[PATCH] esp: Make driver SMP-correct

[PATCH] esp: Make driver SMP-correct

[James Nelson]

This is an attempt to make the esp serial driver SMP-correct.  It also removes
some cruft left over from the serial_write() conversion.

Diffstat output:
 drivers/char/Kconfig     |    2
 drivers/char/esp.c       |  299 ++++++++++++++++++++++++-----------------------
 include/linux/hayesesp.h |    1
 3 files changed, 160 insertions(+), 142 deletions(-)

Signed-off-by: James Nelson <james4765@gmail.com>

diff -urN --exclude='*~' linux-2.6.10-original/drivers/char/esp.c linux-2.6.10/drivers/char/esp.c
--- linux-2.6.10-original/drivers/char/esp.c	2004-12-24 16:33:48.000000000 -0500
+++ linux-2.6.10/drivers/char/esp.c	2004-12-30 20:33:46.243770922 -0500
@@ -152,18 +152,6 @@
 /* Standard COM flags (except for COM4, because of the 8514 problem) */
 #define STD_COM_FLAGS (ASYNC_BOOT_AUTOCONF | ASYNC_SKIP_TEST)
 
-/*
- * tmp_buf is used as a temporary buffer by serial_write.  We need to
- * lock it in case the memcpy_fromfs blocks while swapping in a page,
- * and some other program tries to do a serial write at the same time.
- * Since the lock will only come under contention when the system is
- * swapping and available memory is low, it makes sense to share one
- * buffer across all the serial ports, since it significantly saves
- * memory if large numbers of serial ports are open.
- */
-static unsigned char *tmp_buf;
-static DECLARE_MUTEX(tmp_buf_sem);
-
 static inline int serial_paranoia_check(struct esp_struct *info,
 					char *name, const char *routine)
 {
@@ -209,34 +197,38 @@
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
 	unsigned long flags;
 
-	if (serial_paranoia_check(info, tty->name, "rs_stop"))
-		return;
+	spin_lock_irqsave(&info->irq_lock, flags);
+
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
 	
-	save_flags(flags); cli();
 	if (info->IER & UART_IER_THRI) {
 		info->IER &= ~UART_IER_THRI;
 		serial_out(info, UART_ESI_CMD1, ESI_SET_SRV_MASK);
 		serial_out(info, UART_ESI_CMD2, info->IER);
 	}
 
-	restore_flags(flags);
+out:
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 static void rs_start(struct tty_struct *tty)
 {
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
 	unsigned long flags;
-	
-	if (serial_paranoia_check(info, tty->name, "rs_start"))
-		return;
-	
-	save_flags(flags); cli();
+
+	spin_lock_irqsave(&info->irq_lock, flags);
+
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
+
 	if (info->xmit_cnt && info->xmit_buf && !(info->IER & UART_IER_THRI)) {
 		info->IER |= UART_IER_THRI;
 		serial_out(info, UART_ESI_CMD1, ESI_SET_SRV_MASK);
 		serial_out(info, UART_ESI_CMD2, info->IER);
 	}
-	restore_flags(flags);
+out:
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 /*
@@ -311,7 +303,7 @@
 		return;
 	}
 
-	sti();
+	spin_unlock_irq(&info->irq_lock);
 	
 	status_mask = (info->read_status_mask >> 2) & 0x07;
 		
@@ -329,7 +321,7 @@
 			(serial_in(info, UART_ESI_RWS) >> 3) & status_mask;
 	}
 
-	cli();
+	spin_lock_irq(&info->irq_lock);
 
 	/* make sure everything is still ok since interrupts were enabled */
 	tty = info->tty;
@@ -478,7 +470,7 @@
 		info->xmit_tail = (info->xmit_tail + space_avail) &
 			(ESP_XMIT_SIZE - 1);
 
-		sti();
+		spin_unlock_irq(&info->irq_lock);
 
 		for (i = 0; i < space_avail - 1; i += 2) {
 			outw(*((unsigned short *)(pio_buf->data + i)),
@@ -489,7 +481,7 @@
 			serial_out(info, UART_ESI_TX,
 				   pio_buf->data[space_avail - 1]);
 
-		cli();
+		spin_lock_irq(&info->irq_lock);
 
 		if (info->xmit_cnt) {
 			serial_out(info, UART_ESI_CMD1, ESI_NO_COMMAND);
@@ -654,10 +646,10 @@
 	err_status = 0;
 	scratch = serial_in(info, UART_ESI_SID);
 
-	cli();
+	spin_lock_irq(&info->irq_lock);
 	
 	if (!info->tty) {
-		sti();
+		spin_unlock_irq(&info->irq_lock);
 		return IRQ_NONE;
 	}
 
@@ -740,7 +732,7 @@
 #ifdef SERIAL_DEBUG_INTR
 	printk("end.\n");
 #endif
-	sti();
+	spin_unlock_irq(&info->irq_lock);
 	return IRQ_HANDLED;
 }
 
@@ -859,7 +851,7 @@
 	int	retval=0;
         unsigned int num_chars;
 
-	save_flags(flags); cli();
+	spin_lock_irqsave(&info->irq_lock, flags);
 
 	if (info->flags & ASYNC_INITIALIZED)
 		goto out;
@@ -973,7 +965,7 @@
 
 	info->flags |= ASYNC_INITIALIZED;
 	retval = 0;
-out:	restore_flags(flags);
+out:	spin_unlock_irqrestore(&info->irq_lock, flags);
 	return retval;
 }
 
@@ -993,7 +985,7 @@
 	       info->irq);
 #endif
 	
-	save_flags(flags); cli(); /* Disable interrupts */
+	spin_lock_irqsave(&info->irq_lock, flags); /* Disable interrupts */
 
 	/*
 	 * clear delta_msr_wait queue to avoid mem leaks: we may free the irq
@@ -1058,7 +1050,7 @@
 		set_bit(TTY_IO_ERROR, &info->tty->flags);
 	
 	info->flags &= ~ASYNC_INITIALIZED;
-	restore_flags(flags);
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 /*
@@ -1172,7 +1164,7 @@
 	if (I_IXOFF(info->tty))
 		flow1 |= 0x81;
 
-	save_flags(flags); cli();
+	spin_lock_irqsave(&info->irq_lock, flags);
 	/* set baud */
 	serial_out(info, UART_ESI_CMD1, ESI_SET_BAUD);
 	serial_out(info, UART_ESI_CMD2, quot >> 8);
@@ -1219,7 +1211,7 @@
 	serial_out(info, UART_ESI_CMD2, info->config.flow_on >> 8);
 	serial_out(info, UART_ESI_CMD2, info->config.flow_on);
 
-	restore_flags(flags);
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 static void rs_put_char(struct tty_struct *tty, unsigned char ch)
@@ -1227,42 +1219,44 @@
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
 	unsigned long flags;
 
-	if (serial_paranoia_check(info, tty->name, "rs_put_char"))
-		return;
+	spin_lock_irqsave(&info->irq_lock, flags);
+
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
 
 	if (!tty || !info->xmit_buf)
-		return;
+		goto out;
 
-	save_flags(flags); cli();
-	if (info->xmit_cnt >= ESP_XMIT_SIZE - 1) {
-		restore_flags(flags);
-		return;
-	}
+	if (info->xmit_cnt >= ESP_XMIT_SIZE - 1)
+		goto out;
 
 	info->xmit_buf[info->xmit_head++] = ch;
 	info->xmit_head &= ESP_XMIT_SIZE-1;
 	info->xmit_cnt++;
-	restore_flags(flags);
+out:
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 static void rs_flush_chars(struct tty_struct *tty)
 {
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
 	unsigned long flags;
-				
-	if (serial_paranoia_check(info, tty->name, "rs_flush_chars"))
-		return;
+
+	spin_lock_irqsave(&info->irq_lock, flags);
+
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
 
 	if (info->xmit_cnt <= 0 || tty->stopped || !info->xmit_buf)
-		return;
+		goto out;
 
