linux-bluetooth.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Peter Hurley <peter@hurleysoftware.com>
To: Marcel Holtmann <marcel@holtmann.org>
Cc: Gustavo Padovan <gustavo@padovan.org>,
	Johan Hedberg <johan.hedberg@gmail.com>,
	Gianluca Anzolin <gianluca@sottospazio.it>,
	Alexander Holler <holler@ahsoftware.de>,
	Andrey Vihrov <andrey.vihrov@gmail.com>,
	Sander Eikelenboom <linux@eikelenboom.it>,
	linux-bluetooth@vger.kernel.org, linux-kernel@vger.kernel.org,
	Peter Hurley <peter@hurleysoftware.com>
Subject: [PATCH 24/24] Bluetooth: Fix write_room() calculation
Date: Sun,  9 Feb 2014 20:59:24 -0500	[thread overview]
Message-ID: <1391997564-1805-25-git-send-email-peter@hurleysoftware.com> (raw)
In-Reply-To: <1391997564-1805-1-git-send-email-peter@hurleysoftware.com>

The skb truesize of a 12-byte payload with a 10-byte head/tail
reserve is 768 bytes. Consequently, even with 40 tx_credits, at
most 6 packets could be queued at any one time:

  40 tx_credits * 127-byte mtu < 768-byte truesize * 7

This error could also cause the tx queue to apparently stall if
credit flow control is disabled (where tx_credits is fixed at 5),
or if the receiver only granted a limited number of tx credits
(eg., less than 7).

Instead, track the outstanding number of queued packets not yet sent
in wmem_alloc and allow for a maximum of 40 queued packets. Report
the space avail for a single write() as the mtu * number of packets
left before reaching the maximum.

Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
 net/bluetooth/rfcomm/tty.c | 14 +++++---------
 1 file changed, 5 insertions(+), 9 deletions(-)

diff --git a/net/bluetooth/rfcomm/tty.c b/net/bluetooth/rfcomm/tty.c
index 3f44195..403ec09 100644
--- a/net/bluetooth/rfcomm/tty.c
+++ b/net/bluetooth/rfcomm/tty.c
@@ -352,20 +352,16 @@ static inline unsigned int rfcomm_room(struct rfcomm_dev *dev)
 {
 	struct rfcomm_dlc *dlc = dev->dlc;
 
-	/* The limit is bogus; the number of packets which can
-	 * currently be sent by the krfcommd thread has no relevance
-	 * to the number of packets which can be queued on the dlc's
-	 * tx queue.
-	 */
-	int limit = dlc->mtu * (dlc->tx_credits?:1);
+	/* Limit the outstanding number of packets not yet sent to 40 */
+	int pending = 40 - atomic_read(&dev->wmem_alloc);
 
-	return max(0, limit - atomic_read(&dev->wmem_alloc));
+	return max(0, pending) * dlc->mtu;
 }
 
 static void rfcomm_wfree(struct sk_buff *skb)
 {
 	struct rfcomm_dev *dev = (void *) skb->sk;
-	atomic_sub(skb->truesize, &dev->wmem_alloc);
+	atomic_dec(&dev->wmem_alloc);
 	if (test_bit(RFCOMM_TTY_ATTACHED, &dev->flags))
 		tty_port_tty_wakeup(&dev->port);
 	tty_port_put(&dev->port);
@@ -374,7 +370,7 @@ static void rfcomm_wfree(struct sk_buff *skb)
 static void rfcomm_set_owner_w(struct sk_buff *skb, struct rfcomm_dev *dev)
 {
 	tty_port_get(&dev->port);
-	atomic_add(skb->truesize, &dev->wmem_alloc);
+	atomic_inc(&dev->wmem_alloc);
 	skb->sk = (void *) dev;
 	skb->destructor = rfcomm_wfree;
 }
-- 
1.8.1.2

  parent reply	other threads:[~2014-02-10  1:59 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-02-10  1:59 [PATCH 00/24] rfcomm fixes Peter Hurley
2014-02-10  1:59 ` [PATCH 01/24] Revert "Bluetooth: Remove rfcomm_carrier_raised()" Peter Hurley
2014-02-10  1:59 ` [PATCH 02/24] Revert "Bluetooth: Always wait for a connection on RFCOMM open()" Peter Hurley
2014-02-10  1:59 ` [PATCH 03/24] Revert "Bluetooth: Move rfcomm_get_device() before rfcomm_dev_activate()" Peter Hurley
2014-02-10  1:59 ` [PATCH 04/24] tty: Fix ref counting for port krefs Peter Hurley
2014-02-13 18:36   ` Greg Kroah-Hartman
2014-02-10  1:59 ` [PATCH 05/24] Bluetooth: Fix racy acquire of rfcomm_dev reference Peter Hurley
2014-02-10  1:59 ` [PATCH 06/24] Bluetooth: Exclude released devices from RFCOMMGETDEVLIST ioctl Peter Hurley
2014-02-10  1:59 ` [PATCH 07/24] Bluetooth: Release rfcomm_dev only once Peter Hurley
2014-02-10  1:59 ` [PATCH 08/24] Bluetooth: Fix unreleased rfcomm_dev reference Peter Hurley
2014-02-10  1:59 ` [PATCH 09/24] Bluetooth: Fix RFCOMM tty teardown race Peter Hurley
2014-02-10  1:59 ` [PATCH 10/24] Bluetooth: Verify dlci not in use before rfcomm_dev create Peter Hurley
2014-02-10  1:59 ` [PATCH 11/24] Bluetooth: Simplify RFCOMM session state eval Peter Hurley
2014-02-10  1:59 ` [PATCH 12/24] Bluetooth: Refactor deferred setup test in rfcomm_dlc_close() Peter Hurley
2014-02-10  1:59 ` [PATCH 13/24] Bluetooth: Refactor dlc disconnect logic " Peter Hurley
2014-02-10  1:59 ` [PATCH 14/24] Bluetooth: Directly close dlc for not yet started RFCOMM session Peter Hurley
2014-02-10  1:59 ` [PATCH 15/24] Bluetooth: Fix unsafe RFCOMM device parenting Peter Hurley
2014-02-10  1:59 ` [PATCH 16/24] Bluetooth: Fix RFCOMM parent device for reused dlc Peter Hurley
2014-02-10  1:59 ` [PATCH 17/24] Bluetooth: Rename __rfcomm_dev_get() to __rfcomm_dev_lookup() Peter Hurley
2014-02-10  1:59 ` [PATCH 18/24] Bluetooth: Serialize RFCOMMCREATEDEV and RFCOMMRELEASEDEV ioctls Peter Hurley
2014-02-10  1:59 ` [PATCH 19/24] Bluetooth: Refactor rfcomm_dev_add() Peter Hurley
2014-02-10  1:59 ` [PATCH 20/24] Bluetooth: Cleanup RFCOMM device registration error handling Peter Hurley
2014-02-10  1:59 ` [PATCH 21/24] Bluetooth: Force -EIO from tty read/write if .activate() fails Peter Hurley
2014-02-10  1:59 ` [PATCH 22/24] Bluetooth: Don't fail RFCOMM tty writes Peter Hurley
2014-02-10  1:59 ` [PATCH 23/24] Bluetooth: Refactor write_room() calculation Peter Hurley
2014-02-10  1:59 ` Peter Hurley [this message]
2014-02-10 22:09 ` [PATCH 00/24] rfcomm fixes Marcel Holtmann
2014-02-10 23:00   ` Peter Hurley
2014-02-12 22:58     ` Marcel Holtmann
2014-02-13  0:38       ` Peter Hurley
2014-02-13 21:48         ` Alexander Holler
2014-02-12 11:06   ` Sander Eikelenboom
2014-03-03 19:38     ` Sander Eikelenboom
2014-03-10  8:38       ` [RC6 Bell Chime] " Sander Eikelenboom
2014-03-10 15:08         ` John W. Linville
2014-03-11 15:14           ` [RC6 Bell Chime] " Marcel Holtmann
2014-03-14  0:49             ` Sander Eikelenboom
2014-03-14  1:28               ` Marcel Holtmann
2014-03-14  1:29               ` Peter Hurley
2014-03-15 13:51                 ` Sander Eikelenboom
2014-03-15 17:53                   ` Linus Torvalds
2014-03-15 20:45                     ` Peter Hurley
2014-03-15 22:20                       ` Sander Eikelenboom
2014-03-16  0:16                       ` Linus Torvalds
2014-02-13 21:41 ` Alexander Holler
2014-02-14 21:45 ` Marcel Holtmann

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=1391997564-1805-25-git-send-email-peter@hurleysoftware.com \
    --to=peter@hurleysoftware.com \
    --cc=andrey.vihrov@gmail.com \
    --cc=gianluca@sottospazio.it \
    --cc=gustavo@padovan.org \
    --cc=holler@ahsoftware.de \
    --cc=johan.hedberg@gmail.com \
    --cc=linux-bluetooth@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux@eikelenboom.it \
    --cc=marcel@holtmann.org \
    /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).