* Re: [PATCH 2/2] IP_GRE: Linearize skb before csum.
From: David Miller @ 2013-01-23 4:22 UTC (permalink / raw)
To: eric.dumazet; +Cc: jesse, pshelar, netdev
In-Reply-To: <20130122.204629.607263597214211092.davem@davemloft.net>
From: David Miller <davem@davemloft.net>
Date: Tue, 22 Jan 2013 20:46:29 -0500 (EST)
> We can't emit bogus checksums on the wire when this is completely
> under our control.
BTW consider a legitimate use case where an application has a shared
memory page between threads that keeps track of statistic counters,
and it uses sendfile() over an fd backing that mmap()'d area to
send the statistics out remotely over TCP.
That's completely legitimate.
As is an application constantly overwriting a page in the filesystem
page cache, and another app doing sendfile() over TCP from that
area.
In such a situation everyone of your magic retransmits will have a
bad checksum too.
You can't say that this behavior is anything other than broken, and
I don't want to be responsible for a system that's doing broken
things like that.
If your GSO over GRE tunnels stuff fundamentally depends upon this
magic retransmit fantasy working, then yes we have to disable it
completely.
^ permalink raw reply
* Re: [Patch net-next] ipv6: remove duplicated declaration of ip6_fragment()
From: David Miller @ 2013-01-23 4:22 UTC (permalink / raw)
To: amwang; +Cc: netdev, yoshfuji
In-Reply-To: <1358911327-12122-1-git-send-email-amwang@redhat.com>
From: Cong Wang <amwang@redhat.com>
Date: Wed, 23 Jan 2013 11:22:07 +0800
> From: Cong Wang <amwang@redhat.com>
>
> It is declared in:
> include/net/ip6_route.h:187:int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));
>
> and net/ip6_route.h is already included.
>
> Cc: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
> Cc: David S. Miller <davem@davemloft.net>
> Signed-off-by: Cong Wang <amwang@redhat.com>
Applied.
^ permalink raw reply
* Re: [Patch net-next] netpoll: fix an uninitialized variable
From: David Miller @ 2013-01-23 4:22 UTC (permalink / raw)
To: amwang; +Cc: netdev
In-Reply-To: <1358912351-14835-1-git-send-email-amwang@redhat.com>
From: Cong Wang <amwang@redhat.com>
Date: Wed, 23 Jan 2013 11:39:11 +0800
> From: Cong Wang <amwang@redhat.com>
>
> Fengguang reported:
>
> net/core/netpoll.c: In function 'netpoll_setup':
> net/core/netpoll.c:1049:6: warning: 'err' may be used uninitialized in this function [-Wmaybe-uninitialized]
>
> in !CONFIG_IPV6 case, we may error out without initializing
> 'err'.
>
> Reported-by: Fengguang Wu <fengguang.wu@intel.com>
> Cc: David S. Miller <davem@davemloft.net>
> Signed-off-by: Cong Wang <amwang@redhat.com>
Applied.
^ permalink raw reply
* Re: [PATCH net-next] drivers/net/ethernet/micrel/ks8851_mll: Implement basic statistics
From: Joe Perches @ 2013-01-23 4:27 UTC (permalink / raw)
To: David Miller; +Cc: David.Choi, netdev, Ping.Doong
In-Reply-To: <20130122.231754.2210645494690728329.davem@davemloft.net>
On Tue, 2013-01-22 at 23:17 -0500, David Miller wrote:
> Be also careful to not match the ones that are done to
> do something like the initial RX ring filling during
> device open.
>
> If such an allocation failure causes device open to fail,
> that's a legitimate situation in which to log.
True.
These are only the alloc_skb()s with rx_dropped++
instances that aren't init cases.
Here's the patch if anyone wants to do
something with it.
unsigned, uncompiled for !x86, untested...
drivers/net/appletalk/cops.c | 2 --
drivers/net/can/grcan.c | 2 --
drivers/net/can/mcp251x.c | 1 -
drivers/net/can/mscan/mscan.c | 2 --
drivers/net/ethernet/adi/bfin_mac.c | 1 -
drivers/net/ethernet/aeroflex/greth.c | 7 -------
drivers/net/ethernet/amd/7990.c | 2 --
drivers/net/ethernet/amd/a2065.c | 1 -
drivers/net/ethernet/amd/am79c961a.c | 1 -
drivers/net/ethernet/amd/au1000_eth.c | 1 -
drivers/net/ethernet/amd/declance.c | 2 --
drivers/net/ethernet/amd/ni65.c | 1 -
drivers/net/ethernet/amd/sunlance.c | 4 ----
drivers/net/ethernet/cadence/at91_ether.c | 1 -
drivers/net/ethernet/cirrus/cs89x0.c | 3 ---
drivers/net/ethernet/cirrus/mac89x0.c | 1 -
drivers/net/ethernet/ethoc.c | 4 ----
drivers/net/ethernet/freescale/fec.c | 2 --
drivers/net/ethernet/freescale/fec_mpc52xx.c | 1 -
drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c | 4 ----
drivers/net/ethernet/fujitsu/fmvj18x_cs.c | 2 --
drivers/net/ethernet/i825xx/82596.c | 1 -
drivers/net/ethernet/i825xx/lib82596.c | 3 ---
drivers/net/ethernet/microchip/enc28j60.c | 3 ---
drivers/net/ethernet/natsemi/sonic.c | 1 -
drivers/net/ethernet/netx-eth.c | 2 --
drivers/net/ethernet/nuvoton/w90p910_ether.c | 1 -
drivers/net/ethernet/realtek/8139too.c | 2 --
drivers/net/ethernet/realtek/atp.c | 2 --
drivers/net/ethernet/seeq/sgiseeq.c | 2 --
drivers/net/ethernet/sis/sis900.c | 4 ----
drivers/net/ethernet/smsc/smc9194.c | 1 -
drivers/net/ethernet/smsc/smc91c92_cs.c | 1 -
drivers/net/ethernet/smsc/smc91x.c | 2 --
drivers/net/ethernet/xilinx/xilinx_emaclite.c | 1 -
drivers/net/ethernet/xircom/xirc2ps_cs.c | 1 -
drivers/net/hamradio/baycom_epp.c | 1 -
drivers/net/hamradio/hdlcdrv.c | 1 -
drivers/net/hamradio/mkiss.c | 2 --
drivers/net/hippi/rrunner.c | 3 ---
drivers/net/irda/pxaficp_ir.c | 1 -
drivers/net/sb1000.c | 3 ---
drivers/net/slip/slip.c | 1 -
drivers/net/usb/ipheth.c | 2 --
drivers/net/wan/cosa.c | 1 -
drivers/net/wan/sdla.c | 1 -
drivers/net/wan/x25_asy.c | 1 -
drivers/net/wan/z85230.c | 1 -
drivers/net/wimax/i2400m/netdev.c | 1 -
drivers/net/wireless/ray_cs.c | 1 -
drivers/net/wireless/wl3501_cs.c | 2 --
51 files changed, 94 deletions(-)
diff --git a/drivers/net/appletalk/cops.c b/drivers/net/appletalk/cops.c
index cff6f02..bd01b72 100644
--- a/drivers/net/appletalk/cops.c
+++ b/drivers/net/appletalk/cops.c
@@ -790,8 +790,6 @@ static void cops_rx(struct net_device *dev)
skb = dev_alloc_skb(pkt_len);
if(skb == NULL)
{
- printk(KERN_WARNING "%s: Memory squeeze, dropping packet.\n",
- dev->name);
dev->stats.rx_dropped++;
while(pkt_len--) /* Discard packet */
inb(ioaddr);
diff --git a/drivers/net/can/grcan.c b/drivers/net/can/grcan.c
index 17fbc7a..e7277e4 100644
--- a/drivers/net/can/grcan.c
+++ b/drivers/net/can/grcan.c
@@ -1191,8 +1191,6 @@ static int grcan_receive(struct net_device *dev, int budget)
/* Take care of packet */
skb = alloc_can_skb(dev, &cf);
if (skb == NULL) {
- netdev_err(dev,
- "dropping frame: skb allocation failed\n");
stats->rx_dropped++;
continue;
}
diff --git a/drivers/net/can/mcp251x.c b/drivers/net/can/mcp251x.c
index 5eaf47b..ec1b04f 100644
--- a/drivers/net/can/mcp251x.c
+++ b/drivers/net/can/mcp251x.c
@@ -460,7 +460,6 @@ static void mcp251x_hw_rx(struct spi_device *spi, int buf_idx)
skb = alloc_can_skb(priv->net, &frame);
if (!skb) {
- dev_err(&spi->dev, "cannot allocate RX skb\n");
priv->net->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/can/mscan/mscan.c b/drivers/net/can/mscan/mscan.c
index e6b4095..0789c81 100644
--- a/drivers/net/can/mscan/mscan.c
+++ b/drivers/net/can/mscan/mscan.c
@@ -426,8 +426,6 @@ static int mscan_rx_poll(struct napi_struct *napi, int quota)
skb = alloc_can_skb(dev, &frame);
if (!skb) {
- if (printk_ratelimit())
- netdev_notice(dev, "packet dropped\n");
stats->rx_dropped++;
out_8(®s->canrflg, canrflg);
continue;
diff --git a/drivers/net/ethernet/adi/bfin_mac.c b/drivers/net/ethernet/adi/bfin_mac.c
index a175d0b..266f832 100644
--- a/drivers/net/ethernet/adi/bfin_mac.c
+++ b/drivers/net/ethernet/adi/bfin_mac.c
@@ -1236,7 +1236,6 @@ static void bfin_mac_rx(struct net_device *dev)
new_skb = netdev_alloc_skb(dev, PKT_BUF_SZ + NET_IP_ALIGN);
if (!new_skb) {
- netdev_notice(dev, "rx: low on mem - packet dropped\n");
dev->stats.rx_dropped++;
goto out;
}
diff --git a/drivers/net/ethernet/aeroflex/greth.c b/drivers/net/ethernet/aeroflex/greth.c
index 0be2195..2d75b2f 100644
--- a/drivers/net/ethernet/aeroflex/greth.c
+++ b/drivers/net/ethernet/aeroflex/greth.c
@@ -777,12 +777,7 @@ static int greth_rx(struct net_device *dev, int limit)
skb = netdev_alloc_skb(dev, pkt_len + NET_IP_ALIGN);
if (unlikely(skb == NULL)) {
-
- if (net_ratelimit())
- dev_warn(&dev->dev, "low on memory - " "packet dropped\n");
-
dev->stats.rx_dropped++;
-
} else {
skb_reserve(skb, NET_IP_ALIGN);
@@ -934,8 +929,6 @@ static int greth_rx_gbit(struct net_device *dev, int limit)
* table handling should be divided into cleaning and
* filling as the TX part of the driver
*/
- if (net_ratelimit())
- dev_warn(greth->dev, "Could not allocate SKB, dropping packet\n");
/* reusing current skb, so it is a drop */
dev->stats.rx_dropped++;
}
diff --git a/drivers/net/ethernet/amd/7990.c b/drivers/net/ethernet/amd/7990.c
index 6e722dc..65926a9 100644
--- a/drivers/net/ethernet/amd/7990.c
+++ b/drivers/net/ethernet/amd/7990.c
@@ -318,8 +318,6 @@ static int lance_rx (struct net_device *dev)
struct sk_buff *skb = netdev_alloc_skb(dev, len + 2);
if (!skb) {
- printk ("%s: Memory squeeze, deferring packet.\n",
- dev->name);
dev->stats.rx_dropped++;
rd->mblength = 0;
rd->rmd1_bits = LE_R1_OWN;
diff --git a/drivers/net/ethernet/amd/a2065.c b/drivers/net/ethernet/amd/a2065.c
index 3789aff..0866e76 100644
--- a/drivers/net/ethernet/amd/a2065.c
+++ b/drivers/net/ethernet/amd/a2065.c
@@ -293,7 +293,6 @@ static int lance_rx(struct net_device *dev)
struct sk_buff *skb = netdev_alloc_skb(dev, len + 2);
if (!skb) {
- netdev_warn(dev, "Memory squeeze, deferring packet\n");
dev->stats.rx_dropped++;
rd->mblength = 0;
rd->rmd1_bits = LE_R1_OWN;
diff --git a/drivers/net/ethernet/amd/am79c961a.c b/drivers/net/ethernet/amd/am79c961a.c
index 60e2b70..9793767 100644
--- a/drivers/net/ethernet/amd/am79c961a.c
+++ b/drivers/net/ethernet/amd/am79c961a.c
@@ -528,7 +528,6 @@ am79c961_rx(struct net_device *dev, struct dev_priv *priv)
dev->stats.rx_packets++;
} else {
am_writeword (dev, hdraddr + 2, RMD_OWN);
- printk (KERN_WARNING "%s: memory squeeze, dropping packet.\n", dev->name);
dev->stats.rx_dropped++;
break;
}
diff --git a/drivers/net/ethernet/amd/au1000_eth.c b/drivers/net/ethernet/amd/au1000_eth.c
index de774d4..688aede 100644
--- a/drivers/net/ethernet/amd/au1000_eth.c
+++ b/drivers/net/ethernet/amd/au1000_eth.c
@@ -727,7 +727,6 @@ static int au1000_rx(struct net_device *dev)
frmlen -= 4; /* Remove FCS */
skb = netdev_alloc_skb(dev, frmlen + 2);
if (skb == NULL) {
- netdev_err(dev, "Memory squeeze, dropping packet.\n");
dev->stats.rx_dropped++;
continue;
}
diff --git a/drivers/net/ethernet/amd/declance.c b/drivers/net/ethernet/amd/declance.c
index baca0bd..3d86ffe 100644
--- a/drivers/net/ethernet/amd/declance.c
+++ b/drivers/net/ethernet/amd/declance.c
@@ -607,8 +607,6 @@ static int lance_rx(struct net_device *dev)
skb = netdev_alloc_skb(dev, len + 2);
if (skb == 0) {
- printk("%s: Memory squeeze, deferring packet.\n",
- dev->name);
dev->stats.rx_dropped++;
*rds_ptr(rd, mblength, lp->type) = 0;
*rds_ptr(rd, rmd1, lp->type) =
diff --git a/drivers/net/ethernet/amd/ni65.c b/drivers/net/ethernet/amd/ni65.c
index 013b651..9443eff 100644
--- a/drivers/net/ethernet/amd/ni65.c
+++ b/drivers/net/ethernet/amd/ni65.c
@@ -1118,7 +1118,6 @@ static void ni65_recv_intr(struct net_device *dev,int csr0)
}
else
{
- printk(KERN_ERR "%s: can't alloc new sk_buff\n",dev->name);
dev->stats.rx_dropped++;
}
}
diff --git a/drivers/net/ethernet/amd/sunlance.c b/drivers/net/ethernet/amd/sunlance.c
index 6a40290..70d5430 100644
--- a/drivers/net/ethernet/amd/sunlance.c
+++ b/drivers/net/ethernet/amd/sunlance.c
@@ -536,8 +536,6 @@ static void lance_rx_dvma(struct net_device *dev)
skb = netdev_alloc_skb(dev, len + 2);
if (skb == NULL) {
- printk(KERN_INFO "%s: Memory squeeze, deferring packet.\n",
- dev->name);
dev->stats.rx_dropped++;
rd->mblength = 0;
rd->rmd1_bits = LE_R1_OWN;
@@ -708,8 +706,6 @@ static void lance_rx_pio(struct net_device *dev)
skb = netdev_alloc_skb(dev, len + 2);
if (skb == NULL) {
- printk(KERN_INFO "%s: Memory squeeze, deferring packet.\n",
- dev->name);
dev->stats.rx_dropped++;
sbus_writew(0, &rd->mblength);
sbus_writeb(LE_R1_OWN, &rd->rmd1_bits);
diff --git a/drivers/net/ethernet/cadence/at91_ether.c b/drivers/net/ethernet/cadence/at91_ether.c
index 3becdb2..81764bf 100644
--- a/drivers/net/ethernet/cadence/at91_ether.c
+++ b/drivers/net/ethernet/cadence/at91_ether.c
@@ -209,7 +209,6 @@ static void at91ether_rx(struct net_device *dev)
netif_rx(skb);
} else {
lp->stats.rx_dropped++;
- netdev_notice(dev, "Memory squeeze, dropping packet.\n");
}
if (lp->rx_ring[lp->rx_tail].ctrl & MACB_BIT(RX_MHASH_MATCH))
diff --git a/drivers/net/ethernet/cirrus/cs89x0.c b/drivers/net/ethernet/cirrus/cs89x0.c
index 1384469..8fbbee3 100644
--- a/drivers/net/ethernet/cirrus/cs89x0.c
+++ b/drivers/net/ethernet/cirrus/cs89x0.c
@@ -731,9 +731,6 @@ net_rx(struct net_device *dev)
/* Malloc up new buffer. */
skb = netdev_alloc_skb(dev, length + 2);
if (skb == NULL) {
-#if 0 /* Again, this seems a cruel thing to do */
- pr_warn("%s: Memory squeeze, dropping packet\n", dev->name);
-#endif
dev->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/ethernet/cirrus/mac89x0.c b/drivers/net/ethernet/cirrus/mac89x0.c
index e285f38..719b1cb 100644
--- a/drivers/net/ethernet/cirrus/mac89x0.c
+++ b/drivers/net/ethernet/cirrus/mac89x0.c
@@ -509,7 +509,6 @@ net_rx(struct net_device *dev)
/* Malloc up new buffer. */
skb = alloc_skb(length, GFP_ATOMIC);
if (skb == NULL) {
- printk("%s: Memory squeeze, dropping packet.\n", dev->name);
dev->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/ethernet/ethoc.c b/drivers/net/ethernet/ethoc.c
index aa47ef9..3cb0e53 100644
--- a/drivers/net/ethernet/ethoc.c
+++ b/drivers/net/ethernet/ethoc.c
@@ -446,10 +446,6 @@ static int ethoc_rx(struct net_device *dev, int limit)
dev->stats.rx_bytes += size;
netif_receive_skb(skb);
} else {
- if (net_ratelimit())
- dev_warn(&dev->dev, "low on memory - "
- "packet dropped\n");
-
dev->stats.rx_dropped++;
break;
}
diff --git a/drivers/net/ethernet/freescale/fec.c b/drivers/net/ethernet/freescale/fec.c
index f52ba33..22cef5d 100644
--- a/drivers/net/ethernet/freescale/fec.c
+++ b/drivers/net/ethernet/freescale/fec.c
@@ -736,8 +736,6 @@ fec_enet_rx(struct net_device *ndev)
skb = netdev_alloc_skb(ndev, pkt_len - 4 + NET_IP_ALIGN);
if (unlikely(!skb)) {
- printk("%s: Memory squeeze, dropping packet.\n",
- ndev->name);
ndev->stats.rx_dropped++;
} else {
skb_reserve(skb, NET_IP_ALIGN);
diff --git a/drivers/net/ethernet/freescale/fec_mpc52xx.c b/drivers/net/ethernet/freescale/fec_mpc52xx.c
