[PATCH] TTY layer buffering revamp

The API and code have been through various bits of initial review by
serial driver people but they definitely need to live somewhere for a
while so the unconverted drivers can get knocked into shape, existing
drivers that have been updated can be better tuned and bugs whacked out.

This replaces the tty flip buffers with kmalloc objects in rings. In the
normal situation for an IRQ driven serial port at typical speeds the
behaviour is pretty much the same, two buffers end up allocated and the
kernel cycles between them as before.

When there are delays or at high speed we now behave far better as the
buffer pool can grow a bit rather than lose characters. This also means
that we can operate at higher speeds reliably.

For drivers that receive characters in blocks (DMA based, USB and
especially virtualisation) the layer allows a lot of driver specific
code that works around the tty layer with private secondary queues to be
removed. The IBM folks need this sort of layer, the smart serial port
people do, the virtualisers do (because a virtualised tty typically
operates at infinite speed rather than emulating 9600 baud).

Finally many drivers had invalid and unsafe attempts to avoid buffer
overflows by directly invoking tty methods extracted out of the innards
of work queue structs. These are no longer needed and all go away. That
fixes various random hangs with serial ports on overflow.

The other change in here is to optimise the receive_room path that is
used by some callers. It turns out that only one ldisc uses receive room
except asa constant and it updates it far far less than the value is
read. We thus make it a variable not a function call.

I expect the code to contain bugs due to the size alone but I'll be
watching and squashing them and feeding out new patches as it goes.

Because the buffers now dynamically expand you should only run out of
buffering when the kernel runs out of memory for real.  That means a lot of
the horrible hacks high performance drivers used to do just aren't needed any
more.

Description:

tty_insert_flip_char is an old API and continues to work as before, as does
tty_flip_buffer_push() [this is why many drivers dont need modification].  It
does now also return the number of chars inserted

There are also

tty_buffer_request_room(tty, len)

which asks for a buffer block of the length requested and returns the space
found.  This improves efficiency with hardware that knows how much to
transfer.

and tty_insert_flip_string_flags(tty, str, flags, len)

to insert a string of characters and flags

For a smart interface the usual code is

    len = tty_request_buffer_room(tty, amount_hardware_says);
    tty_insert_flip_string(tty, buffer_from_card, len);

More description!

At the moment tty buffers are attached directly to the tty.  This is causing a
lot of the problems related to tty layer locking, also problems at high speed
and also with bursty data (such as occurs in virtualised environments)

I'm working on ripping out the flip buffers and replacing them with a pool of
dynamically allocated buffers.  This allows both for old style "byte I/O"
devices and also helps virtualisation and smart devices where large blocks of
data suddenely materialise and need storing.

So far so good.  Lots of drivers reference tty->flip.*.  Several of them also
call directly and unsafely into function pointers it provides.  This will all
break.  Most drivers can use tty_insert_flip_char which can be kept as an API
but others need more.

At the moment I've added the following interfaces, if people think more will
be needed now is a good time to say

 int tty_buffer_request_room(tty, size)

Try and ensure at least size bytes are available, returns actual room (may be
zero).  At the moment it just uses the flipbuf space but that will change.
Repeated calls without characters being added are not cumulative.  (ie if you
call it with 1, 1, 1, and then 4 you'll have four characters of space.  The
other functions will also try and grow buffers in future but this will be a
more efficient way when you know block sizes.

 int tty_insert_flip_char(tty, ch, flag)

As before insert a character if there is room.  Now returns 1 for success, 0
for failure.

 int tty_insert_flip_string(tty, str, len)

Insert a block of non error characters.  Returns the number inserted.

 int tty_prepare_flip_string(tty, strptr, len)

Adjust the buffer to allow len characters to be added.  Returns a buffer
pointer in strptr and the length available.  This allows for hardware that
needs to use functions like insl or mencpy_fromio.

Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Paul Fulghum <paulkf@microgate.com>
Signed-off-by: Hirokazu Takata <takata@linux-m32r.org>
Signed-off-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: Jeff Dike <jdike@addtoit.com>
Signed-off-by: John Hawkes <hawkes@sgi.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

1 files changed, 234 insertions, 32 deletions

diff --git a/drivers/char/tty_io.c b/drivers/char/tty_io.c
index 4b1eef51ec59..1eda82b31a61 100644
--- a/drivers/char/tty_io.c
+++ b/drivers/char/tty_io.c
@@ -166,9 +166,12 @@ static struct tty_struct *alloc_tty_struct(void)
 	return tty;
 }
 
+static void tty_buffer_free_all(struct tty_struct *);
+
 static inline void free_tty_struct(struct tty_struct *tty)
 {
 	kfree(tty->write_buf);
+	tty_buffer_free_all(tty);
 	kfree(tty);
 }
 
@@ -231,6 +234,201 @@ static int check_tty_count(struct tty_struct *tty, const char *routine)
 }
 
 /*
+ * Tty buffer allocation management
+ */
+
+static void tty_buffer_free_all(struct tty_struct *tty)
+{
+	struct tty_buffer *thead;
+	while((thead = tty->buf.head) != NULL) {
+		tty->buf.head = thead->next;
+		kfree(thead);
+	}
+	while((thead = tty->buf.free) != NULL) {
+		tty->buf.free = thead->next;
+		kfree(thead);
+	}
+	tty->buf.tail = NULL;
+}
+
+static void tty_buffer_init(struct tty_struct *tty)
+{
+	tty->buf.head = NULL;
+	tty->buf.tail = NULL;
+	tty->buf.free = NULL;
+}
+
+static struct tty_buffer *tty_buffer_alloc(size_t size)
+{
+	struct tty_buffer *p = kmalloc(sizeof(struct tty_buffer) + 2 * size, GFP_ATOMIC);
+	if(p == NULL)
+		return NULL;
+	p->used = 0;
+	p->size = size;
+	p->next = NULL;
+	p->char_buf_ptr = (char *)(p->data);
+	p->flag_buf_ptr = (unsigned char *)p->char_buf_ptr + size;
+/* 	printk("Flip create %p\n", p); */
+	return p;
+}
+
+/* Must be called with the tty_read lock held. This needs to acquire strategy
+   code to decide if we should kfree or relink a given expired buffer */
+
+static void tty_buffer_free(struct tty_struct *tty, struct tty_buffer *b)
+{
+	/* Dumb strategy for now - should keep some stats */
+/* 	printk("Flip dispose %p\n", b); */
+	if(b->size >= 512)
+		kfree(b);
+	else {
+		b->next = tty->buf.free;
+		tty->buf.free = b;
+	}
+}
+
+static struct tty_buffer *tty_buffer_find(struct tty_struct *tty, size_t size)
+{
+	struct tty_buffer **tbh = &tty->buf.free;
+	while((*tbh) != NULL) {
+		struct tty_buffer *t = *tbh;
+		if(t->size >= size) {
+			*tbh = t->next;
+			t->next = NULL;
+			t->used = 0;
+			/* DEBUG ONLY */
+			memset(t->data, '*', size);
+/* 			printk("Flip recycle %p\n", t); */
+			return t;
+		}
+		tbh = &((*tbh)->next);
+	}
+	/* Round the buffer size out */
+	size = (size + 0xFF) & ~ 0xFF;
+	return tty_buffer_alloc(size);
+	/* Should possibly check if this fails for the largest buffer we
+	   have queued and recycle that ? */
+}
+
+int tty_buffer_request_room(struct tty_struct *tty, size_t size)
+{
+	struct tty_buffer *b = tty->buf.head, *n;
+	int left = 0;
+
+	/* OPTIMISATION: We could keep a per tty "zero" sized buffer to
+	   remove this conditional if its worth it. This would be invisible
+	   to the callers */
+	if(b != NULL)
+		left = b->size - b->used;
+	if(left >= size)
+		return size;
+	/* This is the slow path - looking for new buffers to use */
+	n = tty_buffer_find(tty, size);
+	if(n == NULL)
+		return left;
+	n->next = b;
+	if(b != NULL)
+		b->next = n;
+	else
+		tty->buf.head = n;
+	tty->buf.tail = n;
+	return size;
+}
+
+EXPORT_SYMBOL_GPL(tty_buffer_request_room);
+
+int tty_insert_flip_string(struct tty_struct *tty, unsigned char *chars, size_t size)
+{
+	int copied = 0;
+	do {
+		int space = tty_buffer_request_room(tty, size - copied);
+		struct tty_buffer *tb = tty->buf.tail;
+		/* If there is no space then tb may be NULL */
+		if(unlikely(space == 0))
+			break;
+		memcpy(tb->char_buf_ptr + tb->used, chars, space);
+		memset(tb->flag_buf_ptr + tb->used, TTY_NORMAL, space);
+		tb->used += space;
+		copied += space;
+		chars += space;
+/* 		printk("Flip insert %d.\n", space); */
+	}
+	/* There is a small chance that we need to split the data over
+	   several buffers. If this is the case we must loop */
+	while (unlikely(size > copied));
+	return copied;
+}
+
+EXPORT_SYMBOL_GPL(tty_insert_flip_string);
+
+int tty_insert_flip_string_flags(struct tty_struct *tty, unsigned char *chars, char *flags, size_t size)
+{
+	int copied = 0;
+	do {
+		int space = tty_buffer_request_room(tty, size - copied);
+		struct tty_buffer *tb = tty->buf.tail;
+		/* If there is no space then tb may be NULL */
+		if(unlikely(space == 0))
+			break;
+		memcpy(tb->char_buf_ptr + tb->used, chars, space);
+		memcpy(tb->flag_buf_ptr + tb->used, flags, space);
+		tb->used += space;
+		copied += space;
+		chars += space;
+		flags += space;
+	}
+	/* There is a small chance that we need to split the data over
+	   several buffers. If this is the case we must loop */
+	while (unlikely(size > copied));
+	return copied;
+}
+
+EXPORT_SYMBOL_GPL(tty_insert_flip_string_flags);
+
+
+/*
+ *	Prepare a block of space in the buffer for data. Returns the length
+ *	available and buffer pointer to the space which is now allocated and
+ *	accounted for as ready for normal characters. This is used for drivers
+ *	that need their own block copy routines into the buffer. There is no
+ *	guarantee the buffer is a DMA target!
+ */
+
+int tty_prepare_flip_string(struct tty_struct *tty, unsigned char **chars, size_t size)
+{
+	int space = tty_buffer_request_room(tty, size);
+	struct tty_buffer *tb = tty->buf.tail;
+	*chars = tb->char_buf_ptr + tb->used;
+	memset(tb->flag_buf_ptr + tb->used, TTY_NORMAL, space);
+	tb->used += space;
+	return space;
+}
+
+EXPORT_SYMBOL_GPL(tty_prepare_flip_string);
+
+/*
+ *	Prepare a block of space in the buffer for data. Returns the length
+ *	available and buffer pointer to the space which is now allocated and
+ *	accounted for as ready for characters. This is used for drivers
+ *	that need their own block copy routines into the buffer. There is no
+ *	guarantee the buffer is a DMA target!
+ */
+
+int tty_prepare_flip_string_flags(struct tty_struct *tty, unsigned char **chars, char **flags, size_t size)
+{
+	int space = tty_buffer_request_room(tty, size);
+	struct tty_buffer *tb = tty->buf.tail;
+	*chars = tb->char_buf_ptr + tb->used;
+	*flags = tb->flag_buf_ptr + tb->used;
+	tb->used += space;
+	return space;
+}
+
+EXPORT_SYMBOL_GPL(tty_prepare_flip_string_flags);
+
+
+
+/*
  *	This is probably overkill for real world processors but
  *	they are not on hot paths so a little discipline won't do 
  *	any harm.
@@ -492,6 +690,17 @@ restart:
 	if (ld == NULL)
 		return -EINVAL;
 
+	/*
+	 *	No more input please, we are switching. The new ldisc
+	 *	will update this value in the ldisc open function
+	 */
+
+	tty->receive_room = 0;
+
+	/*
+	 *	Problem: What do we do if this blocks ?
+	 */
+
 	tty_wait_until_sent(tty, 0);
 
 	if (tty->ldisc.num == ldisc) {
@@ -560,9 +769,9 @@ restart:
 	 *	we say so later on.
 	 */
 
-	work = cancel_delayed_work(&tty->flip.work);
+	work = cancel_delayed_work(&tty->buf.work);
 	/*
-	 * Wait for ->hangup_work and ->flip.work handlers to terminate
+	 * Wait for ->hangup_work and ->buf.work handlers to terminate
 	 */
 	 
 	flush_scheduled_work();
@@ -616,7 +825,7 @@ restart:
 	/* Restart it in case no characters kick it off. Safe if
 	   already running */
 	if (work)
-		schedule_delayed_work(&tty->flip.work, 1);
+		schedule_delayed_work(&tty->buf.work, 1);
 	return retval;
 }
 
@@ -1721,10 +1930,10 @@ static void release_dev(struct file * filp)
 	 */
 	clear_bit(TTY_LDISC, &tty->flags);
 	clear_bit(TTY_DONT_FLIP, &tty->flags);
-	cancel_delayed_work(&tty->flip.work);
+	cancel_delayed_work(&tty->buf.work);
 
 	/*
-	 * Wait for ->hangup_work and ->flip.work handlers to terminate
+	 * Wait for ->hangup_work and ->buf.work handlers to terminate
 	 */
 	 
 	flush_scheduled_work();
@@ -2518,17 +2727,15 @@ EXPORT_SYMBOL(do_SAK);
 
 /*
  * This routine is called out of the software interrupt to flush data
- * from the flip buffer to the line discipline. 
+ * from the buffer chain to the line discipline.
  */
  
 static void flush_to_ldisc(void *private_)
 {
 	struct tty_struct *tty = (struct tty_struct *) private_;
-	unsigned char	*cp;
-	char		*fp;
-	int		count;
 	unsigned long 	flags;
 	struct tty_ldisc *disc;
+	struct tty_buffer *tbuf;
 
 	disc = tty_ldisc_ref(tty);
 	if (disc == NULL)	/*  !TTY_LDISC */
@@ -2538,28 +2745,22 @@ static void flush_to_ldisc(void *private_)
 		/*
 		 * Do it after the next timer tick:
 		 */
-		schedule_delayed_work(&tty->flip.work, 1);
+		schedule_delayed_work(&tty->buf.work, 1);
 		goto out;
 	}
 	spin_lock_irqsave(&tty->read_lock, flags);
-	if (tty->flip.buf_num) {
-		cp = tty->flip.char_buf + TTY_FLIPBUF_SIZE;
-		fp = tty->flip.flag_buf + TTY_FLIPBUF_SIZE;
-		tty->flip.buf_num = 0;
-		tty->flip.char_buf_ptr = tty->flip.char_buf;
-		tty->flip.flag_buf_ptr = tty->flip.flag_buf;
-	} else {
-		cp = tty->flip.char_buf;
-		fp = tty->flip.flag_buf;
-		tty->flip.buf_num = 1;
-		tty->flip.char_buf_ptr = tty->flip.char_buf + TTY_FLIPBUF_SIZE;
-		tty->flip.flag_buf_ptr = tty->flip.flag_buf + TTY_FLIPBUF_SIZE;
-	}
-	count = tty->flip.count;
-	tty->flip.count = 0;
+	while((tbuf = tty->buf.head) != NULL) {
+		tty->buf.head = tbuf->next;
+		spin_unlock_irqrestore(&tty->read_lock, flags);
+		/* printk("Process buffer %p for %d\n", tbuf, tbuf->used); */
+		disc->receive_buf(tty, tbuf->char_buf_ptr,
+				       tbuf->flag_buf_ptr,
+				       tbuf->used);
+		spin_lock_irqsave(&tty->read_lock, flags);
+		tty_buffer_free(tty, tbuf);
+	}
+	tty->buf.tail = NULL;
 	spin_unlock_irqrestore(&tty->read_lock, flags);
-
-	disc->receive_buf(tty, cp, fp, count);
 out:
 	tty_ldisc_deref(disc);
 }
@@ -2654,11 +2855,12 @@ void tty_flip_buffer_push(struct tty_struct *tty)
 	if (tty->low_latency)
 		flush_to_ldisc((void *) tty);
 	else
-		schedule_delayed_work(&tty->flip.work, 1);
+		schedule_delayed_work(&tty->buf.work, 1);
 }
 
 EXPORT_SYMBOL(tty_flip_buffer_push);
 
+
 /*
  * This subroutine initializes a tty structure.
  */
@@ -2669,10 +2871,10 @@ static void initialize_tty_struct(struct tty_struct *tty)
 	tty_ldisc_assign(tty, tty_ldisc_get(N_TTY));
 	tty->pgrp = -1;
 	tty->overrun_time = jiffies;
-	tty->flip.char_buf_ptr = tty->flip.char_buf;
-	tty->flip.flag_buf_ptr = tty->flip.flag_buf;
-	INIT_WORK(&tty->flip.work, flush_to_ldisc, tty);
-	init_MUTEX(&tty->flip.pty_sem);
+	tty->buf.head = tty->buf.tail = NULL;
+	tty_buffer_init(tty);
+	INIT_WORK(&tty->buf.work, flush_to_ldisc, tty);
+	init_MUTEX(&tty->buf.pty_sem);
 	init_MUTEX(&tty->termios_sem);
 	init_waitqueue_head(&tty->write_wait);
 	init_waitqueue_head(&tty->read_wait);