[RFC PATCH 0/21] relay revamp v6

[Tom Zanussi]

Here's the last version of the 'relay revamp' patchset.  When I started
out making these changes, I said I'd continue until I got rid of all the
unnecessary requirements and assumptions that hindered the current
relay, and see where it ended up.  At this point there's nothing else I
can think of to remove or simplify, so this patchset basically
represents where it 'ended up'.

In the end, I think the exercise was worth it - relay ends up with a new
design that I think is much cleaner and simpler, and maybe because of
that, more powerful.

Besides simplifying relay in a way that makes a lot of sense on its own,
I think it might also represent a new way of tracing.  The main idea
this patchset ends up enabling and supporting is the view that if trace
data is nothing more than independent and self-contained pages of data,
a tracer's job could be to simply fill up pages and send them off to
some other component that really doesn't do anything but provide a way
to relay them to userspace, which is exactly what this version of relay
does.  The 'new' part is that unlike the current relay (or other
existing tracers) the pages aren't bound up and tied to an actual
buffer, but are free to be 'peeled off' and sent elsewhere when filled.
This fits nicely with the capabilities of splice(2), especially if
SPLICE_F_MOVE is used - in that case, the trace data would move directly
from the tracer to the destination without any other copies being made.
If each page (or contiguous set of pages) was indeed self-contained, and
there was a common timestamp for all events, multiple tracers could
meaningfully send their pages off to a single relay channel and a
demultiplexing program could sort the events based on the common
timestamp and dispatch each to a special-purpose event formatter
tailored to each event type.  This was something I had actually started
working on for the old relay (libutt + tracer-specific plugins).
Another interesting possibility splice-based tracing opens up would be
merging trace events generated in userspace with the kernel trace data,
via vmsplice (also zero-copy if SPLICE_F_GIFT is used) and again tying
it together in post-processing if a common timestamp was used.

All of the above, with the exception of the userspace tracing part
(which would require a splice_write() implementation for relay), should
be possible with this current patchset.  That said, current users don't
need or use any of that, which doesn't matter since the new version
retains all of the functionality of the original relay and in fact
current users such as blktrace still work without any changes to their
userspace code.  Also, it should be noted that read(2) is still
supported and the pagewriter write (but not reserve) functions don't
create 'self-contained' pages, but split events across pages, so it's
still easy to do printk-style tracing and use cat on the relay files,
without needing any special userspace program.

I've done some pretty heavy pounding on the basic stuff via blktrace,
which seems to still perform well and without any problems, but other
things still need more testing.

I'm posting the complete set of patches mainly to show exactly how I
morphed the old relay into the new through a progression of discrete
changes aimed at removing dependencies on certain requirements and
assumptions; some of them are canceled out by subsequent patches, but in
fact you should be able to stop at any point in the series and have a
fully functional and improved relay.  I'm also appending the end-result
relay and relay_pagewriter at the end of this mail since it may be hard
to see the result through all the individual changes.

overview
========

Basically, the overall effect of the patchset is to split the current
relay into two separate and completely independent pieces: relay and
relay_pagewriter.  Here's a one-line description of each piece:

relay - the new relay simply accepts pages of trace data from anywhere
in the kernel and collects them into per-cpu lists which can be sent by
userspace to disk or network destinations via standard read(2),
splice(2) and sendfile(2) file operations, after which the pages are
returned to their owner(s).

relay_pagewriter -  provides a small set of page-writing functions that
allow tracers to write into pages it grabs one-by-one from a per-cpu
pool of pages and sends them to relay when each is completely written.

There are of course other features e.g. if you don't want to deal with
multiple per-cpu files, there's also a 'global' mode, as there always
was.  relay_pagewriter also supports a 'flight-recorder' mode where the
pages aren't sent to relay but rather returned to the end of the pool;
at some later point they can be sent to relay via a simple function
designed for that purpose.  Similarly, the 'early tracing/late setup'
functionality has been retained, but is now much simpler in
implementation; the fact that these and other features just seem to
'fall out' of the implementation seems to point to the end design being
a little nicer.

Note that because of this separation, there's nothing special about
relay_pagewriter - it becomes just another source of trace data (via an
actual tracer such as blktrace).

Here's an overview of relay and relay_pagewriter:

relay
=====

With the new changes, relay really does become just what its name says
and and nothing more - it accepts pages from tracers, and relays the
data to userspace via read(2) or splice(2) (and therefore sendfile(2)).
It doesn't allocate any buffer space and provides no write functions -
those are expected to be supplied by any tracer that might want relay
pages of trace data to userspace.

After applying the full patch you should be able to use it to stream
your trace data to disk or over the network...

Anyway, here's a brief overview of the new API (see code for details):

- relay_open():

Creates a per-cpu relay channel and by default associates debugfs files
with each per-cpu 'buffer'.  No buffer space is allocated for the
'buffers', rather they collect pages added by tracers in a list which is
drained by read()/splice(), etc.  Tracers add pages to the 'buffers'
using relay_write_page() and relay_write_pages().  One of the parameters
to relay_open() is n_pages_wakeup, which specifies that readers should
be woken up every time n_pages have been added; if this is 0, readers
are never woken up.

- relay_add_page():

Adds a page of trace data to relay.  After it has been consumed by
userspace, the tracer is notified by the 'relay page' callback function
page_released().  The page passed via the callback can then be re-used
by the tracer (see for example the pagewriter code, which simply adds
the page back into pagewriter's per-cpu page pool).  If the page has
been stolen instead (if SPLICE_F_MOVE succeeded, which can't happen in
current kernels since support for it isn't there), the page_stolen()
callback is called, at which point the tracer can allocate a new page to
replace the stolen page (see the pagewriter code, which does this too).

- relay_add_pages():

The same as relay_add_page(), but adds a set of pages to relay and
guarantees that they'll stay together and remain in the same order they
were added.

- relay_close():

Releases unread pages to the tracer(s) and frees the channel.

- relay_flush():

Wakes up readers.

- relay_reset():

Releases unread pages to the tracer(s) and resets the channel state.

That's basically the entire kernel API for relay; the userspace API is
of course just read(), splice(), and sendfile().


relay_pagewriter
================

One example of a component that can write pages of trace data and send
them to relay would be the original relay write functions and buffers
(the no-vmap page-based versions), which have been split out into a new
file called relay_pagewriter.c and provide one means of writing into
pages and feeding them into relay.  blktrace and kvmtrace have been
'ported' over to using pagewriter instead of relay directly.

- pagewriter_open():

Allocates a pool of pages for each cpu and creates a relay channel to
send them to.  Whenever a new page is needed for writing into, the
pagewriter object pulls a page out of the appropriate per-cpu pool and
makes it available to tracers to write/reserve into via the
pagewriter_write() and pagewriter_reserve() functions.  When a page is
full, the pagewriter object sends the page to relay via relay_add_page()
and grabs another page for the next writes.  When relay has finished
relaying the pages to userspace, it returns the page via pagewriter's
page_released() or page_stolen() callbacks; pagewriter simply returns
(and allocates a new page if stolen) the page to the pool.

The number of pages allocated per-cpu is specified by the n_pages param.
The n_pages_wakeup param specifies that readers should be woken up every
n_pages_wakeup pages; pagewriter doesn't actually do anything with this
other than pass it on to the relay channel.

The flags param to pagewriter_open() is also passed on to relay_open();
the bottom half is used by relay for its flags and the top half by
pagewriter.  Here are the pagewriter flags and what they mean:

- PAGEWRITER_PAD_WRITES - this means that the writer doesn't want writes
split across pages - if the write won't fit completely within the space
remaining in the page, the event will be written in the next page, but
before that happens, the write_padding() pagewriter callback is called
to give the tracer a chance to write a 'padding event' in the unused
space.  Currently, this flag only make sense for a tracer using
pagewriter_reserve() and must be set if using pagewriter_reserve()
(since there's no way to reserve across pages).  Also currently,
pagewriter_write() always writes across pages, and for tracers using
pagewriter_write(), this flag should not be set.

- PAGEWRITER_FLIGHT_MODE - if this flag is set, pagewriter won't send
pages to relay as they become full, but will simply send them to the end
of the appropriate per-cpu pool and keep track of how many have been
written.  This continues until the tracer explicitly requests that they
be sent to relay and made available to userspace.  The function that
does this is pagewriter_save_flight_data().  After the pages have been
read by userspace, they're returned to pagewriter as usual.

- PAGEWRITER_LATE_SETUP - if this flag is set, pagewriter_open() won't
create the relay channel until pagewriter_late_setup() is called.
Before that call is made, written pages are tracked in the per-cpu pool
as in the PAGEWRITER_FLIGHT_MODE case and similarly, but after the relay
channel has been created by pagewriter_late_setup(), the pages are sent
to relay and on to userspace (and returned) as usual.

pagewriter callbacks
--------------------

The pagewriter_callbacks param of pagewriter_open() refers to the three
available pagewriter callbacks, none of which are required:

- new_page() - gives the tracer a chance to write page headers, track
pages, etc.

- switch_page() - allows a tracer to define its own switch_page()
'slow-path' function and therefore gain control over the complete write
path.  Wouldn't typically be used by normal tracers.

- write_padding() - called when PAGEWRITER_PAD_WRITES is used - allows
the tracer to write a 'padding event' if a normal event won't fit.
Currently only makes sense when using pagewriter_reserve().

write/reserve functions
-----------------------

- pagewriter_reserve(pagewriter, length, end_reserve):

Reserve length bytes in the current page.  If length bytes won't fit
into the current page, the reserve happens in the next page.  If length
bytes won't fit, at least end_reserve bytes are guaranteed to be
available at the end of the page for the tracer to write a 'padding
event' that presumably will fit in end_reserve bytes.  The tracer is
given an opportunity to write its padding event via the write_padding()
callback, which is passed a pointer to start of padding along with the
length of the padding.  Currently, if using pagewriter_reserve(), the
pagewriter should also be opened using the PAGEWRITER_PAD_WRITES flag.

- pagewriter_write(pagewriter, data, length):

Write length bytes into the current page.  If the event won't fit into
the current page, whatever will fit is written into the current page,
and the remainder is written into the next page.  If there's no page
available for the remainder, the first part won't be written either.
pagewriter_write() disables irqs for synchronization;
__pagewriter_write() is essentially the same, but only disables
preemption.

misc functions
--------------

- pagewriter_flush():

Switches all partially written pages and calls relay_flush().

- pagewriter_close():

Frees the per-cpu pools, closes the associated relay channel, and
destroys the pagewriter.

- pagewriter_reset():

Resets the pagewriter to its initial state.