index 817d081..bdfbf59 100644
--- a/drivers/net/ethernet/freescale/fec_mpc52xx.c
+++ b/drivers/net/ethernet/freescale/fec_mpc52xx.c
@@ -419,7 +419,6 @@ static irqreturn_t mpc52xx_fec_rx_interrupt(int irq, void *dev_id)
skb = netdev_alloc_skb(dev, FEC_RX_BUFFER_SIZE);
if (!skb) {
/* Can't get a new one : reuse the same & drop pkt */
- dev_notice(&dev->dev, "Low memory - dropped packet.\n");
mpc52xx_fec_rx_submit(dev, rskb);
dev->stats.rx_dropped++;
continue;
diff --git a/drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c b/drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c
index 46df288..02edb0f 100644
--- a/drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c
+++ b/drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c
@@ -177,8 +177,6 @@ static int fs_enet_rx_napi(struct napi_struct *napi, int budget)
received++;
netif_receive_skb(skb);
} else {
- dev_warn(fep->dev,
- "Memory squeeze, dropping packet.\n");
fep->stats.rx_dropped++;
skbn = skb;
}
@@ -309,8 +307,6 @@ static int fs_enet_rx_non_napi(struct net_device *dev)
received++;
netif_rx(skb);
} else {
- dev_warn(fep->dev,
- "Memory squeeze, dropping packet.\n");
fep->stats.rx_dropped++;
skbn = skb;
}
diff --git a/drivers/net/ethernet/fujitsu/fmvj18x_cs.c b/drivers/net/ethernet/fujitsu/fmvj18x_cs.c
index 2418faf..8412570 100644
--- a/drivers/net/ethernet/fujitsu/fmvj18x_cs.c
+++ b/drivers/net/ethernet/fujitsu/fmvj18x_cs.c
@@ -1003,8 +1003,6 @@ static void fjn_rx(struct net_device *dev)
}
skb = netdev_alloc_skb(dev, pkt_len + 2);
if (skb == NULL) {
- netdev_notice(dev, "Memory squeeze, dropping packet (len %d)\n",
- pkt_len);
outb(F_SKP_PKT, ioaddr + RX_SKIP);
dev->stats.rx_dropped++;
break;
diff --git a/drivers/net/ethernet/i825xx/82596.c b/drivers/net/ethernet/i825xx/82596.c
index 1c54e22..1760080 100644
--- a/drivers/net/ethernet/i825xx/82596.c
+++ b/drivers/net/ethernet/i825xx/82596.c
@@ -804,7 +804,6 @@ static inline int i596_rx(struct net_device *dev)
memory_squeeze:
if (skb == NULL) {
/* XXX tulip.c can defer packets here!! */
- printk(KERN_WARNING "%s: i596_rx Memory squeeze, dropping packet.\n", dev->name);
dev->stats.rx_dropped++;
}
else {
diff --git a/drivers/net/ethernet/i825xx/lib82596.c b/drivers/net/ethernet/i825xx/lib82596.c
index f045ea4..fc8da33 100644
--- a/drivers/net/ethernet/i825xx/lib82596.c
+++ b/drivers/net/ethernet/i825xx/lib82596.c
@@ -720,9 +720,6 @@ static inline int i596_rx(struct net_device *dev)
memory_squeeze:
if (skb == NULL) {
/* XXX tulip.c can defer packets here!! */
- printk(KERN_ERR
- "%s: i596_rx Memory squeeze, dropping packet.\n",
- dev->name);
dev->stats.rx_dropped++;
} else {
if (!rx_in_place) {
diff --git a/drivers/net/ethernet/microchip/enc28j60.c b/drivers/net/ethernet/microchip/enc28j60.c
index 5d98a9f..e8ed5da 100644
--- a/drivers/net/ethernet/microchip/enc28j60.c
+++ b/drivers/net/ethernet/microchip/enc28j60.c
@@ -956,9 +956,6 @@ static void enc28j60_hw_rx(struct net_device *ndev)
} else {
skb = netdev_alloc_skb(ndev, len + NET_IP_ALIGN);
if (!skb) {
- if (netif_msg_rx_err(priv))
- dev_err(&ndev->dev,
- "out of memory for Rx'd frame\n");
ndev->stats.rx_dropped++;
} else {
skb_reserve(skb, NET_IP_ALIGN);
diff --git a/drivers/net/ethernet/natsemi/sonic.c b/drivers/net/ethernet/natsemi/sonic.c
index 46795e4..1bd419d 100644
--- a/drivers/net/ethernet/natsemi/sonic.c
+++ b/drivers/net/ethernet/natsemi/sonic.c
@@ -424,7 +424,6 @@ static void sonic_rx(struct net_device *dev)
/* Malloc up new buffer. */
new_skb = netdev_alloc_skb(dev, SONIC_RBSIZE + 2);
if (new_skb == NULL) {
- printk(KERN_ERR "%s: Memory squeeze, dropping packet.\n", dev->name);
lp->stats.rx_dropped++;
break;
}
diff --git a/drivers/net/ethernet/netx-eth.c b/drivers/net/ethernet/netx-eth.c
index 63e7af4..cb9e638 100644
--- a/drivers/net/ethernet/netx-eth.c
+++ b/drivers/net/ethernet/netx-eth.c
@@ -152,8 +152,6 @@ static void netx_eth_receive(struct net_device *ndev)
skb = netdev_alloc_skb(ndev, len);
if (unlikely(skb == NULL)) {
- printk(KERN_NOTICE "%s: Low memory, packet dropped.\n",
- ndev->name);
ndev->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/ethernet/nuvoton/w90p910_ether.c b/drivers/net/ethernet/nuvoton/w90p910_ether.c
index 162da89..539d202 100644
--- a/drivers/net/ethernet/nuvoton/w90p910_ether.c
+++ b/drivers/net/ethernet/nuvoton/w90p910_ether.c
@@ -737,7 +737,6 @@ static void netdev_rx(struct net_device *dev)
data = ether->rdesc->recv_buf[ether->cur_rx];
skb = netdev_alloc_skb(dev, length + 2);
if (!skb) {
- dev_err(&pdev->dev, "get skb buffer error\n");
ether->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/ethernet/realtek/8139too.c b/drivers/net/ethernet/realtek/8139too.c
index 1276ac7..3ccedeb 100644
--- a/drivers/net/ethernet/realtek/8139too.c
+++ b/drivers/net/ethernet/realtek/8139too.c
@@ -2041,8 +2041,6 @@ keep_pkt:
netif_receive_skb (skb);
} else {
- if (net_ratelimit())
- netdev_warn(dev, "Memory squeeze, dropping packet\n");
dev->stats.rx_dropped++;
}
received++;
diff --git a/drivers/net/ethernet/realtek/atp.c b/drivers/net/ethernet/realtek/atp.c
index 9f2d416..d77d60e 100644
--- a/drivers/net/ethernet/realtek/atp.c
+++ b/drivers/net/ethernet/realtek/atp.c
@@ -782,8 +782,6 @@ static void net_rx(struct net_device *dev)
skb = netdev_alloc_skb(dev, pkt_len + 2);
if (skb == NULL) {
- printk(KERN_ERR "%s: Memory squeeze, dropping packet.\n",
- dev->name);
dev->stats.rx_dropped++;
goto done;
}
diff --git a/drivers/net/ethernet/seeq/sgiseeq.c b/drivers/net/ethernet/seeq/sgiseeq.c
index 0fde9ca..0ad5694 100644
--- a/drivers/net/ethernet/seeq/sgiseeq.c
+++ b/drivers/net/ethernet/seeq/sgiseeq.c
@@ -381,8 +381,6 @@ memory_squeeze:
dev->stats.rx_packets++;
dev->stats.rx_bytes += len;
} else {
- printk(KERN_NOTICE "%s: Memory squeeze, deferring packet.\n",
- dev->name);
dev->stats.rx_dropped++;
}
} else {
diff --git a/drivers/net/ethernet/sis/sis900.c b/drivers/net/ethernet/sis/sis900.c
index efca14e..32714db 100644
--- a/drivers/net/ethernet/sis/sis900.c
+++ b/drivers/net/ethernet/sis/sis900.c
@@ -1846,10 +1846,6 @@ refill_rx_ring:
* "hole" on the buffer ring, it is not clear
* how the hardware will react to this kind
* of degenerated buffer */
- if (netif_msg_rx_err(sis_priv))
- printk(KERN_INFO "%s: Memory squeeze, "
- "deferring packet.\n",
- net_dev->name);
net_dev->stats.rx_dropped++;
break;
}
diff --git a/drivers/net/ethernet/smsc/smc9194.c b/drivers/net/ethernet/smsc/smc9194.c
index 50823da..4ff72a6 100644
--- a/drivers/net/ethernet/smsc/smc9194.c
+++ b/drivers/net/ethernet/smsc/smc9194.c
@@ -1225,7 +1225,6 @@ static void smc_rcv(struct net_device *dev)
skb = netdev_alloc_skb(dev, packet_length + 5);
if ( skb == NULL ) {
- printk(KERN_NOTICE CARDNAME ": Low memory, packet dropped.\n");
dev->stats.rx_dropped++;
goto done;
}
diff --git a/drivers/net/ethernet/smsc/smc91c92_cs.c b/drivers/net/ethernet/smsc/smc91c92_cs.c
index 04393b5..9dfdf46 100644
--- a/drivers/net/ethernet/smsc/smc91c92_cs.c
+++ b/drivers/net/ethernet/smsc/smc91c92_cs.c
@@ -1502,7 +1502,6 @@ static void smc_rx(struct net_device *dev)
skb = netdev_alloc_skb(dev, packet_length+2);
if (skb == NULL) {
- pr_debug("%s: Low memory, packet dropped.\n", dev->name);
dev->stats.rx_dropped++;
outw(MC_RELEASE, ioaddr + MMU_CMD);
return;
diff --git a/drivers/net/ethernet/smsc/smc91x.c b/drivers/net/ethernet/smsc/smc91x.c
index 591650a..dfbf978 100644
--- a/drivers/net/ethernet/smsc/smc91x.c
+++ b/drivers/net/ethernet/smsc/smc91x.c
@@ -465,8 +465,6 @@ static inline void smc_rcv(struct net_device *dev)
*/
skb = netdev_alloc_skb(dev, packet_len);
if (unlikely(skb == NULL)) {
- printk(KERN_NOTICE "%s: Low memory, packet dropped.\n",
- dev->name);
SMC_WAIT_MMU_BUSY(lp);
SMC_SET_MMU_CMD(lp, MC_RELEASE);
dev->stats.rx_dropped++;
diff --git a/drivers/net/ethernet/xilinx/xilinx_emaclite.c b/drivers/net/ethernet/xilinx/xilinx_emaclite.c
index 919b983..97bf214 100644
--- a/drivers/net/ethernet/xilinx/xilinx_emaclite.c
+++ b/drivers/net/ethernet/xilinx/xilinx_emaclite.c
@@ -617,7 +617,6 @@ static void xemaclite_rx_handler(struct net_device *dev)
if (!skb) {
/* Couldn't get memory. */
dev->stats.rx_dropped++;
- dev_err(&lp->ndev->dev, "Could not allocate receive buffer\n");
return;
}
diff --git a/drivers/net/ethernet/xircom/xirc2ps_cs.c b/drivers/net/ethernet/xircom/xirc2ps_cs.c
index 98e09d0..76210ab 100644
--- a/drivers/net/ethernet/xircom/xirc2ps_cs.c
+++ b/drivers/net/ethernet/xircom/xirc2ps_cs.c
@@ -1041,7 +1041,6 @@ xirc2ps_interrupt(int irq, void *dev_id)
/* 1 extra so we can use insw */
skb = netdev_alloc_skb(dev, pktlen + 3);
if (!skb) {
- pr_notice("low memory, packet dropped (size=%u)\n", pktlen);
dev->stats.rx_dropped++;
} else { /* okay get the packet */
skb_reserve(skb, 2);
diff --git a/drivers/net/hamradio/baycom_epp.c b/drivers/net/hamradio/baycom_epp.c
index 49b8b58..dc1a80b 100644
--- a/drivers/net/hamradio/baycom_epp.c
+++ b/drivers/net/hamradio/baycom_epp.c
@@ -546,7 +546,6 @@ static void do_rxpacket(struct net_device *dev)
return;
pktlen = bc->hdlcrx.bufcnt-2+1; /* KISS kludge */
if (!(skb = dev_alloc_skb(pktlen))) {
- printk("%s: memory squeeze, dropping packet\n", dev->name);
dev->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/hamradio/hdlcdrv.c b/drivers/net/hamradio/hdlcdrv.c
index a4a3516..7fbc1ad 100644
--- a/drivers/net/hamradio/hdlcdrv.c
+++ b/drivers/net/hamradio/hdlcdrv.c
@@ -153,7 +153,6 @@ static void hdlc_rx_flag(struct net_device *dev, struct hdlcdrv_state *s)
return;
pkt_len = s->hdlcrx.len - 2 + 1; /* KISS kludge */
if (!(skb = dev_alloc_skb(pkt_len))) {
- printk("%s: memory squeeze, dropping packet\n", dev->name);
dev->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/hamradio/mkiss.c b/drivers/net/hamradio/mkiss.c
index 8e01c45..f9f7cdb 100644
--- a/drivers/net/hamradio/mkiss.c
+++ b/drivers/net/hamradio/mkiss.c
@@ -292,8 +292,6 @@ static void ax_bump(struct mkiss *ax)
count = ax->rcount;
if ((skb = dev_alloc_skb(count)) == NULL) {
- printk(KERN_ERR "mkiss: %s: memory squeeze, dropping packet.\n",
- ax->dev->name);
ax->dev->stats.rx_dropped++;
spin_unlock_bh(&ax->buflock);
return;
diff --git a/drivers/net/hippi/rrunner.c b/drivers/net/hippi/rrunner.c
index e5b19b0..5e52b25 100644
--- a/drivers/net/hippi/rrunner.c
+++ b/drivers/net/hippi/rrunner.c
@@ -954,7 +954,6 @@ static void rx_int(struct net_device *dev, u32 rxlimit, u32 index)
if (pkt_len < PKT_COPY_THRESHOLD) {
skb = alloc_skb(pkt_len, GFP_ATOMIC);
if (skb == NULL){
- printk(KERN_WARNING "%s: Unable to allocate skb (%i bytes), deferring packet\n", dev->name, pkt_len);
dev->stats.rx_dropped++;
goto defer;
} else {
@@ -991,8 +990,6 @@ static void rx_int(struct net_device *dev, u32 rxlimit, u32 index)
PCI_DMA_FROMDEVICE);
set_rraddr(&desc->addr, addr);
} else {
- printk("%s: Out of memory, deferring "
- "packet\n", dev->name);
dev->stats.rx_dropped++;
goto defer;
}
diff --git a/drivers/net/irda/pxaficp_ir.c b/drivers/net/irda/pxaficp_ir.c
index 858de05..5b021cf 100644
--- a/drivers/net/irda/pxaficp_ir.c
+++ b/drivers/net/irda/pxaficp_ir.c
@@ -443,7 +443,6 @@ static void pxa_irda_fir_irq_eif(struct pxa_irda *si, struct net_device *dev, in
skb = alloc_skb(len+1,GFP_ATOMIC);
if (!skb) {
- printk(KERN_ERR "pxa_ir: fir out of memory for receive skb\n");
dev->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/sb1000.c b/drivers/net/sb1000.c
index 66c2f1a..573a99f 100644
--- a/drivers/net/sb1000.c
+++ b/drivers/net/sb1000.c
@@ -835,9 +835,6 @@ printk("cm0: IP identification: %02x%02x fragment offset: %02x%02x\n", buffer[3
/* compute size to allocate for datagram */
skbsize = dlen + FrameSize;
if ((skb = alloc_skb(skbsize, GFP_ATOMIC)) == NULL) {
- if (sb1000_debug > 1)
- printk(KERN_WARNING "%s: can't allocate %d bytes long "
- "skbuff\n", dev->name, skbsize);
stats->rx_dropped++;
insw(ioaddr, buffer, NewDatagramDataSize / 2);
goto dropped_frame;
diff --git a/drivers/net/slip/slip.c b/drivers/net/slip/slip.c
index a34d6bf..eb1da6c 100644
--- a/drivers/net/slip/slip.c
+++ b/drivers/net/slip/slip.c
@@ -358,7 +358,6 @@ static void sl_bump(struct slip *sl)
skb = dev_alloc_skb(count);
if (skb == NULL) {
- printk(KERN_WARNING "%s: memory squeeze, dropping packet.\n", dev->name);
dev->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/usb/ipheth.c b/drivers/net/usb/ipheth.c
index 534d8be..74138bc 100644
--- a/drivers/net/usb/ipheth.c
+++ b/drivers/net/usb/ipheth.c
@@ -233,8 +233,6 @@ static void ipheth_rcvbulk_callback(struct urb *urb)
skb = dev_alloc_skb(len);
if (!skb) {
- dev_err(&dev->intf->dev, "%s: dev_alloc_skb: -ENOMEM\n",
- __func__);
dev->net->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/wan/cosa.c b/drivers/net/wan/cosa.c
index 6aed238..827c984 100644
--- a/drivers/net/wan/cosa.c
+++ b/drivers/net/wan/cosa.c
@@ -733,7 +733,6 @@ static char *cosa_net_setup_rx(struct channel_data *chan, int size)
kfree_skb(chan->rx_skb);
chan->rx_skb = dev_alloc_skb(size);
if (chan->rx_skb == NULL) {
- pr_notice("%s: Memory squeeze, dropping packet\n", chan->name);
chan->netdev->stats.rx_dropped++;
return NULL;
}
diff --git a/drivers/net/wan/sdla.c b/drivers/net/wan/sdla.c
index de3bbf4..400bdbb 100644
--- a/drivers/net/wan/sdla.c
+++ b/drivers/net/wan/sdla.c
@@ -827,7 +827,6 @@ static void sdla_receive(struct net_device *dev)
skb = dev_alloc_skb(len + sizeof(struct frhdr));
if (skb == NULL)
{
- netdev_notice(dev, "Memory squeeze, dropping packet\n");
dev->stats.rx_dropped++;
success = 0;
}
diff --git a/drivers/net/wan/x25_asy.c b/drivers/net/wan/x25_asy.c
index 44db8b7..6a6cb97 100644
--- a/drivers/net/wan/x25_asy.c
+++ b/drivers/net/wan/x25_asy.c
@@ -197,7 +197,6 @@ static void x25_asy_bump(struct x25_asy *sl)
skb = dev_alloc_skb(count+1);
if (skb == NULL) {
- netdev_warn(sl->dev, "memory squeeze, dropping packet\n");
dev->stats.rx_dropped++;
return;
}
diff --git a/drivers/net/wan/z85230.c b/drivers/net/wan/z85230.c
index feacc3b..ccc57e9 100644
--- a/drivers/net/wan/z85230.c
+++ b/drivers/net/wan/z85230.c
@@ -1639,7 +1639,6 @@ static void z8530_rx_done(struct z8530_channel *c)
skb = dev_alloc_skb(ct);
if (skb == NULL) {
c->netdevice->stats.rx_dropped++;
- netdev_warn(c->netdevice, "Memory squeeze\n");
} else {
skb_put(skb, ct);
skb_copy_to_linear_data(skb, rxb, ct);
diff --git a/drivers/net/wimax/i2400m/netdev.c b/drivers/net/wimax/i2400m/netdev.c
index cedd4d3..7318c4f 100644
--- a/drivers/net/wimax/i2400m/netdev.c
+++ b/drivers/net/wimax/i2400m/netdev.c
@@ -501,7 +501,6 @@ void i2400m_net_rx(struct i2400m *i2400m, struct sk_buff *skb_rx,
* comments at the top of the file */
skb = __netdev_alloc_skb(net_dev, buf_len, GFP_KERNEL);
if (skb == NULL) {
- dev_err(dev, "NETRX: no memory to realloc skb\n");
net_dev->stats.rx_dropped++;
goto error_skb_realloc;
}
diff --git a/drivers/net/wireless/ray_cs.c b/drivers/net/wireless/ray_cs.c
index 598ca1c..68b8353 100644
--- a/drivers/net/wireless/ray_cs.c
+++ b/drivers/net/wireless/ray_cs.c
@@ -2172,7 +2172,6 @@ static void rx_data(struct net_device *dev, struct rcs __iomem *prcs,
skb = dev_alloc_skb(total_len + 5);
if (skb == NULL) {
- pr_debug("ray_cs rx_data could not allocate skb\n");
local->stats.rx_dropped++;
if (readb(&prcs->var.rx_packet.next_frag_rcs_index) != 0xFF)
release_frag_chain(local, prcs);
diff --git a/drivers/net/wireless/wl3501_cs.c b/drivers/net/wireless/wl3501_cs.c
index 730186d..f36aacb 100644
--- a/drivers/net/wireless/wl3501_cs.c
+++ b/drivers/net/wireless/wl3501_cs.c
@@ -964,8 +964,6 @@ static inline void wl3501_md_ind_interrupt(struct net_device *dev,
skb = dev_alloc_skb(pkt_len + 5);
if (!skb) {
- printk(KERN_WARNING "%s: Can't alloc a sk_buff of size %d.\n",
- dev->name, pkt_len);
dev->stats.rx_dropped++;
} else {
skb->dev = dev;
^ permalink raw reply related
* [Patch net-next] net: add RCU annotation to sk_dst_cache field
From: Cong Wang @ 2013-01-23 4:43 UTC (permalink / raw)
To: netdev; +Cc: David S. Miller, Cong Wang
From: Cong Wang <amwang@redhat.com>
sock->sk_dst_cache is protected by RCU.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <amwang@redhat.com>
---
diff --git a/include/net/sock.h b/include/net/sock.h
index 5a34e2f..c4b64c6 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -337,7 +337,7 @@ struct sock {
#endif
unsigned long sk_flags;
struct dst_entry *sk_rx_dst;
- struct dst_entry *sk_dst_cache;
+ struct dst_entry __rcu *sk_dst_cache;
spinlock_t sk_dst_lock;
atomic_t sk_wmem_alloc;
atomic_t sk_omem_alloc;
^ permalink raw reply related
* linux-next: manual merge of the staging tree with the net-next tree
From: Stephen Rothwell @ 2013-01-23 4:45 UTC (permalink / raw)
To: Greg KH
Cc: linux-next, linux-kernel, Florian Fainelli, David Miller, netdev,
Mark Einon
[-- Attachment #1: Type: text/plain, Size: 598 bytes --]
Hi Greg,
Today's linux-next merge of the staging tree got a conflict in
drivers/staging/et131x/et131x.c between commit f9a8f83b04e0 ("net: phy:
remove flags argument from phy_{attach, connect, connect_direct}") from
the net-next tree and commit 26ef1021c1c7 ("staging: et131x: Modify block
comments to fit with networking style") from the staging tree.
I fixed it up (the former adjusted the white space in the same was as the
latter, so I used thet) and can carry the fix as necessary (no action is
required).
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]
^ permalink raw reply
* [PATCH 2/2 net-next] doc: dt: fsl: fec: add napi optional properties
From: Frank Li @ 2013-01-23 4:12 UTC (permalink / raw)
To: lznuaa, shawn.guo, B38611, davem, linux-arm-kernel, netdev,
bhutchings
Cc: s.hauer, Frank Li
Signed-off-by: Frank Li <Frank.Li@freescale.com>
---
Documentation/devicetree/bindings/net/fsl-fec.txt | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/Documentation/devicetree/bindings/net/fsl-fec.txt b/Documentation/devicetree/bindings/net/fsl-fec.txt
index d536392..4a94090 100644
--- a/Documentation/devicetree/bindings/net/fsl-fec.txt
+++ b/Documentation/devicetree/bindings/net/fsl-fec.txt
@@ -15,6 +15,8 @@ Optional properties:
only if property "phy-reset-gpios" is available. Missing the property
will have the duration be 1 millisecond. Numbers greater than 1000 are
invalid and 1 millisecond will be used instead.
+- disable_napi : disable napi support
+- napi_weight : napi weight number, default value is 64
Example:
--
1.7.1
^ permalink raw reply related
* [PATCH 1/2 net-next] net: fec: add napi support to improve proformance
From: Frank Li @ 2013-01-23 4:12 UTC (permalink / raw)
To: lznuaa, shawn.guo, B38611, davem, linux-arm-kernel, netdev,
bhutchings
Cc: s.hauer, Frank Li
Add napi support
Before this patch
iperf -s -i 1
------------------------------------------------------------
Server listening on TCP port 5001
TCP window size: 85.3 KByte (default)
------------------------------------------------------------
[ 4] local 10.192.242.153 port 5001 connected with 10.192.242.138 port 50004
[ ID] Interval Transfer Bandwidth
[ 4] 0.0- 1.0 sec 41.2 MBytes 345 Mbits/sec
[ 4] 1.0- 2.0 sec 43.7 MBytes 367 Mbits/sec
[ 4] 2.0- 3.0 sec 42.8 MBytes 359 Mbits/sec
[ 4] 3.0- 4.0 sec 43.7 MBytes 367 Mbits/sec
[ 4] 4.0- 5.0 sec 42.7 MBytes 359 Mbits/sec
[ 4] 5.0- 6.0 sec 43.8 MBytes 367 Mbits/sec
[ 4] 6.0- 7.0 sec 43.0 MBytes 361 Mbits/sec
After this patch
[ 4] 2.0- 3.0 sec 51.6 MBytes 433 Mbits/sec
[ 4] 3.0- 4.0 sec 51.8 MBytes 435 Mbits/sec
[ 4] 4.0- 5.0 sec 52.2 MBytes 438 Mbits/sec
[ 4] 5.0- 6.0 sec 52.1 MBytes 437 Mbits/sec
[ 4] 6.0- 7.0 sec 52.1 MBytes 437 Mbits/sec
[ 4] 7.0- 8.0 sec 52.3 MBytes 439 Mbits/sec
Signed-off-by: Frank Li <Frank.Li@freescale.com>
Signed-off-by: Fugang Duan <B38611@freescale.com>
---
drivers/net/ethernet/freescale/fec.c | 68 ++++++++++++++++++++++++++++++---
drivers/net/ethernet/freescale/fec.h | 4 ++
2 files changed, 65 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/freescale/fec.c b/drivers/net/ethernet/freescale/fec.c
index f52ba33..8fa420c 100644
--- a/drivers/net/ethernet/freescale/fec.c
+++ b/drivers/net/ethernet/freescale/fec.c
@@ -67,6 +67,8 @@
#endif
#define DRIVER_NAME "fec"
+#define FEC_NAPI_WEIGHT 64
+#define INT32_MAX 0x7FFFFFFF
/* Pause frame feild and FIFO threshold */
#define FEC_ENET_FCE (1 << 5)
@@ -565,6 +567,20 @@ fec_timeout(struct net_device *ndev)
}
static void
+fec_enet_rx_int_is_enabled(struct net_device *ndev, bool enabled)
+{
+ struct fec_enet_private *fep = netdev_priv(ndev);
+ uint int_events;
+
+ int_events = readl(fep->hwp + FEC_IMASK);
+ if (enabled)
+ int_events |= FEC_ENET_RXF;
+ else
+ int_events &= ~FEC_ENET_RXF;
+ writel(int_events, fep->hwp + FEC_IMASK);
+}
+
+static void
fec_enet_tx(struct net_device *ndev)
{
struct fec_enet_private *fep;
@@ -656,8 +672,8 @@ fec_enet_tx(struct net_device *ndev)
* not been given to the system, we just set the empty indicator,
* effectively tossing the packet.
*/
-static void
-fec_enet_rx(struct net_device *ndev)
+static int
+fec_enet_rx(struct net_device *ndev, int budget)
{
struct fec_enet_private *fep = netdev_priv(ndev);
const struct platform_device_id *id_entry =
@@ -667,13 +683,12 @@ fec_enet_rx(struct net_device *ndev)
struct sk_buff *skb;
ushort pkt_len;
__u8 *data;
+ int pkt_received = 0;
#ifdef CONFIG_M532x
flush_cache_all();
#endif
- spin_lock(&fep->hw_lock);
-
/* First, grab all of the stats for the incoming packet.
* These get messed up if we get called due to a busy condition.
*/
@@ -681,6 +696,10 @@ fec_enet_rx(struct net_device *ndev)
while (!((status = bdp->cbd_sc) & BD_ENET_RX_EMPTY)) {
+ if (pkt_received >= budget)
+ break;
+ pkt_received++;
+
/* Since we have allocated space to hold a complete frame,
* the last indicator should be set.
*/
@@ -796,7 +815,7 @@ rx_processing_done:
}
fep->cur_rx = bdp;
- spin_unlock(&fep->hw_lock);
+ return pkt_received;
}
static irqreturn_t
@@ -805,6 +824,7 @@ fec_enet_interrupt(int irq, void *dev_id)
struct net_device *ndev = dev_id;
struct fec_enet_private *fep = netdev_priv(ndev);
uint int_events;
+ ulong flags;
irqreturn_t ret = IRQ_NONE;
do {
@@ -813,7 +833,18 @@ fec_enet_interrupt(int irq, void *dev_id)
if (int_events & FEC_ENET_RXF) {
ret = IRQ_HANDLED;
- fec_enet_rx(ndev);
+ spin_lock_irqsave(&fep->hw_lock, flags);
+
+ if (fep->use_napi) {
+ /* Disable the RX interrupt */
+ if (napi_schedule_prep(&fep->napi)) {
+ fec_enet_rx_int_is_enabled(ndev, false);
+ __napi_schedule(&fep->napi);
+ }
+ } else
+ fec_enet_rx(ndev, INT32_MAX);
+
+ spin_unlock_irqrestore(&fep->hw_lock, flags);
}
/* Transmit OK, or non-fatal error. Update the buffer
@@ -834,7 +865,16 @@ fec_enet_interrupt(int irq, void *dev_id)
return ret;
}
-
+static int fec_enet_rx_napi(struct napi_struct *napi, int budget)
+{
+ struct net_device *ndev = napi->dev;
+ int pkgs = fec_enet_rx(ndev, budget);
+ if (pkgs < budget) {
+ napi_complete(napi);
+ fec_enet_rx_int_is_enabled(ndev, true);
+ }
+ return pkgs;
+}
/* ------------------------------------------------------------------------- */
static void fec_get_mac(struct net_device *ndev)
@@ -1392,6 +1432,9 @@ fec_enet_open(struct net_device *ndev)
struct fec_enet_private *fep = netdev_priv(ndev);
int ret;
+ if (fep->use_napi)
+ napi_enable(&fep->napi);
+
/* I should reset the ring buffers here, but I don't yet know
* a simple way to do that.
*/
@@ -1604,6 +1647,11 @@ static int fec_enet_init(struct net_device *ndev)
ndev->netdev_ops = &fec_netdev_ops;
ndev->ethtool_ops = &fec_enet_ethtool_ops;
+ if (fep->use_napi) {
+ fec_enet_rx_int_is_enabled(ndev, false);
+ netif_napi_add(ndev, &fep->napi, fec_enet_rx_napi, fep->napi_weight);
+ }
+
/* Initialize the receive buffer descriptors. */
bdp = fep->rx_bd_base;
for (i = 0; i < RX_RING_SIZE; i++) {
@@ -1698,6 +1746,7 @@ fec_probe(struct platform_device *pdev)
static int dev_id;
struct pinctrl *pinctrl;
struct regulator *reg_phy;
+ struct device_node *np = pdev->dev.of_node;
of_id = of_match_device(fec_dt_ids, &pdev->dev);
if (of_id)
@@ -1811,6 +1860,11 @@ fec_probe(struct platform_device *pdev)
}
}
+ fep->use_napi = !of_property_read_bool(np, "disable_napi");
+
+ if (of_property_read_u32(np, "napi_weight", &fep->napi_weight))
+ fep->napi_weight = FEC_NAPI_WEIGHT; /*using default value*/
+
fec_reset_phy(pdev);
ret = fec_enet_init(ndev);
diff --git a/drivers/net/ethernet/freescale/fec.h b/drivers/net/ethernet/freescale/fec.h
index 2ebedaf..31fcdd0 100644
--- a/drivers/net/ethernet/freescale/fec.h
+++ b/drivers/net/ethernet/freescale/fec.h
@@ -249,6 +249,10 @@ struct fec_enet_private {
int bufdesc_ex;
int pause_flag;
+ struct napi_struct napi;
+ int napi_weight;
+ bool use_napi;
+
struct ptp_clock *ptp_clock;
struct ptp_clock_info ptp_caps;
unsigned long last_overflow_check;
--
1.7.1
^ permalink raw reply related
* Re: [Patch net-next] net: add RCU annotation to sk_dst_cache field
From: Eric Dumazet @ 2013-01-23 4:50 UTC (permalink / raw)
To: Cong Wang; +Cc: netdev, David S. Miller
In-Reply-To: <1358916229-19004-1-git-send-email-amwang@redhat.com>
On Wed, 2013-01-23 at 12:43 +0800, Cong Wang wrote:
> From: Cong Wang <amwang@redhat.com>
>
> sock->sk_dst_cache is protected by RCU.
>
> Reported-by: Fengguang Wu <fengguang.wu@intel.com>
> Cc: David S. Miller <davem@davemloft.net>
> Signed-off-by: Cong Wang <amwang@redhat.com>
>
> ---
> diff --git a/include/net/sock.h b/include/net/sock.h
> index 5a34e2f..c4b64c6 100644
> --- a/include/net/sock.h
> +++ b/include/net/sock.h
> @@ -337,7 +337,7 @@ struct sock {
> #endif
> unsigned long sk_flags;
> struct dst_entry *sk_rx_dst;
> - struct dst_entry *sk_dst_cache;
> + struct dst_entry __rcu *sk_dst_cache;
> spinlock_t sk_dst_lock;
> atomic_t sk_wmem_alloc;
> atomic_t sk_omem_alloc;
Thats not a complete patch.
You'll have to fix the errors it brings.
^ permalink raw reply
* Re: linux-next: manual merge of the staging tree with the net-next tree
From: Greg KH @ 2013-01-23 4:52 UTC (permalink / raw)
To: Stephen Rothwell
Cc: linux-next, linux-kernel, Florian Fainelli, David Miller, netdev,
Mark Einon
In-Reply-To: <20130123154525.9f1bd6652f13f2a7a8202bf6@canb.auug.org.au>
On Wed, Jan 23, 2013 at 03:45:25PM +1100, Stephen Rothwell wrote:
> Hi Greg,
>
> Today's linux-next merge of the staging tree got a conflict in
> drivers/staging/et131x/et131x.c between commit f9a8f83b04e0 ("net: phy:
> remove flags argument from phy_{attach, connect, connect_direct}") from
> the net-next tree and commit 26ef1021c1c7 ("staging: et131x: Modify block
> comments to fit with networking style") from the staging tree.
>
> I fixed it up (the former adjusted the white space in the same was as the
> latter, so I used thet) and can carry the fix as necessary (no action is
> required).
That sounds good to me, thanks.
greg k-h
^ permalink raw reply
* Re: [Patch net-next] net: add RCU annotation to sk_dst_cache field
From: Cong Wang @ 2013-01-23 4:58 UTC (permalink / raw)
To: Eric Dumazet; +Cc: netdev, David S. Miller
In-Reply-To: <1358916659.12374.684.camel@edumazet-glaptop>
On Tue, 2013-01-22 at 20:50 -0800, Eric Dumazet wrote:
> On Wed, 2013-01-23 at 12:43 +0800, Cong Wang wrote:
> > From: Cong Wang <amwang@redhat.com>
> >
> > sock->sk_dst_cache is protected by RCU.
> >
> > Reported-by: Fengguang Wu <fengguang.wu@intel.com>
> > Cc: David S. Miller <davem@davemloft.net>
> > Signed-off-by: Cong Wang <amwang@redhat.com>
> >
> > ---
> > diff --git a/include/net/sock.h b/include/net/sock.h
> > index 5a34e2f..c4b64c6 100644
> > --- a/include/net/sock.h
> > +++ b/include/net/sock.h
> > @@ -337,7 +337,7 @@ struct sock {
> > #endif
> > unsigned long sk_flags;
> > struct dst_entry *sk_rx_dst;
> > - struct dst_entry *sk_dst_cache;
> > + struct dst_entry __rcu *sk_dst_cache;
> > spinlock_t sk_dst_lock;
> > atomic_t sk_wmem_alloc;
> > atomic_t sk_omem_alloc;
>
> Thats not a complete patch.
>
> You'll have to fix the errors it brings.
It does fix this warning:
net/core/flow_dissector.c:280:37: sparse: incompatible types in
comparison expression (different address spaces)
^ permalink raw reply
* Re: [PATCH 2/2] IP_GRE: Linearize skb before csum.
From: Eric Dumazet @ 2013-01-23 5:02 UTC (permalink / raw)
To: David Miller; +Cc: jesse, pshelar, netdev
In-Reply-To: <20130122.232235.643904201225800004.davem@davemloft.net>
On Tue, 2013-01-22 at 23:22 -0500, David Miller wrote:
> From: David Miller <davem@davemloft.net>
> Date: Tue, 22 Jan 2013 20:46:29 -0500 (EST)
>
> > We can't emit bogus checksums on the wire when this is completely
> > under our control.
>
> BTW consider a legitimate use case where an application has a shared
> memory page between threads that keeps track of statistic counters,
> and it uses sendfile() over an fd backing that mmap()'d area to
> send the statistics out remotely over TCP.
>
> That's completely legitimate.
>
> As is an application constantly overwriting a page in the filesystem
> page cache, and another app doing sendfile() over TCP from that
> area.
>
> In such a situation everyone of your magic retransmits will have a
> bad checksum too.
>
> You can't say that this behavior is anything other than broken, and
> I don't want to be responsible for a system that's doing broken
> things like that.
>
> If your GSO over GRE tunnels stuff fundamentally depends upon this
> magic retransmit fantasy working, then yes we have to disable it
> completely.
> --
But then GSO should be disabled right now, GRE or not.
We have a generic problem.
GSO was a bad idea from the very beginning, if you want
such use case.
^ permalink raw reply
* Re: [Patch net-next] net: add RCU annotation to sk_dst_cache field
From: Eric Dumazet @ 2013-01-23 5:06 UTC (permalink / raw)
To: Cong Wang; +Cc: netdev, David S. Miller
In-Reply-To: <1358917086.4235.12.camel@cr0>
On Wed, 2013-01-23 at 12:58 +0800, Cong Wang wrote:
> It does fix this warning:
>
> net/core/flow_dissector.c:280:37: sparse: incompatible types in
> comparison expression (different address spaces)
Thats not a good enough reason.
It adds 10 new warnings.
^ permalink raw reply
* RE: [PATCH] skb: add a comment to skb_csum_unnecessary to avoid miuse
From: Sanagi, Koki @ 2013-01-23 5:11 UTC (permalink / raw)
To: David Miller; +Cc: netdev@vger.kernel.org
In-Reply-To: <20130122.231353.2100388787067184936.davem@davemloft.net>
> From: "Sanagi, Koki" <sanagi.koki@jp.fujitsu.com>
> Date: Wed, 23 Jan 2013 02:32:52 +0000
>
> >> From: Koki Sanagi <sanagi.koki@jp.fujitsu.com>
> >> Date: Wed, 23 Jan 2013 09:30:16 +0900
> >>
> >> > Due to its name and appearance, someone thinks this only checks if
> >> > ip_summed is CHECKSUM_UNNECESARRY. But actually, this returns true
> >> > even if ip_summed is CHECKSUM_PARTIAL. To avoid misuse, this patch
> >> > a comment which specifies that CHECKSUM_PARTIAL is OK.
> >> >
> >> > Signed-off-by: Koki Sanagi <sanagi.koki@jp.fujitsu.com>
> >>
> >> I'm not applying this, sorry. It's a one line function and it's not
> >> so non-obvious that it deserves an 8 line comment.
> >
> > OK. I just felt weird that CHECKSUM_* is not bit flag but this
> > function handles it as if it was bit flag.
>
> The function name says what it does, it determines whether a checksum is
> necessary or not. How that is implemented is another issue.
As for name of the function which confused me, I'm ok now. But as for how to implement,
I still think it is inappropriate because of the above reason. So, how about introducing
CHECKSUM_UNNECESSARY_BIT like below ?
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 8b2256e..b1dbd36 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -39,6 +39,8 @@
#define CHECKSUM_COMPLETE 2
#define CHECKSUM_PARTIAL 3
+#define CHECKSUM_UNNECESSARY_BIT (1 << 0)
+
#define SKB_DATA_ALIGN(X) (((X) + (SMP_CACHE_BYTES - 1)) & \
~(SMP_CACHE_BYTES - 1))
#define SKB_WITH_OVERHEAD(X) \
@@ -2524,7 +2526,7 @@ extern __sum16 __skb_checksum_complete(struct sk_buff *skb);
static inline int skb_csum_unnecessary(const struct sk_buff *skb)
{
- return skb->ip_summed & CHECKSUM_UNNECESSARY;
+ return skb->ip_summed & CHECKSUM_UNNECESSARY_BIT;
}
/**
^ permalink raw reply related
* Re: [PATCH 2/2] IP_GRE: Linearize skb before csum.
From: David Miller @ 2013-01-23 5:19 UTC (permalink / raw)
To: eric.dumazet; +Cc: jesse, pshelar, netdev
In-Reply-To: <1358917373.12374.707.camel@edumazet-glaptop>
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Tue, 22 Jan 2013 21:02:53 -0800
> GSO was a bad idea from the very beginning, if you want
> such use case.
It's perfectly fine when the card checksums the packet. Because it
will checksum the exact data that will hit the wire and thus the
checksum will be correct always.
It's only buggy when we do pure software GSO and have page frags.
^ permalink raw reply
* Re: [PATCH v2] IP_GRE: Fix kernel panic in IP_GRE with GRE csum.
From: David Miller @ 2013-01-23 5:20 UTC (permalink / raw)
To: blp; +Cc: pshelar, netdev, jesse
In-Reply-To: <877gn4iizy.fsf@blp.benpfaff.org>
From: Ben Pfaff <blp@cs.stanford.edu>
Date: Tue, 22 Jan 2013 21:08:33 -0800
> David Miller <davem@davemloft.net> writes:
>
>> From: Pravin B Shelar <pshelar@nicira.com>
>> Date: Tue, 22 Jan 2013 14:56:38 -0800
>>
>>> Fixed code indentation according to comments from David Miller.
>>>
>>> --8<--------------------------cut here-------------------------->8--
>>
>> This doesn't do what you think it does. No automated tool is going
>> to edit the commit message the way you think this will.
>
> "git am" does if you run with --scissors, see git-am(1):
>
> -c, --scissors
> Remove everything in body before a scissors line (see git-
> mailinfo(1)).
I'd rather not have to do that for submissions, thanks.
Furthermore patchwork doesn't parse this convention either.
^ permalink raw reply
* Re: [PATCH] skb: add a comment to skb_csum_unnecessary to avoid miuse
From: David Miller @ 2013-01-23 5:21 UTC (permalink / raw)
To: sanagi.koki; +Cc: netdev
In-Reply-To: <F6CB40852BC6EF489E0AA83EFF537F470876BBCF@G01JPEXMBYT04>
From: "Sanagi, Koki" <sanagi.koki@jp.fujitsu.com>
Date: Wed, 23 Jan 2013 05:11:14 +0000
> As for name of the function which confused me, I'm ok now. But as
> for how to implement, I still think it is inappropriate because of
> the above reason. So, how about introducing
> CHECKSUM_UNNECESSARY_BIT like below ?
If the user has to look at the implementation, they will see
the "&" and there is no confusion.
Please stop splitting hairs, this is a waste of time.
^ permalink raw reply
* Re: [PATCH 2/2] IP_GRE: Linearize skb before csum.
From: Eric Dumazet @ 2013-01-23 5:30 UTC (permalink / raw)
To: David Miller; +Cc: jesse, pshelar, netdev
In-Reply-To: <20130123.001956.1988771856502410155.davem@davemloft.net>
On Wed, 2013-01-23 at 00:19 -0500, David Miller wrote:
> From: Eric Dumazet <eric.dumazet@gmail.com>
> Date: Tue, 22 Jan 2013 21:02:53 -0800
>
> > GSO was a bad idea from the very beginning, if you want
> > such use case.
>
> It's perfectly fine when the card checksums the packet. Because it
> will checksum the exact data that will hit the wire and thus the
> checksum will be correct always.
>
> It's only buggy when we do pure software GSO and have page frags.
Right.
We could add a new SKB_GSO_SHARED_FRAG bit for that, and force an extra
copy of frags, instead of a linearize call.
Basically, splice(pipe -> socket)/sendfile()/vmsplice() should set this
flag.
Its a bit late here, I'll do that tomorrow.
^ permalink raw reply
* Re: [Patch net-next] net: add RCU annotation to sk_dst_cache field
From: Cong Wang @ 2013-01-23 5:30 UTC (permalink / raw)
To: Eric Dumazet; +Cc: netdev, David S. Miller
In-Reply-To: <1358917603.12374.713.camel@edumazet-glaptop>
On Tue, 2013-01-22 at 21:06 -0800, Eric Dumazet wrote:
> On Wed, 2013-01-23 at 12:58 +0800, Cong Wang wrote:
>
> > It does fix this warning:
> >
> > net/core/flow_dissector.c:280:37: sparse: incompatible types in
> > comparison expression (different address spaces)
>
> Thats not a good enough reason.
>
> It adds 10 new warnings.
>
>
I expect most of them are from decnet, which is (partially) deprecated?
^ permalink raw reply
* Re: [Patch net-next] net: add RCU annotation to sk_dst_cache field
From: Eric Dumazet @ 2013-01-23 5:34 UTC (permalink / raw)
To: Cong Wang; +Cc: netdev, David S. Miller
In-Reply-To: <1358919057.4235.17.camel@cr0>
On Wed, 2013-01-23 at 13:30 +0800, Cong Wang wrote:
>
> I expect most of them are from decnet, which is (partially) deprecated?
How many answers to you want exactly from me ?
^ permalink raw reply
* Re: [PATCH 2/2] IP_GRE: Linearize skb before csum.
From: David Miller @ 2013-01-23 5:36 UTC (permalink / raw)
To: eric.dumazet; +Cc: jesse, pshelar, netdev
In-Reply-To: <1358919050.12374.733.camel@edumazet-glaptop>
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Tue, 22 Jan 2013 21:30:50 -0800
> We could add a new SKB_GSO_SHARED_FRAG bit for that, and force an extra
> copy of frags, instead of a linearize call.
>
> Basically, splice(pipe -> socket)/sendfile()/vmsplice() should set this
> flag.
Right.
> Its a bit late here, I'll do that tomorrow.
Thank you.
^ permalink raw reply
* Re: [Patch net-next] net: add RCU annotation to sk_dst_cache field
From: David Miller @ 2013-01-23 5:36 UTC (permalink / raw)
To: amwang; +Cc: eric.dumazet, netdev
In-Reply-To: <1358919057.4235.17.camel@cr0>
From: Cong Wang <amwang@redhat.com>
Date: Wed, 23 Jan 2013 13:30:57 +0800
> I expect most of them are from decnet, which is (partially) deprecated?
decnet is not deprecated, people use and test it, just fix the
warnings, implement your patch properly, and stop arguing.
^ permalink raw reply
* Re: [PATCH v2] IP_GRE: Fix kernel panic in IP_GRE with GRE csum.
From: Ben Pfaff @ 2013-01-23 5:08 UTC (permalink / raw)
To: David Miller; +Cc: pshelar, netdev, jesse
In-Reply-To: <20130122.205507.339713494072219737.davem@davemloft.net>
David Miller <davem@davemloft.net> writes:
> From: Pravin B Shelar <pshelar@nicira.com>
> Date: Tue, 22 Jan 2013 14:56:38 -0800
>
>> Fixed code indentation according to comments from David Miller.
>>
>> --8<--------------------------cut here-------------------------->8--
>
> This doesn't do what you think it does. No automated tool is going
> to edit the commit message the way you think this will.
"git am" does if you run with --scissors, see git-am(1):
-c, --scissors
Remove everything in body before a scissors line (see git-
mailinfo(1)).
and git-mailinfo(1);
--scissors
Remove everything in body before a scissors line. A line that
mainly consists of scissors (either ">8" or "8<") and perforation
(dash "-") marks is called a scissors line, and is used to request
the reader to cut the message at that line. If such a line appears
in the body of the message before the patch, everything before it
(including the scissors line itself) is ignored when this option is
used.
This is useful if you want to begin your message in a discussion
thread with comments and suggestions on the message you are
responding to, and to conclude it with a patch submission,
separating the discussion and the beginning of the proposed commit
log message with a scissors line.
This can enabled by default with the configuration option
mailinfo.scissors.
^ permalink raw reply
* Re: 3.7.3+: Bad paging request in ip_rcv_finish while running NFS traffic.
From: Eric Dumazet @ 2013-01-23 6:11 UTC (permalink / raw)
To: Ben Greear; +Cc: netdev, linux-nfs@vger.kernel.org
In-Reply-To: <50FF4BC9.1060206@candelatech.com>
On Tue, 2013-01-22 at 18:32 -0800, Ben Greear wrote:
> diff --git a/net/core/dst.c b/net/core/dst.c
> index ee6153e..234b168 100644
> --- a/net/core/dst.c
> +++ b/net/core/dst.c
> @@ -245,6 +245,7 @@ again:
> dst->ops->destroy(dst);
> if (dst->dev)
> dev_put(dst->dev);
> + dst->input = dst->output = 0xdeadbeef;
> kmem_cache_free(dst->ops->kmem_cachep, dst);
Great !
You could comment the kmem_cache_free() as well to get better chances to
hit the bug, and maybe start a bisection to find the faulty commit ?
^ permalink raw reply
* Re: v3 for tcp friends?
From: Li Yu @ 2013-01-23 6:12 UTC (permalink / raw)
To: Bruce Curtis; +Cc: David Miller, netdev
In-Reply-To: <50FF5592.60008@gmail.com>
Oops, this hang is not since TCP friends patch!
sk_sndbuf_get() is broken by 32 bits integer overflow
because of so large value in net.ipv4.tcp_{rmem,wmem}.
but this hang also can be found in net-next.git
(3.8.0-rc3+), if we run below commands, then all new
TCP connections stop working!
# when TCP friends is disabled
sysctl -w net.ipv4.tcp_rmem="4096 4294967296 4294967296" # 4GB
sysctl -w net.ipv4.tcp_wmem="4096 4294967296 4294967296"
Thanks
Yu
于 2013年01月23日 11:14, Li Yu 写道:
> 于 2013年01月23日 05:08, Bruce Curtis 写道:
>> Thanks, Li
>>
>> Started working on friends again, v4, more soon.
>>
>>
>
> :)
>
> I found another odd bug in TCP friends v3, the clients
> may hang at tcp_sendmsg() -> sk_stream_wait_memory() with or
> without my refcnt fix patch.
>
> Below shell script can reproduce this bug:
>
> #! /bin/sh -x
>
> sysctl -w net.ipv4.tcp_rmem="4096 1073741824 1073741824"
> sysctl -w net.ipv4.tcp_wmem="4096 1073741824 1073741824"
>
> sysctl -w net.ipv4.tcp_friends=1
>
> msg=64K
> buf=256M
>
> pkill -9 netserver
> netserver
> netperf -t TCP_STREAM -l 1 -- -m ${msg} -M ${msg} -s ${buf} -S ${buf} -4
>
> sysctl -w net.ipv4.tcp_friends=0
> pkill -9 netserver
> netserver
> netperf -t TCP_STREAM -l 1 -- -m ${msg} -M ${msg} -s ${buf} -S ${buf} -4
> ##################SCRIPT END###################
>
> netperf kernel stack is (by cat /proc/$netperf_pid/stack)
>
> [<ffffffff812ce939>] sk_stream_wait_memory+0x2d9/0x2f0
> [<ffffffff8131460c>] tcp_sendmsg+0xf6c/0x1240
> [<ffffffff8133c117>] inet_sendmsg+0xf7/0x110
> [<ffffffff812bedfd>] sock_sendmsg+0x7d/0xa0
> [<ffffffff812c0e4d>] sys_sendto+0x13d/0x190
> [<ffffffff8138a6c2>] system_call_fastpath+0x16/0x1b
> [<ffffffffffffffff>] 0xffffffffffffffff
>
> netserver kernel stack is :
>
> [<ffffffff812c46ae>] sk_wait_data+0x8e/0xe0
> [<ffffffff81315993>] tcp_recvmsg+0x5c3/0xbe0
> [<ffffffff8133aefd>] inet_recvmsg+0xed/0x110
> [<ffffffff812becf4>] sock_recvmsg+0x84/0xb0
> [<ffffffff812c0fae>] sys_recvfrom+0xee/0x170
> [<ffffffff8138a6c2>] system_call_fastpath+0x16/0x1b
> [<ffffffffffffffff>] 0xffffffffffffffff
>
> And, "netstat -tnp" give us below results:
>
> Active Internet connections (w/o servers)
> Proto Recv-Q Send-Q Local Address Foreign Address
> State PID/Program name
> tcp 0 0 127.0.0.1:35451 127.0.0.1:12865
> ESTABLISHED 2316/netperf
> tcp6 0 0 127.0.0.1:12865 127.0.0.1:35451
> ESTABLISHED 2317/netserver
>
> (It seems that netperf hangs on the control connection of benchmark)
>
> I also try to fix this ...
>
> Thanks
>
> Yu
>
>> On Mon, Jan 21, 2013 at 12:55 AM, Li Yu <raise.sail@gmail.com
>> <mailto:raise.sail@gmail.com>> wrote:
>>
>> 2013/1/21 Li Yu <raise.sail@gmail.com <mailto:raise.sail@gmail.com>>
>>
>> 于 2013年01月21日 15:29, Li Yu 写道:
>>
>> 于 2012年09月05日 00:58, David Miller 写道:
>>
>> From: Bruce Curtis <brutus@google.com
>> <mailto:brutus@google.com>>
>> Date: Tue, 4 Sep 2012 08:10:23 -0700
>>
>> Will do, issues addressed, I'll get the patch out
>> later today or
>> tomorrow at the latest.
>>
>>
>> Thanks a lot Bruce.
>> --
>> To unsubscribe from this list: send the line
>> "unsubscribe netdev" in
>> the body of a message to majordomo@vger.kernel.org
>> <mailto:majordomo@vger.kernel.org>
>> More majordomo info at
>> http://vger.kernel.org/__majordomo-info.html
>> <http://vger.kernel.org/majordomo-info.html>
>>
>>
>>
>> Hi, Bruce,
>>
>> I tested the TCP friends, found a bug here:
>>
>> [ 106.541372] Pid: 1765, comm: client Not tainted
>> 3.7.0-rc1+ #25
>> [ 106.543121] Call Trace:
>> [ 106.543950] [<ffffffff8133d212>]
>> inet_sock_destruct+0x102/0x1f0
>> [ 106.545687] [<ffffffff812c38ad>] __sk_free+0x1d/0x110
>> [ 106.547209] [<ffffffff812c3a1c>] sk_free+0x1c/0x20
>> [ 106.548611] [<ffffffff8131680c>] tcp_close+0x6c/0x3f0
>> [ 106.549863] [<ffffffff8133caea>] inet_release+0xda/0xf0
>> [ 106.551134] [<ffffffff8133ca30>] ? inet_release+0x20/0xf0
>> [ 106.552419] [<ffffffff8137f3de>] ? mutex_unlock+0xe/0x10
>> [ 106.553658] [<ffffffff812bf948>] sock_release+0x28/0xa0
>> [ 106.557366] [<ffffffff812bfd69>] sock_close+0x29/0x30
>> [ 106.558831] [<ffffffff81128972>] __fput+0x122/0x210
>> [ 106.560541] [<ffffffff81128a6e>] ____fput+0xe/0x10
>> [ 106.562006] [<ffffffff8105354e>] task_work_run+0x9e/0xd0
>> [ 106.563285] [<ffffffff810027e1>] do_notify_resume+0x61/0x70
>> [ 106.564582] [<ffffffff8138a908>] int_signal+0x12/0x17
>>
>>
>> I also backported and tested it on stable kernel
>> 3.7.3/RHEL6
>> kernel, this bug still exists. It seem that client may close
>> listening
>> sockets, may we need to add one reference count for listen
>> socket
>> before send its address to peer?
>>
>>
>> Sorry, I lost an important line of kernel log before above them:
>>
>> [ 106.539367] IPv4: Attempt to release TCP socket in state 10
>> ffff880074abb5c0
>>
>> BTW: state 10 = TCP_LISTEN
>>
>>
>> It seem this patch works for me.
>>
>> diff --git a/net/ipv4/inet_connection_sock.c
>> b/net/ipv4/inet_connection_sock.c
>> index 9641215..a625c02 100644
>> --- a/net/ipv4/inet_connection_sock.c
>> +++ b/net/ipv4/inet_connection_sock.c
>> @@ -623,8 +623,11 @@ struct sock *inet_csk_clone(struct sock *sk,
>> const struct request_sock *req,
>> sock_hold(newsk);
>> was = xchg(&req->friend->sk_friend, newsk);
>> /* If requester already connect()ed, maybe
>> sleeping */
>> - if (was && !sock_flag(req->friend, SOCK_DEAD))
>> - sk->sk_state_change(req->friend);
>> + if (was) {
>> + if (!sock_flag(req->friend, SOCK_DEAD))
>> +
>> sk->sk_state_change(req->friend);
>> + sock_put(was);
>> + }
>> }
>> newsk->sk_state = TCP_SYN_RECV;
>> diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
>> index 5917485..7a63245 100644
>> --- a/net/ipv4/tcp_output.c
>> +++ b/net/ipv4/tcp_output.c
>> @@ -2277,8 +2277,10 @@ struct sk_buff *tcp_make_synack(struct sock
>> *sk, struct dst_entry *dst,
>> memset(&opts, 0, sizeof(opts));
>> /* Only try to make friends if enabled */
>> - if (sysctl_tcp_friends)
>> + if (sysctl_tcp_friends) {
>> + sock_hold(sk);
>> skb->friend = sk;
>> + }
>> #ifdef CONFIG_SYN_COOKIES
>> if (unlikely(req->cookie_ts))
>>
>>
>> And, our TCP friends v4? :)
>>
>> Thanks
>>
>> Yu
>>
>>
>>
>
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox