[1/1] Kevent subsystem.

[1/1] Kevent subsystem.

[Evgeniy Polyakov]

[-- Attachment #1: Type: text/plain, Size: 1157 bytes --]

Hello.

Kevent subsystem incorporates several AIO/kqueue design notes and ideas.
Kevent can be used both for edge and level notifications. It supports
socket notifications, network AIO (aio_send(), aio_recv() and 
aio_sendfile()), inode notifications (create/remove),
generic poll()/select() notifications and timer notifications.

It was tested against FreeBSD kqueue and Linux epoll and showed
noticeble performance win.

Network asynchronous IO operations were tested against Linux synchronous 
socket code and showed noticeble performance win.

Patch against linux-2.6.17-git tree attached (gzipped).
I would like to hear some comments about the overall design,
implementation and plans about it's usefullness for generic kernel.

Design notes, patches, userspace application and perfomance tests can be
found at project's homepages.

1. Kevent subsystem.
http://tservice.net.ru/~s0mbre/old/?section=projects&item=kevent

2. Network AIO.
http://tservice.net.ru/~s0mbre/old/?section=projects&item=naio

3. LWN article about kevent.
http://lwn.net/Articles/172844/

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>

Thank you.

-- 
	Evgeniy Polyakov

[-- Attachment #2: kevent-2.6.17-git.diff.gz --]
[-- Type: application/x-gunzip, Size: 24054 bytes --]

Re: [1/1] Kevent subsystem.

[James Morris]

On Thu, 22 Jun 2006, Evgeniy Polyakov wrote:

> Patch against linux-2.6.17-git tree attached (gzipped).
> I would like to hear some comments about the overall design,
> implementation and plans about it's usefullness for generic kernel.

Please send patches as in-line ascii text, along with documentation.


If they're too big, split them up logically into smaller pieces.


- James
-- 
James Morris
<jmorris@namei.org>

Re: [1/1] Kevent subsystem.

[Evgeniy Polyakov]

On Thu, Jun 22, 2006 at 03:01:24PM -0400, James Morris (jmorris@namei.org) wrote:
> On Thu, 22 Jun 2006, Evgeniy Polyakov wrote:
> 
> > Patch against linux-2.6.17-git tree attached (gzipped).
> > I would like to hear some comments about the overall design,
> > implementation and plans about it's usefullness for generic kernel.
> 
> Please send patches as in-line ascii text, along with documentation.
> 
> 
> If they're too big, split them up logically into smaller pieces.

Hmm... It is too big (>100kb) to send as plain text.
I will try to split it to several pieces and resend in a couple of
moments, although they are quite strongly logically connected.
 
> - James
> -- 
> James Morris
> <jmorris@namei.org>

-- 
	Evgeniy Polyakov

Re: [1/1] Kevent subsystem.

[Robert Iakobashvili]

Evgeniy,

On 6/22/06, Evgeniy Polyakov <johnpol@2ka.mipt.ru> wrote:

> Kevent subsystem incorporates several AIO/kqueue design notes and ideas.
> Kevent can be used both for edge and level notifications. It supports
> socket notifications, network AIO (aio_send(), aio_recv() and
> aio_sendfile()), inode notifications (create/remove),
> generic poll()/select() notifications and timer notifications.

Great job!
Smooth integration with userland asynch POSIX frameworks
(e.g. ACE POSIX_Proactor) may require syscalls (or their emulation)
with POSIX interface:

 * POSIX_API
 * aio_read
 * aio_write
 * aio_suspend
 * aio_error
 * aio_return
 * aio_cancel

where aio_suspend is very important.
-- 
Sincerely,
------------------------------------------------------------------
Robert Iakobashvili, coroberti at gmail dot com
Navigare necesse est, vivere non est necesse.
------------------------------------------------------------------

Re: [1/1] Kevent subsystem.

[Evgeniy Polyakov]

On Thu, Jun 22, 2006 at 09:53:38PM +0200, Robert Iakobashvili (coroberti@gmail.com) wrote:
> Evgeniy,
> 
> On 6/22/06, Evgeniy Polyakov <johnpol@2ka.mipt.ru> wrote:
> 
> >Kevent subsystem incorporates several AIO/kqueue design notes and ideas.
> >Kevent can be used both for edge and level notifications. It supports
> >socket notifications, network AIO (aio_send(), aio_recv() and
> >aio_sendfile()), inode notifications (create/remove),
> >generic poll()/select() notifications and timer notifications.
> 
> Great job!
> Smooth integration with userland asynch POSIX frameworks
> (e.g. ACE POSIX_Proactor) may require syscalls (or their emulation)
> with POSIX interface:
> 
> * POSIX_API
> * aio_read
> * aio_write
> * aio_suspend
> * aio_error
> * aio_return
> * aio_cancel
> 
> where aio_suspend is very important.

I've designed network AIO differently as how posix declared it's
functions. POSIX AIO is like rocket science - while one tries to
implement them all there is no time to do the real work.
Although all of them including suspend can be emulated on top of
existing calls to kevent.

> -- 
> Sincerely,
> ------------------------------------------------------------------
> Robert Iakobashvili, coroberti at gmail dot com
> Navigare necesse est, vivere non est necesse.
> ------------------------------------------------------------------

-- 
	Evgeniy Polyakov

Re: [1/1] Kevent subsystem.

[YOSHIFUJI Hideaki / 吉藤英明]

In article <20060622171436.GA26161@2ka.mipt.ru> (at Thu, 22 Jun 2006 21:14:37 +0400), Evgeniy Polyakov <johnpol@2ka.mipt.ru> says:

> Patch against linux-2.6.17-git tree attached (gzipped).
> I would like to hear some comments about the overall design,
> implementation and plans about it's usefullness for generic kernel.
> 
> Design notes, patches, userspace application and perfomance tests can be
> found at project's homepages.

| diff --git a/arch/i386/kernel/syscall_table.S b/arch/i386/kernel/syscall_table.S
| index af56987..93e23ff 100644
| --- a/arch/i386/kernel/syscall_table.S
| +++ b/arch/i386/kernel/syscall_table.S
| @@ -316,3 +316,7 @@ ENTRY(sys_call_table)
|  	.long sys_sync_file_range
|  	.long sys_tee			/* 315 */
|  	.long sys_vmsplice
| +	.long sys_aio_recv
| +	.long sys_aio_send
| +	.long sys_aio_sendfile
| +	.long sys_kevent_ctl
:

We do not waste syscall entries; we can probably use socketcall
for sys_aio_recv, sys_aio_send (and maybe, sys_aio_sendfile).


| diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
| index 25ecc6e..05d7086 100644
| --- a/net/ipv4/tcp_ipv4.c
| +++ b/net/ipv4/tcp_ipv4.c
:

Please do for tcp_ipv6.c as well.  Thank you.

-- 
YOSHIFUJI Hideaki @ USAGI Project  <yoshfuji@linux-ipv6.org>
GPG-FP  : 9022 65EB 1ECF 3AD1 0BDF  80D8 4807 F894 E062 0EEA

Re: [1/1] Kevent subsystem.

[David Miller]

From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Fri, 23 Jun 2006 15:12:44 +0900 (JST)

> We do not waste syscall entries; we can probably use socketcall
> for sys_aio_recv, sys_aio_send (and maybe, sys_aio_sendfile).

socketcall is deprecated and some architectures do not even
have a slot for it in their system call tables, one example
is x86_64

Re: [1/1] Kevent subsystem.

[YOSHIFUJI Hideaki / 吉藤英明]

In article <20060622.231407.59469243.davem@davemloft.net> (at Thu, 22 Jun 2006 23:14:07 -0700 (PDT)), David Miller <davem@davemloft.net> says:

> From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
> Date: Fri, 23 Jun 2006 15:12:44 +0900 (JST)
> 
> > We do not waste syscall entries; we can probably use socketcall
> > for sys_aio_recv, sys_aio_send (and maybe, sys_aio_sendfile).
> 
> socketcall is deprecated and some architectures do not even
> have a slot for it in their system call tables, one example
> is x86_64

Ah, okay.  Thanks.

--yoshfuji

[1/4] kevent: core files.

[Evgeniy Polyakov]

This patch includes core kevent files:
 - userspace controlling
 - kernelspace interfaces
 - initialisation
 - notification state machines

It might also inlclude parts from other subsystem (like network related
syscalls so it is possible that it will not compile without other
patches applied).

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>

diff --git a/arch/i386/kernel/syscall_table.S b/arch/i386/kernel/syscall_table.S
index af56987..93e23ff 100644
--- a/arch/i386/kernel/syscall_table.S
+++ b/arch/i386/kernel/syscall_table.S
@@ -316,3 +316,7 @@ ENTRY(sys_call_table)
 	.long sys_sync_file_range
 	.long sys_tee			/* 315 */
 	.long sys_vmsplice
+	.long sys_aio_recv
+	.long sys_aio_send
+	.long sys_aio_sendfile
+	.long sys_kevent_ctl
diff --git a/arch/x86_64/ia32/ia32entry.S b/arch/x86_64/ia32/ia32entry.S
index 5a92fed..534d516 100644
--- a/arch/x86_64/ia32/ia32entry.S
+++ b/arch/x86_64/ia32/ia32entry.S
@@ -696,4 +696,8 @@ #endif
 	.quad sys_sync_file_range
 	.quad sys_tee
 	.quad compat_sys_vmsplice
+	.quad sys_aio_recv
+	.quad sys_aio_send
+	.quad sys_aio_sendfile
+	.quad sys_kevent_ctl
 ia32_syscall_end:		
diff --git a/include/asm-i386/unistd.h b/include/asm-i386/unistd.h
index de2ccc1..52f8642 100644
--- a/include/asm-i386/unistd.h
+++ b/include/asm-i386/unistd.h
@@ -322,10 +322,14 @@ #define __NR_splice		313
 #define __NR_sync_file_range	314
 #define __NR_tee		315
 #define __NR_vmsplice		316
+#define __NR_aio_recv		317
+#define __NR_aio_send		318
+#define __NR_aio_sendfile	319
+#define __NR_kevent_ctl		320
 
 #ifdef __KERNEL__
 
-#define NR_syscalls 317
+#define NR_syscalls 321
 
 /*
  * user-visible error numbers are in the range -1 - -128: see
diff --git a/include/asm-x86_64/socket.h b/include/asm-x86_64/socket.h
index f2cdbea..1f31f86 100644
--- a/include/asm-x86_64/socket.h
+++ b/include/asm-x86_64/socket.h
@@ -49,4 +49,6 @@ #define SO_ACCEPTCONN		30
 
 #define SO_PEERSEC             31
 
+#define SO_ASYNC_SOCK		34
+
 #endif /* _ASM_SOCKET_H */
diff --git a/include/asm-x86_64/unistd.h b/include/asm-x86_64/unistd.h
index 0aff22b..352c34b 100644
--- a/include/asm-x86_64/unistd.h
+++ b/include/asm-x86_64/unistd.h
@@ -617,11 +617,18 @@ #define __NR_sync_file_range	277
 __SYSCALL(__NR_sync_file_range, sys_sync_file_range)
 #define __NR_vmsplice		278
 __SYSCALL(__NR_vmsplice, sys_vmsplice)
+#define __NR_aio_recv		279
+__SYSCALL(__NR_aio_recv, sys_aio_recv)
+#define __NR_aio_send		280
+__SYSCALL(__NR_aio_send, sys_aio_send)
+#define __NR_aio_sendfile	281
+__SYSCALL(__NR_aio_sendfile, sys_aio_sendfile)
+#define __NR_kevent_ctl		282
+__SYSCALL(__NR_kevent_ctl, sys_kevent_ctl)
 
 #ifdef __KERNEL__
 
-#define __NR_syscall_max __NR_vmsplice
-
+#define __NR_syscall_max __NR_kevent_ctl
 #ifndef __NO_STUBS
 
 /* user-visible error numbers are in the range -1 - -4095 */

diff --git a/include/linux/kevent.h b/include/linux/kevent.h
new file mode 100644
index 0000000..e94a7bf
--- /dev/null
+++ b/include/linux/kevent.h
@@ -0,0 +1,263 @@
+/*
+ * 	kevent.h
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef __KEVENT_H
+#define __KEVENT_H
+
+/*
+ * Kevent request flags.
+ */
+
+#define KEVENT_REQ_ONESHOT	0x1		/* Process this event only once and then dequeue. */
+
+/*
+ * Kevent return flags.
+ */
+#define KEVENT_RET_BROKEN	0x1		/* Kevent is broken. */
+#define KEVENT_RET_DONE		0x2		/* Kevent processing was finished successfully. */
+
+/*
+ * Kevent type set.
+ */
+#define KEVENT_SOCKET 		0
+#define KEVENT_INODE		1
+#define KEVENT_TIMER		2
+#define KEVENT_POLL		3
+#define KEVENT_NAIO		4
+#define KEVENT_AIO		5
+#define	KEVENT_MAX		6
+
+/*
+ * Per-type event sets.
+ * Number of per-event sets should be exactly as number of kevent types.
+ */
+
+/*
+ * Timer events.
+ */
+#define	KEVENT_TIMER_FIRED	0x1
+
+/*
+ * Socket/network asynchronous IO events.
+ */
+#define	KEVENT_SOCKET_RECV	0x1
+#define	KEVENT_SOCKET_ACCEPT	0x2
+#define	KEVENT_SOCKET_SEND	0x4
+
+/*
+ * Inode events.
+ */
+#define	KEVENT_INODE_CREATE	0x1
+#define	KEVENT_INODE_REMOVE	0x2
+
+/*
+ * Poll events.
+ */
+#define	KEVENT_POLL_POLLIN	0x0001
+#define	KEVENT_POLL_POLLPRI	0x0002
+#define	KEVENT_POLL_POLLOUT	0x0004
+#define	KEVENT_POLL_POLLERR	0x0008
+#define	KEVENT_POLL_POLLHUP	0x0010
+#define	KEVENT_POLL_POLLNVAL	0x0020
+
+#define	KEVENT_POLL_POLLRDNORM	0x0040
+#define	KEVENT_POLL_POLLRDBAND	0x0080
+#define	KEVENT_POLL_POLLWRNORM	0x0100
+#define	KEVENT_POLL_POLLWRBAND	0x0200
+#define	KEVENT_POLL_POLLMSG	0x0400
+#define	KEVENT_POLL_POLLREMOVE	0x1000
+
+/*
+ * Asynchronous IO events.
+ */
+#define	KEVENT_AIO_BIO		0x1
+
+#define KEVENT_MASK_ALL		0xffffffff	/* Mask of all possible event values. */
+#define KEVENT_MASK_EMPTY	0x0		/* Empty mask of ready events. */
+
+struct kevent_id
+{
+	__u32		raw[2];
+};
+
+struct ukevent
+{
+	struct kevent_id	id;			/* Id of this request, e.g. socket number, file descriptor and so on... */
+	__u32			type;			/* Event type, e.g. KEVENT_SOCK, KEVENT_INODE, KEVENT_TIMER and so on... */
+	__u32			event;			/* Event itself, e.g. SOCK_ACCEPT, INODE_CREATED, TIMER_FIRED... */
+	__u32			req_flags;		/* Per-event request flags */
+	__u32			ret_flags;		/* Per-event return flags */
+	__u32			ret_data[2];		/* Event return data. Event originator fills it with anything it likes. */
+	union {
+		__u32		user[2];		/* User's data. It is not used, just copied to/from user. */
+		void		*ptr;
+	};
+};
+
+#define	KEVENT_CTL_ADD 		0
+#define	KEVENT_CTL_REMOVE	1
+#define	KEVENT_CTL_MODIFY	2
+#define	KEVENT_CTL_WAIT		3
+#define	KEVENT_CTL_INIT		4
+
+struct kevent_user_control
+{
+	unsigned int		cmd;			/* Control command, e.g. KEVENT_ADD, KEVENT_REMOVE... */
+	unsigned int		num;			/* Number of ukevents this strucutre controls. */
+	unsigned int		timeout;		/* Timeout in milliseconds waiting for "num" events to become ready. */
+};
+
+#define KEVENT_USER_SYMBOL	'K'
+#define KEVENT_USER_CTL		_IOWR(KEVENT_USER_SYMBOL, 0, struct kevent_user_control)
+#define KEVENT_USER_WAIT	_IOWR(KEVENT_USER_SYMBOL, 1, struct kevent_user_control)
+
+#ifdef __KERNEL__
+
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/kevent_storage.h>
+#include <asm/semaphore.h>
+
+struct inode;
+struct dentry;
+struct sock;
+
+struct kevent;
+struct kevent_storage;
+typedef int (* kevent_callback_t)(struct kevent *);
+
+struct kevent
+{
+	struct ukevent		event;
+	spinlock_t		lock;			/* This lock protects ukevent manipulations, e.g. ret_flags changes. */
+
+	struct list_head	kevent_entry;		/* Entry of user's queue. */
+	struct list_head	storage_entry;		/* Entry of origin's queue. */
+	struct list_head	ready_entry;		/* Entry of user's ready. */
+
+	struct kevent_user	*user;			/* User who requested this kevent. */
+	struct kevent_storage	*st;			/* Kevent container. */
+
+	kevent_callback_t	callback;		/* Is called each time new event has been caught. */
+	kevent_callback_t	enqueue;		/* Is called each time new event is queued. */
+	kevent_callback_t	dequeue;		/* Is called each time event is dequeued. */
+
+	void			*priv;			/* Private data for different storages. 
+							 * poll()/select storage has a list of wait_queue_t containers 
+							 * for each ->poll() { poll_wait()' } here.
+							 */
+};
+
+#define KEVENT_HASH_MASK	0xff
+
+struct kevent_list
+{
+	struct list_head	kevent_list;		/* List of all kevents. */
+	spinlock_t 		kevent_lock;		/* Protects all manipulations with queue of kevents. */
+};
+
+struct kevent_user
+{
+	struct kevent_list	kqueue[KEVENT_HASH_MASK+1];
+	unsigned int		kevent_num;		/* Number of queued kevents. */
+
+	struct list_head	ready_list;		/* List of ready kevents. */
+	unsigned int		ready_num;		/* Number of ready kevents. */
+	spinlock_t 		ready_lock;		/* Protects all manipulations with ready queue. */
+
+	unsigned int		max_ready_num;		/* Requested number of kevents. */
+
+	struct semaphore	ctl_mutex;		/* Protects against simultaneous kevent_user control manipulations. */
+	struct semaphore	wait_mutex;		/* Protects against simultaneous kevent_user waits. */
+	wait_queue_head_t	wait;			/* Wait until some events are ready. */
+
+	atomic_t		refcnt;			/* Reference counter, increased for each new kevent. */
+#ifdef CONFIG_KEVENT_USER_STAT
+	unsigned long		im_num;
+	unsigned long		wait_num;
+	unsigned long		total;
+#endif
+};
+
+#define KEVENT_MAX_REQUESTS		PAGE_SIZE/sizeof(struct kevent)
+
+struct kevent *kevent_alloc(gfp_t mask);
+void kevent_free(struct kevent *k);
+int kevent_enqueue(struct kevent *k);
+int kevent_dequeue(struct kevent *k);
+int kevent_init(struct kevent *k);
+void kevent_requeue(struct kevent *k);
+
+#define list_for_each_entry_reverse_safe(pos, n, head, member)		\
+	for (pos = list_entry((head)->prev, typeof(*pos), member),	\
+		n = list_entry(pos->member.prev, typeof(*pos), member);	\
+	     prefetch(pos->member.prev), &pos->member != (head); 	\
+	     pos = n, n = list_entry(pos->member.prev, typeof(*pos), member))
+
+int kevent_break(struct kevent *k);
+int kevent_init(struct kevent *k);
+
+int kevent_init_socket(struct kevent *k);
+int kevent_init_inode(struct kevent *k);
+int kevent_init_timer(struct kevent *k);
+int kevent_init_poll(struct kevent *k);
+int kevent_init_naio(struct kevent *k);
+int kevent_init_aio(struct kevent *k);
+
+void kevent_storage_ready(struct kevent_storage *st, 
+		kevent_callback_t ready_callback, u32 event);
+int kevent_storage_init(void *origin, struct kevent_storage *st);
+void kevent_storage_fini(struct kevent_storage *st);
+int kevent_storage_enqueue(struct kevent_storage *st, struct kevent *k);
+void kevent_storage_dequeue(struct kevent_storage *st, struct kevent *k);
+
+int kevent_user_add_ukevent(struct ukevent *uk, struct kevent_user *u);
+
+#ifdef CONFIG_KEVENT_INODE
+void kevent_inode_notify(struct inode *inode, u32 event);
+void kevent_inode_notify_parent(struct dentry *dentry, u32 event);
+void kevent_inode_remove(struct inode *inode);
+#else
+static inline void kevent_inode_notify(struct inode *inode, u32 event)
+{
+}
+static inline void kevent_inode_notify_parent(struct dentry *dentry, u32 event)
+{
+}
+static inline void kevent_inode_remove(struct inode *inode)
+{
+}
+#endif /* CONFIG_KEVENT_INODE */
+#ifdef CONFIG_KEVENT_SOCKET
+
+void kevent_socket_notify(struct sock *sock, u32 event);
+int kevent_socket_dequeue(struct kevent *k);
+int kevent_socket_enqueue(struct kevent *k);
+#define sock_async(__sk) sock_flag(__sk, SOCK_ASYNC)
+#else
+static inline void kevent_socket_notify(struct sock *sock, u32 event)
+{
+}
+#define sock_async(__sk)	0
+#endif
+#endif /* __KERNEL__ */
+#endif /* __KEVENT_H */
diff --git a/include/linux/kevent_storage.h b/include/linux/kevent_storage.h
new file mode 100644
index 0000000..bd891f0
--- /dev/null
+++ b/include/linux/kevent_storage.h
@@ -0,0 +1,12 @@
+#ifndef __KEVENT_STORAGE_H
+#define __KEVENT_STORAGE_H
+
+struct kevent_storage
+{
+	void			*origin;		/* Originator's pointer, e.g. struct sock or struct file. Can be NULL. */
+	struct list_head	list;			/* List of queued kevents. */
+	unsigned int		qlen;			/* Number of queued kevents. */
+	spinlock_t		lock;			/* Protects users queue. */
+};
+
+#endif /* __KEVENT_STORAGE_H */
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
index bd67a44..33d436e 100644
--- a/include/linux/syscalls.h
+++ b/include/linux/syscalls.h
@@ -587,4 +587,8 @@ asmlinkage long sys_get_robust_list(int 
 asmlinkage long sys_set_robust_list(struct robust_list_head __user *head,
 				    size_t len);
 
+asmlinkage long sys_aio_recv(int ctl_fd, int s, void __user *buf, size_t size, unsigned flags);
+asmlinkage long sys_aio_send(int ctl_fd, int s, void __user *buf, size_t size, unsigned flags);
+asmlinkage long sys_aio_sendfile(int ctl_fd, int fd, int s, size_t size, unsigned flags);
+asmlinkage long sys_kevent_ctl(int ctl_fd, void __user *buf);
 #endif
diff --git a/init/Kconfig b/init/Kconfig
index df864a3..6135afc 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -185,6 +185,8 @@ config AUDITSYSCALL
 	  such as SELinux.  To use audit's filesystem watch feature, please
 	  ensure that INOTIFY is configured.
 
+source "kernel/kevent/Kconfig"
+
 config IKCONFIG
 	bool "Kernel .config support"
 	---help---
diff --git a/kernel/Makefile b/kernel/Makefile
index f6ef00f..eb057ea 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -36,6 +36,7 @@ obj-$(CONFIG_DETECT_SOFTLOCKUP) += softl
 obj-$(CONFIG_GENERIC_HARDIRQS) += irq/
 obj-$(CONFIG_SECCOMP) += seccomp.o
 obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o
+obj-$(CONFIG_KEVENT) += kevent/
 obj-$(CONFIG_RELAY) += relay.o
 