Below is the current relay/relay_pagewriter code after applying all
patches in the patchset, for convenience.

Tom


--- /dev/null	2007-10-15 18:18:04.000000000 -0500
+++ include/linux/relay.h	2008-10-14 20:28:47.000000000 -0500
@@ -0,0 +1,170 @@
+/*
+ * linux/include/linux/relay.h
+ *
+ * Copyright (C) 2002, 2003 - Tom Zanussi (zanussi@xxxxxxxxxx), IBM Corp
+ * Copyright (C) 1999, 2000, 2001, 2002 - Karim Yaghmour (karim@xxxxxxxxxxx)
+ * Copyright (C) 2008 - Tom Zanussi (tzanussi@xxxxxxxxx)
+ *
+ * CONFIG_RELAY definitions and declarations
+ */
+
+#ifndef _LINUX_RELAY_H
+#define _LINUX_RELAY_H
+
+#include <linux/types.h>
+#include <linux/sched.h>
+#include <linux/timer.h>
+#include <linux/wait.h>
+#include <linux/list.h>
+#include <linux/fs.h>
+#include <linux/poll.h>
+#include <linux/kref.h>
+#include <linux/pagevec.h>
+
+/*
+ * relay channel flags
+ */
+#define RCHAN_GLOBAL_BUFFER		0x00000001	/* not using per-cpu */
+
+/*
+ * For page lists
+ */
+struct relay_page {
+	struct page *page;
+	struct list_head list;
+	struct relay_page_callbacks *cb;
+	void *private_data;
+};
+
+/*
+ * Per-cpu relay channel buffer
+ */
+struct rchan_buf {
+	struct rchan *chan;		/* associated channel */
+	wait_queue_head_t read_wait;	/* reader wait queue */
+	struct timer_list timer; 	/* reader wake-up timer */
+	struct dentry *dentry;		/* channel file dentry */
+	struct kref kref;		/* channel buffer refcount */
+	struct list_head pages;		/* current set of unconsumed pages */
+	size_t nr_pages;		/* number of unconsumed pages */
+	spinlock_t lock;		/* protect pages list */
+	size_t consumed_offset;		/* bytes consumed in cur page */
+	unsigned int finalized;		/* buffer has been finalized */
+	unsigned int cpu;		/* this buf's cpu */
+} ____cacheline_aligned;
+
+/*
+ * Relay channel data structure
+ */
+struct rchan
+{
+	size_t n_pages_wakeup;		/* wake up readers after filling n */
+	struct rchan_callbacks *cb;	/* client callbacks */
+	struct kref kref;		/* channel refcount */
+	void *private_data;		/* for user-defined data */
+	struct rchan_buf *buf[NR_CPUS]; /* per-cpu channel buffers */
+	struct list_head list;		/* for channel list */
+	struct dentry *parent;		/* parent dentry passed to open */
+	char base_filename[NAME_MAX];	/* saved base filename */
+	unsigned long flags;		/* relay flags for this channel */
+};
+
+/*
+ * Relay channel client callbacks
+ */
+struct rchan_callbacks
+{
+	/*
+	 * create_buf_file - create file to represent a relay channel buffer
+	 * @filename: the name of the file to create
+	 * @parent: the parent of the file to create
+	 * @mode: the mode of the file to create
+	 * @buf: the channel buffer
+	 *
+	 * Called during relay_open(), once for each per-cpu buffer,
+	 * to allow the client to create a file to be used to
+	 * represent the corresponding channel buffer.  If the file is
+	 * created outside of relay, the parent must also exist in
+	 * that filesystem.
+	 *
+	 * The callback should return the dentry of the file created
+	 * to represent the relay buffer.
+	 *
+	 * See Documentation/filesystems/relayfs.txt for more info.
+	 */
+	struct dentry *(*create_buf_file)(const char *filename,
+					  struct dentry *parent,
+					  int mode,
+					  struct rchan_buf *buf);
+
+	/*
+	 * remove_buf_file - remove file representing a relay channel buffer
+	 * @dentry: the dentry of the file to remove
+	 *
+	 * Called during relay_close(), once for each per-cpu buffer,
+	 * to allow the client to remove a file used to represent a
+	 * channel buffer.
+	 *
+	 * The callback should return 0 if successful, negative if not.
+	 */
+	int (*remove_buf_file)(struct dentry *dentry);
+};
+
+/*
+ * Relay page callbacks
+ */
+struct relay_page_callbacks
+{
+	/*
+	 * page_released - notification that a page is ready for re-use
+	 * @page: the released page
+	 * @private_data: user-defined data associated with the page
+	 *
+	 * This callback is a notification that a given page has been
+	 * read by userspace and can be re-used.  Always called in
+	 * user context.
+	 */
+	void (*page_released) (struct page *page, void *private_data);
+
+	/*
+	 * page_released - notification that a page has been stolen
+	 * @page: the stolen page
+	 * @private_data: user-defined data associated with the page
+	 *
+	 * This callback is a notification that a given page has been
+	 * stolen by userspace.  The owner may wish to replace it;
+	 * this gives it the opportunity to do so.  Always called in
+	 * user context.
+	 */
+	void (*page_stolen) (struct page *page, void *private_data);
+};
+
+/*
+ * CONFIG_RELAY kernel API, kernel/relay.c
+ */
+
+extern struct rchan *relay_open(const char *base_filename,
+				struct dentry *parent,
+				size_t n_pages_wakeup,
+				struct rchan_callbacks *cb,
+				void *private_data,
+				unsigned long rchan_flags);
+extern void relay_add_page(struct rchan *chan,
+			   struct page *page,
+			   struct relay_page_callbacks *cb,
+			   void *private_data);
+extern void relay_add_pages(struct rchan *chan,
+			    struct pagevec *pages,
+			    struct relay_page_callbacks *cb,
+			    void *private_data);
+extern void relay_flush(struct rchan *chan);
+extern void relay_close(struct rchan *chan);
+extern void relay_reset(struct rchan *chan);
+
+/*
+ * exported relay file operations, kernel/relay.c
+ */
+extern const struct file_operations relay_file_operations;
+
+#endif /* _LINUX_RELAY_H */
+
--- /dev/null	2007-10-15 18:18:04.000000000 -0500
+++ kernel/relay.c	2008-10-14 22:05:41.000000000 -0500
@@ -0,0 +1,969 @@
+/*
+ * Public API and common code for kernel->userspace relay file support.
+ *
+ * See Documentation/filesystems/relay.txt for an overview.
+ *
+ * Copyright (C) 2002-2005 - Tom Zanussi (zanussi@xxxxxxxxxx), IBM Corp
+ * Copyright (C) 1999-2005 - Karim Yaghmour (karim@xxxxxxxxxxx)
+ * Copyright (C) 2008 - Tom Zanussi (tzanussi@xxxxxxxxx)
+ *
+ * Moved to kernel/relay.c by Paul Mundt, 2006.
+ * November 2006 - CPU hotplug support by Mathieu Desnoyers
+ * 	(mathieu.desnoyers@xxxxxxxxxx)
+ *
+ * This file is released under the GPL.
+ */
+#include <linux/errno.h>
+#include <linux/stddef.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <linux/string.h>
+#include <linux/relay.h>
+#include <linux/mm.h>
+#include <linux/cpu.h>
+#include <linux/splice.h>
+#include <linux/debugfs.h>
+
+/* list of open channels, for cpu hotplug */
+static DEFINE_MUTEX(relay_channels_mutex);
+static LIST_HEAD(relay_channels);
+
+/* forward declarations */
+static void setup_callbacks(struct rchan *chan, struct rchan_callbacks *cb);
+static struct rchan_buf *relay_open_buf(struct rchan *chan, unsigned int cpu);
+static inline void relay_wakeup_readers(struct rchan_buf *buf);
+static void relay_close_buf(struct rchan_buf *buf);
+static void relay_destroy_channel(struct kref *kref);
+static inline struct relay_page *__relay_get_rpage(struct rchan_buf *buf);
+static inline void __relay_add_page(struct rchan_buf *buf,
+				    struct relay_page *rpage);
+static inline void __relay_add_page_nolock(struct rchan_buf *buf,
+					   struct relay_page *rpage);
+static void __relay_reset(struct rchan_buf *buf, unsigned int init);
+
+/*
+ * relay kernel API
+ */
+
+/**
+ *	relay_open - create a new relay channel
+ *	@base_filename: base name of files to create, %NULL for buffering only
+ *	@parent: dentry of parent directory, %NULL for root directory or buffer
+ *	@n_pages_wakeup: wakeup readers after this many pages, 0 means never
+ *	@cb: client callback functions
+ *	@private_data: user-defined data
+ *	@flags: relay channel flags
+ *
+ *	Returns channel pointer if successful, %NULL otherwise.
+ *
+ *	Creates per-cpu channel lists (or a single list if the
+ *	RCHAN_GLOBAL_BUFFER flag is used) to receive pages from
+ *	tracers via relay_add_page()/relay_add_pages().  These lists
+ *	will be drained by userspace via read(2), splice(2), or
+ *	sendfile(2).  Pages added to relay will be either returned to
+ *	their owners after userspace has finished reading them or the
+ *	owners will be notified if they've been stolen (see
+ *	relay_add_page).
+ *
+ *	buffer files will be named base_filename0...base_filenameN-1.
+ *	File permissions will be %S_IRUSR.
+ */
+struct rchan *relay_open(const char *base_filename,
+			 struct dentry *parent,
+			 size_t n_pages_wakeup,
+			 struct rchan_callbacks *cb,
+			 void *private_data,
+			 unsigned long rchan_flags)
+{
+	unsigned int i;
+	struct rchan *chan;
+
+	chan = kzalloc(sizeof(struct rchan), GFP_KERNEL);
+	if (!chan)
+		return NULL;
+
+	chan->n_pages_wakeup = n_pages_wakeup;
+	chan->parent = parent;
+	chan->flags = rchan_flags;
+
+	chan->private_data = private_data;
+	strlcpy(chan->base_filename, base_filename, NAME_MAX);
+
+	setup_callbacks(chan, cb);
+	kref_init(&chan->kref);
+
+	mutex_lock(&relay_channels_mutex);
+	for_each_online_cpu(i) {
+		chan->buf[i] = relay_open_buf(chan, i);
+		if (!chan->buf[i])
+			goto free_bufs;
+	}
+	list_add(&chan->list, &relay_channels);
+	mutex_unlock(&relay_channels_mutex);
+
+	return chan;
+
+free_bufs:
+	for_each_online_cpu(i) {
+		if (!chan->buf[i])
+			break;
+		relay_close_buf(chan->buf[i]);
+	}
+
+	kref_put(&chan->kref, relay_destroy_channel);
+	mutex_unlock(&relay_channels_mutex);
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(relay_open);
+
+/**
+ *	relay_add_page - add a page to relay
+ *	@chan: the relay channel
+ *	@page: the page to add
+ *	@cb: relay_page callbacks associated with the page
+ *	@private_data: user data to be associated with the relay_page
+ *
+ *	Add a page to relay.  When the page has been read by
+ *	userspace, the owner will be notified.  If the page has been
+ *	copied and is available for re-use by the owner, the
+ *	relay_page_callbacks page_released() callback will be invoked.
+ *	If the page has been stolen, the owner will be notified of
+ *	this fact via the page_stolen() callback; because the
+ *	page_stolen() (and page_released()) callbacks are called from
+ *	user context, the owner can allocate a new page using
+ *	GFP_KERNEL if it wants to.
+ */
+void relay_add_page(struct rchan *chan,
+		    struct page *page,
+		    struct relay_page_callbacks *cb,
+		    void *private_data)
+{
+	struct relay_page *rpage;
+	struct rchan_buf *buf;
+
+	buf = chan->buf[get_cpu()];
+	rpage = __relay_get_rpage(buf);
+
+	if (likely(rpage)) {
+		rpage->page = page;
+		set_page_private(rpage->page, (unsigned long)buf);
+		rpage->cb = cb;
+		rpage->private_data = private_data;
+		__relay_add_page(buf, rpage);
+	}
+	put_cpu();
+}
+EXPORT_SYMBOL_GPL(relay_add_page);
+
+/**
+ *	relay_add_pages - add a set of pages to relay
+ *	@chan: the relay channel
+ *	@pages: the pages to add
+ *	@cb: relay_page callbacks associated with the pages
+ *	@private_data: user data to be associated with the relay_pages
+ *
+ *	Add a set of pages to relay.  The added pages are guaranteed
+ *	to be inserted together as a group and in the same order as in
+ *	the pagevec.  The comments for relay_add_page() apply in the
+ *	same way to relay_add_pages().
+ */
+void relay_add_pages(struct rchan *chan,
+		     struct pagevec *pages,
+		     struct relay_page_callbacks *cb,
+		     void *private_data)
+{
+	struct relay_page *rpage;
+	struct rchan_buf *buf;
+	unsigned long flags;
+	int i, nr_pages = pagevec_count(pages);
+
+	buf = chan->buf[get_cpu()];
+	spin_lock_irqsave(&buf->lock, flags);
+	for (i = 0; i < nr_pages; i--) {
+		rpage = __relay_get_rpage(buf);
+
+		if (likely(rpage)) {
+			rpage->page = pages->pages[i];
+			set_page_private(rpage->page, (unsigned long)buf);
+			rpage->cb = cb;
+			rpage->private_data = private_data;
+			__relay_add_page_nolock(buf, rpage);
+		}
+	}
+	spin_unlock_irqrestore(&buf->lock, flags);
+	put_cpu();
+
+	relay_wakeup_readers(buf);
+}
+EXPORT_SYMBOL_GPL(relay_add_pages);
+
+/**
+ *	relay_flush - flush the channel
+ *	@chan: the channel
+ *
+ *	Flushes all channel buffers, i.e. wakes up readers
+ */
+void relay_flush(struct rchan *chan)
+{
+	unsigned int i;
+	size_t prev_wakeup = chan->n_pages_wakeup;
+
+	if (!chan)
+		return;
+
+	if (prev_wakeup)
+		chan->n_pages_wakeup = 1;
+
+	if (chan->flags & RCHAN_GLOBAL_BUFFER && chan->buf[0]) {
+		chan->n_pages_wakeup = prev_wakeup;
+		return;
+	}
+
+	mutex_lock(&relay_channels_mutex);
+	for_each_possible_cpu(i)
+		if (chan->buf[i])
+			relay_wakeup_readers(chan->buf[i]);
+	mutex_unlock(&relay_channels_mutex);
+	chan->n_pages_wakeup = prev_wakeup;
+}
+EXPORT_SYMBOL_GPL(relay_flush);
+
+/**
+ *	relay_close - close the channel
+ *	@chan: the channel
+ *
+ *	Closes all channel buffers and frees the channel.
+ */
+void relay_close(struct rchan *chan)
+{
+	unsigned int i;
+
+	if (!chan)
+		return;
+
+	mutex_lock(&relay_channels_mutex);
+	if (chan->flags & RCHAN_GLOBAL_BUFFER && chan->buf[0])
+		relay_close_buf(chan->buf[0]);
+	else
+		for_each_possible_cpu(i)
+			if (chan->buf[i])
+				relay_close_buf(chan->buf[i]);
+
+	list_del(&chan->list);
+	kref_put(&chan->kref, relay_destroy_channel);
+	mutex_unlock(&relay_channels_mutex);
+}
+EXPORT_SYMBOL_GPL(relay_close);
+
+/**
+ *	relay_reset - reset the channel
+ *	@chan: the channel
+ *
+ *	This has the effect of erasing all data from all channel buffers
+ *	and restarting the channel in its initial state.
+ *
+ *	NOTE. Care should be taken that the channel isn't actually
+ *	being used by anything when this call is made.
+ */
+void relay_reset(struct rchan *chan)
+{
+	unsigned int i;
+
+	if (!chan)
+		return;
+
+	if (chan->flags & RCHAN_GLOBAL_BUFFER && chan->buf[0]) {
+		__relay_reset(chan->buf[0], 0);
+		return;
+	}
+
+	mutex_lock(&relay_channels_mutex);
+	for_each_online_cpu(i)
+		if (chan->buf[i])
+			__relay_reset(chan->buf[i], 0);
+	mutex_unlock(&relay_channels_mutex);
+}
+EXPORT_SYMBOL_GPL(relay_reset);
+
+/*
+ * end relay kernel API
+ */
+
+/**
+ *	relay_update_filesize - increase relay file i_size by length
+ *	@buf: relay channel buffer
+ *	@length: length to add
+ */
+static inline void relay_update_filesize(struct rchan_buf *buf, size_t length)
+{
+	buf->dentry->d_inode->i_size +=	length;
+}
+
+/**
+ *	__relay_get_rpage - get an empty relay page struct
+ *	@buf: the buffer struct
+ */
+static inline struct relay_page *__relay_get_rpage(struct rchan_buf *buf)
+{
+	return kmalloc(sizeof(struct relay_page), GFP_ATOMIC);
+}
+
+static inline void __relay_add_page_nolock(struct rchan_buf *buf,
+					   struct relay_page *rpage)
+{
+	list_add_tail(&rpage->list, &buf->pages);
+	buf->nr_pages++;
+	relay_update_filesize(buf, PAGE_SIZE);
+}
+
+static inline void __relay_add_page(struct rchan_buf *buf,
+				    struct relay_page *rpage)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&buf->lock, flags);
+	__relay_add_page_nolock(buf, rpage);
+	spin_unlock_irqrestore(&buf->lock, flags);
+
+	relay_wakeup_readers(buf);
+}
+
+/**
+ *	__relay_remove_page - remove a page from relay
+ *	@buf: the buffer struct
+ *	@rpage: struct relay_page
+ */
+static void __relay_remove_page(struct rchan_buf *buf,
+				struct relay_page *rpage)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&buf->lock, flags);
+	list_del(&rpage->list);
+	buf->nr_pages--;
+	spin_unlock_irqrestore(&buf->lock, flags);
+
+	kfree(rpage);
+}
+
+/**
+ *	__relay_release_page - remove page from relay and notify owner
+ *	@buf: the buffer struct
+ *	@rpage: struct relay_page
+ */
+static void __relay_release_page(struct rchan_buf *buf,
+				 struct relay_page *rpage)
+{
+	if (rpage->cb && rpage->cb->page_released)
+		rpage->cb->page_released(rpage->page, rpage->private_data);
+
+	__relay_remove_page(buf, rpage);
+}
+
+/**
+ *	relay_destroy_channel - free the channel struct
+ *	@kref: target kernel reference that contains the relay channel
+ *
+ *	Should only be called from kref_put().
+ */
+static void relay_destroy_channel(struct kref *kref)
+{
+	struct rchan *chan = container_of(kref, struct rchan, kref);
+	kfree(chan);
+}
+
+/**
+ *	relay_destroy_buf - destroy an rchan_buf struct and release pages
+ *	@buf: the buffer struct
+ */
+static void relay_destroy_buf(struct rchan_buf *buf)
+{
+	struct rchan *chan = buf->chan;
+	struct relay_page *rpage, *rpage2;
+
+	list_for_each_entry_safe(rpage, rpage2, &buf->pages, list)
+		__relay_release_page(buf, rpage);
+
+	chan->buf[buf->cpu] = NULL;
+	kfree(buf);
+	kref_put(&chan->kref, relay_destroy_channel);
+}
+
+/**
+ *	relay_remove_buf - remove a channel buffer
+ *	@kref: target kernel reference that contains the relay buffer
+ *
+ *	Removes the file from the fileystem, which also frees the
+ *	rchan_buf_struct and the channel buffer.  Should only be called from
+ *	kref_put().
+ */
+static void relay_remove_buf(struct kref *kref)
+{
+	struct rchan_buf *buf = container_of(kref, struct rchan_buf, kref);
+	buf->chan->cb->remove_buf_file(buf->dentry);
+	relay_destroy_buf(buf);
+}
+
+/**
+ *	relay_close_buf - close a channel buffer
+ *	@buf: channel buffer
+ *
+ *	Marks the buffer finalized.  The channel buffer and channel
+ *	buffer data structure are then freed automatically when the
+ *	last reference is given up.
+ */
+static void relay_close_buf(struct rchan_buf *buf)
+{
+	buf->finalized = 1;
+	del_timer_sync(&buf->timer);
+	kref_put(&buf->kref, relay_remove_buf);
+}
+
+static struct dentry *relay_create_buf_file(struct rchan *chan,
+					    struct rchan_buf *buf,
+					    unsigned int cpu)
+{
+	struct dentry *dentry;
+	char *tmpname;
+
+	tmpname = kzalloc(NAME_MAX + 1, GFP_KERNEL);
+	if (!tmpname)
+		return NULL;
+	snprintf(tmpname, NAME_MAX, "%s%d", chan->base_filename, cpu);
+
+	/* Create file in fs */
+	dentry = chan->cb->create_buf_file(tmpname, chan->parent,
+					   S_IRUSR, buf);
+
+	kfree(tmpname);
+
+	return dentry;
+}
+
+/**
+ *	relay_create_buf - allocate and initialize a channel buffer
+ *	@chan: the relay channel
+ *
+ *	Returns channel buffer if successful, %NULL otherwise.
+ */
+static struct rchan_buf *relay_create_buf(struct rchan *chan)
+{
+	struct rchan_buf *buf = kzalloc(sizeof(struct rchan_buf), GFP_KERNEL);
+	if (!buf)
+		return NULL;
+
+	spin_lock_init(&buf->lock);
+	INIT_LIST_HEAD(&buf->pages);
+	buf->chan = chan;
+	kref_get(&buf->chan->kref);
+
+	return buf;
+}
+
+/*
+ *	relay_open_buf - create a new relay channel buffer
+ *
+ *	used by relay_open() and CPU hotplug.
+ */
+static struct rchan_buf *relay_open_buf(struct rchan *chan, unsigned int cpu)
+{
+ 	struct rchan_buf *buf = NULL;
+	struct dentry *dentry;
+
+	if (chan->flags & RCHAN_GLOBAL_BUFFER)
+		return chan->buf[0];
+
+	buf = relay_create_buf(chan);
+	if (!buf)
+		return NULL;
+
+	dentry = relay_create_buf_file(chan, buf, cpu);
+	if (!dentry)
+		goto free_buf;
+	buf->dentry = dentry;
+	buf->dentry->d_inode->i_size = 0;
+
+ 	buf->cpu = cpu;
+ 	__relay_reset(buf, 1);
+
+	if (chan->flags & RCHAN_GLOBAL_BUFFER) {
+ 		chan->buf[0] = buf;
+ 		buf->cpu = 0;
+  	}
+
+	return buf;
+
+free_buf:
+ 	relay_destroy_buf(buf);
+	return NULL;
+}
+
+/**
+ *	relay_wakeup_readers - wake up readers if applicable
+ *	@buf: relay channel buffer
+ *
+ *	Will wake up readers after each buf->n_pages_wakeup pages have
+ *	been produced.  To do no waking up, simply pass 0 into relay
+ *	open for this value.
+ */
+static inline void relay_wakeup_readers(struct rchan_buf *buf)
+{
+	size_t wakeup = buf->chan->n_pages_wakeup;
+
+	if (wakeup && (buf->nr_pages % wakeup == 0) &&
+	    (waitqueue_active(&buf->read_wait)))
+		/*
+		 * Calling wake_up_interruptible() from here
+		 * will deadlock if we happen to be logging
+		 * from the scheduler (trying to re-grab
+		 * rq->lock), so defer it.
+		 */
+		__mod_timer(&buf->timer, jiffies + 1);
+}
+
+/**
+ *	wakeup_readers - wake up readers waiting on a channel
+ *	@data: contains the channel buffer
+ *
+ *	This is the timer function used to defer reader waking.
+ */
+static void wakeup_readers(unsigned long data)
+{
+	struct rchan_buf *buf = (struct rchan_buf *)data;
+	wake_up_interruptible(&buf->read_wait);
+}
+
+/**
+ *	__relay_reset - reset a channel buffer
+ *	@buf: the channel buffer
+ *	@init: 1 if this is a first-time initialization
+ *
+ *	See relay_reset() for description of effect.
+ */
+static void __relay_reset(struct rchan_buf *buf, unsigned int init)
+{
+	struct relay_page *rpage, *rpage2;
+
+	if (init) {
+		init_waitqueue_head(&buf->read_wait);
+		kref_init(&buf->kref);
+		setup_timer(&buf->timer, wakeup_readers, (unsigned long)buf);
+	} else
+		del_timer_sync(&buf->timer);
+
+	list_for_each_entry_safe(rpage, rpage2, &buf->pages, list)
+		__relay_release_page(buf, rpage);
+
+	buf->consumed_offset = 0;
+	buf->finalized = 0;
+}
+
+/*
+ * create_buf_file_create() default callback.  Creates debugfs file.
+ */
+static struct dentry *create_buf_file_default_callback(const char *filename,
+						       struct dentry *parent,
+						       int mode,
+						       struct rchan_buf *buf)
+{
+	return debugfs_create_file(filename, mode, parent, buf,
+				   &relay_file_operations);
+}
+
+/*
+ * remove_buf_file() default callback.  Removes debugfs file.
+ */
+static int remove_buf_file_default_callback(struct dentry *dentry)
+{
+	debugfs_remove(dentry);
+	return 0;
+}
+
+/* relay channel default callbacks */
+static struct rchan_callbacks default_channel_callbacks = {
+	.create_buf_file = create_buf_file_default_callback,
+	.remove_buf_file = remove_buf_file_default_callback,
+};
+
+static void setup_callbacks(struct rchan *chan, struct rchan_callbacks *cb)
+{
+	if (!cb) {
+		chan->cb = &default_channel_callbacks;
+		return;
+	}
+
+	if (!cb->create_buf_file)
+		cb->create_buf_file = create_buf_file_default_callback;
+	if (!cb->remove_buf_file)
+		cb->remove_buf_file = remove_buf_file_default_callback;
+	chan->cb = cb;
+}
+
+/*
+ * relay userspace implementations
+ */
+
+/**
+ *	relay_file_open - open file op for relay files
+ *	@inode: the inode
+ *	@filp: the file
+ *
+ *	Increments the channel buffer refcount.
+ */
+static int relay_file_open(struct inode *inode, struct file *filp)
+{
+	struct rchan_buf *buf = inode->i_private;
+	kref_get(&buf->kref);
+	filp->private_data = buf;
+
+	return nonseekable_open(inode, filp);
+}
+
+/**
+ *	relay_file_poll - poll file op for relay files
+ *	@filp: the file
+ *	@wait: poll table
+ *
+ *	Poll implemention.
+ */
+static unsigned int relay_file_poll(struct file *filp, poll_table *wait)
+{
+	unsigned int mask = 0;
+	struct rchan_buf *buf = filp->private_data;
+
+	if (buf->finalized)
+		return POLLERR;
+
+	if (filp->f_mode & FMODE_READ) {
+		poll_wait(filp, &buf->read_wait, wait);
+		if (buf->nr_pages)
+			mask |= POLLIN | POLLRDNORM;
+	}
+
+	return mask;
+}
+
+/**
+ *	relay_file_release - release file op for relay files
+ *	@inode: the inode
+ *	@filp: the file
+ *
+ *	Decrements the channel refcount, as the filesystem is
+ *	no longer using it.
+ */
+static int relay_file_release(struct inode *inode, struct file *filp)
+{
+	struct rchan_buf *buf = filp->private_data;
+	kref_put(&buf->kref, relay_remove_buf);
+
+	return 0;
+}
+
+/**
+ *	relay_file_read_page_avail - return bytes available in next page
+ *	@buf: relay channel buffer
+ */
+static size_t relay_file_read_page_avail(struct rchan_buf *buf)
+{
+	size_t avail = 0;
+
+	if (!list_empty(&buf->pages))
+		avail = PAGE_SIZE - buf->consumed_offset;
+
+	return avail;
+}
+
+/*
+ *	relay_consume - update the consumed count for the buffer
+ */
+static void relay_consume(struct rchan_buf *buf, int bytes_consumed)
+{
+	buf->consumed_offset += bytes_consumed;
+
+	if (buf->consumed_offset == PAGE_SIZE) {
+		struct relay_page *rpage;
+		rpage = list_first_entry(&buf->pages, struct relay_page, list);
+		__relay_release_page(buf, rpage);
+
+		buf->consumed_offset = 0;
+	}
+}
+
+/*
+ *	page_read_actor - read up to one page's worth of data
+ */
+static int page_read_actor(struct rchan_buf *buf,
+			   size_t avail,
+			   read_descriptor_t *desc,
+			   read_actor_t actor)
+{
+	void *from;
+	int ret = 0;
+	struct relay_page *rpage;
+
+	rpage = list_first_entry(&buf->pages, struct relay_page, list);
+
+	from = page_address(rpage->page);
+	from += PAGE_SIZE - avail;
+	ret = avail;
+	if (copy_to_user(desc->arg.buf, from, avail)) {
+		desc->error = -EFAULT;
+		ret = 0;
+	}
+	desc->arg.data += ret;
+	desc->written += ret;
+	desc->count -= ret;
+
+	return ret;
+}
+
+typedef int (*page_actor_t) (struct rchan_buf *buf,
+			     size_t avail,
+			     read_descriptor_t *desc,
+			     read_actor_t actor);
+
+/*
+ *	relay_file_read_pages - read count bytes, bridging page boundaries
+ */
+static ssize_t relay_file_read_pages(struct file *filp, loff_t *ppos,
+				     page_actor_t page_actor,
+				     read_actor_t actor,
+				     read_descriptor_t *desc)
+{
+	struct rchan_buf *buf = filp->private_data;
+	size_t avail;
+	int ret;
+
+	if (!desc->count)
+		return 0;
+
+	mutex_lock(&filp->f_path.dentry->d_inode->i_mutex);
+	do {
+		avail = relay_file_read_page_avail(buf);
+		if (!avail)
+			break;
+		avail = min(desc->count, avail);
+		ret = page_actor(buf, avail, desc, actor);
+		if (desc->error < 0)
+			break;
+		if (ret) {
+			relay_consume(buf, ret);
+			*ppos += ret;
+		}
+	} while (desc->count && ret);
+	mutex_unlock(&filp->f_path.dentry->d_inode->i_mutex);
+
+	return desc->written;
+}
+
+static ssize_t relay_file_read(struct file *filp,
+			       char __user *buffer,
+			       size_t count,
+			       loff_t *ppos)
+{
+	read_descriptor_t desc;
+	desc.written = 0;
+	desc.count = count;
+	desc.arg.buf = buffer;
+	desc.error = 0;
+	return relay_file_read_pages(filp, ppos, page_read_actor,
+				     NULL, &desc);
+}
+
+static void relay_pipe_buf_release(struct pipe_inode_info *pipe,
+				   struct pipe_buffer *pipe_buf)
+{
+	struct rchan_buf *buf;
+
+	buf = (struct rchan_buf *)page_private(pipe_buf->page);
+	relay_consume(buf, pipe_buf->private);
+}
+
+static int relay_pipe_buf_steal(struct pipe_inode_info *pipe,
+				struct pipe_buffer *pipe_buf)
+{
+	int ret;
+	struct rchan_buf *buf;
+
