Message-ID: <send-serie.davidel@xmailserver.org.16816.1175371774.10>
Date:	Sat, 31 Mar 2007 13:09:29 -0700
From:	Davide Libenzi <davidel@...ilserver.org>
To:	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Cc:	Andrew Morton <akpm@...ux-foundation.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Ingo Molnar <mingo@...e.hu>,
	Suparna Bhattacharya <suparna@...ibm.com>,
	Zach Brown <zach.brown@...cle.com>,
	Benjamin LaHaise <bcrl@...ck.org>
Subject: [patch 10/13] signal/timer/event fds v9 - eventfd core ...

This is a very simple and light file descriptor, that can be used as
event wait/dispatch by userspace (both wait and dispatch) and by the
kernel (dispatch only). It can be used instead of pipe(2) in all cases
where those would simply be used to signal events. Their kernel overhead
is much lower than pipes, and they do not consume two fds. When used in
the kernel, it can offer an fd-bridge to enable, for example, functionalities
like KAIO or syslets/threadlets to signal to an fd the completion of certain
operations. But more in general, an eventfd can be used by the kernel to
signal readiness, in a POSIX poll/select way, of interfaces that would
otherwise be incompatible with it. The API is:

int eventfd(unsigned int count);

The eventfd API accepts an initial "count" parameter, and returns an
eventfd fd. It supports poll(2) (POLLIN, POLLOUT, POLLERR), read(2) and write(2).
The POLLIN flag is raised when the internal counter is greater than zero.
The POLLOUT flag is raised when at least a value of "1" can be written to
the internal counter.
The POLLERR flag is raised when an overflow in the counter value is detected.
The write(2) operation can never overflow the counter, since it blocks
(unless O_NONBLOCK is set, in which case -EAGAIN is returned).
But the eventfd_signal() function can do it, since it's supposed to not
sleep during its operation.
The read(2) function reads the __u64 counter value, and reset the internal
value to zero. If the value read is equal to (__u64) -1, an overflow
happened on the internal counter (due to 2^64 eventfd_signal() posts
that has never been retired - unlickely, but possible).
The write(2) call writes an __u64 count value, and adds it
to the current counter. The eventfd fd supports O_NONBLOCK also.
On the kernel side, we have:

struct file *eventfd_fget(int fd);
int eventfd_signal(struct file *file, unsigned int n);

The eventfd_fget() should be called to get a struct file* from an eventfd
fd (this is an fget() + check of f_op being an eventfd fops pointer).
The kernel can then call eventfd_signal() every time it wants to post
an event to userspace. The eventfd_signal() function can be called from any
context.
An eventfd() simple test and bench is available here:

http://www.xmailserver.org/eventfd-bench.c

This is the eventfd-based version of pipetest-4 (pipe(2) based):

http://www.xmailserver.org/pipetest-4.c

Not that performance matters much in the eventfd case, but eventfd-bench
shows almost as double as performance than pipetest-4.




Signed-off-by: Davide Libenzi <davidel@...ilserver.org>



- Davide



Index: linux-2.6.21-rc5.fds/include/linux/syscalls.h
===================================================================
--- linux-2.6.21-rc5.fds.orig/include/linux/syscalls.h	2007-03-31 12:30:36.000000000 -0700
+++ linux-2.6.21-rc5.fds/include/linux/syscalls.h	2007-03-31 12:36:11.000000000 -0700
@@ -605,6 +605,7 @@
 asmlinkage long sys_signalfd(int ufd, sigset_t __user *user_mask, size_t sizemask);
 asmlinkage long sys_timerfd(int ufd, int clockid, int flags,
 			    const struct itimerspec __user *utmr);
+asmlinkage long sys_eventfd(unsigned int count);
 
 int kernel_execve(const char *filename, char *const argv[], char *const envp[]);
 
