+	if (data->descriptor >= 0) {
+		close(data->descriptor);
+		data->descriptor = -1;
+	}
+}
+
+void Event_Selector_KQueue_Type_free(void *_data)
+{
+	struct Event_Selector_KQueue *data = _data;
+	
+	close_internal(data);
+	
+	free(data);
+}
+
+size_t Event_Selector_KQueue_Type_size(const void *data)
+{
+	return sizeof(struct Event_Selector_KQueue);
+}
+
+static const rb_data_type_t Event_Selector_KQueue_Type = {
+	.wrap_struct_name = "Event::Backend::KQueue",
+	.function = {
+		.dmark = Event_Selector_KQueue_Type_mark,
+		.dfree = Event_Selector_KQueue_Type_free,
+		.dsize = Event_Selector_KQueue_Type_size,
+	},
+	.data = NULL,
+	.flags = RUBY_TYPED_FREE_IMMEDIATELY,
+};
+
+VALUE Event_Selector_KQueue_allocate(VALUE self) {
+	struct Event_Selector_KQueue *data = NULL;
+	VALUE instance = TypedData_Make_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	Event_Selector_initialize(&data->backend, Qnil);
+	data->descriptor = -1;
+	
+	return instance;
+}
+
+VALUE Event_Selector_KQueue_initialize(VALUE self, VALUE loop) {
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	Event_Selector_initialize(&data->backend, loop);
+	int result = kqueue();
+	
+	if (result == -1) {
+		rb_sys_fail("kqueue");
+	} else {
+		ioctl(result, FIOCLEX);
+		data->descriptor = result;
+		
+		rb_update_max_fd(data->descriptor);
+	}
+	
+	return self;
+}
+
+VALUE Event_Selector_KQueue_close(VALUE self) {
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	close_internal(data);
+	
+	return Qnil;
+}
+
+VALUE Event_Selector_KQueue_transfer(int argc, VALUE *argv, VALUE self)
+{
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	return Event_Selector_wait_and_transfer(&data->backend, argc, argv);
+}
+
+VALUE Event_Selector_KQueue_yield(VALUE self)
+{
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	Event_Selector_yield(&data->backend);
+	
+	return Qnil;
+}
+
+VALUE Event_Selector_KQueue_push(VALUE self, VALUE fiber)
+{
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	Event_Selector_queue_push(&data->backend, fiber);
+	
+	return Qnil;
+}
+
+VALUE Event_Selector_KQueue_raise(int argc, VALUE *argv, VALUE self)
+{
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	return Event_Selector_wait_and_raise(&data->backend, argc, argv);
+}
+
+VALUE Event_Selector_KQueue_ready_p(VALUE self) {
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	return data->backend.ready ? Qtrue : Qfalse;
+}
+
+struct process_wait_arguments {
+	struct Event_Selector_KQueue *data;
+	pid_t pid;
+	int flags;
+};
+
+static
+int process_add_filters(int descriptor, int ident, VALUE fiber) {
+	struct kevent event = {0};
+	
+	event.ident = ident;
+	event.filter = EVFILT_PROC;
+	event.flags = EV_ADD | EV_ENABLE | EV_ONESHOT;
+	event.fflags = NOTE_EXIT;
+	event.udata = (void*)fiber;
+	
+	int result = kevent(descriptor, &event, 1, NULL, 0, NULL);
+	
+	if (result == -1) {
+		// No such process - the process has probably already terminated:
+		if (errno == ESRCH) {
+			return 0;
+		}
+		
+		rb_sys_fail("kevent(process_add_filters)");
+	}
+	
+	return 1;
+}
+
+static
+void process_remove_filters(int descriptor, int ident) {
+	struct kevent event = {0};
+	
+	event.ident = ident;
+	event.filter = EVFILT_PROC;
+	event.flags = EV_DELETE;
+	event.fflags = NOTE_EXIT;
+	
+	// Ignore the result.