 ifneq ($(CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER),y)
diff --git a/kernel/kevent/Kconfig b/kernel/kevent/Kconfig
new file mode 100644
index 0000000..88b35af
--- /dev/null
+++ b/kernel/kevent/Kconfig
@@ -0,0 +1,57 @@
+config KEVENT
+	bool "Kernel event notification mechanism"
+	help
+	  This option enables event queue mechanism.
+	  It can be used as replacement for poll()/select(), AIO callback invocations,
+	  advanced timer notifications and other kernel object status changes.
+
+config KEVENT_USER_STAT
+	bool "Kevent user statistic"
+	depends on KEVENT
+	default N
+	help
+	  This option will turn kevent_user statistic collection on.
+	  Statistic data includes total number of kevent, number of kevents which are ready
+	  immediately at insertion time and number of kevents which were removed through
+	  readiness completion. It will be printed each time control kevent descriptor
+	  is closed.
+
+config KEVENT_SOCKET
+	bool "Kernel event notifications for sockets"
+	depends on NET && KEVENT
+	help
+	  This option enables notifications through KEVENT subsystem of 
+	  sockets operations, like new packet receiving conditions, ready for accept
+  	  conditions and so on.
+	
+config KEVENT_INODE
+	bool "Kernel event notifications for inodes"
+	depends on KEVENT
+	help
+	  This option enables notifications through KEVENT subsystem of 
+	  inode operations, like file creation, removal and so on.
+
+config KEVENT_TIMER
+	bool "Kernel event notifications for timers"
+	depends on KEVENT
+	help
+	  This option allows to use timers through KEVENT subsystem.
+
+config KEVENT_POLL
+	bool "Kernel event notifications for poll()/select()"
+	depends on KEVENT
+	help
+	  This option allows to use kevent subsystem for poll()/select() notifications.
+
+config KEVENT_NAIO
+	bool "Network asynchronous IO"
+	depends on KEVENT && KEVENT_SOCKET
+	help
+	  This option enables kevent based network asynchronous IO subsystem.
+
+config KEVENT_AIO
+	bool "Asynchronous IO"
+	depends on KEVENT
+	help
+	  This option allows to use kevent subsystem for AIO operations.
+	  AIO read is currently supported.
diff --git a/kernel/kevent/Makefile b/kernel/kevent/Makefile
new file mode 100644
index 0000000..7dcd651
--- /dev/null
+++ b/kernel/kevent/Makefile
@@ -0,0 +1,7 @@
+obj-y := kevent.o kevent_user.o kevent_init.o
+obj-$(CONFIG_KEVENT_SOCKET) += kevent_socket.o
+obj-$(CONFIG_KEVENT_INODE) += kevent_inode.o
+obj-$(CONFIG_KEVENT_TIMER) += kevent_timer.o
+obj-$(CONFIG_KEVENT_POLL) += kevent_poll.o
+obj-$(CONFIG_KEVENT_NAIO) += kevent_naio.o
+obj-$(CONFIG_KEVENT_AIO) += kevent_aio.o
diff --git a/kernel/kevent/kevent.c b/kernel/kevent/kevent.c
new file mode 100644
index 0000000..f699a13
--- /dev/null
+++ b/kernel/kevent/kevent.c
@@ -0,0 +1,260 @@
+/*
+ * 	kevent.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/mempool.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
+#include <linux/kevent.h>
+
+static kmem_cache_t *kevent_cache;
+
+/*
+ * Attempts to add an event into appropriate origin's queue.
+ * Returns positive value if this event is ready immediately,
+ * negative value in case of error and zero if event has been queued.
+ * ->enqueue() callback must increase origin's reference counter.
+ */
+int kevent_enqueue(struct kevent *k)
+{
+	if (k->event.type >= KEVENT_MAX)
+		return -E2BIG;
+
+	if (!k->enqueue) {
+		kevent_break(k);
+		return -EINVAL;
+	}
+	
+	return k->enqueue(k);
+}
+
+/*
+ * Remove event from the appropriate queue.
+ * ->dequeue() callback must decrease origin's reference counter.
+ */
+int kevent_dequeue(struct kevent *k)
+{
+	if (k->event.type >= KEVENT_MAX)
+		return -E2BIG;
+	
+	if (!k->dequeue) {
+		kevent_break(k);
+		return -EINVAL;
+	}
+
+	return k->dequeue(k);
+}
+
+/*
+ * Must be called before event is going to be added into some origin's queue.
+ * Initializes ->enqueue(), ->dequeue() and ->callback() callbacks.
+ * If failed, kevent should not be used or kevent_enqueue() will fail to add
+ * this kevent into origin's queue with setting
+ * KEVENT_RET_BROKEN flag in kevent->event.ret_flags.
+ */
+int kevent_init(struct kevent *k)
+{
+	int err;
+
+	spin_lock_init(&k->lock);
+	k->kevent_entry.next = LIST_POISON1;
+	k->storage_entry.next = LIST_POISON1;
+	k->ready_entry.next = LIST_POISON1;
+
+	if (k->event.type >= KEVENT_MAX)
+		return -E2BIG;
+	
+	switch (k->event.type) {
+		case KEVENT_NAIO:
+			err = kevent_init_naio(k);
+			break;
+		case KEVENT_SOCKET:
+			err = kevent_init_socket(k);
+			break;
+		case KEVENT_INODE:
+			err = kevent_init_inode(k);
+			break;
+		case KEVENT_TIMER:
+			err = kevent_init_timer(k);
+			break;
+		case KEVENT_POLL:
+			err = kevent_init_poll(k);
+			break;
+		case KEVENT_AIO:
+			err = kevent_init_aio(k);
+			break;
+		default:
+			err = -ENODEV;
+	}
+
+	return err;
+}
+
+/*
+ * Called from ->enqueue() callback when reference counter for given
+ * origin (socket, inode...) has been increased.
+ */
+int kevent_storage_enqueue(struct kevent_storage *st, struct kevent *k)
+{
+	unsigned long flags;
+
+	k->st = st;
+	spin_lock_irqsave(&st->lock, flags);
+	list_add_tail(&k->storage_entry, &st->list);
+	st->qlen++;
+	spin_unlock_irqrestore(&st->lock, flags);
+	return 0;
+}
+
+/*
+ * Dequeue kevent from origin's queue. 
+ * It does not decrease origin's reference counter in any way 
+ * and must be called before it, so storage itself must be valid.
+ * It is called from ->dequeue() callback.
+ */
+void kevent_storage_dequeue(struct kevent_storage *st, struct kevent *k)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&st->lock, flags);
+	if (k->storage_entry.next != LIST_POISON1) {
+		list_del(&k->storage_entry);
+		st->qlen--;
+	}
+	spin_unlock_irqrestore(&st->lock, flags);
+}
+
+static void __kevent_requeue(struct kevent *k, u32 event)
+{
+	int err, rem = 0;
+	unsigned long flags;
+
+	err = k->callback(k);
+
+	spin_lock_irqsave(&k->lock, flags);
+	if (err > 0) {
+		k->event.ret_flags |= KEVENT_RET_DONE;
+	} else if (err < 0) {
+		k->event.ret_flags |= KEVENT_RET_BROKEN;
+		k->event.ret_flags |= KEVENT_RET_DONE;
+	}
+	rem = (k->event.req_flags & KEVENT_REQ_ONESHOT);
+	if (!err)
+		err = (k->event.ret_flags & (KEVENT_RET_BROKEN|KEVENT_RET_DONE));
+	spin_unlock_irqrestore(&k->lock, flags);
+
+	if (err) {
+		if (rem) {
+			list_del(&k->storage_entry);
+			k->st->qlen--;
+		}
+		
+		spin_lock_irqsave(&k->user->ready_lock, flags);
+		if (k->ready_entry.next == LIST_POISON1) {
+			list_add_tail(&k->ready_entry, &k->user->ready_list);
+			k->user->ready_num++;
+		}
+		spin_unlock_irqrestore(&k->user->ready_lock, flags);
+		wake_up(&k->user->wait);
+	}
+}
+
+void kevent_requeue(struct kevent *k)
+{
+	unsigned long flags;
+	
+	spin_lock_irqsave(&k->st->lock, flags);
+	__kevent_requeue(k, 0);
+	spin_unlock_irqrestore(&k->st->lock, flags);
+}
+
+/*
+ * Called each time some activity in origin (socket, inode...) is noticed.
+ */
+void kevent_storage_ready(struct kevent_storage *st, 
+		kevent_callback_t ready_callback, u32 event)
+{
+	struct kevent *k, *n;
+
+	spin_lock(&st->lock);
+	list_for_each_entry_safe(k, n, &st->list, storage_entry) {
+		if (ready_callback)
+			ready_callback(k);
+
+		if (event & k->event.event)
+			__kevent_requeue(k, event);
+	}
+	spin_unlock(&st->lock);
+}
+
+int kevent_storage_init(void *origin, struct kevent_storage *st)
+{
+	spin_lock_init(&st->lock);
+	st->origin = origin;
+	st->qlen = 0;
+	INIT_LIST_HEAD(&st->list);
+	return 0;
+}
+
+void kevent_storage_fini(struct kevent_storage *st)
+{
+	kevent_storage_ready(st, kevent_break, KEVENT_MASK_ALL);
+}
+
+struct kevent *kevent_alloc(gfp_t mask)
+{
+	struct kevent *k;
+	
+	if (kevent_cache)
+		k = kmem_cache_alloc(kevent_cache, mask);
+	else
+		k = kzalloc(sizeof(struct kevent), mask);
+
+	return k;
+}
+
+void kevent_free(struct kevent *k)
+{
+	memset(k, 0xab, sizeof(struct kevent));
+
+	if (kevent_cache)
+		kmem_cache_free(kevent_cache, k);
+	else
+		kfree(k);
+}
+
+int __init kevent_sys_init(void)
+{
+	int err = 0;
+
+	kevent_cache = kmem_cache_create("kevent_cache", 
+			sizeof(struct kevent), 0, 0, NULL, NULL);
+	if (!kevent_cache)
+		err = -ENOMEM;
+	
+	return err;
+}
+
+late_initcall(kevent_sys_init);
diff --git a/kernel/kevent/kevent_init.c b/kernel/kevent/kevent_init.c
new file mode 100644
index 0000000..ec95114
--- /dev/null
+++ b/kernel/kevent/kevent_init.c
@@ -0,0 +1,85 @@
+/*
+ * 	kevent_init.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/errno.h>
+#include <linux/kevent.h>
+
+int kevent_break(struct kevent *k)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&k->lock, flags);
+	k->event.ret_flags |= KEVENT_RET_BROKEN;
+	spin_unlock_irqrestore(&k->lock, flags);
+	return 0;
+}
+
+#ifndef CONFIG_KEVENT_SOCKET
+int kevent_init_socket(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_INODE
+int kevent_init_inode(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_TIMER
+int kevent_init_timer(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_POLL
+int kevent_init_poll(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_NAIO
+int kevent_init_naio(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_AIO
+int kevent_init_aio(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
diff --git a/kernel/kevent/kevent_user.c b/kernel/kevent/kevent_user.c
new file mode 100644
index 0000000..566b62b
--- /dev/null
+++ b/kernel/kevent/kevent_user.c
@@ -0,0 +1,728 @@
+/*
+ * 	kevent_user.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/fs.h>
+#include <linux/file.h>
+#include <linux/mount.h>
+#include <linux/device.h>
+#include <linux/poll.h>
+#include <linux/kevent.h>
+#include <linux/jhash.h>
+#include <asm/uaccess.h>
+#include <asm/semaphore.h>
+
+static struct class *kevent_user_class;
+static char kevent_name[] = "kevent";
+static int kevent_user_major;
+
+static int kevent_user_open(struct inode *, struct file *);
+static int kevent_user_release(struct inode *, struct file *);
+static int kevent_user_ioctl(struct inode *, struct file *, 
+		unsigned int, unsigned long);
+static unsigned int kevent_user_poll(struct file *, struct poll_table_struct *);
+
+static struct file_operations kevent_user_fops = {
+	.open		= kevent_user_open,
+	.release	= kevent_user_release,
+	.ioctl		= kevent_user_ioctl,
+	.poll		= kevent_user_poll,
+	.owner		= THIS_MODULE,
+};
+
+static struct super_block *kevent_get_sb(struct file_system_type *fs_type, 
+		int flags, const char *dev_name, void *data)
+{
+	/* So original magic... */
+	return get_sb_pseudo(fs_type, kevent_name, NULL, 0xabcdef);	
+}
+
+static struct file_system_type kevent_fs_type = {
+	.name		= kevent_name,
+	.get_sb		= kevent_get_sb,
+	.kill_sb	= kill_anon_super,
+};
+
+static struct vfsmount *kevent_mnt;
+
+static unsigned int kevent_user_poll(struct file *file, struct poll_table_struct *wait)
+{
+	struct kevent_user *u = file->private_data;
+	unsigned int mask;
+	
+	poll_wait(file, &u->wait, wait);
+	mask = 0;
+
+	if (u->ready_num)
+		mask |= POLLIN | POLLRDNORM;
+
+	return mask;
+}
+
+static struct kevent_user *kevent_user_alloc(void)
+{
+	struct kevent_user *u;
+	int i;
+
+	u = kzalloc(sizeof(struct kevent_user), GFP_KERNEL);
+	if (!u)
+		return NULL;
+
+	INIT_LIST_HEAD(&u->ready_list);
+	spin_lock_init(&u->ready_lock);
+	u->ready_num = 0;
+#ifdef CONFIG_KEVENT_USER_STAT
+	u->wait_num = u->im_num = u->total = 0;
+#endif
+	for (i=0; i<KEVENT_HASH_MASK+1; ++i) {
+		INIT_LIST_HEAD(&u->kqueue[i].kevent_list);
+		spin_lock_init(&u->kqueue[i].kevent_lock);
+	}
+	u->kevent_num = 0;
+	
+	init_MUTEX(&u->ctl_mutex);
+	init_MUTEX(&u->wait_mutex);
+	init_waitqueue_head(&u->wait);
+	u->max_ready_num = 0;
+
+	atomic_set(&u->refcnt, 1);
+
+	return u;
+}
+
+static int kevent_user_open(struct inode *inode, struct file *file)
+{
+	struct kevent_user *u = kevent_user_alloc();
+	
+	if (!u)
+		return -ENOMEM;
+
+	file->private_data = u;
+	
+	return 0;
+}
+
+static inline void kevent_user_get(struct kevent_user *u)
+{
+	atomic_inc(&u->refcnt);
+}
+
+static inline void kevent_user_put(struct kevent_user *u)
+{
+	if (atomic_dec_and_test(&u->refcnt)) {
+#ifdef CONFIG_KEVENT_USER_STAT
+		printk("%s: u=%p, wait=%lu, immediately=%lu, total=%lu.\n", 
+				__func__, u, u->wait_num, u->im_num, u->total);
+#endif
+		kfree(u);
+	}
+}
+
+#if 0
+static inline unsigned int kevent_user_hash(struct ukevent *uk)
+{
+	unsigned int h = (uk->user[0] ^ uk->user[1]) ^ (uk->id.raw[0] ^ uk->id.raw[1]);
+	
+	h = (((h >> 16) & 0xffff) ^ (h & 0xffff)) & 0xffff;
+	h = (((h >> 8) & 0xff) ^ (h & 0xff)) & KEVENT_HASH_MASK;
+
+	return h;
+}
+#else
+static inline unsigned int kevent_user_hash(struct ukevent *uk)
+{
+	return jhash_1word(uk->id.raw[0], 0) & KEVENT_HASH_MASK;
+}
+#endif
+
+/*
+ * Remove kevent from user's list of all events, 
+ * dequeue it from storage and decrease user's reference counter,
+ * since this kevent does not exist anymore. That is why it is freed here.
+ */
+static void kevent_finish_user(struct kevent *k, int lock, int deq)
+{
+	struct kevent_user *u = k->user;
+	unsigned long flags;
+
+	if (lock) {
+		unsigned int hash = kevent_user_hash(&k->event);
+		struct kevent_list *l = &u->kqueue[hash];
+
+		spin_lock_irqsave(&l->kevent_lock, flags);
+		list_del(&k->kevent_entry);
+		u->kevent_num--;
+		spin_unlock_irqrestore(&l->kevent_lock, flags);
+	} else {
+		list_del(&k->kevent_entry);
+		u->kevent_num--;
+	}
+
+	if (deq)
+		kevent_dequeue(k);
+
+	spin_lock_irqsave(&u->ready_lock, flags);
+	if (k->ready_entry.next != LIST_POISON1) {
+		list_del(&k->ready_entry);
+		u->ready_num--;
+	}
+	spin_unlock_irqrestore(&u->ready_lock, flags);
+	
+	kevent_user_put(u);
+	kevent_free(k);
+}
+
+/*
+ * Dequeue one entry from user's ready queue.
+ */
+static struct kevent *__kqueue_dequeue_one_ready(struct list_head *q, 
+		unsigned int *qlen)
+{
+	struct kevent *k = NULL;
+	unsigned int len = *qlen;
+	
+	if (len && !list_empty(q)) {
+		k = list_entry(q->next, struct kevent, ready_entry);
+		list_del(&k->ready_entry);
+		*qlen = len - 1;
+	}
+	
+	return k;
+}
+
+static struct kevent *kqueue_dequeue_ready(struct kevent_user *u)
+{
+	unsigned long flags;
+	struct kevent *k;
+
+	spin_lock_irqsave(&u->ready_lock, flags);
+	k = __kqueue_dequeue_one_ready(&u->ready_list, &u->ready_num);
+	spin_unlock_irqrestore(&u->ready_lock, flags);
+
+	return k;
+}
+
+static struct kevent *__kevent_search(struct kevent_list *l, struct ukevent *uk, 
+		struct kevent_user *u)
+{
+	struct kevent *k;
+	int found = 0;
+	
+	list_for_each_entry(k, &l->kevent_list, kevent_entry) {
+		spin_lock(&k->lock);
+		if (k->event.user[0] == uk->user[0] && k->event.user[1] == uk->user[1] &&
+				k->event.id.raw[0] == uk->id.raw[0] && 
+				k->event.id.raw[1] == uk->id.raw[1]) {
+			found = 1;
+			spin_unlock(&k->lock);
+			break;
+		}
+		spin_unlock(&k->lock);
+	}
+
+	return (found)?k:NULL;
+}
+
+static int kevent_modify(struct ukevent *uk, struct kevent_user *u)
+{
+	struct kevent *k;
+	unsigned int hash = kevent_user_hash(uk);
+	struct kevent_list *l = &u->kqueue[hash];
+	int err = -ENODEV;
+	unsigned long flags;
+	
+	spin_lock_irqsave(&l->kevent_lock, flags);
+	k = __kevent_search(l, uk, u);
+	if (k) {
+		spin_lock(&k->lock);
+		k->event.event = uk->event;
+		k->event.req_flags = uk->req_flags;
+		k->event.ret_flags = 0;
+		spin_unlock(&k->lock);
+		kevent_requeue(k);
+		err = 0;
+	}
+	spin_unlock_irqrestore(&l->kevent_lock, flags);
+	
+	return err;
+}
+
+static int kevent_remove(struct ukevent *uk, struct kevent_user *u)
+{
+	int err = -ENODEV;
+	struct kevent *k;
+	unsigned int hash = kevent_user_hash(uk);
+	struct kevent_list *l = &u->kqueue[hash];
+	unsigned long flags;
+
+	spin_lock_irqsave(&l->kevent_lock, flags);
+	k = __kevent_search(l, uk, u);
+	if (k) {
+		kevent_finish_user(k, 0, 1);
+		err = 0;
+	}
+	spin_unlock_irqrestore(&l->kevent_lock, flags);
+
+	return err;
+}
+
+/*
+ * No new entry can be added or removed from any list at this point.
+ * It is not permitted to call ->ioctl() and ->release() in parallel.
+ */
+static int kevent_user_release(struct inode *inode, struct file *file)
+{
+	struct kevent_user *u = file->private_data;
+	struct kevent *k, *n;
+	int i;
+
+	for (i=0; i<KEVENT_HASH_MASK+1; ++i) {
+		struct kevent_list *l = &u->kqueue[i];
+		
+		list_for_each_entry_safe(k, n, &l->kevent_list, kevent_entry)
+			kevent_finish_user(k, 1, 1);
+	}
+
+	kevent_user_put(u);
+	file->private_data = NULL;
+
+	return 0;
+}
+
+static int kevent_user_ctl_modify(struct kevent_user *u, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	int err = 0, i;
+	struct ukevent uk;
+
+	if (down_interruptible(&u->ctl_mutex))
+		return -ERESTARTSYS;
+
+	for (i=0; i<ctl->num; ++i) {
+		if (copy_from_user(&uk, arg, sizeof(struct ukevent))) {
+			err = -EINVAL;
+			break;
+		}
+
+		if (kevent_modify(&uk, u))
+			uk.ret_flags |= KEVENT_RET_BROKEN;
+		uk.ret_flags |= KEVENT_RET_DONE;
+
+		if (copy_to_user(arg, &uk, sizeof(struct ukevent))) {
+			err = -EINVAL;
+			break;
+		}
+
+		arg += sizeof(struct ukevent);
+	}
+
+	up(&u->ctl_mutex);
+
+	return err;
+}
+
+static int kevent_user_ctl_remove(struct kevent_user *u, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	int err = 0, i;
+	struct ukevent uk;
+
+	if (down_interruptible(&u->ctl_mutex))
+		return -ERESTARTSYS;
+
+	for (i=0; i<ctl->num; ++i) {
+		if (copy_from_user(&uk, arg, sizeof(struct ukevent))) {
+			err = -EINVAL;
+			break;
+		}
+
+		if (kevent_remove(&uk, u))
+			uk.ret_flags |= KEVENT_RET_BROKEN;
+
+		uk.ret_flags |= KEVENT_RET_DONE;
+
+		if (copy_to_user(arg, &uk, sizeof(struct ukevent))) {
+			err = -EINVAL;
+			break;
+		}
+
+		arg += sizeof(struct ukevent);
+	}
+
+	up(&u->ctl_mutex);
+
+	return err;
+}
+
+int kevent_user_add_ukevent(struct ukevent *uk, struct kevent_user *u)
+{
+	struct kevent *k;
+	int err;
+
+	k = kevent_alloc(GFP_KERNEL);
+	if (!k) {
+		err = -ENOMEM;
+		goto err_out_exit;
+	}
+
+	memcpy(&k->event, uk, sizeof(struct ukevent));
+
+	k->event.ret_flags = 0;
+
+	err = kevent_init(k);
+	if (err) {
+		kevent_free(k);
+		goto err_out_exit;
+	}
+	k->user = u;
+#ifdef CONFIG_KEVENT_USER_STAT
+	u->total++;
+#endif
+	{
+		unsigned long flags;
+		unsigned int hash = kevent_user_hash(&k->event);
+		struct kevent_list *l = &u->kqueue[hash];
+		
+		spin_lock_irqsave(&l->kevent_lock, flags);
+		list_add_tail(&k->kevent_entry, &l->kevent_list);
+		u->kevent_num++;
+		kevent_user_get(u);
+		spin_unlock_irqrestore(&l->kevent_lock, flags);
+	}
+
+	err = kevent_enqueue(k);
+	if (err) {
+		memcpy(uk, &k->event, sizeof(struct ukevent));
+		if (err < 0)
+			uk->ret_flags |= KEVENT_RET_BROKEN;
+		uk->ret_flags |= KEVENT_RET_DONE;
+		kevent_finish_user(k, 1, 0);
+	} 
+
+err_out_exit:
+	return err;
+}
+
+/*
+ * Copy all ukevents from userspace, allocate kevent for each one 
+ * and add them into appropriate kevent_storages, 
+ * e.g. sockets, inodes and so on...
+ * If something goes wrong, all events will be dequeued and 
+ * negative error will be returned. 
+ * On success zero is returned and 
+ * ctl->num will be a number of finished events, either completed or failed. 
+ * Array of finished events (struct ukevent) will be placed behind 
+ * kevent_user_control structure. User must run through that array and check 
+ * ret_flags field of each ukevent structure to determine if it is fired or failed event.
+ */
+static int kevent_user_ctl_add(struct kevent_user *u, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	int err = 0, cerr = 0, num = 0, knum = 0, i;
+	void __user *orig, *ctl_addr;
+	struct ukevent uk;
+
+	if (down_interruptible(&u->ctl_mutex))
+		return -ERESTARTSYS;
+
+	orig = arg;
+	ctl_addr = arg - sizeof(struct kevent_user_control);
+#if 1
+	err = -ENFILE;
+	if (u->kevent_num + ctl->num >= 1024)
+		goto err_out_remove;
+#endif
+	for (i=0; i<ctl->num; ++i) {
+		if (copy_from_user(&uk, arg, sizeof(struct ukevent))) {
+			cerr = -EINVAL;
+			break;
+		}
+		arg += sizeof(struct ukevent);
+
+		err = kevent_user_add_ukevent(&uk, u);
+		if (err) {
+#ifdef CONFIG_KEVENT_USER_STAT
+			u->im_num++;
+#endif
+			if (copy_to_user(orig, &uk, sizeof(struct ukevent)))
+				cerr = -EINVAL;
+			orig += sizeof(struct ukevent);
+			num++;
+		} else
+			knum++;
+	}
+
+	if (cerr < 0)
+		goto err_out_remove;
+
+	ctl->num = num;
+	if (copy_to_user(ctl_addr, ctl, sizeof(struct kevent_user_control)))
+		cerr = -EINVAL;
+
+	if (cerr)
+		err = cerr;
+	if (!err)
+		err = num;
+
+err_out_remove:
+	up(&u->ctl_mutex);
+
+	return err;
+}
+
+/*
+ * Waits until at least ctl->ready_num events are ready or timeout and returns 
+ * number of ready events (in case of timeout) or number of requested events.
+ */
+static int kevent_user_wait(struct file *file, struct kevent_user *u, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	struct kevent *k;
+	int cerr = 0, num = 0;
+	void __user *ptr = arg + sizeof(struct kevent_user_control);
+
+	if (down_interruptible(&u->ctl_mutex))
+		return -ERESTARTSYS;
+
+	if (!(file->f_flags & O_NONBLOCK)) {
+		if (ctl->timeout)
+			wait_event_interruptible_timeout(u->wait, 
+				u->ready_num >= ctl->num, msecs_to_jiffies(ctl->timeout));
+		else
+			wait_event_interruptible_timeout(u->wait, 
+					u->ready_num > 0, msecs_to_jiffies(1000));
+	}
+	while (num < ctl->num && ((k = kqueue_dequeue_ready(u)) != NULL)) {
+		if (copy_to_user(ptr + num*sizeof(struct ukevent), 
+					&k->event, sizeof(struct ukevent)))
+			cerr = -EINVAL;
+
+		/*
+		 * If it is one-shot kevent, it has been removed already from
+		 * origin's queue, so we can easily free it here.
+		 */
+		if (k->event.req_flags & KEVENT_REQ_ONESHOT)
+			kevent_finish_user(k, 1, 1);
+		++num;
+#ifdef CONFIG_KEVENT_USER_STAT
+		u->wait_num++;
+#endif
+	}
+
+	ctl->num = num;
+	if (copy_to_user(arg, ctl, sizeof(struct kevent_user_control)))
+		cerr = -EINVAL;
+
+	up(&u->ctl_mutex);
+
+	return (cerr)?cerr:num;
+}
+
+static int kevent_ctl_init(void)
+{
+	struct kevent_user *u;
+	struct file *file;	
+	int fd, ret;
+
+	fd = get_unused_fd();
+	if (fd < 0)
+		return fd;
+
+	file = get_empty_filp();
+	if (!file) {
+		ret = -ENFILE;
+		goto out_put_fd;
+	}
+
+	u = kevent_user_alloc();
+	if (unlikely(!u)) {
+		ret = -ENOMEM;
+		goto out_put_file;
+	}
+
+	file->f_op = &kevent_user_fops;
+	file->f_vfsmnt = mntget(kevent_mnt);
+	file->f_dentry = dget(kevent_mnt->mnt_root);
+	file->f_mapping = file->f_dentry->d_inode->i_mapping;
+	file->f_mode = FMODE_READ;
+	file->f_flags = O_RDONLY;
+	file->private_data = u;
+	
+	fd_install(fd, file);
+
+	return fd;
+
+out_put_file:
+	put_filp(file);
+out_put_fd:
+	put_unused_fd(fd);
+	return ret;
+}
+
+static int kevent_ctl_process(struct file *file, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	int err;
+	struct kevent_user *u = file->private_data;
+
+	if (!u)
+		return -EINVAL;
+
+	switch (ctl->cmd) {
+		case KEVENT_CTL_ADD:
+			err = kevent_user_ctl_add(u, ctl, 
+					arg+sizeof(struct kevent_user_control));
+			break;
+		case KEVENT_CTL_REMOVE:
+			err = kevent_user_ctl_remove(u, ctl, 
+					arg+sizeof(struct kevent_user_control));
+			break;
+		case KEVENT_CTL_MODIFY:
+			err = kevent_user_ctl_modify(u, ctl, 
+					arg+sizeof(struct kevent_user_control));
+			break;
+		case KEVENT_CTL_WAIT:
+			err = kevent_user_wait(file, u, ctl, arg);
+			break;
+		case KEVENT_CTL_INIT:
+			err = kevent_ctl_init();
+		default:
+			err = -EINVAL;
+			break;
+	}
+
+	return err;
+}
+
+asmlinkage long sys_kevent_ctl(int fd, void __user *arg)
+{
+	int err, fput_needed;
+	struct kevent_user_control ctl;
+	struct file *file;
+
+	if (copy_from_user(&ctl, arg, sizeof(struct kevent_user_control)))
+		return -EINVAL;
+
+	if (ctl.cmd == KEVENT_CTL_INIT)
+		return kevent_ctl_init();
+
+	file = fget_light(fd, &fput_needed);
+	if (!file)
+		return -ENODEV;
+
+	err = kevent_ctl_process(file, &ctl, arg);
+
+	fput_light(file, fput_needed);
+	return err;
+}
+
+static int kevent_user_ioctl(struct inode *inode, struct file *file, 
+		unsigned int cmd, unsigned long arg)
+{
+	int err = -ENODEV;
+	struct kevent_user_control ctl;
+	struct kevent_user *u = file->private_data;
+	void __user *ptr = (void __user *)arg;
+
+	if (copy_from_user(&ctl, ptr, sizeof(struct kevent_user_control)))
+		return -EINVAL;
+
+	switch (cmd) {
+		case KEVENT_USER_CTL:
+			err = kevent_ctl_process(file, &ctl, ptr);
+			break;
+		case KEVENT_USER_WAIT:
+			err = kevent_user_wait(file, u, &ctl, ptr);
+			break;
+		default:
+			break;
+	}
+
+	return err;
+}
+
+static int __devinit kevent_user_init(void)
+{
+	struct class_device *dev;
+	int err = 0;
+	
+	err = register_filesystem(&kevent_fs_type);
+	if (err)
+		panic("%s: failed to register filesystem: err=%d.\n",
+			       kevent_name, err);
+
+	kevent_mnt = kern_mount(&kevent_fs_type);
+	if (IS_ERR(kevent_mnt))
+		panic("%s: failed to mount silesystem: err=%ld.\n", 
+				kevent_name, PTR_ERR(kevent_mnt));
+	
+	kevent_user_major = register_chrdev(0, kevent_name, &kevent_user_fops);
+	if (kevent_user_major < 0) {
+		printk(KERN_ERR "Failed to register \"%s\" char device: err=%d.\n", 
+				kevent_name, kevent_user_major);
+		return -ENODEV;
+	}
+
+	kevent_user_class = class_create(THIS_MODULE, "kevent");
+	if (IS_ERR(kevent_user_class)) {
+		printk(KERN_ERR "Failed to register \"%s\" class: err=%ld.\n", 
+				kevent_name, PTR_ERR(kevent_user_class));
+		err = PTR_ERR(kevent_user_class);
+		goto err_out_unregister;
+	}
+
+	dev = class_device_create(kevent_user_class, NULL, 
+			MKDEV(kevent_user_major, 0), NULL, kevent_name);
+	if (IS_ERR(dev)) {
+		printk(KERN_ERR "Failed to create %d.%d class device in \"%s\" class: err=%ld.\n", 
+				kevent_user_major, 0, kevent_name, PTR_ERR(dev));
+		err = PTR_ERR(dev);
+		goto err_out_class_destroy;
+	}
+
+	printk("KEVENT subsystem: chardev helper: major=%d.\n", kevent_user_major);
+
+	return 0;
+
+err_out_class_destroy:
+	class_destroy(kevent_user_class);
+err_out_unregister:
+	unregister_chrdev(kevent_user_major, kevent_name);
+
+	return err;
+}
+
+static void __devexit kevent_user_fini(void)
+{
+	class_device_destroy(kevent_user_class, MKDEV(kevent_user_major, 0));
+	class_destroy(kevent_user_class);
+	unregister_chrdev(kevent_user_major, kevent_name);
+	mntput(kevent_mnt);
+	unregister_filesystem(&kevent_fs_type);
+}
+
+module_init(kevent_user_init);
+module_exit(kevent_user_fini);
diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c
index 5433195..dcbacf5 100644
--- a/kernel/sys_ni.c
+++ b/kernel/sys_ni.c
@@ -121,6 +121,11 @@ cond_syscall(ppc_rtas);
 cond_syscall(sys_spu_run);
 cond_syscall(sys_spu_create);
 
+cond_syscall(sys_aio_recv);
+cond_syscall(sys_aio_send);
+cond_syscall(sys_aio_sendfile);
+cond_syscall(sys_kevent_ctl);
+
 /* mmu depending weak syscall entries */
 cond_syscall(sys_mprotect);
 cond_syscall(sys_msync);

-- 
	Evgeniy Polyakov

Re: [1/4] kevent: core files.

[Benjamin LaHaise]

On Fri, Jun 23, 2006 at 11:09:34AM +0400, Evgeniy Polyakov wrote:
> This patch includes core kevent files:
>  - userspace controlling
>  - kernelspace interfaces
>  - initialisation
>  - notification state machines

We don't need yet another event mechanism in the kernel, so I don't see 
why the new syscalls should be added when they don't interoperate with 
existing solutions.  If your results are enough to sway akpm that it is 
worth taking the patches, then it would make sense to merge the code with 
the already in-tree APIs.

		-ben

Re: [1/4] kevent: core files.