+	buf = (struct rchan_buf *)page_private(pipe_buf->page);
+	ret = generic_pipe_buf_steal(pipe, pipe_buf);
+	if (!ret) {
+		struct relay_page *rpage;
+		rpage = list_first_entry(&buf->pages, struct relay_page, list);
+		__relay_remove_page(buf, rpage);
+		if (rpage->cb && rpage->cb->page_stolen)
+			rpage->cb->page_stolen(pipe_buf->page,
+					       rpage->private_data);
+	}
+
+	return ret;
+}
+
+static struct pipe_buf_operations relay_pipe_buf_ops = {
+	.can_merge = 0,
+	.map = generic_pipe_buf_map,
+	.unmap = generic_pipe_buf_unmap,
+	.confirm = generic_pipe_buf_confirm,
+	.release = relay_pipe_buf_release,
+	.steal = relay_pipe_buf_steal,
+	.get = generic_pipe_buf_get,
+};
+
+static void relay_page_release(struct splice_pipe_desc *spd, unsigned int i)
+{
+}
+
+/*
+ *	page_splice_actor - splice available data
+ */
+static int page_splice_actor(struct file *in,
+			     struct pipe_inode_info *pipe,
+			     size_t len,
+			     unsigned int flags)
+{
+	unsigned int poff, total_len, nr_pages, ret;
+	struct rchan_buf *buf = in->private_data;
+	struct relay_page *rpage;
+	struct page *pages[PIPE_BUFFERS];
+	struct partial_page partial[PIPE_BUFFERS];
+	struct splice_pipe_desc spd = {
+		.pages = pages,
+		.nr_pages = 0,
+		.partial = partial,
+		.flags = flags,
+		.ops = &relay_pipe_buf_ops,
+		.spd_release = relay_page_release,
+	};
+
+	if (list_empty(&buf->pages))
+		return 0;
+
+	poff = buf->consumed_offset;
+	nr_pages = min_t(unsigned int, buf->nr_pages, PIPE_BUFFERS);
+	total_len = 0;
+
+	list_for_each_entry(rpage, &buf->pages, list) {
+		unsigned int this_len;
+
+		if (spd.nr_pages >= nr_pages)
+			break;
+
+		if (!len)
+			break;
+
+		this_len = min_t(unsigned long, len, PAGE_SIZE - poff);
+
+		spd.pages[spd.nr_pages] = rpage->page;
+		spd.partial[spd.nr_pages].offset = poff;
+		spd.partial[spd.nr_pages].len = this_len;
+		spd.partial[spd.nr_pages].private = this_len;
+
+		len -= this_len;
+		total_len += this_len;
+		poff = 0;
+		spd.nr_pages++;
+	}
+
+	ret = splice_to_pipe(pipe, &spd);
+
+        return ret;
+}
+
+static ssize_t relay_file_splice_read(struct file *in,
+				      loff_t *ppos,
+				      struct pipe_inode_info *pipe,
+				      size_t len,
+				      unsigned int flags)
+{
+	ssize_t spliced;
+	int ret;
+
+	ret = 0;
+	spliced = 0;
+
+	while (len && !spliced) {
+		ret = page_splice_actor(in, pipe, len, flags);
+		if (ret < 0)
+			break;
+		else if (!ret) {
+			if (spliced)
+				break;
+			if (flags & SPLICE_F_NONBLOCK) {
+				ret = -EAGAIN;
+				break;
+			}
+		}
+
+		*ppos += ret;
+		if (ret > len)
+			len = 0;
+		else
+			len -= ret;
+		spliced += ret;
+	}
+
+	if (spliced)
+		return spliced;
+
+	return ret;
+}
+
+const struct file_operations relay_file_operations = {
+	.open		= relay_file_open,
+	.poll		= relay_file_poll,
+	.read		= relay_file_read,
+	.llseek		= no_llseek,
+	.release	= relay_file_release,
+	.splice_read	= relay_file_splice_read,
+};
+EXPORT_SYMBOL_GPL(relay_file_operations);
+
+/**
+ * 	relay_hotcpu_callback - CPU hotplug callback
+ * 	@nb: notifier block
+ * 	@action: hotplug action to take
+ * 	@hcpu: CPU number
+ *
+ * 	Returns the success/failure of the operation. (%NOTIFY_OK, %NOTIFY_BAD)
+ */
+static int __cpuinit relay_hotcpu_callback(struct notifier_block *nb,
+				unsigned long action,
+				void *hcpu)
+{
+	unsigned int hotcpu = (unsigned long)hcpu;
+	struct rchan *chan;
+
+	switch (action) {
+	case CPU_UP_PREPARE:
+	case CPU_UP_PREPARE_FROZEN:
+		mutex_lock(&relay_channels_mutex);
+		list_for_each_entry(chan, &relay_channels, list) {
+			if (chan->buf[hotcpu])
+				continue;
+			chan->buf[hotcpu] = relay_open_buf(chan, hotcpu);
+			if (!chan->buf[hotcpu]) {
+				printk(KERN_ERR
+					"relay_hotcpu_callback: cpu %d buffer "
+					"creation failed\n", hotcpu);
+				mutex_unlock(&relay_channels_mutex);
+				return NOTIFY_BAD;
+			}
+		}
+		mutex_unlock(&relay_channels_mutex);
+		break;
+	case CPU_DEAD:
+	case CPU_DEAD_FROZEN:
+		/* No need to flush the cpu : will be flushed upon
+		 * final relay_flush() call. */
+		break;
+	}
+	return NOTIFY_OK;
+}
+
+static __init int relay_init(void)
+{
+	hotcpu_notifier(relay_hotcpu_callback, 0);
+	return 0;
+}
+
+early_initcall(relay_init);
--- /dev/null	2007-10-15 18:18:04.000000000 -0500
+++ include/linux/relay_pagewriter.h	2008-10-14 22:05:42.000000000 -0500
@@ -0,0 +1,295 @@
+/*
+ * linux/include/linux/relay_pagewriter.h
+ *
+ * Copyright (C) 2002, 2003 - Tom Zanussi (zanussi@xxxxxxxxxx), IBM Corp
+ * Copyright (C) 1999, 2000, 2001, 2002 - Karim Yaghmour (karim@xxxxxxxxxxx)
+ * Copyright (C) 2008 - Tom Zanussi (tzanussi@xxxxxxxxx)
+ *
+ * CONFIG_RELAY definitions and declarations
+ */
+
+#ifndef _LINUX_RELAY_PAGEWRITER_H
+#define _LINUX_RELAY_PAGEWRITER_H
+
+#include <linux/types.h>
+#include <linux/sched.h>
+#include <linux/timer.h>
+#include <linux/wait.h>
+#include <linux/list.h>
+#include <linux/fs.h>
+#include <linux/poll.h>
+#include <linux/kref.h>
+#include <linux/relay.h>
+
+/*
+ * pagewriter flags
+ */
+#define PAGEWRITER_PAD_WRITES		0x00010000	/* don't cross pages */
+#define PAGEWRITER_FLIGHT_MODE		0x00020000	/* n_pages page ring */
+#define PAGEWRITER_LATE_SETUP		0x00040000	/* delay chan create */
+
+/*
+ * Per-cpu pagewriter buffer
+ */
+struct pagewriter_buf {
+	struct relay_page *page;	/* current write page */
+	void *data;			/* address of current page */
+	size_t offset;			/* current offset into page */
+	struct pagewriter *pagewriter;	/* associated pagewriter */
+	struct kref kref;		/* channel buffer refcount */
+	struct list_head pool;		/* current set of unused pages */
+	struct list_head empty_rpage_structs;	/* cached rpage structs */
+	size_t n_pages_flight;		/* number full flight pages written */
+	unsigned int cpu;		/* this buf's cpu */
+} ____cacheline_aligned;
+
+/*
+ * Pagewriter data structure
+ */
+struct pagewriter {
+	struct rchan *rchan;		/* associated relay channel */
+	struct pagewriter_callbacks *cb;	/* client callbacks */
+	size_t n_pages;			/* number of pages per buffer */
+	size_t n_pages_wakeup;		/* save for LATE */
+	struct kref kref;		/* channel refcount */
+	void *private_data;		/* for user-defined data */
+	size_t last_toobig;		/* tried to log event > page size */
+	struct pagewriter_buf *buf[NR_CPUS]; /* per-cpu channel buffers */
+	struct list_head list;		/* for channel list */
+	atomic_t dropped;		/* dropped events due to buffer-full */
+	char base_filename[NAME_MAX];	/* saved base filename, for LATE */
+	unsigned long flags;		/* pagewriter flags for this channel */
+};
+
+extern size_t pagewriter_switch_page_default_callback(struct pagewriter_buf *b,
+						      size_t length,
+						      void **reserved);
+
+/**
+ *	pagewriter_event_toobig - is event too big to fit in a page?
+ *	@buf: pagewriter channel buffer
+ *	@length: length of event
+ *
+ *	Returns 1 if too big, 0 otherwise.
+ *
+ *	switch_page() helper function.
+ */
+static inline int pagewriter_event_toobig(struct pagewriter_buf *buf,
+					  size_t length)
+{
+	return length > PAGE_SIZE;
+}
+
+/*
+ * Pagewriter client callbacks
+ */
+struct pagewriter_callbacks {
+	/*
+	 * new_page - called on switch to a new page
+	 * @buf: the channel buffer containing the new page
+	 * @page_data: the start of the new page
+	 *
+	 * This is simply a notification that a new page has been
+	 * switched to.  The default version does nothing.  Clients
+	 * can use the channel private_data to track previous pages,
+	 * determine whether this is the first page, etc.
+	 *
+	 * NOTE: the client can reserve bytes at the beginning of the new
+	 *       page by calling page_start_reserve() in this callback.
+	 */
+	void (*new_page) (struct pagewriter_buf *buf,
+			  void *page_data);
+
+	/*
+	 * switch_page - page switch callback
+	 * @buf: the channel buffer
+	 * @length: size of current event
+	 * @reserved: a pointer to the space reserved
+	 *
+	 * This callback can be used to replace the complete write
+	 * path.  Normally clients wouldn't override this and would
+	 * use the default version instead.
+	 *
+	 * Returns either the length passed in or 0 if full.
+	 *
+	 * Performs page-switch tasks.
+	 */
+	size_t (*switch_page)(struct pagewriter_buf *buf,
+			      size_t length,
+			      void **reserved);
+
+	/*
+	 * write_padding - callback for writing padding events
+	 * @buf: the channel buffer
+	 * @length: the length of the padding
+	 * @reserved: a pointer to the start of padding
+	 *
+	 * This callback can be used to write a padding event when
+	 * pagewriter_reserve can't write a complete event.  The
+	 * length of the padding is guaranteed to be at least as large