+	kevent(descriptor, &event, 1, NULL, 0, NULL);
+}
+
+static
+VALUE process_wait_transfer(VALUE _arguments) {
+	struct process_wait_arguments *arguments = (struct process_wait_arguments *)_arguments;
+	
+	Event_Selector_fiber_transfer(arguments->data->backend.loop, 0, NULL);
+	
+	return Event_Selector_process_status_wait(arguments->pid);
+}
+
+static
+VALUE process_wait_rescue(VALUE _arguments, VALUE exception) {
+	struct process_wait_arguments *arguments = (struct process_wait_arguments *)_arguments;
+	
+	process_remove_filters(arguments->data->descriptor, arguments->pid);
+	
+	rb_exc_raise(exception);
+}
+
+VALUE Event_Selector_KQueue_process_wait(VALUE self, VALUE fiber, VALUE pid, VALUE flags) {
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	struct process_wait_arguments process_wait_arguments = {
+		.data = data,
+		.pid = NUM2PIDT(pid),
+		.flags = RB_NUM2INT(flags),
+	};
+	
+	int waiting = process_add_filters(data->descriptor, process_wait_arguments.pid, fiber);
+	
+	if (waiting) {
+		return rb_rescue(process_wait_transfer, (VALUE)&process_wait_arguments, process_wait_rescue, (VALUE)&process_wait_arguments);
+	} else {
+		return Event_Selector_process_status_wait(process_wait_arguments.pid);
+	}
+}
+
+static
+int io_add_filters(int descriptor, int ident, int events, VALUE fiber) {
+	int count = 0;
+	struct kevent kevents[2] = {0};
+	
+	if (events & EVENT_READABLE) {
+		kevents[count].ident = ident;
+		kevents[count].filter = EVFILT_READ;
+		kevents[count].flags = EV_ADD | EV_ENABLE | EV_ONESHOT;
+		kevents[count].udata = (void*)fiber;
+		
+// #ifdef EV_OOBAND
+// 		if (events & PRIORITY) {
+// 			kevents[count].flags |= EV_OOBAND;
+// 		}
+// #endif
+		
+		count++;
+	}
+	
+	if (events & EVENT_WRITABLE) {
+		kevents[count].ident = ident;
+		kevents[count].filter = EVFILT_WRITE;
+		kevents[count].flags = EV_ADD | EV_ENABLE | EV_ONESHOT;
+		kevents[count].udata = (void*)fiber;
+		count++;
+	}
+	
+	int result = kevent(descriptor, kevents, count, NULL, 0, NULL);
+	
+	if (result == -1) {
+		rb_sys_fail("kevent(io_add_filters)");
+	}
+	
+	return events;
+}
+
+static
+void io_remove_filters(int descriptor, int ident, int events) {
+	int count = 0;
+	struct kevent kevents[2] = {0};
+	
+	if (events & EVENT_READABLE) {
+		kevents[count].ident = ident;
+		kevents[count].filter = EVFILT_READ;
+		kevents[count].flags = EV_DELETE;
+		
+		count++;
+	}
+	
+	if (events & EVENT_WRITABLE) {
+		kevents[count].ident = ident;
+		kevents[count].filter = EVFILT_WRITE;
+		kevents[count].flags = EV_DELETE;
+		count++;
+	}
+	
+	// Ignore the result.