[Evgeniy Polyakov]

On Fri, Jun 23, 2006 at 02:44:57PM -0400, Benjamin LaHaise (bcrl@kvack.org) wrote:
> On Fri, Jun 23, 2006 at 11:09:34AM +0400, Evgeniy Polyakov wrote:
> > This patch includes core kevent files:
> >  - userspace controlling
> >  - kernelspace interfaces
> >  - initialisation
> >  - notification state machines
> 
> We don't need yet another event mechanism in the kernel, so I don't see 
> why the new syscalls should be added when they don't interoperate with 
> existing solutions.  If your results are enough to sway akpm that it is 
> worth taking the patches, then it would make sense to merge the code with 
> the already in-tree APIs.

What API are you talking about?
There is only epoll(), which is 40% slower than kevent, and AIO, which
works not as state machine, but as repeated call for the same work.
There is also inotify, which allocates new message each time event
occurs, which is not a good solution for every situation.

Linux just does not have unified event processing mechanism, which was
pointed to many times in AIO mail list and when epoll() was only
introduced. I would even say, that Linux does not have such mechanism at
all, since every potential user implements it's own, which can not be
used with others.

Kevent fixes that. Although implementation itself can be suboptimal for
some cases or even unacceptible at all, but it is really needed
functionality.

Every existing notification can be built on top of kevent. One can find
how easy it was to implement generic poll/select notifications (what
epoll() does) or socket notifications (which are similar to epoll(), but
are called from inside socket state machine, thus improving processing
performance).

> 		-ben

-- 
	Evgeniy Polyakov

Re: [1/4] kevent: core files.

[Benjamin LaHaise]

On Fri, Jun 23, 2006 at 11:24:29PM +0400, Evgeniy Polyakov wrote:
> What API are you talking about?
> There is only epoll(), which is 40% slower than kevent, and AIO, which
> works not as state machine, but as repeated call for the same work.
> There is also inotify, which allocates new message each time event
> occurs, which is not a good solution for every situation.

AIO can be implemented as a state machine.  Nothing in the API stops 
you from doing that, and in fact there was code which was implemented as 
a state machine used on 2.4 kernels.

> Linux just does not have unified event processing mechanism, which was
> pointed to many times in AIO mail list and when epoll() was only
> introduced. I would even say, that Linux does not have such mechanism at
> all, since every potential user implements it's own, which can not be
> used with others.

The epoll event API doesn't have space in the event fields for result codes 
as needed for AIO.  The AIO API does -- how is it lacking in this regard?

> Kevent fixes that. Although implementation itself can be suboptimal for
> some cases or even unacceptible at all, but it is really needed
> functionality.

At the expense of adding another API?  How is this a good thing?  Why 
not spit out events in the existing format?

> Every existing notification can be built on top of kevent. One can find
> how easy it was to implement generic poll/select notifications (what
> epoll() does) or socket notifications (which are similar to epoll(), but
> are called from inside socket state machine, thus improving processing
> performance).

So far your code is adding a lot without unifying anything.

		-ben
-- 
"Time is of no importance, Mr. President, only life is important."
Don't Email: <dont@kvack.org>.

Re: [1/4] kevent: core files.

[Evgeniy Polyakov]

On Fri, Jun 23, 2006 at 03:55:13PM -0400, Benjamin LaHaise (bcrl@kvack.org) wrote:
> On Fri, Jun 23, 2006 at 11:24:29PM +0400, Evgeniy Polyakov wrote:
> > What API are you talking about?
> > There is only epoll(), which is 40% slower than kevent, and AIO, which
> > works not as state machine, but as repeated call for the same work.
> > There is also inotify, which allocates new message each time event
> > occurs, which is not a good solution for every situation.
> 
> AIO can be implemented as a state machine.  Nothing in the API stops 
> you from doing that, and in fact there was code which was implemented as 
> a state machine used on 2.4 kernels.

But now it is implemented as repeated call for the same work, which does
not look like it can be used for any other types of work.
And repeated work introduce latencies.
As far as I recall, it is you who wanted to remove thread based approach
from AIO subsystem.

> > Linux just does not have unified event processing mechanism, which was
> > pointed to many times in AIO mail list and when epoll() was only
> > introduced. I would even say, that Linux does not have such mechanism at
> > all, since every potential user implements it's own, which can not be
> > used with others.
> 
> The epoll event API doesn't have space in the event fields for result codes 
> as needed for AIO.  The AIO API does -- how is it lacking in this regard?

AIO completion approach was designed to be used with process context VFS
update. read/write approach can not cover other types of notifications,
like inode updates or timers.

> > Kevent fixes that. Although implementation itself can be suboptimal for
> > some cases or even unacceptible at all, but it is really needed
> > functionality.
> 
> At the expense of adding another API?  How is this a good thing?  Why 
> not spit out events in the existing format?

Format of the structure transferred between the objects does not matter
at all. We can create a wrapper on kevent structures or kevent can
transform data from AIO objects.
The main design goal of kevent is to provide easy connected hooks into
any state machine, which might be used by kernelspace to notify about
any kind of events without any knowledge of it's background nature.
Kevent can be used for example as notification blocks for address
changes or it can replace netlink completely (it can even emulate
event multicasting).

Kevent is queue of events, which can be transferred from any object to
any destination.

> > Every existing notification can be built on top of kevent. One can find
> > how easy it was to implement generic poll/select notifications (what
> > epoll() does) or socket notifications (which are similar to epoll(), but
> > are called from inside socket state machine, thus improving processing
> > performance).
> 
> So far your code is adding a lot without unifying anything.

Not at all!
Kevent is a mechanism, which allows to impleement AIO, network AIO, poll
and select, timer control, adaptive readhead (as example of AIO VFS
update). All the code I present shows how to use kevent, it is not part
of the kevent. One can find Makefile in kevent dir to check what is the
core of the subsystem, which allows to be used as a transport for
events.

AIO, NAIO, poll/select, socket and timer notifications are just users.
One can add it's own usage as easy as to call kevent_storage
initialization function and event generation function. All other pieces
are hidded in the implementation.

> 		-ben
> -- 
> "Time is of no importance, Mr. President, only life is important."
> Don't Email: <dont@kvack.org>.

-- 
	Evgeniy Polyakov

Re: [1/4] kevent: core files.

[Benjamin LaHaise]

On Sat, Jun 24, 2006 at 12:17:17AM +0400, Evgeniy Polyakov wrote:
> But now it is implemented as repeated call for the same work, which does
> not look like it can be used for any other types of work.

Given an iocb, you do not have to return -EIOCBRETRY, instead return 
-EIOCBQUEUED and then from whatever context do an aio_complete() with the 
result for that iocb.

> And repeated work introduce latencies.
> As far as I recall, it is you who wanted to remove thread based approach
> from AIO subsystem.

I have essentially given up on trying to get the filesystem AIO patches 
in given that the concerns against them are "woe complexity" with no real 
recourse for inclusion being open.  If David is open to changes in the 
networking area, I'd love to see it built on top of your code.

> AIO completion approach was designed to be used with process context VFS
> update. read/write approach can not cover other types of notifications,
> like inode updates or timers.

The completion event is 100% generic and does not need to come from process 
context.  Calling aio_complete() from irq context is entirely valid.

> Format of the structure transferred between the objects does not matter
> at all. We can create a wrapper on kevent structures or kevent can
> transform data from AIO objects.

> The main design goal of kevent is to provide easy connected hooks into
> any state machine, which might be used by kernelspace to notify about
> any kind of events without any knowledge of it's background nature.
> Kevent can be used for example as notification blocks for address
> changes or it can replace netlink completely (it can even emulate
> event multicasting).
> 
> Kevent is queue of events, which can be transferred from any object to
> any destination.

And io_getevents() reads a queue of events, so I'm not sure why you need 
a new syscall.

> Not at all!
> Kevent is a mechanism, which allows to impleement AIO, network AIO, poll
> and select, timer control, adaptive readhead (as example of AIO VFS
> update). All the code I present shows how to use kevent, it is not part
> of the kevent. One can find Makefile in kevent dir to check what is the
> core of the subsystem, which allows to be used as a transport for
> events.
> 
> AIO, NAIO, poll/select, socket and timer notifications are just users.
> One can add it's own usage as easy as to call kevent_storage
> initialization function and event generation function. All other pieces
> are hidded in the implementation.

I'll look at adapting your code to use the existing syscalls.  Maybe code 
will be better at expressing my concerns.

		-ben
-- 
"Time is of no importance, Mr. President, only life is important."
Don't Email: <dont@kvack.org>.

Re: [1/4] kevent: core files.

[Evgeniy Polyakov]

On Fri, Jun 23, 2006 at 04:44:42PM -0400, Benjamin LaHaise (bcrl@kvack.org) wrote:
> > AIO completion approach was designed to be used with process context VFS
> > update. read/write approach can not cover other types of notifications,
> > like inode updates or timers.
> 
> The completion event is 100% generic and does not need to come from process 
> context.  Calling aio_complete() from irq context is entirely valid.

put_ioctx() can sleep.
And the whole approach is different: AIO just wakes up requesting
thread, so user must provide a lot to be able to work with AIO.
It perfectly fits VFS design, but it is not acceptible for generic event
notifications.

> > Format of the structure transferred between the objects does not matter
> > at all. We can create a wrapper on kevent structures or kevent can
> > transform data from AIO objects.
> 
> > The main design goal of kevent is to provide easy connected hooks into
> > any state machine, which might be used by kernelspace to notify about
> > any kind of events without any knowledge of it's background nature.
> > Kevent can be used for example as notification blocks for address
> > changes or it can replace netlink completely (it can even emulate
> > event multicasting).
> > 
> > Kevent is queue of events, which can be transferred from any object to
> > any destination.
> 
> And io_getevents() reads a queue of events, so I'm not sure why you need 
> a new syscall.

It is not syscall, but overall design should be analyzed.
It is possible to use existing ssycalls, kevent design does not care
about how it's data structures are delivered to the internal
"processor".

> > Not at all!
> > Kevent is a mechanism, which allows to impleement AIO, network AIO, poll
> > and select, timer control, adaptive readhead (as example of AIO VFS
> > update). All the code I present shows how to use kevent, it is not part
> > of the kevent. One can find Makefile in kevent dir to check what is the
> > core of the subsystem, which allows to be used as a transport for
> > events.
> > 
> > AIO, NAIO, poll/select, socket and timer notifications are just users.
> > One can add it's own usage as easy as to call kevent_storage
> > initialization function and event generation function. All other pieces
> > are hidded in the implementation.
> 
> I'll look at adapting your code to use the existing syscalls.  Maybe code 
> will be better at expressing my concerns.

That would be great.

> 		-ben

-- 
	Evgeniy Polyakov

Re: [1/4] kevent: core files.

[Benjamin LaHaise]

On Sat, Jun 24, 2006 at 01:08:27AM +0400, Evgeniy Polyakov wrote:
> On Fri, Jun 23, 2006 at 04:44:42PM -0400, Benjamin LaHaise (bcrl@kvack.org) wrote:
> > > AIO completion approach was designed to be used with process context VFS
> > > update. read/write approach can not cover other types of notifications,
> > > like inode updates or timers.
> > 
> > The completion event is 100% generic and does not need to come from process 
> > context.  Calling aio_complete() from irq context is entirely valid.
> 
> put_ioctx() can sleep.

Err, no, that should definately not be the case.  If it can, someone has 
completely broken aio.

> It is not syscall, but overall design should be analyzed.
> It is possible to use existing ssycalls, kevent design does not care
> about how it's data structures are delivered to the internal
> "processor".

Okay, that's good to hear. =-)

		-ben
-- 
"Time is of no importance, Mr. President, only life is important."
Don't Email: <dont@kvack.org>.

Re: [1/4] kevent: core files.

[Evgeniy Polyakov]

On Fri, Jun 23, 2006 at 05:31:44PM -0400, Benjamin LaHaise (bcrl@kvack.org) wrote:
> On Sat, Jun 24, 2006 at 01:08:27AM +0400, Evgeniy Polyakov wrote:
> > On Fri, Jun 23, 2006 at 04:44:42PM -0400, Benjamin LaHaise (bcrl@kvack.org) wrote:
> > > > AIO completion approach was designed to be used with process context VFS
> > > > update. read/write approach can not cover other types of notifications,
> > > > like inode updates or timers.
> > > 
> > > The completion event is 100% generic and does not need to come from process 
> > > context.  Calling aio_complete() from irq context is entirely valid.
> > 
> > put_ioctx() can sleep.
> 
> Err, no, that should definately not be the case.  If it can, someone has 
> completely broken aio.

When reference counter hits zero it flushes aio workqueue, which can
sleep.
put_ioctx() -> __put_ioctx() -> cancel_delayed_work()/flush_workqueue().

It is there at least from 2.6.15 days (it is the oldest tree I can
access using my extremely slow GPRS link).

Hang the looter!

> 		-ben

-- 
	Evgeniy Polyakov

Re: [1/4] kevent: core files.

[David Miller]

From: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Date: Fri, 23 Jun 2006 23:24:29 +0400

> Linux just does not have unified event processing mechanism, which was
> pointed to many times in AIO mail list and when epoll() was only
> introduced. I would even say, that Linux does not have such mechanism at
> all, since every potential user implements it's own, which can not be
> used with others.
> 
> Kevent fixes that. Although implementation itself can be suboptimal for
> some cases or even unacceptible at all, but it is really needed
> functionality.

I completely agree with Evgeniy here.

There is nothing in the kernel today that provides integrated event
handling.  Nothing.  So when someone says to use the "existing" stuff,
they need to have their head examined.

The existing AIO stuff stinks as a set of interfaces.  It was designed
by a standards committee, not by people truly interested in a good
performing event processing design.  It is especially poorly suited
for networking, and any networking developer understands this.

It is pretty much a foregone conclusion that we will need new
APIs to get good networking performance.  Every existing interface
has one limitation or another.

So we should be happy people like Evgeniy try to work on this stuff,
instead of discouraging them.

Re: [1/4] kevent: core files.

[Benjamin LaHaise]

On Fri, Jun 23, 2006 at 01:19:40PM -0700, David Miller wrote:
> I completely agree with Evgeniy here.
> 
> There is nothing in the kernel today that provides integrated event
> handling.  Nothing.  So when someone says to use the "existing" stuff,
> they need to have their head examined.

The existing AIO events are *events*, with the syscalls providing the 
reading of events.

> The existing AIO stuff stinks as a set of interfaces.  It was designed
> by a standards committee, not by people truly interested in a good
> performing event processing design.  It is especially poorly suited
> for networking, and any networking developer understands this.

I disagree.  Stuffing an event that a read or write is complete/ready is a 
good way of handling things, even more so with hardware that will perform 
the memory copies to/from user buffers.

> It is pretty much a foregone conclusion that we will need new
> APIs to get good networking performance.  Every existing interface
> has one limitation or another.

Eh?  Nobody has posted any numbers comparing the approaches yet, so this 
is pure handwaving, unless you have real concrete results?

> So we should be happy people like Evgeniy try to work on this stuff,
> instead of discouraging them.

I would like to encourage him, but at the same time I don't want to see 
creating APIs that essentially duplicate existing work and needlessly 
break compatibility.  I completely agree that the in-kernel APIs are not 
as encompassing as they should be, and within the kernel Evgeniy's work 
may well be the way to go.  What I do not agree is that we need new 
syscalls at this point.  I'm perfectly willing to accept proof that change 
is needed if we do a proper comparision between any new syscall API and the 
use of the existing syscall API, but the pain of introducing a new API is 
sufficiently large that I think it is worth looking at the numbers.

		-ben
-- 
"Time is of no importance, Mr. President, only life is important."
Don't Email: <dont@kvack.org>.

Re: [1/4] kevent: core files.

[Evgeniy Polyakov]

On Fri, Jun 23, 2006 at 04:31:14PM -0400, Benjamin LaHaise (bcrl@kvack.org) wrote:
> may well be the way to go.  What I do not agree is that we need new 
> syscalls at this point.  I'm perfectly willing to accept proof that change 
> is needed if we do a proper comparision between any new syscall API and the 
> use of the existing syscall API, but the pain of introducing a new API is 
> sufficiently large that I think it is worth looking at the numbers.

New syscall is just an interface. Originally kevent (and it still can)
use char device and it's ioctl method.
It is perfectly possible to create wrappers for posix aio_* calls,
although I do not see why it is needed.
No need to concentrate on end-users interface at this point - it can be
changed at any time since design allows it, we should think about
overall design and if it is ok, move forward in implementation.

Btw, new API adds only one syscall for userspace kevent processing (and
three for send/recv/sendfile for network AIO).

According to numbers: kevent compared to epoll resulted in the
folllowing numbers:
kevent: more than 2600 requests per second (trivial web server)
epoll: about 1600-1800 requests.
Number of errors for 3k bursts of connections with 30K connections total
in 10seconds:
kevent: about 2k errors.
epoll: upto 15k errors.

More detailed results can be found on project's homepage at:
tservice.net.ru/~s0mbre/old/?section=projects&item=kevent

> 		-ben

-- 
	Evgeniy Polyakov

Re: [1/4] kevent: core files.

[Robert Iakobashvili]

Hi,

> According to numbers: kevent compared to epoll resulted in the
> folllowing numbers:
> kevent: more than 2600 requests per second (trivial web server)
> epoll: about 1600-1800 requests.
> Number of errors for 3k bursts of connections with 30K connections total
> in 10seconds:
> kevent: about 2k errors.
> epoll: upto 15k errors.

If it beats the great epoll, it means a real business case for kevent.

All previous attempts in kernel as well as by glibc and other userland
emulations to provide some real AIO infrastructure and API for server
applications with performance benefits, were not too much successful.
Heavy load networking servers are normally not using AIOs on linux due
to low performance.

>From another side Windows have a very strong I/O completion ports API,
which are widely used for the most heavy load applications.

Kevent may take linux servers productivity forward in general as well as
encourage moving aio-applications from windows to linux.

-- 
Sincerely,
------------------------------------------------------------------
Robert Iakobashvili, coroberti at gmail dot com
Navigare necesse est, vivere non est necesse.
------------------------------------------------------------------

Re: [1/4] kevent: core files.

[David Miller]

From: Benjamin LaHaise <bcrl@kvack.org>
Date: Fri, 23 Jun 2006 16:31:14 -0400

> Eh?  Nobody has posted any numbers comparing the approaches yet, so this 
> is pure handwaving, unless you have real concrete results?

Evgeniy posts numbers and performance graphs on his kevent work all
the time.

Van Jacobson did in his LCA2006 net channel slides too, perhaps you
missed that.

Re: [1/4] kevent: core files.

[Benjamin LaHaise]

On Fri, Jun 23, 2006 at 01:54:23PM -0700, David Miller wrote:
> From: Benjamin LaHaise <bcrl@kvack.org>
> Date: Fri, 23 Jun 2006 16:31:14 -0400
> 
> > Eh?  Nobody has posted any numbers comparing the approaches yet, so this 
> > is pure handwaving, unless you have real concrete results?
> 
> Evgeniy posts numbers and performance graphs on his kevent work all
> the time.

But you're argueing that the performance of something that hasn't been 
tested is worse simply by nature of it not having been tested.  That's a 
fallacy of omission, iiuc.

> Van Jacobson did in his LCA2006 net channel slides too, perhaps you
> missed that.

I have yet to be convinced that the layering violation known as net channels 
is the right way to go, mostly because it breaks horribly in a few cases -- 
think what happens during periods of CPU overcommit, in which case doing too 
much in interrupt context will kill a system (which is why softirqs are 
needed).  The effect of doing all processing in user context creates issues 
with delayed acks (due to context switching to other tasks in the system), 
which will cause excess retransmits.  The hard problems associated with 
packet filtering and security are also still unresolved, which is okay for 
a paper, but a concern in real life.

There are also a number of performance flaws in the current stack that 
show up under profiling, some of which I posted fixes for, some of which 
have yet to be fixed.  The pushf/popf pipeline stall was one of the bigger 
instances of CPU wastage that Van Jacobson noticed (it shows up as bottom 
halves using lots of CPU).  Iirc, Ingo's real time patches may avoid that 
by way of reworking the irq disable/enable mechanism, which would mean the 
results need retesting.  Using the cr8 register to enable/disable 
interrupts on x86-64 might also improve things, as that would eliminate the 
flags dependancy of cli/sti...

In short, there's a lot of work that still has to be done.

		-ben
-- 
"Time is of no importance, Mr. President, only life is important."
Don't Email: <dont@kvack.org>.

Re: [1/4] kevent: core files.

[David Miller]

From: Benjamin LaHaise <bcrl@kvack.org>
Date: Fri, 23 Jun 2006 17:53:14 -0400

> The effect of doing all processing in user context creates issues 
> with delayed acks (due to context switching to other tasks in the system), 

The Linux TCP stack does this today.

Full TCP input protocol processing is done in the user process
context.

What you are not understanding is that process scheduling helps TCP,
it does not hinder it.  If the system is loaded, we want the senders
to pace themselves to the rate at which the kernel can schedule the
abundance of receiver work it has.  And this happens naturally when
the TCP protocol input processing operates in process context.

Your fear of cpu overcommit in interrupt handlers is also heavily
flawed.  Net channels do a socket demux, and a queue entail plus
ring a doorbell if necessary, nothing more.

[2/4] kevent: network notifications.

[Evgeniy Polyakov]

This patchset includes socket notifications and network asynchronous IO.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>

diff --git a/include/asm-i386/socket.h b/include/asm-i386/socket.h
index 802ae76..3473f5c 100644
--- a/include/asm-i386/socket.h
+++ b/include/asm-i386/socket.h
@@ -49,4 +49,6 @@ #define SO_ACCEPTCONN		30
 
 #define SO_PEERSEC		31
 
+#define SO_ASYNC_SOCK		34
+
 #endif /* _ASM_SOCKET_H */

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 66f8819..ea914c3 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -1269,6 +1269,8 @@ extern struct sk_buff *skb_recv_datagram
 					 int noblock, int *err);
 extern unsigned int    datagram_poll(struct file *file, struct socket *sock,
 				     struct poll_table_struct *wait);
+extern int	       skb_copy_datagram(const struct sk_buff *from, 
+					 int offset, void *dst, int size);
 extern int	       skb_copy_datagram_iovec(const struct sk_buff *from,
 					       int offset, struct iovec *to,
 					       int size);
diff --git a/include/net/sock.h b/include/net/sock.h
index d10dfec..7a2bee3 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -47,6 +47,7 @@ #include <linux/module.h>
 #include <linux/netdevice.h>
 #include <linux/skbuff.h>	/* struct sk_buff */
 #include <linux/security.h>
+#include <linux/kevent.h>
 
 #include <linux/filter.h>
 
@@ -386,6 +387,8 @@ enum sock_flags {
 	SOCK_NO_LARGESEND, /* whether to sent large segments or not */
 	SOCK_LOCALROUTE, /* route locally only, %SO_DONTROUTE setting */
 	SOCK_QUEUE_SHRUNK, /* write queue has been shrunk recently */
+	SOCK_ASYNC,
+	SOCK_ASYNC_INUSE,
 };
 
 static inline void sock_copy_flags(struct sock *nsk, struct sock *osk)
@@ -445,6 +448,21 @@ static inline int sk_stream_memory_free(
 
 extern void sk_stream_rfree(struct sk_buff *skb);
 
+struct socket_alloc {
+	struct socket socket;
+	struct inode vfs_inode;
+};
+
+static inline struct socket *SOCKET_I(struct inode *inode)
+{
+	return &container_of(inode, struct socket_alloc, vfs_inode)->socket;
+}
+
+static inline struct inode *SOCK_INODE(struct socket *socket)
+{
+	return &container_of(socket, struct socket_alloc, socket)->vfs_inode;
+}
+
 static inline void sk_stream_set_owner_r(struct sk_buff *skb, struct sock *sk)
 {
 	skb->sk = sk;
@@ -472,6 +490,7 @@ static inline void sk_add_backlog(struct
 		sk->sk_backlog.tail = skb;
 	}
 	skb->next = NULL;
+	kevent_socket_notify(sk, KEVENT_SOCKET_RECV);
 }
 
 #define sk_wait_event(__sk, __timeo, __condition)		\
@@ -543,6 +562,12 @@ struct proto {
 
 	int			(*backlog_rcv) (struct sock *sk, 
 						struct sk_buff *skb);
+	
+	int			(*async_recv) (struct sock *sk, 
+						void *dst, size_t size);
+	int			(*async_send) (struct sock *sk, 
+						struct page **pages, unsigned int poffset, 
+						size_t size);
 
 	/* Keeping track of sk's, looking them up, and port selection methods. */
 	void			(*hash)(struct sock *sk);
@@ -674,21 +699,6 @@ static inline struct kiocb *siocb_to_kio
 	return si->kiocb;
 }
 
-struct socket_alloc {
-	struct socket socket;
-	struct inode vfs_inode;
-};
-
-static inline struct socket *SOCKET_I(struct inode *inode)
-{
-	return &container_of(inode, struct socket_alloc, vfs_inode)->socket;
-}
-
-static inline struct inode *SOCK_INODE(struct socket *socket)
-{
-	return &container_of(socket, struct socket_alloc, socket)->vfs_inode;
-}
-
 extern void __sk_stream_mem_reclaim(struct sock *sk);
 extern int sk_stream_mem_schedule(struct sock *sk, int size, int kind);
 
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 5f4eb5c..820cd5a 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -364,6 +364,8 @@ extern int			compat_tcp_setsockopt(struc
 					int level, int optname,
 					char __user *optval, int optlen);
 extern void			tcp_set_keepalive(struct sock *sk, int val);
+extern int			tcp_async_recv(struct sock *sk, void *dst, size_t size);
+extern int			tcp_async_send(struct sock *sk, struct page **pages, unsigned int poffset, size_t size);
 extern int			tcp_recvmsg(struct kiocb *iocb, struct sock *sk,
 					    struct msghdr *msg,
 					    size_t len, int nonblock, 
@@ -857,6 +859,7 @@ static inline int tcp_prequeue(struct so
 			tp->ucopy.memory = 0;
 		} else if (skb_queue_len(&tp->ucopy.prequeue) == 1) {
 			wake_up_interruptible(sk->sk_sleep);
+			kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
 			if (!inet_csk_ack_scheduled(sk))
 				inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
 						          (3 * TCP_RTO_MIN) / 4,
diff --git a/kernel/kevent/kevent_naio.c b/kernel/kevent/kevent_naio.c
new file mode 100644
index 0000000..1c71021
--- /dev/null
+++ b/kernel/kevent/kevent_naio.c
@@ -0,0 +1,239 @@
+/*
+ * 	kevent_naio.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/file.h>
+#include <linux/pagemap.h>
+#include <linux/kevent.h>
+
+#include <net/sock.h>
+#include <net/tcp_states.h>
+
+static int kevent_naio_enqueue(struct kevent *k);
+static int kevent_naio_dequeue(struct kevent *k);
+static int kevent_naio_callback(struct kevent *k);
+
+static int kevent_naio_setup_aio(int ctl_fd, int s, void __user *buf, 
+		size_t size, u32 event)
+{
+	struct kevent_user *u;
+	struct file *file;
+	int err, fput_needed;
+	struct ukevent uk;
+
+	file = fget_light(ctl_fd, &fput_needed);
+	if (!file)
+		return -ENODEV;
+
+	u = file->private_data;
+	if (!u) {
+		err = -EINVAL;
+		goto err_out_fput;
+	}
+
+	memset(&uk, 0, sizeof(struct ukevent));
+	uk.type = KEVENT_NAIO;
+	uk.ptr = buf;
+	uk.req_flags = KEVENT_REQ_ONESHOT;
+	uk.event = event;
+	uk.id.raw[0] = s;
+	uk.id.raw[1] = size;
+
+	err = kevent_user_add_ukevent(&uk, u);
+
+err_out_fput:
+	fput_light(file, fput_needed);
+	return err;
+}
+
+asmlinkage long sys_aio_recv(int ctl_fd, int s, void __user *buf, 
+		size_t size, unsigned flags)
+{
+	return kevent_naio_setup_aio(ctl_fd, s, buf, size, KEVENT_SOCKET_RECV);
+}
+
+asmlinkage long sys_aio_send(int ctl_fd, int s, void __user *buf, 
+		size_t size, unsigned flags)
+{
+	return kevent_naio_setup_aio(ctl_fd, s, buf, size, KEVENT_SOCKET_SEND);
+}
+
+static int kevent_naio_enqueue(struct kevent *k)
+{
+	int err, i;
+	struct page **page;
+	void *addr;
+	unsigned int size = k->event.id.raw[1];
+	int num = size/PAGE_SIZE;
+	struct file *file;
+	struct sock *sk = NULL;
+	int fput_needed;
+
+	file = fget_light(k->event.id.raw[0], &fput_needed);
+	if (!file)
+		return -ENODEV;
+
+	err = -EINVAL;
+	if (!file->f_dentry || !file->f_dentry->d_inode)
+		goto err_out_fput;
+
+	sk = SOCKET_I(file->f_dentry->d_inode)->sk;
+
+	err = -ESOCKTNOSUPPORT;
+	if (!sk || !sk->sk_prot->async_recv || !sk->sk_prot->async_send || 
+		!sock_flag(sk, SOCK_ASYNC))
+		goto err_out_fput;
+	
+	addr = k->event.ptr;
+	if (((unsigned long)addr & PAGE_MASK) != (unsigned long)addr)
+		num++;
+
+	page = kmalloc(sizeof(struct page *) * num, GFP_KERNEL);
+	if (!page)
+		return -ENOMEM;
+
+	down_read(&current->mm->mmap_sem);
+	err = get_user_pages(current, current->mm, (unsigned long)addr, 
+			num, 1, 0, page, NULL);
+	up_read(&current->mm->mmap_sem);
+	if (err <= 0)
+		goto err_out_free;
+	num = err;
+
+	k->event.ret_data[0] = num;
+	k->event.ret_data[1] = offset_in_page(k->event.ptr);
+	k->priv = page;
+
+	sk->sk_allocation = GFP_ATOMIC;
+
+	spin_lock_bh(&sk->sk_lock.slock);
+	err = kevent_socket_enqueue(k);
+	spin_unlock_bh(&sk->sk_lock.slock);
+	if (err)
+		goto err_out_put_pages;
+
+	fput_light(file, fput_needed);
+
+	return err;
+
+err_out_put_pages:
+	for (i=0; i<num; ++i)
+		page_cache_release(page[i]);
+err_out_free:
+	kfree(page);
+err_out_fput:
+	fput_light(file, fput_needed);
+
+	return err;
+}
+
+static int kevent_naio_dequeue(struct kevent *k)
+{
+	int err, i, num;
+	struct page **page = k->priv;
+
+	num = k->event.ret_data[0];
+
+	err = kevent_socket_dequeue(k);
+
+	for (i=0; i<num; ++i)
+		page_cache_release(page[i]);
+
+	kfree(k->priv);
+	k->priv = NULL;
+
+	return err;
+}
+
+static int kevent_naio_callback(struct kevent *k)
+{
+	struct inode *inode = k->st->origin;
+	struct sock *sk = SOCKET_I(inode)->sk;
+	unsigned int size = k->event.id.raw[1];
+	unsigned int off = k->event.ret_data[1];
+	struct page **pages = k->priv, *page;
+	int ready = 0, num = off/PAGE_SIZE, err = 0, send = 0;
+	void *ptr, *optr;
+	unsigned int len;
+
+	if (!sock_flag(sk, SOCK_ASYNC))
+		return -1;
+
+	if (k->event.event & KEVENT_SOCKET_SEND)
+		send = 1;
+	else if (!(k->event.event & KEVENT_SOCKET_RECV))
+		return -EINVAL;
+
+	/*
+	 * sk_prot->async_*() can return either number of bytes processed,
+	 * or negative error value, or zero if socket is closed.
+	 */
+
+	if (!send) {
+		page = pages[num];
+
+		optr = ptr = kmap_atomic(page, KM_IRQ0);
+		if (!ptr)
+			return -ENOMEM;
+
+		ptr += off % PAGE_SIZE;
+		len = min_t(unsigned int, PAGE_SIZE - (ptr - optr), size);
+
+		err = sk->sk_prot->async_recv(sk, ptr, len);
+
+		kunmap_atomic(optr, KM_IRQ0);
+	} else {
+		len = size;
+		err = sk->sk_prot->async_send(sk, pages, off, size);
+	}
+
+	if (err > 0) {
+		num++;
+		size -= err;
+		off += err;
+	}
+
+	k->event.ret_data[1] = off;
+	k->event.id.raw[1] = size;
+
+	if (err == 0 || (err < 0 && err != -EAGAIN))
+		ready = -1;
+
+	if (!size)
+		ready = 1;
+#if 0
+	printk("%s: sk=%p, k=%p, size=%4u, off=%4u, err=%3d, ready=%1d.\n",
+			__func__, sk, k, size, off, err, ready);
+#endif
+
+	return ready;
+}
+
+int kevent_init_naio(struct kevent *k)
+{
+	k->enqueue = &kevent_naio_enqueue;
+	k->dequeue = &kevent_naio_dequeue;
+	k->callback = &kevent_naio_callback;
+	return 0;
+}
diff --git a/kernel/kevent/kevent_socket.c b/kernel/kevent/kevent_socket.c
new file mode 100644
index 0000000..c230aaa
--- /dev/null
+++ b/kernel/kevent/kevent_socket.c
@@ -0,0 +1,125 @@
+/*
+ * 	kevent_socket.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/timer.h>
+#include <linux/file.h>
+#include <linux/tcp.h>
+#include <linux/kevent.h>
+
+#include <net/sock.h>
+#include <net/request_sock.h>
+#include <net/inet_connection_sock.h>
+
+static int kevent_socket_callback(struct kevent *k)
+{
+	struct inode *inode = k->st->origin;
+	struct sock *sk = SOCKET_I(inode)->sk;
+	int rmem;
+	
+	if (k->event.event & KEVENT_SOCKET_RECV) {
+		int ret = 0;
+		
+		if ((rmem = atomic_read(&sk->sk_rmem_alloc)) > 0 || 
+				!skb_queue_empty(&sk->sk_receive_queue))
+			ret = 1;
+		if (sk->sk_shutdown & RCV_SHUTDOWN)
+			ret = 1;
+		if (ret)
+			return ret;
+	}
+	if ((k->event.event & KEVENT_SOCKET_ACCEPT) && 
+		(!reqsk_queue_empty(&inet_csk(sk)->icsk_accept_queue) || 
+		 	reqsk_queue_len_young(&inet_csk(sk)->icsk_accept_queue))) {
+		k->event.ret_data[1] = reqsk_queue_len(&inet_csk(sk)->icsk_accept_queue);
+		return 1;
+	}
+
+	return 0;
+}
+
+int kevent_socket_enqueue(struct kevent *k)
+{
+	struct file *file;
+	struct inode *inode;
+	int err, fput_needed;
+
+	file = fget_light(k->event.id.raw[0], &fput_needed);
+	if (!file)
+		return -ENODEV;
+
+	err = -EINVAL;
+	if (!file->f_dentry || !file->f_dentry->d_inode)
+		goto err_out_fput;
+
+	inode = igrab(file->f_dentry->d_inode);
+	if (!inode)
+		goto err_out_fput;
+
+	err = kevent_storage_enqueue(&inode->st, k);
+	if (err)
+		goto err_out_iput;
+
+	err = k->callback(k);
+	if (err)
+		goto err_out_dequeue;
+
+	fput_light(file, fput_needed);
+	return err;
+
+err_out_dequeue:
+	kevent_storage_dequeue(k->st, k);
+err_out_iput:
+	iput(inode);
+err_out_fput:
+	fput_light(file, fput_needed);
+	return err;
+}
+
+int kevent_socket_dequeue(struct kevent *k)
+{
+	struct inode *inode = k->st->origin;
+
+	kevent_storage_dequeue(k->st, k);
+	iput(inode);
+
+	return 0;
+}
+
+int kevent_init_socket(struct kevent *k)
+{
+	k->enqueue = &kevent_socket_enqueue;
+	k->dequeue = &kevent_socket_dequeue;
+	k->callback = &kevent_socket_callback;
+	return 0;
+}
+
+void kevent_socket_notify(struct sock *sk, u32 event)
+{
+	if (sk->sk_socket && !test_and_set_bit(SOCK_ASYNC_INUSE, &sk->sk_flags)) {
+		kevent_storage_ready(&SOCK_INODE(sk->sk_socket)->st, NULL, event);
+		sock_reset_flag(sk, SOCK_ASYNC_INUSE);
+	}
+}
diff --git a/net/core/datagram.c b/net/core/datagram.c
index aecddcc..493245b 100644
--- a/net/core/datagram.c
+++ b/net/core/datagram.c
@@ -236,6 +236,60 @@ void skb_kill_datagram(struct sock *sk, 
 EXPORT_SYMBOL(skb_kill_datagram);
 
 /**
+ *	skb_copy_datagram - Copy a datagram.
+ *	@skb: buffer to copy
+ *	@offset: offset in the buffer to start copying from
+ *	@to: pointer to copy to
+ *	@len: amount of data to copy from buffer to iovec
+ */
+int skb_copy_datagram(const struct sk_buff *skb, int offset,
+			    void *to, int len)
+{
+	int i, fraglen, end = 0;
+	struct sk_buff *next = skb_shinfo(skb)->frag_list;
+
+	if (!len)
+		return 0;
+
+next_skb:
+	fraglen = skb_headlen(skb);
+	i = -1;
+
+	while (1) {
+		int start = end;
+
+		if ((end += fraglen) > offset) {
+			int copy = end - offset, o = offset - start;
+
+			if (copy > len)
+				copy = len;
+			if (i == -1)
+				memcpy(to, skb->data + o, copy);
+			else {
+				skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
+				struct page *page = frag->page;
+				void *p = kmap(page) + frag->page_offset + o;
+				memcpy(to, p, copy);
+				kunmap(page);
+			}
+			if (!(len -= copy))
+				return 0;
+			offset += copy;
+		}
+		if (++i >= skb_shinfo(skb)->nr_frags)
+			break;
+		fraglen = skb_shinfo(skb)->frags[i].size;
+	}
+	if (next) {
+		skb = next;
+		BUG_ON(skb_shinfo(skb)->frag_list);
+		next = skb->next;
+		goto next_skb;
+	}
+	return -EFAULT;
+}
+
+/**
  *	skb_copy_datagram_iovec - Copy a datagram to an iovec.
  *	@skb: buffer to copy
  *	@offset: offset in the buffer to start copying from
@@ -530,6 +584,7 @@ unsigned int datagram_poll(struct file *
 
 EXPORT_SYMBOL(datagram_poll);
 EXPORT_SYMBOL(skb_copy_and_csum_datagram_iovec);
+EXPORT_SYMBOL(skb_copy_datagram);
 EXPORT_SYMBOL(skb_copy_datagram_iovec);
 EXPORT_SYMBOL(skb_free_datagram);
 EXPORT_SYMBOL(skb_recv_datagram);
diff --git a/net/core/sock.c b/net/core/sock.c
index 5d820c3..3345048 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -564,6 +564,16 @@ #endif
 			spin_unlock_bh(&sk->sk_lock.slock);
 			ret = -ENONET;
 			break;
+#ifdef CONFIG_KEVENT_SOCKET
+		case SO_ASYNC_SOCK:
+			spin_lock_bh(&sk->sk_lock.slock);
+			if (valbool)
+				sock_set_flag(sk, SOCK_ASYNC);
+			else
+				sock_reset_flag(sk, SOCK_ASYNC);
+			spin_unlock_bh(&sk->sk_lock.slock);
+			break;
+#endif
 
 		/* We implement the SO_SNDLOWAT etc to
 		   not be settable (1003.1g 5.3) */
@@ -1313,6 +1323,7 @@ static void sock_def_wakeup(struct sock 
 	if (sk->sk_sleep && waitqueue_active(sk->sk_sleep))
 		wake_up_interruptible_all(sk->sk_sleep);
 	read_unlock(&sk->sk_callback_lock);
+	kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
 }
 
 static void sock_def_error_report(struct sock *sk)
@@ -1322,6 +1333,7 @@ static void sock_def_error_report(struct
 		wake_up_interruptible(sk->sk_sleep);
 	sk_wake_async(sk,0,POLL_ERR); 
 	read_unlock(&sk->sk_callback_lock);
+	kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
 }
 
 static void sock_def_readable(struct sock *sk, int len)
@@ -1331,6 +1343,7 @@ static void sock_def_readable(struct soc
 		wake_up_interruptible(sk->sk_sleep);
 	sk_wake_async(sk,1,POLL_IN);
 	read_unlock(&sk->sk_callback_lock);
+	kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
 }
 
 static void sock_def_write_space(struct sock *sk)
@@ -1350,6 +1363,7 @@ static void sock_def_write_space(struct 
 	}
 
 	read_unlock(&sk->sk_callback_lock);
+	kevent_socket_notify(sk, KEVENT_SOCKET_SEND|KEVENT_SOCKET_RECV);
 }
 
 static void sock_def_destruct(struct sock *sk)
@@ -1454,8 +1468,10 @@ void fastcall release_sock(struct sock *
 	if (sk->sk_backlog.tail)
 		__release_sock(sk);
 	sk->sk_lock.owner = NULL;
-        if (waitqueue_active(&(sk->sk_lock.wq)))
+        if (waitqueue_active(&(sk->sk_lock.wq))) {
 		wake_up(&(sk->sk_lock.wq));
+		kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
+	}
 	spin_unlock_bh(&(sk->sk_lock.slock));
 }
 EXPORT_SYMBOL(release_sock);
diff --git a/net/core/stream.c b/net/core/stream.c
index e948969..91e2e07 100644
--- a/net/core/stream.c
+++ b/net/core/stream.c
@@ -36,6 +36,7 @@ void sk_stream_write_space(struct sock *
 			wake_up_interruptible(sk->sk_sleep);
 		if (sock->fasync_list && !(sk->sk_shutdown & SEND_SHUTDOWN))
 			sock_wake_async(sock, 2, POLL_OUT);
+		kevent_socket_notify(sk, KEVENT_SOCKET_SEND|KEVENT_SOCKET_RECV);
 	}
 }
 
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 74998f2..403d33e 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -206,6 +206,7 @@
  *					lingertime == 0 (RFC 793 ABORT Call)
  *	Hirokazu Takahashi	:	Use copy_from_user() instead of
  *					csum_and_copy_from_user() if possible.
+ *	Evgeniy Polyakov	:	Network asynchronous IO.
  *
  *		This program is free software; you can redistribute it and/or
  *		modify it under the terms of the GNU General Public License
@@ -1085,6 +1086,275 @@ int tcp_read_sock(struct sock *sk, read_
 }
 
 /*
+ * Must be called with locked sock.
+ */
+int tcp_async_send(struct sock *sk, struct page **pages, unsigned int poffset, size_t len)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+	int mss_now, size_goal;
+	int err = -EAGAIN;
+	ssize_t copied;
+
+	/* Wait for a connection to finish. */
+	if ((1 << sk->sk_state) & ~(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT))
+		goto out_err;
+
+	clear_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags);
+
+	mss_now = tcp_current_mss(sk, 1);
+	size_goal = tp->xmit_size_goal;
+	copied = 0;
+
+	err = -EPIPE;
+	if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN) || sock_flag(sk, SOCK_DONE) ||
+			(sk->sk_state == TCP_CLOSE) || (atomic_read(&sk->sk_refcnt) == 1))
+		goto do_error;
+
+	while (len > 0) {
+		struct sk_buff *skb = sk->sk_write_queue.prev;
+		struct page *page = pages[poffset / PAGE_SIZE];
+		int copy, i, can_coalesce;
+		int offset = poffset % PAGE_SIZE;
+		int size = min_t(size_t, len, PAGE_SIZE - offset);
+
+		if (!sk->sk_send_head || (copy = size_goal - skb->len) <= 0) {
+new_segment:
+			if (!sk_stream_memory_free(sk))
+				goto wait_for_sndbuf;
+
+			skb = sk_stream_alloc_pskb(sk, 0, 0,
+						   sk->sk_allocation);
+			if (!skb)
+				goto wait_for_memory;
+
+			skb_entail(sk, tp, skb);
+			copy = size_goal;
+		}
+
+		if (copy > size)
+			copy = size;
+
+		i = skb_shinfo(skb)->nr_frags;
+		can_coalesce = skb_can_coalesce(skb, i, page, offset);
+		if (!can_coalesce && i >= MAX_SKB_FRAGS) {
+			tcp_mark_push(tp, skb);
+			goto new_segment;
+		}
+		if (!sk_stream_wmem_schedule(sk, copy))
+			goto wait_for_memory;
+		
+		if (can_coalesce) {
+			skb_shinfo(skb)->frags[i - 1].size += copy;
+		} else {
+			get_page(page);
+			skb_fill_page_desc(skb, i, page, offset, copy);
+		}
+
+		skb->len += copy;
+		skb->data_len += copy;
+		skb->truesize += copy;
+		sk->sk_wmem_queued += copy;
+		sk->sk_forward_alloc -= copy;
+		skb->ip_summed = CHECKSUM_HW;
+		tp->write_seq += copy;
+		TCP_SKB_CB(skb)->end_seq += copy;
+		skb_shinfo(skb)->tso_segs = 0;
+
+		if (!copied)
+			TCP_SKB_CB(skb)->flags &= ~TCPCB_FLAG_PSH;
+
+		copied += copy;
+		poffset += copy;
+		if (!(len -= copy))
+			goto out;
+
+		if (skb->len < mss_now)
+			continue;
+
+		if (forced_push(tp)) {
+			tcp_mark_push(tp, skb);
+			__tcp_push_pending_frames(sk, tp, mss_now, TCP_NAGLE_PUSH);
+		} else if (skb == sk->sk_send_head)
+			tcp_push_one(sk, mss_now);
+		continue;
+
+wait_for_sndbuf:
+		set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
+wait_for_memory:
+		if (copied)
+			tcp_push(sk, tp, 0, mss_now, TCP_NAGLE_PUSH);
+
+		err = -EAGAIN;
+		goto do_error;
+	}
+
+out:
+	if (copied)
+		tcp_push(sk, tp, 0, mss_now, tp->nonagle);
+	return copied;
+
+do_error:
+	if (copied)
+		goto out;
+out_err:
+	return sk_stream_error(sk, 0, err);
+}
+
+/*
+ * Must be called with locked sock.
+ */
+int tcp_async_recv(struct sock *sk, void *dst, size_t len)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+	int copied = 0;
+	u32 *seq;
+	unsigned long used;
+	int err;
+	int target;		/* Read at least this many bytes */
+
+	TCP_CHECK_TIMER(sk);
+
+	err = -ENOTCONN;
+	if (sk->sk_state == TCP_LISTEN)
+		goto out;
+
+	seq = &tp->copied_seq;
+
+	target = sock_rcvlowat(sk, 0, len);
+
+	do {
+		struct sk_buff *skb;
+		u32 offset;
+
+		/* Are we at urgent data? Stop if we have read anything or have SIGURG pending. */
+		if (tp->urg_data && tp->urg_seq == *seq) {
+			if (copied)
+				break;
+		}
+
+		/* Next get a buffer. */
+
+		skb = skb_peek(&sk->sk_receive_queue);
+		do {
+			if (!skb)
+				break;
+
+			/* Now that we have two receive queues this
+			 * shouldn't happen.
+			 */
+			if (before(*seq, TCP_SKB_CB(skb)->seq)) {
+				printk(KERN_INFO "async_recv bug: copied %X "
+				       "seq %X\n", *seq, TCP_SKB_CB(skb)->seq);
+				break;
+			}
+			offset = *seq - TCP_SKB_CB(skb)->seq;
+			if (skb->h.th->syn)
+				offset--;
+			if (offset < skb->len)
+				goto found_ok_skb;
+			if (skb->h.th->fin)
+				goto found_fin_ok;
+			skb = skb->next;
+		} while (skb != (struct sk_buff *)&sk->sk_receive_queue);
+
+		if (copied)
+			break;
+
+		if (sock_flag(sk, SOCK_DONE))
+			break;
+
+		if (sk->sk_err) {
+			copied = sock_error(sk);
+			break;
+		}
+
+		if (sk->sk_shutdown & RCV_SHUTDOWN)
+			break;
+
+		if (sk->sk_state == TCP_CLOSE) {
+			if (!sock_flag(sk, SOCK_DONE)) {
+				/* This occurs when user tries to read
+				 * from never connected socket.
+				 */
+				copied = -ENOTCONN;
+				break;
+			}
+			break;
+		}
+
+		copied = -EAGAIN;
+		break;
+
+	found_ok_skb:
+		/* Ok so how much can we use? */
+		used = skb->len - offset;
+		if (len < used)
+			used = len;
+
+		/* Do we have urgent data here? */
+		if (tp->urg_data) {
+			u32 urg_offset = tp->urg_seq - *seq;
+			if (urg_offset < used) {
+				if (!urg_offset) {
+					if (!sock_flag(sk, SOCK_URGINLINE)) {
+						++*seq;
+						offset++;
+						used--;
+						if (!used)
+							goto skip_copy;
+					}
+				} else
+					used = urg_offset;
+			}
+		}
+
+		err = skb_copy_datagram(skb, offset, dst, used);
+		if (err) {
+			/* Exception. Bailout! */
+			if (!copied)
+				copied = -EFAULT;
+			break;
+		}
+
+		*seq += used;
+		copied += used;
+		len -= used;
+		dst += used;
+
+		tcp_rcv_space_adjust(sk);
+
+skip_copy:
+		if (tp->urg_data && after(tp->copied_seq, tp->urg_seq)) {
+			tp->urg_data = 0;
+			tcp_fast_path_check(sk, tp);
+		}
+		if (used + offset < skb->len)
+			continue;
+
+		if (skb->h.th->fin)
+			goto found_fin_ok;
+		sk_eat_skb(sk, skb);
+		continue;
+
+	found_fin_ok:
+		/* Process the FIN. */
+		++*seq;
+		sk_eat_skb(sk, skb);
+		break;
+	} while (len > 0);
+
+	/* Clean up data we have read: This will do ACK frames. */
+	cleanup_rbuf(sk, copied);
+
+	TCP_CHECK_TIMER(sk);
+	return copied;
+
+out:
+	TCP_CHECK_TIMER(sk);
+	return err;
+}
+
+/*
  *	This routine copies from a sock struct into the user buffer.
  *
  *	Technical note: in 2.3 we work on _locked_ socket, so that
@@ -2259,6 +2529,8 @@ EXPORT_SYMBOL(tcp_getsockopt);
 EXPORT_SYMBOL(tcp_ioctl);
 EXPORT_SYMBOL(tcp_poll);
 EXPORT_SYMBOL(tcp_read_sock);
+EXPORT_SYMBOL(tcp_async_recv);
+EXPORT_SYMBOL(tcp_async_send);
 EXPORT_SYMBOL(tcp_recvmsg);
 EXPORT_SYMBOL(tcp_sendmsg);
 EXPORT_SYMBOL(tcp_sendpage);
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index e08245b..5655b1e 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -3113,6 +3113,7 @@ static void tcp_ofo_queue(struct sock *s
 
 		__skb_unlink(skb, &tp->out_of_order_queue);
 		__skb_queue_tail(&sk->sk_receive_queue, skb);
+		kevent_socket_notify(sk, KEVENT_SOCKET_RECV);
 		tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq;
 		if(skb->h.th->fin)
 			tcp_fin(skb, sk, skb->h.th);
@@ -3956,7 +3957,8 @@ int tcp_rcv_established(struct sock *sk,
 			int copied_early = 0;
 
 			if (tp->copied_seq == tp->rcv_nxt &&
-			    len - tcp_header_len <= tp->ucopy.len) {
+			    len - tcp_header_len <= tp->ucopy.len &&
+			    !sock_async(sk)) {
 #ifdef CONFIG_NET_DMA
 				if (tcp_dma_try_early_copy(sk, skb, tcp_header_len)) {
 					copied_early = 1;
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 25ecc6e..05d7086 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -62,6 +62,7 @@ #include <linux/cache.h>
 #include <linux/jhash.h>
 #include <linux/init.h>
 #include <linux/times.h>
+#include <linux/kevent.h>
 
 #include <net/icmp.h>
 #include <net/inet_hashtables.h>
@@ -850,6 +851,7 @@ #endif
 	   	reqsk_free(req);
 	} else {
 		inet_csk_reqsk_queue_hash_add(sk, req, TCP_TIMEOUT_INIT);
+		kevent_socket_notify(sk, KEVENT_SOCKET_ACCEPT);
 	}
 	return 0;
 
@@ -1089,24 +1091,30 @@ process:
 
 	skb->dev = NULL;
 
-	bh_lock_sock(sk);
 	ret = 0;
-	if (!sock_owned_by_user(sk)) {
+	if (sock_async(sk)) {
+		spin_lock_bh(&sk->sk_lock.slock);
+		ret = tcp_v4_do_rcv(sk, skb);
+		spin_unlock_bh(&sk->sk_lock.slock);
+	} else {
+		bh_lock_sock(sk);
+		if (!sock_owned_by_user(sk)) {
 #ifdef CONFIG_NET_DMA
-		struct tcp_sock *tp = tcp_sk(sk);
-		if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
-			tp->ucopy.dma_chan = get_softnet_dma();
-		if (tp->ucopy.dma_chan)
-			ret = tcp_v4_do_rcv(sk, skb);
-		else
+			struct tcp_sock *tp = tcp_sk(sk);
+			if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
+				tp->ucopy.dma_chan = get_softnet_dma();
+			if (tp->ucopy.dma_chan)
+				ret = tcp_v4_do_rcv(sk, skb);
+			else
 #endif
-		{
-			if (!tcp_prequeue(sk, skb))
-			ret = tcp_v4_do_rcv(sk, skb);
-		}
-	} else
-		sk_add_backlog(sk, skb);
-	bh_unlock_sock(sk);
+			{
+				if (!tcp_prequeue(sk, skb))
+				ret = tcp_v4_do_rcv(sk, skb);
+			}
+		} else
+			sk_add_backlog(sk, skb);
+		bh_unlock_sock(sk);
+	}
 
 	sock_put(sk);
 
@@ -1830,6 +1838,8 @@ struct proto tcp_prot = {
 	.getsockopt		= tcp_getsockopt,
 	.sendmsg		= tcp_sendmsg,
 	.recvmsg		= tcp_recvmsg,
+	.async_recv		= tcp_async_recv,
+	.async_send		= tcp_async_send,
 	.backlog_rcv		= tcp_v4_do_rcv,
 	.hash			= tcp_v4_hash,
 	.unhash			= tcp_unhash,


-- 
	Evgeniy Polyakov

[3/4] kevent: fs/aio notifications.

