linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Peter Hurley <peter@hurleysoftware.com>
To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: linux1394-devel@lists.sourceforge.net,
	linux-kernel@vger.kernel.org,
	Stefan Richter <stefanr@s5r6.in-berlin.de>,
	Peter Hurley <peter@hurleysoftware.com>
Subject: [PATCH 10/11] staging/fwserial: Fix premature unthrottle
Date: Mon, 28 Jan 2013 22:34:44 -0500	[thread overview]
Message-ID: <1359430485-5301-10-git-send-email-peter@hurleysoftware.com> (raw)
In-Reply-To: <1359430485-5301-1-git-send-email-peter@hurleysoftware.com>

The unthrottle may restart the writer before the rx push
has a chance to start emptying the rx buffer, resulting in an
overflowed rx buffer and lost data.

Perform the actual device unthrottle with the rx push instead.

Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
 drivers/staging/fwserial/fwserial.c | 30 ++++++++++++++----------------
 1 file changed, 14 insertions(+), 16 deletions(-)

diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index 912ab77..121beff 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -463,6 +463,14 @@ static void __fwtty_throttle(struct fwtty_port *port, struct tty_struct *tty)
 		__fwtty_write_port_status(port);
 }
 
+static void __fwtty_unthrottle(struct fwtty_port *port, struct tty_struct *tty)
+{
+	port->mctrl &= ~OOB_RX_THROTTLE;
+	if (C_CRTSCTS(tty))
+		port->mctrl |= TIOCM_RTS;
+	__fwtty_write_port_status(port);
+}
+
 /**
  * fwtty_do_hangup - wait for ldisc to deliver all pending rx; only then hangup
  *
@@ -527,14 +535,15 @@ static void fwtty_emit_breaks(struct work_struct *work)
 	port->icount.brk += brk;
 }
 
-static void fwtty_pushrx(struct work_struct *work)
+static void fwtty_pushrx(struct fwtty_port *port, struct tty_struct *tty)
 {
-	struct fwtty_port *port = to_port(work, push);
-	struct tty_struct *tty;
 	struct buffered_rx *buf, *next;
 	int n, c = 0;
 
 	spin_lock_bh(&port->lock);
+
+	__fwtty_unthrottle(port, tty);
+
 	list_for_each_entry_safe(buf, next, &port->buf_list, list) {
 		n = tty_insert_flip_string_fixed_flag(&port->port, buf->data,
 						      TTY_NORMAL, buf->n);
@@ -545,11 +554,7 @@ static void fwtty_pushrx(struct work_struct *work)
 				memmove(buf->data, buf->data + n, buf->n - n);
 				buf->n -= n;
 			}
-			tty = tty_port_tty_get(&port->port);
-			if (tty) {
-				__fwtty_throttle(port, tty);
-				tty_kref_put(tty);
-			}
+			__fwtty_throttle(port, tty);
 			break;
 		} else {
 			list_del(&buf->list);
@@ -1264,14 +1269,7 @@ static void fwtty_unthrottle(struct tty_struct *tty)
 
 	profile_fifo_avail(port, port->stats.unthrottle);
 
-	schedule_work(&port->push);
-
-	spin_lock_bh(&port->lock);
-	port->mctrl &= ~OOB_RX_THROTTLE;
-	if (C_CRTSCTS(tty))
-		port->mctrl |= TIOCM_RTS;
-	__fwtty_write_port_status(port);
-	spin_unlock_bh(&port->lock);
+	fwtty_pushrx(port, tty);
 }
 
 static int check_msr_delta(struct fwtty_port *port, unsigned long mask,
-- 
1.8.1.1


  parent reply	other threads:[~2013-01-29  3:36 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-01-29  3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
2013-01-29  3:34 ` [PATCH 02/11] staging/fwserial: Release port regardless of unplug response code Peter Hurley
2013-01-29  3:34 ` [PATCH 03/11] staging/fwserial: Fix sparse build warnings Peter Hurley
2013-01-29  3:34 ` [PATCH 04/11] staging/fwserial: Create loop device the 'tty' way Peter Hurley
2013-01-29  3:34 ` [PATCH 05/11] staging/fwserial: Cleanup /proc/tty/driver/ file Peter Hurley
2013-01-29  3:34 ` [PATCH 06/11] staging/fwserial: Factor unstable stats/debug/status info to debugfs Peter Hurley
2013-01-29  3:34 ` [PATCH 07/11] staging/fwserial: Don't use deprecated alloc_tty_driver() Peter Hurley
2013-01-29  3:34 ` [PATCH 08/11] staging/fwserial: Remove reference to removed constant Peter Hurley
2013-01-29  3:34 ` [PATCH 09/11] staging/fwserial: add diagnostic for buffer overflow Peter Hurley
2013-01-29  3:34 ` Peter Hurley [this message]
2013-01-30  4:29   ` [PATCH 10/11] staging/fwserial: Fix premature unthrottle Greg Kroah-Hartman
2013-01-30 13:11     ` Peter Hurley
2013-01-30 13:32       ` Greg Kroah-Hartman
2013-01-30 13:37         ` Peter Hurley
2013-01-30 16:25           ` Greg Kroah-Hartman
2013-01-29  3:34 ` [PATCH 11/11] staging/fwserial: Remove unneeded push work Peter Hurley

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1359430485-5301-10-git-send-email-peter@hurleysoftware.com \
    --to=peter@hurleysoftware.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux1394-devel@lists.sourceforge.net \
    --cc=stefanr@s5r6.in-berlin.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).