+	kevent(descriptor, kevents, count, NULL, 0, NULL);
+}
+
+struct io_wait_arguments {
+	struct Event_Selector_KQueue *data;
+	int events;
+	int descriptor;
+};
+
+static
+VALUE io_wait_rescue(VALUE _arguments, VALUE exception) {
+	struct io_wait_arguments *arguments = (struct io_wait_arguments *)_arguments;
+	
+	io_remove_filters(arguments->data->descriptor, arguments->descriptor, arguments->events);
+	
+	rb_exc_raise(exception);
+}
+
+static inline
+int events_from_kqueue_filter(int filter) {
+	if (filter == EVFILT_READ) return EVENT_READABLE;
+	if (filter == EVFILT_WRITE) return EVENT_WRITABLE;
+	
+	return 0;
+}
+
+static
+VALUE io_wait_transfer(VALUE _arguments) {
+	struct io_wait_arguments *arguments = (struct io_wait_arguments *)_arguments;
+	
+	VALUE result = Event_Selector_fiber_transfer(arguments->data->backend.loop, 0, NULL);
+	
+	return INT2NUM(events_from_kqueue_filter(RB_NUM2INT(result)));
+}
+
+VALUE Event_Selector_KQueue_io_wait(VALUE self, VALUE fiber, VALUE io, VALUE events) {
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	int descriptor = Event_Selector_io_descriptor(io);
+	
+	struct io_wait_arguments io_wait_arguments = {
+		.events = io_add_filters(data->descriptor, descriptor, RB_NUM2INT(events), fiber),
+		.data = data,
+		.descriptor = descriptor,
+	};
+	
+	return rb_rescue(io_wait_transfer, (VALUE)&io_wait_arguments, io_wait_rescue, (VALUE)&io_wait_arguments);
+}
+
+#ifdef HAVE_RUBY_IO_BUFFER_H
+
+struct io_read_arguments {
+	VALUE self;
+	VALUE fiber;
+	VALUE io;
+	
+	int flags;
+	
+	int descriptor;
+	
+	VALUE buffer;
+	size_t length;
+};
+
+static
+VALUE io_read_loop(VALUE _arguments) {
+	struct io_read_arguments *arguments = (struct io_read_arguments *)_arguments;
+	
+	void *base;
+	size_t size;
+	rb_io_buffer_get_mutable(arguments->buffer, &base, &size);
+	
+	size_t offset = 0;
+	size_t length = arguments->length;
+	
+	while (length > 0) {
+		size_t maximum_size = size - offset;
+		ssize_t result = read(arguments->descriptor, (char*)base+offset, maximum_size);
+		
+		if (result == 0) {
+			break;
+		} else if (result > 0) {
+			offset += result;
+			length -= result;
+		} else if (errno == EAGAIN || errno == EWOULDBLOCK) {
+			Event_Selector_KQueue_io_wait(arguments->self, arguments->fiber, arguments->io, RB_INT2NUM(READABLE));
+		} else {
+			rb_sys_fail("Event_Selector_KQueue_io_read");
+		}
+	}
+	
+	return SIZET2NUM(offset);
+}
+
+static
+VALUE io_read_ensure(VALUE _arguments) {
+	struct io_read_arguments *arguments = (struct io_read_arguments *)_arguments;
+	
+	Event_Selector_nonblock_restore(arguments->descriptor, arguments->flags);
+	
+	return Qnil;
+}
+
+VALUE Event_Selector_KQueue_io_read(VALUE self, VALUE fiber, VALUE io, VALUE buffer, VALUE _length) {
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	int descriptor = Event_Selector_io_descriptor(io);
+	
+	size_t length = NUM2SIZET(_length);
+	
+	struct io_read_arguments io_read_arguments = {
+		.self = self,
+		.fiber = fiber,
+		.io = io,
+		
+		.flags = Event_Selector_nonblock_set(descriptor),
+		.descriptor = descriptor,
+		.buffer = buffer,
+		.length = length,
+	};
+	
+	return rb_ensure(io_read_loop, (VALUE)&io_read_arguments, io_read_ensure, (VALUE)&io_read_arguments);
+}
+
+struct io_write_arguments {
+	VALUE self;
+	VALUE fiber;
+	VALUE io;
+	
+	int flags;
+	
+	int descriptor;
+	
+	VALUE buffer;
+	size_t length;
+};
+
+static
+VALUE io_write_loop(VALUE _arguments) {
+	struct io_write_arguments *arguments = (struct io_write_arguments *)_arguments;
+	
+	const void *base;
+	size_t size;