[Evgeniy Polyakov]

This patch includes asynchronous propagation of file's data into VFS
cache and aio_sendfile() implementation.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>

diff --git a/fs/bio.c b/fs/bio.c
index 6a0b9ad..a3ee530 100644
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -119,7 +119,7 @@ void bio_free(struct bio *bio, struct bi
 /*
  * default destructor for a bio allocated with bio_alloc_bioset()
  */
-static void bio_fs_destructor(struct bio *bio)
+void bio_fs_destructor(struct bio *bio)
 {
 	bio_free(bio, fs_bio_set);
 }
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index 04af9c4..295fce9 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -685,6 +685,7 @@ ext2_writepages(struct address_space *ma
 }
 
 struct address_space_operations ext2_aops = {
+	.get_block		= ext2_get_block,
 	.readpage		= ext2_readpage,
 	.readpages		= ext2_readpages,
 	.writepage		= ext2_writepage,
diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c
index 2edd7ee..e44f5ad 100644
--- a/fs/ext3/inode.c
+++ b/fs/ext3/inode.c
@@ -1700,6 +1700,7 @@ static int ext3_journalled_set_page_dirt
 }
 
 static struct address_space_operations ext3_ordered_aops = {
+	.get_block	= ext3_get_block,
 	.readpage	= ext3_readpage,
 	.readpages	= ext3_readpages,
 	.writepage	= ext3_ordered_writepage,
diff --git a/fs/file_table.c b/fs/file_table.c
index bcea199..8759479 100644
--- a/fs/file_table.c
+++ b/fs/file_table.c
@@ -113,6 +113,9 @@ struct file *get_empty_filp(void)
 	if (security_file_alloc(f))
 		goto fail_sec;
 
+#ifdef CONFIG_KEVENT_POLL
+	kevent_storage_init(f, &f->st);
+#endif
 	tsk = current;
 	INIT_LIST_HEAD(&f->f_u.fu_list);
 	atomic_set(&f->f_count, 1);
@@ -160,6 +163,9 @@ void fastcall __fput(struct file *file)
 	might_sleep();
 
 	fsnotify_close(file);
+#ifdef CONFIG_KEVENT_POLL
+	kevent_storage_fini(&file->st);
+#endif
 	/*
 	 * The function eventpoll_release() should be the first called
 	 * in the file cleanup chain.
diff --git a/fs/inode.c b/fs/inode.c
index 3a2446a..0493935 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -22,6 +22,7 @@ #include <linux/pagemap.h>
 #include <linux/cdev.h>
 #include <linux/bootmem.h>
 #include <linux/inotify.h>
+#include <linux/kevent.h>
 #include <linux/mount.h>
 
 /*
@@ -166,12 +167,18 @@ #endif
 		}
 		memset(&inode->u, 0, sizeof(inode->u));
 		inode->i_mapping = mapping;
+#if defined CONFIG_KEVENT
+		kevent_storage_init(inode, &inode->st);
+#endif
 	}
 	return inode;
 }
 
 void destroy_inode(struct inode *inode) 
 {
+#if defined CONFIG_KEVENT_INODE || defined CONFIG_KEVENT_SOCKET
+	kevent_storage_fini(&inode->st);
+#endif
 	BUG_ON(inode_has_buffers(inode));
 	security_inode_free(inode);
 	if (inode->i_sb->s_op->destroy_inode)
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 9857e50..bbbb578 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -2997,6 +2997,7 @@ int reiserfs_setattr(struct dentry *dent
 }
 
 struct address_space_operations reiserfs_address_space_operations = {
+	.get_block = reiserfs_get_block,
 	.writepage = reiserfs_writepage,
 	.readpage = reiserfs_readpage,
 	.readpages = reiserfs_readpages,
diff --git a/include/linux/fs.h b/include/linux/fs.h
index ecc8c2c..248f6a1 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -236,6 +236,9 @@ #include <linux/mutex.h>
 #include <asm/atomic.h>
 #include <asm/semaphore.h>
 #include <asm/byteorder.h>
+#ifdef CONFIG_KEVENT
+#include <linux/kevent_storage.h>
+#endif
 
 struct hd_geometry;
 struct iovec;
@@ -348,6 +351,8 @@ struct address_space;
 struct writeback_control;
 
 struct address_space_operations {
+	int  (*get_block)(struct inode *inode, sector_t iblock,
+			struct buffer_head *bh_result, int create);
 	int (*writepage)(struct page *page, struct writeback_control *wbc);
 	int (*readpage)(struct file *, struct page *);
 	void (*sync_page)(struct page *);
@@ -526,6 +531,10 @@ #ifdef CONFIG_INOTIFY
 	struct mutex		inotify_mutex;	/* protects the watches list */
 #endif
 
+#ifdef CONFIG_KEVENT_INODE
+	struct kevent_storage	st;
+#endif
+
 	unsigned long		i_state;
 	unsigned long		dirtied_when;	/* jiffies of first dirtying */
 
@@ -659,6 +668,9 @@ #ifdef CONFIG_EPOLL
 	struct list_head	f_ep_links;
 	spinlock_t		f_ep_lock;
 #endif /* #ifdef CONFIG_EPOLL */
+#ifdef CONFIG_KEVENT_POLL
+	struct kevent_storage	st;
+#endif
 	struct address_space	*f_mapping;
 };
 extern spinlock_t files_lock;
diff --git a/include/linux/fsnotify.h b/include/linux/fsnotify.h
index cc5dec7..0acc8db 100644
--- a/include/linux/fsnotify.h
+++ b/include/linux/fsnotify.h
@@ -15,6 +15,7 @@ #ifdef __KERNEL__
 
 #include <linux/dnotify.h>
 #include <linux/inotify.h>
+#include <linux/kevent.h>
 #include <linux/audit.h>
 
 /*
@@ -79,6 +80,7 @@ static inline void fsnotify_nameremove(s
 		isdir = IN_ISDIR;
 	dnotify_parent(dentry, DN_DELETE);
 	inotify_dentry_parent_queue_event(dentry, IN_DELETE|isdir, 0, dentry->d_name.name);
+	kevent_inode_notify_parent(dentry, KEVENT_INODE_REMOVE);
 }
 
 /*
@@ -88,6 +90,7 @@ static inline void fsnotify_inoderemove(
 {
 	inotify_inode_queue_event(inode, IN_DELETE_SELF, 0, NULL, NULL);
 	inotify_inode_is_dead(inode);
+	kevent_inode_remove(inode);
 }
 
 /*
@@ -96,6 +99,7 @@ static inline void fsnotify_inoderemove(
 static inline void fsnotify_create(struct inode *inode, struct dentry *dentry)
 {
 	inode_dir_notify(inode, DN_CREATE);
+	kevent_inode_notify(inode, KEVENT_INODE_CREATE);
 	inotify_inode_queue_event(inode, IN_CREATE, 0, dentry->d_name.name,
 				  dentry->d_inode);
 	audit_inode_child(dentry->d_name.name, dentry->d_inode, inode->i_ino);
@@ -107,6 +111,7 @@ static inline void fsnotify_create(struc
 static inline void fsnotify_mkdir(struct inode *inode, struct dentry *dentry)
 {
 	inode_dir_notify(inode, DN_CREATE);
+	kevent_inode_notify(inode, KEVENT_INODE_CREATE);
 	inotify_inode_queue_event(inode, IN_CREATE | IN_ISDIR, 0, 
 				  dentry->d_name.name, dentry->d_inode);
 	audit_inode_child(dentry->d_name.name, dentry->d_inode, inode->i_ino);
diff --git a/kernel/kevent/kevent_aio.c b/kernel/kevent/kevent_aio.c
new file mode 100644
index 0000000..d4132a3
--- /dev/null
+++ b/kernel/kevent/kevent_aio.c
@@ -0,0 +1,580 @@
+/*
+ * 	kevent_aio.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/file.h>
+#include <linux/fs.h>
+#include <linux/swap.h>
+#include <linux/pagemap.h>
+#include <linux/bio.h>
+#include <linux/buffer_head.h>
+#include <linux/kevent.h>
+
+#include <net/sock.h>
+
+#define KEVENT_AIO_DEBUG
+
+#ifdef KEVENT_AIO_DEBUG
+#define dprintk(f, a...) printk(f, ##a)
+#else
+#define dprintk(f, a...) do {} while (0)
+#endif
+
+struct kevent_aio_private
+{
+	int			pg_num;
+	size_t			size;
+	loff_t			offset;
+	loff_t			processed;
+	atomic_t		bio_page_num;
+	struct completion	bio_complete;
+	struct file		*file, *sock;
+	struct work_struct	work;
+};
+
+static int kevent_aio_dequeue(struct kevent *k);
+static int kevent_aio_enqueue(struct kevent *k);
+static int kevent_aio_callback(struct kevent *k);
+
+extern void bio_fs_destructor(struct bio *bio);
+
+static void kevent_aio_bio_destructor(struct bio *bio)
+{
+	struct kevent *k = bio->bi_private;
+	struct kevent_aio_private *priv = k->priv;
+
+	dprintk("%s: bio=%p, num=%u, k=%p, inode=%p.\n", __func__, bio, bio->bi_vcnt, k, k->st->origin);
+	schedule_work(&priv->work);
+	bio_fs_destructor(bio);
+}
+
+static void kevent_aio_bio_put(struct kevent *k)
+{
+	struct kevent_aio_private *priv = k->priv;
+	
+	if (atomic_dec_and_test(&priv->bio_page_num))
+		complete(&priv->bio_complete);
+}
+
+static int kevent_mpage_end_io_read(struct bio *bio, unsigned int bytes_done, int err)
+{
+	const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
+	struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1;
+	struct kevent *k = bio->bi_private;
+
+	if (bio->bi_size)
+		return 1;
+
+	do {
+		struct page *page = bvec->bv_page;
+
+		if (--bvec >= bio->bi_io_vec)
+			prefetchw(&bvec->bv_page->flags);
+
+		if (uptodate) {
+			SetPageUptodate(page);
+		} else {
+			ClearPageUptodate(page);
+			SetPageError(page);
+		}
+
+		unlock_page(page);
+		kevent_aio_bio_put(k);
+	} while (bvec >= bio->bi_io_vec);
+
+	bio_put(bio);
+	return 0;
+}
+
+static inline struct bio *kevent_mpage_bio_submit(int rw, struct bio *bio)
+{
+	if (bio) {
+		bio->bi_end_io = kevent_mpage_end_io_read;
+		dprintk("%s: bio=%p, num=%u.\n", __func__, bio, bio->bi_vcnt);
+		submit_bio(READ, bio);
+	}
+	return NULL;
+}
+
+static struct bio *kevent_mpage_readpage(struct kevent *k, struct bio *bio,
+		struct page *page, unsigned nr_pages, get_block_t get_block, 
+		loff_t *offset, sector_t *last_block_in_bio)
+{
+	struct inode *inode = k->st->origin;
+	const unsigned blkbits = inode->i_blkbits;
+	const unsigned blocks_per_page = PAGE_CACHE_SIZE >> blkbits;
+	const unsigned blocksize = 1 << blkbits;
+	sector_t block_in_file;
+	sector_t last_block;
+	struct block_device *bdev = NULL;
+	unsigned first_hole = blocks_per_page;
+	unsigned page_block;
+	sector_t blocks[MAX_BUF_PER_PAGE];
+	struct buffer_head bh;
+	int fully_mapped = 1, length;
+
+	block_in_file = (*offset + blocksize - 1) >> blkbits;
+	last_block = (i_size_read(inode) + blocksize - 1) >> blkbits;
+
+	bh.b_page = page;
+	for (page_block = 0; page_block < blocks_per_page; page_block++, block_in_file++) {
+		bh.b_state = 0;
+		if (block_in_file < last_block) {
+			if (get_block(inode, block_in_file, &bh, 0))
+				goto confused;
+		}
+
+		if (!buffer_mapped(&bh)) {
+			fully_mapped = 0;
+			if (first_hole == blocks_per_page)
+				first_hole = page_block;
+			continue;
+		}
+
+		/* some filesystems will copy data into the page during
+		 * the get_block call, in which case we don't want to
+		 * read it again.  map_buffer_to_page copies the data
+		 * we just collected from get_block into the page's buffers
+		 * so readpage doesn't have to repeat the get_block call
+		 */
+		if (buffer_uptodate(&bh)) {
+			BUG();
+			//map_buffer_to_page(page, &bh, page_block);
+			goto confused;
+		}
+	
+		if (first_hole != blocks_per_page)
+			goto confused;		/* hole -> non-hole */
+
+		/* Contiguous blocks? */
+		if (page_block && blocks[page_block-1] != bh.b_blocknr-1)
+			goto confused;
+		blocks[page_block] = bh.b_blocknr;
+		bdev = bh.b_bdev;
+	}
+
+	if (!bdev)
+		goto confused;
+
+	if (first_hole != blocks_per_page) {
+		char *kaddr = kmap_atomic(page, KM_USER0);
+		memset(kaddr + (first_hole << blkbits), 0,
+				PAGE_CACHE_SIZE - (first_hole << blkbits));
+		flush_dcache_page(page);
+		kunmap_atomic(kaddr, KM_USER0);
+		if (first_hole == 0) {
+			SetPageUptodate(page);
+			goto out;
+		}
+	} else if (fully_mapped) {
+		SetPageMappedToDisk(page);
+	}
+	
+	/*
+	 * This page will go to BIO.  Do we need to send this BIO off first?
+	 */
+	if (bio && (*last_block_in_bio != blocks[0] - 1))
+		bio = kevent_mpage_bio_submit(READ, bio);
+
+alloc_new:
+	if (bio == NULL) {
+		nr_pages = min_t(unsigned, nr_pages, bio_get_nr_vecs(bdev));
+		bio = bio_alloc(GFP_KERNEL, nr_pages);
+		if (bio == NULL)
+			goto confused;
+
+		bio->bi_destructor = kevent_aio_bio_destructor;
+		bio->bi_bdev = bdev;
+		bio->bi_sector = blocks[0] << (blkbits - 9);
+		bio->bi_private = k;
+	}
+
+	length = first_hole << blkbits;
+	if (bio_add_page(bio, page, length, 0) < length) {
+		bio = kevent_mpage_bio_submit(READ, bio);
+		dprintk("%s: Failed to add a page: nr_pages=%d, length=%d, page=%p.\n", 
+				__func__, nr_pages, length, page);
+		goto alloc_new;
+	}
+	
+	dprintk("%s: bio=%p, b=%d, m=%d, u=%d, nr_pages=%d, offset=%Lu, "
+			"size=%Lu. page_block=%u, page=%p.\n", 
+			__func__, bio, buffer_boundary(&bh), buffer_mapped(&bh), 
+			buffer_uptodate(&bh), nr_pages, *offset, i_size_read(inode), 
+			page_block, page);
+	
+	*offset = *offset + length;
+
+	if (buffer_boundary(&bh) || (first_hole != blocks_per_page))
+		bio = kevent_mpage_bio_submit(READ, bio);
+	else
+		*last_block_in_bio = blocks[blocks_per_page - 1];
+
+out:
+	return bio;
+
+confused:
+	dprintk("%s: confused. bio=%p, nr_pages=%d.\n", __func__, bio, nr_pages);
+	if (bio)
+		bio = kevent_mpage_bio_submit(READ, bio);
+	kevent_aio_bio_put(k);
+	SetPageUptodate(page);
+
+	if (nr_pages == 1) {
+		struct kevent_aio_private *priv = k->priv;
+
+		wait_for_completion(&priv->bio_complete);
+		kevent_storage_ready(k->st, NULL, KEVENT_AIO_BIO);
+		init_completion(&priv->bio_complete);
+		complete(&priv->bio_complete);
+	}
+	goto out;
+}
+
+static int kevent_aio_alloc_cached_page(struct kevent *k, struct page **cached_page)
+{
+	struct kevent_aio_private *priv = k->priv;
+	struct address_space *mapping = priv->file->f_mapping;
+	struct page *page;
+	int err = 0;
+	pgoff_t index = priv->offset >> PAGE_CACHE_SHIFT;
+
+	page = page_cache_alloc_cold(mapping);
+	if (!page) {
+		err = -ENOMEM;
+		goto out;
+	}
+
+	err = add_to_page_cache_lru(page, mapping, index, GFP_KERNEL);
+	if (err) {
+		if (err == -EEXIST)
+			err = 0;
+		page_cache_release(page);
+		goto out;
+	}
+
+	dprintk("%s: page=%p, offset=%Lu, processed=%Lu, index=%lu, size=%zu.\n",
+			__func__, page, priv->offset, priv->processed, index, priv->size);
+
+	*cached_page = page;
+
+out:
+	return err;
+}
+
+static int kevent_mpage_readpages(struct kevent *k, int first,
+		int (* get_block)(struct inode *inode, sector_t iblock,	
+			struct buffer_head *bh_result, int create))
+{
+	struct bio *bio = NULL;
+	struct kevent_aio_private *priv = k->priv;
+	sector_t last_block_in_bio = 0;
+	int i, err = 0;
+
+	atomic_set(&priv->bio_page_num, priv->pg_num);
+
+	for (i=first; i<priv->pg_num; ++i) {
+		struct page *page;
+		
+		err = kevent_aio_alloc_cached_page(k, &page);
+		if (err)
+			break;
+
+		/*
+		 * If there is no error and page is NULL, this means
+		 * that someone added a page into VFS cache.
+		 * We will not process this page, since it is that who
+		 * added a page must read data from disk.
+		 */
+		if (!page)
+			continue;
+
+		bio = kevent_mpage_readpage(k, bio, page, priv->pg_num - i, 
+				get_block, &priv->offset, &last_block_in_bio);
+	}
+
+	if (bio)
+		bio = kevent_mpage_bio_submit(READ, bio);
+
+	return err;
+}
+
+static size_t kevent_aio_vfs_read_actor(struct kevent *k, struct page *kpage, size_t len)
+{
+	struct kevent_aio_private *priv = k->priv;
+	size_t ret;
+	
+	ret = priv->sock->f_op->sendpage(priv->sock, kpage, 0, len, &priv->sock->f_pos, 1);
+
+	dprintk("%s: k=%p, page=%p, len=%zu, ret=%zd.\n", 
+			__func__, k, kpage, len, ret);
+
+	return ret;
+}
+
+static int kevent_aio_vfs_read(struct kevent *k, 
+		size_t (*actor)(struct kevent *, struct page *, size_t))
+{
+	struct kevent_aio_private *priv = k->priv;
+	struct address_space *mapping;
+	size_t isize, actor_size;
+	int i;
+
+	mapping = priv->file->f_mapping;
+	isize = i_size_read(priv->file->f_dentry->d_inode);
+	
+	dprintk("%s: start: size_left=%zd, offset=%Lu, processed=%Lu, isize=%zu, pg_num=%d.\n", 
+			__func__, priv->size, priv->offset, priv->processed, isize, priv->pg_num);
+
+	for (i=0; i<priv->pg_num && priv->size; ++i) {
+		struct page *page;
+		size_t nr = PAGE_CACHE_SIZE;
+
+		cond_resched();
+		page = find_get_page(mapping, priv->processed >> PAGE_CACHE_SHIFT);
+		if (unlikely(page == NULL))
+			break;
+		if (!PageUptodate(page)) {
+			dprintk("%s: %2d: page=%p, processed=%Lu, size=%zu not uptodate.\n", 
+					__func__, i, page, priv->processed, priv->size);
+			page_cache_release(page);
+			break;
+		}
+
+		if (mapping_writably_mapped(mapping))
+			flush_dcache_page(page);
+
+		mark_page_accessed(page);
+
+		if (nr + priv->processed > isize)
+			nr = isize - priv->processed;
+		if (nr > priv->size)
+			nr = priv->size;
+
+		actor_size = actor(k, page, nr);
+		if (actor_size < 0) {
+			page_cache_release(page);
+			break;
+		}
+
+		page_cache_release(page);
+
+		priv->processed += actor_size;
+		priv->size -= actor_size;
+	}
+
+	if (!priv->size)
+		i = priv->pg_num;
+
+	if (i != priv->pg_num)
+		priv->offset = priv->processed;
+
+	dprintk("%s: end: next=%d, num=%d, left=%zu, offset=%Lu, procesed=%Lu, ret=%d.\n", 
+			__func__, i, priv->pg_num, 
+			priv->size, priv->offset, priv->processed, i);
+
+	return i;
+}
+
+static int kevent_aio_callback(struct kevent *k)
+{
+	return 1;
+}
+
+static void kevent_aio_work(void *data)
+{
+	struct kevent *k = data;
+	struct kevent_aio_private *priv = k->priv;
+	struct inode *inode = k->st->origin;
+	struct address_space *mapping = priv->file->f_mapping;
+	int err, ready = 0, num;
+
+	dprintk("%s: k=%p, priv=%p, inode=%p.\n", __func__, k, priv, inode);
+
+	init_completion(&priv->bio_complete);
+	
+	num = ready = kevent_aio_vfs_read(k, &kevent_aio_vfs_read_actor);
+	if (ready > 0 && ready != priv->pg_num)
+		ready = 0;
+
+	dprintk("%s: k=%p, ready=%d, size=%zd.\n", __func__, k, ready, priv->size);
+
+	if (!ready) {
+		err = kevent_mpage_readpages(k, num, mapping->a_ops->get_block);
+		if (err) {
+			dprintk("%s: kevent_mpage_readpages failed: err=%d, k=%p, size=%zd.\n",
+					__func__, err, k, priv->size);
+			kevent_break(k);
+			kevent_storage_ready(k->st, NULL, KEVENT_MASK_ALL);
+		}
+	} else {
+		dprintk("%s: next k=%p, size=%zd.\n", __func__, k, priv->size);
+
+		if (priv->size)
+			schedule_work(&priv->work);
+		else {
+			kevent_storage_ready(k->st, NULL, KEVENT_MASK_ALL);
+		}
+
+		complete(&priv->bio_complete);
+	}
+}
+
+static int kevent_aio_enqueue(struct kevent *k)
+{
+	int err;
+	struct file *file, *sock;
+	struct inode *inode;
+	struct kevent_aio_private *priv;
+	struct address_space *mapping;
+	int fd = k->event.id.raw[0];
+	int num = k->event.id.raw[1];
+	int s = k->event.ret_data[0];
+	size_t size;
+
+	err = -ENODEV;
+	file = fget(fd);
+	if (!file)
+		goto err_out_exit;
+	
+	sock = fget(s);
+	if (!sock)
+		goto err_out_fput_file;
+	
+	mapping = file->f_mapping;
+
+	err = -EINVAL;
+	if (!file->f_dentry || !file->f_dentry->d_inode || !mapping->a_ops->get_block)
+		goto err_out_fput;
+	if (!sock->f_dentry || !sock->f_dentry->d_inode)
+		goto err_out_fput;
+
+	inode = igrab(file->f_dentry->d_inode);
+	if (!inode)
+		goto err_out_fput;
+
+	size = i_size_read(inode);
+	
+	num = (size > num << PAGE_SHIFT) ? num : (size >> PAGE_SHIFT);
+
+	err = -ENOMEM;
+	priv = kzalloc(sizeof(struct kevent_aio_private), GFP_KERNEL);
+	if (!priv)
+		goto err_out_iput;
+
+	priv->pg_num = num;
+	priv->size = size;
+	priv->offset = 0;
+	priv->file = file;
+	priv->sock = sock;
+	INIT_WORK(&priv->work, kevent_aio_work, k);
+	k->priv = priv;
+
+	dprintk("%s: read: k=%p, priv=%p, inode=%p, num=%u, size=%zu, off=%Lu.\n", 
+			__func__, k, priv, inode, priv->pg_num, priv->size, priv->offset);
+	
+	init_completion(&priv->bio_complete);
+	kevent_storage_enqueue(&inode->st, k);
+	schedule_work(&priv->work);
+	
+	return 0;
+
+err_out_iput:
+	iput(inode);
+err_out_fput:
+	fput(sock);
+err_out_fput_file:
+	fput(file);
+err_out_exit:
+
+	return err;
+}
+
+static int kevent_aio_dequeue(struct kevent *k)
+{
+	struct kevent_aio_private *priv = k->priv;
+	struct inode *inode = k->st->origin;
+	struct file *file = priv->file;
+	struct file *sock = priv->sock;
+
+	kevent_storage_dequeue(k->st, k);
+	flush_scheduled_work();
+	wait_for_completion(&priv->bio_complete);
+
+	kfree(k->priv);
+	k->priv = NULL;
+	iput(inode);
+	fput(file);
+	fput(sock);
+
+	return 0;
+}
+
+asmlinkage long sys_aio_sendfile(int ctl_fd, int fd, int s, 
+		size_t size, unsigned flags)
+{
+	struct ukevent ukread, uksend;
+	struct kevent_user *u;
+	struct file *file;
+	int err, fput_needed;
+	int num = (flags & 7)?(flags & 7):8;
+
+	memset(&ukread, 0, sizeof(struct ukevent));
+	memset(&uksend, 0, sizeof(struct ukevent));
+
+	ukread.type = KEVENT_AIO;
+	ukread.event = KEVENT_AIO_BIO;
+
+	ukread.id.raw[0] = fd;
+	ukread.id.raw[1] = num;
+	ukread.ret_data[0] = s;
+
+	dprintk("%s: fd=%d, s=%d, num=%d.\n", __func__, fd, s, num);
+
+	file = fget_light(ctl_fd, &fput_needed);
+	if (!file)
+		return -ENODEV;
+
+	u = file->private_data;
+	if (!u) {
+		err = -EINVAL;
+		goto err_out_fput;
+	}
+
+	err = kevent_user_add_ukevent(&ukread, u);
+	if (err < 0)
+		goto err_out_fput;
+
+err_out_fput:
+	fput_light(file, fput_needed);
+	return err;
+}
+
+int kevent_init_aio(struct kevent *k)
+{
+	k->enqueue = &kevent_aio_enqueue;
+	k->dequeue = &kevent_aio_dequeue;
+	k->callback = &kevent_aio_callback;
+	return 0;
+}
diff --git a/kernel/kevent/kevent_inode.c b/kernel/kevent/kevent_inode.c
new file mode 100644
index 0000000..3af0e11
--- /dev/null
+++ b/kernel/kevent/kevent_inode.c
@@ -0,0 +1,110 @@
+/*
+ * 	kevent_inode.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/timer.h>
+#include <linux/file.h>
+#include <linux/kevent.h>
+#include <linux/fs.h>
+
+static int kevent_inode_enqueue(struct kevent *k)
+{
+	struct file *file;
+	struct inode *inode;
+	int err, fput_needed;
+
+	file = fget_light(k->event.id.raw[0], &fput_needed);
+	if (!file)
+		return -ENODEV;
+
+	err = -EINVAL;
+	if (!file->f_dentry || !file->f_dentry->d_inode)
+		goto err_out_fput;
+	
+	inode = igrab(file->f_dentry->d_inode);
+	if (!inode)
+		goto err_out_fput;
+
+	err = kevent_storage_enqueue(&inode->st, k);
+	if (err)
+		goto err_out_iput;
+
+	fput_light(file, fput_needed);
+	return 0;
+
+err_out_iput:
+	iput(inode);
+err_out_fput:
+	fput_light(file, fput_needed);
+	return err;
+}
+
+static int kevent_inode_dequeue(struct kevent *k)
+{
+	struct inode *inode = k->st->origin;
+
+	kevent_storage_dequeue(k->st, k);
+	iput(inode);
+
+	return 0;
+}
+
+static int kevent_inode_callback(struct kevent *k)
+{
+	return 1;
+}
+
+int kevent_init_inode(struct kevent *k)
+{
+	k->enqueue = &kevent_inode_enqueue;
+	k->dequeue = &kevent_inode_dequeue;
+	k->callback = &kevent_inode_callback;
+	return 0;
+}
+
+void kevent_inode_notify_parent(struct dentry *dentry, u32 event)
+{
+	struct dentry *parent;
+	struct inode *inode;
+	
+	spin_lock(&dentry->d_lock);
+	parent = dentry->d_parent;
+	inode = parent->d_inode;
+
+	dget(parent);
+	spin_unlock(&dentry->d_lock);
+	kevent_inode_notify(inode, KEVENT_INODE_REMOVE);
+	dput(parent);
+}
+	
+void kevent_inode_remove(struct inode *inode)
+{
+	kevent_storage_fini(&inode->st);
+}
+	
+void kevent_inode_notify(struct inode *inode, u32 event)
+{
+	kevent_storage_ready(&inode->st, NULL, event);
+}



-- 
	Evgeniy Polyakov

[4/4] kevent: generic poll and timer notifications.

[Evgeniy Polyakov]

This patch includes generic poll/select and timer notifications.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mitp.ru>

diff --git a/kernel/kevent/kevent_poll.c b/kernel/kevent/kevent_poll.c
new file mode 100644
index 0000000..4950e7c
--- /dev/null
+++ b/kernel/kevent/kevent_poll.c
@@ -0,0 +1,223 @@
+/*
+ * 	kevent_poll.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/timer.h>
+#include <linux/file.h>
+#include <linux/kevent.h>
+#include <linux/poll.h>
+#include <linux/fs.h>
+
+static kmem_cache_t *kevent_poll_container_cache;
+static kmem_cache_t *kevent_poll_priv_cache;
+
+struct kevent_poll_ctl
+{
+	struct poll_table_struct 	pt;
+	struct kevent			*k;
+};
+
+struct kevent_poll_wait_container
+{
+	struct list_head		container_entry;
+	wait_queue_head_t		*whead;
+	wait_queue_t			wait;
+	struct kevent			*k;
+};
+
+struct kevent_poll_private
+{
+	struct list_head		container_list;
+	spinlock_t			container_lock;
+};
+
+static int kevent_poll_enqueue(struct kevent *k);
+static int kevent_poll_dequeue(struct kevent *k);
+static int kevent_poll_callback(struct kevent *k);
+
+static int kevent_poll_wait_callback(wait_queue_t *wait, 
+		unsigned mode, int sync, void *key)
+{
+	struct kevent_poll_wait_container *cont = 
+		container_of(wait, struct kevent_poll_wait_container, wait);
+	struct kevent *k = cont->k;
+	struct file *file = k->st->origin;
+	unsigned long flags;
+	u32 revents, event;
+
+	revents = file->f_op->poll(file, NULL);
+	spin_lock_irqsave(&k->lock, flags);
+	event = k->event.event;
+	spin_unlock_irqrestore(&k->lock, flags);
+
+	kevent_storage_ready(k->st, NULL, revents);
+
+	return 0;
+}
+
+static void kevent_poll_qproc(struct file *file, wait_queue_head_t *whead, 
+		struct poll_table_struct *poll_table)
+{
+	struct kevent *k = 
+		container_of(poll_table, struct kevent_poll_ctl, pt)->k;
+	struct kevent_poll_private *priv = k->priv;
+	struct kevent_poll_wait_container *cont;
+	unsigned long flags;
+
+	cont = kmem_cache_alloc(kevent_poll_container_cache, SLAB_KERNEL);
+	if (!cont) {
+		kevent_break(k);
+		return;
+	}
+		
+	cont->k = k;
+	init_waitqueue_func_entry(&cont->wait, kevent_poll_wait_callback);
+	cont->whead = whead;
+
+	spin_lock_irqsave(&priv->container_lock, flags);
+	list_add_tail(&cont->container_entry, &priv->container_list);
+	spin_unlock_irqrestore(&priv->container_lock, flags);
+
+	add_wait_queue(whead, &cont->wait);
+}
+
+static int kevent_poll_enqueue(struct kevent *k)
+{
+	struct file *file;
+	int err, ready = 0;
+	unsigned int revents;
+	struct kevent_poll_ctl ctl;
+	struct kevent_poll_private *priv;
+
+	file = fget(k->event.id.raw[0]);
+	if (!file)
+		return -ENODEV;
+
+	err = -EINVAL;
+	if (!file->f_op || !file->f_op->poll)
+		goto err_out_fput;
+
+	err = -ENOMEM;
+	priv = kmem_cache_alloc(kevent_poll_priv_cache, SLAB_KERNEL);
+	if (!priv)
+		goto err_out_fput;
+
+	spin_lock_init(&priv->container_lock);
+	INIT_LIST_HEAD(&priv->container_list);
+
+	k->priv = priv;
+
+	ctl.k = k;
+	init_poll_funcptr(&ctl.pt, &kevent_poll_qproc);
+
+	err = kevent_storage_enqueue(&file->st, k);
+	if (err)
+		goto err_out_free;
+
+	revents = file->f_op->poll(file, &ctl.pt);
+	if (revents & k->event.event) {
+		ready = 1;
+		kevent_poll_dequeue(k);
+	}
+	
+	return ready;
+
+err_out_free:
+	kmem_cache_free(kevent_poll_priv_cache, priv);
+err_out_fput:
+	fput(file);
+	return err;
+}
+
+static int kevent_poll_dequeue(struct kevent *k)
+{
+	struct file *file = k->st->origin;
+	struct kevent_poll_private *priv = k->priv;
+	struct kevent_poll_wait_container *w, *n;
+	unsigned long flags;
+
+	kevent_storage_dequeue(k->st, k);
+
+	spin_lock_irqsave(&priv->container_lock, flags);
+	list_for_each_entry_safe(w, n, &priv->container_list, container_entry) {
+		list_del(&w->container_entry);
+		remove_wait_queue(w->whead, &w->wait);
+		kmem_cache_free(kevent_poll_container_cache, w);
+	}
+	spin_unlock_irqrestore(&priv->container_lock, flags);
+	
+	kmem_cache_free(kevent_poll_priv_cache, priv);
+	k->priv = NULL;
+	
+	fput(file);
+
+	return 0;
+}
+
+static int kevent_poll_callback(struct kevent *k)
+{
+	struct file *file = k->st->origin;
+	unsigned int revents = file->f_op->poll(file, NULL);
+	return (revents & k->event.event);
+}
+
+int kevent_init_poll(struct kevent *k)
+{
+	if (!kevent_poll_container_cache || !kevent_poll_priv_cache)
+		return -ENOMEM;
+
+	k->enqueue = &kevent_poll_enqueue;
+	k->dequeue = &kevent_poll_dequeue;
+	k->callback = &kevent_poll_callback;
+	return 0;
+}
+
+
+static int __init kevent_poll_sys_init(void)
+{
+	kevent_poll_container_cache = kmem_cache_create("kevent_poll_container_cache", 
+			sizeof(struct kevent_poll_wait_container), 0, 0, NULL, NULL);
+	if (!kevent_poll_container_cache) {
+		printk(KERN_ERR "Failed to create kevent poll container cache.\n");
+		return -ENOMEM;
+	}
+	
+	kevent_poll_priv_cache = kmem_cache_create("kevent_poll_priv_cache", 
+			sizeof(struct kevent_poll_private), 0, 0, NULL, NULL);
+	if (!kevent_poll_priv_cache) {
+		printk(KERN_ERR "Failed to create kevent poll private data cache.\n");
+		kmem_cache_destroy(kevent_poll_container_cache);
+		kevent_poll_container_cache = NULL;
+		return -ENOMEM;
+	}
+
+	printk(KERN_INFO "Kevent poll()/select() subsystem has been initialized.\n");
+	return 0;
+}
+
+static void __exit kevent_poll_sys_fini(void)
+{
+	kmem_cache_destroy(kevent_poll_priv_cache);
+	kmem_cache_destroy(kevent_poll_container_cache);
+}
+
+module_init(kevent_poll_sys_init);
+module_exit(kevent_poll_sys_fini);
diff --git a/kernel/kevent/kevent_timer.c b/kernel/kevent/kevent_timer.c
new file mode 100644
index 0000000..53d3bdf
--- /dev/null
+++ b/kernel/kevent/kevent_timer.c
@@ -0,0 +1,112 @@
+/*
+ * 	kevent_timer.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/timer.h>
+#include <linux/jiffies.h>
+#include <linux/kevent.h>
+
+static void kevent_timer_func(unsigned long data)
+{
+	struct kevent *k = (struct kevent *)data;
+	struct timer_list *t = k->st->origin;
+
+	kevent_storage_ready(k->st, NULL, KEVENT_MASK_ALL);
+	mod_timer(t, jiffies + msecs_to_jiffies(k->event.id.raw[0]));
+}
+
+static int kevent_timer_enqueue(struct kevent *k)
+{
+	struct timer_list *t;
+	struct kevent_storage *st;
+	int err;
+
+	t = kmalloc(sizeof(struct timer_list) + sizeof(struct kevent_storage), 
+			GFP_KERNEL);
+	if (!t)
+		return -ENOMEM;
+
+	init_timer(t);
+	t->function = kevent_timer_func;
+	t->expires = jiffies + msecs_to_jiffies(k->event.id.raw[0]);
+	t->data = (unsigned long)k;
+
+	st = (struct kevent_storage *)(t+1);
+	err = kevent_storage_init(t, st);
+	if (err)
+		goto err_out_free;
+
+	err = kevent_storage_enqueue(st, k);
+	if (err)
+		goto err_out_st_fini;
+	
+	add_timer(t);
+
+	return 0;
+
+err_out_st_fini:	
+	kevent_storage_fini(st);
+err_out_free:
+	kfree(t);
+
+	return err;
+}
+
+static int kevent_timer_dequeue(struct kevent *k)
+{
+	struct kevent_storage *st = k->st;
+	struct timer_list *t = st->origin;
+
+	if (!t)
+		return -ENODEV;
+
+	del_timer_sync(t);
+	
+	kevent_storage_dequeue(st, k);
+	
+	kfree(t);
+
+	return 0;
+}
+
+static int kevent_timer_callback(struct kevent *k)
+{
+	struct kevent_storage *st = k->st;
+	struct timer_list *t = st->origin;
+
+	if (!t)
+		return -ENODEV;
+	
+	k->event.ret_data[0] = (__u32)jiffies;
+	return 1;
+}
+
+int kevent_init_timer(struct kevent *k)
+{
+	k->enqueue = &kevent_timer_enqueue;
+	k->dequeue = &kevent_timer_dequeue;
+	k->callback = &kevent_timer_callback;
+	return 0;
+}




-- 
	Evgeniy Polyakov

[0/4] kevent: generic event processing subsystem.

[Evgeniy Polyakov]

Kevent subsystem incorporates several AIO/kqueue design notes and ideas.
Kevent can be used both for edge and level notifications. It supports
socket notifications (accept, send, recv), network AIO (aio_send(),
aio_recv() and aio_sendfile()), inode notifications (create/remove),
generic poll()/select() notifications and timer notifications.

There are several object in the kevent system:
storage - each source of events (socket, inode, timer, aio, anything) has
	structure kevent_storage incorporated into it, which is basically a list
	of registered interests for this source of events.
user - it is abstraction which holds all requested kevents. It is
	similar to FreeBSD's kqueue.
kevent - set of interests for given source of events or storage.

When kevent is queued into storage, it will live there until removed by
kevent_dequeue(). When some activity is noticed in given storage, it
scans it's kevent_storage->list for kevents which match activity event.
If kevents are found and they are not already in the
kevent_user->ready_list, they will be added there at the end.

ioctl(WAIT) (or appropriate syscall) will wait until either requested
number of kevents are ready or timeout elapsed or at least one kevent is
ready, it's behaviour depends on parameters.

It is possible to have one-shot kevents, which are automatically removed
when are ready.

Any event can be added/removed/modified by ioctl or special controlling
syscall.

Network AIO is based on kevent and works as usual kevent storage on top
of inode.
When new socket is created it is associated with that inode and when
some activity is detected appropriate notifications are generated and
kevent_naio_callback() is called.
When new kevent is being registered, network AIO ->enqueue() callback
simply marks itself like usual socket event watcher. It also locks
physical userspace pages in memory and stores appropriate pointers in
private kevent structure. I have not created additional DMA memory
allocation methods, like Ulrich described in his article, so I handle it
inside NAIO which has some overhead (I posted get_user_pages()
sclability graph some time ago).
Network AIO callback gets pointers to userspace pages and tries to copy
data from receiving skb queue into them using protocol specific
callback. This callback is very similar to ->recvmsg(), so they could
share a lot in future (as far as I recall it worked only with hardware
capable to do checksumming, I'm a bit lazy).

Both network and aio implementation work on top of hooks inside
appropriate state machines, but not as repeated call design (curect AIO) 
or special thread (SGI AIO). AIO work was stopped, since I was unable to 
achieve the same speed as synchronous read 
(maximum speeds were 2Gb/sec vs. 2.1 GB/sec for aio and sync IO accordingly
when reading data from the cache).
Network aio_sendfile() works lazily - it asynchronously populates pages
into the VFS cache (which can be used for various tricks with adaptive
readahead) and then uses usual ->sendfile() callback.

I have not created an interface for userspace events (like Solaris), 
since right now I do not see it's usefullness, but if there is
requirements for that it is quite easy with kevents.

Patches currently include ifdefs and kevent can be disabled in config,
when things are settled that can be removed.

1. kevent homepage.
http://tservice.net.ru/~s0mbre/old/?section=projects&item=kevent

2. network aio homepage.
http://tservice.net.ru/~s0mbre/old/?section=projects&item=naio

3. LWN.net published a very good article about kevent.
http://lwn.net/Articles/172844/

Thank you.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>

[1/4] kevent: core files.

[Evgeniy Polyakov]

This patch includes core kevent files:
 - userspace controlling
 - kernelspace interfaces
 - initialization
 - notification state machines

It might also inlclude parts from other subsystem (like network related
syscalls, so it is possible that it will not compile without other
patches applied).

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>

diff --git a/arch/i386/kernel/syscall_table.S b/arch/i386/kernel/syscall_table.S
index af56987..93e23ff 100644
--- a/arch/i386/kernel/syscall_table.S
+++ b/arch/i386/kernel/syscall_table.S
@@ -316,3 +316,7 @@ ENTRY(sys_call_table)
 	.long sys_sync_file_range
 	.long sys_tee			/* 315 */
 	.long sys_vmsplice
+	.long sys_aio_recv
+	.long sys_aio_send
+	.long sys_aio_sendfile
+	.long sys_kevent_ctl
diff --git a/arch/x86_64/ia32/ia32entry.S b/arch/x86_64/ia32/ia32entry.S
index 5a92fed..534d516 100644
--- a/arch/x86_64/ia32/ia32entry.S
+++ b/arch/x86_64/ia32/ia32entry.S
@@ -696,4 +696,8 @@ #endif
 	.quad sys_sync_file_range
 	.quad sys_tee
 	.quad compat_sys_vmsplice
+	.quad sys_aio_recv
+	.quad sys_aio_send
+	.quad sys_aio_sendfile
+	.quad sys_kevent_ctl
 ia32_syscall_end:		

diff --git a/include/asm-i386/socket.h b/include/asm-i386/socket.h
index 802ae76..3473f5c 100644
--- a/include/asm-i386/socket.h
+++ b/include/asm-i386/socket.h
@@ -49,4 +49,6 @@ #define SO_ACCEPTCONN		30
 
 #define SO_PEERSEC		31
 
+#define SO_ASYNC_SOCK		34
+
 #endif /* _ASM_SOCKET_H */
diff --git a/include/asm-i386/unistd.h b/include/asm-i386/unistd.h
index de2ccc1..52f8642 100644
--- a/include/asm-i386/unistd.h
+++ b/include/asm-i386/unistd.h
@@ -322,10 +322,14 @@ #define __NR_splice		313
 #define __NR_sync_file_range	314
 #define __NR_tee		315
 #define __NR_vmsplice		316
+#define __NR_aio_recv		317
+#define __NR_aio_send		318
+#define __NR_aio_sendfile	319
+#define __NR_kevent_ctl		320
 
 #ifdef __KERNEL__
 
-#define NR_syscalls 317
+#define NR_syscalls 321
 
 /*
  * user-visible error numbers are in the range -1 - -128: see
diff --git a/include/asm-x86_64/socket.h b/include/asm-x86_64/socket.h
index f2cdbea..1f31f86 100644
--- a/include/asm-x86_64/socket.h
+++ b/include/asm-x86_64/socket.h
@@ -49,4 +49,6 @@ #define SO_ACCEPTCONN		30
 
 #define SO_PEERSEC             31
 
+#define SO_ASYNC_SOCK		34
+
 #endif /* _ASM_SOCKET_H */
diff --git a/include/asm-x86_64/unistd.h b/include/asm-x86_64/unistd.h
index 0aff22b..352c34b 100644
--- a/include/asm-x86_64/unistd.h
+++ b/include/asm-x86_64/unistd.h
@@ -617,11 +617,18 @@ #define __NR_sync_file_range	277
 __SYSCALL(__NR_sync_file_range, sys_sync_file_range)
 #define __NR_vmsplice		278
 __SYSCALL(__NR_vmsplice, sys_vmsplice)
+#define __NR_aio_recv		279
+__SYSCALL(__NR_aio_recv, sys_aio_recv)
+#define __NR_aio_send		280
+__SYSCALL(__NR_aio_send, sys_aio_send)
+#define __NR_aio_sendfile	281
+__SYSCALL(__NR_aio_sendfile, sys_aio_sendfile)
+#define __NR_kevent_ctl		282
+__SYSCALL(__NR_kevent_ctl, sys_kevent_ctl)
 
 #ifdef __KERNEL__
 
-#define __NR_syscall_max __NR_vmsplice
-
+#define __NR_syscall_max __NR_kevent_ctl
 #ifndef __NO_STUBS
 
 /* user-visible error numbers are in the range -1 - -4095 */

diff --git a/include/linux/kevent.h b/include/linux/kevent.h
new file mode 100644
index 0000000..e94a7bf
--- /dev/null
+++ b/include/linux/kevent.h
@@ -0,0 +1,263 @@
+/*
+ * 	kevent.h
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef __KEVENT_H
+#define __KEVENT_H
+
+/*
+ * Kevent request flags.
+ */
+
+#define KEVENT_REQ_ONESHOT	0x1		/* Process this event only once and then dequeue. */
+
+/*
+ * Kevent return flags.
+ */
+#define KEVENT_RET_BROKEN	0x1		/* Kevent is broken. */
+#define KEVENT_RET_DONE		0x2		/* Kevent processing was finished successfully. */
+
+/*
+ * Kevent type set.
+ */
+#define KEVENT_SOCKET 		0
+#define KEVENT_INODE		1
+#define KEVENT_TIMER		2
+#define KEVENT_POLL		3
+#define KEVENT_NAIO		4
+#define KEVENT_AIO		5
+#define	KEVENT_MAX		6
+
+/*
+ * Per-type event sets.
+ * Number of per-event sets should be exactly as number of kevent types.
+ */
+
+/*
+ * Timer events.
+ */
+#define	KEVENT_TIMER_FIRED	0x1
+
+/*
+ * Socket/network asynchronous IO events.
+ */
+#define	KEVENT_SOCKET_RECV	0x1
+#define	KEVENT_SOCKET_ACCEPT	0x2
+#define	KEVENT_SOCKET_SEND	0x4
+
+/*
+ * Inode events.
+ */
+#define	KEVENT_INODE_CREATE	0x1
+#define	KEVENT_INODE_REMOVE	0x2
+
+/*
+ * Poll events.
+ */
+#define	KEVENT_POLL_POLLIN	0x0001
+#define	KEVENT_POLL_POLLPRI	0x0002
+#define	KEVENT_POLL_POLLOUT	0x0004
+#define	KEVENT_POLL_POLLERR	0x0008
+#define	KEVENT_POLL_POLLHUP	0x0010
+#define	KEVENT_POLL_POLLNVAL	0x0020
+
+#define	KEVENT_POLL_POLLRDNORM	0x0040
+#define	KEVENT_POLL_POLLRDBAND	0x0080
+#define	KEVENT_POLL_POLLWRNORM	0x0100
+#define	KEVENT_POLL_POLLWRBAND	0x0200
+#define	KEVENT_POLL_POLLMSG	0x0400
+#define	KEVENT_POLL_POLLREMOVE	0x1000
+
+/*
+ * Asynchronous IO events.
+ */
+#define	KEVENT_AIO_BIO		0x1
+
+#define KEVENT_MASK_ALL		0xffffffff	/* Mask of all possible event values. */
+#define KEVENT_MASK_EMPTY	0x0		/* Empty mask of ready events. */
+
+struct kevent_id
+{
+	__u32		raw[2];
+};
+
+struct ukevent
+{
+	struct kevent_id	id;			/* Id of this request, e.g. socket number, file descriptor and so on... */
+	__u32			type;			/* Event type, e.g. KEVENT_SOCK, KEVENT_INODE, KEVENT_TIMER and so on... */
+	__u32			event;			/* Event itself, e.g. SOCK_ACCEPT, INODE_CREATED, TIMER_FIRED... */
+	__u32			req_flags;		/* Per-event request flags */
+	__u32			ret_flags;		/* Per-event return flags */
+	__u32			ret_data[2];		/* Event return data. Event originator fills it with anything it likes. */
+	union {
+		__u32		user[2];		/* User's data. It is not used, just copied to/from user. */
+		void		*ptr;
+	};
+};
+
+#define	KEVENT_CTL_ADD 		0
+#define	KEVENT_CTL_REMOVE	1
+#define	KEVENT_CTL_MODIFY	2
+#define	KEVENT_CTL_WAIT		3
+#define	KEVENT_CTL_INIT		4
+
+struct kevent_user_control
+{
+	unsigned int		cmd;			/* Control command, e.g. KEVENT_ADD, KEVENT_REMOVE... */
+	unsigned int		num;			/* Number of ukevents this strucutre controls. */
+	unsigned int		timeout;		/* Timeout in milliseconds waiting for "num" events to become ready. */
+};
+
+#define KEVENT_USER_SYMBOL	'K'
+#define KEVENT_USER_CTL		_IOWR(KEVENT_USER_SYMBOL, 0, struct kevent_user_control)
+#define KEVENT_USER_WAIT	_IOWR(KEVENT_USER_SYMBOL, 1, struct kevent_user_control)
+
+#ifdef __KERNEL__
+
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/kevent_storage.h>
+#include <asm/semaphore.h>
+
+struct inode;
+struct dentry;
+struct sock;
+
+struct kevent;
+struct kevent_storage;
+typedef int (* kevent_callback_t)(struct kevent *);
+
+struct kevent
+{
+	struct ukevent		event;
+	spinlock_t		lock;			/* This lock protects ukevent manipulations, e.g. ret_flags changes. */
+
+	struct list_head	kevent_entry;		/* Entry of user's queue. */
+	struct list_head	storage_entry;		/* Entry of origin's queue. */
+	struct list_head	ready_entry;		/* Entry of user's ready. */
+
+	struct kevent_user	*user;			/* User who requested this kevent. */
+	struct kevent_storage	*st;			/* Kevent container. */
+
+	kevent_callback_t	callback;		/* Is called each time new event has been caught. */
+	kevent_callback_t	enqueue;		/* Is called each time new event is queued. */
+	kevent_callback_t	dequeue;		/* Is called each time event is dequeued. */
+
+	void			*priv;			/* Private data for different storages. 
+							 * poll()/select storage has a list of wait_queue_t containers 
+							 * for each ->poll() { poll_wait()' } here.
+							 */
+};
+
+#define KEVENT_HASH_MASK	0xff
+
+struct kevent_list
+{
+	struct list_head	kevent_list;		/* List of all kevents. */
+	spinlock_t 		kevent_lock;		/* Protects all manipulations with queue of kevents. */
+};
+
+struct kevent_user
+{
+	struct kevent_list	kqueue[KEVENT_HASH_MASK+1];
+	unsigned int		kevent_num;		/* Number of queued kevents. */
+
+	struct list_head	ready_list;		/* List of ready kevents. */
+	unsigned int		ready_num;		/* Number of ready kevents. */
+	spinlock_t 		ready_lock;		/* Protects all manipulations with ready queue. */
+
+	unsigned int		max_ready_num;		/* Requested number of kevents. */
+
+	struct semaphore	ctl_mutex;		/* Protects against simultaneous kevent_user control manipulations. */
+	struct semaphore	wait_mutex;		/* Protects against simultaneous kevent_user waits. */
+	wait_queue_head_t	wait;			/* Wait until some events are ready. */
+
+	atomic_t		refcnt;			/* Reference counter, increased for each new kevent. */
+#ifdef CONFIG_KEVENT_USER_STAT
+	unsigned long		im_num;
+	unsigned long		wait_num;
+	unsigned long		total;
+#endif
+};
+
+#define KEVENT_MAX_REQUESTS		PAGE_SIZE/sizeof(struct kevent)
+
+struct kevent *kevent_alloc(gfp_t mask);
+void kevent_free(struct kevent *k);
+int kevent_enqueue(struct kevent *k);
+int kevent_dequeue(struct kevent *k);
+int kevent_init(struct kevent *k);
+void kevent_requeue(struct kevent *k);
+
+#define list_for_each_entry_reverse_safe(pos, n, head, member)		\
+	for (pos = list_entry((head)->prev, typeof(*pos), member),	\
+		n = list_entry(pos->member.prev, typeof(*pos), member);	\
+	     prefetch(pos->member.prev), &pos->member != (head); 	\
+	     pos = n, n = list_entry(pos->member.prev, typeof(*pos), member))
+
+int kevent_break(struct kevent *k);
+int kevent_init(struct kevent *k);
+
+int kevent_init_socket(struct kevent *k);
+int kevent_init_inode(struct kevent *k);
+int kevent_init_timer(struct kevent *k);
+int kevent_init_poll(struct kevent *k);
+int kevent_init_naio(struct kevent *k);
+int kevent_init_aio(struct kevent *k);
+
+void kevent_storage_ready(struct kevent_storage *st, 
+		kevent_callback_t ready_callback, u32 event);
+int kevent_storage_init(void *origin, struct kevent_storage *st);
+void kevent_storage_fini(struct kevent_storage *st);
+int kevent_storage_enqueue(struct kevent_storage *st, struct kevent *k);
+void kevent_storage_dequeue(struct kevent_storage *st, struct kevent *k);
+
+int kevent_user_add_ukevent(struct ukevent *uk, struct kevent_user *u);
+
+#ifdef CONFIG_KEVENT_INODE
+void kevent_inode_notify(struct inode *inode, u32 event);
+void kevent_inode_notify_parent(struct dentry *dentry, u32 event);
+void kevent_inode_remove(struct inode *inode);
+#else
+static inline void kevent_inode_notify(struct inode *inode, u32 event)
+{
+}
+static inline void kevent_inode_notify_parent(struct dentry *dentry, u32 event)
+{
+}
+static inline void kevent_inode_remove(struct inode *inode)
+{
+}
+#endif /* CONFIG_KEVENT_INODE */
+#ifdef CONFIG_KEVENT_SOCKET
+
+void kevent_socket_notify(struct sock *sock, u32 event);
+int kevent_socket_dequeue(struct kevent *k);
+int kevent_socket_enqueue(struct kevent *k);
+#define sock_async(__sk) sock_flag(__sk, SOCK_ASYNC)
+#else
+static inline void kevent_socket_notify(struct sock *sock, u32 event)
+{
+}
+#define sock_async(__sk)	0
+#endif
+#endif /* __KERNEL__ */
+#endif /* __KEVENT_H */
diff --git a/include/linux/kevent_storage.h b/include/linux/kevent_storage.h
new file mode 100644
index 0000000..bd891f0
--- /dev/null
+++ b/include/linux/kevent_storage.h
@@ -0,0 +1,12 @@
+#ifndef __KEVENT_STORAGE_H
+#define __KEVENT_STORAGE_H
+
+struct kevent_storage
+{
+	void			*origin;		/* Originator's pointer, e.g. struct sock or struct file. Can be NULL. */
+	struct list_head	list;			/* List of queued kevents. */
+	unsigned int		qlen;			/* Number of queued kevents. */
+	spinlock_t		lock;			/* Protects users queue. */
+};
+
+#endif /* __KEVENT_STORAGE_H */
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 66f8819..ea914c3 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -1269,6 +1269,8 @@ extern struct sk_buff *skb_recv_datagram
 					 int noblock, int *err);
 extern unsigned int    datagram_poll(struct file *file, struct socket *sock,
 				     struct poll_table_struct *wait);
+extern int	       skb_copy_datagram(const struct sk_buff *from, 
+					 int offset, void *dst, int size);
 extern int	       skb_copy_datagram_iovec(const struct sk_buff *from,
 					       int offset, struct iovec *to,
 					       int size);
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
index bd67a44..33d436e 100644
--- a/include/linux/syscalls.h
+++ b/include/linux/syscalls.h
@@ -587,4 +587,8 @@ asmlinkage long sys_get_robust_list(int 
 asmlinkage long sys_set_robust_list(struct robust_list_head __user *head,
 				    size_t len);
 
+asmlinkage long sys_aio_recv(int ctl_fd, int s, void __user *buf, size_t size, unsigned flags);
+asmlinkage long sys_aio_send(int ctl_fd, int s, void __user *buf, size_t size, unsigned flags);
+asmlinkage long sys_aio_sendfile(int ctl_fd, int fd, int s, size_t size, unsigned flags);
+asmlinkage long sys_kevent_ctl(int ctl_fd, void __user *buf);
 #endif
diff --git a/include/net/sock.h b/include/net/sock.h
index d10dfec..7a2bee3 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -47,6 +47,7 @@ #include <linux/module.h>
 #include <linux/netdevice.h>
 #include <linux/skbuff.h>	/* struct sk_buff */
 #include <linux/security.h>
+#include <linux/kevent.h>
 
 #include <linux/filter.h>
 
@@ -386,6 +387,8 @@ enum sock_flags {
 	SOCK_NO_LARGESEND, /* whether to sent large segments or not */
 	SOCK_LOCALROUTE, /* route locally only, %SO_DONTROUTE setting */
 	SOCK_QUEUE_SHRUNK, /* write queue has been shrunk recently */
+	SOCK_ASYNC,
+	SOCK_ASYNC_INUSE,
 };
 
 static inline void sock_copy_flags(struct sock *nsk, struct sock *osk)
@@ -445,6 +448,21 @@ static inline int sk_stream_memory_free(
 
 extern void sk_stream_rfree(struct sk_buff *skb);
 
+struct socket_alloc {
+	struct socket socket;
+	struct inode vfs_inode;
+};
+
+static inline struct socket *SOCKET_I(struct inode *inode)
+{
+	return &container_of(inode, struct socket_alloc, vfs_inode)->socket;
+}
+
+static inline struct inode *SOCK_INODE(struct socket *socket)
+{
+	return &container_of(socket, struct socket_alloc, socket)->vfs_inode;
+}
+
 static inline void sk_stream_set_owner_r(struct sk_buff *skb, struct sock *sk)
 {
 	skb->sk = sk;
@@ -472,6 +490,7 @@ static inline void sk_add_backlog(struct
 		sk->sk_backlog.tail = skb;
 	}
 	skb->next = NULL;
+	kevent_socket_notify(sk, KEVENT_SOCKET_RECV);
 }
 
 #define sk_wait_event(__sk, __timeo, __condition)		\
@@ -543,6 +562,12 @@ struct proto {
 
 	int			(*backlog_rcv) (struct sock *sk, 
 						struct sk_buff *skb);
+	
+	int			(*async_recv) (struct sock *sk, 
+						void *dst, size_t size);
+	int			(*async_send) (struct sock *sk, 
+						struct page **pages, unsigned int poffset, 
+						size_t size);
 
 	/* Keeping track of sk's, looking them up, and port selection methods. */
 	void			(*hash)(struct sock *sk);
@@ -674,21 +699,6 @@ static inline struct kiocb *siocb_to_kio
 	return si->kiocb;
 }
 
-struct socket_alloc {
-	struct socket socket;
-	struct inode vfs_inode;
-};
-
-static inline struct socket *SOCKET_I(struct inode *inode)
-{
-	return &container_of(inode, struct socket_alloc, vfs_inode)->socket;
-}
-
-static inline struct inode *SOCK_INODE(struct socket *socket)
-{
-	return &container_of(socket, struct socket_alloc, socket)->vfs_inode;
-}
-
 extern void __sk_stream_mem_reclaim(struct sock *sk);
 extern int sk_stream_mem_schedule(struct sock *sk, int size, int kind);
 
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 5f4eb5c..820cd5a 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -364,6 +364,8 @@ extern int			compat_tcp_setsockopt(struc
 					int level, int optname,
 					char __user *optval, int optlen);
 extern void			tcp_set_keepalive(struct sock *sk, int val);
+extern int			tcp_async_recv(struct sock *sk, void *dst, size_t size);
+extern int			tcp_async_send(struct sock *sk, struct page **pages, unsigned int poffset, size_t size);
 extern int			tcp_recvmsg(struct kiocb *iocb, struct sock *sk,
 					    struct msghdr *msg,
 					    size_t len, int nonblock, 
@@ -857,6 +859,7 @@ static inline int tcp_prequeue(struct so
 			tp->ucopy.memory = 0;
 		} else if (skb_queue_len(&tp->ucopy.prequeue) == 1) {
 			wake_up_interruptible(sk->sk_sleep);
+			kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
 			if (!inet_csk_ack_scheduled(sk))
 				inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
 						          (3 * TCP_RTO_MIN) / 4,
diff --git a/init/Kconfig b/init/Kconfig
index df864a3..6135afc 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -185,6 +185,8 @@ config AUDITSYSCALL
 	  such as SELinux.  To use audit's filesystem watch feature, please
 	  ensure that INOTIFY is configured.
 
+source "kernel/kevent/Kconfig"
+
 config IKCONFIG
 	bool "Kernel .config support"
 	---help---
diff --git a/kernel/Makefile b/kernel/Makefile
index f6ef00f..eb057ea 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -36,6 +36,7 @@ obj-$(CONFIG_DETECT_SOFTLOCKUP) += softl
 obj-$(CONFIG_GENERIC_HARDIRQS) += irq/
 obj-$(CONFIG_SECCOMP) += seccomp.o
 obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o
+obj-$(CONFIG_KEVENT) += kevent/
 obj-$(CONFIG_RELAY) += relay.o
 
 ifneq ($(CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER),y)
diff --git a/kernel/kevent/Kconfig b/kernel/kevent/Kconfig
new file mode 100644
index 0000000..88b35af
--- /dev/null
+++ b/kernel/kevent/Kconfig
@@ -0,0 +1,57 @@
+config KEVENT
+	bool "Kernel event notification mechanism"
+	help
+	  This option enables event queue mechanism.
+	  It can be used as replacement for poll()/select(), AIO callback invocations,
+	  advanced timer notifications and other kernel object status changes.
+
+config KEVENT_USER_STAT
+	bool "Kevent user statistic"
+	depends on KEVENT
+	default N
+	help
+	  This option will turn kevent_user statistic collection on.
+	  Statistic data includes total number of kevent, number of kevents which are ready
+	  immediately at insertion time and number of kevents which were removed through
+	  readiness completion. It will be printed each time control kevent descriptor
+	  is closed.
+
+config KEVENT_SOCKET
+	bool "Kernel event notifications for sockets"
+	depends on NET && KEVENT
+	help
+	  This option enables notifications through KEVENT subsystem of 
+	  sockets operations, like new packet receiving conditions, ready for accept
+  	  conditions and so on.
+	
+config KEVENT_INODE
+	bool "Kernel event notifications for inodes"
+	depends on KEVENT
+	help
+	  This option enables notifications through KEVENT subsystem of 
+	  inode operations, like file creation, removal and so on.
+
+config KEVENT_TIMER
+	bool "Kernel event notifications for timers"
+	depends on KEVENT
+	help
+	  This option allows to use timers through KEVENT subsystem.
+
+config KEVENT_POLL
+	bool "Kernel event notifications for poll()/select()"
+	depends on KEVENT
+	help
+	  This option allows to use kevent subsystem for poll()/select() notifications.
+
+config KEVENT_NAIO
+	bool "Network asynchronous IO"
+	depends on KEVENT && KEVENT_SOCKET
+	help
+	  This option enables kevent based network asynchronous IO subsystem.
+
+config KEVENT_AIO
+	bool "Asynchronous IO"
+	depends on KEVENT
+	help
+	  This option allows to use kevent subsystem for AIO operations.
+	  AIO read is currently supported.
diff --git a/kernel/kevent/Makefile b/kernel/kevent/Makefile
new file mode 100644
index 0000000..7dcd651
--- /dev/null
+++ b/kernel/kevent/Makefile
@@ -0,0 +1,7 @@
+obj-y := kevent.o kevent_user.o kevent_init.o
+obj-$(CONFIG_KEVENT_SOCKET) += kevent_socket.o
+obj-$(CONFIG_KEVENT_INODE) += kevent_inode.o
+obj-$(CONFIG_KEVENT_TIMER) += kevent_timer.o
+obj-$(CONFIG_KEVENT_POLL) += kevent_poll.o
+obj-$(CONFIG_KEVENT_NAIO) += kevent_naio.o
+obj-$(CONFIG_KEVENT_AIO) += kevent_aio.o
diff --git a/kernel/kevent/kevent.c b/kernel/kevent/kevent.c
new file mode 100644
index 0000000..f699a13
--- /dev/null
+++ b/kernel/kevent/kevent.c
@@ -0,0 +1,260 @@
+/*
+ * 	kevent.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/mempool.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
+#include <linux/kevent.h>
+
+static kmem_cache_t *kevent_cache;
+
+/*
+ * Attempts to add an event into appropriate origin's queue.
+ * Returns positive value if this event is ready immediately,
+ * negative value in case of error and zero if event has been queued.
+ * ->enqueue() callback must increase origin's reference counter.
+ */
+int kevent_enqueue(struct kevent *k)
+{
+	if (k->event.type >= KEVENT_MAX)
+		return -E2BIG;
+
+	if (!k->enqueue) {
+		kevent_break(k);
+		return -EINVAL;
+	}
+	
+	return k->enqueue(k);
+}
+
+/*
+ * Remove event from the appropriate queue.
+ * ->dequeue() callback must decrease origin's reference counter.
+ */
+int kevent_dequeue(struct kevent *k)
+{
+	if (k->event.type >= KEVENT_MAX)
+		return -E2BIG;
+	
+	if (!k->dequeue) {
+		kevent_break(k);
+		return -EINVAL;
+	}
+
+	return k->dequeue(k);
+}
+
+/*
+ * Must be called before event is going to be added into some origin's queue.
+ * Initializes ->enqueue(), ->dequeue() and ->callback() callbacks.
+ * If failed, kevent should not be used or kevent_enqueue() will fail to add
+ * this kevent into origin's queue with setting
+ * KEVENT_RET_BROKEN flag in kevent->event.ret_flags.
+ */
+int kevent_init(struct kevent *k)
+{
+	int err;
+
+	spin_lock_init(&k->lock);
+	k->kevent_entry.next = LIST_POISON1;
+	k->storage_entry.next = LIST_POISON1;
+	k->ready_entry.next = LIST_POISON1;
+
+	if (k->event.type >= KEVENT_MAX)
+		return -E2BIG;
+	
+	switch (k->event.type) {
+		case KEVENT_NAIO:
+			err = kevent_init_naio(k);
+			break;
+		case KEVENT_SOCKET:
+			err = kevent_init_socket(k);
+			break;
+		case KEVENT_INODE:
+			err = kevent_init_inode(k);
+			break;
+		case KEVENT_TIMER:
+			err = kevent_init_timer(k);
+			break;
+		case KEVENT_POLL:
+			err = kevent_init_poll(k);
+			break;
+		case KEVENT_AIO:
+			err = kevent_init_aio(k);
+			break;
+		default:
+			err = -ENODEV;
+	}
+
+	return err;
+}
+
+/*
+ * Called from ->enqueue() callback when reference counter for given
+ * origin (socket, inode...) has been increased.
+ */
+int kevent_storage_enqueue(struct kevent_storage *st, struct kevent *k)
+{
+	unsigned long flags;
+
+	k->st = st;
+	spin_lock_irqsave(&st->lock, flags);
+	list_add_tail(&k->storage_entry, &st->list);
+	st->qlen++;
+	spin_unlock_irqrestore(&st->lock, flags);
+	return 0;
+}
+
+/*
+ * Dequeue kevent from origin's queue. 
+ * It does not decrease origin's reference counter in any way 
+ * and must be called before it, so storage itself must be valid.
+ * It is called from ->dequeue() callback.
+ */
+void kevent_storage_dequeue(struct kevent_storage *st, struct kevent *k)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&st->lock, flags);
+	if (k->storage_entry.next != LIST_POISON1) {
+		list_del(&k->storage_entry);
+		st->qlen--;
+	}
+	spin_unlock_irqrestore(&st->lock, flags);
+}
+
+static void __kevent_requeue(struct kevent *k, u32 event)
+{
+	int err, rem = 0;
+	unsigned long flags;
+
+	err = k->callback(k);
+
+	spin_lock_irqsave(&k->lock, flags);
+	if (err > 0) {
+		k->event.ret_flags |= KEVENT_RET_DONE;
+	} else if (err < 0) {
+		k->event.ret_flags |= KEVENT_RET_BROKEN;
+		k->event.ret_flags |= KEVENT_RET_DONE;
+	}
+	rem = (k->event.req_flags & KEVENT_REQ_ONESHOT);
+	if (!err)
+		err = (k->event.ret_flags & (KEVENT_RET_BROKEN|KEVENT_RET_DONE));
+	spin_unlock_irqrestore(&k->lock, flags);
+
+	if (err) {
+		if (rem) {
+			list_del(&k->storage_entry);
+			k->st->qlen--;
+		}
+		
+		spin_lock_irqsave(&k->user->ready_lock, flags);
+		if (k->ready_entry.next == LIST_POISON1) {
+			list_add_tail(&k->ready_entry, &k->user->ready_list);
+			k->user->ready_num++;
+		}
+		spin_unlock_irqrestore(&k->user->ready_lock, flags);
+		wake_up(&k->user->wait);
+	}
+}
+
+void kevent_requeue(struct kevent *k)
+{
+	unsigned long flags;
+	
+	spin_lock_irqsave(&k->st->lock, flags);
+	__kevent_requeue(k, 0);
+	spin_unlock_irqrestore(&k->st->lock, flags);
+}
+
+/*
+ * Called each time some activity in origin (socket, inode...) is noticed.
+ */
+void kevent_storage_ready(struct kevent_storage *st, 
+		kevent_callback_t ready_callback, u32 event)
+{
+	struct kevent *k, *n;
+
+	spin_lock(&st->lock);
+	list_for_each_entry_safe(k, n, &st->list, storage_entry) {
+		if (ready_callback)
+			ready_callback(k);
+
+		if (event & k->event.event)
+			__kevent_requeue(k, event);
+	}
+	spin_unlock(&st->lock);
+}
+
+int kevent_storage_init(void *origin, struct kevent_storage *st)
+{
+	spin_lock_init(&st->lock);
+	st->origin = origin;
+	st->qlen = 0;
+	INIT_LIST_HEAD(&st->list);
+	return 0;
+}
+
+void kevent_storage_fini(struct kevent_storage *st)
+{
+	kevent_storage_ready(st, kevent_break, KEVENT_MASK_ALL);
+}
+
+struct kevent *kevent_alloc(gfp_t mask)
+{
+	struct kevent *k;
+	
+	if (kevent_cache)
+		k = kmem_cache_alloc(kevent_cache, mask);
+	else
+		k = kzalloc(sizeof(struct kevent), mask);
+
+	return k;
+}
+
+void kevent_free(struct kevent *k)
+{
+	memset(k, 0xab, sizeof(struct kevent));
+
+	if (kevent_cache)
+		kmem_cache_free(kevent_cache, k);
+	else
+		kfree(k);
+}
+
+int __init kevent_sys_init(void)
+{
+	int err = 0;
+
+	kevent_cache = kmem_cache_create("kevent_cache", 
+			sizeof(struct kevent), 0, 0, NULL, NULL);
+	if (!kevent_cache)
+		err = -ENOMEM;
+	
+	return err;
+}
+
+late_initcall(kevent_sys_init);

diff --git a/kernel/kevent/kevent_init.c b/kernel/kevent/kevent_init.c
new file mode 100644
index 0000000..ec95114
--- /dev/null
+++ b/kernel/kevent/kevent_init.c
@@ -0,0 +1,85 @@
+/*
+ * 	kevent_init.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/errno.h>
+#include <linux/kevent.h>
+
+int kevent_break(struct kevent *k)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&k->lock, flags);
+	k->event.ret_flags |= KEVENT_RET_BROKEN;
+	spin_unlock_irqrestore(&k->lock, flags);
+	return 0;
+}
+
+#ifndef CONFIG_KEVENT_SOCKET
+int kevent_init_socket(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_INODE
+int kevent_init_inode(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_TIMER
+int kevent_init_timer(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_POLL
+int kevent_init_poll(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_NAIO
+int kevent_init_naio(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif
+
+#ifndef CONFIG_KEVENT_AIO
+int kevent_init_aio(struct kevent *k)
+{
+	kevent_break(k);
+	return -ENODEV;
+}
+#endif

diff --git a/kernel/kevent/kevent_user.c b/kernel/kevent/kevent_user.c
new file mode 100644
index 0000000..2f71fe4
--- /dev/null
+++ b/kernel/kevent/kevent_user.c
@@ -0,0 +1,728 @@
+/*
+ * 	kevent_user.c
+ * 
+ * 2006 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/fs.h>
+#include <linux/file.h>
+#include <linux/mount.h>
+#include <linux/device.h>
+#include <linux/poll.h>
+#include <linux/kevent.h>
+#include <linux/jhash.h>
+#include <asm/uaccess.h>
+#include <asm/semaphore.h>
+
+static struct class *kevent_user_class;
+static char kevent_name[] = "kevent";
+static int kevent_user_major;
+
+static int kevent_user_open(struct inode *, struct file *);
+static int kevent_user_release(struct inode *, struct file *);
+static int kevent_user_ioctl(struct inode *, struct file *, 
+		unsigned int, unsigned long);
+static unsigned int kevent_user_poll(struct file *, struct poll_table_struct *);
+
+static struct file_operations kevent_user_fops = {
+	.open		= kevent_user_open,
+	.release	= kevent_user_release,
+	.ioctl		= kevent_user_ioctl,
+	.poll		= kevent_user_poll,
+	.owner		= THIS_MODULE,
+};
+
+static struct super_block *kevent_get_sb(struct file_system_type *fs_type, 
+		int flags, const char *dev_name, void *data)
+{
+	/* So original magic... */
+	return get_sb_pseudo(fs_type, kevent_name, NULL, 0xabcdef);	
+}
+
+static struct file_system_type kevent_fs_type = {
+	.name		= kevent_name,
+	.get_sb		= kevent_get_sb,
+	.kill_sb	= kill_anon_super,
+};
+
+static struct vfsmount *kevent_mnt;
+
+static unsigned int kevent_user_poll(struct file *file, struct poll_table_struct *wait)
+{
+	struct kevent_user *u = file->private_data;
+	unsigned int mask;
+	
+	poll_wait(file, &u->wait, wait);
+	mask = 0;
+
+	if (u->ready_num)
+		mask |= POLLIN | POLLRDNORM;
+
+	return mask;
+}
+
+static struct kevent_user *kevent_user_alloc(void)
+{
+	struct kevent_user *u;
+	int i;
+
+	u = kzalloc(sizeof(struct kevent_user), GFP_KERNEL);
+	if (!u)
+		return NULL;
+
+	INIT_LIST_HEAD(&u->ready_list);
+	spin_lock_init(&u->ready_lock);
+	u->ready_num = 0;
+#ifdef CONFIG_KEVENT_USER_STAT
+	u->wait_num = u->im_num = u->total = 0;
+#endif
+	for (i=0; i<KEVENT_HASH_MASK+1; ++i) {
+		INIT_LIST_HEAD(&u->kqueue[i].kevent_list);
+		spin_lock_init(&u->kqueue[i].kevent_lock);
+	}
+	u->kevent_num = 0;
+	
+	init_MUTEX(&u->ctl_mutex);
+	init_MUTEX(&u->wait_mutex);
+	init_waitqueue_head(&u->wait);
+	u->max_ready_num = 0;
+
+	atomic_set(&u->refcnt, 1);
+
+	return u;
+}
+
+static int kevent_user_open(struct inode *inode, struct file *file)
+{
+	struct kevent_user *u = kevent_user_alloc();
+	
+	if (!u)
+		return -ENOMEM;
+
+	file->private_data = u;
+	
+	return 0;
+}
+
+static inline void kevent_user_get(struct kevent_user *u)
+{
+	atomic_inc(&u->refcnt);
+}
+
+static inline void kevent_user_put(struct kevent_user *u)
+{
+	if (atomic_dec_and_test(&u->refcnt)) {
+#ifdef CONFIG_KEVENT_USER_STAT
+		printk("%s: u=%p, wait=%lu, immediately=%lu, total=%lu.\n", 
+				__func__, u, u->wait_num, u->im_num, u->total);
+#endif
+		kfree(u);
+	}
+}
+
+#if 0
+static inline unsigned int kevent_user_hash(struct ukevent *uk)
+{
+	unsigned int h = (uk->user[0] ^ uk->user[1]) ^ (uk->id.raw[0] ^ uk->id.raw[1]);
+	
+	h = (((h >> 16) & 0xffff) ^ (h & 0xffff)) & 0xffff;
+	h = (((h >> 8) & 0xff) ^ (h & 0xff)) & KEVENT_HASH_MASK;
+
+	return h;
+}
+#else
+static inline unsigned int kevent_user_hash(struct ukevent *uk)
+{
+	return jhash_1word(uk->id.raw[0], 0) & KEVENT_HASH_MASK;
+}
+#endif
+
+/*
+ * Remove kevent from user's list of all events, 
+ * dequeue it from storage and decrease user's reference counter,
+ * since this kevent does not exist anymore. That is why it is freed here.
+ */
+static void kevent_finish_user(struct kevent *k, int lock, int deq)
+{
+	struct kevent_user *u = k->user;
+	unsigned long flags;
+
+	if (lock) {
+		unsigned int hash = kevent_user_hash(&k->event);
+		struct kevent_list *l = &u->kqueue[hash];
+
+		spin_lock_irqsave(&l->kevent_lock, flags);
+		list_del(&k->kevent_entry);
+		u->kevent_num--;
+		spin_unlock_irqrestore(&l->kevent_lock, flags);
+	} else {
+		list_del(&k->kevent_entry);
+		u->kevent_num--;
+	}
+
+	if (deq)
+		kevent_dequeue(k);
+
+	spin_lock_irqsave(&u->ready_lock, flags);
+	if (k->ready_entry.next != LIST_POISON1) {
+		list_del(&k->ready_entry);
+		u->ready_num--;
+	}
+	spin_unlock_irqrestore(&u->ready_lock, flags);
+	
+	kevent_user_put(u);
+	kevent_free(k);
+}
+
+/*
+ * Dequeue one entry from user's ready queue.
+ */
+static struct kevent *__kqueue_dequeue_one_ready(struct list_head *q, 
+		unsigned int *qlen)
+{
+	struct kevent *k = NULL;
+	unsigned int len = *qlen;
+	
+	if (len && !list_empty(q)) {
+		k = list_entry(q->next, struct kevent, ready_entry);
+		list_del(&k->ready_entry);
+		*qlen = len - 1;
+	}
+	
+	return k;
+}
+
+static struct kevent *kqueue_dequeue_ready(struct kevent_user *u)
+{
+	unsigned long flags;
+	struct kevent *k;
+
+	spin_lock_irqsave(&u->ready_lock, flags);
+	k = __kqueue_dequeue_one_ready(&u->ready_list, &u->ready_num);
+	spin_unlock_irqrestore(&u->ready_lock, flags);
+
+	return k;
+}
+
+static struct kevent *__kevent_search(struct kevent_list *l, struct ukevent *uk, 
+		struct kevent_user *u)
+{
+	struct kevent *k;
+	int found = 0;
+	
+	list_for_each_entry(k, &l->kevent_list, kevent_entry) {
+		spin_lock(&k->lock);
+		if (k->event.user[0] == uk->user[0] && k->event.user[1] == uk->user[1] &&
+				k->event.id.raw[0] == uk->id.raw[0] && 
+				k->event.id.raw[1] == uk->id.raw[1]) {
+			found = 1;
+			spin_unlock(&k->lock);
+			break;
+		}
+		spin_unlock(&k->lock);
+	}
+
+	return (found)?k:NULL;
+}
+
+static int kevent_modify(struct ukevent *uk, struct kevent_user *u)
+{
+	struct kevent *k;
+	unsigned int hash = kevent_user_hash(uk);
+	struct kevent_list *l = &u->kqueue[hash];
+	int err = -ENODEV;
+	unsigned long flags;
+	
+	spin_lock_irqsave(&l->kevent_lock, flags);
+	k = __kevent_search(l, uk, u);
+	if (k) {
+		spin_lock(&k->lock);
+		k->event.event = uk->event;
+		k->event.req_flags = uk->req_flags;
+		k->event.ret_flags = 0;
+		spin_unlock(&k->lock);
+		kevent_requeue(k);
+		err = 0;
+	}
+	spin_unlock_irqrestore(&l->kevent_lock, flags);
+	
+	return err;
+}
+
+static int kevent_remove(struct ukevent *uk, struct kevent_user *u)
+{
+	int err = -ENODEV;
+	struct kevent *k;
+	unsigned int hash = kevent_user_hash(uk);
+	struct kevent_list *l = &u->kqueue[hash];
+	unsigned long flags;
+
+	spin_lock_irqsave(&l->kevent_lock, flags);
+	k = __kevent_search(l, uk, u);
+	if (k) {
+		kevent_finish_user(k, 0, 1);
+		err = 0;
+	}
+	spin_unlock_irqrestore(&l->kevent_lock, flags);
+
+	return err;
+}
+
+/*
+ * No new entry can be added or removed from any list at this point.
+ * It is not permitted to call ->ioctl() and ->release() in parallel.
+ */
+static int kevent_user_release(struct inode *inode, struct file *file)
+{
+	struct kevent_user *u = file->private_data;
+	struct kevent *k, *n;
+	int i;
+
+	for (i=0; i<KEVENT_HASH_MASK+1; ++i) {
+		struct kevent_list *l = &u->kqueue[i];
+		
+		list_for_each_entry_safe(k, n, &l->kevent_list, kevent_entry)
+			kevent_finish_user(k, 1, 1);
+	}
+
+	kevent_user_put(u);
+	file->private_data = NULL;
+
+	return 0;
+}
+
+static int kevent_user_ctl_modify(struct kevent_user *u, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	int err = 0, i;
+	struct ukevent uk;
+
+	if (down_interruptible(&u->ctl_mutex))
+		return -ERESTARTSYS;
+
+	for (i=0; i<ctl->num; ++i) {
+		if (copy_from_user(&uk, arg, sizeof(struct ukevent))) {
+			err = -EINVAL;
+			break;
+		}
+
+		if (kevent_modify(&uk, u))
+			uk.ret_flags |= KEVENT_RET_BROKEN;
+		uk.ret_flags |= KEVENT_RET_DONE;
+
+		if (copy_to_user(arg, &uk, sizeof(struct ukevent))) {
+			err = -EINVAL;
+			break;
+		}
+
+		arg += sizeof(struct ukevent);
+	}
+
+	up(&u->ctl_mutex);
+
+	return err;
+}
+
+static int kevent_user_ctl_remove(struct kevent_user *u, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	int err = 0, i;
+	struct ukevent uk;
+
+	if (down_interruptible(&u->ctl_mutex))
+		return -ERESTARTSYS;
+
+	for (i=0; i<ctl->num; ++i) {
+		if (copy_from_user(&uk, arg, sizeof(struct ukevent))) {
+			err = -EINVAL;
+			break;
+		}
+
+		if (kevent_remove(&uk, u))
+			uk.ret_flags |= KEVENT_RET_BROKEN;
+
+		uk.ret_flags |= KEVENT_RET_DONE;
+
+		if (copy_to_user(arg, &uk, sizeof(struct ukevent))) {
+			err = -EINVAL;
+			break;
+		}
+
+		arg += sizeof(struct ukevent);
+	}
+
+	up(&u->ctl_mutex);
+
+	return err;
+}
+
+int kevent_user_add_ukevent(struct ukevent *uk, struct kevent_user *u)
+{
+	struct kevent *k;
+	int err;
+
+	k = kevent_alloc(GFP_KERNEL);
+	if (!k) {
+		err = -ENOMEM;
+		goto err_out_exit;
+	}
+
+	memcpy(&k->event, uk, sizeof(struct ukevent));
+
+	k->event.ret_flags = 0;
+
+	err = kevent_init(k);
+	if (err) {
+		kevent_free(k);
+		goto err_out_exit;
+	}
+	k->user = u;
+#ifdef CONFIG_KEVENT_USER_STAT
+	u->total++;
+#endif
+	{
+		unsigned long flags;
+		unsigned int hash = kevent_user_hash(&k->event);
+		struct kevent_list *l = &u->kqueue[hash];
+		
+		spin_lock_irqsave(&l->kevent_lock, flags);
+		list_add_tail(&k->kevent_entry, &l->kevent_list);
+		u->kevent_num++;
+		kevent_user_get(u);
+		spin_unlock_irqrestore(&l->kevent_lock, flags);
+	}
+
+	err = kevent_enqueue(k);
+	if (err) {
+		memcpy(uk, &k->event, sizeof(struct ukevent));
+		if (err < 0)
+			uk->ret_flags |= KEVENT_RET_BROKEN;
+		uk->ret_flags |= KEVENT_RET_DONE;
+		kevent_finish_user(k, 1, 0);
+	} 
+
+err_out_exit:
+	return err;
+}
+
+/*
+ * Copy all ukevents from userspace, allocate kevent for each one 
+ * and add them into appropriate kevent_storages, 
+ * e.g. sockets, inodes and so on...
+ * If something goes wrong, all events will be dequeued and 
+ * negative error will be returned. 
+ * On success zero is returned and 
+ * ctl->num will be a number of finished events, either completed or failed. 
+ * Array of finished events (struct ukevent) will be placed behind 
+ * kevent_user_control structure. User must run through that array and check 
+ * ret_flags field of each ukevent structure to determine if it is fired or failed event.
+ */
+static int kevent_user_ctl_add(struct kevent_user *u, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	int err = 0, cerr = 0, num = 0, knum = 0, i;
+	void __user *orig, *ctl_addr;
+	struct ukevent uk;
+
+	if (down_interruptible(&u->ctl_mutex))
+		return -ERESTARTSYS;
+
+	orig = arg;
+	ctl_addr = arg - sizeof(struct kevent_user_control);
+#if 1
+	err = -ENFILE;
+	if (u->kevent_num + ctl->num >= 1024)
+		goto err_out_remove;
+#endif
+	for (i=0; i<ctl->num; ++i) {
+		if (copy_from_user(&uk, arg, sizeof(struct ukevent))) {
+			cerr = -EINVAL;
+			break;
+		}
+		arg += sizeof(struct ukevent);
+
+		err = kevent_user_add_ukevent(&uk, u);
+		if (err) {
+#ifdef CONFIG_KEVENT_USER_STAT
+			u->im_num++;
+#endif
+			if (copy_to_user(orig, &uk, sizeof(struct ukevent)))
+				cerr = -EINVAL;
+			orig += sizeof(struct ukevent);
+			num++;
+		} else
+			knum++;
+	}
+
+	if (cerr < 0)
+		goto err_out_remove;
+
+	ctl->num = num;
+	if (copy_to_user(ctl_addr, ctl, sizeof(struct kevent_user_control)))
+		cerr = -EINVAL;
+
+	if (cerr)
+		err = cerr;
+	if (!err)
+		err = num;
+
+err_out_remove:
+	up(&u->ctl_mutex);
+
+	return err;
+}
+
+/*
+ * Waits until at least ctl->ready_num events are ready or timeout and returns 
+ * number of ready events (in case of timeout) or number of requested events.
+ */
+static int kevent_user_wait(struct file *file, struct kevent_user *u, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	struct kevent *k;
+	int cerr = 0, num = 0;
+	void __user *ptr = arg + sizeof(struct kevent_user_control);
+
+	if (down_interruptible(&u->ctl_mutex))
+		return -ERESTARTSYS;
+
+	if (!(file->f_flags & O_NONBLOCK)) {
+		if (ctl->timeout)
+			wait_event_interruptible_timeout(u->wait, 
+				u->ready_num >= ctl->num, msecs_to_jiffies(ctl->timeout));
+		else
+			wait_event_interruptible_timeout(u->wait, 
+					u->ready_num > 0, msecs_to_jiffies(1000));
+	}
+	while (num < ctl->num && ((k = kqueue_dequeue_ready(u)) != NULL)) {
+		if (copy_to_user(ptr + num*sizeof(struct ukevent), 
+					&k->event, sizeof(struct ukevent)))
+			cerr = -EINVAL;
+
+		/*
+		 * If it is one-shot kevent, it has been removed already from
+		 * origin's queue, so we can easily free it here.
+		 */
+		if (k->event.req_flags & KEVENT_REQ_ONESHOT)
+			kevent_finish_user(k, 1, 1);
+		++num;
+#ifdef CONFIG_KEVENT_USER_STAT
+		u->wait_num++;
+#endif
+	}
+
+	ctl->num = num;
+	if (copy_to_user(arg, ctl, sizeof(struct kevent_user_control)))
+		cerr = -EINVAL;
+
+	up(&u->ctl_mutex);
+
+	return (cerr)?cerr:num;
+}
+
+static int kevent_ctl_init(void)
+{
+	struct kevent_user *u;
+	struct file *file;
+	int fd, ret;
+
+	fd = get_unused_fd();
+	if (fd < 0)
+		return fd;
+
+	file = get_empty_filp();
+	if (!file) {
+		ret = -ENFILE;
+		goto out_put_fd;
+	}
+
+	u = kevent_user_alloc();
+	if (unlikely(!u)) {
+		ret = -ENOMEM;
+		goto out_put_file;
+	}
+
+	file->f_op = &kevent_user_fops;
+	file->f_vfsmnt = mntget(kevent_mnt);
+	file->f_dentry = dget(kevent_mnt->mnt_root);
+	file->f_mapping = file->f_dentry->d_inode->i_mapping;
+	file->f_mode = FMODE_READ;
+	file->f_flags = O_RDONLY;
+	file->private_data = u;
+	
+	fd_install(fd, file);
+
+	return fd;
+
+out_put_file:
+	put_filp(file);
+out_put_fd:
+	put_unused_fd(fd);
+	return ret;
+}
+
+static int kevent_ctl_process(struct file *file, 
+		struct kevent_user_control *ctl, void __user *arg)
+{
+	int err;
+	struct kevent_user *u = file->private_data;
+
+	if (!u)
+		return -EINVAL;
+
+	switch (ctl->cmd) {
+		case KEVENT_CTL_ADD:
+			err = kevent_user_ctl_add(u, ctl, 
+					arg+sizeof(struct kevent_user_control));
+			break;
+		case KEVENT_CTL_REMOVE:
+			err = kevent_user_ctl_remove(u, ctl, 
+					arg+sizeof(struct kevent_user_control));
+			break;
+		case KEVENT_CTL_MODIFY:
+			err = kevent_user_ctl_modify(u, ctl, 
+					arg+sizeof(struct kevent_user_control));
+			break;
+		case KEVENT_CTL_WAIT:
+			err = kevent_user_wait(file, u, ctl, arg);
+			break;
+		case KEVENT_CTL_INIT:
+			err = kevent_ctl_init();
+		default:
+			err = -EINVAL;
+			break;
+	}
+
+	return err;
+}
+
+asmlinkage long sys_kevent_ctl(int fd, void __user *arg)
+{
+	int err, fput_needed;
+	struct kevent_user_control ctl;
+	struct file *file;
+
+	if (copy_from_user(&ctl, arg, sizeof(struct kevent_user_control)))
+		return -EINVAL;
+
+	if (ctl.cmd == KEVENT_CTL_INIT)
+		return kevent_ctl_init();
+
+	file = fget_light(fd, &fput_needed);
+	if (!file)
+		return -ENODEV;
+
+	err = kevent_ctl_process(file, &ctl, arg);
+
+	fput_light(file, fput_needed);
+	return err;
+}
+
+static int kevent_user_ioctl(struct inode *inode, struct file *file, 
+		unsigned int cmd, unsigned long arg)
+{
+	int err = -ENODEV;
+	struct kevent_user_control ctl;
+	struct kevent_user *u = file->private_data;
+	void __user *ptr = (void __user *)arg;
+
+	if (copy_from_user(&ctl, ptr, sizeof(struct kevent_user_control)))
+		return -EINVAL;
+
+	switch (cmd) {
+		case KEVENT_USER_CTL:
+			err = kevent_ctl_process(file, &ctl, ptr);
+			break;
+		case KEVENT_USER_WAIT:
+			err = kevent_user_wait(file, u, &ctl, ptr);
+			break;
+		default:
+			break;
+	}
+
+	return err;
+}
+
+static int __devinit kevent_user_init(void)
+{
+	struct class_device *dev;
+	int err = 0;
+	
+	err = register_filesystem(&kevent_fs_type);
+	if (err)
+		panic("%s: failed to register filesystem: err=%d.\n",
+			       kevent_name, err);
+
+	kevent_mnt = kern_mount(&kevent_fs_type);
+	if (IS_ERR(kevent_mnt))
+		panic("%s: failed to mount silesystem: err=%ld.\n", 
+				kevent_name, PTR_ERR(kevent_mnt));
+	
+	kevent_user_major = register_chrdev(0, kevent_name, &kevent_user_fops);
+	if (kevent_user_major < 0) {
+		printk(KERN_ERR "Failed to register \"%s\" char device: err=%d.\n", 
+				kevent_name, kevent_user_major);
+		return -ENODEV;
+	}
+
+	kevent_user_class = class_create(THIS_MODULE, "kevent");
+	if (IS_ERR(kevent_user_class)) {
+		printk(KERN_ERR "Failed to register \"%s\" class: err=%ld.\n", 
+				kevent_name, PTR_ERR(kevent_user_class));
+		err = PTR_ERR(kevent_user_class);
+		goto err_out_unregister;
+	}
+
+	dev = class_device_create(kevent_user_class, NULL, 
+			MKDEV(kevent_user_major, 0), NULL, kevent_name);
+	if (IS_ERR(dev)) {
+		printk(KERN_ERR "Failed to create %d.%d class device in \"%s\" class: err=%ld.\n", 
+				kevent_user_major, 0, kevent_name, PTR_ERR(dev));
+		err = PTR_ERR(dev);
+		goto err_out_class_destroy;
+	}
+
+	printk("KEVENT subsystem: chardev helper: major=%d.\n", kevent_user_major);
+
+	return 0;
+
+err_out_class_destroy:
+	class_destroy(kevent_user_class);
+err_out_unregister:
+	unregister_chrdev(kevent_user_major, kevent_name);
+
+	return err;
+}
+
+static void __devexit kevent_user_fini(void)
+{
+	class_device_destroy(kevent_user_class, MKDEV(kevent_user_major, 0));
+	class_destroy(kevent_user_class);
+	unregister_chrdev(kevent_user_major, kevent_name);
+	mntput(kevent_mnt);
+	unregister_filesystem(&kevent_fs_type);
+}
+
+module_init(kevent_user_init);
+module_exit(kevent_user_fini);
diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c
index 5433195..dcbacf5 100644
--- a/kernel/sys_ni.c
+++ b/kernel/sys_ni.c
@@ -121,6 +121,11 @@ cond_syscall(ppc_rtas);
 cond_syscall(sys_spu_run);
 cond_syscall(sys_spu_create);
 
+cond_syscall(sys_aio_recv);
+cond_syscall(sys_aio_send);
+cond_syscall(sys_aio_sendfile);
+cond_syscall(sys_kevent_ctl);
+
 /* mmu depending weak syscall entries */
 cond_syscall(sys_mprotect);
 cond_syscall(sys_msync);
diff --git a/net/core/datagram.c b/net/core/datagram.c
index aecddcc..493245b 100644
--- a/net/core/datagram.c
+++ b/net/core/datagram.c
@@ -236,6 +236,60 @@ void skb_kill_datagram(struct sock *sk, 
 EXPORT_SYMBOL(skb_kill_datagram);
 
 /**
+ *	skb_copy_datagram - Copy a datagram.
+ *	@skb: buffer to copy
+ *	@offset: offset in the buffer to start copying from
+ *	@to: pointer to copy to
+ *	@len: amount of data to copy from buffer to iovec
+ */
+int skb_copy_datagram(const struct sk_buff *skb, int offset,
+			    void *to, int len)
+{
+	int i, fraglen, end = 0;
+	struct sk_buff *next = skb_shinfo(skb)->frag_list;
+
+	if (!len)
+		return 0;
+
+next_skb:
+	fraglen = skb_headlen(skb);
+	i = -1;
+
+	while (1) {
+		int start = end;
+
+		if ((end += fraglen) > offset) {
+			int copy = end - offset, o = offset - start;
+
+			if (copy > len)
+				copy = len;
+			if (i == -1)
+				memcpy(to, skb->data + o, copy);
+			else {
+				skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
+				struct page *page = frag->page;
+				void *p = kmap(page) + frag->page_offset + o;
+				memcpy(to, p, copy);
+				kunmap(page);
+			}
+			if (!(len -= copy))
+				return 0;
+			offset += copy;
+		}
+		if (++i >= skb_shinfo(skb)->nr_frags)
+			break;
+		fraglen = skb_shinfo(skb)->frags[i].size;
+	}
+	if (next) {
+		skb = next;
+		BUG_ON(skb_shinfo(skb)->frag_list);
+		next = skb->next;
+		goto next_skb;
+	}
+	return -EFAULT;
+}
+
+/**
  *	skb_copy_datagram_iovec - Copy a datagram to an iovec.
  *	@skb: buffer to copy
  *	@offset: offset in the buffer to start copying from
@@ -530,6 +584,7 @@ unsigned int datagram_poll(struct file *
 
 EXPORT_SYMBOL(datagram_poll);
 EXPORT_SYMBOL(skb_copy_and_csum_datagram_iovec);
+EXPORT_SYMBOL(skb_copy_datagram);
 EXPORT_SYMBOL(skb_copy_datagram_iovec);
 EXPORT_SYMBOL(skb_free_datagram);
 EXPORT_SYMBOL(skb_recv_datagram);
diff --git a/net/core/sock.c b/net/core/sock.c
index 5d820c3..3345048 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -564,6 +564,16 @@ #endif
 			spin_unlock_bh(&sk->sk_lock.slock);
 			ret = -ENONET;
 			break;
+#ifdef CONFIG_KEVENT_SOCKET
+		case SO_ASYNC_SOCK:
+			spin_lock_bh(&sk->sk_lock.slock);
+			if (valbool)
+				sock_set_flag(sk, SOCK_ASYNC);
+			else
+				sock_reset_flag(sk, SOCK_ASYNC);
+			spin_unlock_bh(&sk->sk_lock.slock);
+			break;
+#endif
 
 		/* We implement the SO_SNDLOWAT etc to
 		   not be settable (1003.1g 5.3) */
@@ -1313,6 +1323,7 @@ static void sock_def_wakeup(struct sock 
 	if (sk->sk_sleep && waitqueue_active(sk->sk_sleep))
 		wake_up_interruptible_all(sk->sk_sleep);
 	read_unlock(&sk->sk_callback_lock);
+	kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
 }
 
 static void sock_def_error_report(struct sock *sk)
@@ -1322,6 +1333,7 @@ static void sock_def_error_report(struct
 		wake_up_interruptible(sk->sk_sleep);
 	sk_wake_async(sk,0,POLL_ERR); 
 	read_unlock(&sk->sk_callback_lock);
+	kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
 }
 
 static void sock_def_readable(struct sock *sk, int len)
@@ -1331,6 +1343,7 @@ static void sock_def_readable(struct soc
 		wake_up_interruptible(sk->sk_sleep);
 	sk_wake_async(sk,1,POLL_IN);
 	read_unlock(&sk->sk_callback_lock);
+	kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
 }
 
 static void sock_def_write_space(struct sock *sk)
@@ -1350,6 +1363,7 @@ static void sock_def_write_space(struct 
 	}
 
 	read_unlock(&sk->sk_callback_lock);
+	kevent_socket_notify(sk, KEVENT_SOCKET_SEND|KEVENT_SOCKET_RECV);
 }
 
 static void sock_def_destruct(struct sock *sk)
@@ -1454,8 +1468,10 @@ void fastcall release_sock(struct sock *
 	if (sk->sk_backlog.tail)
 		__release_sock(sk);
 	sk->sk_lock.owner = NULL;
-        if (waitqueue_active(&(sk->sk_lock.wq)))
+        if (waitqueue_active(&(sk->sk_lock.wq))) {
 		wake_up(&(sk->sk_lock.wq));
+		kevent_socket_notify(sk, KEVENT_SOCKET_RECV|KEVENT_SOCKET_SEND);
+	}
 	spin_unlock_bh(&(sk->sk_lock.slock));
 }
 EXPORT_SYMBOL(release_sock);
diff --git a/net/core/stream.c b/net/core/stream.c
index e948969..91e2e07 100644
--- a/net/core/stream.c
+++ b/net/core/stream.c
@@ -36,6 +36,7 @@ void sk_stream_write_space(struct sock *
 			wake_up_interruptible(sk->sk_sleep);
 		if (sock->fasync_list && !(sk->sk_shutdown & SEND_SHUTDOWN))
 			sock_wake_async(sock, 2, POLL_OUT);
+		kevent_socket_notify(sk, KEVENT_SOCKET_SEND|KEVENT_SOCKET_RECV);
 	}
 }
 
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 74998f2..403d33e 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -206,6 +206,7 @@
  *					lingertime == 0 (RFC 793 ABORT Call)
  *	Hirokazu Takahashi	:	Use copy_from_user() instead of
  *					csum_and_copy_from_user() if possible.
+ *	Evgeniy Polyakov	:	Network asynchronous IO.
  *
  *		This program is free software; you can redistribute it and/or
  *		modify it under the terms of the GNU General Public License
@@ -1085,6 +1086,275 @@ int tcp_read_sock(struct sock *sk, read_
 }
 
 /*
+ * Must be called with locked sock.
+ */
+int tcp_async_send(struct sock *sk, struct page **pages, unsigned int poffset, size_t len)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+	int mss_now, size_goal;
+	int err = -EAGAIN;
+	ssize_t copied;
+
+	/* Wait for a connection to finish. */
+	if ((1 << sk->sk_state) & ~(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT))
+		goto out_err;
+
+	clear_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags);
+
+	mss_now = tcp_current_mss(sk, 1);
+	size_goal = tp->xmit_size_goal;
+	copied = 0;
+
+	err = -EPIPE;
+	if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN) || sock_flag(sk, SOCK_DONE) ||
+			(sk->sk_state == TCP_CLOSE) || (atomic_read(&sk->sk_refcnt) == 1))
+		goto do_error;
+
+	while (len > 0) {
+		struct sk_buff *skb = sk->sk_write_queue.prev;
+		struct page *page = pages[poffset / PAGE_SIZE];
+		int copy, i, can_coalesce;
+		int offset = poffset % PAGE_SIZE;
+		int size = min_t(size_t, len, PAGE_SIZE - offset);
+
+		if (!sk->sk_send_head || (copy = size_goal - skb->len) <= 0) {
+new_segment:
+			if (!sk_stream_memory_free(sk))
+				goto wait_for_sndbuf;
+
+			skb = sk_stream_alloc_pskb(sk, 0, 0,
+						   sk->sk_allocation);
+			if (!skb)
+				goto wait_for_memory;
+
+			skb_entail(sk, tp, skb);
+			copy = size_goal;
+		}
+
+		if (copy > size)
+			copy = size;
+
+		i = skb_shinfo(skb)->nr_frags;
+		can_coalesce = skb_can_coalesce(skb, i, page, offset);
+		if (!can_coalesce && i >= MAX_SKB_FRAGS) {
+			tcp_mark_push(tp, skb);
+			goto new_segment;
+		}
+		if (!sk_stream_wmem_schedule(sk, copy))
+			goto wait_for_memory;
+		
+		if (can_coalesce) {
+			skb_shinfo(skb)->frags[i - 1].size += copy;
+		} else {
+			get_page(page);
+			skb_fill_page_desc(skb, i, page, offset, copy);
+		}
+
+		skb->len += copy;
+		skb->data_len += copy;
+		skb->truesize += copy;
+		sk->sk_wmem_queued += copy;
+		sk->sk_forward_alloc -= copy;
+		skb->ip_summed = CHECKSUM_HW;
+		tp->write_seq += copy;
+		TCP_SKB_CB(skb)->end_seq += copy;
+		skb_shinfo(skb)->tso_segs = 0;
+
+		if (!copied)
+			TCP_SKB_CB(skb)->flags &= ~TCPCB_FLAG_PSH;
+
+		copied += copy;
+		poffset += copy;
+		if (!(len -= copy))
+			goto out;
+
+		if (skb->len < mss_now)
+			continue;
+
+		if (forced_push(tp)) {
+			tcp_mark_push(tp, skb);
+			__tcp_push_pending_frames(sk, tp, mss_now, TCP_NAGLE_PUSH);
+		} else if (skb == sk->sk_send_head)
+			tcp_push_one(sk, mss_now);
+		continue;
+
+wait_for_sndbuf:
+		set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
+wait_for_memory:
+		if (copied)
+			tcp_push(sk, tp, 0, mss_now, TCP_NAGLE_PUSH);
+
+		err = -EAGAIN;
+		goto do_error;
+	}
+
+out:
+	if (copied)
+		tcp_push(sk, tp, 0, mss_now, tp->nonagle);
+	return copied;
+
+do_error:
+	if (copied)
+		goto out;
+out_err:
+	return sk_stream_error(sk, 0, err);
+}
+
+/*
+ * Must be called with locked sock.
+ */
+int tcp_async_recv(struct sock *sk, void *dst, size_t len)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+	int copied = 0;
+	u32 *seq;
+	unsigned long used;
+	int err;
+	int target;		/* Read at least this many bytes */
+
+	TCP_CHECK_TIMER(sk);
+
+	err = -ENOTCONN;
+	if (sk->sk_state == TCP_LISTEN)
+		goto out;
+
+	seq = &tp->copied_seq;
+
+	target = sock_rcvlowat(sk, 0, len);
+
+	do {
+		struct sk_buff *skb;
+		u32 offset;
+
+		/* Are we at urgent data? Stop if we have read anything or have SIGURG pending. */
+		if (tp->urg_data && tp->urg_seq == *seq) {
+			if (copied)
+				break;
+		}
+
+		/* Next get a buffer. */
+
+		skb = skb_peek(&sk->sk_receive_queue);
+		do {
+			if (!skb)
+				break;
+
+			/* Now that we have two receive queues this
+			 * shouldn't happen.
+			 */
+			if (before(*seq, TCP_SKB_CB(skb)->seq)) {
+				printk(KERN_INFO "async_recv bug: copied %X "
+				       "seq %X\n", *seq, TCP_SKB_CB(skb)->seq);
+				break;
+			}
+			offset = *seq - TCP_SKB_CB(skb)->seq;
+			if (skb->h.th->syn)
+				offset--;
+			if (offset < skb->len)
+				goto found_ok_skb;
+			if (skb->h.th->fin)
+				goto found_fin_ok;
+			skb = skb->next;
+		} while (skb != (struct sk_buff *)&sk->sk_receive_queue);
+
+		if (copied)
+			break;
+
+		if (sock_flag(sk, SOCK_DONE))
+			break;
+
+		if (sk->sk_err) {
+			copied = sock_error(sk);
+			break;
+		}
+
+		if (sk->sk_shutdown & RCV_SHUTDOWN)
+			break;
+
+		if (sk->sk_state == TCP_CLOSE) {
+			if (!sock_flag(sk, SOCK_DONE)) {
+				/* This occurs when user tries to read
+				 * from never connected socket.
+				 */
+				copied = -ENOTCONN;
+				break;
+			}
+			break;
+		}
+
+		copied = -EAGAIN;
+		break;
+
+	found_ok_skb:
+		/* Ok so how much can we use? */
+		used = skb->len - offset;
+		if (len < used)
+			used = len;
+
+		/* Do we have urgent data here? */
+		if (tp->urg_data) {
+			u32 urg_offset = tp->urg_seq - *seq;
+			if (urg_offset < used) {
+				if (!urg_offset) {
+					if (!sock_flag(sk, SOCK_URGINLINE)) {
+						++*seq;
+						offset++;
+						used--;
+						if (!used)
+							goto skip_copy;
+					}
+				} else
+					used = urg_offset;
+			}
+		}
+
+		err = skb_copy_datagram(skb, offset, dst, used);
+		if (err) {
+			/* Exception. Bailout! */
+			if (!copied)
+				copied = -EFAULT;
+			break;
+		}
+
+		*seq += used;
+		copied += used;
+		len -= used;
+		dst += used;
+
+		tcp_rcv_space_adjust(sk);
+
+skip_copy:
+		if (tp->urg_data && after(tp->copied_seq, tp->urg_seq)) {
+			tp->urg_data = 0;
+			tcp_fast_path_check(sk, tp);
+		}
+		if (used + offset < skb->len)
+			continue;
+
+		if (skb->h.th->fin)
+			goto found_fin_ok;
+		sk_eat_skb(sk, skb);
+		continue;
+
+	found_fin_ok:
+		/* Process the FIN. */
+		++*seq;
+		sk_eat_skb(sk, skb);
+		break;
+	} while (len > 0);
+
+	/* Clean up data we have read: This will do ACK frames. */
+	cleanup_rbuf(sk, copied);
+
+	TCP_CHECK_TIMER(sk);
+	return copied;
+
+out:
+	TCP_CHECK_TIMER(sk);
+	return err;
+}
+
+/*
  *	This routine copies from a sock struct into the user buffer.
  *
  *	Technical note: in 2.3 we work on _locked_ socket, so that
@@ -2259,6 +2529,8 @@ EXPORT_SYMBOL(tcp_getsockopt);
 EXPORT_SYMBOL(tcp_ioctl);
 EXPORT_SYMBOL(tcp_poll);
 EXPORT_SYMBOL(tcp_read_sock);
+EXPORT_SYMBOL(tcp_async_recv);
+EXPORT_SYMBOL(tcp_async_send);
 EXPORT_SYMBOL(tcp_recvmsg);
 EXPORT_SYMBOL(tcp_sendmsg);
 EXPORT_SYMBOL(tcp_sendpage);
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index e08245b..5655b1e 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -3113,6 +3113,7 @@ static void tcp_ofo_queue(struct sock *s
 
 		__skb_unlink(skb, &tp->out_of_order_queue);
 		__skb_queue_tail(&sk->sk_receive_queue, skb);
+		kevent_socket_notify(sk, KEVENT_SOCKET_RECV);
 		tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq;
 		if(skb->h.th->fin)
 			tcp_fin(skb, sk, skb->h.th);
@@ -3956,7 +3957,8 @@ int tcp_rcv_established(struct sock *sk,
 			int copied_early = 0;
 
 			if (tp->copied_seq == tp->rcv_nxt &&
-			    len - tcp_header_len <= tp->ucopy.len) {
+			    len - tcp_header_len <= tp->ucopy.len &&
+			    !sock_async(sk)) {
 #ifdef CONFIG_NET_DMA
 				if (tcp_dma_try_early_copy(sk, skb, tcp_header_len)) {
 					copied_early = 1;
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 25ecc6e..05d7086 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -62,6 +62,7 @@ #include <linux/cache.h>
 #include <linux/jhash.h>
 #include <linux/init.h>
 #include <linux/times.h>
+#include <linux/kevent.h>
 
 #include <net/icmp.h>
 #include <net/inet_hashtables.h>
@@ -850,6 +851,7 @@ #endif
 	   	reqsk_free(req);
 	} else {
 		inet_csk_reqsk_queue_hash_add(sk, req, TCP_TIMEOUT_INIT);
+		kevent_socket_notify(sk, KEVENT_SOCKET_ACCEPT);
 	}
 	return 0;
 
@@ -1089,24 +1091,30 @@ process:
 
 	skb->dev = NULL;
 
-	bh_lock_sock(sk);
 	ret = 0;
-	if (!sock_owned_by_user(sk)) {
+	if (sock_async(sk)) {
+		spin_lock_bh(&sk->sk_lock.slock);
+		ret = tcp_v4_do_rcv(sk, skb);
+		spin_unlock_bh(&sk->sk_lock.slock);
+	} else {
+		bh_lock_sock(sk);
+		if (!sock_owned_by_user(sk)) {
 #ifdef CONFIG_NET_DMA
-		struct tcp_sock *tp = tcp_sk(sk);
-		if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
-			tp->ucopy.dma_chan = get_softnet_dma();
-		if (tp->ucopy.dma_chan)
-			ret = tcp_v4_do_rcv(sk, skb);
-		else
+			struct tcp_sock *tp = tcp_sk(sk);
+			if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
+				tp->ucopy.dma_chan = get_softnet_dma();
+			if (tp->ucopy.dma_chan)
+				ret = tcp_v4_do_rcv(sk, skb);
+			else
 #endif
-		{
-			if (!tcp_prequeue(sk, skb))
-			ret = tcp_v4_do_rcv(sk, skb);
-		}
-	} else
-		sk_add_backlog(sk, skb);
-	bh_unlock_sock(sk);
+			{
+				if (!tcp_prequeue(sk, skb))
+				ret = tcp_v4_do_rcv(sk, skb);
+			}
+		} else
+			sk_add_backlog(sk, skb);
+		bh_unlock_sock(sk);
+	}
 
 	sock_put(sk);
 
@@ -1830,6 +1838,8 @@ struct proto tcp_prot = {
 	.getsockopt		= tcp_getsockopt,
 	.sendmsg		= tcp_sendmsg,
 	.recvmsg		= tcp_recvmsg,
+	.async_recv		= tcp_async_recv,
+	.async_send		= tcp_async_send,
 	.backlog_rcv		= tcp_v4_do_rcv,
 	.hash			= tcp_v4_hash,
 	.unhash			= tcp_unhash,
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index a50eb30..e27e231 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1215,22 +1215,28 @@ process:
 
 	skb->dev = NULL;
 
-	bh_lock_sock(sk);
 	ret = 0;
-	if (!sock_owned_by_user(sk)) {
+	if (sock_async(sk)) {
+		spin_lock_bh(&sk->sk_lock.slock);
+		ret = tcp_v4_do_rcv(sk, skb);
+		spin_unlock_bh(&sk->sk_lock.slock);
+	} else {
+		bh_lock_sock(sk);
+		if (!sock_owned_by_user(sk)) {
 #ifdef CONFIG_NET_DMA
-                struct tcp_sock *tp = tcp_sk(sk);
-                if (tp->ucopy.dma_chan)
-                        ret = tcp_v6_do_rcv(sk, skb);
-                else
-#endif
-		{
-			if (!tcp_prequeue(sk, skb))
+			struct tcp_sock *tp = tcp_sk(sk);
+			if (tp->ucopy.dma_chan)
 				ret = tcp_v6_do_rcv(sk, skb);
-		}
-	} else
-		sk_add_backlog(sk, skb);
-	bh_unlock_sock(sk);
+			else
+#endif
+			{
+				if (!tcp_prequeue(sk, skb))
+					ret = tcp_v6_do_rcv(sk, skb);
+			}
+		} else
+			sk_add_backlog(sk, skb);
+		bh_unlock_sock(sk);
+	}
 
 	sock_put(sk);
 	return ret ? -1 : 0;
@@ -1580,6 +1586,8 @@ struct proto tcpv6_prot = {
 	.getsockopt		= tcp_getsockopt,
 	.sendmsg		= tcp_sendmsg,
 	.recvmsg		= tcp_recvmsg,
+	.async_recv		= tcp_async_recv,
+	.async_send		= tcp_async_send,
 	.backlog_rcv		= tcp_v6_do_rcv,
 	.hash			= tcp_v6_hash,
 	.unhash			= tcp_unhash,

Re: [1/4] kevent: core files.

[Andrew Morton]

On Wed, 26 Jul 2006 13:18:15 +0400
Evgeniy Polyakov <johnpol@2ka.mipt.ru> wrote:

> +static int kevent_ctl_process(struct file *file, 
> +		struct kevent_user_control *ctl, void __user *arg)
> +{
> +	int err;
> +	struct kevent_user *u = file->private_data;
> +
> +	if (!u)
> +		return -EINVAL;
> +
> +	switch (ctl->cmd) {
> +		case KEVENT_CTL_ADD:
> +			err = kevent_user_ctl_add(u, ctl, 
> +					arg+sizeof(struct kevent_user_control));
> +			break;
> +		case KEVENT_CTL_REMOVE:
> +			err = kevent_user_ctl_remove(u, ctl, 
> +					arg+sizeof(struct kevent_user_control));
> +			break;
> +		case KEVENT_CTL_MODIFY:
> +			err = kevent_user_ctl_modify(u, ctl, 
> +					arg+sizeof(struct kevent_user_control));
> +			break;
> +		case KEVENT_CTL_WAIT:
> +			err = kevent_user_wait(file, u, ctl, arg);
> +			break;
> +		case KEVENT_CTL_INIT:
> +			err = kevent_ctl_init();
> +		default:
> +			err = -EINVAL;
> +			break;
> +	}
> +
> +	return err;
> +}

Please indent the body of the switch one tabstop to the left.

> +asmlinkage long sys_kevent_ctl(int fd, void __user *arg)
> +{
> +	int err, fput_needed;
> +	struct kevent_user_control ctl;
> +	struct file *file;
> +
> +	if (copy_from_user(&ctl, arg, sizeof(struct kevent_user_control)))
> +		return -EINVAL;
> +
> +	if (ctl.cmd == KEVENT_CTL_INIT)
> +		return kevent_ctl_init();
> +
> +	file = fget_light(fd, &fput_needed);
> +	if (!file)
> +		return -ENODEV;
> +
> +	err = kevent_ctl_process(file, &ctl, arg);
> +
> +	fput_light(file, fput_needed);
> +	return err;
> +}

If the user passes this an fd which was obtained via means other than
kevent_ctl_init(), the kernel will explode.  Do

	if (file->f_fop != &kevent_user_fops)
		return -EINVAL;

Re: [1/4] kevent: core files.

[Evgeniy Polyakov]

On Wed, Jul 26, 2006 at 03:31:05AM -0700, Andrew Morton (akpm@osdl.org) wrote:
> Please indent the body of the switch one tabstop to the left.
..
> If the user passes this an fd which was obtained via means other than
> kevent_ctl_init(), the kernel will explode.  Do
> 
> 	if (file->f_fop != &kevent_user_fops)
> 		return -EINVAL;

Thanks, I will implement both.

-- 
	Evgeniy Polyakov

Re: [1/4] kevent: core files.

[Evgeniy Polyakov]

On Wed, Jul 26, 2006 at 01:18:15PM +0400, Evgeniy Polyakov (johnpol@2ka.mipt.ru) wrote:
> +struct kevent *kevent_alloc(gfp_t mask)
> +{
> +	struct kevent *k;
> +	
> +	if (kevent_cache)
> +		k = kmem_cache_alloc(kevent_cache, mask);
> +	else
> +		k = kzalloc(sizeof(struct kevent), mask);
> +
> +	return k;
> +}
> +

Sorry for that.
It is fixed already to always use cache, but I forget to commit that
change before I created pachset.

-- 
	Evgeniy Polyakov