-	save_flags(flags); cli();
 	if (!(info->IER & UART_IER_THRI)) {
 		info->IER |= UART_IER_THRI;
 		serial_out(info, UART_ESI_CMD1, ESI_SET_SRV_MASK);
 		serial_out(info, UART_ESI_CMD2, info->IER);
 	}
-	restore_flags(flags);
+out:
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 static int rs_write(struct tty_struct * tty,
@@ -1272,16 +1266,25 @@
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
 	unsigned long flags;
 
-	if (serial_paranoia_check(info, tty->name, "rs_write"))
-		return 0;
+	spin_lock_irqsave(&info->irq_lock, flags);
 
-	if (!tty || !info->xmit_buf || !tmp_buf)
-		return 0;
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
+
+	if (!tty || !info->xmit_buf)
+		goto out;
 	    
 	while (1) {
 		/* Thanks to R. Wolff for suggesting how to do this with */
 		/* interrupts enabled */
 
+		/*
+		 * Sorry about the interrupt-enabled trick, but it is not
+		 * SMP-safe, since this is no longer protected by the 2.4-era
+		 * write buffer semaphore. It needs to be done with the
+		 * interrupt spinlock aquired - Jim Nelson
+		 */
+
 		c = count;
 		t = ESP_XMIT_SIZE - info->xmit_cnt - 1;
 		
@@ -1305,15 +1308,14 @@
 		ret += c;
 	}
 
-	save_flags(flags); cli();
-
 	if (info->xmit_cnt && !tty->stopped && !(info->IER & UART_IER_THRI)) {
 		info->IER |= UART_IER_THRI;
 		serial_out(info, UART_ESI_CMD1, ESI_SET_SRV_MASK);
 		serial_out(info, UART_ESI_CMD2, info->IER);
 	}
 
-	restore_flags(flags);
+out:
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 	return ret;
 }
 
@@ -1322,7 +1324,7 @@
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
 	int	ret;
 				
-	if (serial_paranoia_check(info, tty->name, "rs_write_room"))
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
 		return 0;
 	ret = ESP_XMIT_SIZE - info->xmit_cnt - 1;
 	if (ret < 0)
@@ -1334,7 +1336,7 @@
 {
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
 				
-	if (serial_paranoia_check(info, tty->name, "rs_chars_in_buffer"))
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
 		return 0;
 	return info->xmit_cnt;
 }
@@ -1342,12 +1344,15 @@
 static void rs_flush_buffer(struct tty_struct *tty)
 {
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
-				
-	if (serial_paranoia_check(info, tty->name, "rs_flush_buffer"))
+	unsigned long flags;
+
+	spin_lock_irqsave(&info->irq_lock, flags);
+	if (serial_paranoia_check(info, tty->name, __FUNCTION__)) {
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 		return;
-	cli();
+	}
 	info->xmit_cnt = info->xmit_head = info->xmit_tail = 0;
-	sti();
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 	tty_wakeup(tty);
 }
 
@@ -1362,6 +1367,7 @@
 static void rs_throttle(struct tty_struct * tty)
 {
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
+	unsigned long flags;
 #ifdef SERIAL_DEBUG_THROTTLE
 	char	buf[64];
 	
@@ -1369,38 +1375,43 @@
 	       tty->ldisc.chars_in_buffer(tty));
 #endif
 
-	if (serial_paranoia_check(info, tty->name, "rs_throttle"))
-		return;
+	spin_lock_irqsave(&info->irq_lock, flags);
+
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
 	
-	cli();
 	info->IER &= ~UART_IER_RDI;
 	serial_out(info, UART_ESI_CMD1, ESI_SET_SRV_MASK);
 	serial_out(info, UART_ESI_CMD2, info->IER);
 	serial_out(info, UART_ESI_CMD1, ESI_SET_RX_TIMEOUT);
 	serial_out(info, UART_ESI_CMD2, 0x00);
-	sti();
+out:
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 static void rs_unthrottle(struct tty_struct * tty)
 {
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
+	unsigned long flags;
 #ifdef SERIAL_DEBUG_THROTTLE
 	char	buf[64];
 	
 	printk("unthrottle %s: %d....\n", tty_name(tty, buf),
 	       tty->ldisc.chars_in_buffer(tty));
 #endif
+	
+	spin_lock_irqsave(&info->irq_lock, flags);
 
-	if (serial_paranoia_check(info, tty->name, "rs_unthrottle"))
-		return;
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
 	
-	cli();
 	info->IER |= UART_IER_RDI;
 	serial_out(info, UART_ESI_CMD1, ESI_SET_SRV_MASK);
 	serial_out(info, UART_ESI_CMD2, info->IER);
 	serial_out(info, UART_ESI_CMD1, ESI_SET_RX_TIMEOUT);
 	serial_out(info, UART_ESI_CMD2, info->config.rx_timeout);
-	sti();
+out:
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 /*
@@ -1573,6 +1584,7 @@
 	unsigned int change_dma;
 	int retval = 0;
 	struct esp_struct *current_async;
+	unsigned long flags;
 
 	/* Perhaps a non-sysadmin user should be able to do some of these */
 	/* operations.  I haven't decided yet. */
@@ -1650,26 +1662,24 @@
 
 	if ((new_config.flow_off != info->config.flow_off) ||
 	    (new_config.flow_on != info->config.flow_on)) {
-		unsigned long flags;
 
+		spin_lock_irqsave(&info->irq_lock, flags);
 		info->config.flow_off = new_config.flow_off;
 		info->config.flow_on = new_config.flow_on;
-		save_flags(flags); cli();
 		serial_out(info, UART_ESI_CMD1, ESI_SET_FLOW_LVL);
 		serial_out(info, UART_ESI_CMD2, new_config.flow_off >> 8);
 		serial_out(info, UART_ESI_CMD2, new_config.flow_off);
 		serial_out(info, UART_ESI_CMD2, new_config.flow_on >> 8);
 		serial_out(info, UART_ESI_CMD2, new_config.flow_on);
-		restore_flags(flags);
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 	}
 
 	if ((new_config.rx_trigger != info->config.rx_trigger) ||
 	    (new_config.tx_trigger != info->config.tx_trigger)) {
-		unsigned long flags;
 
+		spin_lock_irqsave(&info->irq_lock, flags);
 		info->config.rx_trigger = new_config.rx_trigger;
 		info->config.tx_trigger = new_config.tx_trigger;
-		save_flags(flags); cli();
 		serial_out(info, UART_ESI_CMD1, ESI_SET_TRIGGER);
 		serial_out(info, UART_ESI_CMD2,
 			   new_config.rx_trigger >> 8);
@@ -1677,14 +1687,13 @@
 		serial_out(info, UART_ESI_CMD2,
 			   new_config.tx_trigger >> 8);
 		serial_out(info, UART_ESI_CMD2, new_config.tx_trigger);
-		restore_flags(flags);
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 	}
 
 	if (new_config.rx_timeout != info->config.rx_timeout) {
-		unsigned long flags;
 
+		spin_lock_irqsave(&info->irq_lock, flags);
 		info->config.rx_timeout = new_config.rx_timeout;
-		save_flags(flags); cli();
 
 		if (info->IER & UART_IER_RDI) {
 			serial_out(info, UART_ESI_CMD1,
@@ -1693,7 +1702,7 @@
 				   new_config.rx_timeout);
 		}
 
-		restore_flags(flags);
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 	}
 
 	if (!(info->flags & ASYNC_INITIALIZED))
@@ -1716,11 +1725,12 @@
 {
 	unsigned char status;
 	unsigned int result;
+	unsigned long flags;
 
-	cli();
+	spin_lock_irqsave(&info->irq_lock, flags);
 	serial_out(info, UART_ESI_CMD1, ESI_GET_UART_STAT);
 	status = serial_in(info, UART_ESI_STAT1);
-	sti();
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 	result = ((status & UART_LSR_TEMT) ? TIOCSER_TEMT : 0);
 	return put_user(result,value);
 }
@@ -1730,17 +1740,23 @@
 {
 	struct esp_struct * info = (struct esp_struct *)tty->driver_data;
 	unsigned char control, status;
+	unsigned long flags;
 
-	if (serial_paranoia_check(info, tty->name, __FUNCTION__))
+	spin_lock_irqsave(&info->irq_lock, flags);
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__))) {
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 		return -ENODEV;
-	if (tty->flags & (1 << TTY_IO_ERROR))
+	}
+	if (unlikely(tty->flags & (1 << TTY_IO_ERROR))) {
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 		return -EIO;
+	}
 
 	control = info->MCR;
-	cli();
+
 	serial_out(info, UART_ESI_CMD1, ESI_GET_UART_STAT);
 	status = serial_in(info, UART_ESI_STAT2);
-	sti();
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 	return    ((control & UART_MCR_RTS) ? TIOCM_RTS : 0)
 		| ((control & UART_MCR_DTR) ? TIOCM_DTR : 0)
 		| ((status  & UART_MSR_DCD) ? TIOCM_CAR : 0)
@@ -1753,13 +1769,17 @@
 			unsigned int set, unsigned int clear)
 {
 	struct esp_struct * info = (struct esp_struct *)tty->driver_data;
+	unsigned long flags;
 
-	if (serial_paranoia_check(info, tty->name, __FUNCTION__))
+	spin_lock_irqsave(&info->irq_lock, flags);
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__))) {
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 		return -ENODEV;
-	if (tty->flags & (1 << TTY_IO_ERROR))
+	}
+	if (unlikely(tty->flags & (1 << TTY_IO_ERROR))) {
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 		return -EIO;
-
-	cli();
+	}
 
 	if (set & TIOCM_RTS)
 		info->MCR |= UART_MCR_RTS;
@@ -1774,7 +1794,7 @@
 	serial_out(info, UART_ESI_CMD1, ESI_WRITE_UART);
 	serial_out(info, UART_ESI_CMD2, UART_MCR);
 	serial_out(info, UART_ESI_CMD2, info->MCR);
-	sti();
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 	return 0;
 }
 
@@ -1785,21 +1805,25 @@
 {
 	struct esp_struct * info = (struct esp_struct *)tty->driver_data;
 	unsigned long flags;
-	
-	if (serial_paranoia_check(info, tty->name, "esp_break"))
+
+	spin_lock_irqsave(&info->irq_lock, flags);
+
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__))) {
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 		return;
+	}
 
-	save_flags(flags); cli();
 	if (break_state == -1) {
 		serial_out(info, UART_ESI_CMD1, ESI_ISSUE_BREAK);
 		serial_out(info, UART_ESI_CMD2, 0x01);
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 
 		interruptible_sleep_on(&info->break_wait);
 	} else {
 		serial_out(info, UART_ESI_CMD1, ESI_ISSUE_BREAK);
 		serial_out(info, UART_ESI_CMD2, 0x00);
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 	}
-	restore_flags(flags);
 }
 
 static int rs_ioctl(struct tty_struct *tty, struct file * file,
@@ -1809,8 +1833,9 @@
 	struct async_icount cprev, cnow;	/* kernel counter temps */
 	struct serial_icounter_struct __user *p_cuser;	/* user space */
 	void __user *argp = (void __user *)arg;
+	unsigned long flags;
 
-	if (serial_paranoia_check(info, tty->name, "rs_ioctl"))
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
 		return -ENODEV;
 
 	if ((cmd != TIOCGSERIAL) && (cmd != TIOCSSERIAL) &&
@@ -1849,17 +1874,17 @@
 		 * Caller should use TIOCGICOUNT to see which one it was
 		 */
 		 case TIOCMIWAIT:
-			cli();
+			spin_lock_irqsave(&info->irq_lock, flags);
 			cprev = info->icount;	/* note the counters on entry */
-			sti();
+			spin_unlock_irqrestore(&info->irq_lock, flags);
 			while (1) {
 				interruptible_sleep_on(&info->delta_msr_wait);
 				/* see if a signal did it */
 				if (signal_pending(current))
 					return -ERESTARTSYS;
-				cli();
+				spin_lock_irqsave(&info->irq_lock, flags);
 				cnow = info->icount;	/* atomic copy */
-				sti();
+				spin_unlock_irqrestore(&info->irq_lock, flags);
 				if (cnow.rng == cprev.rng &&
 				    cnow.dsr == cprev.dsr && 
 				    cnow.dcd == cprev.dcd &&
@@ -1886,9 +1911,9 @@
 		 *     RI where only 0->1 is counted.
 		 */
 		case TIOCGICOUNT:
-			cli();
+			spin_lock_irqsave(&info->irq_lock, flags);
 			cnow = info->icount;
-			sti();
+			spin_unlock_irqrestore(&info->irq_lock, flags);
 			p_cuser = argp;
 			if (put_user(cnow.cts, &p_cuser->cts) ||
 			    put_user(cnow.dsr, &p_cuser->dsr) ||
@@ -1911,6 +1936,7 @@
 static void rs_set_termios(struct tty_struct *tty, struct termios *old_termios)
 {
 	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
+	unsigned long flags;
 
 	if (   (tty->termios->c_cflag == old_termios->c_cflag)
 	    && (   RELEVANT_IFLAG(tty->termios->c_iflag) 
@@ -1922,23 +1948,23 @@
 	/* Handle transition to B0 status */
 	if ((old_termios->c_cflag & CBAUD) &&
 		!(tty->termios->c_cflag & CBAUD)) {
+		spin_lock_irqsave(&info->irq_lock, flags);
 		info->MCR &= ~(UART_MCR_DTR|UART_MCR_RTS);
-		cli();
 		serial_out(info, UART_ESI_CMD1, ESI_WRITE_UART);
 		serial_out(info, UART_ESI_CMD2, UART_MCR);
 		serial_out(info, UART_ESI_CMD2, info->MCR);
-		sti();
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 	}
 
 	/* Handle transition away from B0 status */
 	if (!(old_termios->c_cflag & CBAUD) &&
 		(tty->termios->c_cflag & CBAUD)) {
+		spin_lock_irqsave(&info->irq_lock, flags);
 		info->MCR |= (UART_MCR_DTR | UART_MCR_RTS);
-		cli();
 		serial_out(info, UART_ESI_CMD1, ESI_WRITE_UART);
 		serial_out(info, UART_ESI_CMD2, UART_MCR);
 		serial_out(info, UART_ESI_CMD2, info->MCR);
-		sti();
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 	}
 
 	/* Handle turning of CRTSCTS */
@@ -1975,10 +2001,11 @@
 	struct esp_struct * info = (struct esp_struct *)tty->driver_data;
 	unsigned long flags;
 