+	rb_io_buffer_get_immutable(arguments->buffer, &base, &size);
+	
+	size_t offset = 0;
+	size_t length = arguments->length;
+	
+	if (length > size) {
+		rb_raise(rb_eRuntimeError, "Length exceeds size of buffer!");
+	}
+	
+	while (length > 0) {
+		ssize_t result = write(arguments->descriptor, (char*)base+offset, length);
+		
+		if (result >= 0) {
+			offset += result;
+			length -= result;
+		} else if (errno == EAGAIN || errno == EWOULDBLOCK) {
+			Event_Selector_KQueue_io_wait(arguments->self, arguments->fiber, arguments->io, RB_INT2NUM(WRITABLE));
+		} else {
+			rb_sys_fail("Event_Selector_KQueue_io_write");
+		}
+	}
+	
+	return SIZET2NUM(offset);
+};
+
+static
+VALUE io_write_ensure(VALUE _arguments) {
+	struct io_write_arguments *arguments = (struct io_write_arguments *)_arguments;
+	
+	Event_Selector_nonblock_restore(arguments->descriptor, arguments->flags);
+	
+	return Qnil;
+};
+
+VALUE Event_Selector_KQueue_io_write(VALUE self, VALUE fiber, VALUE io, VALUE buffer, VALUE _length) {
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	int descriptor = Event_Selector_io_descriptor(io);
+	
+	size_t length = NUM2SIZET(_length);
+	
+	struct io_write_arguments io_write_arguments = {
+		.self = self,
+		.fiber = fiber,
+		.io = io,
+		
+		.flags = Event_Selector_nonblock_set(descriptor),
+		.descriptor = descriptor,
+		.buffer = buffer,
+		.length = length,
+	};
+	
+	return rb_ensure(io_write_loop, (VALUE)&io_write_arguments, io_write_ensure, (VALUE)&io_write_arguments);
+}
+
+#endif
+
+static
+struct timespec * make_timeout(VALUE duration, struct timespec * storage) {
+	if (duration == Qnil) {
+		return NULL;
+	}
+	
+	if (FIXNUM_P(duration)) {
+		storage->tv_sec = NUM2TIMET(duration);
+		storage->tv_nsec = 0;
+		
+		return storage;
+	}
+	
+	else if (RB_FLOAT_TYPE_P(duration)) {
+		double value = RFLOAT_VALUE(duration);
+		time_t seconds = value;
+		
+		storage->tv_sec = seconds;
+		storage->tv_nsec = (value - seconds) * 1000000000L;
+		
+		return storage;
+	}
+	
+	rb_raise(rb_eRuntimeError, "unable to convert timeout");
+}
+
+static
+int timeout_nonblocking(struct timespec * timespec) {
+	return timespec && timespec->tv_sec == 0 && timespec->tv_nsec == 0;
+}
+
+struct select_arguments {
+	struct Event_Selector_KQueue *data;
+	
+	int count;
+	struct kevent events[KQUEUE_MAX_EVENTS];
+	
+	struct timespec storage;
+	struct timespec *timeout;
+};
+
+static
+void * select_internal(void *_arguments) {
+	struct select_arguments * arguments = (struct select_arguments *)_arguments;
+	
+	arguments->count = kevent(arguments->data->descriptor, NULL, 0, arguments->events, arguments->count, arguments->timeout);
+	
+	return NULL;
+}
+
+static
+void select_internal_without_gvl(struct select_arguments *arguments) {
+	rb_thread_call_without_gvl(select_internal, (void *)arguments, RUBY_UBF_IO, 0);
+	
+	if (arguments->count == -1) {
+		rb_sys_fail("select_internal_without_gvl:kevent");
+	}
+}
+
+static
+void select_internal_with_gvl(struct select_arguments *arguments) {
+	select_internal((void *)arguments);
+	
+	if (arguments->count == -1) {
+		rb_sys_fail("select_internal_with_gvl:kevent");
+	}
+}
+
+VALUE Event_Selector_KQueue_select(VALUE self, VALUE duration) {
+	struct Event_Selector_KQueue *data = NULL;
+	TypedData_Get_Struct(self, struct Event_Selector_KQueue, &Event_Selector_KQueue_Type, data);
+	
+	int ready = Event_Selector_queue_flush(&data->backend);
+	
+	struct select_arguments arguments = {
+		.data = data,
+		.count = KQUEUE_MAX_EVENTS,
+		.storage = {
+			.tv_sec = 0,
+			.tv_nsec = 0
+		}
+	};
+	
+	// We break this implementation into two parts.
+	// (1) count = kevent(..., timeout = 0)
+	// (2) without gvl: kevent(..., timeout = 0) if count == 0 and timeout != 0
+	// This allows us to avoid releasing and reacquiring the GVL.
+	// Non-comprehensive testing shows this gives a 1.5x speedup.
+	arguments.timeout = &arguments.storage;
+	
+	// First do the syscall with no timeout to get any immediately available events:
+	select_internal_with_gvl(&arguments);
+	
+	// If there were no pending events, if we have a timeout, wait for more events:
+	if (arguments.count == 0 && !ready) {
+		arguments.timeout = make_timeout(duration, &arguments.storage);
+		
+		if (!timeout_nonblocking(arguments.timeout)) {
+			arguments.count = KQUEUE_MAX_EVENTS;
+			
+			select_internal_without_gvl(&arguments);
+		}
+	}
+	
+	for (int i = 0; i < arguments.count; i += 1) {
+		VALUE fiber = (VALUE)arguments.events[i].udata;
+		VALUE result = INT2NUM(arguments.events[i].filter);
+		
+		Event_Selector_fiber_transfer(fiber, 1, &result);
+	}
+	
+	return INT2NUM(arguments.count);
+}
+
+void Init_Event_Selector_KQueue(VALUE Event_Selector) {
+	Event_Selector_KQueue = rb_define_class_under(Event_Selector, "KQueue", rb_cObject);
+	
+	rb_define_alloc_func(Event_Selector_KQueue, Event_Selector_KQueue_allocate);
+	rb_define_method(Event_Selector_KQueue, "initialize", Event_Selector_KQueue_initialize, 1);
+	
+	rb_define_method(Event_Selector_KQueue, "transfer", Event_Selector_KQueue_transfer, -1);
+	rb_define_method(Event_Selector_KQueue, "yield", Event_Selector_KQueue_yield, 0);
+	rb_define_method(Event_Selector_KQueue, "push", Event_Selector_KQueue_push, 1);
+	rb_define_method(Event_Selector_KQueue, "raise", Event_Selector_KQueue_raise, -1);
+	
+	rb_define_method(Event_Selector_KQueue, "ready?", Event_Selector_KQueue_ready_p, 0);
+	
+	rb_define_method(Event_Selector_KQueue, "select", Event_Selector_KQueue_select, 1);
+	rb_define_method(Event_Selector_KQueue, "close", Event_Selector_KQueue_close, 0);
+	
+	rb_define_method(Event_Selector_KQueue, "io_wait", Event_Selector_KQueue_io_wait, 3);
+	
+#ifdef HAVE_RUBY_IO_BUFFER_H
+	rb_define_method(Event_Selector_KQueue, "io_read", Event_Selector_KQueue_io_read, 4);
+	rb_define_method(Event_Selector_KQueue, "io_write", Event_Selector_KQueue_io_write, 4);
+#endif
+	
+	rb_define_method(Event_Selector_KQueue, "process_wait", Event_Selector_KQueue_process_wait, 3);
+}