Index: linux-2.6.21-rc5.fds/fs/eventfd.c
===================================================================
--- /dev/null	1970-01-01 00:00:00.000000000 +0000
+++ linux-2.6.21-rc5.fds/fs/eventfd.c	2007-03-31 12:36:11.000000000 -0700
@@ -0,0 +1,237 @@
+/*
+ *  fs/eventfd.c
+ *
+ *  Copyright (C) 2007  Davide Libenzi <davidel@...ilserver.org>
+ *
+ */
+
+#include <linux/file.h>
+#include <linux/poll.h>
+#include <linux/slab.h>
+#include <linux/init.h>
+#include <linux/fs.h>
+#include <linux/mount.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/jiffies.h>
+#include <linux/anon_inodes.h>
+#include <linux/eventfd.h>
+
+#include <asm/uaccess.h>
+
+
+struct eventfd_ctx {
+	spinlock_t lock;
+	wait_queue_head_t wqh;
+	/*
+	 * Every time that a write(2) is performed on an eventfd, the
+	 * value of the __u64 being written is added to "count" and a
+	 * wakeup is performed on "wqh". A read(2) will return the "count"
+	 * value to userspace, and will reset "count" to zero. The kernel
+	 * size eventfd_signal() also, adds to the "count" counter and
+	 * issue a wakeup.
+	 */
+	__u64 count;
+};
+
+
+/*
+ * Adds "n" to the eventfd counter "count". Returns "n" in case of
+ * success, or a value lower then "n" in case of coutner overflow.
+ * This function is supposed to be called by the kernel in paths
+ * that do not allow sleeping. In this function we allow the counter
+ * to reach the ULLONG_MAX value, and we signal this as overflow
+ * condition by returining a POLLERR to poll(2).
+ */
+int eventfd_signal(struct file *file, int n)
+{
+	struct eventfd_ctx *ctx = file->private_data;
+	unsigned long flags;
+
+	if (n < 0)
+		return -EINVAL;
+	spin_lock_irqsave(&ctx->lock, flags);
+	if (ULLONG_MAX - ctx->count < n)
+		n = (int) (ULLONG_MAX - ctx->count);
+	ctx->count += n;
+	if (waitqueue_active(&ctx->wqh))
+		wake_up_locked(&ctx->wqh);
+	spin_unlock_irqrestore(&ctx->lock, flags);
+
+	return n;
+}
+
+static int eventfd_release(struct inode *inode, struct file *file)
+{
+	kfree(file->private_data);
+	return 0;
+}
+
+static unsigned int eventfd_poll(struct file *file, poll_table *wait)
+{
+	struct eventfd_ctx *ctx = file->private_data;
+	unsigned int events = 0;
+	unsigned long flags;
+
+	poll_wait(file, &ctx->wqh, wait);
+
+	spin_lock_irqsave(&ctx->lock, flags);
+	if (ctx->count > 0)
+		events |= POLLIN;
+	if (ctx->count == ULLONG_MAX)
+		events |= POLLERR;
+	if (ULLONG_MAX - 1 > ctx->count)
+		events |= POLLOUT;
+	spin_unlock_irqrestore(&ctx->lock, flags);
+
+	return events;
+}
+
+static ssize_t eventfd_read(struct file *file, char __user *buf, size_t count,
+			    loff_t *ppos)
+{
+	struct eventfd_ctx *ctx = file->private_data;
+	ssize_t res;
+	__u64 ucnt;
+	DECLARE_WAITQUEUE(wait, current);
+
+	if (count < sizeof(ucnt))
+		return -EINVAL;
+	spin_lock_irq(&ctx->lock);
+	res = -EAGAIN;
+	ucnt = ctx->count;
+	if (ucnt > 0)
+		res = sizeof(ucnt);
+	else if (!(file->f_flags & O_NONBLOCK)) {
+		__add_wait_queue(&ctx->wqh, &wait);
+		for (res = 0;;) {
+			set_current_state(TASK_INTERRUPTIBLE);
+			if (ctx->count > 0) {
+				ucnt = ctx->count;
+				res = sizeof(ucnt);
+				break;
+			}
+			if (signal_pending(current)) {
+				res = -ERESTARTSYS;
+				break;
+			}
+			spin_unlock_irq(&ctx->lock);
+			schedule();
+			spin_lock_irq(&ctx->lock);
+		}
+		__remove_wait_queue(&ctx->wqh, &wait);
+		__set_current_state(TASK_RUNNING);
+	}
+	if (res > 0) {
+		ctx->count = 0;
+		if (waitqueue_active(&ctx->wqh))
+			wake_up_locked(&ctx->wqh);
+	}
+	spin_unlock_irq(&ctx->lock);
+	if (res > 0 && put_user(ucnt, (__u64 __user *) buf))
+		return -EFAULT;
+
+	return res;
+}
+
+static ssize_t eventfd_write(struct file *file, const char __user *buf, size_t count,
+			     loff_t *ppos)
+{
+	struct eventfd_ctx *ctx = file->private_data;
+	ssize_t res;
+	__u64 ucnt;
+	DECLARE_WAITQUEUE(wait, current);
+
+	if (count < sizeof(ucnt))
+		return -EINVAL;
+	if (get_user(ucnt, (const __u64 __user *) buf))
+		return -EFAULT;
+	if (ucnt == ULLONG_MAX)
+		return -EINVAL;
+	spin_lock_irq(&ctx->lock);
+	res = -EAGAIN;
+	if (ULLONG_MAX - ctx->count > ucnt)
+		res = sizeof(ucnt);
+	else if (!(file->f_flags & O_NONBLOCK)) {
+		__add_wait_queue(&ctx->wqh, &wait);
+		for (res = 0;;) {
+			set_current_state(TASK_INTERRUPTIBLE);
+			if (ULLONG_MAX - ctx->count > ucnt) {
+				res = sizeof(ucnt);
+				break;
+			}
+			if (signal_pending(current)) {
+				res = -ERESTARTSYS;
+				break;
+			}
+			spin_unlock_irq(&ctx->lock);
+			schedule();
+			spin_lock_irq(&ctx->lock);
+		}
+		__remove_wait_queue(&ctx->wqh, &wait);
+		__set_current_state(TASK_RUNNING);
+	}
+	if (res > 0) {
+		ctx->count += ucnt;
+		if (waitqueue_active(&ctx->wqh))
+			wake_up_locked(&ctx->wqh);
+	}
+	spin_unlock_irq(&ctx->lock);
+
+	return res;
+}
+
+
+static const struct file_operations eventfd_fops = {
+	.release	= eventfd_release,
+	.poll		= eventfd_poll,
+	.read		= eventfd_read,
+	.write		= eventfd_write,
+};
+
+
+struct file *eventfd_fget(int fd)
+{
+	struct file *file;
+
+	file = fget(fd);
+	if (!file)
+		return ERR_PTR(-EBADF);
+	if (file->f_op != &eventfd_fops) {
+		fput(file);
+		return ERR_PTR(-EINVAL);
+	}
+
+	return file;
+}
+
+asmlinkage long sys_eventfd(unsigned int count)
+{
+	int error, fd;
+	struct eventfd_ctx *ctx;
+	struct file *file;
+	struct inode *inode;
+
+	ctx = kmalloc(sizeof(*ctx), GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+
+	init_waitqueue_head(&ctx->wqh);
+	spin_lock_init(&ctx->lock);
+	ctx->count = count;
+
+	/*
+	 * When we call this, the initialization must be complete, since
+	 * aino_getfd() will install the fd.
+	 */
+	error = aino_getfd(&fd, &inode, &file, "[eventfd]",
+			   &eventfd_fops, ctx);
+	if (!error)
+		return fd;
+
+	kfree(ctx);
+	return error;
+}
+
Index: linux-2.6.21-rc5.fds/include/linux/eventfd.h
===================================================================
--- /dev/null	1970-01-01 00:00:00.000000000 +0000
+++ linux-2.6.21-rc5.fds/include/linux/eventfd.h	2007-03-31 12:36:11.000000000 -0700
@@ -0,0 +1,29 @@
+/*
+ *  include/linux/eventfd.h
+ *
+ *  Copyright (C) 2007  Davide Libenzi <davidel@...ilserver.org>
+ *
+ */
+
+#ifndef _LINUX_EVENTFD_H
+#define _LINUX_EVENTFD_H
+
+
+#ifdef __KERNEL__
+
+#ifdef CONFIG_EVENTFD
+
+struct file *eventfd_fget(int fd);
+int eventfd_signal(struct file *file, int n);
+
+#else /* CONFIG_EVENTFD */
+
+#define eventfd_fget(fd) ERR_PTR(-ENOSYS)
+#define eventfd_signal(f, n) 0
+
+#endif /* CONFIG_EVENTFD */
+
+#endif /* __KERNEL__ */
+
+#endif /* _LINUX_EVENTFD_H */
+
Index: linux-2.6.21-rc5.fds/fs/Makefile
===================================================================
--- linux-2.6.21-rc5.fds.orig/fs/Makefile	2007-03-31 12:30:36.000000000 -0700
+++ linux-2.6.21-rc5.fds/fs/Makefile	2007-03-31 12:36:11.000000000 -0700
@@ -25,6 +25,7 @@
 obj-$(CONFIG_ANON_INODES)	+= anon_inodes.o
 obj-$(CONFIG_SIGNALFD)		+= signalfd.o
 obj-$(CONFIG_TIMERFD)		+= timerfd.o
+obj-$(CONFIG_EVENTFD)		+= eventfd.o
 obj-$(CONFIG_COMPAT)		+= compat.o compat_ioctl.o
 
 nfsd-$(CONFIG_NFSD)		:= nfsctl.o
Index: linux-2.6.21-rc5.fds/init/Kconfig
===================================================================
--- linux-2.6.21-rc5.fds.orig/init/Kconfig	2007-03-31 12:31:08.000000000 -0700
+++ linux-2.6.21-rc5.fds/init/Kconfig	2007-03-31 12:36:38.000000000 -0700
@@ -493,6 +493,16 @@
 
 	  If unsure, say Y.
 
+config EVENTFD
+	bool "Eanble eventfd() system call" if EMBEDDED
+	depends on ANON_INODES
+	default y
+	help
+	  Enable the eventfd() system call that allows to receive both
+	  kernel notification (ie. KAIO) or userspace notifications.
+
+	  If unsure, say Y.
+
 config SHMEM
 	bool "Use full shmem filesystem" if EMBEDDED
 	default y

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists