* Re: [PATCH] bonding: bond_select_queue off by one
From: Jay Vosburgh @ 2011-02-18 4:41 UTC (permalink / raw)
To: Andy Gospodarek, Phil Oester; +Cc: netdev
In-Reply-To: <20110218020713.GA9696@linuxace.com>
Phil Oester <kernel@linuxace.com> wrote:
>The bonding driver's bond_select_queue function simply returns
>skb->queue_mapping. However queue_mapping could be == 16
>for queue #16. This causes the following message to be flooded
>to syslog:
>
>kernel: bondx selects TX queue 16, but real number of TX queues is 16
>
>ndo_select_queue wants a zero-based number, so bonding driver needs
>to subtract one to return the proper queue number. Also fix grammar in
>a comment while in the vicinity.
Andy, can you comment on this?
If memory serves, the omission of queue ID zero was on purpose;
is this patch going to break any of the functionality added by:
commit bb1d912323d5dd50e1079e389f4e964be14f0ae3
Author: Andy Gospodarek <andy@greyhouse.net>
Date: Wed Jun 2 08:40:18 2010 +0000
bonding: allow user-controlled output slave selection
Ben Hutchings <bhutchings@solarflare.com> wrote:
>This looks basically correct, but it should use the proper functions:
>
> skb_rx_queue_recorded(skb) ? skb_get_rx_queue(skb) : 0;
As Ben points out, skb_rx_queue_recorded, skb_record_rx_queue,
et al, do the offset by one internally, but the bond_slave_override
function is comparing the slave's queue_id to the skb->queue_mapping.
That makes me wonder if this patch is going to mess things up,
and if bond_slave_override should also use the skb_rx_queue_recorded, et
al, functions.
-J
>Phil Oester
>
>Signed-off-by: Phil Oester <kernel@linuxace.com>
>
>
>--- linux-2.6/drivers/net/bonding/bond_main.c.orig 2011-01-30 09:15:09.813843817 -0800
>+++ linux-2.6/drivers/net/bonding/bond_main.c 2011-02-17 18:02:46.919050909 -0800
>@@ -4537,11 +4537,11 @@
> {
> /*
> * This helper function exists to help dev_pick_tx get the correct
>- * destination queue. Using a helper function skips the a call to
>+ * destination queue. Using a helper function skips a call to
> * skb_tx_hash and will put the skbs in the queue we expect on their
> * way down to the bonding driver.
> */
>- return skb->queue_mapping;
>+ return skb->queue_mapping ? skb->queue_mapping - 1 : 0;
> }
>
> static netdev_tx_t bond_start_xmit(struct sk_buff *skb, struct net_device *dev)
---
-Jay Vosburgh, IBM Linux Technology Center, fubar@us.ibm.com
^ permalink raw reply
* [PATCH 0/2] Add ipv4 if addr hash.
From: David Miller @ 2011-02-18 4:46 UTC (permalink / raw)
To: netdev
These changes are based upon something I brought up the other day,
namely that we do a full FIB lookup just to find whether a particular
unicast ipv4 address is attached to an interface.
Like ipv6, I added an RCU/RTNL protected hash table and reimplemented
__ip_dev_find() in terms of it.
These patches apply both with and without the route cache deletion
patch applied, but with the route cache removal applied the udpflood
test gets faster by ~13 seconds:
real 2m57.500s
user 0m10.640s
sys 2m46.910s
We can probably use this new table for other things as well.
^ permalink raw reply
* [PATCH 1/2] ipv4: Add hash table of interface addresses.
From: David Miller @ 2011-02-18 4:46 UTC (permalink / raw)
To: netdev
This will be used to optimize __ip_dev_find() and friends.
Signed-off-by: David S. Miller <davem@davemloft.net>
---
include/linux/inetdevice.h | 1 +
net/ipv4/devinet.c | 44 ++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 45 insertions(+), 0 deletions(-)
diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h
index ae8fdc5..5f81466 100644
--- a/include/linux/inetdevice.h
+++ b/include/linux/inetdevice.h
@@ -144,6 +144,7 @@ static inline void ipv4_devconf_setall(struct in_device *in_dev)
#define IN_DEV_ARP_NOTIFY(in_dev) IN_DEV_MAXCONF((in_dev), ARP_NOTIFY)
struct in_ifaddr {
+ struct hlist_node hash;
struct in_ifaddr *ifa_next;
struct in_device *ifa_dev;
struct rcu_head rcu_head;
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 748cb5b..c1f2552 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -92,6 +92,38 @@ static const struct nla_policy ifa_ipv4_policy[IFA_MAX+1] = {
[IFA_LABEL] = { .type = NLA_STRING, .len = IFNAMSIZ - 1 },
};
+/* inet_addr_hash's shifting is dependent upon this IN4_ADDR_HSIZE
+ * value. So if you change this define, make appropriate changes to
+ * inet_addr_hash as well.
+ */
+#define IN4_ADDR_HSIZE 256
+static struct hlist_head inet_addr_lst[IN4_ADDR_HSIZE];
+static DEFINE_SPINLOCK(inet_addr_hash_lock);
+
+static inline unsigned int inet_addr_hash(__be32 addr)
+{
+ u32 val = (__force u32) addr;
+
+ return ((val ^ (val >> 8) ^ (val >> 16) ^ (val >> 24)) &
+ (IN4_ADDR_HSIZE - 1));
+}
+
+static void inet_hash_insert(struct in_ifaddr *ifa)
+{
+ unsigned int hash = inet_addr_hash(ifa->ifa_address);
+
+ spin_lock(&inet_addr_hash_lock);
+ hlist_add_head_rcu(&ifa->hash, &inet_addr_lst[hash]);
+ spin_unlock(&inet_addr_hash_lock);
+}
+
+static void inet_hash_remove(struct in_ifaddr *ifa)
+{
+ spin_lock(&inet_addr_hash_lock);
+ hlist_del_init_rcu(&ifa->hash);
+ spin_unlock(&inet_addr_hash_lock);
+}
+
static void rtmsg_ifa(int event, struct in_ifaddr *, struct nlmsghdr *, u32);
static BLOCKING_NOTIFIER_HEAD(inetaddr_chain);
@@ -265,6 +297,7 @@ static void __inet_del_ifa(struct in_device *in_dev, struct in_ifaddr **ifap,
}
if (!do_promote) {
+ inet_hash_remove(ifa);
*ifap1 = ifa->ifa_next;
rtmsg_ifa(RTM_DELADDR, ifa, nlh, pid);
@@ -281,6 +314,7 @@ static void __inet_del_ifa(struct in_device *in_dev, struct in_ifaddr **ifap,
/* 2. Unlink it */
*ifap = ifa1->ifa_next;
+ inet_hash_remove(ifa1);
/* 3. Announce address deletion */
@@ -368,6 +402,8 @@ static int __inet_insert_ifa(struct in_ifaddr *ifa, struct nlmsghdr *nlh,
ifa->ifa_next = *ifap;
*ifap = ifa;
+ inet_hash_insert(ifa);
+
/* Send message first, then call notifier.
Notifier will trigger FIB update, so that
listeners of netlink will know about new ifaddr */
@@ -521,6 +557,7 @@ static struct in_ifaddr *rtm_to_ifaddr(struct net *net, struct nlmsghdr *nlh)
if (tb[IFA_ADDRESS] == NULL)
tb[IFA_ADDRESS] = tb[IFA_LOCAL];
+ INIT_HLIST_NODE(&ifa->hash);
ifa->ifa_prefixlen = ifm->ifa_prefixlen;
ifa->ifa_mask = inet_make_mask(ifm->ifa_prefixlen);
ifa->ifa_flags = ifm->ifa_flags;
@@ -728,6 +765,7 @@ int devinet_ioctl(struct net *net, unsigned int cmd, void __user *arg)
if (!ifa) {
ret = -ENOBUFS;
ifa = inet_alloc_ifa();
+ INIT_HLIST_NODE(&ifa->hash);
if (!ifa)
break;
if (colon)
@@ -1069,6 +1107,7 @@ static int inetdev_event(struct notifier_block *this, unsigned long event,
struct in_ifaddr *ifa = inet_alloc_ifa();
if (ifa) {
+ INIT_HLIST_NODE(&ifa->hash);
ifa->ifa_local =
ifa->ifa_address = htonl(INADDR_LOOPBACK);
ifa->ifa_prefixlen = 8;
@@ -1710,6 +1749,11 @@ static struct rtnl_af_ops inet_af_ops = {
void __init devinet_init(void)
{
+ int i;
+
+ for (i = 0; i < IN4_ADDR_HSIZE; i++)
+ INIT_HLIST_HEAD(&inet_addr_lst[i]);
+
register_pernet_subsys(&devinet_ops);
register_gifconf(PF_INET, inet_gifconf);
--
1.7.4.1
^ permalink raw reply related
* [PATCH 2/2] ipv4: Implement __ip_dev_find using new interface address hash.
From: David Miller @ 2011-02-18 4:46 UTC (permalink / raw)
To: netdev
Much quicker than going through the FIB tables.
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/ipv4/devinet.c | 33 +++++++++++++++++++++++++++++++++
net/ipv4/fib_frontend.c | 40 ----------------------------------------
2 files changed, 33 insertions(+), 40 deletions(-)
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index c1f2552..1550881 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -124,6 +124,39 @@ static void inet_hash_remove(struct in_ifaddr *ifa)
spin_unlock(&inet_addr_hash_lock);
}
+/**
+ * __ip_dev_find - find the first device with a given source address.
+ * @net: the net namespace
+ * @addr: the source address
+ * @devref: if true, take a reference on the found device
+ *
+ * If a caller uses devref=false, it should be protected by RCU, or RTNL
+ */
+struct net_device *__ip_dev_find(struct net *net, __be32 addr, bool devref)
+{
+ unsigned int hash = inet_addr_hash(addr);
+ struct net_device *result = NULL;
+ struct in_ifaddr *ifa;
+ struct hlist_node *node;
+
+ rcu_read_lock();
+ hlist_for_each_entry_rcu(ifa, node, &inet_addr_lst[hash], hash) {
+ struct net_device *dev = ifa->ifa_dev->dev;
+
+ if (!net_eq(dev_net(dev), net))
+ continue;
+ if (ifa->ifa_address == addr) {
+ result = dev;
+ break;
+ }
+ }
+ if (result && devref)
+ dev_hold(result);
+ rcu_read_unlock();
+ return result;
+}
+EXPORT_SYMBOL(__ip_dev_find);
+
static void rtmsg_ifa(int event, struct in_ifaddr *, struct nlmsghdr *, u32);
static BLOCKING_NOTIFIER_HEAD(inetaddr_chain);
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index 2a49c06..ad0778a 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -132,46 +132,6 @@ static void fib_flush(struct net *net)
rt_cache_flush(net, -1);
}
-/**
- * __ip_dev_find - find the first device with a given source address.
- * @net: the net namespace
- * @addr: the source address
- * @devref: if true, take a reference on the found device
- *
- * If a caller uses devref=false, it should be protected by RCU, or RTNL
- */
-struct net_device *__ip_dev_find(struct net *net, __be32 addr, bool devref)
-{
- struct flowi fl = {
- .fl4_dst = addr,
- };
- struct fib_result res = { 0 };
- struct net_device *dev = NULL;
- struct fib_table *local_table;
-
-#ifdef CONFIG_IP_MULTIPLE_TABLES
- res.r = NULL;
-#endif
-
- rcu_read_lock();
- local_table = fib_get_table(net, RT_TABLE_LOCAL);
- if (!local_table ||
- fib_table_lookup(local_table, &fl, &res, FIB_LOOKUP_NOREF)) {
- rcu_read_unlock();
- return NULL;
- }
- if (res.type != RTN_LOCAL)
- goto out;
- dev = FIB_RES_DEV(res);
-
- if (dev && devref)
- dev_hold(dev);
-out:
- rcu_read_unlock();
- return dev;
-}
-EXPORT_SYMBOL(__ip_dev_find);
-
/*
* Find address type as if only "dev" was present in the system. If
* on_dev is NULL then all interfaces are taken into consideration.
--
1.7.4.1
^ permalink raw reply related
* Re: [PATCH 1/2] ipv4: Add hash table of interface addresses.
From: Eric Dumazet @ 2011-02-18 6:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20110217.204638.104041410.davem@davemloft.net>
Le jeudi 17 février 2011 à 20:46 -0800, David Miller a écrit :
> This will be used to optimize __ip_dev_find() and friends.
>
> Signed-off-by: David S. Miller <davem@davemloft.net>
> ---
> include/linux/inetdevice.h | 1 +
> net/ipv4/devinet.c | 44 ++++++++++++++++++++++++++++++++++++++++++++
> 2 files changed, 45 insertions(+), 0 deletions(-)
>
> diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h
> index ae8fdc5..5f81466 100644
> --- a/include/linux/inetdevice.h
> +++ b/include/linux/inetdevice.h
> @@ -144,6 +144,7 @@ static inline void ipv4_devconf_setall(struct in_device *in_dev)
> #define IN_DEV_ARP_NOTIFY(in_dev) IN_DEV_MAXCONF((in_dev), ARP_NOTIFY)
>
> struct in_ifaddr {
> + struct hlist_node hash;
> struct in_ifaddr *ifa_next;
> struct in_device *ifa_dev;
> struct rcu_head rcu_head;
> diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
> index 748cb5b..c1f2552 100644
> --- a/net/ipv4/devinet.c
> +++ b/net/ipv4/devinet.c
> @@ -92,6 +92,38 @@ static const struct nla_policy ifa_ipv4_policy[IFA_MAX+1] = {
> [IFA_LABEL] = { .type = NLA_STRING, .len = IFNAMSIZ - 1 },
> };
>
> +/* inet_addr_hash's shifting is dependent upon this IN4_ADDR_HSIZE
> + * value. So if you change this define, make appropriate changes to
> + * inet_addr_hash as well.
> + */
> +#define IN4_ADDR_HSIZE 256
> +static struct hlist_head inet_addr_lst[IN4_ADDR_HSIZE];
> +static DEFINE_SPINLOCK(inet_addr_hash_lock);
> +
> +static inline unsigned int inet_addr_hash(__be32 addr)
> +{
> + u32 val = (__force u32) addr;
> +
> + return ((val ^ (val >> 8) ^ (val >> 16) ^ (val >> 24)) &
> + (IN4_ADDR_HSIZE - 1));
> +}
Maybe you should take into account net pointer here, or machines with
many net namespaces will hash collide for 127.0.0.1
^ permalink raw reply
* Re: 2.6.38-rc5 tcp_connect oops: EIP = 0x0
From: Eric Dumazet @ 2011-02-18 6:32 UTC (permalink / raw)
To: George Spelvin; +Cc: linux-kernel, netdev
In-Reply-To: <20110218050321.10415.qmail@science.horizon.com>
Le vendredi 18 février 2011 à 00:03 -0500, George Spelvin a écrit :
> But wonder of wonders, kernel mode switching worked so I got to see the
> oops on the text-mode console. It happened just as I tried to ssh out.
>
CC netdev (removed linux-netdev)
I'll take a look this morning, thanks for the report.
> It's a Core 2 duo laptop (Dell E1405), 2 GB RAM, running a 32-bit kernel.
> It's worth noting that I was using wired internet (b44 driver) and
> not wireless.
>
> I've been having a lot of weird lockups with 2.6.38-rcX, quite a change
> from the very stable 2.6.36, but this is the first time I booted -rc5.
> Also, the symptoms are very different; before the lockup did not seen
> correlated with any particular activity, but the "lockup" was more like
> something getting wedged in the kernel that more and more tasks would
> get stuck on until everything stopped responding.
>
> I should mention that this is transcribed by hand from the screen.
> Oh, and also, it is far from the first time I ran ssh this boot.
> (I re-tested it ater rebooting, just to be sure. Not a consistent
> crash.)
>
> Anyway, jumping to address 0 looks "interesting", so it seems worth reporting.
>
> BUG: unable to handle kernel NULL pointer dereference at (null)
> IP: [< (null)>] (null)
> *pde = 00000000
> Oops: 0000 [#1] SMP
> last sysfs file: /sys/devices/pci0000:00/0000:00:1f.2/host0/target0:0:0/0:0:0:0/block/sda/uevent
> Modules linked in: rfcomm btusb sco l2cap crc16 bluetooth b43 mac80211 cfg80211 [last unloaded: sha256_generic]
>
> Pid: 12178, comm: ssh Not tainted 2.6.38-rc5 #227 Dell Inc. MXC061 /0MG532
> EIP: 0060:[<00000000>] EFLAGS: 0021246 CPU: 0
> EIP is at 0x0
> EAX: f5947e00 EBX: f5982f80 ECX: 00000024 EDX: c148fe00
> ESI: f5947e00 EDI: ebf29e54 EBP: ebf29ee0 ESP: ebf29dd0
> DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
> Process ssh (pid: 12178, ti=ebf28000 task=ebefc6c0 task.ti = ebef28000)
> Stack:
> c12c83ea 07e2427d 02000000 7b018f79 c117b827 0b996609 7b018f79 5f5f6644
> f5982f80 00000000 ebf29e58 ebf29ee0 c12cc0e0 00001600 00000000 00000001
> 03641600 00000000 00000000 00000000 00000000 036423c0 3e6423c0 00000000
> Call Trace:
> [<c12c83ea>] ? tcp_connect+0xdd/0x3fd
> [<c117b827>] ? secure_tcp_sequence_number+0x4f/0x65
> [<c12cc0e0>] ? tcp_v4_connect+0x3c1/0x417
> [<c12d6726>] ? inet_stream_connect+0x88/0x1fc
> [<c1110705>] ? _copy_from_user+0x2b/0x10e
> [<c129307d>] ? sys_connect+0x70/0x98
> [<c108df0f>] ? get_empty_filp+0x9f/0x121
> [<c108dfa0>] ? alloc_file+0xf/0x85
> [<c1293287>] ? sock_alloc_file+0x97/0xeb
> [<c108b758>] ? fd_install+0x1b/0x38
> [<c12932f6>] ? sock_map_fd+0x1b/0x20
> [<c1293bdf>] ? sys_socketcall+0x9d/0x291
> [<c1002750>] ? sysenter_do_call+0x12/0x26
> Code: Bad EIP value
> EIP: [<00000000>] 0x0 SS:ESP 0068:ebf29dd0
> CR2: 0000000000000000
^ permalink raw reply
* Re: [PATCH v2 09/13] can: pruss CAN driver.
From: Subhasish Ghosh @ 2011-02-18 7:07 UTC (permalink / raw)
To: Kurt Van Dijck
Cc: sachi-EvXpCiN+lbve9wHmmfpqLFaTQe2KTcn/,
davinci-linux-open-source-VycZQUHpC/PFrsHnngEfi1aTQe2KTcn/,
open list:CAN NETWORK DRIVERS, nsekhar-l0cyMroinI0, open list,
open list:CAN NETWORK DRIVERS,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
m-watkins-l0cyMroinI0, Wolfgang Grandegger
In-Reply-To: <20110211152026.GC373-MxZ6Iy/zr/UdbCeoMzGj59i2O/JbrIOy@public.gmane.org>
--------------------------------------------------
From: "Kurt Van Dijck" <kurt.van.dijck-/BeEPy95v10@public.gmane.org>
Sent: Friday, February 11, 2011 8:50 PM
To: "Subhasish Ghosh" <subhasish-EvXpCiN+lbve9wHmmfpqLFaTQe2KTcn/@public.gmane.org>
Cc: <davinci-linux-open-source-VycZQUHpC/PFrsHnngEfi1aTQe2KTcn/@public.gmane.org>;
<linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org>; <m-watkins-l0cyMroinI0@public.gmane.org>;
<nsekhar-l0cyMroinI0@public.gmane.org>; <sachi-EvXpCiN+lbve9wHmmfpqLFaTQe2KTcn/@public.gmane.org>; "Wolfgang Grandegger"
<wg-5Yr1BZd7O62+XT7JhA+gdA@public.gmane.org>; "open list:CAN NETWORK DRIVERS"
<socketcan-core-0fE9KPoRgkgATYTw5x5z8w@public.gmane.org>; "open list:CAN NETWORK DRIVERS"
<netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>; "open list" <linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
Subject: Re: [PATCH v2 09/13] can: pruss CAN driver.
> Hi,
>
> I looked a bit at the TX path:
>
> On Fri, Feb 11, 2011 at 08:21:28PM +0530, Subhasish Ghosh wrote:
>> +static int omapl_pru_can_set_bittiming(struct net_device *ndev)
>> +{
>> + struct omapl_pru_can_priv *priv = netdev_priv(ndev);
>> + struct can_bittiming *bt = &priv->can.bittiming;
>> + long bit_error = 0;
>> +
>> + if (priv->can.ctrlmode & CAN_CTRLMODE_3_SAMPLES) {
>> + dev_warn(priv->dev, "WARN: Triple"
>> + "sampling not set due to h/w limitations");
> You should not have enabled CAN_CTRLMODE_3_SAMPLES in the first place?
SG - Ok Will remove.
>> + }
>> + if (pru_can_calc_timing(priv->dev, priv->can.clock.freq,
>> + bt->bitrate) != 0)
>> + return -EINVAL;
>> + bit_error =
>> + (((priv->timer_freq / (priv->timer_freq / bt->bitrate)) -
>> + bt->bitrate) * 1000) / bt->bitrate;
>> + if (bit_error) {
>> + bit_error =
>> + (((priv->timer_freq / (priv->timer_freq / bt->bitrate)) -
>> + bt->bitrate) * 1000000) / bt->bitrate;
>> + printk(KERN_INFO "\nBitrate error %ld.%ld%%\n",
>> + bit_error / 10000, bit_error % 1000);
>> + } else
>> + printk(KERN_INFO "\nBitrate error 0.0%%\n");
>> +
>> + return 0;
>> +}
> I wonder how much of this code is duplicated from drivers/net/can/dev.c ?
SG - Well, I just followed ti_hecc.c :-)
>
>> +static netdev_tx_t omapl_pru_can_start_xmit(struct sk_buff *skb,
>> + struct net_device *ndev)
>> +{
>> + struct omapl_pru_can_priv *priv = netdev_priv(ndev);
>> + struct can_frame *cf = (struct can_frame *)skb->data;
>> + int count;
>> + u8 *data = cf->data;
>> + u8 dlc = cf->can_dlc;
>> + u8 *ptr8data = NULL;
>> +
> most drivers start with:
> if (can_dropped_invalid_skb(dev, skb))
> return NETDEV_TX_OK;
SG - Will do.
>
>> + netif_stop_queue(ndev);
> why would you stop when you just resumed the queue?
SG - I do not want more than one transmit request at one time. Hence, on
entering the transmit
I am using netif_stop_queue to disable tx.
>> + if (cf->can_id & CAN_EFF_FLAG) /* Extended frame format */
>> + *((u32 *) &priv->can_tx_hndl.strcanmailbox) =
>> + (cf->can_id & CAN_EFF_MASK) | PRU_CANMID_IDE;
>> + else /* Standard frame format */
>> + *((u32 *) &priv->can_tx_hndl.strcanmailbox) =
>> + (cf->can_id & CAN_SFF_MASK) << 18;
>> +
>> + if (cf->can_id & CAN_RTR_FLAG) /* Remote transmission request */
>> + *((u32 *) &priv->can_tx_hndl.strcanmailbox) |= CAN_RTR_FLAG;
>> +
>> + ptr8data = &priv->can_tx_hndl.strcanmailbox.u8data7 + (dlc - 1);
>> + for (count = 0; count < (u8) dlc; count++) {
>> + *ptr8data-- = *data++;
>> + }
>> + *((u32 *) &priv->can_tx_hndl.strcanmailbox.u16datalength) = (u32) dlc;
>> +/*
>> + * search for the next available mbx
>> + * if the next mbx is busy, then try the next + 1
>> + * do this until the head is reached.
>> + * if still unable to tx, stop accepting any packets
>> + * if able to tx and the head is reached, then reset next to tail, i.e
>> mbx0
>> + * if head is not reached, then just point to the next mbx
>> + */
>> + for (; priv->tx_next <= priv->tx_head; priv->tx_next++) {
>> + priv->can_tx_hndl.ecanmailboxnumber =
>> + (can_mailbox_number) priv->tx_next;
>> + if (-1 == pru_can_write_data_to_mailbox(priv->dev,
>> + &priv->can_tx_hndl)) {
>> + if (priv->tx_next == priv->tx_head) {
>> + priv->tx_next = priv->tx_tail;
>> + if (!netif_queue_stopped(ndev))
> If you get here, the queue is not stopped. This test is therefore useless.
SG -Ok, will remove
>> + netif_stop_queue(ndev); /* IF stalled */
>> + dev_err(priv->dev,
>> + "%s: no tx mbx available", __func__);
>> + return NETDEV_TX_BUSY;
>> + } else
>> + continue;
>> + } else {
>> + /* set transmit request */
>> + pru_can_tx(priv->dev, priv->tx_next, CAN_TX_PRU_1);
>> + pru_can_tx_mode_set(priv->dev, false, ecanreceive);
>> + pru_can_tx_mode_set(priv->dev, true, ecantransmit);
>> + pru_can_start_abort_tx(priv->dev, PRU_CAN_START);
>> + priv->tx_next++;
>> + can_put_echo_skb(skb, ndev, 0);
>> + break;
>> + }
>> + }
>> + if (priv->tx_next > priv->tx_head) {
>> + priv->tx_next = priv->tx_tail;
>> + }
>> + return NETDEV_TX_OK;
>> +}
>> +
>> +
>
>> +irqreturn_t omapl_tx_can_intr(int irq, void *dev_id)
>> +{
>> + struct net_device *ndev = dev_id;
>> + struct omapl_pru_can_priv *priv = netdev_priv(ndev);
>> + struct net_device_stats *stats = &ndev->stats;
>> + u32 bit_set, mbxno;
>> +
>> + pru_can_get_intr_status(priv->dev, &priv->can_tx_hndl);
>> + if ((PRU_CAN_ISR_BIT_CCI & priv->can_tx_hndl.u32interruptstatus)
>> + || (PRU_CAN_ISR_BIT_SRDI & priv->can_tx_hndl.u32interruptstatus)) {
>> + __can_debug("tx_int_status = 0x%X\n",
>> + priv->can_tx_hndl.u32interruptstatus);
>> + can_free_echo_skb(ndev, 0);
>> + } else {
>> + for (bit_set = 0; ((priv->can_tx_hndl.u32interruptstatus & 0xFF)
>> + >> bit_set != 0); bit_set++)
>> + ;
>> + if (0 == bit_set) {
>> + __can_err("%s: invalid mailbox number\n", __func__);
>> + can_free_echo_skb(ndev, 0);
>> + } else {
>> + mbxno = bit_set - 1; /* mail box numbering starts from 0 */
>> + if (PRU_CAN_ISR_BIT_ESI & priv->can_tx_hndl.
>> + u32interruptstatus) {
>> + /* read gsr and ack pru */
>> + pru_can_get_global_status(priv->dev, &priv->can_tx_hndl);
>> + omapl_pru_can_err(ndev,
>> + priv->can_tx_hndl.
>> + u32interruptstatus,
>> + priv->can_tx_hndl.
>> + u32globalstatus);
>> + } else {
>> + stats->tx_packets++;
>> + /* stats->tx_bytes += dlc; */
>> + /*can_get_echo_skb(ndev, 0);*/
>> + }
>> + }
>> + }
>> + if (netif_queue_stopped(ndev))
> you can call netif_wake_queue(ndev) multiple times, so there is no need
> for netif_queue_stopped()
SG -Ok, will remove
>> + netif_wake_queue(ndev);
>> +
>> + can_get_echo_skb(ndev, 0);
>> + pru_can_tx_mode_set(priv->dev, true, ecanreceive);
>> + return IRQ_HANDLED;
>> +}
>> +
>> +static int omapl_pru_can_open(struct net_device *ndev)
>> +{
>> + struct omapl_pru_can_priv *priv = netdev_priv(ndev);
>> + int err;
>> +
>> + /* register interrupt handler */
>> + err = request_irq(priv->trx_irq, &omapl_rx_can_intr, IRQF_SHARED,
>> + "pru_can_irq", ndev);
> you're doing a lot of work _in_ the irq handler. Maybe threaded irq?
>
SG -Ok, will do
>> +static int omapl_pru_can_close(struct net_device *ndev)
>> +{
>> + struct omapl_pru_can_priv *priv = netdev_priv(ndev);
>> +
>> + if (!netif_queue_stopped(ndev))
> check is not needed.
SG -Ok, will remove
>> + netif_stop_queue(ndev);
>> +
>> + close_candev(ndev);
>> +
>> + free_irq(priv->trx_irq, ndev);
>> + return 0;
>> +}
>> +
>
> Regards,
> Kurt
^ permalink raw reply
* Re: IGMP and rwlock: Dead ocurred again on TILEPro
From: Cypher Wu @ 2011-02-18 7:08 UTC (permalink / raw)
To: Chris Metcalf
Cc: David Miller, xiyou.wangcong, linux-kernel, eric.dumazet, netdev
In-Reply-To: <4D5DACC5.60105@tilera.com>
On Fri, Feb 18, 2011 at 7:18 AM, Chris Metcalf <cmetcalf@tilera.com> wrote:
> On 2/17/2011 6:11 PM, David Miller wrote:
>> From: Chris Metcalf <cmetcalf@tilera.com>
>> Date: Thu, 17 Feb 2011 18:04:13 -0500
>>
>>> On 2/17/2011 5:53 PM, David Miller wrote:
>>>> From: Chris Metcalf <cmetcalf@tilera.com>
>>>> Date: Thu, 17 Feb 2011 17:49:46 -0500
>>>>
>>>>> The fix is to disable interrupts for the arch_read_lock family of methods.
>>>> How does that help handle the race when it happens between different
>>>> cpus, instead of between IRQ and non-IRQ context on the same CPU?
>>> There's no race in that case, since the lock code properly backs off and
>>> retries until the other cpu frees it. The distinction here is that the
>>> non-IRQ context is "wedged" by the IRQ context.
>>>
>>>> Why don't you just use the generic spinlock based rwlock code on Tile,
>>>> since that is all that your atomic instructions can handle
>>>> sufficiently?
>>> The tile-specific code encodes reader/writer information in the same 32-bit
>>> word that the test-and-set instruction manipulates, so it's more efficient
>>> both in space and time. This may not really matter for rwlocks, since no
>>> one cares much about them any more, but that was the motivation.
>> Ok, but IRQ disabling is going to be very expensive.
>
> The interrupt architecture on Tile allows a write to a special-purpose
> register to put you into a "critical section" where no interrupts or faults
> are delivered. So we just need to bracket the read_lock operations with
> two SPR writes; each takes six machine cycles, so we're only adding 12
> cycles to the total cost of taking or releasing a read lock on an rwlock.
>
> --
> Chris Metcalf, Tilera Corp.
> http://www.tilera.com
>
>
Adding that to SPR writes should be fine, but it may cause interrupt
delay a little more that other platform's read_lock()?
Another question: What NMI in the former mail means?
Looking forward to your patch.
Regards
--
Cyberman Wu
^ permalink raw reply
* Re: [PATCH v2 09/13] can: pruss CAN driver.
From: Wolfgang Grandegger @ 2011-02-18 7:53 UTC (permalink / raw)
To: Subhasish Ghosh
Cc: sachi-EvXpCiN+lbve9wHmmfpqLFaTQe2KTcn/,
davinci-linux-open-source-VycZQUHpC/PFrsHnngEfi1aTQe2KTcn/,
CAN NETWORK DRIVERS, nsekhar-l0cyMroinI0, open list,
CAN NETWORK DRIVERS, m-watkins-l0cyMroinI0,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r
In-Reply-To: <32A5399EB727427C98185089E5DBFA65@subhasishg>
On 02/18/2011 08:07 AM, Subhasish Ghosh wrote:
> --------------------------------------------------
> From: "Kurt Van Dijck" <kurt.van.dijck-/BeEPy95v10@public.gmane.org>
...
>>> + /* register interrupt handler */
>>> + err = request_irq(priv->trx_irq, &omapl_rx_can_intr, IRQF_SHARED,
>>> + "pru_can_irq", ndev);
>> you're doing a lot of work _in_ the irq handler. Maybe threaded irq?
>>
> SG -Ok, will do
No, please use NAPI instead.
Wolfgang
^ permalink raw reply
* Re: [PATCH v2 09/13] can: pruss CAN driver.