-	if (!info || serial_paranoia_check(info, tty->name, "rs_close"))
-		return;
-	
-	save_flags(flags); cli();
+	spin_lock_irqsave(&info->irq_lock, flags);
+
+	if (unlikely(!info ||
+			serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
 	
 	if (tty_hung_up_p(filp)) {
 		DBG_CNT("before DEC-hung");
@@ -2058,7 +2085,7 @@
 	info->flags &= ~(ASYNC_NORMAL_ACTIVE|ASYNC_CLOSING);
 	wake_up_interruptible(&info->close_wait);
 out:
-	restore_flags(flags);
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 }
 
 static void rs_wait_until_sent(struct tty_struct *tty, int timeout)
@@ -2067,8 +2094,9 @@
 	unsigned long orig_jiffies, char_time;
 	unsigned long flags;
 
-	if (serial_paranoia_check(info, tty->name, "rs_wait_until_sent"))
-		return;
+	spin_lock_irqsave(&info->irq_lock, flags);
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
+		goto out;
 
 	orig_jiffies = jiffies;
 	char_time = ((info->timeout - HZ / 50) / 1024) / 5;
@@ -2076,13 +2104,15 @@
 	if (!char_time)
 		char_time = 1;
 
-	save_flags(flags); cli();
 	serial_out(info, UART_ESI_CMD1, ESI_NO_COMMAND);
 	serial_out(info, UART_ESI_CMD1, ESI_GET_TX_AVAIL);
 
 	while ((serial_in(info, UART_ESI_STAT1) != 0x03) ||
 		(serial_in(info, UART_ESI_STAT2) != 0xff)) {
+		
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 		msleep_interruptible(jiffies_to_msecs(char_time));
+		spin_lock_irqsave(&info->irq_lock, flags);
 
 		if (signal_pending(current))
 			break;
@@ -2093,8 +2123,8 @@
 		serial_out(info, UART_ESI_CMD1, ESI_NO_COMMAND);
 		serial_out(info, UART_ESI_CMD1, ESI_GET_TX_AVAIL);
 	}
-	
-	restore_flags(flags);
+
+out:	spin_unlock_irqrestore(&info->irq_lock, flags);
 	set_current_state(TASK_RUNNING);
 }
 
@@ -2105,7 +2135,7 @@
 {
 	struct esp_struct * info = (struct esp_struct *)tty->driver_data;
 	
-	if (serial_paranoia_check(info, tty->name, "esp_hangup"))
+	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
 		return;
 	
 	rs_flush_buffer(tty);
@@ -2174,15 +2204,13 @@
 	printk("block_til_ready before block: ttys%d, count = %d\n",
 	       info->line, info->count);
 #endif
-	save_flags(flags);
-	cli();
+	spin_lock_irqsave(&info->irq_lock, flags);
 	if (!tty_hung_up_p(filp)) 
 		info->count--;
-	restore_flags(flags);
+	spin_unlock_irqrestore(&info->irq_lock, flags);
 	info->blocked_open++;
 	while (1) {
-		save_flags(flags);
-		cli();
+		spin_lock_irqsave(&info->irq_lock, flags);
 		if ((tty->termios->c_cflag & CBAUD)) {
 			unsigned int scratch;
 
@@ -2194,7 +2222,7 @@
 			serial_out(info, UART_ESI_CMD2,
 				scratch | UART_MCR_DTR | UART_MCR_RTS);
 		}
-		restore_flags(flags);
+		spin_unlock_irqrestore(&info->irq_lock, flags);
 		set_current_state(TASK_INTERRUPTIBLE);
 		if (tty_hung_up_p(filp) ||
 		    !(info->flags & ASYNC_INITIALIZED)) {
@@ -2264,7 +2292,7 @@
 		info = info->next_port;
 
 	if (!info) {
-		serial_paranoia_check(info, tty->name, "esp_open");
+		serial_paranoia_check(info, tty->name, __FUNCTION__);
 		return -ENODEV;
 	}
 
@@ -2275,12 +2303,6 @@
 	tty->driver_data = info;
 	info->tty = tty;
 
-	if (!tmp_buf) {
-		tmp_buf = (unsigned char *) get_zeroed_page(GFP_KERNEL);
-		if (!tmp_buf)
-			return -ENOMEM;
-	}
-	
 	/*
 	 * Start up serial port
 	 */
@@ -2335,7 +2357,7 @@
 	if (!request_region(info->port, REGION_SIZE, "esp serial"))
 		return -EIO;
 
-	save_flags(flags); cli();
+	spin_lock_irqsave(&info->irq_lock, flags);
 	
 	/*
 	 * Check for ESP card
@@ -2369,10 +2391,11 @@
 			serial_out(info, UART_ESI_CMD2, 0x00);
 		}
 	}
+	spin_unlock_irqrestore(&info->irq_lock, flags);
+
 	if (!port_detected)
 		release_region(info->port, REGION_SIZE);
 
-	restore_flags(flags);
 	return (port_detected);
 }
 
@@ -2513,6 +2536,7 @@
 		init_waitqueue_head(&info->close_wait);
 		init_waitqueue_head(&info->delta_msr_wait);
 		init_waitqueue_head(&info->break_wait);
+		spin_lock_init(&info->irq_lock);
 		ports = info;
 		printk(KERN_INFO "ttyP%d at 0x%04x (irq = %d) is an ESP ",
 			info->line, info->port, info->irq);
@@ -2563,18 +2587,14 @@
 
 static void __exit espserial_exit(void) 
 {
-	unsigned long flags;
 	int e1;
 	struct esp_struct *temp_async;
 	struct esp_pio_buffer *pio_buf;
 
 	/* printk("Unloading %s: version %s\n", serial_name, serial_version); */
-	save_flags(flags);
-	cli();
 	if ((e1 = tty_unregister_driver(esp_driver)))
-		printk("SERIAL: failed to unregister serial driver (%d)\n",
-		       e1);
-	restore_flags(flags);
+		printk(KERN_ERR "SERIAL: failed to unregister serial driver"
+				" (%d)\n", e1);
 	put_tty_driver(esp_driver);
 
 	while (ports) {
@@ -2590,9 +2610,6 @@
 		free_pages((unsigned long)dma_buffer,
 			get_order(DMA_BUFFER_SZ));
 
-	if (tmp_buf)
-		free_page((unsigned long)tmp_buf);
-
 	while (free_pio_buf) {
 		pio_buf = free_pio_buf->next;
 		kfree(free_pio_buf);
diff -urN --exclude='*~' linux-2.6.10-original/drivers/char/Kconfig linux-2.6.10/drivers/char/Kconfig
--- linux-2.6.10-original/drivers/char/Kconfig	2004-12-24 16:33:49.000000000 -0500
+++ linux-2.6.10/drivers/char/Kconfig	2004-12-30 20:25:12.945068082 -0500
@@ -170,7 +170,7 @@
 
 config ESPSERIAL
 	tristate "Hayes ESP serial port support"
-	depends on SERIAL_NONSTANDARD && ISA && BROKEN_ON_SMP
+	depends on SERIAL_NONSTANDARD && ISA
 	help
 	  This is a driver which supports Hayes ESP serial ports.  Both single
 	  port cards and multiport cards are supported.  Make sure to read
diff -urN --exclude='*~' linux-2.6.10-original/include/linux/hayesesp.h linux-2.6.10/include/linux/hayesesp.h
--- linux-2.6.10-original/include/linux/hayesesp.h	2004-12-24 16:35:24.000000000 -0500
+++ linux-2.6.10/include/linux/hayesesp.h	2004-12-30 18:15:06.541960410 -0500
@@ -108,6 +108,7 @@
 	wait_queue_head_t	break_wait;
 	struct async_icount	icount;	/* kernel counters for the 4 input interrupts */
 	struct hayes_esp_config config; /* port configuration */
+	spinlock_t		irq_lock;
 	struct esp_struct	*next_port; /* For the linked list */
 };
 

Re: [PATCH] esp: Make driver SMP-correct

[Jan Dittmer]

James Nelson wrote:
>  static inline int serial_paranoia_check(struct esp_struct *info,
>  					char *name, const char *routine)
>  {
> @@ -209,34 +197,38 @@
>  	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
>  	unsigned long flags;
>  
> -	if (serial_paranoia_check(info, tty->name, "rs_stop"))
> -		return;
> +	spin_lock_irqsave(&info->irq_lock, flags);
> +
> +	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
> +		goto out;

>  static void rs_start(struct tty_struct *tty)
>  {
>  	struct esp_struct *info = (struct esp_struct *)tty->driver_data;
>  	unsigned long flags;
> -	
> -	if (serial_paranoia_check(info, tty->name, "rs_start"))
> -		return;
> -	
> -	save_flags(flags); cli();
> +
> +	spin_lock_irqsave(&info->irq_lock, flags);
> +
> +	if (unlikely(serial_paranoia_check(info, tty->name, __FUNCTION__)))
> +		goto out;
> +

This will deadlock. serial_paranoia_check also unconditionally tries to
take the lock.

