From: Florian Fainelli <f.fainelli@gmail.com>
To: netdev@vger.kernel.org
Cc: davem@davemloft.net, Florian Fainelli <f.fainelli@gmail.com>
Subject: [PATCH net-next 1/4] net: systemport: fix transmit locking scheme
Date: Thu, 5 Jun 2014 10:22:15 -0700 [thread overview]
Message-ID: <1401988938-25940-2-git-send-email-f.fainelli@gmail.com> (raw)
In-Reply-To: <1401988938-25940-1-git-send-email-f.fainelli@gmail.com>
Our transmit locking scheme did not account for the TX ring full
interrupt. If a TX ring full interrupt fires while we are attempting to
transmit, we will cause a deadlock to occur. Fix this by making sure
that we properly disable interrupts while acquiring the spinlock.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/ethernet/broadcom/bcmsysport.c | 12 +++++++-----
1 file changed, 7 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c
index 25982b02f0ea..4dfc93fe9744 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.c
+++ b/drivers/net/ethernet/broadcom/bcmsysport.c
@@ -637,10 +637,11 @@ static unsigned int bcm_sysport_tx_reclaim(struct bcm_sysport_priv *priv,
struct bcm_sysport_tx_ring *ring)
{
unsigned int released;
+ unsigned long flags;
- spin_lock(&ring->lock);
+ spin_lock_irqsave(&ring->lock, flags);
released = __bcm_sysport_tx_reclaim(priv, ring);
- spin_unlock(&ring->lock);
+ spin_unlock_irqrestore(&ring->lock, flags);
return released;
}
@@ -822,6 +823,7 @@ static netdev_tx_t bcm_sysport_xmit(struct sk_buff *skb,
struct netdev_queue *txq;
struct dma_desc *desc;
unsigned int skb_len;
+ unsigned long flags;
dma_addr_t mapping;
u32 len_status;
u16 queue;
@@ -831,8 +833,8 @@ static netdev_tx_t bcm_sysport_xmit(struct sk_buff *skb,
txq = netdev_get_tx_queue(dev, queue);
ring = &priv->tx_rings[queue];
- /* lock against tx reclaim in BH context */
- spin_lock(&ring->lock);
+ /* lock against tx reclaim in BH context and TX ring full interrupt */
+ spin_lock_irqsave(&ring->lock, flags);
if (unlikely(ring->desc_count == 0)) {
netif_tx_stop_queue(txq);
netdev_err(dev, "queue %d awake and ring full!\n", queue);
@@ -914,7 +916,7 @@ static netdev_tx_t bcm_sysport_xmit(struct sk_buff *skb,
ret = NETDEV_TX_OK;
out:
- spin_unlock(&ring->lock);
+ spin_unlock_irqrestore(&ring->lock, flags);
return ret;
}
--
1.9.1
next prev parent reply other threads:[~2014-06-05 17:22 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-06-05 17:22 [PATCH net-next 0/4] net: systemport: misc fixes Florian Fainelli
2014-06-05 17:22 ` Florian Fainelli [this message]
2014-06-06 9:29 ` [PATCH net-next 1/4] net: systemport: fix transmit locking scheme David Laight
2014-06-06 17:00 ` Florian Fainelli
2014-06-05 17:22 ` [PATCH net-next 2/4] net: systemport: correctly check for RX_STATUS_OVFLOW flag Florian Fainelli
2014-06-05 17:22 ` [PATCH net-next 3/4] net: systemport: fix comment about HW prepending 2bytes Florian Fainelli
2014-06-05 17:22 ` [PATCH net-next 4/4] net: systemport: start with carrier off Florian Fainelli
2014-06-05 22:37 ` [PATCH net-next 0/4] net: systemport: misc fixes David Miller
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=1401988938-25940-2-git-send-email-f.fainelli@gmail.com \
--to=f.fainelli@gmail.com \
--cc=davem@davemloft.net \
--cc=netdev@vger.kernel.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).