From: Subhasish Ghosh @ 2011-02-18 8:15 UTC (permalink / raw)
To: Wolfgang Grandegger
Cc: sachi-EvXpCiN+lbve9wHmmfpqLFaTQe2KTcn/,
davinci-linux-open-source-VycZQUHpC/PFrsHnngEfi1aTQe2KTcn/,
open list:CAN NETWORK DRIVERS, nsekhar-l0cyMroinI0, open list,
open list:CAN NETWORK DRIVERS, m-watkins-l0cyMroinI0,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r
In-Reply-To: <4D5E2570.10108-5Yr1BZd7O62+XT7JhA+gdA@public.gmane.org>
> On 02/18/2011 08:07 AM, Subhasish Ghosh wrote:
>> --------------------------------------------------
>> From: "Kurt Van Dijck" <kurt.van.dijck-/BeEPy95v10@public.gmane.org>
>
> ...
>>>> + /* register interrupt handler */
>>>> + err = request_irq(priv->trx_irq, &omapl_rx_can_intr, IRQF_SHARED,
>>>> + "pru_can_irq", ndev);
>>> you're doing a lot of work _in_ the irq handler. Maybe threaded irq?
>>>
>> SG -Ok, will do
>
> No, please use NAPI instead.
We are using h/w filters, so the number of interrupts coming into the
processor are not hogging it.
I feel that we may not require an interrupt mitigation.
-Subhasish
^ permalink raw reply
* [PATCH v2 1/2] r8169: Correct settings of rtl8102e
From: Hayes Wang @ 2011-02-18 8:34 UTC (permalink / raw)
To: romieu; +Cc: netdev, linux-kernel, Hayes Wang
Adjust and remove certain settings of RTL8102E which are for previous chips.
Signed-off-by: Hayes Wang <hayeswang@realtek.com>
---
drivers/net/r8169.c | 20 ++++++--------------
1 files changed, 6 insertions(+), 14 deletions(-)
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 59ccf0c..9eaf78f 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -3042,7 +3042,7 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
goto err_out_mwi_2;
}
- tp->cp_cmd = PCIMulRW | RxChkSum;
+ tp->cp_cmd = RxChkSum;
if ((sizeof(dma_addr_t) > 4) &&
!pci_set_dma_mask(pdev, DMA_BIT_MASK(64)) && use_dac) {
@@ -3845,8 +3845,7 @@ static void rtl_hw_start_8168(struct net_device *dev)
Cxpl_dbg_sel | \
ASF | \
PktCntrDisable | \
- PCIDAC | \
- PCIMulRW)
+ Mac_dbgo_sel)
static void rtl_hw_start_8102e_1(void __iomem *ioaddr, struct pci_dev *pdev)
{
@@ -3876,8 +3875,6 @@ static void rtl_hw_start_8102e_1(void __iomem *ioaddr, struct pci_dev *pdev)
if ((cfg1 & LEDS0) && (cfg1 & LEDS1))
RTL_W8(Config1, cfg1 & ~LEDS0);
- RTL_W16(CPlusCmd, RTL_R16(CPlusCmd) & ~R810X_CPCMD_QUIRK_MASK);
-
rtl_ephy_init(ioaddr, e_info_8102e_1, ARRAY_SIZE(e_info_8102e_1));
}
@@ -3889,8 +3886,6 @@ static void rtl_hw_start_8102e_2(void __iomem *ioaddr, struct pci_dev *pdev)
RTL_W8(Config1, MEMMAP | IOMAP | VPD | PMEnable);
RTL_W8(Config3, RTL_R8(Config3) & ~Beacon_en);
-
- RTL_W16(CPlusCmd, RTL_R16(CPlusCmd) & ~R810X_CPCMD_QUIRK_MASK);
}
static void rtl_hw_start_8102e_3(void __iomem *ioaddr, struct pci_dev *pdev)
@@ -3916,6 +3911,8 @@ static void rtl_hw_start_8101(struct net_device *dev)
}
}
+ RTL_W8(Cfg9346, Cfg9346_Unlock);
+
switch (tp->mac_version) {
case RTL_GIGA_MAC_VER_07:
rtl_hw_start_8102e_1(ioaddr, pdev);
@@ -3930,14 +3927,13 @@ static void rtl_hw_start_8101(struct net_device *dev)
break;
}
- RTL_W8(Cfg9346, Cfg9346_Unlock);
+ RTL_W8(Cfg9346, Cfg9346_Lock);
RTL_W8(MaxTxPacketSize, TxPacketMax);
rtl_set_rx_max_size(ioaddr, rx_buf_sz);
- tp->cp_cmd |= rtl_rw_cpluscmd(ioaddr) | PCIMulRW;
-
+ tp->cp_cmd &= ~R810X_CPCMD_QUIRK_MASK;
RTL_W16(CPlusCmd, tp->cp_cmd);
RTL_W16(IntrMitigate, 0x0000);
@@ -3947,14 +3943,10 @@ static void rtl_hw_start_8101(struct net_device *dev)
RTL_W8(ChipCmd, CmdTxEnb | CmdRxEnb);
rtl_set_rx_tx_config_registers(tp);
- RTL_W8(Cfg9346, Cfg9346_Lock);
-
RTL_R8(IntrMask);
rtl_set_rx_mode(dev);
- RTL_W8(ChipCmd, CmdTxEnb | CmdRxEnb);
-
RTL_W16(MultiIntr, RTL_R16(MultiIntr) & 0xf000);
RTL_W16(IntrMask, tp->intr_event);
--
1.7.3.2
^ permalink raw reply related
* [PATCH v2 2/2] r8169: Support RTL8105E
From: Hayes Wang @ 2011-02-18 8:34 UTC (permalink / raw)
To: romieu; +Cc: netdev, linux-kernel, Hayes Wang
In-Reply-To: <1298018085-1370-1-git-send-email-hayeswang@realtek.com>
Support the new chips for RTL8105E
Signed-off-by: Hayes Wang <hayeswang@realtek.com>
---
drivers/net/r8169.c | 92 +++++++++++++++++++++++++++++++++++++++++++++++++-
1 files changed, 90 insertions(+), 2 deletions(-)
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 9eaf78f..ffe6b00 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -36,6 +36,7 @@
#define FIRMWARE_8168D_1 "rtl_nic/rtl8168d-1.fw"
#define FIRMWARE_8168D_2 "rtl_nic/rtl8168d-2.fw"
+#define FIRMWARE_8105E_1 "rtl_nic/rtl8105e-1.fw"
#ifdef RTL8169_DEBUG
#define assert(expr) \
@@ -123,6 +124,8 @@ enum mac_version {
RTL_GIGA_MAC_VER_26 = 0x1a, // 8168D
RTL_GIGA_MAC_VER_27 = 0x1b, // 8168DP
RTL_GIGA_MAC_VER_28 = 0x1c, // 8168DP
+ RTL_GIGA_MAC_VER_29 = 0x1d, // 8105E
+ RTL_GIGA_MAC_VER_30 = 0x1e, // 8105E
};
#define _R(NAME,MAC,MASK) \
@@ -160,7 +163,9 @@ static const struct {
_R("RTL8168d/8111d", RTL_GIGA_MAC_VER_25, 0xff7e1880), // PCI-E
_R("RTL8168d/8111d", RTL_GIGA_MAC_VER_26, 0xff7e1880), // PCI-E
_R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_27, 0xff7e1880), // PCI-E
- _R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_28, 0xff7e1880) // PCI-E
+ _R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_28, 0xff7e1880), // PCI-E
+ _R("RTL8105e", RTL_GIGA_MAC_VER_29, 0xff7e1880), // PCI-E
+ _R("RTL8105e", RTL_GIGA_MAC_VER_30, 0xff7e1880) // PCI-E
};
#undef _R
@@ -267,6 +272,12 @@ enum rtl8168_8101_registers {
#define EPHYAR_REG_MASK 0x1f
#define EPHYAR_REG_SHIFT 16
#define EPHYAR_DATA_MASK 0xffff
+ DLLPR = 0xd0,
+#define PM_SWITCH (1 << 6)
+ TWSI = 0xd2,
+ MCU = 0xd3,
+#define EN_NDP (1 << 3)
+#define EN_OOB_RESET (1 << 2)
DBG_REG = 0xd1,
#define FIX_NAK_1 (1 << 4)
#define FIX_NAK_2 (1 << 3)
@@ -568,6 +579,7 @@ MODULE_LICENSE("GPL");
MODULE_VERSION(RTL8169_VERSION);
MODULE_FIRMWARE(FIRMWARE_8168D_1);
MODULE_FIRMWARE(FIRMWARE_8168D_2);
+MODULE_FIRMWARE(FIRMWARE_8105E_1);
static int rtl8169_open(struct net_device *dev);
static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb,
@@ -1143,7 +1155,9 @@ static int rtl8169_set_speed_xmii(struct net_device *dev,
(tp->mac_version != RTL_GIGA_MAC_VER_13) &&
(tp->mac_version != RTL_GIGA_MAC_VER_14) &&
(tp->mac_version != RTL_GIGA_MAC_VER_15) &&
- (tp->mac_version != RTL_GIGA_MAC_VER_16)) {
+ (tp->mac_version != RTL_GIGA_MAC_VER_16) &&
+ (tp->mac_version != RTL_GIGA_MAC_VER_29) &&
+ (tp->mac_version != RTL_GIGA_MAC_VER_30)) {
giga_ctrl |= ADVERTISE_1000FULL | ADVERTISE_1000HALF;
} else {
netif_info(tp, link, dev,
@@ -1559,6 +1573,9 @@ static void rtl8169_get_mac_version(struct rtl8169_private *tp,
{ 0x7c800000, 0x30000000, RTL_GIGA_MAC_VER_11 },
/* 8101 family. */
+ { 0x7cf00000, 0x40a00000, RTL_GIGA_MAC_VER_30 },
+ { 0x7cf00000, 0x40900000, RTL_GIGA_MAC_VER_29 },
+ { 0x7c800000, 0x40800000, RTL_GIGA_MAC_VER_30 },
{ 0x7cf00000, 0x34a00000, RTL_GIGA_MAC_VER_09 },
{ 0x7cf00000, 0x24a00000, RTL_GIGA_MAC_VER_09 },
{ 0x7cf00000, 0x34900000, RTL_GIGA_MAC_VER_08 },
@@ -2435,6 +2452,33 @@ static void rtl8102e_hw_phy_config(struct rtl8169_private *tp)
rtl_writephy_batch(tp, phy_reg_init, ARRAY_SIZE(phy_reg_init));
}
+static void rtl8105e_hw_phy_config(struct rtl8169_private *tp)
+{
+ static const struct phy_reg phy_reg_init[] = {
+ {0x1f, 0x0005},
+ {0x1a, 0x0000},
+ {0x1f, 0x0000},
+
+ {0x1f, 0x0004},
+ {0x1c, 0x0000},
+ {0x1f, 0x0000},
+
+ {0x1f, 0x0001},
+ {0x15, 0x7701},
+ {0x1f, 0x0000}
+ };
+
+ /* Diable ALDPS before ram code */
+ rtl_writephy(tp, 0x1f, 0x0000);
+ rtl_writephy(tp, 0x18, 0x0310);
+ msleep(100);
+
+ if (rtl_apply_firmware(tp, FIRMWARE_8105E_1) < 0)
+ netif_warn(tp, probe, tp->dev, "unable to apply firmware patch\n");
+
+ rtl_writephy_batch(tp, phy_reg_init, ARRAY_SIZE(phy_reg_init));
+}
+
static void rtl_hw_phy_config(struct net_device *dev)
{
struct rtl8169_private *tp = netdev_priv(dev);
@@ -2502,6 +2546,10 @@ static void rtl_hw_phy_config(struct net_device *dev)
case RTL_GIGA_MAC_VER_28:
rtl8168d_4_hw_phy_config(tp);
break;
+ case RTL_GIGA_MAC_VER_29:
+ case RTL_GIGA_MAC_VER_30:
+ rtl8105e_hw_phy_config(tp);
+ break;
default:
break;
@@ -2940,6 +2988,8 @@ static void __devinit rtl_init_pll_power_ops(struct rtl8169_private *tp)
case RTL_GIGA_MAC_VER_09:
case RTL_GIGA_MAC_VER_10:
case RTL_GIGA_MAC_VER_16:
+ case RTL_GIGA_MAC_VER_29:
+ case RTL_GIGA_MAC_VER_30:
ops->down = r810x_pll_power_down;
ops->up = r810x_pll_power_up;
break;
@@ -3895,6 +3945,37 @@ static void rtl_hw_start_8102e_3(void __iomem *ioaddr, struct pci_dev *pdev)
rtl_ephy_write(ioaddr, 0x03, 0xc2f9);
}
+static void rtl_hw_start_8105e_1(void __iomem *ioaddr, struct pci_dev *pdev)
+{
+ static const struct ephy_info e_info_8105e_1[] = {
+ { 0x07, 0, 0x4000 },
+ { 0x19, 0, 0x0200 },
+ { 0x19, 0, 0x0020 },
+ { 0x1e, 0, 0x2000 },
+ { 0x03, 0, 0x0001 },
+ { 0x19, 0, 0x0100 },
+ { 0x19, 0, 0x0004 },
+ { 0x0a, 0, 0x0020 }
+ };
+
+ /* Force LAN exit from ASPM if Rx/Tx are not idel */
+ RTL_W32(FuncEvent, RTL_R32(FuncEvent) | 0x002800);
+
+ /* disable Early Tally Counter */
+ RTL_W32(FuncEvent, RTL_R32(FuncEvent) & ~0x010000);
+
+ RTL_W8(MCU, RTL_R8(MCU) | EN_NDP | EN_OOB_RESET);
+ RTL_W8(DLLPR, RTL_R8(DLLPR) | PM_SWITCH);
+
+ rtl_ephy_init(ioaddr, e_info_8105e_1, ARRAY_SIZE(e_info_8105e_1));
+}
+
+static void rtl_hw_start_8105e_2(void __iomem *ioaddr, struct pci_dev *pdev)
+{
+ rtl_hw_start_8105e_1(ioaddr, pdev);
+ rtl_ephy_write(ioaddr, 0x1e, rtl_ephy_read(ioaddr, 0x1e) | 0x8000);
+}
+
static void rtl_hw_start_8101(struct net_device *dev)
{
struct rtl8169_private *tp = netdev_priv(dev);
@@ -3925,6 +4006,13 @@ static void rtl_hw_start_8101(struct net_device *dev)
case RTL_GIGA_MAC_VER_09:
rtl_hw_start_8102e_2(ioaddr, pdev);
break;
+
+ case RTL_GIGA_MAC_VER_29:
+ rtl_hw_start_8105e_1(ioaddr, pdev);
+ break;
+ case RTL_GIGA_MAC_VER_30:
+ rtl_hw_start_8105e_2(ioaddr, pdev);
+ break;
}
RTL_W8(Cfg9346, Cfg9346_Lock);
--
1.7.3.2
^ permalink raw reply related
* Re: [PATCH v2 09/13] can: pruss CAN driver.
From: Marc Kleine-Budde @ 2011-02-18 8:36 UTC (permalink / raw)
To: Subhasish Ghosh
Cc: sachi-EvXpCiN+lbve9wHmmfpqLFaTQe2KTcn/,
davinci-linux-open-source-VycZQUHpC/PFrsHnngEfi1aTQe2KTcn/,
CAN NETWORK DRIVERS, nsekhar-l0cyMroinI0, LKML,
CAN NETWORK DRIVERS,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
m-watkins-l0cyMroinI0, Wolfgang Grandegger
In-Reply-To: <184FA07B783D4F5FAAC09D5C11737A3F@subhasishg>
[-- Attachment #1.1: Type: text/plain, Size: 1058 bytes --]
On 02/18/2011 09:15 AM, Subhasish Ghosh wrote:
>> On 02/18/2011 08:07 AM, Subhasish Ghosh wrote:
>>> --------------------------------------------------
>>> From: "Kurt Van Dijck" <kurt.van.dijck-/BeEPy95v10@public.gmane.org>
>>
>> ...
>>>>> + /* register interrupt handler */
>>>>> + err = request_irq(priv->trx_irq, &omapl_rx_can_intr, IRQF_SHARED,
>>>>> + "pru_can_irq", ndev);
>>>> you're doing a lot of work _in_ the irq handler. Maybe threaded irq?
>>>>
>>> SG -Ok, will do
>>
>> No, please use NAPI instead.
>
> We are using h/w filters, so the number of interrupts coming into the
> processor are not hogging it.
> I feel that we may not require an interrupt mitigation.
As davem stated the other day, all new drivers should use NAPI.
regards, Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]
[-- Attachment #2: Type: text/plain, Size: 188 bytes --]
_______________________________________________
Socketcan-core mailing list
Socketcan-core-0fE9KPoRgkgATYTw5x5z8w@public.gmane.org
https://lists.berlios.de/mailman/listinfo/socketcan-core
^ permalink raw reply
* Transfer Notice
From: Western Union @ 2011-02-18 8:44 UTC (permalink / raw)
To: Info
We have $100,000usd Winning Fund in our possession for transfer to you
from BLACK BERRY MOBILE COMPANY. Please send the following to us.
Receivers name: Address: Phone No: Country: Regards James Ferguson
^ permalink raw reply
* [PATCH 1/2] net: dont leave active on stack LIST_HEAD
From: Eric Dumazet @ 2011-02-18 8:54 UTC (permalink / raw)
To: David Miller
Cc: torvalds, ebiederm, opurdila, mingo, mhocko, linux-mm,
linux-kernel, netdev
In-Reply-To: <1298014191.2642.11.camel@edumazet-laptop>
From: Linus Torvalds <torvalds@linux-foundation.org>
Eric W. Biderman and Michal Hocko reported various memory corruptions
that we suspected to be related to a LIST head located on stack, that
was manipulated after thread left function frame (and eventually exited,
so its stack was freed and reused).
Eric Dumazet suggested the problem was probably coming from commit
443457242beb (net: factorize
sync-rcu call in unregister_netdevice_many)
This patch fixes __dev_close() and dev_close() to properly deinit their
respective LIST_HEAD(single) before exiting.
References: https://lkml.org/lkml/2011/2/16/304
References: https://lkml.org/lkml/2011/2/14/223
Reported-by: Michal Hocko <mhocko@suse.cz>
Reported-by: Eric W. Biderman <ebiderman@xmission.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Ingo Molnar <mingo@elte.hu>
CC: Octavian Purdila <opurdila@ixiacom.com>
CC: Eric W. Biderman <ebiderman@xmission.com>
---
net/core/dev.c | 7 +++++--
1 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/net/core/dev.c b/net/core/dev.c
index 8e726cb..a18c164 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1280,10 +1280,13 @@ static int __dev_close_many(struct list_head *head)
static int __dev_close(struct net_device *dev)
{
+ int retval;
LIST_HEAD(single);
list_add(&dev->unreg_list, &single);
- return __dev_close_many(&single);
+ retval = __dev_close_many(&single);
+ list_del(&single);
+ return retval;
}
int dev_close_many(struct list_head *head)
@@ -1325,7 +1328,7 @@ int dev_close(struct net_device *dev)
list_add(&dev->unreg_list, &single);
dev_close_many(&single);
-
+ list_del(&single);
return 0;
}
EXPORT_SYMBOL(dev_close);
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply related
* [PATCH 2/2] net: deinit automatic LIST_HEAD
From: Eric Dumazet @ 2011-02-18 8:59 UTC (permalink / raw)
To: David Miller
Cc: torvalds, ebiederm, mingo, opurdila, linux-kernel, linux-mm,
mhocko, netdev, stable
In-Reply-To: <m17hcx7wca.fsf@fess.ebiederm.org>
commit 9b5e383c11b08784 (net: Introduce
unregister_netdevice_many()) left an active LIST_HEAD() in
rollback_registered(), with possible memory corruption.
Even if device is freed without touching its unreg_list (and therefore
touching the previous memory location holding LISTE_HEAD(single), better
close the bug for good, since its really subtle.
(Same fix for default_device_exit_batch() for completeness)
Reported-by: Michal Hocko <mhocko@suse.cz>
Reported-by: Eric W. Biderman <ebiderman@xmission.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Ingo Molnar <mingo@elte.hu>
CC: Octavian Purdila <opurdila@ixiacom.com>
CC: Eric W. Biderman <ebiderman@xmission.com>
CC: stable <stable@kernel.org> [.33+]
---
net/core/dev.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/net/core/dev.c b/net/core/dev.c
index a18c164..8ae6631 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -5066,6 +5066,7 @@ static void rollback_registered(struct net_device *dev)
list_add(&dev->unreg_list, &single);
rollback_registered_many(&single);
+ list_del(&single);
}
unsigned long netdev_fix_features(unsigned long features, const char *name)
@@ -6219,6 +6220,7 @@ static void __net_exit default_device_exit_batch(struct list_head *net_list)
}
}
unregister_netdevice_many(&dev_kill_list);
+ list_del(&dev_kill_list);
rtnl_unlock();
}
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply related
* Re: [PATCH v2 09/13] can: pruss CAN driver.
From: Subhasish Ghosh @ 2011-02-18 9:09 UTC (permalink / raw)
To: Marc Kleine-Budde
Cc: sachi-EvXpCiN+lbve9wHmmfpqLFaTQe2KTcn/,
davinci-linux-open-source-VycZQUHpC/PFrsHnngEfi1aTQe2KTcn/,
CAN NETWORK DRIVERS, nsekhar-l0cyMroinI0, LKML,
CAN NETWORK DRIVERS,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
m-watkins-l0cyMroinI0, Wolfgang Grandegger
In-Reply-To: <4D5E2F86.1020400-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
Ok, will do NAPI
On 02/18/2011 09:15 AM, Subhasish Ghosh wrote:
>> On 02/18/2011 08:07 AM, Subhasish Ghosh wrote:
>>> --------------------------------------------------
>>> From: "Kurt Van Dijck" <kurt.van.dijck-/BeEPy95v10@public.gmane.org>
>>
>> ...
>>>>> + /* register interrupt handler */
>>>>> + err = request_irq(priv->trx_irq, &omapl_rx_can_intr, IRQF_SHARED,
>>>>> + "pru_can_irq", ndev);
>>>> you're doing a lot of work _in_ the irq handler. Maybe threaded irq?
>>>>
>>> SG -Ok, will do
>>
>> No, please use NAPI instead.
>
> We are using h/w filters, so the number of interrupts coming into the
> processor are not hogging it.
> I feel that we may not require an interrupt mitigation.
As davem stated the other day, all new drivers should use NAPI.
regards, Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
^ permalink raw reply
* Re: Mass udp flow reboot linux with RealTek RTL-8169 Gigabit
From: Francois Romieu @ 2011-02-18 9:30 UTC (permalink / raw)
To: Seblu; +Cc: Eric Dumazet, lkml, netdev, Ivan Vecera
In-Reply-To: <AANLkTi=PrOXMD54hSbEwUtmzXiwsSwiSijE3AJ9n=WDe@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 813 bytes --]
Seblu <seblu@seblu.net> :
[...]
> I've applyed your patch on 2.6.38-rc5. Host have rebooted 2mn after udp start.
> After this reboot, host is still on after 2 hour under a 1Gbit/s udp flow.
Thanks for testing.
> I attached a dmesg output before reboot. Do you need anything else?
Mostly :
1. .config
2. the size of the udp packets and the mtu
As an option :
3. a few seconds of 'vmstat 1' from the host under test
4. an 'ethtool -s eth0' from the host under test
5. /proc/interrupts from the host under test
6. lspci -tv
Can you apply the two attached patches on top of the previous ones and
give it a try ? The debug should not be too verbose if things are stationary
enough.
Do you have a serial cable and a second computer at hand by chance (don't
go for netconsole with the r8169 driver) ?
--
Ueimor
[-- Attachment #2: 0001-r8169-TX-stop-queue-race-fix.patch --]
[-- Type: text/plain, Size: 1129 bytes --]
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 59ccf0c..712231f 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -4361,13 +4361,13 @@ static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb,
tp->cur_tx += frags + 1;
- wmb();
-
RTL_W8(TxPoll, NPQ); /* set polling bit */
+ mmiowb();
+
if (TX_BUFFS_AVAIL(tp) < MAX_SKB_FRAGS) {
netif_stop_queue(dev);
- smp_rmb();
+ smp_mb();
if (TX_BUFFS_AVAIL(tp) >= MAX_SKB_FRAGS)
netif_wake_queue(dev);
}
@@ -4468,10 +4468,14 @@ static void rtl8169_tx_interrupt(struct net_device *dev,
if (tp->dirty_tx != dirty_tx) {
tp->dirty_tx = dirty_tx;
- smp_wmb();
- if (netif_queue_stopped(dev) &&
- (TX_BUFFS_AVAIL(tp) >= MAX_SKB_FRAGS)) {
- netif_wake_queue(dev);
+ smp_mb();
+ if (unlikely(netif_queue_stopped(dev) &&
+ (TX_BUFFS_AVAIL(tp) >= (NUM_TX_DESC / 4)))) {
+ netif_tx_lock(dev);
+ if (netif_queue_stopped(dev) &&
+ (TX_BUFFS_AVAIL(tp) >= (NUM_TX_DESC / 4)))
+ netif_wake_queue(dev);
+ netif_tx_unlock(dev);
}
/*
* 8168 hack: TxPoll requests are lost when the Tx packets are
--
1.7.3.4
[-- Attachment #3: 0002-r8169-debuginator.patch --]
[-- Type: text/plain, Size: 856 bytes --]
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 712231f..5eaccbb 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -4622,6 +4622,11 @@ static int rtl8169_rx_interrupt(struct net_device *dev,
return count;
}
+static struct {
+ u16 status[4];
+ u16 idx;
+} x = { { 0 }, 0 };
+
static irqreturn_t rtl8169_interrupt(int irq, void *dev_instance)
{
struct net_device *dev = dev_instance;
@@ -4637,6 +4642,12 @@ static irqreturn_t rtl8169_interrupt(int irq, void *dev_instance)
while (status && status != 0xffff) {
handled = 1;
+ x.status[x.idx++ % 4] = status;
+ if (net_ratelimit()) {
+ printk(KERN_INFO "%04x %04x %04x %04x\n",
+ x.status[0], x.status[1],
+ x.status[2], x.status[3]);
+ }
/* Handle all of the error cases first. These will reset
* the chip, so just exit the loop.
*/
--
1.7.3.4
^ permalink raw reply related
* [V4 PATCH 1/3] bonding: sync netpoll code with bridge
From: Amerigo Wang @ 2011-02-18 9:43 UTC (permalink / raw)
To: linux-kernel; +Cc: Neil Horman, Herbert Xu, Amerigo Wang, Jay Vosburgh, netdev
V4: rebase to net-next-2.6
V3: remove an useless #ifdef.
This patch unifies the netpoll code in bonding with netpoll code in bridge,
thanks to Herbert that code is much cleaner now.
Signed-off-by: WANG Cong <amwang@redhat.com>
Cc: Neil Horman <nhorman@redhat.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
---
drivers/net/bonding/bond_main.c | 155 ++++++++++++++++++++++++--------------
drivers/net/bonding/bonding.h | 20 +++++
2 files changed, 118 insertions(+), 57 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 77e3c6a..2ed6624 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -59,7 +59,6 @@
#include <linux/uaccess.h>
#include <linux/errno.h>
#include <linux/netdevice.h>
-#include <linux/netpoll.h>
#include <linux/inetdevice.h>
#include <linux/igmp.h>
#include <linux/etherdevice.h>
@@ -424,15 +423,11 @@ int bond_dev_queue_xmit(struct bonding *bond, struct sk_buff *skb,
{
skb->dev = slave_dev;
skb->priority = 1;
-#ifdef CONFIG_NET_POLL_CONTROLLER
- if (unlikely(bond->dev->priv_flags & IFF_IN_NETPOLL)) {
- struct netpoll *np = bond->dev->npinfo->netpoll;
- slave_dev->npinfo = bond->dev->npinfo;
+ if (unlikely(netpoll_tx_running(slave_dev))) {
slave_dev->priv_flags |= IFF_IN_NETPOLL;
- netpoll_send_skb_on_dev(np, skb, slave_dev);
+ bond_netpoll_send_skb(bond_get_slave_by_dev(bond, slave_dev), skb);
slave_dev->priv_flags &= ~IFF_IN_NETPOLL;
} else
-#endif
dev_queue_xmit(skb);
return 0;
@@ -1288,63 +1283,113 @@ static void bond_detach_slave(struct bonding *bond, struct slave *slave)
}
#ifdef CONFIG_NET_POLL_CONTROLLER
-/*
- * You must hold read lock on bond->lock before calling this.
- */
-static bool slaves_support_netpoll(struct net_device *bond_dev)
+static inline int slave_enable_netpoll(struct slave *slave)
{
- struct bonding *bond = netdev_priv(bond_dev);
- struct slave *slave;
- int i = 0;
- bool ret = true;
+ struct netpoll *np;
+ int err = 0;
- bond_for_each_slave(bond, slave, i) {
- if ((slave->dev->priv_flags & IFF_DISABLE_NETPOLL) ||
- !slave->dev->netdev_ops->ndo_poll_controller)
- ret = false;
+ np = kzalloc(sizeof(*np), GFP_KERNEL);
+ err = -ENOMEM;
+ if (!np)
+ goto out;
+
+ np->dev = slave->dev;
+ err = __netpoll_setup(np);
+ if (err) {
+ kfree(np);
+ goto out;
}
- return i != 0 && ret;
+ slave->np = np;
+out:
+ return err;
+}
+static inline void slave_disable_netpoll(struct slave *slave)
+{
+ struct netpoll *np = slave->np;
+
+ if (!np)
+ return;
+
+ slave->np = NULL;
+ synchronize_rcu_bh();
+ __netpoll_cleanup(np);
+ kfree(np);
+}
+static inline bool slave_dev_support_netpoll(struct net_device *slave_dev)
+{
+ if (slave_dev->priv_flags & IFF_DISABLE_NETPOLL)
+ return false;
+ if (!slave_dev->netdev_ops->ndo_poll_controller)
+ return false;
+ return true;
}
static void bond_poll_controller(struct net_device *bond_dev)
{
- struct bonding *bond = netdev_priv(bond_dev);
+}
+
+static void __bond_netpoll_cleanup(struct bonding *bond)
+{
struct slave *slave;
int i;
- bond_for_each_slave(bond, slave, i) {
- if (slave->dev && IS_UP(slave->dev))
- netpoll_poll_dev(slave->dev);
- }
+ bond_for_each_slave(bond, slave, i)
+ if (IS_UP(slave->dev))
+ slave_disable_netpoll(slave);
}
-
static void bond_netpoll_cleanup(struct net_device *bond_dev)
{
struct bonding *bond = netdev_priv(bond_dev);
+
+ read_lock(&bond->lock);
+ __bond_netpoll_cleanup(bond);
+ read_unlock(&bond->lock);
+}
+
+static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni)
+{
+ struct bonding *bond = netdev_priv(dev);
struct slave *slave;
- const struct net_device_ops *ops;
- int i;
+ int i, err = 0;
read_lock(&bond->lock);
- bond_dev->npinfo = NULL;
bond_for_each_slave(bond, slave, i) {
- if (slave->dev) {
- ops = slave->dev->netdev_ops;
- if (ops->ndo_netpoll_cleanup)
- ops->ndo_netpoll_cleanup(slave->dev);
- else
- slave->dev->npinfo = NULL;
+ if (!IS_UP(slave->dev))
+ continue;
+ err = slave_enable_netpoll(slave);
+ if (err) {
+ __bond_netpoll_cleanup(bond);
+ break;
}
}
read_unlock(&bond->lock);
+ return err;
}
-#else
+static struct netpoll_info *bond_netpoll_info(struct bonding *bond)
+{
+ return bond->dev->npinfo;
+}
+#else
+static inline int slave_enable_netpoll(struct slave *slave)
+{
+ return 0;
+}
+static inline void slave_disable_netpoll(struct slave *slave)
+{
+}
static void bond_netpoll_cleanup(struct net_device *bond_dev)
{
}
-
+static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni)
+{
+ return 0;
+}
+static struct netpoll_info *bond_netpoll_info(struct bonding *bond)
+{
+ return NULL;
+}
#endif
/*---------------------------------- IOCTL ----------------------------------*/
@@ -1782,17 +1827,19 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev)
bond_set_carrier(bond);
#ifdef CONFIG_NET_POLL_CONTROLLER
- if (slaves_support_netpoll(bond_dev)) {
- bond_dev->priv_flags &= ~IFF_DISABLE_NETPOLL;
- if (bond_dev->npinfo)
- slave_dev->npinfo = bond_dev->npinfo;
- } else if (!(bond_dev->priv_flags & IFF_DISABLE_NETPOLL)) {
- bond_dev->priv_flags |= IFF_DISABLE_NETPOLL;
- pr_info("New slave device %s does not support netpoll\n",
- slave_dev->name);
- pr_info("Disabling netpoll support for %s\n", bond_dev->name);
+ slave_dev->npinfo = bond_netpoll_info(bond);
+ if (slave_dev->npinfo) {
+ if (slave_enable_netpoll(new_slave)) {
+ read_unlock(&bond->lock);
+ pr_info("Error, %s: master_dev is using netpoll, "
+ "but new slave device does not support netpoll.\n",
+ bond_dev->name);
+ res = -EBUSY;
+ goto err_close;
+ }
}
#endif
+
read_unlock(&bond->lock);
res = bond_create_slave_symlinks(bond_dev, slave_dev);
@@ -1994,17 +2041,7 @@ int bond_release(struct net_device *bond_dev, struct net_device *slave_dev)
netdev_set_bond_master(slave_dev, NULL);
-#ifdef CONFIG_NET_POLL_CONTROLLER
- read_lock_bh(&bond->lock);
-
- if (slaves_support_netpoll(bond_dev))
- bond_dev->priv_flags &= ~IFF_DISABLE_NETPOLL;
- read_unlock_bh(&bond->lock);
- if (slave_dev->netdev_ops->ndo_netpoll_cleanup)
- slave_dev->netdev_ops->ndo_netpoll_cleanup(slave_dev);
- else
- slave_dev->npinfo = NULL;
-#endif
+ slave_disable_netpoll(slave);
/* close slave before restoring its mac address */
dev_close(slave_dev);
@@ -2039,6 +2076,7 @@ static int bond_release_and_destroy(struct net_device *bond_dev,
ret = bond_release(bond_dev, slave_dev);
if ((ret == 0) && (bond->slave_cnt == 0)) {
+ bond_dev->priv_flags |= IFF_DISABLE_NETPOLL;
pr_info("%s: destroying bond %s.\n",
bond_dev->name, bond_dev->name);
unregister_netdevice(bond_dev);
@@ -2116,6 +2154,8 @@ static int bond_release_all(struct net_device *bond_dev)
netdev_set_bond_master(slave_dev, NULL);
+ slave_disable_netpoll(slave);
+
/* close slave before restoring its mac address */
dev_close(slave_dev);
@@ -4654,6 +4694,7 @@ static const struct net_device_ops bond_netdev_ops = {
.ndo_vlan_rx_add_vid = bond_vlan_rx_add_vid,
.ndo_vlan_rx_kill_vid = bond_vlan_rx_kill_vid,
#ifdef CONFIG_NET_POLL_CONTROLLER
+ .ndo_netpoll_setup = bond_netpoll_setup,
.ndo_netpoll_cleanup = bond_netpoll_cleanup,
.ndo_poll_controller = bond_poll_controller,
#endif
diff --git a/drivers/net/bonding/bonding.h b/drivers/net/bonding/bonding.h
index 31fe980..0a3e00b 100644
--- a/drivers/net/bonding/bonding.h
+++ b/drivers/net/bonding/bonding.h
@@ -20,6 +20,7 @@
#include <linux/if_bonding.h>
#include <linux/cpumask.h>
#include <linux/in6.h>
+#include <linux/netpoll.h>
#include "bond_3ad.h"
#include "bond_alb.h"
@@ -198,6 +199,9 @@ struct slave {
u16 queue_id;
struct ad_slave_info ad_info; /* HUGE - better to dynamically alloc */
struct tlb_slave_info tlb_info;
+#ifdef CONFIG_NET_POLL_CONTROLLER
+ struct netpoll *np;
+#endif
};
/*
@@ -323,6 +327,22 @@ static inline unsigned long slave_last_rx(struct bonding *bond,
return slave->dev->last_rx;
}
+#ifdef CONFIG_NET_POLL_CONTROLLER
+static inline void bond_netpoll_send_skb(const struct slave *slave,
+ struct sk_buff *skb)
+{
+ struct netpoll *np = slave->np;
+
+ if (np)
+ netpoll_send_skb(np, skb);
+}
+#else
+static inline void bond_netpoll_send_skb(const struct slave *slave,
+ struct sk_buff *skb)
+{
+}
+#endif
+
static inline void bond_set_slave_inactive_flags(struct slave *slave)
{
struct bonding *bond = netdev_priv(slave->dev->master);
--
1.7.1
^ permalink raw reply related
* [V4 PATCH 2/3] netpoll: remove IFF_IN_NETPOLL flag
From: Amerigo Wang @ 2011-02-18 9:43 UTC (permalink / raw)
To: linux-kernel
Cc: Neil Horman, Herbert Xu, Amerigo Wang, Jay Vosburgh,
David S. Miller, Simon Horman, Jiri Pirko, Patrick McHardy,
Neil Horman, Eric Dumazet, netdev
In-Reply-To: <1298022215-21059-1-git-send-email-amwang@redhat.com>
V4: rebase to net-next-2.6
This patch removes the flag IFF_IN_NETPOLL, we don't need it any more since
we have netpoll_tx_running() now.
Signed-off-by: WANG Cong <amwang@redhat.com>
Cc: Neil Horman <nhorman@redhat.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
---
drivers/net/bonding/bond_main.c | 6 ++----
drivers/net/bonding/bonding.h | 2 +-
include/linux/if.h | 9 ++++-----
net/core/netpoll.c | 2 --
4 files changed, 7 insertions(+), 12 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 2ed6624..c75126d 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -423,11 +423,9 @@ int bond_dev_queue_xmit(struct bonding *bond, struct sk_buff *skb,
{
skb->dev = slave_dev;
skb->priority = 1;
- if (unlikely(netpoll_tx_running(slave_dev))) {
- slave_dev->priv_flags |= IFF_IN_NETPOLL;
+ if (unlikely(netpoll_tx_running(slave_dev)))
bond_netpoll_send_skb(bond_get_slave_by_dev(bond, slave_dev), skb);
- slave_dev->priv_flags &= ~IFF_IN_NETPOLL;
- } else
+ else
dev_queue_xmit(skb);
return 0;
diff --git a/drivers/net/bonding/bonding.h b/drivers/net/bonding/bonding.h
index 0a3e00b..a401b8d 100644
--- a/drivers/net/bonding/bonding.h
+++ b/drivers/net/bonding/bonding.h
@@ -133,7 +133,7 @@ static inline void unblock_netpoll_tx(void)
static inline int is_netpoll_tx_blocked(struct net_device *dev)
{
- if (unlikely(dev->priv_flags & IFF_IN_NETPOLL))
+ if (unlikely(netpoll_tx_running(dev)))
return atomic_read(&netpoll_block_tx);
return 0;
}
diff --git a/include/linux/if.h b/include/linux/if.h
index 1239599..3bc63e6 100644
--- a/include/linux/if.h
+++ b/include/linux/if.h
@@ -71,11 +71,10 @@
* release skb->dst
*/
#define IFF_DONT_BRIDGE 0x800 /* disallow bridging this ether dev */
-#define IFF_IN_NETPOLL 0x1000 /* whether we are processing netpoll */
-#define IFF_DISABLE_NETPOLL 0x2000 /* disable netpoll at run-time */
-#define IFF_MACVLAN_PORT 0x4000 /* device used as macvlan port */
-#define IFF_BRIDGE_PORT 0x8000 /* device used as bridge port */
-#define IFF_OVS_DATAPATH 0x10000 /* device used as Open vSwitch
+#define IFF_DISABLE_NETPOLL 0x1000 /* disable netpoll at run-time */
+#define IFF_MACVLAN_PORT 0x2000 /* device used as macvlan port */
+#define IFF_BRIDGE_PORT 0x4000 /* device used as bridge port */
+#define IFF_OVS_DATAPATH 0x8000 /* device used as Open vSwitch
* datapath port */
#define IF_GET_IFACE 0x0001 /* for querying only */
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index 02dc2cb..f68e694 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -313,9 +313,7 @@ void netpoll_send_skb_on_dev(struct netpoll *np, struct sk_buff *skb,
tries > 0; --tries) {
if (__netif_tx_trylock(txq)) {
if (!netif_tx_queue_stopped(txq)) {
- dev->priv_flags |= IFF_IN_NETPOLL;
status = ops->ndo_start_xmit(skb, dev);
- dev->priv_flags &= ~IFF_IN_NETPOLL;
if (status == NETDEV_TX_OK)
txq_trans_update(txq);
}
--
1.7.1
^ permalink raw reply related
* [V4 PATCH 3/3] bond: service netpoll arp queue on master device
From: Amerigo Wang @ 2011-02-18 9:43 UTC (permalink / raw)
To: linux-kernel
Cc: Neil Horman, Herbert Xu, Amerigo Wang, David S. Miller,
Neil Horman, Eric Dumazet, netdev
In-Reply-To: <1298022215-21059-1-git-send-email-amwang@redhat.com>
Neil pointed out that we can't send ARP reply on behalf of slaves,
we need to move the arp queue to their bond device.
Signed-off-by: WANG Cong <amwang@redhat.com>
Cc: Neil Horman <nhorman@redhat.com>
---
net/core/netpoll.c | 9 +++++++++
1 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index f68e694..013e04a 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -193,6 +193,15 @@ void netpoll_poll_dev(struct net_device *dev)
poll_napi(dev);
+ if (dev->priv_flags & IFF_SLAVE) {
+ if (dev->npinfo) {
+ struct net_device *bond_dev = dev->master;
+ struct sk_buff *skb;
+ while ((skb = skb_dequeue(&dev->npinfo->arp_tx)))
+ skb_queue_tail(&bond_dev->npinfo->arp_tx, skb);
+ }
+ }
+
service_arp_queue(dev->npinfo);
zap_completion_queue();
--
1.7.1
^ permalink raw reply related
* [PATCH] net: Add default_advmss() methods to blackhole dst_ops
From: Eric Dumazet @ 2011-02-18 9:57 UTC (permalink / raw)
To: George Spelvin, David Miller; +Cc: linux-kernel, netdev, Roland Dreier
In-Reply-To: <1298010735.2642.9.camel@edumazet-laptop>
Le vendredi 18 février 2011 à 07:32 +0100, Eric Dumazet a écrit :
> Le vendredi 18 février 2011 à 00:03 -0500, George Spelvin a écrit :
> > But wonder of wonders, kernel mode switching worked so I got to see the
> > oops on the text-mode console. It happened just as I tried to ssh out.
> >
>
> CC netdev (removed linux-netdev)
>
> I'll take a look this morning, thanks for the report.
>
> > It's a Core 2 duo laptop (Dell E1405), 2 GB RAM, running a 32-bit kernel.
> > It's worth noting that I was using wired internet (b44 driver) and
> > not wireless.
> >
> > I've been having a lot of weird lockups with 2.6.38-rcX, quite a change
> > from the very stable 2.6.36, but this is the first time I booted -rc5.
> > Also, the symptoms are very different; before the lockup did not seen
> > correlated with any particular activity, but the "lockup" was more like
> > something getting wedged in the kernel that more and more tasks would
> > get stuck on until everything stopped responding.
> >
> > I should mention that this is transcribed by hand from the screen.
> > Oh, and also, it is far from the first time I ran ssh this boot.
> > (I re-tested it ater rebooting, just to be sure. Not a consistent
> > crash.)
> >
> > Anyway, jumping to address 0 looks "interesting", so it seems worth reporting.
> >
> > BUG: unable to handle kernel NULL pointer dereference at (null)
> > IP: [< (null)>] (null)
> > *pde = 00000000
> > Oops: 0000 [#1] SMP
> > last sysfs file: /sys/devices/pci0000:00/0000:00:1f.2/host0/target0:0:0/0:0:0:0/block/sda/uevent
> > Modules linked in: rfcomm btusb sco l2cap crc16 bluetooth b43 mac80211 cfg80211 [last unloaded: sha256_generic]
> >
> > Pid: 12178, comm: ssh Not tainted 2.6.38-rc5 #227 Dell Inc. MXC061 /0MG532
> > EIP: 0060:[<00000000>] EFLAGS: 0021246 CPU: 0
> > EIP is at 0x0
> > EAX: f5947e00 EBX: f5982f80 ECX: 00000024 EDX: c148fe00
> > ESI: f5947e00 EDI: ebf29e54 EBP: ebf29ee0 ESP: ebf29dd0
> > DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
> > Process ssh (pid: 12178, ti=ebf28000 task=ebefc6c0 task.ti = ebef28000)
> > Stack:
> > c12c83ea 07e2427d 02000000 7b018f79 c117b827 0b996609 7b018f79 5f5f6644
> > f5982f80 00000000 ebf29e58 ebf29ee0 c12cc0e0 00001600 00000000 00000001
> > 03641600 00000000 00000000 00000000 00000000 036423c0 3e6423c0 00000000
> > Call Trace:
> > [<c12c83ea>] ? tcp_connect+0xdd/0x3fd
> > [<c117b827>] ? secure_tcp_sequence_number+0x4f/0x65
> > [<c12cc0e0>] ? tcp_v4_connect+0x3c1/0x417
> > [<c12d6726>] ? inet_stream_connect+0x88/0x1fc
> > [<c1110705>] ? _copy_from_user+0x2b/0x10e
> > [<c129307d>] ? sys_connect+0x70/0x98
> > [<c108df0f>] ? get_empty_filp+0x9f/0x121
> > [<c108dfa0>] ? alloc_file+0xf/0x85
> > [<c1293287>] ? sock_alloc_file+0x97/0xeb
> > [<c108b758>] ? fd_install+0x1b/0x38
> > [<c12932f6>] ? sock_map_fd+0x1b/0x20
> > [<c1293bdf>] ? sys_socketcall+0x9d/0x291
> > [<c1002750>] ? sysenter_do_call+0x12/0x26
> > Code: Bad EIP value
> > EIP: [<00000000>] 0x0 SS:ESP 0068:ebf29dd0
> > CR2: 0000000000000000
I suspect following patch is needed
CC Roland Dreier <roland@purestorage.com> because he fixed the
default_mtu problem in commit ec831ea72ee5d7d47
(net: Add default_mtu() methods to blackhole dst_ops)
Thanks !
[PATCH] net: Add default_advmss() methods to blackhole dst_ops
Commit 0dbaee3b37e118a (net: Abstract default ADVMSS behind an
accessor.) introduced a possible crash in tcp_connect_init(), when
dst->default_advmss() is called from dst_metric_advmss()
Reported-by: George Spelvin <linux@horizon.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Roland Dreier <roland@purestorage.com>
---
net/ipv4/route.c | 7 +++++++
net/ipv6/route.c | 6 ++++++
2 files changed, 13 insertions(+)
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 788a3e7..5edb605 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -2712,6 +2712,12 @@ static unsigned int ipv4_blackhole_default_mtu(const struct dst_entry *dst)
return 0;
}
+static unsigned int ipv4_blackhole_default_advmss(const struct dst_entry *dst)
+{
+ return 256;
+}
+
+
static void ipv4_rt_blackhole_update_pmtu(struct dst_entry *dst, u32 mtu)
{
}
@@ -2722,6 +2728,7 @@ static struct dst_ops ipv4_dst_blackhole_ops = {
.destroy = ipv4_dst_destroy,
.check = ipv4_blackhole_dst_check,
.default_mtu = ipv4_blackhole_default_mtu,
+ .default_advmss = ipv4_blackhole_default_advmss,
.update_pmtu = ipv4_rt_blackhole_update_pmtu,
};
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 1c29f95..2eeeabb 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -118,6 +118,11 @@ static unsigned int ip6_blackhole_default_mtu(const struct dst_entry *dst)
return 0;
}
+static unsigned int ip6_blackhole_default_advmss(const struct dst_entry *dst)
+{
+ return 256;
+}
+
static void ip6_rt_blackhole_update_pmtu(struct dst_entry *dst, u32 mtu)
{
}
@@ -128,6 +133,7 @@ static struct dst_ops ip6_dst_blackhole_ops = {
.destroy = ip6_dst_destroy,
.check = ip6_dst_check,
.default_mtu = ip6_blackhole_default_mtu,
+ .default_advmss = ip6_blackhole_default_advmss,
.update_pmtu = ip6_rt_blackhole_update_pmtu,
};
^ permalink raw reply related
* [PATCH] stmmac: fixed dma lib build when turn-on the debug option
From: Peppe CAVALLARO @ 2011-02-18 10:21 UTC (permalink / raw)
To: netdev@vger.kernel.org; +Cc: Peppe CAVALLARO
This patch fixes a compilation error when build the
dwmac_lib with the DEBUG option enabled.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/stmmac/dwmac_lib.c | 28 ++++++++++++++--------------
1 files changed, 14 insertions(+), 14 deletions(-)
diff --git a/drivers/net/stmmac/dwmac_lib.c b/drivers/net/stmmac/dwmac_lib.c
index d65fab1..e250935 100644
--- a/drivers/net/stmmac/dwmac_lib.c
+++ b/drivers/net/stmmac/dwmac_lib.c
@@ -26,9 +26,9 @@
#undef DWMAC_DMA_DEBUG
#ifdef DWMAC_DMA_DEBUG
-#define DBG(fmt, args...) printk(fmt, ## args)
+#define DWMAC_LIB_DBG(fmt, args...) printk(fmt, ## args)
#else
-#define DBG(fmt, args...) do { } while (0)
+#define DWMAC_LIB_DBG(fmt, args...) do { } while (0)
#endif
/* CSR1 enables the transmit DMA to check for new descriptor */
@@ -152,7 +152,7 @@ int dwmac_dma_interrupt(void __iomem *ioaddr,
/* read the status register (CSR5) */
u32 intr_status = readl(ioaddr + DMA_STATUS);
- DBG(INFO, "%s: [CSR5: 0x%08x]\n", __func__, intr_status);
+ DWMAC_LIB_DBG(KERN_INFO "%s: [CSR5: 0x%08x]\n", __func__, intr_status);
#ifdef DWMAC_DMA_DEBUG
/* It displays the DMA process states (CSR5 register) */
show_tx_process_state(intr_status);
@@ -160,43 +160,43 @@ int dwmac_dma_interrupt(void __iomem *ioaddr,
#endif
/* ABNORMAL interrupts */
if (unlikely(intr_status & DMA_STATUS_AIS)) {
- DBG(INFO, "CSR5[15] DMA ABNORMAL IRQ: ");
+ DWMAC_LIB_DBG(KERN_INFO "CSR5[15] DMA ABNORMAL IRQ: ");
if (unlikely(intr_status & DMA_STATUS_UNF)) {
- DBG(INFO, "transmit underflow\n");
+ DWMAC_LIB_DBG(KERN_INFO "transmit underflow\n");
ret = tx_hard_error_bump_tc;
x->tx_undeflow_irq++;
}
if (unlikely(intr_status & DMA_STATUS_TJT)) {
- DBG(INFO, "transmit jabber\n");
+ DWMAC_LIB_DBG(KERN_INFO "transmit jabber\n");
x->tx_jabber_irq++;
}
if (unlikely(intr_status & DMA_STATUS_OVF)) {
- DBG(INFO, "recv overflow\n");
+ DWMAC_LIB_DBG(KERN_INFO "recv overflow\n");
x->rx_overflow_irq++;
}
if (unlikely(intr_status & DMA_STATUS_RU)) {
- DBG(INFO, "receive buffer unavailable\n");
+ DWMAC_LIB_DBG(KERN_INFO "receive buffer unavailable\n");
x->rx_buf_unav_irq++;
}
if (unlikely(intr_status & DMA_STATUS_RPS)) {
- DBG(INFO, "receive process stopped\n");
+ DWMAC_LIB_DBG(KERN_INFO "receive process stopped\n");
x->rx_process_stopped_irq++;
}
if (unlikely(intr_status & DMA_STATUS_RWT)) {
- DBG(INFO, "receive watchdog\n");
+ DWMAC_LIB_DBG(KERN_INFO "receive watchdog\n");
x->rx_watchdog_irq++;
}
if (unlikely(intr_status & DMA_STATUS_ETI)) {
- DBG(INFO, "transmit early interrupt\n");
+ DWMAC_LIB_DBG(KERN_INFO "transmit early interrupt\n");
x->tx_early_irq++;
}
if (unlikely(intr_status & DMA_STATUS_TPS)) {
- DBG(INFO, "transmit process stopped\n");
+ DWMAC_LIB_DBG(KERN_INFO "transmit process stopped\n");
x->tx_process_stopped_irq++;
ret = tx_hard_error;
}
if (unlikely(intr_status & DMA_STATUS_FBI)) {
- DBG(INFO, "fatal bus error\n");
+ DWMAC_LIB_DBG(KERN_INFO "fatal bus error\n");
x->fatal_bus_error_irq++;
ret = tx_hard_error;
}
@@ -215,7 +215,7 @@ int dwmac_dma_interrupt(void __iomem *ioaddr,
/* Clear the interrupt by writing a logic 1 to the CSR5[15-0] */
writel((intr_status & 0x1ffff), ioaddr + DMA_STATUS);
- DBG(INFO, "\n\n");
+ DWMAC_LIB_DBG(KERN_INFO "\n\n");
return ret;
}
--
1.7.4
^ permalink raw reply related
* You have reached the limit of your email quota.
From: WED MAIL HELP DESK @ 2011-02-18 11:52 UTC (permalink / raw)
To: 111
You have reached the limit of your email quota.
You will not be able to send or receive new mail until you boost
your mailbox size.
Click the below link and fill the form to upgrade your account.
http://www.psanthony.com/contact/use/accountupgrading/form1.html
Technical Support
192.168.0.1
______________ ______________ ______________ ______________
Sent via the KillerWebMail system at cyril.k12.ok.us
^ permalink raw reply
* Re: KIND
From: Mr.David Gurupatham @ 2011-02-18 9:07 UTC (permalink / raw)
To: netdev@vger.kernel.org
My name is David Gurupatham a legal practitioner with David Gurupatham & Associates
in Kuala Lumpur Msia.
I found your contact/profile some where over the Internet and it gave me the
greatest joy, that you are the one I have been looking for. Whom I strongly believe could
execute this project with me. Kindly get back to me for more information
Best regards,
Mr David Gurupatham{Esq}
^ 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