Jan

Re: [PATCH] esp: Make driver SMP-correct

[Andrew Morton]

James Nelson <james4765@verizon.net> wrote:
>
> This is an attempt to make the esp serial driver SMP-correct.  It also removes
>  some cruft left over from the serial_write() conversion.

>From a quick scan:

- startup() does multiple sleeping allocations and request_irq() under
  spin_lock_irqsave().  Maybe fixed by this:

--- 25/drivers/char/esp.c~esp-make-driver-smp-correct-fixes	2004-12-31 01:40:57.987232152 -0800
+++ 25-akpm/drivers/char/esp.c	2004-12-31 01:42:21.444544712 -0800
@@ -851,16 +851,14 @@ static int startup(struct esp_struct * i
 	int	retval=0;
         unsigned int num_chars;
 
-	spin_lock_irqsave(&info->irq_lock, flags);
-
 	if (info->flags & ASYNC_INITIALIZED)
-		goto out;
+		goto out_unlocked;
 
 	if (!info->xmit_buf) {
 		info->xmit_buf = (unsigned char *)get_zeroed_page(GFP_KERNEL);
 		retval = -ENOMEM;
 		if (!info->xmit_buf)
-			goto out;
+			goto out_unlocked;
 	}
 
 #ifdef SERIAL_DEBUG_OPEN
@@ -907,7 +905,7 @@ static int startup(struct esp_struct * i
 					&info->tty->flags);
 			retval = 0;
 		}
-		goto out;
+		goto out_unocked;
 	}
 
 	if (!(info->stat_flags & ESP_STAT_USE_PIO) && !dma_buffer) {
@@ -926,6 +924,8 @@ static int startup(struct esp_struct * i
 			
 	}
 
+	spin_lock_irqsave(&info->irq_lock, flags);
+
 	info->MCR = UART_MCR_DTR | UART_MCR_RTS | UART_MCR_OUT2;
 	serial_out(info, UART_ESI_CMD1, ESI_WRITE_UART);
 	serial_out(info, UART_ESI_CMD2, UART_MCR);
@@ -965,7 +965,9 @@ static int startup(struct esp_struct * i
 
 	info->flags |= ASYNC_INITIALIZED;
 	retval = 0;
-out:	spin_unlock_irqrestore(&info->irq_lock, flags);
+out:
+	spin_unlock_irqrestore(&info->irq_lock, flags);
+out_unocked:
 	return retval;
 }
 
_


- startup() calls change_speed() under info->irq_lock, but change_speed()
  also takes info->irq_lock.  Instant deadlock on driver initialisation.

The driver needs more serious surgery than this.

Re: [PATCH] esp: Make driver SMP-correct

[Russell King]

On Fri, Dec 31, 2004 at 01:46:11AM -0800, Andrew Morton wrote:
> James Nelson <james4765@verizon.net> wrote:
> >
> > This is an attempt to make the esp serial driver SMP-correct.  It also removes
> >  some cruft left over from the serial_write() conversion.
> 
> >From a quick scan:
> 
> - startup() does multiple sleeping allocations and request_irq() under
>   spin_lock_irqsave().  Maybe fixed by this:

However, can you guarantee that two threads won't enter startup() at
the same time?  (that's what ASYNC_INITIALIZED is protecting the
function against, and the corresponding shutdown() as well.)

It's probably better to port ESP to the serial_core structure where
this type of thing is already taken care of.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core

Re: [PATCH] esp: Make driver SMP-correct

[Jim Nelson]

Russell King wrote:
> On Fri, Dec 31, 2004 at 01:46:11AM -0800, Andrew Morton wrote:
> 
>>James Nelson <james4765@verizon.net> wrote:
>>
>>>This is an attempt to make the esp serial driver SMP-correct.  It also removes
>>> some cruft left over from the serial_write() conversion.
>>
>>>From a quick scan:
>>
>>- startup() does multiple sleeping allocations and request_irq() under
>>  spin_lock_irqsave().  Maybe fixed by this:
> 
> 
> However, can you guarantee that two threads won't enter startup() at
> the same time?  (that's what ASYNC_INITIALIZED is protecting the
> function against, and the corresponding shutdown() as well.)
> 
> It's probably better to port ESP to the serial_core structure where
> this type of thing is already taken care of.
> 

You are right.  Quite a bit of work, if I am looking at it correctly, though.  Any 
pointers/URLs to help a newbie get their head wrapped around the process?

Re: [PATCH] esp: Make driver SMP-correct

[Russell King]

On Fri, Dec 31, 2004 at 10:00:37AM +0000, Russell King wrote:
> On Fri, Dec 31, 2004 at 01:46:11AM -0800, Andrew Morton wrote:
> > James Nelson <james4765@verizon.net> wrote:
> > >
> > > This is an attempt to make the esp serial driver SMP-correct.  It also removes
> > >  some cruft left over from the serial_write() conversion.
> > 
> > >From a quick scan:
> > 
> > - startup() does multiple sleeping allocations and request_irq() under
> >   spin_lock_irqsave().  Maybe fixed by this:
> 
> However, can you guarantee that two threads won't enter startup() at
> the same time?  (that's what ASYNC_INITIALIZED is protecting the
> function against, and the corresponding shutdown() as well.)
> 
> It's probably better to port ESP to the serial_core structure where
> this type of thing is already taken care of.

For the record, Verizon appear to have adopted silly policies.

>From now on, I will be removing the CC: line containing any verizon
email address until further notice, or just plain ignoring mails
containing such addresses.  Why?  To prevent the inevitable bounce
caused by their misconfigured systems.  None of the servers I have
access to on several different ISPs can connect to Verizon's incoming
mail server.

See:
 http://www.broadbandreports.com/forum/remark,12116645~mode=flat~days=9999

particularly the last post by techie68, who claims to be a Verizon
tech support person.

I encourage James Nelson to find another provider without silly policies
in the mean time.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Jim Nelson]

Russell King wrote:
> On Fri, Dec 31, 2004 at 10:00:37AM +0000, Russell King wrote:
> 
>>On Fri, Dec 31, 2004 at 01:46:11AM -0800, Andrew Morton wrote:
>>
>>>James Nelson <james4765@verizon.net> wrote:
>>>
>>>>This is an attempt to make the esp serial driver SMP-correct.  It also removes
>>>> some cruft left over from the serial_write() conversion.
>>>
>>>>From a quick scan:
>>>
>>>- startup() does multiple sleeping allocations and request_irq() under
>>>  spin_lock_irqsave().  Maybe fixed by this:
>>
>>However, can you guarantee that two threads won't enter startup() at
>>the same time?  (that's what ASYNC_INITIALIZED is protecting the
>>function against, and the corresponding shutdown() as well.)
>>
>>It's probably better to port ESP to the serial_core structure where
>>this type of thing is already taken care of.
> 
> 
> For the record, Verizon appear to have adopted silly policies.
> 
>>From now on, I will be removing the CC: line containing any verizon
> email address until further notice, or just plain ignoring mails
> containing such addresses.  Why?  To prevent the inevitable bounce
> caused by their misconfigured systems.  None of the servers I have
> access to on several different ISPs can connect to Verizon's incoming
> mail server.
> 
> See:
>  http://www.broadbandreports.com/forum/remark,12116645~mode=flat~days=9999
> 
> particularly the last post by techie68, who claims to be a Verizon
> tech support person.
> 
> I encourage James Nelson to find another provider without silly policies
> in the mean time.
> 
> 

Another mail provider has been found.  Bloody Verizon.

Documentation/serial/driver mentions a reference implementation - serial_amba.c. 
Couldn't find it in the 2.6.10 kernel sources - was it removed, and what would be 
a good refernce implementation in the current sources?

Jim

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Gene Heskett]

On Friday 31 December 2004 17:16, Jim Nelson wrote:
>Russell King wrote:
>> On Fri, Dec 31, 2004 at 10:00:37AM +0000, Russell King wrote:
>>>On Fri, Dec 31, 2004 at 01:46:11AM -0800, Andrew Morton wrote:
>>>>James Nelson <james4765@verizon.net> wrote:
>>>>>This is an attempt to make the esp serial driver SMP-correct. 
>>>>> It also removes some cruft left over from the serial_write()
>>>>> conversion.
>>>>>
>>>>>>From a quick scan:
>>>>
>>>>- startup() does multiple sleeping allocations and request_irq()
>>>> under spin_lock_irqsave().  Maybe fixed by this:
>>>
>>>However, can you guarantee that two threads won't enter startup()
>>> at the same time?  (that's what ASYNC_INITIALIZED is protecting
>>> the function against, and the corresponding shutdown() as well.)
>>>
>>>It's probably better to port ESP to the serial_core structure
>>> where this type of thing is already taken care of.
>>
>> For the record, Verizon appear to have adopted silly policies.
>>
>>>>From now on, I will be removing the CC: line containing any
>>> verizon
>>
>> email address until further notice, or just plain ignoring mails
>> containing such addresses.  Why?  To prevent the inevitable bounce
>> caused by their misconfigured systems.  None of the servers I have
>> access to on several different ISPs can connect to Verizon's
>> incoming mail server.
>>
>> See:
>> 
>> http://www.broadbandreports.com/forum/remark,12116645~mode=flat~da
>>ys=9999
>>
>> particularly the last post by techie68, who claims to be a Verizon
>> tech support person.
>>
>> I encourage James Nelson to find another provider without silly
>> policies in the mean time.
>
>Another mail provider has been found.  Bloody Verizon.
>
This strikes me as being some damn petty politics on verizons part.  
I'm a customer because its the only game in town with affordable 
broadband, by at least 30 dollars a month.  So I'm stuck with it, or 
a friggin dialup that is also about 30 a month more expensive.

And it seems that everytime I use the "reply all" button on this list, 
I'll get at least 2 bounces because somebody elses server has 
blacklisted verizon for running an open relay 2 years ago, & never 
got around to lifting the ban when verizon rebuilt the systems last 
spring.

One should be carefull as that shoe can fit many feet. The CC: line of 
this message:

Jim Nelson <james4765@cwazy.co.uk>, Russell King 
<rmk+lkml@arm.linux.org.uk>, Andrew Morton <akpm@osdl.org>, 
kernel-janitors@lists.osdl.org

I'm curious as to who I will get a bounce from.

>Documentation/serial/driver mentions a reference implementation -
> serial_amba.c. Couldn't find it in the 2.6.10 kernel sources - was
> it removed, and what would be a good refernce implementation in the
> current sources?
>
>Jim
>-
>To unsubscribe from this list: send the line "unsubscribe
> linux-kernel" in the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/

-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.31% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com attorneys please note, additions to this message
by Gene Heskett are:
Copyright 2004 by Maurice Eugene Heskett, all rights reserved.

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Russell King]

On Fri, Dec 31, 2004 at 07:01:50PM -0500, Gene Heskett wrote:
> One should be carefull as that shoe can fit many feet. The CC: line of 
> this message:
> 
> Jim Nelson <james4765@cwazy.co.uk>, Russell King 
> <rmk+lkml@arm.linux.org.uk>, Andrew Morton <akpm@osdl.org>, 
> kernel-janitors@lists.osdl.org
> 
> I'm curious as to who I will get a bounce from.

Well, you'll get one from me, but not because _I've_ blocked verizon (I've
done no such thing.)  Since I'm unable to verify verizon's addresses, all
mail from verizon is unable to be received here anymore.

(and yes, you get to read my reply via lkml...)

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Gene Heskett]

On Friday 31 December 2004 19:13, Russell King wrote:
>To: no To-header on input <>

Thats the To: line of this message Russell, as it came in here.  I 
assume it was originally filled in to be to me and that you cleaned 
that to prevent your getting a bounce from verizon?

And no, it hasn't bounced yet as it typically will lay in the queue 
somewhere in lala land for anywhere from 4 hours to 6 or 7 days.  By 
that time the friggin message is no longer germain to the 
conversation, so they get deleted here.

-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.31% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com attorneys please note, additions to this message
by Gene Heskett are:
Copyright 2004 by Maurice Eugene Heskett, all rights reserved.

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Russell King]

On Fri, Dec 31, 2004 at 08:14:39PM -0500, Gene Heskett wrote:
> On Friday 31 December 2004 19:13, Russell King wrote:
> >To: no To-header on input <>
> 
> Thats the To: line of this message Russell, as it came in here.  I 
> assume it was originally filled in to be to me and that you cleaned 
> that to prevent your getting a bounce from verizon?

Indeed, and I now have a better solution which I don't even have to
think about anymore, and doesn't damage the headers in the message. 8)

Let me know if and when verizon comes back to reality please.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Gene Heskett]

On Friday 31 December 2004 20:14, Gene Heskett wrote:
>On Friday 31 December 2004 19:13, Russell King wrote:
>>To: no To-header on input <>
>
>Thats the To: line of this message Russell, as it came in here.  I
>assume it was originally filled in to be to me and that you cleaned
>that to prevent your getting a bounce from verizon?
>
>And no, it hasn't bounced yet as it typically will lay in the queue
>somewhere in lala land for anywhere from 4 hours to 6 or 7 days.  By
>that time the friggin message is no longer germain to the
>conversation, so they get deleted here.

Now, this message did bounce, and the bounce message is damned 
confusing...

From: Mail Administrator <Postmaster@verizon.net>
 To: gene.heskett@verizon.net
 
This Message was undeliverable due to the following reason:

Your message was not delivered because the destination computer was
not found.  Carefully check that it was spelled correctly and try
sending it again if there were any mistakes.

It is also possible that a network problem caused this situation,
so if you are sure the address is correct you might want to try to
send it again.  If the problem continues, contact your friendly
system administrator.

     Host coyote.coyote.den not found

The following recipients did not receive this message:

     <""@coyote.coyote.den>

Please reply to Postmaster@verizon.net
if you feel this message to be in error.
--------------------------
coyote.coyote.den is indeed the name of this machine, but I should be 
known to the outside world as gene.heskett AT verizon.net.

So the $64,000 question is how did that domain name even get to the 
outside world.

Here is the complete header from the message that elicited that 
response from verizons servers.
From: Gene Heskett <gene.heskett@verizon.net>
 Reply-To: gene.heskett@verizon.net
 Organization: Organization: None, detectable by casual observers
 To: linux-kernel@vger.kernel.org
 Subject: Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct
 Date: Fri, 31 Dec 2004 20:14:39 -0500
 User-Agent: KMail/1.7
 Cc: Russell King <rmk+lkml@arm.linux.org.uk>,
 Jim Nelson <james4765@cwazy.co.uk>,
 Andrew Morton <akpm@osdl.org>,
 kernel-janitors@lists.osdl.org
 References: <20041231014403.3309.58245.96163@localhost.localdomain> 
<200412311901.50638.gene.heskett@verizon.net> 
<20050101001311.D10216@flint.arm.linux.org.uk>
 In-Reply-To: <20050101001311.D10216@flint.arm.linux.org.uk>
 X-KMail-Link-Message: 676118
 X-KMail-Link-Type: reply
 MIME-Version: 1.0
 Content-Type: text/plain;
  charset="us-ascii"
 Content-Transfer-Encoding: 7bit
 Content-Disposition: inline
 Message-Id: <200412312014.39618.gene.heskett@verizon.net>
 Status: RO
 X-Status: RSC
 X-KMail-EncryptionState: 
 X-KMail-SignatureState: 
 X-KMail-MDN-Sent: 
 
On Friday 31 December 2004 19:13, Russell King wrote:
>To: no To-header on input <>

And there sure as heck isn't any mention of 'coyote.coyote.den' in 
that.

So whats the deal verizon???  But asking for an answer is like talking 
to a stone wall, verizon has been congenitally (or surgically) 
rendered mute by their legal dept.  I've yet to get a reply to any of 
the several questions I've sent to the Postmaster@verizon.net black 
hole.

They have a private newsgroup I used to inhabit, but the snr was so 
horrible I gave it up, my remaining lifetime is too short to spend it 
argueing east bergen nj politics.

-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.31% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com attorneys please note, additions to this message
by Gene Heskett are:
Copyright 2004 by Maurice Eugene Heskett, all rights reserved.

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Jim Nelson]

Gene Heskett wrote:
> On Friday 31 December 2004 20:14, Gene Heskett wrote:
> 
>>On Friday 31 December 2004 19:13, Russell King wrote:
>>
>>>To: no To-header on input <>
>>
>>Thats the To: line of this message Russell, as it came in here.  I
>>assume it was originally filled in to be to me and that you cleaned
>>that to prevent your getting a bounce from verizon?
>>
>>And no, it hasn't bounced yet as it typically will lay in the queue
>>somewhere in lala land for anywhere from 4 hours to 6 or 7 days.  By
>>that time the friggin message is no longer germain to the
>>conversation, so they get deleted here.
> 
> 
> Now, this message did bounce, and the bounce message is damned 
> confusing...
> 
> From: Mail Administrator <Postmaster@verizon.net>
>  To: gene.heskett@verizon.net
>  
> This Message was undeliverable due to the following reason:
> 
> Your message was not delivered because the destination computer was
> not found.  Carefully check that it was spelled correctly and try
> sending it again if there were any mistakes.
> 
> It is also possible that a network problem caused this situation,
> so if you are sure the address is correct you might want to try to
> send it again.  If the problem continues, contact your friendly
> system administrator.
> 
>      Host coyote.coyote.den not found
> 
> The following recipients did not receive this message:
> 
>      <""@coyote.coyote.den>
> 
> Please reply to Postmaster@verizon.net
> if you feel this message to be in error.
> --------------------------
> coyote.coyote.den is indeed the name of this machine, but I should be 
> known to the outside world as gene.heskett AT verizon.net.
> 
> So the $64,000 question is how did that domain name even get to the 
> outside world.
> 
> Here is the complete header from the message that elicited that 
> response from verizons servers.
> From: Gene Heskett <gene.heskett@verizon.net>
>  Reply-To: gene.heskett@verizon.net
>  Organization: Organization: None, detectable by casual observers
>  To: linux-kernel@vger.kernel.org
>  Subject: Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct
>  Date: Fri, 31 Dec 2004 20:14:39 -0500
>  User-Agent: KMail/1.7
>  Cc: Russell King <rmk+lkml@arm.linux.org.uk>,
>  Jim Nelson <james4765@cwazy.co.uk>,
>  Andrew Morton <akpm@osdl.org>,
>  kernel-janitors@lists.osdl.org
>  References: <20041231014403.3309.58245.96163@localhost.localdomain> 
> <200412311901.50638.gene.heskett@verizon.net> 
> <20050101001311.D10216@flint.arm.linux.org.uk>
>  In-Reply-To: <20050101001311.D10216@flint.arm.linux.org.uk>
>  X-KMail-Link-Message: 676118
>  X-KMail-Link-Type: reply
>  MIME-Version: 1.0
>  Content-Type: text/plain;
>   charset="us-ascii"
>  Content-Transfer-Encoding: 7bit
>  Content-Disposition: inline
>  Message-Id: <200412312014.39618.gene.heskett@verizon.net>
>  Status: RO
>  X-Status: RSC
>  X-KMail-EncryptionState: 
>  X-KMail-SignatureState: 
>  X-KMail-MDN-Sent: 
>  
> On Friday 31 December 2004 19:13, Russell King wrote:
> 
>>To: no To-header on input <>
> 
> 
> And there sure as heck isn't any mention of 'coyote.coyote.den' in 
> that.
> 

 From my perspective:
 From - Fri Dec 31 20:42:48 2004
X-UIDL: <59229-2004-1231-193921-22204@st108mss.verizon.net>
X-Mozilla-Status: 0011
X-Mozilla-Status2: 00000000
Return-Path: 
<linux-kernel-owner+james4765=40verizon.net-S262173AbVAABfz@vger.kernel.org>
Received: from vger.kernel.org ([206.46.170.120]) by mta008.verizon.net
           (InterMail vM.5.01.06.06 201-253-122-130-106-20030910) with ESMTP
           id <20050101013921.PWOD28241.mta008.verizon.net@vger.kernel.org>
           for <james4765@verizon.net>; Fri, 31 Dec 2004 19:39:21 -0600
Received: from vger.kernel.org (12.107.209.244) by sc009pub.verizon.net (MailPass 
SMTP server v1.1.1 - 121803235448JY) with  ESMTP id 
<1-995-136-995-258990-6-1104543560> for mta008.verizon.net; Fri, 31 Dec 2004 
19:39:21 -0600
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S262173AbVAABfz (ORCPT <rfc822;james4765@verizon.net>);
	Fri, 31 Dec 2004 20:35:55 -0500
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262174AbVAABfy
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Fri, 31 Dec 2004 20:35:54 -0500
Received: from out003pub.verizon.net ([206.46.170.103]:39854 "EHLO
	out003.verizon.net") by vger.kernel.org with ESMTP id S262173AbVAABfE
	(ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Fri, 31 Dec 2004 20:35:04 -0500
Received: from coyote.coyote.den ([151.205.52.185]) by out003.verizon.net
           (InterMail vM.5.01.06.06 201-253-122-130-106-20030910) with ESMTP
           id <20050101013503.LVAF1106.out003.verizon.net@coyote.coyote.den>;
           Fri, 31 Dec 2004 19:35:03 -0600
From:	Gene Heskett <gene.heskett@verizon.net>
Reply-To: gene.heskett@verizon.net
Organization: Organization: None, detectable by casual observers
To:	linux-kernel@vger.kernel.org, Postmaster@verizon.net
Subject: Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct
Date:	Fri, 31 Dec 2004 20:35:02 -0500
User-Agent: KMail/1.7
Cc:	Russell King <rmk+lkml@arm.linux.org.uk>,
	Jim Nelson <james4765@cwazy.co.uk>,
	Andrew Morton <akpm@osdl.org>, kernel-janitors@lists.osdl.org
References: <20041231014403.3309.58245.96163@localhost.localdomain> 
<20050101001311.D10216@flint.arm.linux.org.uk> 
<200412312014.39618.gene.heskett@verizon.net>
In-Reply-To: <200412312014.39618.gene.heskett@verizon.net>
MIME-Version: 1.0
Content-Type: text/plain;
   charset="us-ascii"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200412312035.02761.gene.heskett@verizon.net>
X-Authentication-Info: Submitted using SMTP AUTH at out003.verizon.net from 
[151.205.52.185] at Fri, 31 Dec 2004 19:35:03 -0600
Sender:	linux-kernel-owner@vger.kernel.org
Precedence: bulk
X-Mailing-List:	linux-kernel@vger.kernel.org

Are you running your own MTA?  That would explain a lot - Verizon's mailservers 
are touchy at best.

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Jim Nelson]

Russell King wrote:
> On Fri, Dec 31, 2004 at 08:14:39PM -0500, Gene Heskett wrote:
> 
>>On Friday 31 December 2004 19:13, Russell King wrote:
>>
>>>To: no To-header on input <>
>>
>>Thats the To: line of this message Russell, as it came in here.  I 
>>assume it was originally filled in to be to me and that you cleaned 
>>that to prevent your getting a bounce from verizon?
> 
> 
> Indeed, and I now have a better solution which I don't even have to
> think about anymore, and doesn't damage the headers in the message. 8)
> 
> Let me know if and when verizon comes back to reality please.
> 

Not anytime soon, methinks.  Verizon was one of the first ISP's to block Gnutella 
on their networks, and are unresponsive to any appeals to unblock it.  I've sent a 
nastygram to their security "experts", and am actively seeking another ISP - I 
think there are 1 or 2 other DSL providers in my area.

It's a shame, really - their field techs are top-notch and I've had no 
connectivity issues in 3 years, but the PHBs who set policy are as bad as my 
bosses at Greyhound - which is a major insult, in my book.

In the meantime, I'll just be a Virginia boy with a .co.uk e-mail address :)

Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct

[Gene Heskett]

On Friday 31 December 2004 20:47, Jim Nelson wrote:
>Gene Heskett wrote:
>> On Friday 31 December 2004 20:14, Gene Heskett wrote:
>>>On Friday 31 December 2004 19:13, Russell King wrote:
>>>>To: no To-header on input <>
>>>
>>>Thats the To: line of this message Russell, as it came in here.  I
>>>assume it was originally filled in to be to me and that you
>>> cleaned that to prevent your getting a bounce from verizon?
>>>
>>>And no, it hasn't bounced yet as it typically will lay in the
>>> queue somewhere in lala land for anywhere from 4 hours to 6 or 7
>>> days.  By that time the friggin message is no longer germain to
>>> the conversation, so they get deleted here.
>>
>> Now, this message did bounce, and the bounce message is damned
>> confusing...
>>
>> From: Mail Administrator <Postmaster@verizon.net>
>>  To: gene.heskett@verizon.net
>>
>> This Message was undeliverable due to the following reason:
>>
>> Your message was not delivered because the destination computer
>> was not found.  Carefully check that it was spelled correctly and
>> try sending it again if there were any mistakes.
>>
>> It is also possible that a network problem caused this situation,
>> so if you are sure the address is correct you might want to try to
>> send it again.  If the problem continues, contact your friendly
>> system administrator.
>>
>>      Host coyote.coyote.den not found
>>
>> The following recipients did not receive this message:
>>
>>      <""@coyote.coyote.den>
>>
>> Please reply to Postmaster@verizon.net
>> if you feel this message to be in error.
>> --------------------------
>> coyote.coyote.den is indeed the name of this machine, but I should
>> be known to the outside world as gene.heskett AT verizon.net.
>>
>> So the $64,000 question is how did that domain name even get to
>> the outside world.
>>
>> Here is the complete header from the message that elicited that
>> response from verizons servers.
>> From: Gene Heskett <gene.heskett@verizon.net>
>>  Reply-To: gene.heskett@verizon.net
>>  Organization: Organization: None, detectable by casual observers
>>  To: linux-kernel@vger.kernel.org
>>  Subject: Re: [KJ] Re: [PATCH] esp: Make driver SMP-correct
>>  Date: Fri, 31 Dec 2004 20:14:39 -0500
>>  User-Agent: KMail/1.7
>>  Cc: Russell King <rmk+lkml@arm.linux.org.uk>,
>>  Jim Nelson <james4765@cwazy.co.uk>,
>>  Andrew Morton <akpm@osdl.org>,
>>  kernel-janitors@lists.osdl.org
>>  References:
>> <20041231014403.3309.58245.96163@localhost.localdomain>
>> <200412311901.50638.gene.heskett@verizon.net>
>> <20050101001311.D10216@flint.arm.linux.org.uk>
>>  In-Reply-To: <20050101001311.D10216@flint.arm.linux.org.uk>
>>  X-KMail-Link-Message: 676118
>>  X-KMail-Link-Type: reply
>>  MIME-Version: 1.0
>>  Content-Type: text/plain;
>>   charset="us-ascii"
>>  Content-Transfer-Encoding: 7bit
>>  Content-Disposition: inline
>>  Message-Id: <200412312014.39618.gene.heskett@verizon.net>
>>  Status: RO
>>  X-Status: RSC
>>  X-KMail-EncryptionState:
>>  X-KMail-SignatureState:
>>  X-KMail-MDN-Sent:
>>
>> On Friday 31 December 2004 19:13, Russell King wrote:
>>>To: no To-header on input <>
>>
>> And there sure as heck isn't any mention of 'coyote.coyote.den' in
>> that.
>
> From my perspective:
[...]
------------------------
>Received: from coyote.coyote.den ([151.205.52.185]) by
> out003.verizon.net (InterMail vM.5.01.06.06
> 201-253-122-130-106-20030910) with ESMTP id
> <20050101013503.LVAF1106.out003.verizon.net@coyote.coyote.den>;
--------
I still wonder.  And no, AFAIK I am not running a mail server, just 
kmail from a kde 3.3.0 build by konstruct.

The actual path from coyote.coyote.den at 192.168.xx.xx goes thru an 8 
port netgear switch, to my firewall, gene.coyote.den via the eth1 
interface in that machine, gets noted by iptables, and forwarded to 
router.coyote.den in a different subnet of the 192.168.xx.xx block.  
router.coyote.den is a linksys BEFSR41 set for gateway mode, which 
sends it thru its NAT and PPPoE to the wan port and the westel dsl 
modem.  So the quoted address in the 151.205.xx.xx block above 
doesn't exist on my side of the BEFSR41.  Kmails sending is 
configured for smtp, with the output address & password direct to 
incoming.verizon.net.

If somehow, the smtp is adding my local hostname so that my regular 
address is munged, how can I fix that?  Or is it my problem since the 
header of the outgoing message copy in kmails sent mail folder 
contains no such data.  Is that not a verbatum copy of whats sent.  
If not, then it sounds like its time to file a bug against kmail.

[...]

>Are you running your own MTA?  That would explain a lot - Verizon's
> mailservers are touchy at best.

See above description and tell me, I am apparently a big dummy.

-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.31% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com attorneys please note, additions to this message
by Gene Heskett are:
Copyright 2004 by Maurice Eugene Heskett, all rights reserved.