+	 * as the end_reserve size passed into pagewriter_reserve().
+	 */
+	void (*write_padding)(struct pagewriter_buf *buf,
+			      size_t length,
+			      void *reserved);
+};
+
+/**
+ *	pagewriter_write - write data into the channel, without padding
+ *	@pagewriter: pagewriter
+ *	@data: data to be written
+ *	@length: number of bytes to write
+ *
+ *	Writes data into the current cpu's channel buffer, crossing
+ *	page boundaries.
+ *
+ *	Protects the buffer by disabling interrupts.  Use this if you
+ *	might be logging from interrupt context.  Try
+ *	__pagewriter_write() if you know you won't be logging from
+ *	interrupt context.
+ */
+static inline void pagewriter_write(struct pagewriter *pagewriter,
+				    const void *data,
+				    size_t length)
+{
+	size_t remainder = length;
+	struct pagewriter_buf *buf;
+	unsigned long flags;
+	void *reserved, *reserved2;
+
+	local_irq_save(flags);
+	buf = pagewriter->buf[smp_processor_id()];
+	reserved = buf->data + buf->offset;
+	if (unlikely(buf->offset + length > PAGE_SIZE)) {
+		remainder = pagewriter->cb->switch_page(buf, length,
+							&reserved2);
+		if (unlikely(!reserved2)) {
+			local_irq_restore(flags);
+			return;
+		}
+		length -= remainder;
+		memcpy(reserved2, data + length, remainder);
+	}
+	memcpy(reserved, data, length);
+	buf->offset += remainder;
+	local_irq_restore(flags);
+}
+
+/**
+ *	__pagewriter_write - write data into the channel, without padding
+ *	@pagewriter: pagewriter
+ *	@data: data to be written
+ *	@length: number of bytes to write
+ *
+ *	Writes data into the current cpu's channel buffer, crossing
+ *	page boundaries.
+ *
+ *	Protects the buffer by disabling preemption.  Use
+ *	pagewriter_write() if you might be logging from interrupt
+ *	context.
+ */
+static inline void __pagewriter_write(struct pagewriter *pagewriter,
+				      const void *data,
+				      size_t length)
+{
+	size_t remainder = length;
+	struct pagewriter_buf *buf;
+	void *reserved, *reserved2;
+
+	buf = pagewriter->buf[get_cpu()];
+	reserved = buf->data + buf->offset;
+	if (unlikely(buf->offset + length > PAGE_SIZE)) {
+		remainder = pagewriter->cb->switch_page(buf, length,
+							&reserved2);
+		if (unlikely(!reserved2))
+			return;
+		length -= remainder;
+		memcpy(reserved2, data + length, remainder);
+	}
+	memcpy(reserved, data, length);
+	buf->offset += remainder;
+	put_cpu();
+}
+
+/**
+ *	pagewriter_reserve - reserve slot in channel buffer
+ *	@pagewriter: pagewriter
+ *	@length: number of bytes to reserve
+ *	@end_reserve: reserve at least this much for a padding event, if needed
+ *
+ *	Returns pointer to reserved slot, NULL if full.
+ *
+ *	Reserves a slot in the current cpu's channel buffer.
+ *	Does not protect the buffer at all - caller must provide
+ *	appropriate synchronization.
+ *
+ *	If the event won't fit, at least end_reserve bytes are
+ *	reserved for a padding event, and the write_padding() callback
+ *	function is called to allow the client to write the padding
+ *	event before switching to the next page.  The write_padding()
+ *	callback is passed a pointer to the start of the padding along
+ *	with its length.
+ */
+
+static inline void *pagewriter_reserve(struct pagewriter *pagewriter,
+				       size_t length,
+				       size_t end_reserve)
+{
+	struct pagewriter_buf *buf;
+	unsigned long flags;
+	void *reserved;
+
+	local_irq_save(flags);
+	buf = pagewriter->buf[smp_processor_id()];
+	reserved = buf->data + buf->offset;
+	if (unlikely(buf->offset + length + end_reserve > PAGE_SIZE)) {
+		void *prev_reserved = reserved;
+		size_t new_offset = buf->offset + length;
+		if (likely(new_offset != PAGE_SIZE)) {
+			size_t padding = PAGE_SIZE - buf->offset;
+			pagewriter->cb->write_padding(buf, padding, reserved);
+		}
+		pagewriter->cb->switch_page(buf, length, &reserved);
+		if (unlikely(!reserved))
+			length = 0;
+		if (unlikely(new_offset == PAGE_SIZE)) {
+			reserved = prev_reserved;
+			length = 0;
+		}
+	}
+	buf->offset += length;
+	local_irq_restore(flags);
+
+	return reserved;
+}
+
+/**
+ *	page_start_reserve - reserve bytes at the start of a page
+ *	@buf: pagewriter channel buffer
+ *	@length: number of bytes to reserve
+ *
+ *	Helper function used to reserve bytes at the beginning of
+ *	a page in the new_page() callback.
+ */
+static inline void page_start_reserve(struct pagewriter_buf *buf,
+				      size_t length)
+{
+	BUG_ON(length >= PAGE_SIZE - 1);
+	buf->offset = length;
+}
+
+extern struct pagewriter *pagewriter_open(const char *base_filename,
+					  struct dentry *parent,
+					  size_t n_pages,
+					  size_t n_pages_wakeup,
+					  struct pagewriter_callbacks *cb,
+					  void *private_data,
+					  unsigned long rchan_flags);
+extern void pagewriter_flush(struct pagewriter *pagewriter);
+extern void pagewriter_close(struct pagewriter *pagewriter);
+extern void pagewriter_reset(struct pagewriter *pagewriter);
+extern void pagewriter_save_flight_data(struct pagewriter *pagewriter);
+extern int pagewriter_late_setup(struct pagewriter *pagewriter,
+				 struct dentry *parent);
+
+#endif /* _LINUX_RELAY_PAGEWRITER_H */
--- /dev/null	2007-10-15 18:18:04.000000000 -0500
+++ kernel/relay_pagewriter.c	2008-10-14 22:05:42.000000000 -0500
@@ -0,0 +1,742 @@
+/*
+ * Provides per-cpu page writers and page pool management for current
+ * users of the relay interface.  Basically this provides functions to
+ * write into pages, feed them into a relay object for consumption by
+ * usespace, and reclaim them after they've been read.
+ *
+ * See Documentation/filesystems/relay.txt for an overview.
+ *
+ * Copyright (C) 2002-2005 - Tom Zanussi (zanussi@xxxxxxxxxx), IBM Corp
+ * Copyright (C) 1999-2005 - Karim Yaghmour (karim@xxxxxxxxxxx)
+ * Copyright (C) 2008 - Tom Zanussi (tzanussi@xxxxxxxxx)
+ *
+ * Moved to kernel/relay.c by Paul Mundt, 2006.
+ * November 2006 - CPU hotplug support by Mathieu Desnoyers
+ * 	(mathieu.desnoyers@xxxxxxxxxx)
+ *
+ * This file is released under the GPL.
+ */
+#include <linux/errno.h>
+#include <linux/stddef.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <linux/string.h>
+#include <linux/relay.h>
+#include <linux/vmalloc.h>
+#include <linux/mm.h>
+#include <linux/cpu.h>
+#include <linux/splice.h>
+#include <linux/relay_pagewriter.h>
+#include <linux/debugfs.h>
+
+/* list of open pagewriters, for cpu hotplug */
+static DEFINE_MUTEX(pagewriters_mutex);
+static LIST_HEAD(pagewriters);
+
+/* forward declarations */
+static void setup_callbacks(struct pagewriter *pagewriter,
+			    struct pagewriter_callbacks *cb);
+static void pagewriter_close_buf(struct pagewriter_buf *buf);
+static struct pagewriter_buf *pagewriter_open_buf(struct pagewriter *pw,
+						  unsigned int cpu);
+static void pagewriter_destroy(struct kref *kref);
+static void __pagewriter_reset(struct pagewriter_buf *buf, unsigned int init);
+static void pagewriter_save_flight_buf(struct pagewriter_buf *buf);
+
+/*
+ * pagewriter kernel API
+ */
+
+/**
+ *	pagewriter_open - create a new pagewriter
+ *	@base_filename: base name of files to create, %NULL for buffering only
+ *	@parent: dentry of parent directory, %NULL for root directory or buffer
+ *	@n_pages: number of pages to use for each buffer
+ *	@n_pages_wakeup: wakeup readers after this many pages, 0 means never
+ *	@cb: client callback functions
+ *	@private_data: user-defined data
+ *	@flags: channel flags, top half for pagewriter, bottom half for relay
+ *
+ *	Returns pagewriter pointer if successful, %NULL otherwise.
+ *
+ *	Creates a pagewriter page pool for each cpu using the sizes and
+ *	attributes specified.
+ */
+struct pagewriter *pagewriter_open(const char *base_filename,
+				   struct dentry *parent,
+				   size_t n_pages,
+				   size_t n_pages_wakeup,
+				   struct pagewriter_callbacks *cb,
+				   void *private_data,
+				   unsigned long flags)
+{
+	unsigned int i;
+	struct pagewriter *pagewriter;
+
+	if (!n_pages)
+		return NULL;
+
+	pagewriter = kzalloc(sizeof(struct pagewriter), GFP_KERNEL);
+	if (!pagewriter)
+		return NULL;
+
+	if (flags & PAGEWRITER_LATE_SETUP) {
+		strlcpy(pagewriter->base_filename, base_filename, NAME_MAX);
+		pagewriter->n_pages_wakeup = n_pages_wakeup;
+	} else {
+		pagewriter->rchan = relay_open(base_filename, parent,
+					       n_pages_wakeup, NULL,
+					       private_data, flags);
+		if (!pagewriter->rchan) {
+			kfree(pagewriter);
+			return NULL;
+		}
+	}
+
+	pagewriter->flags = flags;
+	pagewriter->n_pages = n_pages;
+	atomic_set(&pagewriter->dropped, 0);
+
+	pagewriter->private_data = private_data;
+	setup_callbacks(pagewriter, cb);
+	kref_init(&pagewriter->kref);
+
+	mutex_lock(&pagewriters_mutex);
+	for_each_online_cpu(i) {
+		pagewriter->buf[i] = pagewriter_open_buf(pagewriter, i);
+		if (!pagewriter->buf[i])
+			goto free_bufs;
+	}
+	list_add(&pagewriter->list, &pagewriters);
+	mutex_unlock(&pagewriters_mutex);
+
+	return pagewriter;
+
+free_bufs:
+	for_each_online_cpu(i) {
+		if (!pagewriter->buf[i])
+			break;
+		pagewriter_close_buf(pagewriter->buf[i]);
+	}
+
+	relay_close(pagewriter->rchan);
+	kref_put(&pagewriter->kref, pagewriter_destroy);
+	kfree(pagewriter);
+	mutex_unlock(&pagewriters_mutex);
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(pagewriter_open);
+
+/**
+ *	pagewriter_flush - close the pagewriter
+ *	@pagewriter: the pagewriter
+ *
+ *	Flushes all channel buffers, i.e. forces page switch.
+ */
+void pagewriter_flush(struct pagewriter *pagewriter)
+{
+	unsigned int i;
+
+	if (!pagewriter)
+		return;
+
+	mutex_lock(&pagewriters_mutex);
+	for_each_possible_cpu(i) {
+		struct pagewriter_buf *buf = pagewriter->buf[i];
+		if (buf) {
+			if (buf->pagewriter->flags & PAGEWRITER_PAD_WRITES) {
+				void *padding = buf->data + buf->offset;
+				size_t padlen = PAGE_SIZE - buf->offset;
+				pagewriter->cb->write_padding(buf, padlen,
+							      padding);
+			}
+			pagewriter->cb->switch_page(buf, 0, NULL);
+		}
+	}
+	relay_flush(pagewriter->rchan);
+	mutex_unlock(&pagewriters_mutex);
+}
+EXPORT_SYMBOL_GPL(pagewriter_flush);
+
+/**
+ *	pagewriter_close - close the pagewriter
+ *	@pagewriter: the pagewriter
+ *
+ *	Closes all buffers and frees their page pools, and also frees
+ *	the pagewriter.
+ */
+void pagewriter_close(struct pagewriter *pagewriter)
+{
+	unsigned int i;
+
+	if (!pagewriter)
+		return;
+
+	mutex_lock(&pagewriters_mutex);
+	for_each_possible_cpu(i)
+		if (pagewriter->buf[i])
+			pagewriter_close_buf(pagewriter->buf[i]);
+
+	relay_close(pagewriter->rchan);
+	if (pagewriter->last_toobig)
+		printk(KERN_WARNING "pagewriter: one or more items not logged "
+		       "[item size (%Zd) > PAGE_SIZE (%lu)]\n",
+		       pagewriter->last_toobig, PAGE_SIZE);
+
+	list_del(&pagewriter->list);
+	kref_put(&pagewriter->kref, pagewriter_destroy);
+	mutex_unlock(&pagewriters_mutex);
+}
+EXPORT_SYMBOL_GPL(pagewriter_close);
+
+/**
+ *	pagewriter_reset - reset the pagewriter
+ *	@pagewriter: the pagewriter
+ *
+ *	This has the effect of erasing all data from the current page
+ *	and restarting the pagewriter in its initial state.
+ *
+ *	NOTE. Care should be taken that the pagewriter isn't actually
+ *	being used by anything when this call is made.
+ */
+void pagewriter_reset(struct pagewriter *pagewriter)
+{
+	unsigned int i;
+
+	if (!pagewriter)
+		return;
+
+	mutex_lock(&pagewriters_mutex);
+	for_each_online_cpu(i)
+		if (pagewriter->buf[i])
+			__pagewriter_reset(pagewriter->buf[i], 0);
+	mutex_unlock(&pagewriters_mutex);
+}
+EXPORT_SYMBOL_GPL(pagewriter_reset);
+
+/**
+ *	pagewriter_save_flight_data - log all pages dirtied in flight mode
+ *	@pagewriter: pagewriter
+ *
+ *	In flight mode (PAGEWRITER_FLIGHT_MODE), the pages written to
+ *	via the pagewriter_write/reserve functions are simply cycled
+ *	around the per-cpu page pools, and not sent to relay.  This
+ *	function provides a way, at the user's request, to simply
+ *	sends all the dirty pages in the page pools to relay and
+ *	therefore onto their final destination e.g. disk or network.
+ *
+ *	The pagewriter and associated buffers will be in the same
+ *	state as if hey were reset after this call.
+ */
+void pagewriter_save_flight_data(struct pagewriter *pagewriter)
+{
+	unsigned int i;
+
+	if (!pagewriter)
+		return;
+
+	mutex_lock(&pagewriters_mutex);
+	for_each_possible_cpu(i)
+		if (pagewriter->buf[i])
+			pagewriter_save_flight_buf(pagewriter->buf[i]);
+	relay_flush(pagewriter->rchan);
+	mutex_unlock(&pagewriters_mutex);
+}
+EXPORT_SYMBOL_GPL(pagewriter_save_flight_data);
+
+/**
+ *	pagewriter_late_setup - create relay channel and log early pages
+ *	@pagewriter: pagewriter
+ *	@parent: dentry of parent directory, %NULL for root directory
+ *
+ *	If the pagewriter was initially created in early mode
+ *	(PAGEWRITER_LATE_SETUP), this creates the relay channel and
+ *	sends all the early pages in the page pools to relay and
+ *	therefore onto their final destination e.g. disk or network.
+ *
+ *	Returns 0 if successful, non-zero otherwise.
+ *
+ *	Use to setup files for a previously buffer-only channel.
+ *	Useful to do early tracing in kernel, before VFS is up, for example.
+ */
+int pagewriter_late_setup(struct pagewriter *pagewriter,
+			  struct dentry *parent)
+{
+	if (!pagewriter)
+		return -EINVAL;
+
+	pagewriter->rchan = relay_open(pagewriter->base_filename,
+				       parent,
+				       pagewriter->n_pages_wakeup,
+				       NULL,
+				       pagewriter->private_data,
+				       pagewriter->flags);
+	if (!pagewriter->rchan)
+		return -ENOMEM;
+
+	pagewriter->flags &= ~PAGEWRITER_LATE_SETUP;
+	pagewriter_save_flight_data(pagewriter);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(pagewriter_late_setup);
+
+/*
+ * end relay kernel API
+ */
+
+/**
+ *	pagewriter_get_free_page - get a free relay_page from the pool
+ *	@buf: the buffer struct
+ *
+ *	Returns relay page if successful, NULL if not.
+ */
+static struct relay_page *pagewriter_get_free_page(struct pagewriter_buf *buf)
+{
+	struct relay_page *rpage = NULL;
+
+	if (!list_empty(&buf->pool)) {
+		rpage = list_first_entry(&buf->pool, struct relay_page, list);
+		list_del(&rpage->list);
+	}
+
+	return rpage;
+}
+
+/**
+ *	pagewriter_add_free_page - add/return a free relay_page to the pool
+ *	@buf: the buffer struct
+ *	@rpage: relay_page to add
+ *
+ *	Returns relay page if successful, NULL if not.
+ */
+static void pagewriter_add_free_page(struct pagewriter_buf *buf,
+				     struct relay_page *rpage)
+{
+	list_add_tail(&rpage->list, &buf->pool);
+}
+
+/**
+ *	get_empty_rpage_struct - get an empty rpage_struct to hold a page
+ *	@buf: the buffer struct
+ *
+ *	Returns an rpage_struct if successful, NULL if not.
+ */
+static struct relay_page *get_empty_rpage_struct(struct pagewriter_buf *buf)
+{
+	struct relay_page *rpage = NULL;
+
+	if (!list_empty(&buf->empty_rpage_structs)) {
+		rpage = list_first_entry(&buf->empty_rpage_structs,
+					 struct relay_page, list);
+		list_del(&rpage->list);
+	}
+
+	return rpage;
+}
+
+/**
+ *	add_empty_rpage_struct - add/return a free rpage_struct to the pool
+ *	@buf: the buffer struct
+ *	@rpage: struct relay_page
+ */
+static void add_empty_rpage_struct(struct pagewriter_buf *buf,
+				   struct relay_page *rpage)
+{
+	list_add_tail(&rpage->list, &buf->empty_rpage_structs);
+}
+
+/**
+ *	pagewriter_destroy - free the pagewriter struct
+ *	@kref: target kernel reference that contains the relay channel
+ *
+ *	Should only be called from kref_put().
+ */
+static void pagewriter_destroy(struct kref *kref)
+{
+	struct pagewriter *pagewriter = container_of(kref, struct pagewriter,
+						     kref);
+	kfree(pagewriter);
+}
+
+/**
+ *	pagewriter_destroy_buf - destroy a pagewriter_buf struct and page pool
+ *	@buf: the buffer struct
+ */
+static void pagewriter_destroy_buf(struct pagewriter_buf *buf)
+{
+	struct pagewriter *pagewriter = buf->pagewriter;
+	struct relay_page *rpage, *rpage2;
+
+	list_for_each_entry_safe(rpage, rpage2, &buf->pool, list) {
+		__free_page(rpage->page);
+		list_del(&rpage->list);
+		kfree(rpage);
+	}
+
+	pagewriter->buf[buf->cpu] = NULL;
+	kfree(buf);
+	kref_put(&pagewriter->kref, pagewriter_destroy);
+}
+
+/**
+ *	pagewriter_remove_buf - remove a pagewriter buffer
+ *	@kref: target kernel reference that contains the relay buffer
+ *
+ *	Frees the pagweriter_buf and the buffer's page pool.  Should
+ *	only be called from kref_put().
+ */
+static void pagewriter_remove_buf(struct kref *kref)
+{
+	struct pagewriter_buf *buf = container_of(kref, struct pagewriter_buf,
+						  kref);
+	pagewriter_destroy_buf(buf);
+}
+
+/**
+ *	pagewriter_close_buf - close a pagewriter buffer
+ *	@buf: channel buffer
+ *
+ *	The channel buffer and channel buffer data structure are freed
+ *	automatically when the last reference is given up.
+ */
+static void pagewriter_close_buf(struct pagewriter_buf *buf)
+{
+	kref_put(&buf->kref, pagewriter_remove_buf);
+}
+
+/**
+ *	pagewriter_alloc_pool - allocate a pool of pages for the buffer
+ *	@buf: the buffer struct
+ *
+ *	Allocates buf->pagewriter->n_pages pages to the buffer.
+ *	Returns 0 if successful.
+ */
+static int pagewriter_alloc_pool(struct pagewriter_buf *buf)
+{
+	unsigned int i;
+	struct relay_page *rpage = NULL;
+
+	for (i = 0; i < buf->pagewriter->n_pages; i++) {
+		rpage = kmalloc(sizeof(struct relay_page), GFP_KERNEL);
+		if (unlikely(!rpage))
+			goto depopulate;
+		rpage->page = alloc_page(GFP_KERNEL | __GFP_ZERO);
+		if (unlikely(!rpage->page))
+			goto depopulate;
+		list_add_tail(&rpage->list, &buf->pool);
+	}
+
+	return 0;
+
+depopulate:
+	list_for_each_entry(rpage, &buf->pool, list) {
+		__free_page(rpage->page);
+		list_del(&rpage->list);
+	}
+
+	return -ENOMEM;
+}
+
+/**
+ *	pagewriter_create_buf - allocate and initialize a buffer's page pool
+ *	@pagewriter: the pagewriter
+ *
+ *	Returns pagewriter buffer if successful, %NULL otherwise.
+ */
+static struct pagewriter_buf *pagewriter_create_buf(struct pagewriter *pw)
+{
+	struct pagewriter_buf *buf = kzalloc(sizeof(struct pagewriter_buf),
+					     GFP_KERNEL);
+	if (!buf)
+		return NULL;
+
+	INIT_LIST_HEAD(&buf->pool);
+	INIT_LIST_HEAD(&buf->empty_rpage_structs);
+	buf->pagewriter = pw;
+	kref_get(&buf->pagewriter->kref);
+
+	if (pagewriter_alloc_pool(buf))
+		goto free_buf;
+
+	return buf;
+
+free_buf:
+	kfree(buf);
+	return NULL;
+}
+
+/*
+ *	pagewriter_open_buf - create a new pagewriter buf with page pool
+ *
+ *	used by pagewriter_open() and CPU hotplug.
+ */
+static struct pagewriter_buf *pagewriter_open_buf(struct pagewriter *pagewriter,
+					     unsigned int cpu)
+{
+	struct pagewriter_buf *buf = NULL;
+
+	buf = pagewriter_create_buf(pagewriter);
+	if (!buf)
+		return NULL;
+
+	buf->cpu = cpu;
+
+	__pagewriter_reset(buf, 1);
+
+	return buf;
+}
+
+/*
+ * new_page() default callback.
+ */
+static void new_page_default_callback(struct pagewriter_buf *buf,
+				      void *page_data)
+{
+}
+
+/*
+ * write_padding() default callback.
+ */
+void pagewriter_write_padding_default_callback(struct pagewriter_buf *buf,
+					       size_t length,
+					       void *reserved)
+{
+}
+
+/* pagewriter default callbacks */
+static struct pagewriter_callbacks default_pagewriter_callbacks = {
+	.new_page = new_page_default_callback,
+	.switch_page = pagewriter_switch_page_default_callback,
+	.write_padding = pagewriter_write_padding_default_callback,
+};
+
+static void setup_callbacks(struct pagewriter *pagewriter,
+			    struct pagewriter_callbacks *cb)
+{
+	if (!cb) {
+		pagewriter->cb = &default_pagewriter_callbacks;
+		return;
+	}
+
+	if (!cb->new_page)
+		cb->new_page = new_page_default_callback;
+	if (!cb->switch_page)
+		cb->switch_page = pagewriter_switch_page_default_callback;
+	if (!cb->write_padding)
+		cb->write_padding = pagewriter_write_padding_default_callback;
+
+	pagewriter->cb = cb;
+}
+
+/**
+ * 	pagewriter_page_released_callback - relay_page page_released impl
+ * 	@page: the page released
+ * 	@private_data: contains associated pagewriter_buf
+ *
+ * 	relay has notified us that a page we gave it has been read and
+ * 	is now available for us to re-use.  We simply add it back to
+ * 	the page pool for that buf.
+ */
+static void pagewriter_page_released_callback(struct page *page,
+					      void *private_data)
+{
+	struct pagewriter_buf *buf = private_data;
+	struct relay_page *rpage = get_empty_rpage_struct(buf);
+
+	rpage->page = page;
+	pagewriter_add_free_page(buf, rpage);
+}
+
+/**
+ * 	pagewriter_page_stolen_callback - relay_page page_stolen impl
+ * 	@page: the page released
+ * 	@private_data: contains associated pagewriter_buf
+ *
+ * 	relay has notified us that a page we gave it has been stolen.
+ * 	We simply allocate a new one and add it to the page pool for
+ * 	that buf.
+ */
+static void pagewriter_page_stolen_callback(struct page *page,
+					    void *private_data)
+{
+	struct pagewriter_buf *buf = private_data;
+	struct relay_page *rpage;
+	struct page *new_page;
+
+	new_page = alloc_page(GFP_KERNEL | __GFP_ZERO);
+	if (unlikely(!new_page))
+		return;
+	set_page_private(new_page, (unsigned long)buf);
+	rpage = get_empty_rpage_struct(buf);
+
+	rpage->page = new_page;
+	pagewriter_add_free_page(buf, rpage);
+}
+
+static struct relay_page_callbacks pagewriter_relay_page_callbacks = {
+	.page_released	= pagewriter_page_released_callback,
+	.page_stolen	= pagewriter_page_stolen_callback,
+};
+
+/**
+ *	pagewriter_switch_page_default_callback - switch to a new page
+ *	@buf: channel buffer
+ *	@length: size of current event
+ *	@reserved: a pointer to the space reserved
+ *
+ *	Page switching function for pagewriter_write() functions,
+ *	which don't use padding because they write across page
+ *	boundaries.  Returns the remainder i.e. the amount that should
+ *	be written into the second page.
+ *
+ *	Performs page-switch tasks.
+ */
+size_t pagewriter_switch_page_default_callback(struct pagewriter_buf *buf,
+					       size_t length,
+					       void **reserved)
+{
+	size_t remainder = length;
+	struct relay_page *new_page;
+
+	if (unlikely(pagewriter_event_toobig(buf, length)))
+		goto toobig;
+
+	/* don't write anything unless we can write it all. */
+	new_page = pagewriter_get_free_page(buf);
+	if (!new_page) {
+		if (reserved)
+			*reserved = NULL;
+		atomic_inc(&buf->pagewriter->dropped);
+		return 0;
+	}
+
+	if (!(buf->pagewriter->flags & PAGEWRITER_PAD_WRITES))
+		remainder = length - (PAGE_SIZE - buf->offset);
+
+	if (buf->pagewriter->flags & PAGEWRITER_FLIGHT_MODE ||
+	    buf->pagewriter->flags & PAGEWRITER_LATE_SETUP) {
+		list_add_tail(&buf->page->list, &buf->pool);
+		buf->n_pages_flight++;
+	} else {
+		relay_add_page(buf->pagewriter->rchan, buf->page->page,
+			       &pagewriter_relay_page_callbacks, (void *)buf);
+		buf->page->page = NULL;
+		add_empty_rpage_struct(buf, buf->page);
+	}
+
+	buf->page = new_page;
+	buf->data = page_address(buf->page->page);
+
+	buf->offset = 0; /* remainder will be added by caller */
+	buf->pagewriter->cb->new_page(buf, buf->data);
+
+	if (unlikely(pagewriter_event_toobig(buf, length + buf->offset)))
+		goto toobig;
+
+	if (reserved)
+		*reserved = buf->data;
+
+	return remainder;
+toobig:
+	buf->pagewriter->last_toobig = length;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(pagewriter_switch_page_default_callback);
+
+/**
+ *	__pagewriter_reset - reset a pagewriter
+ *	@buf: the channel buffer
+ *	@init: 1 if this is a first-time initialization
+ *
+ *	See pagewriter_reset() for description of effect.
+ */
+static void __pagewriter_reset(struct pagewriter_buf *buf, unsigned int init)
+{
+	if (init)
+		kref_init(&buf->kref);
+
+	buf->page = pagewriter_get_free_page(buf);
+	buf->data = page_address(buf->page->page);
+	buf->offset = 0;
+	buf->n_pages_flight = 0;
+
+	buf->pagewriter->cb->new_page(buf, buf->data);
+}
+
+static void pagewriter_save_flight_buf(struct pagewriter_buf *buf)
+{
+	size_t first_page, n_pages = buf->n_pages_flight;
+	struct relay_page *first_rpage;
+
+	buf->pagewriter->cb->switch_page(buf, 0, NULL);
+
+	if(buf->n_pages_flight > buf->pagewriter->n_pages)
+		n_pages = buf->pagewriter->n_pages;
+
+	first_page = buf->pagewriter->n_pages - n_pages;
+	list_for_each_entry(first_rpage, &buf->pool, list)
+		if (!first_page--)
+			break;
+
+	list_for_each_entry_from(first_rpage, &buf->pool, list) {
+		relay_add_page(buf->pagewriter->rchan, buf->page->page,
+			       &pagewriter_relay_page_callbacks, (void *)buf);
+		buf->page->page = NULL;
+		add_empty_rpage_struct(buf, buf->page);
+	}
+
+	__pagewriter_reset(buf, 0);
+}
+
+/**
+ * 	pagewriter_hotcpu_callback - CPU hotplug callback
+ * 	@nb: notifier block
+ * 	@action: hotplug action to take
+ * 	@hcpu: CPU number
+ *
+ * 	Returns the success/failure of the operation. (%NOTIFY_OK, %NOTIFY_BAD)
+ */
+static int __cpuinit pagewriter_hotcpu_callback(struct notifier_block *nb,
+						unsigned long action,
+						void *hcpu)
+{
+	unsigned int hotcpu = (unsigned long)hcpu;
+	struct pagewriter *pagewriter;
+
+	switch (action) {
+	case CPU_UP_PREPARE:
+	case CPU_UP_PREPARE_FROZEN:
+		mutex_lock(&pagewriters_mutex);
+		list_for_each_entry(pagewriter, &pagewriters, list) {
+			if (pagewriter->buf[hotcpu])
+				continue;
+			pagewriter->buf[hotcpu] =
+				pagewriter_open_buf(pagewriter, hotcpu);
+			if (!pagewriter->buf[hotcpu]) {
+				printk(KERN_ERR
+					"pagewriter_hotcpu_callback: cpu %d "
+				       "buffer creation failed\n", hotcpu);
+				mutex_unlock(&pagewriters_mutex);
+				return NOTIFY_BAD;
+			}
+		}
+		mutex_unlock(&pagewriters_mutex);
+		break;
+	case CPU_DEAD:
+	case CPU_DEAD_FROZEN:
+		/* No need to flush the cpu : will be flushed upon
+		 * final relay_flush() call. */
+		break;
+	}
+	return NOTIFY_OK;
+}
+
+static __init int pagewriter_init(void)
+{
+
+	hotcpu_notifier(pagewriter_hotcpu_callback, 0);
+	return 0;
+}
+
+early_initcall(pagewriter_init);





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