* Re: Question about tcp hash function tcp_hashfn()
From: David Miller @ 2006-06-01 6:24 UTC (permalink / raw)
To: bidulock; +Cc: johnpol, draghuram, linux-kernel, netdev
In-Reply-To: <20060601002221.B21730@openss7.org>
From: "Brian F. G. Bidulock" <bidulock@openss7.org>
Date: Thu, 1 Jun 2006 00:22:21 -0600
> I thought you said you were considering jenkins_3word(), not
> jenkins_2word()?
We could xor some of the inputs in order to use jenkins_2word().
^ permalink raw reply
* Re: Question about tcp hash function tcp_hashfn()
From: Evgeniy Polyakov @ 2006-06-01 6:30 UTC (permalink / raw)
To: David Miller, draghuram, linux-kernel, netdev; +Cc: Brian F. G. Bidulock
In-Reply-To: <20060601001825.A21730@openss7.org>
On Thu, Jun 01, 2006 at 12:18:25AM -0600, Brian F. G. Bidulock (bidulock@openss7.org) wrote:
> Evgeniy,
>
> On Thu, 01 Jun 2006, Evgeniy Polyakov wrote:
> >
> > for (i=0; i<hash_size*iter_num; ++i) {
> > saddr = num2ip(get_random_byte(), get_random_byte(), get_random_byte(), get_random_byte());
> > sport = get_random_word();
>
> You still have a problem: you cannot use a pseudo-random number
> generator to generate the sample set as the pseudo-random number
> generator function itself can interact with the hash.
>
> Try iterating through all 2**48 values or at least a sizeable
> representative subset.
Since pseudo-randomness affects both folded and not folded hash
distribution, it can not end up in different results.
You are right that having test with 2^48 values is really interesting,
but it will take ages on my test machine :)
--
Evgeniy Polyakov
^ permalink raw reply
* Re: 2.6.17-rc5-mm1 lockdep output
From: Arjan van de Ven @ 2006-06-01 6:34 UTC (permalink / raw)
To: Jesse Brandeburg; +Cc: NetDEV list, mingo, Andrew Morton, Linux-Kernel,
In-Reply-To: <4807377b0605311704g44fe10f1oc54315276890071@mail.gmail.com>
On Wed, 2006-05-31 at 17:04 -0700, Jesse Brandeburg wrote:
> well, when running e1000 through some code paths on FC4 +
> 2.6.17-rc5-mm1 + ingo's latest rollup patch, with this lockdep debug
> option enabled I got this:
>
> e1000: eth1: e1000_watchdog_task: NIC Link is Up 1000 Mbps Full Duplex
>
> ======================================
> [ BUG: bad unlock ordering detected! ]
> --------------------------------------
> mDNSResponder/2361 is trying to release lock (&in_dev->mc_list_lock) at:
> [<ffffffff81233f5a>] ip_mc_add_src+0x85/0x1f8
> but the next lock to release is:
> (&im->lock){-+..}, at: [<ffffffff81233f52>] ip_mc_add_src+0x7d/0x1f8
>
> other info that might help us debug this:
> 2 locks held by mDNSResponder/2361:
> #0: (rtnl_mutex){--..}, at: [<ffffffff81253741>] mutex_lock+0x27/0x2c
> #1: (&in_dev->mc_list_lock){-.-?}, at: [<ffffffff81233f14>]
> ip_mc_add_src+0x3f/0x1f8
ok another out of order one in igmp ...
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
---
net/ipv4/igmp.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Index: linux-2.6.17-rc5-mm1.5/net/ipv4/igmp.c
===================================================================
--- linux-2.6.17-rc5-mm1.5.orig/net/ipv4/igmp.c
+++ linux-2.6.17-rc5-mm1.5/net/ipv4/igmp.c
@@ -1646,7 +1646,7 @@ static int ip_mc_add_src(struct in_devic
return -ESRCH;
}
spin_lock_bh(&pmc->lock);
- read_unlock(&in_dev->mc_list_lock);
+ read_unlock_non_nested(&in_dev->mc_list_lock);
#ifdef CONFIG_IP_MULTICAST
sf_markstate(pmc);
^ permalink raw reply
* [patch, -rc5-mm1] lock validator: special locking: net/ipv4/igmp.c #2
From: Ingo Molnar @ 2006-06-01 6:35 UTC (permalink / raw)
To: Jesse Brandeburg
Cc: Linux-Kernel,, Andrew Morton, Arjan van de Ven, NetDEV list
In-Reply-To: <4807377b0605311704g44fe10f1oc54315276890071@mail.gmail.com>
* Jesse Brandeburg <jesse.brandeburg@gmail.com> wrote:
> well, when running e1000 through some code paths on FC4 +
> 2.6.17-rc5-mm1 + ingo's latest rollup patch, with this lockdep debug
> option enabled I got this:
>
> e1000: eth1: e1000_watchdog_task: NIC Link is Up 1000 Mbps Full Duplex
>
> ======================================
> [ BUG: bad unlock ordering detected! ]
> --------------------------------------
> mDNSResponder/2361 is trying to release lock (&in_dev->mc_list_lock) at:
> [<ffffffff81233f5a>] ip_mc_add_src+0x85/0x1f8
ok, could you try the patch below? (i also updated the rollup with this
fix)
Ingo
---------------------
Subject: lock validator: special locking: net/ipv4/igmp.c #2
From: Ingo Molnar <mingo@elte.hu>
another case of non-nested unlocking igmp.c.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
net/ipv4/igmp.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Index: linux/net/ipv4/igmp.c
===================================================================
--- linux.orig/net/ipv4/igmp.c
+++ linux/net/ipv4/igmp.c
@@ -1646,7 +1646,7 @@ static int ip_mc_add_src(struct in_devic
return -ESRCH;
}
spin_lock_bh(&pmc->lock);
- read_unlock(&in_dev->mc_list_lock);
+ read_unlock_non_nested(&in_dev->mc_list_lock);
#ifdef CONFIG_IP_MULTICAST
sf_markstate(pmc);
^ permalink raw reply
* Re: Question about tcp hash function tcp_hashfn()
From: Brian F. G. Bidulock @ 2006-06-01 6:46 UTC (permalink / raw)
To: Evgeniy Polyakov; +Cc: David Miller, draghuram, linux-kernel, netdev
In-Reply-To: <20060601063012.GC28087@2ka.mipt.ru>
Evgeniy,
On Thu, 01 Jun 2006, Evgeniy Polyakov wrote:
>
> Since pseudo-randomness affects both folded and not folded hash
> distribution, it can not end up in different results.
Yes it would, so to rule out pseudo-random effects the pseudo-
random number generator must be removed.
>
> You are right that having test with 2^48 values is really interesting,
> but it will take ages on my test machine :)
Try a usable subset; no pseudo-random number generator.
^ permalink raw reply
* Re: Question about tcp hash function tcp_hashfn()
From: Evgeniy Polyakov @ 2006-06-01 7:01 UTC (permalink / raw)
To: David Miller, draghuram, linux-kernel, netdev
In-Reply-To: <20060601004608.C21730@openss7.org>
On Thu, Jun 01, 2006 at 12:46:08AM -0600, Brian F. G. Bidulock (bidulock@openss7.org) wrote:
> > Since pseudo-randomness affects both folded and not folded hash
> > distribution, it can not end up in different results.
>
> Yes it would, so to rule out pseudo-random effects the pseudo-
> random number generator must be removed.
>
> >
> > You are right that having test with 2^48 values is really interesting,
> > but it will take ages on my test machine :)
>
> Try a usable subset; no pseudo-random number generator.
I've run it for 2^30 - the same result: folded and not folded Jenkins
hash behave the same and still both results produce exactly the same
artifacts compared to XOR hash.
Btw, XOR hash, as completely stateless, can be used to show how
Linux pseudo-random generator works for given subset - it's average of
distribution is very good.
--
Evgeniy Polyakov
^ permalink raw reply
* Re: [Bugme-new] [Bug 6613] New: iptables broken on 32-bit PReP (ARCH=ppc)
From: Meelis Roos @ 2006-06-01 7:03 UTC (permalink / raw)
To: Patrick McHardy; +Cc: Andrew Morton, bugme-daemon, netdev
In-Reply-To: <4475FCFC.5000701@trash.net>
>>> modprobe iptable_filter (errors out with Invalid Argument)
>>> iptables -t nat -A POSTROUTING -s 10.0.0.0/8 -j SNAT --to 192.168.1.1 (usually
>>> errors out with Invalid Argument, sometimes succeeds, when succeeds then the
>>> rule works fine)
>
> Meelis, it would really help if you could try 2.6.16 and in case
> that doesn't work 2.6.15 to give an idea about whether this is a
> recent regression or an old problem. We had a number of changes
> in this area in the last two kernel versions that could be related.
Have not gotten 2.6.15 to work with one evening of tinkering - the irq
patch was not sufficent, there is something more broken in booting that
I dodn't figure out yet. So no test results for 2.6.15 yet.
--
Meelis Roos (mroos@linux.ee)
^ permalink raw reply
* Re: Question about tcp hash function tcp_hashfn()
From: Brian F. G. Bidulock @ 2006-06-01 7:11 UTC (permalink / raw)
To: Evgeniy Polyakov; +Cc: David Miller, draghuram, linux-kernel, netdev
In-Reply-To: <20060601070136.GA754@2ka.mipt.ru>
Evgeniy,
On Thu, 01 Jun 2006, Evgeniy Polyakov wrote:
> On Thu, Jun 01, 2006 at 12:46:08AM -0600, Brian F. G. Bidulock (bidulock@openss7.org) wrote:
> > > Since pseudo-randomness affects both folded and not folded hash
> > > distribution, it can not end up in different results.
> >
> > Yes it would, so to rule out pseudo-random effects the pseudo-
> > random number generator must be removed.
> >
> > >
> > > You are right that having test with 2^48 values is really interesting,
> > > but it will take ages on my test machine :)
> >
> > Try a usable subset; no pseudo-random number generator.
>
> I've run it for 2^30 - the same result: folded and not folded Jenkins
> hash behave the same and still both results produce exactly the same
> artifacts compared to XOR hash.
But not without the pseudo-random number generation... ?
>
> Btw, XOR hash, as completely stateless, can be used to show how
> Linux pseudo-random generator works for given subset - it's average of
> distribution is very good.
But its distribution might auto-correlate with the Jenkins function.
The only way to be sure is to remove the pseudo-random number generator.
Just try incrementing from, say, 10.0.0.0:10000 up, resetting port number
to 10000 at 16000, and just incrementing the IP address when the port
number wraps, instead of pseudo-random, through 2^30 loops for both.
If the same artifacts emerge, I give in.
Can you show the same artifacts for jenkins_3word?
^ permalink raw reply
* [PATCH netdev-2.6#upstream] net: au1000_eth: PHY framework conversion
From: Herbert Valerio Riedel @ 2006-06-01 7:41 UTC (permalink / raw)
To: netdev; +Cc: linux-mips, jgarzik, ppopov, sshtylyov, ralf
convert au1000_eth driver to use PHY framework and garbage collected
functions and identifiers that became unused/obsolete in the process
Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
---
this is a resend of
http://marc.theaimsgroup.com/?l=linux-netdev&m=114746547301867&w=2
rebased&handmerged to netdev-2.6#upstream (i.e. on top of sergei's
probe code cleanup patch)
drivers/net/Kconfig | 1
drivers/net/au1000_eth.c | 1602 +++++++++++-----------------------------------
drivers/net/au1000_eth.h | 134 ----
3 files changed, 378 insertions(+), 1359 deletions(-)
616e5a012a927d401befcbed37b77c3ea42da5a5
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index f499a3b..520765d 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -447,6 +447,7 @@ config MIPS_GT96100ETH
config MIPS_AU1X00_ENET
bool "MIPS AU1000 Ethernet support"
depends on NET_ETHERNET && SOC_AU1X00
+ select PHYLIB
select CRC32
help
If you have an Alchemy Semi AU1X00 based system
diff --git a/drivers/net/au1000_eth.c b/drivers/net/au1000_eth.c
index e1fe960..038d5fc 100644
--- a/drivers/net/au1000_eth.c
+++ b/drivers/net/au1000_eth.c
@@ -9,6 +9,9 @@
* Update: 2004 Bjoern Riemer, riemer@fokus.fraunhofer.de
* or riemer@riemer-nt.de: fixed the link beat detection with
* ioctls (SIOCGMIIPHY)
+ * Copyright 2006 Herbert Valerio Riedel <hvr@gnu.org>
+ * converted to use linux-2.6.x's PHY framework
+ *
* Author: MontaVista Software, Inc.
* ppopov@mvista.com or source@mvista.com
*
@@ -53,6 +56,7 @@ #include <linux/mii.h>
#include <linux/skbuff.h>
#include <linux/delay.h>
#include <linux/crc32.h>
+#include <linux/phy.h>
#include <asm/mipsregs.h>
#include <asm/irq.h>
#include <asm/io.h>
@@ -88,17 +92,15 @@ static int au1000_tx(struct sk_buff *, s
static int au1000_rx(struct net_device *);
static irqreturn_t au1000_interrupt(int, void *, struct pt_regs *);
static void au1000_tx_timeout(struct net_device *);
-static int au1000_set_config(struct net_device *dev, struct ifmap *map);
static void set_rx_mode(struct net_device *);
static struct net_device_stats *au1000_get_stats(struct net_device *);
-static void au1000_timer(unsigned long);
static int au1000_ioctl(struct net_device *, struct ifreq *, int);
static int mdio_read(struct net_device *, int, int);
static void mdio_write(struct net_device *, int, int, u16);
-static void dump_mii(struct net_device *dev, int phy_id);
+static void au1000_adjust_link(struct net_device *);
+static void enable_mac(struct net_device *, int);
// externs
-extern void ack_rise_edge_irq(unsigned int);
extern int get_ethernet_addr(char *ethernet_addr);
extern void str2eaddr(unsigned char *ea, unsigned char *str);
extern char * __init prom_getcmdline(void);
@@ -126,705 +128,83 @@ static unsigned char au1000_mac_addr[6]
0x00, 0x50, 0xc2, 0x0c, 0x30, 0x00
};
-#define nibswap(x) ((((x) >> 4) & 0x0f) | (((x) << 4) & 0xf0))
-#define RUN_AT(x) (jiffies + (x))
-
-// For reading/writing 32-bit words from/to DMA memory
-#define cpu_to_dma32 cpu_to_be32
-#define dma32_to_cpu be32_to_cpu
-
struct au1000_private *au_macs[NUM_ETH_INTERFACES];
-/* FIXME
- * All of the PHY code really should be detached from the MAC
- * code.
+/*
+ * board-specific configurations
+ *
+ * PHY detection algorithm
+ *
+ * If AU1XXX_PHY_STATIC_CONFIG is undefined, the PHY setup is
+ * autodetected:
+ *
+ * mii_probe() first searches the current MAC's MII bus for a PHY,
+ * selecting the first (or last, if AU1XXX_PHY_SEARCH_HIGHEST_ADDR is
+ * defined) PHY address not already claimed by another netdev.
+ *
+ * If nothing was found that way when searching for the 2nd ethernet
+ * controller's PHY and AU1XXX_PHY1_SEARCH_ON_MAC0 is defined, then
+ * the first MII bus is searched as well for an unclaimed PHY; this is
+ * needed in case of a dual-PHY accessible only through the MAC0's MII
+ * bus.
+ *
+ * Finally, if no PHY is found, then the corresponding ethernet
+ * controller is not registered to the network subsystem.
*/
-/* Default advertise */
-#define GENMII_DEFAULT_ADVERTISE \
- ADVERTISED_10baseT_Half | ADVERTISED_10baseT_Full | \
- ADVERTISED_100baseT_Half | ADVERTISED_100baseT_Full | \
- ADVERTISED_Autoneg
-
-#define GENMII_DEFAULT_FEATURES \
- SUPPORTED_10baseT_Half | SUPPORTED_10baseT_Full | \
- SUPPORTED_100baseT_Half | SUPPORTED_100baseT_Full | \
- SUPPORTED_Autoneg
-
-int bcm_5201_init(struct net_device *dev, int phy_addr)
-{
- s16 data;
-
- /* Stop auto-negotiation */
- data = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, data & ~MII_CNTL_AUTO);
-
- /* Set advertisement to 10/100 and Half/Full duplex
- * (full capabilities) */
- data = mdio_read(dev, phy_addr, MII_ANADV);
- data |= MII_NWAY_TX | MII_NWAY_TX_FDX | MII_NWAY_T_FDX | MII_NWAY_T;
- mdio_write(dev, phy_addr, MII_ANADV, data);
-
- /* Restart auto-negotiation */
- data = mdio_read(dev, phy_addr, MII_CONTROL);
- data |= MII_CNTL_RST_AUTO | MII_CNTL_AUTO;
- mdio_write(dev, phy_addr, MII_CONTROL, data);
-
- if (au1000_debug > 4)
- dump_mii(dev, phy_addr);
- return 0;
-}
-
-int bcm_5201_reset(struct net_device *dev, int phy_addr)
-{
- s16 mii_control, timeout;
-
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, mii_control | MII_CNTL_RESET);
- mdelay(1);
- for (timeout = 100; timeout > 0; --timeout) {
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- if ((mii_control & MII_CNTL_RESET) == 0)
- break;
- mdelay(1);
- }
- if (mii_control & MII_CNTL_RESET) {
- printk(KERN_ERR "%s PHY reset timeout !\n", dev->name);
- return -1;
- }
- return 0;
-}
-
-int
-bcm_5201_status(struct net_device *dev, int phy_addr, u16 *link, u16 *speed)
-{
- u16 mii_data;
- struct au1000_private *aup;
-
- if (!dev) {
- printk(KERN_ERR "bcm_5201_status error: NULL dev\n");
- return -1;
- }
- aup = (struct au1000_private *) dev->priv;
-
- mii_data = mdio_read(dev, aup->phy_addr, MII_STATUS);
- if (mii_data & MII_STAT_LINK) {
- *link = 1;
- mii_data = mdio_read(dev, aup->phy_addr, MII_AUX_CNTRL);
- if (mii_data & MII_AUX_100) {
- if (mii_data & MII_AUX_FDX) {
- *speed = IF_PORT_100BASEFX;
- dev->if_port = IF_PORT_100BASEFX;
- }
- else {
- *speed = IF_PORT_100BASETX;
- dev->if_port = IF_PORT_100BASETX;
- }
- }
- else {
- *speed = IF_PORT_10BASET;
- dev->if_port = IF_PORT_10BASET;
- }
-
- }
- else {
- *link = 0;
- *speed = 0;
- dev->if_port = IF_PORT_UNKNOWN;
- }
- return 0;
-}
-
-int lsi_80227_init(struct net_device *dev, int phy_addr)
-{
- if (au1000_debug > 4)
- printk("lsi_80227_init\n");
-
- /* restart auto-negotiation */
- mdio_write(dev, phy_addr, MII_CONTROL,
- MII_CNTL_F100 | MII_CNTL_AUTO | MII_CNTL_RST_AUTO); // | MII_CNTL_FDX);
- mdelay(1);
-
- /* set up LEDs to correct display */
-#ifdef CONFIG_MIPS_MTX1
- mdio_write(dev, phy_addr, 17, 0xff80);
-#else
- mdio_write(dev, phy_addr, 17, 0xffc0);
-#endif
-
- if (au1000_debug > 4)
- dump_mii(dev, phy_addr);
- return 0;
-}
-
-int lsi_80227_reset(struct net_device *dev, int phy_addr)
-{
- s16 mii_control, timeout;
-
- if (au1000_debug > 4) {
- printk("lsi_80227_reset\n");
- dump_mii(dev, phy_addr);
- }
-
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, mii_control | MII_CNTL_RESET);
- mdelay(1);
- for (timeout = 100; timeout > 0; --timeout) {
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- if ((mii_control & MII_CNTL_RESET) == 0)
- break;
- mdelay(1);
- }
- if (mii_control & MII_CNTL_RESET) {
- printk(KERN_ERR "%s PHY reset timeout !\n", dev->name);
- return -1;
- }
- return 0;
-}
-
-int
-lsi_80227_status(struct net_device *dev, int phy_addr, u16 *link, u16 *speed)
-{
- u16 mii_data;
- struct au1000_private *aup;
-
- if (!dev) {
- printk(KERN_ERR "lsi_80227_status error: NULL dev\n");
- return -1;
- }
- aup = (struct au1000_private *) dev->priv;
-
- mii_data = mdio_read(dev, aup->phy_addr, MII_STATUS);
- if (mii_data & MII_STAT_LINK) {
- *link = 1;
- mii_data = mdio_read(dev, aup->phy_addr, MII_LSI_PHY_STAT);
- if (mii_data & MII_LSI_PHY_STAT_SPD) {
- if (mii_data & MII_LSI_PHY_STAT_FDX) {
- *speed = IF_PORT_100BASEFX;
- dev->if_port = IF_PORT_100BASEFX;
- }
- else {
- *speed = IF_PORT_100BASETX;
- dev->if_port = IF_PORT_100BASETX;
- }
- }
- else {
- *speed = IF_PORT_10BASET;
- dev->if_port = IF_PORT_10BASET;
- }
-
- }
- else {
- *link = 0;
- *speed = 0;
- dev->if_port = IF_PORT_UNKNOWN;
- }
- return 0;
-}
-
-int am79c901_init(struct net_device *dev, int phy_addr)
-{
- printk("am79c901_init\n");
- return 0;
-}
-
-int am79c901_reset(struct net_device *dev, int phy_addr)
-{
- printk("am79c901_reset\n");
- return 0;
-}
-
-int
-am79c901_status(struct net_device *dev, int phy_addr, u16 *link, u16 *speed)
-{
- return 0;
-}
-
-int am79c874_init(struct net_device *dev, int phy_addr)
-{
- s16 data;
-
- /* 79c874 has quit resembled bit assignments to BCM5201 */
- if (au1000_debug > 4)
- printk("am79c847_init\n");
-
- /* Stop auto-negotiation */
- data = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, data & ~MII_CNTL_AUTO);
-
- /* Set advertisement to 10/100 and Half/Full duplex
- * (full capabilities) */
- data = mdio_read(dev, phy_addr, MII_ANADV);
- data |= MII_NWAY_TX | MII_NWAY_TX_FDX | MII_NWAY_T_FDX | MII_NWAY_T;
- mdio_write(dev, phy_addr, MII_ANADV, data);
-
- /* Restart auto-negotiation */
- data = mdio_read(dev, phy_addr, MII_CONTROL);
- data |= MII_CNTL_RST_AUTO | MII_CNTL_AUTO;
-
- mdio_write(dev, phy_addr, MII_CONTROL, data);
-
- if (au1000_debug > 4) dump_mii(dev, phy_addr);
- return 0;
-}
-
-int am79c874_reset(struct net_device *dev, int phy_addr)
-{
- s16 mii_control, timeout;
-
- if (au1000_debug > 4)
- printk("am79c874_reset\n");
-
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, mii_control | MII_CNTL_RESET);
- mdelay(1);
- for (timeout = 100; timeout > 0; --timeout) {
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- if ((mii_control & MII_CNTL_RESET) == 0)
- break;
- mdelay(1);
- }
- if (mii_control & MII_CNTL_RESET) {
- printk(KERN_ERR "%s PHY reset timeout !\n", dev->name);
- return -1;
- }
- return 0;
-}
-
-int
-am79c874_status(struct net_device *dev, int phy_addr, u16 *link, u16 *speed)
-{
- u16 mii_data;
- struct au1000_private *aup;
-
- // printk("am79c874_status\n");
- if (!dev) {
- printk(KERN_ERR "am79c874_status error: NULL dev\n");
- return -1;
- }
-
- aup = (struct au1000_private *) dev->priv;
- mii_data = mdio_read(dev, aup->phy_addr, MII_STATUS);
+/* autodetection defaults */
+#undef AU1XXX_PHY_SEARCH_HIGHEST_ADDR
+#define AU1XXX_PHY1_SEARCH_ON_MAC0
- if (mii_data & MII_STAT_LINK) {
- *link = 1;
- mii_data = mdio_read(dev, aup->phy_addr, MII_AMD_PHY_STAT);
- if (mii_data & MII_AMD_PHY_STAT_SPD) {
- if (mii_data & MII_AMD_PHY_STAT_FDX) {
- *speed = IF_PORT_100BASEFX;
- dev->if_port = IF_PORT_100BASEFX;
- }
- else {
- *speed = IF_PORT_100BASETX;
- dev->if_port = IF_PORT_100BASETX;
- }
- }
- else {
- *speed = IF_PORT_10BASET;
- dev->if_port = IF_PORT_10BASET;
- }
-
- }
- else {
- *link = 0;
- *speed = 0;
- dev->if_port = IF_PORT_UNKNOWN;
- }
- return 0;
-}
-
-int lxt971a_init(struct net_device *dev, int phy_addr)
-{
- if (au1000_debug > 4)
- printk("lxt971a_init\n");
-
- /* restart auto-negotiation */
- mdio_write(dev, phy_addr, MII_CONTROL,
- MII_CNTL_F100 | MII_CNTL_AUTO | MII_CNTL_RST_AUTO | MII_CNTL_FDX);
-
- /* set up LEDs to correct display */
- mdio_write(dev, phy_addr, 20, 0x0422);
-
- if (au1000_debug > 4)
- dump_mii(dev, phy_addr);
- return 0;
-}
-
-int lxt971a_reset(struct net_device *dev, int phy_addr)
-{
- s16 mii_control, timeout;
-
- if (au1000_debug > 4) {
- printk("lxt971a_reset\n");
- dump_mii(dev, phy_addr);
- }
-
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, mii_control | MII_CNTL_RESET);
- mdelay(1);
- for (timeout = 100; timeout > 0; --timeout) {
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- if ((mii_control & MII_CNTL_RESET) == 0)
- break;
- mdelay(1);
- }
- if (mii_control & MII_CNTL_RESET) {
- printk(KERN_ERR "%s PHY reset timeout !\n", dev->name);
- return -1;
- }
- return 0;
-}
-
-int
-lxt971a_status(struct net_device *dev, int phy_addr, u16 *link, u16 *speed)
-{
- u16 mii_data;
- struct au1000_private *aup;
-
- if (!dev) {
- printk(KERN_ERR "lxt971a_status error: NULL dev\n");
- return -1;
- }
- aup = (struct au1000_private *) dev->priv;
-
- mii_data = mdio_read(dev, aup->phy_addr, MII_STATUS);
- if (mii_data & MII_STAT_LINK) {
- *link = 1;
- mii_data = mdio_read(dev, aup->phy_addr, MII_INTEL_PHY_STAT);
- if (mii_data & MII_INTEL_PHY_STAT_SPD) {
- if (mii_data & MII_INTEL_PHY_STAT_FDX) {
- *speed = IF_PORT_100BASEFX;
- dev->if_port = IF_PORT_100BASEFX;
- }
- else {
- *speed = IF_PORT_100BASETX;
- dev->if_port = IF_PORT_100BASETX;
- }
- }
- else {
- *speed = IF_PORT_10BASET;
- dev->if_port = IF_PORT_10BASET;
- }
-
- }
- else {
- *link = 0;
- *speed = 0;
- dev->if_port = IF_PORT_UNKNOWN;
- }
- return 0;
-}
-
-int ks8995m_init(struct net_device *dev, int phy_addr)
-{
- s16 data;
-
-// printk("ks8995m_init\n");
- /* Stop auto-negotiation */
- data = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, data & ~MII_CNTL_AUTO);
-
- /* Set advertisement to 10/100 and Half/Full duplex
- * (full capabilities) */
- data = mdio_read(dev, phy_addr, MII_ANADV);
- data |= MII_NWAY_TX | MII_NWAY_TX_FDX | MII_NWAY_T_FDX | MII_NWAY_T;
- mdio_write(dev, phy_addr, MII_ANADV, data);
-
- /* Restart auto-negotiation */
- data = mdio_read(dev, phy_addr, MII_CONTROL);
- data |= MII_CNTL_RST_AUTO | MII_CNTL_AUTO;
- mdio_write(dev, phy_addr, MII_CONTROL, data);
-
- if (au1000_debug > 4) dump_mii(dev, phy_addr);
-
- return 0;
-}
-
-int ks8995m_reset(struct net_device *dev, int phy_addr)
-{
- s16 mii_control, timeout;
-
-// printk("ks8995m_reset\n");
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, mii_control | MII_CNTL_RESET);
- mdelay(1);
- for (timeout = 100; timeout > 0; --timeout) {
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- if ((mii_control & MII_CNTL_RESET) == 0)
- break;
- mdelay(1);
- }
- if (mii_control & MII_CNTL_RESET) {
- printk(KERN_ERR "%s PHY reset timeout !\n", dev->name);
- return -1;
- }
- return 0;
-}
-
-int ks8995m_status(struct net_device *dev, int phy_addr, u16 *link, u16 *speed)
-{
- u16 mii_data;
- struct au1000_private *aup;
-
- if (!dev) {
- printk(KERN_ERR "ks8995m_status error: NULL dev\n");
- return -1;
- }
- aup = (struct au1000_private *) dev->priv;
-
- mii_data = mdio_read(dev, aup->phy_addr, MII_STATUS);
- if (mii_data & MII_STAT_LINK) {
- *link = 1;
- mii_data = mdio_read(dev, aup->phy_addr, MII_AUX_CNTRL);
- if (mii_data & MII_AUX_100) {
- if (mii_data & MII_AUX_FDX) {
- *speed = IF_PORT_100BASEFX;
- dev->if_port = IF_PORT_100BASEFX;
- }
- else {
- *speed = IF_PORT_100BASETX;
- dev->if_port = IF_PORT_100BASETX;
- }
- }
- else {
- *speed = IF_PORT_10BASET;
- dev->if_port = IF_PORT_10BASET;
- }
-
- }
- else {
- *link = 0;
- *speed = 0;
- dev->if_port = IF_PORT_UNKNOWN;
- }
- return 0;
-}
-
-int
-smsc_83C185_init (struct net_device *dev, int phy_addr)
-{
- s16 data;
-
- if (au1000_debug > 4)
- printk("smsc_83C185_init\n");
-
- /* Stop auto-negotiation */
- data = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, data & ~MII_CNTL_AUTO);
-
- /* Set advertisement to 10/100 and Half/Full duplex
- * (full capabilities) */
- data = mdio_read(dev, phy_addr, MII_ANADV);
- data |= MII_NWAY_TX | MII_NWAY_TX_FDX | MII_NWAY_T_FDX | MII_NWAY_T;
- mdio_write(dev, phy_addr, MII_ANADV, data);
-
- /* Restart auto-negotiation */
- data = mdio_read(dev, phy_addr, MII_CONTROL);
- data |= MII_CNTL_RST_AUTO | MII_CNTL_AUTO;
-
- mdio_write(dev, phy_addr, MII_CONTROL, data);
-
- if (au1000_debug > 4) dump_mii(dev, phy_addr);
- return 0;
-}
-
-int
-smsc_83C185_reset (struct net_device *dev, int phy_addr)
-{
- s16 mii_control, timeout;
-
- if (au1000_debug > 4)
- printk("smsc_83C185_reset\n");
-
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- mdio_write(dev, phy_addr, MII_CONTROL, mii_control | MII_CNTL_RESET);
- mdelay(1);
- for (timeout = 100; timeout > 0; --timeout) {
- mii_control = mdio_read(dev, phy_addr, MII_CONTROL);
- if ((mii_control & MII_CNTL_RESET) == 0)
- break;
- mdelay(1);
- }
- if (mii_control & MII_CNTL_RESET) {
- printk(KERN_ERR "%s PHY reset timeout !\n", dev->name);
- return -1;
- }
- return 0;
-}
-
-int
-smsc_83C185_status (struct net_device *dev, int phy_addr, u16 *link, u16 *speed)
-{
- u16 mii_data;
- struct au1000_private *aup;
-
- if (!dev) {
- printk(KERN_ERR "smsc_83C185_status error: NULL dev\n");
- return -1;
- }
-
- aup = (struct au1000_private *) dev->priv;
- mii_data = mdio_read(dev, aup->phy_addr, MII_STATUS);
-
- if (mii_data & MII_STAT_LINK) {
- *link = 1;
- mii_data = mdio_read(dev, aup->phy_addr, 0x1f);
- if (mii_data & (1<<3)) {
- if (mii_data & (1<<4)) {
- *speed = IF_PORT_100BASEFX;
- dev->if_port = IF_PORT_100BASEFX;
- }
- else {
- *speed = IF_PORT_100BASETX;
- dev->if_port = IF_PORT_100BASETX;
- }
- }
- else {
- *speed = IF_PORT_10BASET;
- dev->if_port = IF_PORT_10BASET;
- }
- }
- else {
- *link = 0;
- *speed = 0;
- dev->if_port = IF_PORT_UNKNOWN;
- }
- return 0;
-}
-
-
-#ifdef CONFIG_MIPS_BOSPORUS
-int stub_init(struct net_device *dev, int phy_addr)
-{
- //printk("PHY stub_init\n");
- return 0;
-}
-
-int stub_reset(struct net_device *dev, int phy_addr)
-{
- //printk("PHY stub_reset\n");
- return 0;
-}
-
-int
-stub_status(struct net_device *dev, int phy_addr, u16 *link, u16 *speed)
-{
- //printk("PHY stub_status\n");
- *link = 1;
- /* hmmm, revisit */
- *speed = IF_PORT_100BASEFX;
- dev->if_port = IF_PORT_100BASEFX;
- return 0;
-}
-#endif
-
-struct phy_ops bcm_5201_ops = {
- bcm_5201_init,
- bcm_5201_reset,
- bcm_5201_status,
-};
-
-struct phy_ops am79c874_ops = {
- am79c874_init,
- am79c874_reset,
- am79c874_status,
-};
-
-struct phy_ops am79c901_ops = {
- am79c901_init,
- am79c901_reset,
- am79c901_status,
-};
-
-struct phy_ops lsi_80227_ops = {
- lsi_80227_init,
- lsi_80227_reset,
- lsi_80227_status,
-};
-
-struct phy_ops lxt971a_ops = {
- lxt971a_init,
- lxt971a_reset,
- lxt971a_status,
-};
+/* static PHY setup
+ *
+ * most boards PHY setup should be detectable properly with the
+ * autodetection algorithm in mii_probe(), but in some cases (e.g. if
+ * you have a switch attached, or want to use the PHY's interrupt
+ * notification capabilities) you can provide a static PHY
+ * configuration here
+ *
+ * IRQs may only be set, if a PHY address was configured
+ * If a PHY address is given, also a bus id is required to be set
+ *
+ * ps: make sure the used irqs are configured properly in the board
+ * specific irq-map
+ */
-struct phy_ops ks8995m_ops = {
- ks8995m_init,
- ks8995m_reset,
- ks8995m_status,
-};
+#if defined(CONFIG_MIPS_BOSPORUS)
+/*
+ * Micrel/Kendin 5 port switch attached to MAC0,
+ * MAC0 is associated with PHY address 5 (== WAN port)
+ * MAC1 is not associated with any PHY, since it's connected directly
+ * to the switch.
+ * no interrupts are used
+ */
+# define AU1XXX_PHY_STATIC_CONFIG
-struct phy_ops smsc_83C185_ops = {
- smsc_83C185_init,
- smsc_83C185_reset,
- smsc_83C185_status,
-};
+# define AU1XXX_PHY0_ADDR 5
+# define AU1XXX_PHY0_BUSID 0
+# undef AU1XXX_PHY0_IRQ
-#ifdef CONFIG_MIPS_BOSPORUS
-struct phy_ops stub_ops = {
- stub_init,
- stub_reset,
- stub_status,
-};
+# undef AU1XXX_PHY1_ADDR
+# undef AU1XXX_PHY1_BUSID
+# undef AU1XXX_PHY1_IRQ
#endif
-static struct mii_chip_info {
- const char * name;
- u16 phy_id0;
- u16 phy_id1;
- struct phy_ops *phy_ops;
- int dual_phy;
-} mii_chip_table[] = {
- {"Broadcom BCM5201 10/100 BaseT PHY",0x0040,0x6212, &bcm_5201_ops,0},
- {"Broadcom BCM5221 10/100 BaseT PHY",0x0040,0x61e4, &bcm_5201_ops,0},
- {"Broadcom BCM5222 10/100 BaseT PHY",0x0040,0x6322, &bcm_5201_ops,1},
- {"NS DP83847 PHY", 0x2000, 0x5c30, &bcm_5201_ops ,0},
- {"AMD 79C901 HomePNA PHY",0x0000,0x35c8, &am79c901_ops,0},
- {"AMD 79C874 10/100 BaseT PHY",0x0022,0x561b, &am79c874_ops,0},
- {"LSI 80227 10/100 BaseT PHY",0x0016,0xf840, &lsi_80227_ops,0},
- {"Intel LXT971A Dual Speed PHY",0x0013,0x78e2, &lxt971a_ops,0},
- {"Kendin KS8995M 10/100 BaseT PHY",0x0022,0x1450, &ks8995m_ops,0},
- {"SMSC LAN83C185 10/100 BaseT PHY",0x0007,0xc0a3, &smsc_83C185_ops,0},
-#ifdef CONFIG_MIPS_BOSPORUS
- {"Stub", 0x1234, 0x5678, &stub_ops },
+#if defined(AU1XXX_PHY0_BUSID) && (AU1XXX_PHY0_BUSID > 0)
+# error MAC0-associated PHY attached 2nd MACs MII bus not supported yet
#endif
- {0,},
-};
-static int mdio_read(struct net_device *dev, int phy_id, int reg)
+/*
+ * MII operations
+ */
+static int mdio_read(struct net_device *dev, int phy_addr, int reg)
{
struct au1000_private *aup = (struct au1000_private *) dev->priv;
- volatile u32 *mii_control_reg;
- volatile u32 *mii_data_reg;
+ volatile u32 *const mii_control_reg = &aup->mac->mii_control;
+ volatile u32 *const mii_data_reg = &aup->mac->mii_data;
u32 timedout = 20;
u32 mii_control;
- #ifdef CONFIG_BCM5222_DUAL_PHY
- /* First time we probe, it's for the mac0 phy.
- * Since we haven't determined yet that we have a dual phy,
- * aup->mii->mii_control_reg won't be setup and we'll
- * default to the else statement.
- * By the time we probe for the mac1 phy, the mii_control_reg
- * will be setup to be the address of the mac0 phy control since
- * both phys are controlled through mac0.
- */
- if (aup->mii && aup->mii->mii_control_reg) {
- mii_control_reg = aup->mii->mii_control_reg;
- mii_data_reg = aup->mii->mii_data_reg;
- }
- else if (au_macs[0]->mii && au_macs[0]->mii->mii_control_reg) {
- /* assume both phys are controlled through mac0 */
- mii_control_reg = au_macs[0]->mii->mii_control_reg;
- mii_data_reg = au_macs[0]->mii->mii_data_reg;
- }
- else
- #endif
- {
- /* default control and data reg addresses */
- mii_control_reg = &aup->mac->mii_control;
- mii_data_reg = &aup->mac->mii_data;
- }
-
while (*mii_control_reg & MAC_MII_BUSY) {
mdelay(1);
if (--timedout == 0) {
@@ -835,7 +215,7 @@ static int mdio_read(struct net_device *
}
mii_control = MAC_SET_MII_SELECT_REG(reg) |
- MAC_SET_MII_SELECT_PHY(phy_id) | MAC_MII_READ;
+ MAC_SET_MII_SELECT_PHY(phy_addr) | MAC_MII_READ;
*mii_control_reg = mii_control;
@@ -851,32 +231,14 @@ static int mdio_read(struct net_device *
return (int)*mii_data_reg;
}
-static void mdio_write(struct net_device *dev, int phy_id, int reg, u16 value)
+static void mdio_write(struct net_device *dev, int phy_addr, int reg, u16 value)
{
struct au1000_private *aup = (struct au1000_private *) dev->priv;
- volatile u32 *mii_control_reg;
- volatile u32 *mii_data_reg;
+ volatile u32 *const mii_control_reg = &aup->mac->mii_control;
+ volatile u32 *const mii_data_reg = &aup->mac->mii_data;
u32 timedout = 20;
u32 mii_control;
- #ifdef CONFIG_BCM5222_DUAL_PHY
- if (aup->mii && aup->mii->mii_control_reg) {
- mii_control_reg = aup->mii->mii_control_reg;
- mii_data_reg = aup->mii->mii_data_reg;
- }
- else if (au_macs[0]->mii && au_macs[0]->mii->mii_control_reg) {
- /* assume both phys are controlled through mac0 */
- mii_control_reg = au_macs[0]->mii->mii_control_reg;
- mii_data_reg = au_macs[0]->mii->mii_data_reg;
- }
- else
- #endif
- {
- /* default control and data reg addresses */
- mii_control_reg = &aup->mac->mii_control;
- mii_data_reg = &aup->mac->mii_data;
- }
-
while (*mii_control_reg & MAC_MII_BUSY) {
mdelay(1);
if (--timedout == 0) {
@@ -887,165 +249,145 @@ static void mdio_write(struct net_device
}
mii_control = MAC_SET_MII_SELECT_REG(reg) |
- MAC_SET_MII_SELECT_PHY(phy_id) | MAC_MII_WRITE;
+ MAC_SET_MII_SELECT_PHY(phy_addr) | MAC_MII_WRITE;
*mii_data_reg = value;
*mii_control_reg = mii_control;
}
+static int mdiobus_read(struct mii_bus *bus, int phy_addr, int regnum)
+{
+ /* WARNING: bus->phy_map[phy_addr].attached_dev == dev does
+ * _NOT_ hold (e.g. when PHY is accessed through other MAC's MII bus) */
+ struct net_device *const dev = bus->priv;
+
+ enable_mac(dev, 0); /* make sure the MAC associated with this
+ * mii_bus is enabled */
+ return mdio_read(dev, phy_addr, regnum);
+}
-static void dump_mii(struct net_device *dev, int phy_id)
+static int mdiobus_write(struct mii_bus *bus, int phy_addr, int regnum,
+ u16 value)
{
- int i, val;
+ struct net_device *const dev = bus->priv;
- for (i = 0; i < 7; i++) {
- if ((val = mdio_read(dev, phy_id, i)) >= 0)
- printk("%s: MII Reg %d=%x\n", dev->name, i, val);
- }
- for (i = 16; i < 25; i++) {
- if ((val = mdio_read(dev, phy_id, i)) >= 0)
- printk("%s: MII Reg %d=%x\n", dev->name, i, val);
- }
+ enable_mac(dev, 0); /* make sure the MAC associated with this
+ * mii_bus is enabled */
+ mdio_write(dev, phy_addr, regnum, value);
+ return 0;
}
-static int mii_probe (struct net_device * dev)
+static int mdiobus_reset(struct mii_bus *bus)
{
- struct au1000_private *aup = (struct au1000_private *) dev->priv;
- int phy_addr;
-#ifdef CONFIG_MIPS_BOSPORUS
- int phy_found=0;
-#endif
+ struct net_device *const dev = bus->priv;
- /* search for total of 32 possible mii phy addresses */
- for (phy_addr = 0; phy_addr < 32; phy_addr++) {
- u16 mii_status;
- u16 phy_id0, phy_id1;
- int i;
+ enable_mac(dev, 0); /* make sure the MAC associated with this
+ * mii_bus is enabled */
+ return 0;
+}
- #ifdef CONFIG_BCM5222_DUAL_PHY
- /* Mask the already found phy, try next one */
- if (au_macs[0]->mii && au_macs[0]->mii->mii_control_reg) {
- if (au_macs[0]->phy_addr == phy_addr)
- continue;
- }
- #endif
-
- mii_status = mdio_read(dev, phy_addr, MII_STATUS);
- if (mii_status == 0xffff || mii_status == 0x0000)
- /* the mii is not accessable, try next one */
- continue;
-
- phy_id0 = mdio_read(dev, phy_addr, MII_PHY_ID0);
- phy_id1 = mdio_read(dev, phy_addr, MII_PHY_ID1);
-
- /* search our mii table for the current mii */
- for (i = 0; mii_chip_table[i].phy_id1; i++) {
- if (phy_id0 == mii_chip_table[i].phy_id0 &&
- phy_id1 == mii_chip_table[i].phy_id1) {
- struct mii_phy * mii_phy = aup->mii;
-
- printk(KERN_INFO "%s: %s at phy address %d\n",
- dev->name, mii_chip_table[i].name,
- phy_addr);
-#ifdef CONFIG_MIPS_BOSPORUS
- phy_found = 1;
-#endif
- mii_phy->chip_info = mii_chip_table+i;
- aup->phy_addr = phy_addr;
- aup->want_autoneg = 1;
- aup->phy_ops = mii_chip_table[i].phy_ops;
- aup->phy_ops->phy_init(dev,phy_addr);
-
- // Check for dual-phy and then store required
- // values and set indicators. We need to do
- // this now since mdio_{read,write} need the
- // control and data register addresses.
- #ifdef CONFIG_BCM5222_DUAL_PHY
- if ( mii_chip_table[i].dual_phy) {
-
- /* assume both phys are controlled
- * through MAC0. Board specific? */
-
- /* sanity check */
- if (!au_macs[0] || !au_macs[0]->mii)
- return -1;
- aup->mii->mii_control_reg = (u32 *)
- &au_macs[0]->mac->mii_control;
- aup->mii->mii_data_reg = (u32 *)
- &au_macs[0]->mac->mii_data;
- }
- #endif
- goto found;
- }
+static int mii_probe (struct net_device *dev)
+{
+ struct au1000_private *const aup = (struct au1000_private *) dev->priv;
+ struct phy_device *phydev = NULL;
+
+#if defined(AU1XXX_PHY_STATIC_CONFIG)
+ BUG_ON(aup->mac_id < 0 || aup->mac_id > 1);
+
+ if(aup->mac_id == 0) { /* get PHY0 */
+# if defined(AU1XXX_PHY0_ADDR)
+ phydev = au_macs[AU1XXX_PHY0_BUSID]->mii_bus.phy_map[AU1XXX_PHY0_ADDR];
+# else
+ printk (KERN_INFO DRV_NAME ":%s: using PHY-less setup\n",
+ dev->name);
+ return 0;
+# endif /* defined(AU1XXX_PHY0_ADDR) */
+ } else if (aup->mac_id == 1) { /* get PHY1 */
+# if defined(AU1XXX_PHY1_ADDR)
+ phydev = au_macs[AU1XXX_PHY1_BUSID]->mii_bus.phy_map[AU1XXX_PHY1_ADDR];
+# else
+ printk (KERN_INFO DRV_NAME ":%s: using PHY-less setup\n",
+ dev->name);
+ return 0;
+# endif /* defined(AU1XXX_PHY1_ADDR) */
+ }
+
+#else /* defined(AU1XXX_PHY_STATIC_CONFIG) */
+ int phy_addr;
+
+ /* find the first (lowest address) PHY on the current MAC's MII bus */
+ for (phy_addr = 0; phy_addr < PHY_MAX_ADDR; phy_addr++)
+ if (aup->mii_bus.phy_map[phy_addr]) {
+ phydev = aup->mii_bus.phy_map[phy_addr];
+# if !defined(AU1XXX_PHY_SEARCH_HIGHEST_ADDR)
+ break; /* break out with first one found */
+# endif
}
- }
-found:
-
-#ifdef CONFIG_MIPS_BOSPORUS
- /* This is a workaround for the Micrel/Kendin 5 port switch
- The second MAC doesn't see a PHY connected... so we need to
- trick it into thinking we have one.
-
- If this kernel is run on another Au1500 development board
- the stub will be found as well as the actual PHY. However,
- the last found PHY will be used... usually at Addr 31 (Db1500).
- */
- if ( (!phy_found) )
- {
- u16 phy_id0, phy_id1;
- int i;
- phy_id0 = 0x1234;
- phy_id1 = 0x5678;
-
- /* search our mii table for the current mii */
- for (i = 0; mii_chip_table[i].phy_id1; i++) {
- if (phy_id0 == mii_chip_table[i].phy_id0 &&
- phy_id1 == mii_chip_table[i].phy_id1) {
- struct mii_phy * mii_phy;
-
- printk(KERN_INFO "%s: %s at phy address %d\n",
- dev->name, mii_chip_table[i].name,
- phy_addr);
- mii_phy = kmalloc(sizeof(struct mii_phy),
- GFP_KERNEL);
- if (mii_phy) {
- mii_phy->chip_info = mii_chip_table+i;
- aup->phy_addr = phy_addr;
- mii_phy->next = aup->mii;
- aup->phy_ops =
- mii_chip_table[i].phy_ops;
- aup->mii = mii_phy;
- aup->phy_ops->phy_init(dev,phy_addr);
- } else {
- printk(KERN_ERR "%s: out of memory\n",
- dev->name);
- return -1;
- }
- mii_phy->chip_info = mii_chip_table+i;
- aup->phy_addr = phy_addr;
- aup->phy_ops = mii_chip_table[i].phy_ops;
- aup->phy_ops->phy_init(dev,phy_addr);
- break;
- }
+# if defined(AU1XXX_PHY1_SEARCH_ON_MAC0)
+ /* try harder to find a PHY */
+ if (!phydev && (aup->mac_id == 1)) {
+ /* no PHY found, maybe we have a dual PHY? */
+ printk (KERN_INFO DRV_NAME ": no PHY found on MAC1, "
+ "let's see if it's attached to MAC0...\n");
+
+ BUG_ON(!au_macs[0]);
+
+ /* find the first (lowest address) non-attached PHY on
+ * the MAC0 MII bus */
+ for (phy_addr = 0; phy_addr < PHY_MAX_ADDR; phy_addr++) {
+ struct phy_device *const tmp_phydev =
+ au_macs[0]->mii_bus.phy_map[phy_addr];
+
+ if (!tmp_phydev)
+ continue; /* no PHY here... */
+
+ if (tmp_phydev->attached_dev)
+ continue; /* already claimed by MAC0 */
+
+ phydev = tmp_phydev;
+ break; /* found it */
}
}
- if (aup->mac_id == 0) {
- /* the Bosporus phy responds to addresses 0-5 but
- * 5 is the correct one.
- */
- aup->phy_addr = 5;
- }
-#endif
+# endif /* defined(AU1XXX_PHY1_SEARCH_OTHER_BUS) */
- if (aup->mii->chip_info == NULL) {
- printk(KERN_ERR "%s: Au1x No known MII transceivers found!\n",
- dev->name);
+#endif /* defined(AU1XXX_PHY_STATIC_CONFIG) */
+ if (!phydev) {
+ printk (KERN_ERR DRV_NAME ":%s: no PHY found\n", dev->name);
return -1;
}
- printk(KERN_INFO "%s: Using %s as default\n",
- dev->name, aup->mii->chip_info->name);
+ /* now we are supposed to have a proper phydev, to attach to... */
+ BUG_ON(!phydev);
+ BUG_ON(phydev->attached_dev);
+
+ phydev = phy_connect(dev, phydev->dev.bus_id, &au1000_adjust_link, 0);
+
+ if (IS_ERR(phydev)) {
+ printk(KERN_ERR "%s: Could not attach to PHY\n", dev->name);
+ return PTR_ERR(phydev);
+ }
+
+ /* mask with MAC supported features */
+ phydev->supported &= (SUPPORTED_10baseT_Half
+ | SUPPORTED_10baseT_Full
+ | SUPPORTED_100baseT_Half
+ | SUPPORTED_100baseT_Full
+ | SUPPORTED_Autoneg
+ /* | SUPPORTED_Pause | SUPPORTED_Asym_Pause */
+ | SUPPORTED_MII
+ | SUPPORTED_TP);
+
+ phydev->advertising = phydev->supported;
+
+ aup->old_link = 0;
+ aup->old_speed = 0;
+ aup->old_duplex = -1;
+ aup->phy_dev = phydev;
+
+ printk(KERN_INFO "%s: attached PHY driver [%s] "
+ "(mii_bus:phy_addr=%s, irq=%d)\n",
+ dev->name, phydev->drv->name, phydev->dev.bus_id, phydev->irq);
return 0;
}
@@ -1097,35 +439,38 @@ static void hard_stop(struct net_device
au_sync_delay(10);
}
-
-static void reset_mac(struct net_device *dev)
+static void enable_mac(struct net_device *dev, int force_reset)
{
- int i;
- u32 flags;
+ unsigned long flags;
struct au1000_private *aup = (struct au1000_private *) dev->priv;
- if (au1000_debug > 4)
- printk(KERN_INFO "%s: reset mac, aup %x\n",
- dev->name, (unsigned)aup);
-
spin_lock_irqsave(&aup->lock, flags);
- if (aup->timer.function == &au1000_timer) {/* check if timer initted */
- del_timer(&aup->timer);
- }
- hard_stop(dev);
- #ifdef CONFIG_BCM5222_DUAL_PHY
- if (aup->mac_id != 0) {
- #endif
- /* If BCM5222, we can't leave MAC0 in reset because then
- * we can't access the dual phy for ETH1 */
+ if(force_reset || (!aup->mac_enabled)) {
*aup->enable = MAC_EN_CLOCK_ENABLE;
au_sync_delay(2);
- *aup->enable = 0;
+ *aup->enable = (MAC_EN_RESET0 | MAC_EN_RESET1 | MAC_EN_RESET2
+ | MAC_EN_CLOCK_ENABLE);
au_sync_delay(2);
- #ifdef CONFIG_BCM5222_DUAL_PHY
+
+ aup->mac_enabled = 1;
}
- #endif
+
+ spin_unlock_irqrestore(&aup->lock, flags);
+}
+
+static void reset_mac_unlocked(struct net_device *dev)
+{
+ struct au1000_private *const aup = (struct au1000_private *) dev->priv;
+ int i;
+
+ hard_stop(dev);
+
+ *aup->enable = MAC_EN_CLOCK_ENABLE;
+ au_sync_delay(2);
+ *aup->enable = 0;
+ au_sync_delay(2);
+
aup->tx_full = 0;
for (i = 0; i < NUM_RX_DMA; i++) {
/* reset control bits */
@@ -1135,9 +480,26 @@ static void reset_mac(struct net_device
/* reset control bits */
aup->tx_dma_ring[i]->buff_stat &= ~0xf;
}
- spin_unlock_irqrestore(&aup->lock, flags);
+
+ aup->mac_enabled = 0;
+
}
+static void reset_mac(struct net_device *dev)
+{
+ struct au1000_private *const aup = (struct au1000_private *) dev->priv;
+ unsigned long flags;
+
+ if (au1000_debug > 4)
+ printk(KERN_INFO "%s: reset mac, aup %x\n",
+ dev->name, (unsigned)aup);
+
+ spin_lock_irqsave(&aup->lock, flags);
+
+ reset_mac_unlocked (dev);
+
+ spin_unlock_irqrestore(&aup->lock, flags);
+}
/*
* Setup the receive and transmit "rings". These pointers are the addresses
@@ -1208,178 +570,31 @@ static int __init au1000_init_module(voi
return 0;
}
-static int au1000_setup_aneg(struct net_device *dev, u32 advertise)
-{
- struct au1000_private *aup = (struct au1000_private *)dev->priv;
- u16 ctl, adv;
-
- /* Setup standard advertise */
- adv = mdio_read(dev, aup->phy_addr, MII_ADVERTISE);
- adv &= ~(ADVERTISE_ALL | ADVERTISE_100BASE4);
- if (advertise & ADVERTISED_10baseT_Half)
- adv |= ADVERTISE_10HALF;
- if (advertise & ADVERTISED_10baseT_Full)
- adv |= ADVERTISE_10FULL;
- if (advertise & ADVERTISED_100baseT_Half)
- adv |= ADVERTISE_100HALF;
- if (advertise & ADVERTISED_100baseT_Full)
- adv |= ADVERTISE_100FULL;
- mdio_write(dev, aup->phy_addr, MII_ADVERTISE, adv);
-
- /* Start/Restart aneg */
- ctl = mdio_read(dev, aup->phy_addr, MII_BMCR);
- ctl |= (BMCR_ANENABLE | BMCR_ANRESTART);
- mdio_write(dev, aup->phy_addr, MII_BMCR, ctl);
-
- return 0;
-}
+/*
+ * ethtool operations
+ */
-static int au1000_setup_forced(struct net_device *dev, int speed, int fd)
+static int au1000_get_settings(struct net_device *dev, struct ethtool_cmd *cmd)
{
struct au1000_private *aup = (struct au1000_private *)dev->priv;
- u16 ctl;
-
- ctl = mdio_read(dev, aup->phy_addr, MII_BMCR);
- ctl &= ~(BMCR_FULLDPLX | BMCR_SPEED100 | BMCR_ANENABLE);
-
- /* First reset the PHY */
- mdio_write(dev, aup->phy_addr, MII_BMCR, ctl | BMCR_RESET);
-
- /* Select speed & duplex */
- switch (speed) {
- case SPEED_10:
- break;
- case SPEED_100:
- ctl |= BMCR_SPEED100;
- break;
- case SPEED_1000:
- default:
- return -EINVAL;
- }
- if (fd == DUPLEX_FULL)
- ctl |= BMCR_FULLDPLX;
- mdio_write(dev, aup->phy_addr, MII_BMCR, ctl);
-
- return 0;
-}
-
-static void
-au1000_start_link(struct net_device *dev, struct ethtool_cmd *cmd)
-{
- struct au1000_private *aup = (struct au1000_private *)dev->priv;
- u32 advertise;
- int autoneg;
- int forced_speed;
- int forced_duplex;
-
- /* Default advertise */
- advertise = GENMII_DEFAULT_ADVERTISE;
- autoneg = aup->want_autoneg;
- forced_speed = SPEED_100;
- forced_duplex = DUPLEX_FULL;
-
- /* Setup link parameters */
- if (cmd) {
- if (cmd->autoneg == AUTONEG_ENABLE) {
- advertise = cmd->advertising;
- autoneg = 1;
- } else {
- autoneg = 0;
-
- forced_speed = cmd->speed;
- forced_duplex = cmd->duplex;
- }
- }
+ if (aup->phy_dev)
+ return phy_ethtool_gset(aup->phy_dev, cmd);
- /* Configure PHY & start aneg */
- aup->want_autoneg = autoneg;
- if (autoneg)
- au1000_setup_aneg(dev, advertise);
- else
- au1000_setup_forced(dev, forced_speed, forced_duplex);
- mod_timer(&aup->timer, jiffies + HZ);
+ return -EINVAL;
}
-static int au1000_get_settings(struct net_device *dev, struct ethtool_cmd *cmd)
+static int au1000_set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
{
struct au1000_private *aup = (struct au1000_private *)dev->priv;
- u16 link, speed;
-
- cmd->supported = GENMII_DEFAULT_FEATURES;
- cmd->advertising = GENMII_DEFAULT_ADVERTISE;
- cmd->port = PORT_MII;
- cmd->transceiver = XCVR_EXTERNAL;
- cmd->phy_address = aup->phy_addr;
- spin_lock_irq(&aup->lock);
- cmd->autoneg = aup->want_autoneg;
- aup->phy_ops->phy_status(dev, aup->phy_addr, &link, &speed);
- if ((speed == IF_PORT_100BASETX) || (speed == IF_PORT_100BASEFX))
- cmd->speed = SPEED_100;
- else if (speed == IF_PORT_10BASET)
- cmd->speed = SPEED_10;
- if (link && (dev->if_port == IF_PORT_100BASEFX))
- cmd->duplex = DUPLEX_FULL;
- else
- cmd->duplex = DUPLEX_HALF;
- spin_unlock_irq(&aup->lock);
- return 0;
-}
-static int au1000_set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
-{
- struct au1000_private *aup = (struct au1000_private *)dev->priv;
- unsigned long features = GENMII_DEFAULT_FEATURES;
-
- if (!capable(CAP_NET_ADMIN))
- return -EPERM;
-
- if (cmd->autoneg != AUTONEG_ENABLE && cmd->autoneg != AUTONEG_DISABLE)
- return -EINVAL;
- if (cmd->autoneg == AUTONEG_ENABLE && cmd->advertising == 0)
- return -EINVAL;
- if (cmd->duplex != DUPLEX_HALF && cmd->duplex != DUPLEX_FULL)
- return -EINVAL;
- if (cmd->autoneg == AUTONEG_DISABLE)
- switch (cmd->speed) {
- case SPEED_10:
- if (cmd->duplex == DUPLEX_HALF &&
- (features & SUPPORTED_10baseT_Half) == 0)
- return -EINVAL;
- if (cmd->duplex == DUPLEX_FULL &&
- (features & SUPPORTED_10baseT_Full) == 0)
- return -EINVAL;
- break;
- case SPEED_100:
- if (cmd->duplex == DUPLEX_HALF &&
- (features & SUPPORTED_100baseT_Half) == 0)
- return -EINVAL;
- if (cmd->duplex == DUPLEX_FULL &&
- (features & SUPPORTED_100baseT_Full) == 0)
- return -EINVAL;
- break;
- default:
- return -EINVAL;
- }
- else if ((features & SUPPORTED_Autoneg) == 0)
- return -EINVAL;
-
- spin_lock_irq(&aup->lock);
- au1000_start_link(dev, cmd);
- spin_unlock_irq(&aup->lock);
- return 0;
-}
+ if (!capable(CAP_NET_ADMIN))
+ return -EPERM;
-static int au1000_nway_reset(struct net_device *dev)
-{
- struct au1000_private *aup = (struct au1000_private *)dev->priv;
+ if (aup->phy_dev)
+ return phy_ethtool_sset(aup->phy_dev, cmd);
- if (!aup->want_autoneg)
- return -EINVAL;
- spin_lock_irq(&aup->lock);
- au1000_start_link(dev, NULL);
- spin_unlock_irq(&aup->lock);
- return 0;
+ return -EINVAL;
}
static void
@@ -1394,17 +609,11 @@ au1000_get_drvinfo(struct net_device *de
info->regdump_len = 0;
}
-static u32 au1000_get_link(struct net_device *dev)
-{
- return netif_carrier_ok(dev);
-}
-
static struct ethtool_ops au1000_ethtool_ops = {
.get_settings = au1000_get_settings,
.set_settings = au1000_set_settings,
.get_drvinfo = au1000_get_drvinfo,
- .nway_reset = au1000_nway_reset,
- .get_link = au1000_get_link
+ .get_link = ethtool_op_get_link,
};
static struct net_device * au1000_probe(int port_num)
@@ -1499,23 +708,31 @@ static struct net_device * au1000_probe(
memcpy(dev->dev_addr, au1000_mac_addr, sizeof(au1000_mac_addr));
dev->dev_addr[5] += port_num;
- /* Bring the device out of reset, otherwise probing the MII will hang */
- *aup->enable = MAC_EN_CLOCK_ENABLE;
- au_sync_delay(2);
- *aup->enable = MAC_EN_RESET0 | MAC_EN_RESET1 | MAC_EN_RESET2 |
- MAC_EN_CLOCK_ENABLE;
- au_sync_delay(2);
-
- aup->mii = kmalloc(sizeof(struct mii_phy), GFP_KERNEL);
- if (!aup->mii) {
- printk(KERN_ERR "%s: out of memory\n", dev->name);
- goto err_out;
- }
- aup->mii->next = NULL;
- aup->mii->chip_info = NULL;
- aup->mii->status = 0;
- aup->mii->mii_control_reg = 0;
- aup->mii->mii_data_reg = 0;
+ *aup->enable = 0;
+ aup->mac_enabled = 0;
+
+ aup->mii_bus.priv = dev;
+ aup->mii_bus.read = mdiobus_read;
+ aup->mii_bus.write = mdiobus_write;
+ aup->mii_bus.reset = mdiobus_reset;
+ aup->mii_bus.name = "au1000_eth_mii";
+ aup->mii_bus.id = aup->mac_id;
+ aup->mii_bus.irq = kmalloc(sizeof(int)*PHY_MAX_ADDR, GFP_KERNEL);
+ for(i = 0; i < PHY_MAX_ADDR; ++i)
+ aup->mii_bus.irq[i] = PHY_POLL;
+
+ /* if known, set corresponding PHY IRQs */
+#if defined(AU1XXX_PHY_STATIC_CONFIG)
+# if defined(AU1XXX_PHY0_IRQ)
+ if (AU1XXX_PHY0_BUSID == aup->mii_bus.id)
+ aup->mii_bus.irq[AU1XXX_PHY0_ADDR] = AU1XXX_PHY0_IRQ;
+# endif
+# if defined(AU1XXX_PHY1_IRQ)
+ if (AU1XXX_PHY1_BUSID == aup->mii_bus.id)
+ aup->mii_bus.irq[AU1XXX_PHY1_ADDR] = AU1XXX_PHY1_IRQ;
+# endif
+#endif
+ mdiobus_register(&aup->mii_bus);
if (mii_probe(dev) != 0) {
goto err_out;
@@ -1561,7 +778,6 @@ static struct net_device * au1000_probe(
dev->set_multicast_list = &set_rx_mode;
dev->do_ioctl = &au1000_ioctl;
SET_ETHTOOL_OPS(dev, &au1000_ethtool_ops);
- dev->set_config = &au1000_set_config;
dev->tx_timeout = au1000_tx_timeout;
dev->watchdog_timeo = ETH_TX_TIMEOUT;
@@ -1577,7 +793,7 @@ err_out:
/* here we should have a valid dev plus aup-> register addresses
* so we can reset the mac properly.*/
reset_mac(dev);
- kfree(aup->mii);
+
for (i = 0; i < NUM_RX_DMA; i++) {
if (aup->rx_db_inuse[i])
ReleaseDB(aup, aup->rx_db_inuse[i]);
@@ -1610,19 +826,14 @@ static int au1000_init(struct net_device
u32 flags;
int i;
u32 control;
- u16 link, speed;
if (au1000_debug > 4)
printk("%s: au1000_init\n", dev->name);
- spin_lock_irqsave(&aup->lock, flags);
-
/* bring the device out of reset */
- *aup->enable = MAC_EN_CLOCK_ENABLE;
- au_sync_delay(2);
- *aup->enable = MAC_EN_RESET0 | MAC_EN_RESET1 |
- MAC_EN_RESET2 | MAC_EN_CLOCK_ENABLE;
- au_sync_delay(20);
+ enable_mac(dev, 1);
+
+ spin_lock_irqsave(&aup->lock, flags);
aup->mac->control = 0;
aup->tx_head = (aup->tx_dma_ring[0]->buff_stat & 0xC) >> 2;
@@ -1638,12 +849,16 @@ static int au1000_init(struct net_device
}
au_sync();
- aup->phy_ops->phy_status(dev, aup->phy_addr, &link, &speed);
- control = MAC_DISABLE_RX_OWN | MAC_RX_ENABLE | MAC_TX_ENABLE;
+ control = MAC_RX_ENABLE | MAC_TX_ENABLE;
#ifndef CONFIG_CPU_LITTLE_ENDIAN
control |= MAC_BIG_ENDIAN;
#endif
- if (link && (dev->if_port == IF_PORT_100BASEFX)) {
+ if (aup->phy_dev) {
+ if (aup->phy_dev->link && (DUPLEX_FULL == aup->phy_dev->duplex))
+ control |= MAC_FULL_DUPLEX;
+ else
+ control |= MAC_DISABLE_RX_OWN;
+ } else { /* PHY-less op, assume full-duplex */
control |= MAC_FULL_DUPLEX;
}
@@ -1655,57 +870,84 @@ #endif
return 0;
}
-static void au1000_timer(unsigned long data)
+static void
+au1000_adjust_link(struct net_device *dev)
{
- struct net_device *dev = (struct net_device *)data;
struct au1000_private *aup = (struct au1000_private *) dev->priv;
- unsigned char if_port;
- u16 link, speed;
+ struct phy_device *phydev = aup->phy_dev;
+ unsigned long flags;
- if (!dev) {
- /* fatal error, don't restart the timer */
- printk(KERN_ERR "au1000_timer error: NULL dev\n");
- return;
- }
+ int status_change = 0;
- if_port = dev->if_port;
- if (aup->phy_ops->phy_status(dev, aup->phy_addr, &link, &speed) == 0) {
- if (link) {
- if (!netif_carrier_ok(dev)) {
- netif_carrier_on(dev);
- printk(KERN_INFO "%s: link up\n", dev->name);
- }
- }
- else {
- if (netif_carrier_ok(dev)) {
- netif_carrier_off(dev);
- dev->if_port = 0;
- printk(KERN_INFO "%s: link down\n", dev->name);
- }
+ BUG_ON(!aup->phy_dev);
+
+ spin_lock_irqsave(&aup->lock, flags);
+
+ if (phydev->link && (aup->old_speed != phydev->speed)) {
+ // speed changed
+
+ switch(phydev->speed) {
+ case SPEED_10:
+ case SPEED_100:
+ break;
+ default:
+ printk(KERN_WARNING
+ "%s: Speed (%d) is not 10/100 ???\n",
+ dev->name, phydev->speed);
+ break;
}
+
+ aup->old_speed = phydev->speed;
+
+ status_change = 1;
}
- if (link && (dev->if_port != if_port) &&
- (dev->if_port != IF_PORT_UNKNOWN)) {
+ if (phydev->link && (aup->old_duplex != phydev->duplex)) {
+ // duplex mode changed
+
+ /* switching duplex mode requires to disable rx and tx! */
hard_stop(dev);
- if (dev->if_port == IF_PORT_100BASEFX) {
- printk(KERN_INFO "%s: going to full duplex\n",
- dev->name);
- aup->mac->control |= MAC_FULL_DUPLEX;
- au_sync_delay(1);
- }
- else {
- aup->mac->control &= ~MAC_FULL_DUPLEX;
- au_sync_delay(1);
- }
+
+ if (DUPLEX_FULL == phydev->duplex)
+ aup->mac->control = ((aup->mac->control
+ | MAC_FULL_DUPLEX)
+ & ~MAC_DISABLE_RX_OWN);
+ else
+ aup->mac->control = ((aup->mac->control
+ & ~MAC_FULL_DUPLEX)
+ | MAC_DISABLE_RX_OWN);
+ au_sync_delay(1);
+
enable_rx_tx(dev);
+ aup->old_duplex = phydev->duplex;
+
+ status_change = 1;
+ }
+
+ if(phydev->link != aup->old_link) {
+ // link state changed
+
+ if (phydev->link) // link went up
+ netif_schedule(dev);
+ else { // link went down
+ aup->old_speed = 0;
+ aup->old_duplex = -1;
+ }
+
+ aup->old_link = phydev->link;
+ status_change = 1;
}
- aup->timer.expires = RUN_AT((1*HZ));
- aup->timer.data = (unsigned long)dev;
- aup->timer.function = &au1000_timer; /* timer handler */
- add_timer(&aup->timer);
+ spin_unlock_irqrestore(&aup->lock, flags);
+ if (status_change) {
+ if (phydev->link)
+ printk(KERN_INFO "%s: link up (%d/%s)\n",
+ dev->name, phydev->speed,
+ DUPLEX_FULL == phydev->duplex ? "Full" : "Half");
+ else
+ printk(KERN_INFO "%s: link down\n", dev->name);
+ }
}
static int au1000_open(struct net_device *dev)
@@ -1716,25 +958,26 @@ static int au1000_open(struct net_device
if (au1000_debug > 4)
printk("%s: open: dev=%p\n", dev->name, dev);
+ if ((retval = request_irq(dev->irq, &au1000_interrupt, 0,
+ dev->name, dev))) {
+ printk(KERN_ERR "%s: unable to get IRQ %d\n",
+ dev->name, dev->irq);
+ return retval;
+ }
+
if ((retval = au1000_init(dev))) {
printk(KERN_ERR "%s: error in au1000_init\n", dev->name);
free_irq(dev->irq, dev);
return retval;
}
- netif_start_queue(dev);
- if ((retval = request_irq(dev->irq, &au1000_interrupt, 0,
- dev->name, dev))) {
- printk(KERN_ERR "%s: unable to get IRQ %d\n",
- dev->name, dev->irq);
- return retval;
+ if (aup->phy_dev) {
+ /* cause the PHY state machine to schedule a link state check */
+ aup->phy_dev->state = PHY_CHANGELINK;
+ phy_start(aup->phy_dev);
}
- init_timer(&aup->timer); /* used in ioctl() */
- aup->timer.expires = RUN_AT((3*HZ));
- aup->timer.data = (unsigned long)dev;
- aup->timer.function = &au1000_timer; /* timer handler */
- add_timer(&aup->timer);
+ netif_start_queue(dev);
if (au1000_debug > 4)
printk("%s: open: Initialization done.\n", dev->name);
@@ -1744,16 +987,19 @@ static int au1000_open(struct net_device
static int au1000_close(struct net_device *dev)
{
- u32 flags;
- struct au1000_private *aup = (struct au1000_private *) dev->priv;
+ unsigned long flags;
+ struct au1000_private *const aup = (struct au1000_private *) dev->priv;
if (au1000_debug > 4)
printk("%s: close: dev=%p\n", dev->name, dev);
- reset_mac(dev);
+ if (aup->phy_dev)
+ phy_stop(aup->phy_dev);
spin_lock_irqsave(&aup->lock, flags);
-
+
+ reset_mac_unlocked (dev);
+
/* stop the device */
netif_stop_queue(dev);
@@ -1775,7 +1021,6 @@ static void __exit au1000_cleanup_module
if (dev) {
aup = (struct au1000_private *) dev->priv;
unregister_netdev(dev);
- kfree(aup->mii);
for (j = 0; j < NUM_RX_DMA; j++)
if (aup->rx_db_inuse[j])
ReleaseDB(aup, aup->rx_db_inuse[j]);
@@ -1798,7 +1043,7 @@ static void update_tx_stats(struct net_d
struct net_device_stats *ps = &aup->stats;
if (status & TX_FRAME_ABORTED) {
- if (dev->if_port == IF_PORT_100BASEFX) {
+ if (!aup->phy_dev || (DUPLEX_FULL == aup->phy_dev->duplex)) {
if (status & (TX_JAB_TIMEOUT | TX_UNDERRUN)) {
/* any other tx errors are only valid
* in half duplex mode */
@@ -2072,126 +1317,15 @@ static void set_rx_mode(struct net_devic
}
}
-
static int au1000_ioctl(struct net_device *dev, struct ifreq *rq, int cmd)
{
struct au1000_private *aup = (struct au1000_private *)dev->priv;
- u16 *data = (u16 *)&rq->ifr_ifru;
-
- switch(cmd) {
- case SIOCDEVPRIVATE: /* Get the address of the PHY in use. */
- case SIOCGMIIPHY:
- if (!netif_running(dev)) return -EINVAL;
- data[0] = aup->phy_addr;
- case SIOCDEVPRIVATE+1: /* Read the specified MII register. */
- case SIOCGMIIREG:
- data[3] = mdio_read(dev, data[0], data[1]);
- return 0;
- case SIOCDEVPRIVATE+2: /* Write the specified MII register */
- case SIOCSMIIREG:
- if (!capable(CAP_NET_ADMIN))
- return -EPERM;
- mdio_write(dev, data[0], data[1],data[2]);
- return 0;
- default:
- return -EOPNOTSUPP;
- }
-
-}
-
-
-static int au1000_set_config(struct net_device *dev, struct ifmap *map)
-{
- struct au1000_private *aup = (struct au1000_private *) dev->priv;
- u16 control;
- if (au1000_debug > 4) {
- printk("%s: set_config called: dev->if_port %d map->port %x\n",
- dev->name, dev->if_port, map->port);
- }
+ if (!netif_running(dev)) return -EINVAL;
- switch(map->port){
- case IF_PORT_UNKNOWN: /* use auto here */
- printk(KERN_INFO "%s: config phy for aneg\n",
- dev->name);
- dev->if_port = map->port;
- /* Link Down: the timer will bring it up */
- netif_carrier_off(dev);
-
- /* read current control */
- control = mdio_read(dev, aup->phy_addr, MII_CONTROL);
- control &= ~(MII_CNTL_FDX | MII_CNTL_F100);
-
- /* enable auto negotiation and reset the negotiation */
- mdio_write(dev, aup->phy_addr, MII_CONTROL,
- control | MII_CNTL_AUTO |
- MII_CNTL_RST_AUTO);
+ if (!aup->phy_dev) return -EINVAL; // PHY not controllable
- break;
-
- case IF_PORT_10BASET: /* 10BaseT */
- printk(KERN_INFO "%s: config phy for 10BaseT\n",
- dev->name);
- dev->if_port = map->port;
-
- /* Link Down: the timer will bring it up */
- netif_carrier_off(dev);
-
- /* set Speed to 10Mbps, Half Duplex */
- control = mdio_read(dev, aup->phy_addr, MII_CONTROL);
- control &= ~(MII_CNTL_F100 | MII_CNTL_AUTO |
- MII_CNTL_FDX);
-
- /* disable auto negotiation and force 10M/HD mode*/
- mdio_write(dev, aup->phy_addr, MII_CONTROL, control);
- break;
-
- case IF_PORT_100BASET: /* 100BaseT */
- case IF_PORT_100BASETX: /* 100BaseTx */
- printk(KERN_INFO "%s: config phy for 100BaseTX\n",
- dev->name);
- dev->if_port = map->port;
-
- /* Link Down: the timer will bring it up */
- netif_carrier_off(dev);
-
- /* set Speed to 100Mbps, Half Duplex */
- /* disable auto negotiation and enable 100MBit Mode */
- control = mdio_read(dev, aup->phy_addr, MII_CONTROL);
- control &= ~(MII_CNTL_AUTO | MII_CNTL_FDX);
- control |= MII_CNTL_F100;
- mdio_write(dev, aup->phy_addr, MII_CONTROL, control);
- break;
-
- case IF_PORT_100BASEFX: /* 100BaseFx */
- printk(KERN_INFO "%s: config phy for 100BaseFX\n",
- dev->name);
- dev->if_port = map->port;
-
- /* Link Down: the timer will bring it up */
- netif_carrier_off(dev);
-
- /* set Speed to 100Mbps, Full Duplex */
- /* disable auto negotiation and enable 100MBit Mode */
- control = mdio_read(dev, aup->phy_addr, MII_CONTROL);
- control &= ~MII_CNTL_AUTO;
- control |= MII_CNTL_F100 | MII_CNTL_FDX;
- mdio_write(dev, aup->phy_addr, MII_CONTROL, control);
- break;
- case IF_PORT_10BASE2: /* 10Base2 */
- case IF_PORT_AUI: /* AUI */
- /* These Modes are not supported (are they?)*/
- printk(KERN_ERR "%s: 10Base2/AUI not supported",
- dev->name);
- return -EOPNOTSUPP;
- break;
-
- default:
- printk(KERN_ERR "%s: Invalid media selected",
- dev->name);
- return -EINVAL;
- }
- return 0;
+ return phy_mii_ioctl(aup->phy_dev, if_mii(rq), cmd);
}
static struct net_device_stats *au1000_get_stats(struct net_device *dev)
diff --git a/drivers/net/au1000_eth.h b/drivers/net/au1000_eth.h
index 7f9326e..41c2f84 100644
--- a/drivers/net/au1000_eth.h
+++ b/drivers/net/au1000_eth.h
@@ -40,120 +40,6 @@ #define MAC_MIN_PKT_SIZE 64
#define MULTICAST_FILTER_LIMIT 64
-/* FIXME
- * The PHY defines should be in a separate file.
- */
-
-/* MII register offsets */
-#define MII_CONTROL 0x0000
-#define MII_STATUS 0x0001
-#define MII_PHY_ID0 0x0002
-#define MII_PHY_ID1 0x0003
-#define MII_ANADV 0x0004
-#define MII_ANLPAR 0x0005
-#define MII_AEXP 0x0006
-#define MII_ANEXT 0x0007
-#define MII_LSI_PHY_CONFIG 0x0011
-/* Status register */
-#define MII_LSI_PHY_STAT 0x0012
-#define MII_AMD_PHY_STAT MII_LSI_PHY_STAT
-#define MII_INTEL_PHY_STAT 0x0011
-
-#define MII_AUX_CNTRL 0x0018
-/* mii registers specific to AMD 79C901 */
-#define MII_STATUS_SUMMARY = 0x0018
-
-/* MII Control register bit definitions. */
-#define MII_CNTL_FDX 0x0100
-#define MII_CNTL_RST_AUTO 0x0200
-#define MII_CNTL_ISOLATE 0x0400
-#define MII_CNTL_PWRDWN 0x0800
-#define MII_CNTL_AUTO 0x1000
-#define MII_CNTL_F100 0x2000
-#define MII_CNTL_LPBK 0x4000
-#define MII_CNTL_RESET 0x8000
-
-/* MII Status register bit */
-#define MII_STAT_EXT 0x0001
-#define MII_STAT_JAB 0x0002
-#define MII_STAT_LINK 0x0004
-#define MII_STAT_CAN_AUTO 0x0008
-#define MII_STAT_FAULT 0x0010
-#define MII_STAT_AUTO_DONE 0x0020
-#define MII_STAT_CAN_T 0x0800
-#define MII_STAT_CAN_T_FDX 0x1000
-#define MII_STAT_CAN_TX 0x2000
-#define MII_STAT_CAN_TX_FDX 0x4000
-#define MII_STAT_CAN_T4 0x8000
-
-
-#define MII_ID1_OUI_LO 0xFC00 /* low bits of OUI mask */
-#define MII_ID1_MODEL 0x03F0 /* model number */
-#define MII_ID1_REV 0x000F /* model number */
-
-/* MII NWAY Register Bits ...
- valid for the ANAR (Auto-Negotiation Advertisement) and
- ANLPAR (Auto-Negotiation Link Partner) registers */
-#define MII_NWAY_NODE_SEL 0x001f
-#define MII_NWAY_CSMA_CD 0x0001
-#define MII_NWAY_T 0x0020
-#define MII_NWAY_T_FDX 0x0040
-#define MII_NWAY_TX 0x0080
-#define MII_NWAY_TX_FDX 0x0100
-#define MII_NWAY_T4 0x0200
-#define MII_NWAY_PAUSE 0x0400
-#define MII_NWAY_RF 0x2000 /* Remote Fault */
-#define MII_NWAY_ACK 0x4000 /* Remote Acknowledge */
-#define MII_NWAY_NP 0x8000 /* Next Page (Enable) */
-
-/* mii stsout register bits */
-#define MII_STSOUT_LINK_FAIL 0x4000
-#define MII_STSOUT_SPD 0x0080
-#define MII_STSOUT_DPLX 0x0040
-
-/* mii stsics register bits */
-#define MII_STSICS_SPD 0x8000
-#define MII_STSICS_DPLX 0x4000
-#define MII_STSICS_LINKSTS 0x0001
-
-/* mii stssum register bits */
-#define MII_STSSUM_LINK 0x0008
-#define MII_STSSUM_DPLX 0x0004
-#define MII_STSSUM_AUTO 0x0002
-#define MII_STSSUM_SPD 0x0001
-
-/* lsi phy status register */
-#define MII_LSI_PHY_STAT_FDX 0x0040
-#define MII_LSI_PHY_STAT_SPD 0x0080
-
-/* amd phy status register */
-#define MII_AMD_PHY_STAT_FDX 0x0800
-#define MII_AMD_PHY_STAT_SPD 0x0400
-
-/* intel phy status register */
-#define MII_INTEL_PHY_STAT_FDX 0x0200
-#define MII_INTEL_PHY_STAT_SPD 0x4000
-
-/* Auxilliary Control/Status Register */
-#define MII_AUX_FDX 0x0001
-#define MII_AUX_100 0x0002
-#define MII_AUX_F100 0x0004
-#define MII_AUX_ANEG 0x0008
-
-typedef struct mii_phy {
- struct mii_phy * next;
- struct mii_chip_info * chip_info;
- u16 status;
- u32 *mii_control_reg;
- u32 *mii_data_reg;
-} mii_phy_t;
-
-struct phy_ops {
- int (*phy_init) (struct net_device *, int);
- int (*phy_reset) (struct net_device *, int);
- int (*phy_status) (struct net_device *, int, u16 *, u16 *);
-};
-
/*
* Data Buffer Descriptor. Data buffers must be aligned on 32 byte
* boundary for both, receive and transmit.
@@ -200,7 +86,6 @@ typedef struct mac_reg {
struct au1000_private {
-
db_dest_t *pDBfree;
db_dest_t db[NUM_RX_BUFFS+NUM_TX_BUFFS];
volatile rx_dma_t *rx_dma_ring[NUM_RX_DMA];
@@ -213,8 +98,15 @@ struct au1000_private {
u32 tx_full;
int mac_id;
- mii_phy_t *mii;
- struct phy_ops *phy_ops;
+
+ int mac_enabled; /* whether MAC is currently enabled and running (req. for mdio) */
+
+ int old_link; /* used by au1000_adjust_link */
+ int old_speed;
+ int old_duplex;
+
+ struct phy_device *phy_dev;
+ struct mii_bus mii_bus;
/* These variables are just for quick access to certain regs addresses. */
volatile mac_reg_t *mac; /* mac registers */
@@ -223,14 +115,6 @@ struct au1000_private {
u32 vaddr; /* virtual address of rx/tx buffers */
dma_addr_t dma_addr; /* dma address of rx/tx buffers */
- u8 *hash_table;
- u32 hash_mode;
- u32 intr_work_done; /* number of Rx and Tx pkts processed in the isr */
- int phy_addr; /* phy address */
- u32 options; /* User-settable misc. driver options. */
- u32 drv_flags;
- int want_autoneg;
struct net_device_stats stats;
- struct timer_list timer;
spinlock_t lock; /* Serialise access to device */
};
--
1.3.2
^ permalink raw reply related
* Re: Question about tcp hash function tcp_hashfn()
From: Evgeniy Polyakov @ 2006-06-01 8:38 UTC (permalink / raw)
To: David Miller, draghuram, linux-kernel, netdev
In-Reply-To: <20060601011125.C22283@openss7.org>
On Thu, Jun 01, 2006 at 01:11:25AM -0600, Brian F. G. Bidulock (bidulock@openss7.org) wrote:
> Evgeniy,
>
> On Thu, 01 Jun 2006, Evgeniy Polyakov wrote:
>
> > On Thu, Jun 01, 2006 at 12:46:08AM -0600, Brian F. G. Bidulock (bidulock@openss7.org) wrote:
> > > > Since pseudo-randomness affects both folded and not folded hash
> > > > distribution, it can not end up in different results.
> > >
> > > Yes it would, so to rule out pseudo-random effects the pseudo-
> > > random number generator must be removed.
> > >
> > > >
> > > > You are right that having test with 2^48 values is really interesting,
> > > > but it will take ages on my test machine :)
> > >
> > > Try a usable subset; no pseudo-random number generator.
> >
> > I've run it for 2^30 - the same result: folded and not folded Jenkins
> > hash behave the same and still both results produce exactly the same
> > artifacts compared to XOR hash.
>
> But not without the pseudo-random number generation... ?
How can I obtain (2^30)*6 bytes of truly random bytes?
> > Btw, XOR hash, as completely stateless, can be used to show how
> > Linux pseudo-random generator works for given subset - it's average of
> > distribution is very good.
>
> But its distribution might auto-correlate with the Jenkins function.
> The only way to be sure is to remove the pseudo-random number generator.
>
> Just try incrementing from, say, 10.0.0.0:10000 up, resetting port number
> to 10000 at 16000, and just incrementing the IP address when the port
> number wraps, instead of pseudo-random, through 2^30 loops for both.
> If the same artifacts emerge, I give in.
I've run it with following source ip/port selection algo:
if (++sport == 0) {
saddr++;
sport++;
}
Starting IP was 1.1.1.1 and sport was 1.
Destination IP and port are the same 192.168.0.1:80
Jenkins hash started to show different behaviour:
it does not have previous artefacts, but instead it's dispersion is
_much_ wider than in XOR case.
With following ip/port selection algo:
if (++sport == 0) {
//saddr++;
sport += 123;
}
I see yet another jenkins artefacts, but again different from previous
two.
But each time both folded and not folded hashes behave exactly the same.
> Can you show the same artifacts for jenkins_3word?
What should be used as starting point there?
If I use 0 it is the same as jhash_2words().
If I use 123123 - artefacts are the same, just slighly shifted (I tested
only the latest test above though).
Looking into the code we can see that jhash_2words() is jhash_3words()
with zero "C" value, so it will show the same nature.
--
Evgeniy Polyakov
^ permalink raw reply
* Re: 2.6.17-rc4: netfilter LOG messages truncated via NETCONSOLE
From: Frank van Maarseveen @ 2006-06-01 9:11 UTC (permalink / raw)
To: Patrick McHardy; +Cc: linux-kernel, Kernel Netdev Mailing List
In-Reply-To: <447DD66C.30605@trash.net>
On Wed, May 31, 2006 at 07:46:20PM +0200, Patrick McHardy wrote:
[...]
> > ip -s link doesn't show any
> > dropped packets so far with any patch and I don't use traffic control
> > that I'm aware of. But I'm not sure what to make of "tc" output, maybe
> > because CONFIG_SHAPER is not set:
> >
> > # tc -s -d qdisc show
> > RTNETLINK answers: Invalid argument
> > Dump terminated
>
> Thats because you're missing CONFIG_NET_SCHED. Please enable it and
> try the tc command again, without it we can't see whether the qdisc
> (which is present even without CONFIG_NET_SCHED) just dropped the
> packets.
ok, now "tc -s -d qdisc show" says (after noticing missing netconsole
packets):
qdisc pfifo_fast 0: dev eth0 bands 3 priomap 1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
Sent 155031 bytes 2067 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
--
Frank
^ permalink raw reply
* Re: Question about tcp hash function tcp_hashfn()
From: Brian F. G. Bidulock @ 2006-06-01 10:24 UTC (permalink / raw)
To: Evgeniy Polyakov; +Cc: David Miller, draghuram, linux-kernel, netdev
In-Reply-To: <20060601083805.GB754@2ka.mipt.ru>
Evgeniy,
On Thu, 01 Jun 2006, Evgeniy Polyakov wrote:
For purely random numbers you could amplify thermal noise off an
open transitor junction (the audiofile's white noise generator)
and feed it into an analog to digital converter.
>
> I've run it with following source ip/port selection algo:
> if (++sport == 0) {
> saddr++;
> sport++;
> }
>
> Starting IP was 1.1.1.1 and sport was 1.
> Destination IP and port are the same 192.168.0.1:80
>
> Jenkins hash started to show different behaviour:
> it does not have previous artefacts, but instead it's dispersion is
> _much_ wider than in XOR case.
Aha! But perhaps this is too easy a data set. HTTP clients typically
dynamically allocate port numbers within a range and source address
are typically not less than a certain value. That is why I suggested
something like:
sport = 10000;
saddr = 0x0a000000; /* 10.0.0.0 */
...
if (++sport == 16000) {
sport = 10000;
saddr++;
}
If this shows artifacts worse than XOR then more realistic gaps in the
input values will cause artifacts.
>
> With following ip/port selection algo:
> if (++sport == 0) {
> //saddr++;
> sport += 123;
> }
>
> I see yet another jenkins artefacts, but again different from previous
> two.
Adding primes. Again, the arithmetic series of primes might auto-correlate
with the Jenkins function. Or it plain might not like gaps.
>
> But each time both folded and not folded hashes behave exactly the same.
>
> > Can you show the same artifacts for jenkins_3word?
>
> What should be used as starting point there?
> If I use 0 it is the same as jhash_2words().
> If I use 123123 - artefacts are the same, just slighly shifted (I tested
> only the latest test above though).
>
> Looking into the code we can see that jhash_2words() is jhash_3words()
> with zero "C" value, so it will show the same nature.
Skip that then.
^ permalink raw reply
* Re: Refactor Netlink connector?
From: Thomas Graf @ 2006-06-01 10:45 UTC (permalink / raw)
To: James Morris
Cc: jamal, Evgeniy Polyakov, netdev, David S. Miller, Stephen Smalley
In-Reply-To: <Pine.LNX.4.64.0605311141320.4964@d.namei>
* James Morris <jmorris@namei.org> 2006-05-31 11:42
> On Wed, 31 May 2006, jamal wrote:
>
> > To also answer your other email:
> > Look at security/selinux/nlmsgtab.c for example for NETLINK_ROUTE
> > and compare with NETLINK_GENERIC to see the hole. I was suggesting if
> > we started by just adding checks for NETLINK_GENERIC first in those
> > tables (currently lacking), that would be a good start.
>
> They're currently mediated as a generic netlink socket type at a higher
> level: they're not unmediated, just not fine grained enough to know what
> kind of message is being sent.
It shouldn't be hard to split what is implemented in nlmsg_route_perms[]
for NETLINK_ROUTE into the definitions of the generic netlink
operations, could look like this:
struct genl_ops some_op = {
[...]
.perm = NETLINK_GENERIC_SOCKET__NLMSG_READ,
};
int genl_lookup_perm(u16 nlmsg_type, u8 cmd)
{
struct genl_family *family;
struct genl_ops *ops;
family = genl_family_find_byid(nlmsg_type);
ops = genl_get_cmd(cmd, family);
return ops->perm;
}
int genl_peek_cmd(struct nlmsghdr *nlh)
{
struct genlmsghdr *hdr = nlmsg_data(nlh);
if (nlh->nlmsglen < nlmsg_msg_sizeo(GENL_HDRLEN))
return -EINVAL;
return hdr->cmd;
}
selinux_lookup() must take struct nlmsghdr as an additional argument
case SECCLASS_NETLINK_GENERIC_SOCKET:
cmd = genl_peek_cmd(nlh)
*perm = genl_lookup_perm(nlmsg_type, cmd);
break;
^ permalink raw reply
* Re: Question about tcp hash function tcp_hashfn()
From: Evgeniy Polyakov @ 2006-06-01 11:06 UTC (permalink / raw)
To: David Miller, draghuram, linux-kernel, netdev; +Cc: Brian F. G. Bidulock
In-Reply-To: <20060601042457.B25584@openss7.org>
On Thu, Jun 01, 2006 at 04:24:57AM -0600, Brian F. G. Bidulock (bidulock@openss7.org) wrote:
> For purely random numbers you could amplify thermal noise off an
> open transitor junction (the audiofile's white noise generator)
> and feed it into an analog to digital converter.
It is also possible to look into window and count how many times sun
hides and shines... It is quite cloudy in Moscow though.
As Acrypto asynchronous crypto layer author I can use different hardware
crypto accelerators, but it is a topic for another discussion.
> >
> > I've run it with following source ip/port selection algo:
> > if (++sport == 0) {
> > saddr++;
> > sport++;
> > }
> >
> > Starting IP was 1.1.1.1 and sport was 1.
> > Destination IP and port are the same 192.168.0.1:80
> >
> > Jenkins hash started to show different behaviour:
> > it does not have previous artefacts, but instead it's dispersion is
> > _much_ wider than in XOR case.
>
> Aha! But perhaps this is too easy a data set. HTTP clients typically
> dynamically allocate port numbers within a range and source address
> are typically not less than a certain value. That is why I suggested
> something like:
>
> sport = 10000;
> saddr = 0x0a000000; /* 10.0.0.0 */
>
> ...
>
> if (++sport == 16000) {
> sport = 10000;
> saddr++;
> }
>
> If this shows artifacts worse than XOR then more realistic gaps in the
> input values will cause artifacts.
Specially for you :)
It does not have artifacts, but it's dispersion is wider than XOR one.
_Much_ wider, which tends to creation of some specially crafted source
distribution which ends up in totally broken fairness.
As usual folded and not folded versions behave exactly the same.
> > With following ip/port selection algo:
> > if (++sport == 0) {
> > //saddr++;
> > sport += 123;
> > }
> >
> > I see yet another jenkins artefacts, but again different from previous
> > two.
>
> Adding primes. Again, the arithmetic series of primes might auto-correlate
> with the Jenkins function. Or it plain might not like gaps.
>
I want to confirm three things and one state:
1. Jenkins hash has some unacceptible artefacts in some source
address/port distributions, no matter if it has some law embedded or it
is (pseudo)-random set.
If there are bugs, bugs exist.
2. If it does not have artifacts it has unacceptible dispersion.
3. It is 3 times slower than XOR one (28 seconds for XOR for 2^29
iterations vs. 101 seconds jhash nonfolded and 109 jhash folded on my AMD64
3500+ 2.2 Ghz desktop).
4. I believe it can be tuned or has some gaps inside refactoring logic,
which can be fixed, but as is it can not be used for fair hash creation.
--
Evgeniy Polyakov
^ permalink raw reply
* [NET]: Add netif_tx_lock
From: Herbert Xu @ 2006-06-01 11:15 UTC (permalink / raw)
To: David Miller; +Cc: mchan, jgarzik, netdev
In-Reply-To: <20060601002525.GA19220@gondor.apana.org.au>
[-- Attachment #1: Type: text/plain, Size: 2180 bytes --]
On Thu, Jun 01, 2006 at 10:25:25AM +1000, herbert wrote:
>
> > I think this netpoll wrinkle means we also have to make
> > sure to set the xmit_lock_owner across the board.
>
> You're right. In fact this can deadlock today for those drivers that
> already make use of xmit_lock without setting the owner. So I suppose
> something like net_xmit_lock to obtain xmit_lock is called for.
OK, here is a patch which does this.
[NET]: Add netif_tx_lock
Various drivers use xmit_lock internally to synchronise with their
transmission routines. They do so without setting xmit_lock_owner.
This is fine as long as netpoll is not in use.
With netpoll it is possible for deadlocks to occur if xmit_lock_owner
isn't set. This is because if a printk occurs while xmit_lock is held
and xmit_lock_owner is not set can cause netpoll to attempt to take
xmit_lock recursively.
While it is possible to resolve this by getting netpoll to use trylock,
it is suboptimal because netpoll's sole objective is to maximise the
chance of getting the printk out on the wire. So delaying or dropping
the message is to be avoided as much as possible.
So the only alternative is to always set xmit_lock_owner. The following
patch does this by introducing the netif_tx_lock family of functions that
take care of setting/unsetting xmit_lock_owner.
I renamed xmit_lock to _xmit_lock to indicate that it should not be used
directly. I didn't provide irq versions of the netif_tx_lock functions
since xmit_lock is meant to be a BH-disabling lock.
This is pretty much a straight text substitution except for a small bug
fix in winbond. It currently uses netif_stop_queue/spin_unlock_wait to
stop transmission. This is unsafe as an IRQ can potentially wake up the
queue. So it is safer to use netif_tx_disable.
The hamradio bits used spin_lock_irq but it is unnecessary as xmit_lock
must never be taken in an IRQ handler.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
[-- Attachment #2: netif-tx-lock.patch --]
[-- Type: text/plain, Size: 23979 bytes --]
diff --git a/Documentation/networking/netdevices.txt b/Documentation/networking/netdevices.txt
index 3c0a5ba..847cedb 100644
--- a/Documentation/networking/netdevices.txt
+++ b/Documentation/networking/netdevices.txt
@@ -42,9 +42,9 @@ dev->get_stats:
Context: nominally process, but don't sleep inside an rwlock
dev->hard_start_xmit:
- Synchronization: dev->xmit_lock spinlock.
+ Synchronization: netif_tx_lock spinlock.
When the driver sets NETIF_F_LLTX in dev->features this will be
- called without holding xmit_lock. In this case the driver
+ called without holding netif_tx_lock. In this case the driver
has to lock by itself when needed. It is recommended to use a try lock
for this and return -1 when the spin lock fails.
The locking there should also properly protect against
@@ -62,12 +62,12 @@ dev->hard_start_xmit:
Only valid when NETIF_F_LLTX is set.
dev->tx_timeout:
- Synchronization: dev->xmit_lock spinlock.
+ Synchronization: netif_tx_lock spinlock.
Context: BHs disabled
Notes: netif_queue_stopped() is guaranteed true
dev->set_multicast_list:
- Synchronization: dev->xmit_lock spinlock.
+ Synchronization: netif_tx_lock spinlock.
Context: BHs disabled
dev->poll:
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
index 1dae4b2..1d917ed 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
@@ -821,7 +821,8 @@ void ipoib_mcast_restart_task(void *dev_
ipoib_mcast_stop_thread(dev, 0);
- spin_lock_irqsave(&dev->xmit_lock, flags);
+ local_irq_save(flags);
+ netif_tx_lock(dev);
spin_lock(&priv->lock);
/*
@@ -896,7 +897,8 @@ void ipoib_mcast_restart_task(void *dev_
}
spin_unlock(&priv->lock);
- spin_unlock_irqrestore(&dev->xmit_lock, flags);
+ netif_tx_unlock(dev);
+ local_irq_restore(flags);
/* We have to cancel outside of the spinlock */
list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
diff --git a/drivers/media/dvb/dvb-core/dvb_net.c b/drivers/media/dvb/dvb-core/dvb_net.c
index 2f0f358..9fd8752 100644
--- a/drivers/media/dvb/dvb-core/dvb_net.c
+++ b/drivers/media/dvb/dvb-core/dvb_net.c
@@ -1052,7 +1052,7 @@ static void wq_set_multicast_list (void
dvb_net_feed_stop(dev);
priv->rx_mode = RX_MODE_UNI;
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
if (dev->flags & IFF_PROMISC) {
dprintk("%s: promiscuous mode\n", dev->name);
@@ -1077,7 +1077,7 @@ static void wq_set_multicast_list (void
}
}
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
dvb_net_feed_start(dev);
}
diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c
index 54161ae..9c5a884 100644
--- a/drivers/net/bnx2.c
+++ b/drivers/net/bnx2.c
@@ -2009,7 +2009,7 @@ bnx2_poll(struct net_device *dev, int *b
return 1;
}
-/* Called with rtnl_lock from vlan functions and also dev->xmit_lock
+/* Called with rtnl_lock from vlan functions and also netif_tx_lock
* from set_multicast.
*/
static void
@@ -4252,7 +4252,7 @@ bnx2_vlan_rx_kill_vid(struct net_device
}
#endif
-/* Called with dev->xmit_lock.
+/* Called with netif_tx_lock.
* hard_start_xmit is pseudo-lockless - a lock is only required when
* the tx queue is full. This way, we get the benefit of lockless
* operations most of the time without the complexities to handle
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 55d2367..46326cd 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -4191,7 +4191,7 @@ static int bond_init(struct net_device *
*/
bond_dev->features |= NETIF_F_VLAN_CHALLENGED;
- /* don't acquire bond device's xmit_lock when
+ /* don't acquire bond device's netif_tx_lock when
* transmitting */
bond_dev->features |= NETIF_F_LLTX;
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index 705e122..cb1b4d0 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -533,9 +533,9 @@ #define NV_MSI_X_VECTOR_OTHER 0x2
* critical parts:
* - rx is (pseudo-) lockless: it relies on the single-threading provided
* by the arch code for interrupts.
- * - tx setup is lockless: it relies on dev->xmit_lock. Actual submission
+ * - tx setup is lockless: it relies on netif_tx_lock. Actual submission
* needs dev->priv->lock :-(
- * - set_multicast_list: preparation lockless, relies on dev->xmit_lock.
+ * - set_multicast_list: preparation lockless, relies on netif_tx_lock.
*/
/* in dev: base, irq */
@@ -1213,7 +1213,7 @@ static void drain_ring(struct net_device
/*
* nv_start_xmit: dev->hard_start_xmit function
- * Called with dev->xmit_lock held.
+ * Called with netif_tx_lock held.
*/
static int nv_start_xmit(struct sk_buff *skb, struct net_device *dev)
{
@@ -1407,7 +1407,7 @@ static void nv_tx_done(struct net_device
/*
* nv_tx_timeout: dev->tx_timeout function
- * Called with dev->xmit_lock held.
+ * Called with netif_tx_lock held.
*/
static void nv_tx_timeout(struct net_device *dev)
{
@@ -1737,7 +1737,7 @@ static int nv_change_mtu(struct net_devi
* Changing the MTU is a rare event, it shouldn't matter.
*/
nv_disable_irq(dev);
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
spin_lock(&np->lock);
/* stop engines */
nv_stop_rx(dev);
@@ -1768,7 +1768,7 @@ static int nv_change_mtu(struct net_devi
nv_start_rx(dev);
nv_start_tx(dev);
spin_unlock(&np->lock);
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
nv_enable_irq(dev);
}
return 0;
@@ -1803,7 +1803,7 @@ static int nv_set_mac_address(struct net
memcpy(dev->dev_addr, macaddr->sa_data, ETH_ALEN);
if (netif_running(dev)) {
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
spin_lock_irq(&np->lock);
/* stop rx engine */
@@ -1815,7 +1815,7 @@ static int nv_set_mac_address(struct net
/* restart rx engine */
nv_start_rx(dev);
spin_unlock_irq(&np->lock);
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
} else {
nv_copy_mac_to_hw(dev);
}
@@ -1824,7 +1824,7 @@ static int nv_set_mac_address(struct net
/*
* nv_set_multicast: dev->set_multicast function
- * Called with dev->xmit_lock held.
+ * Called with netif_tx_lock held.
*/
static void nv_set_multicast(struct net_device *dev)
{
diff --git a/drivers/net/hamradio/6pack.c b/drivers/net/hamradio/6pack.c
index 102c1f0..d12605f 100644
--- a/drivers/net/hamradio/6pack.c
+++ b/drivers/net/hamradio/6pack.c
@@ -308,9 +308,9 @@ static int sp_set_mac_address(struct net
{
struct sockaddr_ax25 *sa = addr;
- spin_lock_irq(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
memcpy(dev->dev_addr, &sa->sax25_call, AX25_ADDR_LEN);
- spin_unlock_irq(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
return 0;
}
@@ -767,9 +767,9 @@ static int sixpack_ioctl(struct tty_stru
break;
}
- spin_lock_irq(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
memcpy(dev->dev_addr, &addr, AX25_ADDR_LEN);
- spin_unlock_irq(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
err = 0;
break;
diff --git a/drivers/net/hamradio/mkiss.c b/drivers/net/hamradio/mkiss.c
index d81a8e1..3ebbbe5 100644
--- a/drivers/net/hamradio/mkiss.c
+++ b/drivers/net/hamradio/mkiss.c
@@ -357,9 +357,9 @@ static int ax_set_mac_address(struct net
{
struct sockaddr_ax25 *sa = addr;
- spin_lock_irq(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
memcpy(dev->dev_addr, &sa->sax25_call, AX25_ADDR_LEN);
- spin_unlock_irq(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
return 0;
}
@@ -886,9 +886,9 @@ static int mkiss_ioctl(struct tty_struct
break;
}
- spin_lock_irq(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
memcpy(dev->dev_addr, addr, AX25_ADDR_LEN);
- spin_unlock_irq(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
err = 0;
break;
diff --git a/drivers/net/ifb.c b/drivers/net/ifb.c
index 31fb2d7..2e222ef 100644
--- a/drivers/net/ifb.c
+++ b/drivers/net/ifb.c
@@ -76,13 +76,13 @@ static void ri_tasklet(unsigned long dev
dp->st_task_enter++;
if ((skb = skb_peek(&dp->tq)) == NULL) {
dp->st_txq_refl_try++;
- if (spin_trylock(&_dev->xmit_lock)) {
+ if (netif_tx_trylock(_dev)) {
dp->st_rxq_enter++;
while ((skb = skb_dequeue(&dp->rq)) != NULL) {
skb_queue_tail(&dp->tq, skb);
dp->st_rx2tx_tran++;
}
- spin_unlock(&_dev->xmit_lock);
+ netif_tx_unlock(_dev);
} else {
/* reschedule */
dp->st_rxq_notenter++;
@@ -110,7 +110,7 @@ static void ri_tasklet(unsigned long dev
}
}
- if (spin_trylock(&_dev->xmit_lock)) {
+ if (netif_tx_trylock(_dev)) {
dp->st_rxq_check++;
if ((skb = skb_peek(&dp->rq)) == NULL) {
dp->tasklet_pending = 0;
@@ -118,10 +118,10 @@ static void ri_tasklet(unsigned long dev
netif_wake_queue(_dev);
} else {
dp->st_rxq_rsch++;
- spin_unlock(&_dev->xmit_lock);
+ netif_tx_unlock(_dev);
goto resched;
}
- spin_unlock(&_dev->xmit_lock);
+ netif_tx_unlock(_dev);
} else {
resched:
dp->tasklet_pending = 1;
diff --git a/drivers/net/irda/vlsi_ir.c b/drivers/net/irda/vlsi_ir.c
index 97a49e0..d70b9e8 100644
--- a/drivers/net/irda/vlsi_ir.c
+++ b/drivers/net/irda/vlsi_ir.c
@@ -959,7 +959,7 @@ static int vlsi_hard_start_xmit(struct s
|| (now.tv_sec==ready.tv_sec && now.tv_usec>=ready.tv_usec))
break;
udelay(100);
- /* must not sleep here - we are called under xmit_lock! */
+ /* must not sleep here - called under netif_tx_lock! */
}
}
diff --git a/drivers/net/natsemi.c b/drivers/net/natsemi.c
index 9062775..2e4eced 100644
--- a/drivers/net/natsemi.c
+++ b/drivers/net/natsemi.c
@@ -318,12 +318,12 @@ performance critical codepaths:
The rx process only runs in the interrupt handler. Access from outside
the interrupt handler is only permitted after disable_irq().
-The rx process usually runs under the dev->xmit_lock. If np->intr_tx_reap
+The rx process usually runs under the netif_tx_lock. If np->intr_tx_reap
is set, then access is permitted under spin_lock_irq(&np->lock).
Thus configuration functions that want to access everything must call
disable_irq(dev->irq);
- spin_lock_bh(dev->xmit_lock);
+ netif_tx_lock_bh(dev);
spin_lock_irq(&np->lock);
IV. Notes
diff --git a/drivers/net/tulip/winbond-840.c b/drivers/net/tulip/winbond-840.c
index 136a70c..6ff016d 100644
--- a/drivers/net/tulip/winbond-840.c
+++ b/drivers/net/tulip/winbond-840.c
@@ -1605,11 +1605,11 @@ #ifdef CONFIG_PM
* - get_stats:
* spin_lock_irq(np->lock), doesn't touch hw if not present
* - hard_start_xmit:
- * netif_stop_queue + spin_unlock_wait(&dev->xmit_lock);
+ * synchronize_irq + netif_tx_disable;
* - tx_timeout:
- * netif_device_detach + spin_unlock_wait(&dev->xmit_lock);
+ * netif_device_detach + netif_tx_disable;
* - set_multicast_list
- * netif_device_detach + spin_unlock_wait(&dev->xmit_lock);
+ * netif_device_detach + netif_tx_disable;
* - interrupt handler
* doesn't touch hw if not present, synchronize_irq waits for
* running instances of the interrupt handler.
@@ -1635,11 +1635,10 @@ static int w840_suspend (struct pci_dev
netif_device_detach(dev);
update_csr6(dev, 0);
iowrite32(0, ioaddr + IntrEnable);
- netif_stop_queue(dev);
spin_unlock_irq(&np->lock);
- spin_unlock_wait(&dev->xmit_lock);
synchronize_irq(dev->irq);
+ netif_disable_tx(dev);
np->stats.rx_missed_errors += ioread32(ioaddr + RxMissed) & 0xffff;
diff --git a/drivers/net/wireless/orinoco.c b/drivers/net/wireless/orinoco.c
index c2d0b09..a5fcfcd 100644
--- a/drivers/net/wireless/orinoco.c
+++ b/drivers/net/wireless/orinoco.c
@@ -1833,7 +1833,9 @@ static int __orinoco_program_rids(struct
/* Set promiscuity / multicast*/
priv->promiscuous = 0;
priv->mc_count = 0;
- __orinoco_set_multicast_list(dev); /* FIXME: what about the xmit_lock */
+
+ /* FIXME: what about netif_tx_lock */
+ __orinoco_set_multicast_list(dev);
return 0;
}
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index f4169bb..01699eb 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -406,7 +406,7 @@ #define NETIF_F_UFO 8192
* One part is mostly used on xmit path (device)
*/
/* hard_start_xmit synchronizer */
- spinlock_t xmit_lock ____cacheline_aligned_in_smp;
+ spinlock_t _xmit_lock ____cacheline_aligned_in_smp;
/* cpu id of processor entered to hard_start_xmit or -1,
if nobody entered there.
*/
@@ -889,11 +889,43 @@ static inline void __netif_rx_complete(s
clear_bit(__LINK_STATE_RX_SCHED, &dev->state);
}
+static inline void netif_tx_lock(struct net_device *dev)
+{
+ spin_lock(&dev->_xmit_lock);
+ dev->xmit_lock_owner = smp_processor_id();
+}
+
+static inline void netif_tx_lock_bh(struct net_device *dev)
+{
+ spin_lock_bh(&dev->_xmit_lock);
+ dev->xmit_lock_owner = smp_processor_id();
+}
+
+static inline int netif_tx_trylock(struct net_device *dev)
+{
+ int err = spin_trylock(&dev->_xmit_lock);
+ if (!err)
+ dev->xmit_lock_owner = smp_processor_id();
+ return err;
+}
+
+static inline void netif_tx_unlock(struct net_device *dev)
+{
+ dev->xmit_lock_owner = -1;
+ spin_unlock(&dev->_xmit_lock);
+}
+
+static inline void netif_tx_unlock_bh(struct net_device *dev)
+{
+ dev->xmit_lock_owner = -1;
+ spin_unlock_bh(&dev->_xmit_lock);
+}
+
static inline void netif_tx_disable(struct net_device *dev)
{
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
netif_stop_queue(dev);
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
}
/* These functions live elsewhere (drivers/net/net_init.c, but related) */
diff --git a/net/atm/clip.c b/net/atm/clip.c
index 72d8529..f92f9c9 100644
--- a/net/atm/clip.c
+++ b/net/atm/clip.c
@@ -98,7 +98,7 @@ static void unlink_clip_vcc(struct clip_
printk(KERN_CRIT "!clip_vcc->entry (clip_vcc %p)\n", clip_vcc);
return;
}
- spin_lock_bh(&entry->neigh->dev->xmit_lock); /* block clip_start_xmit() */
+ netif_tx_lock_bh(entry->neigh->dev); /* block clip_start_xmit() */
entry->neigh->used = jiffies;
for (walk = &entry->vccs; *walk; walk = &(*walk)->next)
if (*walk == clip_vcc) {
@@ -122,7 +122,7 @@ static void unlink_clip_vcc(struct clip_
printk(KERN_CRIT "ATMARP: unlink_clip_vcc failed (entry %p, vcc "
"0x%p)\n", entry, clip_vcc);
out:
- spin_unlock_bh(&entry->neigh->dev->xmit_lock);
+ netif_tx_unlock_bh(entry->neigh->dev);
}
/* The neighbour entry n->lock is held. */
diff --git a/net/core/dev.c b/net/core/dev.c
index 4fba549..8906f59 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1275,15 +1275,13 @@ int __skb_linearize(struct sk_buff *skb,
#define HARD_TX_LOCK(dev, cpu) { \
if ((dev->features & NETIF_F_LLTX) == 0) { \
- spin_lock(&dev->xmit_lock); \
- dev->xmit_lock_owner = cpu; \
+ netif_tx_lock(dev); \
} \
}
#define HARD_TX_UNLOCK(dev) { \
if ((dev->features & NETIF_F_LLTX) == 0) { \
- dev->xmit_lock_owner = -1; \
- spin_unlock(&dev->xmit_lock); \
+ netif_tx_unlock(dev); \
} \
}
@@ -1382,8 +1380,8 @@ #endif
/* The device has no queue. Common case for software devices:
loopback, all the sorts of tunnels...
- Really, it is unlikely that xmit_lock protection is necessary here.
- (f.e. loopback and IP tunnels are clean ignoring statistics
+ Really, it is unlikely that netif_tx_lock protection is necessary
+ here. (f.e. loopback and IP tunnels are clean ignoring statistics
counters.)
However, it is possible, that they rely on protection
made by us here.
@@ -2785,7 +2783,7 @@ int register_netdevice(struct net_device
BUG_ON(dev->reg_state != NETREG_UNINITIALIZED);
spin_lock_init(&dev->queue_lock);
- spin_lock_init(&dev->xmit_lock);
+ spin_lock_init(&dev->_xmit_lock);
dev->xmit_lock_owner = -1;
#ifdef CONFIG_NET_CLS_ACT
spin_lock_init(&dev->ingress_lock);
diff --git a/net/core/dev_mcast.c b/net/core/dev_mcast.c
index 05d6085..c57d887 100644
--- a/net/core/dev_mcast.c
+++ b/net/core/dev_mcast.c
@@ -62,7 +62,7 @@ #include <net/arp.h>
* Device mc lists are changed by bh at least if IPv6 is enabled,
* so that it must be bh protected.
*
- * We block accesses to device mc filters with dev->xmit_lock.
+ * We block accesses to device mc filters with netif_tx_lock.
*/
/*
@@ -93,9 +93,9 @@ static void __dev_mc_upload(struct net_d
void dev_mc_upload(struct net_device *dev)
{
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
__dev_mc_upload(dev);
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
}
/*
@@ -107,7 +107,7 @@ int dev_mc_delete(struct net_device *dev
int err = 0;
struct dev_mc_list *dmi, **dmip;
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
for (dmip = &dev->mc_list; (dmi = *dmip) != NULL; dmip = &dmi->next) {
/*
@@ -139,13 +139,13 @@ int dev_mc_delete(struct net_device *dev
*/
__dev_mc_upload(dev);
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
return 0;
}
}
err = -ENOENT;
done:
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
return err;
}
@@ -160,7 +160,7 @@ int dev_mc_add(struct net_device *dev, v
dmi1 = kmalloc(sizeof(*dmi), GFP_ATOMIC);
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
for (dmi = dev->mc_list; dmi != NULL; dmi = dmi->next) {
if (memcmp(dmi->dmi_addr, addr, dmi->dmi_addrlen) == 0 &&
dmi->dmi_addrlen == alen) {
@@ -176,7 +176,7 @@ int dev_mc_add(struct net_device *dev, v
}
if ((dmi = dmi1) == NULL) {
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
return -ENOMEM;
}
memcpy(dmi->dmi_addr, addr, alen);
@@ -189,11 +189,11 @@ int dev_mc_add(struct net_device *dev, v
__dev_mc_upload(dev);
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
return 0;
done:
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
kfree(dmi1);
return err;
}
@@ -204,7 +204,7 @@ done:
void dev_mc_discard(struct net_device *dev)
{
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
while (dev->mc_list != NULL) {
struct dev_mc_list *tmp = dev->mc_list;
@@ -215,7 +215,7 @@ void dev_mc_discard(struct net_device *d
}
dev->mc_count = 0;
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
}
#ifdef CONFIG_PROC_FS
@@ -250,7 +250,7 @@ static int dev_mc_seq_show(struct seq_fi
struct dev_mc_list *m;
struct net_device *dev = v;
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
for (m = dev->mc_list; m; m = m->next) {
int i;
@@ -262,7 +262,7 @@ static int dev_mc_seq_show(struct seq_fi
seq_putc(seq, '\n');
}
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
return 0;
}
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index e8e05ce..9cb7818 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -273,24 +273,21 @@ static void netpoll_send_skb(struct netp
do {
npinfo->tries--;
- spin_lock(&np->dev->xmit_lock);
- np->dev->xmit_lock_owner = smp_processor_id();
+ netif_tx_lock(np->dev);
/*
* network drivers do not expect to be called if the queue is
* stopped.
*/
if (netif_queue_stopped(np->dev)) {
- np->dev->xmit_lock_owner = -1;
- spin_unlock(&np->dev->xmit_lock);
+ netif_tx_unlock(np->dev);
netpoll_poll(np);
udelay(50);
continue;
}
status = np->dev->hard_start_xmit(skb, np->dev);
- np->dev->xmit_lock_owner = -1;
- spin_unlock(&np->dev->xmit_lock);
+ netif_tx_unlock(np->dev);
/* success */
if(!status) {
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index c23e9c0..67ed14d 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -2897,7 +2897,7 @@ static __inline__ void pktgen_xmit(struc
}
}
- spin_lock_bh(&odev->xmit_lock);
+ netif_tx_lock_bh(odev);
if (!netif_queue_stopped(odev)) {
atomic_inc(&(pkt_dev->skb->users));
@@ -2942,7 +2942,7 @@ static __inline__ void pktgen_xmit(struc
pkt_dev->next_tx_ns = 0;
}
- spin_unlock_bh(&odev->xmit_lock);
+ netif_tx_unlock_bh(odev);
/* If pkt_dev->count is zero, then run forever */
if ((pkt_dev->count != 0) && (pkt_dev->sofar >= pkt_dev->count)) {
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c
index 138ea92..b1e4c5e 100644
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -72,9 +72,9 @@ void qdisc_unlock_tree(struct net_device
dev->queue_lock serializes queue accesses for this device
AND dev->qdisc pointer itself.
- dev->xmit_lock serializes accesses to device driver.
+ netif_tx_lock serializes accesses to device driver.
- dev->queue_lock and dev->xmit_lock are mutually exclusive,
+ dev->queue_lock and netif_tx_lock are mutually exclusive,
if one is grabbed, another must be free.
*/
@@ -108,7 +108,7 @@ int qdisc_restart(struct net_device *dev
* will be requeued.
*/
if (!nolock) {
- if (!spin_trylock(&dev->xmit_lock)) {
+ if (!netif_tx_trylock(dev)) {
collision:
/* So, someone grabbed the driver. */
@@ -126,8 +126,6 @@ int qdisc_restart(struct net_device *dev
__get_cpu_var(netdev_rx_stat).cpu_collision++;
goto requeue;
}
- /* Remember that the driver is grabbed by us. */
- dev->xmit_lock_owner = smp_processor_id();
}
{
@@ -142,8 +140,7 @@ int qdisc_restart(struct net_device *dev
ret = dev->hard_start_xmit(skb, dev);
if (ret == NETDEV_TX_OK) {
if (!nolock) {
- dev->xmit_lock_owner = -1;
- spin_unlock(&dev->xmit_lock);
+ netif_tx_unlock(dev);
}
spin_lock(&dev->queue_lock);
return -1;
@@ -157,8 +154,7 @@ int qdisc_restart(struct net_device *dev
/* NETDEV_TX_BUSY - we need to requeue */
/* Release the driver */
if (!nolock) {
- dev->xmit_lock_owner = -1;
- spin_unlock(&dev->xmit_lock);
+ netif_tx_unlock(dev);
}
spin_lock(&dev->queue_lock);
q = dev->qdisc;
@@ -187,7 +183,7 @@ static void dev_watchdog(unsigned long a
{
struct net_device *dev = (struct net_device *)arg;
- spin_lock(&dev->xmit_lock);
+ netif_tx_lock(dev);
if (dev->qdisc != &noop_qdisc) {
if (netif_device_present(dev) &&
netif_running(dev) &&
@@ -203,7 +199,7 @@ static void dev_watchdog(unsigned long a
dev_hold(dev);
}
}
- spin_unlock(&dev->xmit_lock);
+ netif_tx_unlock(dev);
dev_put(dev);
}
@@ -227,17 +223,17 @@ void __netdev_watchdog_up(struct net_dev
static void dev_watchdog_up(struct net_device *dev)
{
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
__netdev_watchdog_up(dev);
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
}
static void dev_watchdog_down(struct net_device *dev)
{
- spin_lock_bh(&dev->xmit_lock);
+ netif_tx_lock_bh(dev);
if (del_timer(&dev->watchdog_timer))
dev_put(dev);
- spin_unlock_bh(&dev->xmit_lock);
+ netif_tx_unlock_bh(dev);
}
void netif_carrier_on(struct net_device *dev)
@@ -582,7 +578,7 @@ void dev_deactivate(struct net_device *d
while (test_bit(__LINK_STATE_SCHED, &dev->state))
yield();
- spin_unlock_wait(&dev->xmit_lock);
+ spin_unlock_wait(&dev->_xmit_lock);
}
void dev_init_scheduler(struct net_device *dev)
diff --git a/net/sched/sch_teql.c b/net/sched/sch_teql.c
index 79b8ef3..4c16ad5 100644
--- a/net/sched/sch_teql.c
+++ b/net/sched/sch_teql.c
@@ -302,20 +302,17 @@ restart:
switch (teql_resolve(skb, skb_res, slave)) {
case 0:
- if (spin_trylock(&slave->xmit_lock)) {
- slave->xmit_lock_owner = smp_processor_id();
+ if (netif_tx_trylock(slave)) {
if (!netif_queue_stopped(slave) &&
slave->hard_start_xmit(skb, slave) == 0) {
- slave->xmit_lock_owner = -1;
- spin_unlock(&slave->xmit_lock);
+ netif_tx_unlock(slave);
master->slaves = NEXT_SLAVE(q);
netif_wake_queue(dev);
master->stats.tx_packets++;
master->stats.tx_bytes += len;
return 0;
}
- slave->xmit_lock_owner = -1;
- spin_unlock(&slave->xmit_lock);
+ netif_tx_unlock(slave);
}
if (netif_queue_stopped(dev))
busy = 1;
^ permalink raw reply related
* [RFT] r8169: MAC address change support
From: Francois Romieu @ 2006-05-31 22:59 UTC (permalink / raw)
To: netdev; +Cc: Stefano Cavallari, Tim Mattox
It works fine here (x86 so far) but I would welcome more testers.
The patch applies against 2.6.17-rcX.
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 0ad3310..4208d9a 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -1381,6 +1381,41 @@ static void rtl8169_netpoll(struct net_d
}
#endif
+static void __rtl8169_set_mac_addr(struct net_device *dev, void __iomem *ioaddr)
+{
+ unsigned int i, j;
+
+ RTL_W8(Cfg9346, Cfg9346_Unlock);
+ for (i = 0; i < 2; i++) {
+ __le32 l = 0;
+
+ for (j = 0; j < 4; j++) {
+ l <<= 8;
+ l |= dev->dev_addr[4*i + j];
+ }
+ RTL_W32(MAC0 + 4*i, cpu_to_be32(l));
+ }
+ RTL_W8(Cfg9346, Cfg9346_Lock);
+}
+
+static int rtl8169_set_mac_addr(struct net_device *dev, void *p)
+{
+ struct rtl8169_private *tp = netdev_priv(dev);
+ struct sockaddr *addr = p;
+
+ if (!is_valid_ether_addr(addr->sa_data))
+ return -EINVAL;
+
+ memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
+
+ if (netif_running(dev)) {
+ spin_lock_irq(&tp->lock);
+ __rtl8169_set_mac_addr(dev, tp->mmio_addr);
+ spin_unlock_irq(&tp->lock);
+ }
+ return 0;
+}
+
static void rtl8169_release_board(struct pci_dev *pdev, struct net_device *dev,
void __iomem *ioaddr)
{
@@ -1616,6 +1651,7 @@ rtl8169_init_one(struct pci_dev *pdev, c
dev->stop = rtl8169_close;
dev->tx_timeout = rtl8169_tx_timeout;
dev->set_multicast_list = rtl8169_set_rx_mode;
+ dev->set_mac_address = rtl8169_set_mac_addr;
dev->watchdog_timeo = RTL8169_TX_TIMEOUT;
dev->irq = pdev->irq;
dev->base_addr = (unsigned long) ioaddr;
@@ -1849,6 +1885,8 @@ rtl8169_hw_start(struct net_device *dev)
/* Enable all known interrupts by setting the interrupt mask. */
RTL_W16(IntrMask, rtl8169_intr_mask);
+ __rtl8169_set_mac_addr(dev, ioaddr);
+
netif_start_queue(dev);
}
^ permalink raw reply related
* monitor_during_oper on rt*pci (or any other pcmcia card)?
From: Johannes Berg @ 2006-06-01 11:47 UTC (permalink / raw)
To: netdev; +Cc: Ivo van Doorn, rt2x00-devel, Jean Tourrilhes, Florian.Rampp
[-- Attachment #1: Type: text/plain, Size: 734 bytes --]
Hey,
Florian approached me with a project where you need a pcmcia card that
can have monitor_during_oper. I'd use bcm43xx but unfortunately the
signal strength calculations are way off right now, and that's another
requirement.
Can the rt family of cards be made to support this with reliable signal
strength readings?
Or maybe I should rephrase the questions: which pcmcia cards can support
this in their hardware? I might be willing to port their drivers over to
d80211 just to get some more familiarity with the stack...
Now, if no one can point me to ones that do have working signal
strength, I'll probably have to do more reverse engineering on bcm43xx
to get some reliable readings :/
Thanks,
johannes
[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 793 bytes --]
^ permalink raw reply
* [PATCH 3/7] AMSO1100 WR / Event Definitions.
From: Steve Wise @ 2006-06-01 13:20 UTC (permalink / raw)
To: netdev, linux-kernel
In-Reply-To: <20060531182733.3652.54755.stgit@stevo-desktop>
[-- Attachment #1: Type: text/plain, Size: 221 bytes --]
Sorry for this, but patch 3 is apparently getting eaten by the list
servers for linux-kernel and netdev. Here is patch 3 as a tarball to
see if it goes through...
Steve.
[PATCH 3/7] AMSO1100 WR / Event Definitions.
[-- Attachment #2: amso1100_wr.tgz --]
[-- Type: application/x-compressed-tar, Size: 10461 bytes --]
^ permalink raw reply
* Re: Refactor Netlink connector?
From: James Morris @ 2006-06-01 14:24 UTC (permalink / raw)
To: Thomas Graf
Cc: jamal, Evgeniy Polyakov, netdev, David S. Miller, Stephen Smalley
In-Reply-To: <20060601104506.GF7844@postel.suug.ch>
On Thu, 1 Jun 2006, Thomas Graf wrote:
> It shouldn't be hard to split what is implemented in nlmsg_route_perms[]
> for NETLINK_ROUTE into the definitions of the generic netlink
> operations, could look like this:
>
> struct genl_ops some_op = {
> [...]
> .perm = NETLINK_GENERIC_SOCKET__NLMSG_READ,
> };
We wouldn't need the socket class outside of SELinux, just the perm, so
something like:
NL_PERM_READ
> int genl_peek_cmd(struct nlmsghdr *nlh)
> {
> struct genlmsghdr *hdr = nlmsg_data(nlh);
>
> if (nlh->nlmsglen < nlmsg_msg_sizeo(GENL_HDRLEN))
> return -EINVAL;
>
> return hdr->cmd;
> }
Unless I'm mistaken, people are already multiplexing commands inside genl
commands (and if so, why even bother with registerable ops?).
I'll look at it in more detail soon.
--
James Morris
<jmorris@namei.org>
^ permalink raw reply
* [PATCH] softmac: complete shared key authentication
From: Daniel Drake @ 2006-06-01 14:34 UTC (permalink / raw)
To: linville; +Cc: netdev, johannes, ipw2100-admin
This patch finishes of the partially-complete shared key authentication
implementation in softmac.
The complication here is that we need to encrypt a management frame during
the authentication process. I don't think there are any other scenarios where
this would have to happen.
To get around this without causing too many headaches, we decided to just use
software encryption for this frame. The softmac config option now selects
IEEE80211_CRYPT_WEP so that we can ensure this available. This also involved
a modification to some otherwise unused ieee80211 API.
Signed-off-by: Daniel Drake <dsd@gentoo.org>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Index: linux/net/ieee80211/softmac/ieee80211softmac_auth.c
===================================================================
--- linux.orig/net/ieee80211/softmac/ieee80211softmac_auth.c
+++ linux/net/ieee80211/softmac/ieee80211softmac_auth.c
@@ -107,6 +107,7 @@ ieee80211softmac_auth_queue(void *data)
printkl(KERN_WARNING PFX "Authentication timed out with "MAC_FMT"\n", MAC_ARG(net->bssid));
/* Remove this item from the queue */
spin_lock_irqsave(&mac->lock, flags);
+ net->authenticating = 0;
ieee80211softmac_call_events_locked(mac, IEEE80211SOFTMAC_EVENT_AUTH_TIMEOUT, net);
cancel_delayed_work(&auth->work); /* just to make sure... */
list_del(&auth->list);
@@ -212,13 +213,13 @@ ieee80211softmac_auth_resp(struct net_de
aq->state = IEEE80211SOFTMAC_AUTH_SHARED_RESPONSE;
spin_unlock_irqrestore(&mac->lock, flags);
- /* Switch to correct channel for this network */
- mac->set_channel(mac->dev, net->channel);
-
- /* Send our response (How to encrypt?) */
+ /* Send our response */
ieee80211softmac_send_mgt_frame(mac, aq->net, IEEE80211_STYPE_AUTH, aq->state);
- break;
+ return 0;
case IEEE80211SOFTMAC_AUTH_SHARED_PASS:
+ kfree(net->challenge);
+ net->challenge = NULL;
+ net->challenge_len = 0;
/* Check the status code of the response */
switch(auth->status) {
case WLAN_STATUS_SUCCESS:
@@ -229,6 +230,7 @@ ieee80211softmac_auth_resp(struct net_de
spin_unlock_irqrestore(&mac->lock, flags);
printkl(KERN_NOTICE PFX "Shared Key Authentication completed with "MAC_FMT"\n",
MAC_ARG(net->bssid));
+ ieee80211softmac_call_events(mac, IEEE80211SOFTMAC_EVENT_AUTHENTICATED, net);
break;
default:
printkl(KERN_NOTICE PFX "Shared Key Authentication with "MAC_FMT" failed, error code: %i\n",
Index: linux/net/ieee80211/softmac/ieee80211softmac_io.c
===================================================================
--- linux.orig/net/ieee80211/softmac/ieee80211softmac_io.c
+++ linux/net/ieee80211/softmac/ieee80211softmac_io.c
@@ -268,26 +268,27 @@ ieee80211softmac_reassoc_req(struct ieee
static u32
ieee80211softmac_auth(struct ieee80211_auth **pkt,
struct ieee80211softmac_device *mac, struct ieee80211softmac_network *net,
- u16 transaction, u16 status)
+ u16 transaction, u16 status, int *encrypt_mpdu)
{
u8 *data;
+ int auth_mode = mac->ieee->sec.auth_mode;
+ int is_shared_response = (auth_mode == WLAN_AUTH_SHARED_KEY
+ && transaction == IEEE80211SOFTMAC_AUTH_SHARED_RESPONSE);
+
/* Allocate Packet */
(*pkt) = (struct ieee80211_auth *)ieee80211softmac_alloc_mgt(
2 + /* Auth Algorithm */
2 + /* Auth Transaction Seq */
2 + /* Status Code */
/* Challenge Text IE */
- mac->ieee->open_wep ? 0 :
- 1 + 1 + WLAN_AUTH_CHALLENGE_LEN
- );
+ is_shared_response ? 0 : 1 + 1 + net->challenge_len
+ );
if (unlikely((*pkt) == NULL))
return 0;
ieee80211softmac_hdr_3addr(mac, &((*pkt)->header), IEEE80211_STYPE_AUTH, net->bssid, net->bssid);
/* Algorithm */
- (*pkt)->algorithm = mac->ieee->open_wep ?
- cpu_to_le16(WLAN_AUTH_OPEN) :
- cpu_to_le16(WLAN_AUTH_SHARED_KEY);
+ (*pkt)->algorithm = cpu_to_le16(auth_mode);
/* Transaction */
(*pkt)->transaction = cpu_to_le16(transaction);
/* Status */
@@ -295,18 +296,20 @@ ieee80211softmac_auth(struct ieee80211_a
data = (u8 *)(*pkt)->info_element;
/* Challenge Text */
- if(!mac->ieee->open_wep){
+ if (is_shared_response) {
*data = MFIE_TYPE_CHALLENGE;
data++;
/* Copy the challenge in */
- // *data = challenge length
- // data += sizeof(u16);
- // memcpy(data, challenge, challenge length);
- // data += challenge length;
-
- /* Add the full size to the packet length */
- }
+ *data = net->challenge_len;
+ data++;
+ memcpy(data, net->challenge, net->challenge_len);
+ data += net->challenge_len;
+
+ /* Make sure this frame gets encrypted with the shared key */
+ *encrypt_mpdu = 1;
+ } else
+ *encrypt_mpdu = 0;
/* Return the packet size */
return (data - (u8 *)(*pkt));
@@ -396,6 +399,7 @@ ieee80211softmac_send_mgt_frame(struct i
{
void *pkt = NULL;
u32 pkt_size = 0;
+ int encrypt_mpdu = 0;
switch(type) {
case IEEE80211_STYPE_ASSOC_REQ:
@@ -405,7 +409,7 @@ ieee80211softmac_send_mgt_frame(struct i
pkt_size = ieee80211softmac_reassoc_req((struct ieee80211_reassoc_request **)(&pkt), mac, (struct ieee80211softmac_network *)ptrarg);
break;
case IEEE80211_STYPE_AUTH:
- pkt_size = ieee80211softmac_auth((struct ieee80211_auth **)(&pkt), mac, (struct ieee80211softmac_network *)ptrarg, (u16)(arg & 0xFFFF), (u16) (arg >> 16));
+ pkt_size = ieee80211softmac_auth((struct ieee80211_auth **)(&pkt), mac, (struct ieee80211softmac_network *)ptrarg, (u16)(arg & 0xFFFF), (u16) (arg >> 16), &encrypt_mpdu);
break;
case IEEE80211_STYPE_DISASSOC:
case IEEE80211_STYPE_DEAUTH:
@@ -434,7 +438,8 @@ ieee80211softmac_send_mgt_frame(struct i
* or get rid of it alltogether?
* Does this work for you now?
*/
- ieee80211_tx_frame(mac->ieee, (struct ieee80211_hdr *)pkt, pkt_size);
+ ieee80211_tx_frame(mac->ieee, (struct ieee80211_hdr *)pkt,
+ IEEE80211_3ADDR_LEN, pkt_size, encrypt_mpdu);
kfree(pkt);
return 0;
@@ -479,7 +484,7 @@ ieee80211softmac_send_ctl_frame(struct i
return -ENOMEM;
/* Send the packet to the ieee80211 layer for tx */
- ieee80211_tx_frame(mac->ieee, (struct ieee80211_hdr *) pkt, pkt_size);
+ ieee80211_tx_frame(mac->ieee, (struct ieee80211_hdr *) pkt, IEEE80211_2ADDR_LEN, pkt_size, 0);
kfree(pkt);
return 0;
Index: linux/include/net/ieee80211.h
===================================================================
--- linux.orig/include/net/ieee80211.h
+++ linux/include/net/ieee80211.h
@@ -1241,7 +1241,8 @@ extern int ieee80211_set_encryption(stru
extern int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev);
extern void ieee80211_txb_free(struct ieee80211_txb *);
extern int ieee80211_tx_frame(struct ieee80211_device *ieee,
- struct ieee80211_hdr *frame, int len);
+ struct ieee80211_hdr *frame, int hdr_len,
+ int total_len, int encrypt_mpdu);
/* ieee80211_rx.c */
extern int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb,
Index: linux/net/ieee80211/ieee80211_tx.c
===================================================================
--- linux.orig/net/ieee80211/ieee80211_tx.c
+++ linux/net/ieee80211/ieee80211_tx.c
@@ -516,7 +516,8 @@ int ieee80211_xmit(struct sk_buff *skb,
/* Incoming 802.11 strucure is converted to a TXB
* a block of 802.11 fragment packets (stored as skbs) */
int ieee80211_tx_frame(struct ieee80211_device *ieee,
- struct ieee80211_hdr *frame, int len)
+ struct ieee80211_hdr *frame, int hdr_len, int total_len,
+ int encrypt_mpdu)
{
struct ieee80211_txb *txb = NULL;
unsigned long flags;
@@ -526,6 +527,9 @@ int ieee80211_tx_frame(struct ieee80211_
spin_lock_irqsave(&ieee->lock, flags);
+ if (encrypt_mpdu && !ieee->sec.encrypt)
+ encrypt_mpdu = 0;
+
/* If there is no driver handler to take the TXB, dont' bother
* creating it... */
if (!ieee->hard_start_xmit) {
@@ -533,32 +537,41 @@ int ieee80211_tx_frame(struct ieee80211_
goto success;
}
- if (unlikely(len < 24)) {
+ if (unlikely(total_len < 24)) {
printk(KERN_WARNING "%s: skb too small (%d).\n",
- ieee->dev->name, len);
+ ieee->dev->name, total_len);
goto success;
}
+ if (encrypt_mpdu)
+ frame->frame_ctl |= cpu_to_le16(IEEE80211_FCTL_PROTECTED);
+
/* When we allocate the TXB we allocate enough space for the reserve
* and full fragment bytes (bytes_per_frag doesn't include prefix,
* postfix, header, FCS, etc.) */
- txb = ieee80211_alloc_txb(1, len, ieee->tx_headroom, GFP_ATOMIC);
+ txb = ieee80211_alloc_txb(1, total_len, ieee->tx_headroom, GFP_ATOMIC);
if (unlikely(!txb)) {
printk(KERN_WARNING "%s: Could not allocate TXB\n",
ieee->dev->name);
goto failed;
}
txb->encrypted = 0;
- txb->payload_size = len;
+ txb->payload_size = total_len;
skb_frag = txb->fragments[0];
- memcpy(skb_put(skb_frag, len), frame, len);
+ memcpy(skb_put(skb_frag, total_len), frame, total_len);
if (ieee->config &
(CFG_IEEE80211_COMPUTE_FCS | CFG_IEEE80211_RESERVE_FCS))
skb_put(skb_frag, 4);
+ /* To avoid overcomplicating things, we do the corner-case frame
+ * encryption in software. The only real situation where encryption is
+ * needed here is during software-based shared key authentication. */
+ if (encrypt_mpdu)
+ ieee80211_encrypt_fragment(ieee, skb_frag, hdr_len);
+
success:
spin_unlock_irqrestore(&ieee->lock, flags);
Index: linux/net/ieee80211/softmac/Kconfig
===================================================================
--- linux.orig/net/ieee80211/softmac/Kconfig
+++ linux/net/ieee80211/softmac/Kconfig
@@ -2,6 +2,7 @@ config IEEE80211_SOFTMAC
tristate "Software MAC add-on to the IEEE 802.11 networking stack"
depends on IEEE80211 && EXPERIMENTAL
select WIRELESS_EXT
+ select IEEE80211_CRYPT_WEP
---help---
This option enables the hardware independent software MAC addon
for the IEEE 802.11 networking stack.
^ permalink raw reply
* [PATCH] softmac: Fix handling of authentication failure
From: Daniel Drake @ 2006-06-01 14:37 UTC (permalink / raw)
To: linville; +Cc: netdev, johannes
My router blew up earlier, but exhibited some interesting behaviour during
its dying moments. It was broadcasting beacons but wouldn't respond to
any authentication requests.
I noticed that softmac wasn't playing nice with this, as I couldn't make it try
to connect to other networks after it had timed out authenticating to my ill
router.
To resolve this, I modified the softmac event/notify API to pass the event
code to the callback, so that callbacks being notified from
IEEE80211SOFTMAC_EVENT_ANY masks can make some judgement. In this case, the
ieee80211softmac_assoc callback needs to make a decision based upon whether
the association passed or failed.
Signed-off-by: Daniel Drake <dsd@gentoo.org>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
---
This may be worth considering for 2.6.17. The current code does not check the
response to an authentication request before associating, it just assumes it
was successful. This patch adds validation on that.
Index: linux/include/net/ieee80211softmac.h
===================================================================
--- linux.orig/include/net/ieee80211softmac.h
+++ linux/include/net/ieee80211softmac.h
@@ -359,7 +359,7 @@ extern void ieee80211softmac_stop(struct
* - context set to the context data you want passed
* The return value is 0, or an error.
*/
-typedef void (*notify_function_ptr)(struct net_device *dev, void *context);
+typedef void (*notify_function_ptr)(struct net_device *dev, int event_type, void *context);
#define ieee80211softmac_notify(dev, event, fun, context) ieee80211softmac_notify_gfp(dev, event, fun, context, GFP_KERNEL);
#define ieee80211softmac_notify_atomic(dev, event, fun, context) ieee80211softmac_notify_gfp(dev, event, fun, context, GFP_ATOMIC);
Index: linux/net/ieee80211/softmac/ieee80211softmac_assoc.c
===================================================================
--- linux.orig/net/ieee80211/softmac/ieee80211softmac_assoc.c
+++ linux/net/ieee80211/softmac/ieee80211softmac_assoc.c
@@ -164,12 +164,28 @@ network_matches_request(struct ieee80211
}
static void
-ieee80211softmac_assoc_notify(struct net_device *dev, void *context)
+ieee80211softmac_assoc_notify_scan(struct net_device *dev, int event_type, void *context)
{
struct ieee80211softmac_device *mac = ieee80211_priv(dev);
ieee80211softmac_assoc_work((void*)mac);
}
+static void
+ieee80211softmac_assoc_notify_auth(struct net_device *dev, int event_type, void *context)
+{
+ struct ieee80211softmac_device *mac = ieee80211_priv(dev);
+
+ switch (event_type) {
+ case IEEE80211SOFTMAC_EVENT_AUTHENTICATED:
+ ieee80211softmac_assoc_work((void*)mac);
+ break;
+ case IEEE80211SOFTMAC_EVENT_AUTH_FAILED:
+ case IEEE80211SOFTMAC_EVENT_AUTH_TIMEOUT:
+ ieee80211softmac_disassoc(mac);
+ break;
+ }
+}
+
/* This function is called to handle userspace requests (asynchronously) */
void
ieee80211softmac_assoc_work(void *d)
@@ -249,7 +265,7 @@ ieee80211softmac_assoc_work(void *d)
* Maybe we can hope to have more memory after scanning finishes ;)
*/
dprintk(KERN_INFO PFX "Associate: Scanning for networks first.\n");
- ieee80211softmac_notify(mac->dev, IEEE80211SOFTMAC_EVENT_SCAN_FINISHED, ieee80211softmac_assoc_notify, NULL);
+ ieee80211softmac_notify(mac->dev, IEEE80211SOFTMAC_EVENT_SCAN_FINISHED, ieee80211softmac_assoc_notify_scan, NULL);
if (ieee80211softmac_start_scan(mac))
dprintk(KERN_INFO PFX "Associate: failed to initiate scan. Is device up?\n");
return;
@@ -284,7 +300,7 @@ ieee80211softmac_assoc_work(void *d)
* otherwise adding the notification would be racy. */
if (!ieee80211softmac_auth_req(mac, found)) {
dprintk(KERN_INFO PFX "cannot associate without being authenticated, requested authentication\n");
- ieee80211softmac_notify_internal(mac, IEEE80211SOFTMAC_EVENT_ANY, found, ieee80211softmac_assoc_notify, NULL, GFP_KERNEL);
+ ieee80211softmac_notify_internal(mac, IEEE80211SOFTMAC_EVENT_ANY, found, ieee80211softmac_assoc_notify_auth, NULL, GFP_KERNEL);
} else {
printkl(KERN_WARNING PFX "Not authenticated, but requesting authentication failed. Giving up to associate\n");
ieee80211softmac_call_events(mac, IEEE80211SOFTMAC_EVENT_ASSOCIATE_FAILED, found);
Index: linux/net/ieee80211/softmac/ieee80211softmac_event.c
===================================================================
--- linux.orig/net/ieee80211/softmac/ieee80211softmac_event.c
+++ linux/net/ieee80211/softmac/ieee80211softmac_event.c
@@ -77,7 +77,7 @@ ieee80211softmac_notify_callback(void *d
struct ieee80211softmac_event event = *(struct ieee80211softmac_event*) d;
kfree(d);
- event.fun(event.mac->dev, event.context);
+ event.fun(event.mac->dev, event.event_type, event.context);
}
int
@@ -172,6 +172,9 @@ ieee80211softmac_call_events_locked(stru
if ((eventptr->event_type == event || eventptr->event_type == -1)
&& (eventptr->event_context == NULL || eventptr->event_context == event_ctx)) {
list_del(&eventptr->list);
+ /* User may have subscribed to ANY event, so
+ * we tell them which event triggered it. */
+ eventptr->event_type = event;
schedule_work(&eventptr->work);
}
}
^ permalink raw reply
* [PATCH] softmac: unified capabilities computation
From: Daniel Drake @ 2006-06-01 16:12 UTC (permalink / raw)
To: linville; +Cc: johannes, josejx, netdev
From: Joseph Jezak <josejx@gentoo.org>
This patch moves the capabilities field computation to a function for clarity
and adds some previously unimplemented bits.
Signed off by Joseph Jezak <josejx@gentoo.org>
Signed-off-by: Daniel Drake <dsd@gentoo.org>
Acked-By: Johannes Berg <johannes@sipsolutions.net>
---
Index: linux/net/ieee80211/softmac/ieee80211softmac_io.c
===================================================================
--- linux.orig/net/ieee80211/softmac/ieee80211softmac_io.c
+++ linux/net/ieee80211/softmac/ieee80211softmac_io.c
@@ -149,6 +149,56 @@ ieee80211softmac_hdr_3addr(struct ieee80
* shouldn't the sequence number be in ieee80211? */
}
+static u16
+ieee80211softmac_capabilities(struct ieee80211softmac_device *mac,
+ struct ieee80211softmac_network *net)
+{
+ u16 capability = 0;
+
+ /* ESS and IBSS bits are set according to the current mode */
+ switch (mac->ieee->iw_mode) {
+ case IW_MODE_INFRA:
+ capability = cpu_to_le16(WLAN_CAPABILITY_ESS);
+ break;
+ case IW_MODE_ADHOC:
+ capability = cpu_to_le16(WLAN_CAPABILITY_IBSS);
+ break;
+ case IW_MODE_AUTO:
+ capability = net->capabilities &
+ (WLAN_CAPABILITY_ESS|WLAN_CAPABILITY_IBSS);
+ break;
+ default:
+ /* bleh. we don't ever go to these modes */
+ printk(KERN_ERR PFX "invalid iw_mode!\n");
+ break;
+ }
+
+ /* CF Pollable / CF Poll Request */
+ /* Needs to be implemented, for now, the 0's == not supported */
+
+ /* Privacy Bit */
+ capability |= mac->ieee->sec.level ?
+ cpu_to_le16(WLAN_CAPABILITY_PRIVACY) : 0;
+
+ /* Short Preamble */
+ /* Always supported: we probably won't ever be powering devices which
+ * dont support this... */
+ capability |= WLAN_CAPABILITY_SHORT_PREAMBLE;
+
+ /* PBCC */
+ /* Not widely used */
+
+ /* Channel Agility */
+ /* Not widely used */
+
+ /* Short Slot */
+ /* Will be implemented later */
+
+ /* DSSS-OFDM */
+ /* Not widely used */
+
+ return capability;
+}
/*****************************************************************************
* Create Management packets
@@ -179,27 +229,6 @@ ieee80211softmac_assoc_req(struct ieee80
return 0;
ieee80211softmac_hdr_3addr(mac, &((*pkt)->header), IEEE80211_STYPE_ASSOC_REQ, net->bssid, net->bssid);
- /* Fill in capability Info */
- switch (mac->ieee->iw_mode) {
- case IW_MODE_INFRA:
- (*pkt)->capability = cpu_to_le16(WLAN_CAPABILITY_ESS);
- break;
- case IW_MODE_ADHOC:
- (*pkt)->capability = cpu_to_le16(WLAN_CAPABILITY_IBSS);
- break;
- case IW_MODE_AUTO:
- (*pkt)->capability = net->capabilities & (WLAN_CAPABILITY_ESS|WLAN_CAPABILITY_IBSS);
- break;
- default:
- /* bleh. we don't ever go to these modes */
- printk(KERN_ERR PFX "invalid iw_mode!\n");
- break;
- }
- /* Need to add this
- (*pkt)->capability |= mac->ieee->short_slot ?
- cpu_to_le16(WLAN_CAPABILITY_SHORT_SLOT_TIME) : 0;
- */
- (*pkt)->capability |= mac->ieee->sec.level ? cpu_to_le16(WLAN_CAPABILITY_PRIVACY) : 0;
/* Fill in Listen Interval (?) */
(*pkt)->listen_interval = cpu_to_le16(10);
@@ -239,17 +268,9 @@ ieee80211softmac_reassoc_req(struct ieee
return 0;
ieee80211softmac_hdr_3addr(mac, &((*pkt)->header), IEEE80211_STYPE_REASSOC_REQ, net->bssid, net->bssid);
- /* Fill in capability Info */
- (*pkt)->capability = mac->ieee->iw_mode == IW_MODE_MASTER ?
- cpu_to_le16(WLAN_CAPABILITY_ESS) :
- cpu_to_le16(WLAN_CAPABILITY_IBSS);
- /*
- (*pkt)->capability |= mac->ieee->short_slot ?
- cpu_to_le16(WLAN_CAPABILITY_SHORT_SLOT_TIME) : 0;
- */
- (*pkt)->capability |= mac->ieee->sec.level ?
- cpu_to_le16(WLAN_CAPABILITY_PRIVACY) : 0;
-
+ /* Fill in the capabilities */
+ (*pkt)->capability = ieee80211softmac_capabilities(mac, net);
+
/* Fill in Listen Interval (?) */
(*pkt)->listen_interval = cpu_to_le16(10);
/* Fill in the current AP MAC */
^ permalink raw reply
* Re: [PATCH 1/2] iWARP Connection Manager.
From: Steve Wise @ 2006-06-01 17:00 UTC (permalink / raw)
To: Sean Hefty; +Cc: rdreier, linux-kernel, netdev, openib-general
In-Reply-To: <447E1720.7000307@ichips.intel.com>
On Wed, 2006-05-31 at 15:22 -0700, Sean Hefty wrote:
> Steve Wise wrote:
> > +/*
> > + * Release a reference on cm_id. If the last reference is being removed
> > + * and iw_destroy_cm_id is waiting, wake up the waiting thread.
> > + */
> > +static int iwcm_deref_id(struct iwcm_id_private *cm_id_priv)
> > +{
> > + int ret = 0;
> > +
> > + BUG_ON(atomic_read(&cm_id_priv->refcount)==0);
> > + if (atomic_dec_and_test(&cm_id_priv->refcount)) {
> > + BUG_ON(!list_empty(&cm_id_priv->work_list));
> > + if (waitqueue_active(&cm_id_priv->destroy_wait)) {
> > + BUG_ON(cm_id_priv->state != IW_CM_STATE_DESTROYING);
> > + BUG_ON(test_bit(IWCM_F_CALLBACK_DESTROY,
> > + &cm_id_priv->flags));
> > + ret = 1;
> > + wake_up(&cm_id_priv->destroy_wait);
>
> We recently changed the RDMA CM, IB CM, and a couple of other modules from using
> wait objects to completions. This avoids a race condition between decrementing
> the reference count, which allows destruction to proceed, and calling wake_up on
> a freed cm_id. My guess is that you may need to do the same.
>
Good catch. Yes, the IW CM suffers from the same race condition. I'll
change this to use completions...
> Can you also explain the use of the return value here? It's ignored below in
> rem_ref() and destroy_cm_id().
>
The return value is supposed to indicate whether this call to deref
_may_ have resulted in waking up another thread and the cm_id being
freed. Its used in cm_work_handler(), in conjunction with setting the
IWCM_F_CALLBACK_DESTROY flag to know whether the cm_id needs to be freed
on the callback path.
> > +static void add_ref(struct iw_cm_id *cm_id)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + atomic_inc(&cm_id_priv->refcount);
> > +}
> > +
> > +static void rem_ref(struct iw_cm_id *cm_id)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + iwcm_deref_id(cm_id_priv);
> > +}
> > +
>
> > +/*
> > + * CM_ID <-- CLOSING
> > + *
> > + * Block if a passive or active connection is currenlty being processed. Then
> > + * process the event as follows:
> > + * - If we are ESTABLISHED, move to CLOSING and modify the QP state
> > + * based on the abrupt flag
> > + * - If the connection is already in the CLOSING or IDLE state, the peer is
> > + * disconnecting concurrently with us and we've already seen the
> > + * DISCONNECT event -- ignore the request and return 0
> > + * - Disconnect on a listening endpoint returns -EINVAL
> > + */
> > +int iw_cm_disconnect(struct iw_cm_id *cm_id, int abrupt)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + unsigned long flags;
> > + int ret = 0;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + /* Wait if we're currently in a connect or accept downcall */
> > + wait_event(cm_id_priv->connect_wait,
> > + !test_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags));
>
> Am I understanding this check correctly? You're checking to see if the user has
> called iw_cm_disconnect() at the same time that they called iw_cm_connect() or
> iw_cm_accept(). Are connect / accept blocking, or are you just waiting for an
> event?
The CM must wait for the low level provider to finish a connect() or
accept() operation before telling the low level provider to disconnect
via modifying the iwarp QP. Regardless of whether they block, this
disconnect can happen concurrently with the connect/accept so we need to
hold the disconnect until the connect/accept completes.
>
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_ESTABLISHED:
> > + cm_id_priv->state = IW_CM_STATE_CLOSING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->qp) { /* QP could be <nul> for user-mode client */
> > + if (abrupt)
> > + ret = iwcm_modify_qp_err(cm_id_priv->qp);
> > + else
> > + ret = iwcm_modify_qp_sqd(cm_id_priv->qp);
> > + /*
> > + * If both sides are disconnecting the QP could
> > + * already be in ERR or SQD states
> > + */
> > + ret = 0;
> > + }
> > + else
> > + ret = -EINVAL;
> > + break;
> > + case IW_CM_STATE_LISTEN:
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = -EINVAL;
> > + break;
> > + case IW_CM_STATE_CLOSING:
> > + /* remote peer closed first */
> > + case IW_CM_STATE_IDLE:
> > + /* accept or connect returned !0 */
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + break;
> > + case IW_CM_STATE_CONN_RECV:
> > + /*
> > + * App called disconnect before/without calling accept after
> > + * connect_request event delivered.
> > + */
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + break;
> > + case IW_CM_STATE_CONN_SENT:
> > + /* Can only get here if wait above fails */
> > + default:
> > + BUG_ON(1);
> > + }
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_disconnect);
> > +static void destroy_cm_id(struct iw_cm_id *cm_id)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + unsigned long flags;
> > + int ret;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + /* Wait if we're currently in a connect or accept downcall. A
> > + * listening endpoint should never block here. */
> > + wait_event(cm_id_priv->connect_wait,
> > + !test_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags));
>
> Same question/comment as above.
>
Same answer.
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_LISTEN:
> > + cm_id_priv->state = IW_CM_STATE_DESTROYING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + /* destroy the listening endpoint */
> > + ret = cm_id->device->iwcm->destroy_listen(cm_id);
> > + break;
> > + case IW_CM_STATE_ESTABLISHED:
> > + cm_id_priv->state = IW_CM_STATE_DESTROYING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + /* Abrupt close of the connection */
> > + (void)iwcm_modify_qp_err(cm_id_priv->qp);
> > + break;
> > + case IW_CM_STATE_IDLE:
> > + case IW_CM_STATE_CLOSING:
> > + cm_id_priv->state = IW_CM_STATE_DESTROYING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + break;
> > + case IW_CM_STATE_CONN_RECV:
> > + /*
> > + * App called destroy before/without calling accept after
> > + * receiving connection request event notification.
> > + */
> > + cm_id_priv->state = IW_CM_STATE_DESTROYING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + break;
> > + case IW_CM_STATE_CONN_SENT:
> > + case IW_CM_STATE_DESTROYING:
> > + default:
> > + BUG_ON(1);
> > + break;
> > + }
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
>
> As an alternative, you could hold the lock from above, an let the LISTEN /
> ESTABLISHED state checks release and reacquire.
>
Yes, perhaps that's cleaner.
> > + if (cm_id_priv->qp) {
> > + cm_id_priv->id.device->iwcm->rem_ref(cm_id_priv->qp);
> > + cm_id_priv->qp = NULL;
> > + }
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +
> > + (void)iwcm_deref_id(cm_id_priv);
> > +}
> > +
> > +/*
> > + * This function is only called by the application thread and cannot
> > + * be called by the event thread. The function will wait for all
> > + * references to be released on the cm_id and then kfree the cm_id
> > + * object.
> > + */
> > +void iw_destroy_cm_id(struct iw_cm_id *cm_id)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + BUG_ON(test_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags));
> > +
> > + destroy_cm_id(cm_id);
> > +
> > + wait_event(cm_id_priv->destroy_wait,
> > + !atomic_read(&cm_id_priv->refcount));
> > +
> > + kfree(cm_id_priv);
> > +}
> > +EXPORT_SYMBOL(iw_destroy_cm_id);
> > +
> > +/*
> > + * CM_ID <-- LISTEN
> > + *
> > + * Start listening for connect requests. Generates one CONNECT_REQUEST
> > + * event for each inbound connect request.
> > + */
> > +int iw_cm_listen(struct iw_cm_id *cm_id, int backlog)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + unsigned long flags;
> > + int ret = 0;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_IDLE:
> > + cm_id_priv->state = IW_CM_STATE_LISTEN;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = cm_id->device->iwcm->create_listen(cm_id, backlog);
> > + if (ret)
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + break;
> > + default:
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = -EINVAL;
> > + }
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_listen);
> > +
> > +/*
> > + * CM_ID <-- IDLE
> > + *
> > + * Rejects an inbound connection request. No events are generated.
> > + */
> > +int iw_cm_reject(struct iw_cm_id *cm_id,
> > + const void *private_data,
> > + u8 private_data_len)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + unsigned long flags;
> > + int ret;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->state != IW_CM_STATE_CONN_RECV) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + return -EINVAL;
> > + }
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +
> > + ret = cm_id->device->iwcm->reject(cm_id, private_data,
> > + private_data_len);
> > +
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_reject);
> > +
> > +/*
> > + * CM_ID <-- ESTABLISHED
> > + *
> > + * Accepts an inbound connection request and generates an ESTABLISHED
> > + * event. Callers of iw_cm_disconnect and iw_destroy_cm_id will block
> > + * until the ESTABLISHED event is received from the provider.
> > + */
>
> This makes it sound like we're just waiting for an event.
>
disconnect/destory paths wait for the provider to complete the accept or
connect operation.
> > +int iw_cm_accept(struct iw_cm_id *cm_id,
> > + struct iw_cm_conn_param *iw_param)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + struct ib_qp *qp;
> > + unsigned long flags;
> > + int ret;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->state != IW_CM_STATE_CONN_RECV) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + return -EINVAL;
> > + }
> > + /* Get the ib_qp given the QPN */
> > + qp = cm_id->device->iwcm->get_qp(cm_id->device, iw_param->qpn);
> > + if (!qp) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + return -EINVAL;
> > + }
> > + cm_id->device->iwcm->add_ref(qp);
> > + cm_id_priv->qp = qp;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +
> > + ret = cm_id->device->iwcm->accept(cm_id, iw_param);
> > + if (ret) {
> > + /* An error on accept precludes provider events */
> > + BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_RECV);
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->qp) {
> > + cm_id->device->iwcm->rem_ref(qp);
> > + cm_id_priv->qp = NULL;
> > + }
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + printk("Accept failed, ret=%d\n", ret);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + }
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_accept);
> > +
> > +/*
> > + * Active Side: CM_ID <-- CONN_SENT
> > + *
> > + * If successful, results in the generation of a CONNECT_REPLY
> > + * event. iw_cm_disconnect and iw_cm_destroy will block until the
> > + * CONNECT_REPLY event is received from the provider.
> > + */
> > +int iw_cm_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + int ret = 0;
> > + unsigned long flags;
> > + struct ib_qp *qp;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->state != IW_CM_STATE_IDLE) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + return -EINVAL;
> > + }
> > +
> > + /* Get the ib_qp given the QPN */
> > + qp = cm_id->device->iwcm->get_qp(cm_id->device, iw_param->qpn);
> > + if (!qp) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + return -EINVAL;
> > + }
> > + cm_id->device->iwcm->add_ref(qp);
> > + cm_id_priv->qp = qp;
> > + cm_id_priv->state = IW_CM_STATE_CONN_SENT;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +
> > + ret = cm_id->device->iwcm->connect(cm_id, iw_param);
> > + if (ret) {
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->qp) {
> > + cm_id->device->iwcm->rem_ref(qp);
> > + cm_id_priv->qp = NULL;
> > + }
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_SENT);
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + printk("Connect failed, ret=%d\n", ret);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + }
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_connect);
> > +
> > +/*
> > + * Passive Side: new CM_ID <-- CONN_RECV
> > + *
> > + * Handles an inbound connect request. The function creates a new
> > + * iw_cm_id to represent the new connection and inherits the client
> > + * callback function and other attributes from the listening parent.
> > + *
> > + * The work item contains a pointer to the listen_cm_id and the event. The
> > + * listen_cm_id contains the client cm_handler, context and
> > + * device. These are copied when the device is cloned. The event
> > + * contains the new four tuple.
> > + *
> > + * An error on the child should not affect the parent, so this
> > + * function does not return a value.
> > + */
> > +static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > + struct iw_cm_id *cm_id;
> > + struct iwcm_id_private *cm_id_priv;
> > + int ret;
> > +
> > + /* The provider should never generate a connection request
> > + * event with a bad status.
> > + */
> > + BUG_ON(iw_event->status);
> > +
> > + /* We could be destroying the listening id. If so, ignore this
> > + * upcall. */
> > + spin_lock_irqsave(&listen_id_priv->lock, flags);
> > + if (listen_id_priv->state != IW_CM_STATE_LISTEN) {
> > + spin_unlock_irqrestore(&listen_id_priv->lock, flags);
> > + return;
> > + }
> > + spin_unlock_irqrestore(&listen_id_priv->lock, flags);
> > +
> > + cm_id = iw_create_cm_id(listen_id_priv->id.device,
> > + listen_id_priv->id.cm_handler,
> > + listen_id_priv->id.context);
> > + /* If the cm_id could not be created, ignore the request */
> > + if (IS_ERR(cm_id))
> > + return;
> > +
> > + cm_id->provider_data = iw_event->provider_data;
> > + cm_id->local_addr = iw_event->local_addr;
> > + cm_id->remote_addr = iw_event->remote_addr;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + cm_id_priv->state = IW_CM_STATE_CONN_RECV;
> > +
> > + /* Call the client CM handler */
> > + ret = cm_id->cm_handler(cm_id, iw_event);
> > + if (ret) {
> > + printk("destroying child id %p, ret=%d\n",
> > + cm_id, ret);
>
> We probably don't always want to print a message here.
>
Yes. I'll change this to a pr_debug().
> > + set_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags);
> > + destroy_cm_id(cm_id);
> > + if (atomic_read(&cm_id_priv->refcount)==0)
> > + kfree(cm_id);
> > + }
> > +}
> > +
> > +/*
> > + * Passive Side: CM_ID <-- ESTABLISHED
> > + *
> > + * The provider generated an ESTABLISHED event which means that
> > + * the MPA negotion has completed successfully and we are now in MPA
> > + * FPDU mode.
> > + *
> > + * This event can only be received in the CONN_RECV state. If the
> > + * remote peer closed, the ESTABLISHED event would be received followed
> > + * by the CLOSE event. If the app closes, it will block until we wake
> > + * it up after processing this event.
> > + */
> > +static int cm_conn_est_handler(struct iwcm_id_private *cm_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > + int ret = 0;
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > +
> > + /* We clear the CONNECT_WAIT bit here to allow the callback
> > + * function to call iw_cm_disconnect. Calling iw_destroy_cm_id
> > + * from a callback handler is not allowed */
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_CONN_RECV:
> > + cm_id_priv->state = IW_CM_STATE_ESTABLISHED;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = cm_id_priv->id.cm_handler(&cm_id_priv->id, iw_event);
> > + break;
> > + default:
> > + BUG_ON(1);
>
> Can just BUG_ON the state and avoid the switch. Same comment applies below.
>
ok.
> > + }
> > + wake_up_all(&cm_id_priv->connect_wait);
> > +
> > + return ret;
> > +}
> > +
> > +/*
> > + * Active Side: CM_ID <-- ESTABLISHED
> > + *
> > + * The app has called connect and is waiting for the established event to
> > + * post it's requests to the server. This event will wake up anyone
> > + * blocked in iw_cm_disconnect or iw_destroy_id.
> > + */
> > +static int cm_conn_rep_handler(struct iwcm_id_private *cm_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > + int ret = 0;
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + /* Clear the connect wait bit so a callback function calling
> > + * iw_cm_disconnect will not wait and deadlock this thread */
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_CONN_SENT:
> > + if (iw_event->status == IW_CM_EVENT_STATUS_ACCEPTED) {
> > + cm_id_priv->id.local_addr = iw_event->local_addr;
> > + cm_id_priv->id.remote_addr = iw_event->remote_addr;
> > + cm_id_priv->state = IW_CM_STATE_ESTABLISHED;
> > + } else {
> > + /* REJECTED or RESET */
> > + cm_id_priv->id.device->iwcm->rem_ref(cm_id_priv->qp);
> > + cm_id_priv->qp = NULL;
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + }
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = cm_id_priv->id.cm_handler(&cm_id_priv->id, iw_event);
> > + break;
> > + default:
> > + BUG_ON(1);
> > + }
> > + /* Wake up waiters on connect complete */
> > + wake_up_all(&cm_id_priv->connect_wait);
> > +
> > + return ret;
> > +}
> > +
> > +/*
> > + * CM_ID <-- CLOSING
> > + *
> > + * If in the ESTABLISHED state, move to CLOSING.
> > + */
> > +static void cm_disconnect_handler(struct iwcm_id_private *cm_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->state == IW_CM_STATE_ESTABLISHED)
> > + cm_id_priv->state = IW_CM_STATE_CLOSING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +}
> > +
> > +/*
> > + * CM_ID <-- IDLE
> > + *
> > + * If in the ESTBLISHED or CLOSING states, the QP will have have been
> > + * moved by the provider to the ERR state. Disassociate the CM_ID from
> > + * the QP, move to IDLE, and remove the 'connected' reference.
> > + *
> > + * If in some other state, the cm_id was destroyed asynchronously.
> > + * This is the last reference that will result in waking up
> > + * the app thread blocked in iw_destroy_cm_id.
> > + */
> > +static int cm_close_handler(struct iwcm_id_private *cm_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > + int ret = 0;
> > + /* TT */printk("%s:%d cm_id_priv=%p, state=%d\n",
> > + __FUNCTION__, __LINE__,
> > + cm_id_priv,cm_id_priv->state);
>
> Will want to remove this.
>
oops. yes...
> - Sean
^ permalink raw reply
* Re: [openib-general] Re: [PATCH 1/2] iWARP Connection Manager.
From: Tom Tucker @ 2006-06-01 17:11 UTC (permalink / raw)
To: Sean Hefty; +Cc: Steve Wise, netdev, rdreier, linux-kernel, openib-general
In-Reply-To: <447E1720.7000307@ichips.intel.com>
On Wed, 2006-05-31 at 15:22 -0700, Sean Hefty wrote:
> Steve Wise wrote:
> > +/*
> > + * Release a reference on cm_id. If the last reference is being removed
> > + * and iw_destroy_cm_id is waiting, wake up the waiting thread.
> > + */
> > +static int iwcm_deref_id(struct iwcm_id_private *cm_id_priv)
> > +{
> > + int ret = 0;
> > +
> > + BUG_ON(atomic_read(&cm_id_priv->refcount)==0);
> > + if (atomic_dec_and_test(&cm_id_priv->refcount)) {
> > + BUG_ON(!list_empty(&cm_id_priv->work_list));
> > + if (waitqueue_active(&cm_id_priv->destroy_wait)) {
> > + BUG_ON(cm_id_priv->state != IW_CM_STATE_DESTROYING);
> > + BUG_ON(test_bit(IWCM_F_CALLBACK_DESTROY,
> > + &cm_id_priv->flags));
> > + ret = 1;
> > + wake_up(&cm_id_priv->destroy_wait);
>
> We recently changed the RDMA CM, IB CM, and a couple of other modules from using
> wait objects to completions. This avoids a race condition between decrementing
> the reference count, which allows destruction to proceed, and calling wake_up on
> a freed cm_id. My guess is that you may need to do the same.
>
> Can you also explain the use of the return value here? It's ignored below in
> rem_ref() and destroy_cm_id().
>
> > +static void add_ref(struct iw_cm_id *cm_id)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + atomic_inc(&cm_id_priv->refcount);
> > +}
> > +
> > +static void rem_ref(struct iw_cm_id *cm_id)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + iwcm_deref_id(cm_id_priv);
> > +}
> > +
>
> > +/*
> > + * CM_ID <-- CLOSING
> > + *
> > + * Block if a passive or active connection is currenlty being processed. Then
> > + * process the event as follows:
> > + * - If we are ESTABLISHED, move to CLOSING and modify the QP state
> > + * based on the abrupt flag
> > + * - If the connection is already in the CLOSING or IDLE state, the peer is
> > + * disconnecting concurrently with us and we've already seen the
> > + * DISCONNECT event -- ignore the request and return 0
> > + * - Disconnect on a listening endpoint returns -EINVAL
> > + */
> > +int iw_cm_disconnect(struct iw_cm_id *cm_id, int abrupt)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + unsigned long flags;
> > + int ret = 0;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + /* Wait if we're currently in a connect or accept downcall */
> > + wait_event(cm_id_priv->connect_wait,
> > + !test_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags));
>
> Am I understanding this check correctly? You're checking to see if the user has
> called iw_cm_disconnect() at the same time that they called iw_cm_connect() or
> iw_cm_accept(). Are connect / accept blocking, or are you just waiting for an
> event?
Yes. The application (or the case I saw was user-mode exit logic after
ctrl-C) cleaning up at random times relative to connection
establishment.
>
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_ESTABLISHED:
> > + cm_id_priv->state = IW_CM_STATE_CLOSING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->qp) { /* QP could be <nul> for user-mode client */
> > + if (abrupt)
> > + ret = iwcm_modify_qp_err(cm_id_priv->qp);
> > + else
> > + ret = iwcm_modify_qp_sqd(cm_id_priv->qp);
> > + /*
> > + * If both sides are disconnecting the QP could
> > + * already be in ERR or SQD states
> > + */
> > + ret = 0;
> > + }
> > + else
> > + ret = -EINVAL;
> > + break;
> > + case IW_CM_STATE_LISTEN:
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = -EINVAL;
> > + break;
> > + case IW_CM_STATE_CLOSING:
> > + /* remote peer closed first */
> > + case IW_CM_STATE_IDLE:
> > + /* accept or connect returned !0 */
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + break;
> > + case IW_CM_STATE_CONN_RECV:
> > + /*
> > + * App called disconnect before/without calling accept after
> > + * connect_request event delivered.
> > + */
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + break;
> > + case IW_CM_STATE_CONN_SENT:
> > + /* Can only get here if wait above fails */
> > + default:
> > + BUG_ON(1);
> > + }
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_disconnect);
> > +static void destroy_cm_id(struct iw_cm_id *cm_id)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + unsigned long flags;
> > + int ret;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + /* Wait if we're currently in a connect or accept downcall. A
> > + * listening endpoint should never block here. */
> > + wait_event(cm_id_priv->connect_wait,
> > + !test_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags));
>
> Same question/comment as above.
>
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_LISTEN:
> > + cm_id_priv->state = IW_CM_STATE_DESTROYING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + /* destroy the listening endpoint */
> > + ret = cm_id->device->iwcm->destroy_listen(cm_id);
> > + break;
> > + case IW_CM_STATE_ESTABLISHED:
> > + cm_id_priv->state = IW_CM_STATE_DESTROYING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + /* Abrupt close of the connection */
> > + (void)iwcm_modify_qp_err(cm_id_priv->qp);
> > + break;
> > + case IW_CM_STATE_IDLE:
> > + case IW_CM_STATE_CLOSING:
> > + cm_id_priv->state = IW_CM_STATE_DESTROYING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + break;
> > + case IW_CM_STATE_CONN_RECV:
> > + /*
> > + * App called destroy before/without calling accept after
> > + * receiving connection request event notification.
> > + */
> > + cm_id_priv->state = IW_CM_STATE_DESTROYING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + break;
> > + case IW_CM_STATE_CONN_SENT:
> > + case IW_CM_STATE_DESTROYING:
> > + default:
> > + BUG_ON(1);
> > + break;
> > + }
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
>
> As an alternative, you could hold the lock from above, an let the LISTEN /
> ESTABLISHED state checks release and reacquire.
>
> > + if (cm_id_priv->qp) {
> > + cm_id_priv->id.device->iwcm->rem_ref(cm_id_priv->qp);
> > + cm_id_priv->qp = NULL;
> > + }
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +
> > + (void)iwcm_deref_id(cm_id_priv);
> > +}
> > +
> > +/*
> > + * This function is only called by the application thread and cannot
> > + * be called by the event thread. The function will wait for all
> > + * references to be released on the cm_id and then kfree the cm_id
> > + * object.
> > + */
> > +void iw_destroy_cm_id(struct iw_cm_id *cm_id)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + BUG_ON(test_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags));
> > +
> > + destroy_cm_id(cm_id);
> > +
> > + wait_event(cm_id_priv->destroy_wait,
> > + !atomic_read(&cm_id_priv->refcount));
> > +
> > + kfree(cm_id_priv);
> > +}
> > +EXPORT_SYMBOL(iw_destroy_cm_id);
> > +
> > +/*
> > + * CM_ID <-- LISTEN
> > + *
> > + * Start listening for connect requests. Generates one CONNECT_REQUEST
> > + * event for each inbound connect request.
> > + */
> > +int iw_cm_listen(struct iw_cm_id *cm_id, int backlog)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + unsigned long flags;
> > + int ret = 0;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_IDLE:
> > + cm_id_priv->state = IW_CM_STATE_LISTEN;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = cm_id->device->iwcm->create_listen(cm_id, backlog);
> > + if (ret)
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + break;
> > + default:
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = -EINVAL;
> > + }
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_listen);
> > +
> > +/*
> > + * CM_ID <-- IDLE
> > + *
> > + * Rejects an inbound connection request. No events are generated.
> > + */
> > +int iw_cm_reject(struct iw_cm_id *cm_id,
> > + const void *private_data,
> > + u8 private_data_len)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + unsigned long flags;
> > + int ret;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->state != IW_CM_STATE_CONN_RECV) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + return -EINVAL;
> > + }
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +
> > + ret = cm_id->device->iwcm->reject(cm_id, private_data,
> > + private_data_len);
> > +
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_reject);
> > +
> > +/*
> > + * CM_ID <-- ESTABLISHED
> > + *
> > + * Accepts an inbound connection request and generates an ESTABLISHED
> > + * event. Callers of iw_cm_disconnect and iw_destroy_cm_id will block
> > + * until the ESTABLISHED event is received from the provider.
> > + */
>
> This makes it sound like we're just waiting for an event.
>
> > +int iw_cm_accept(struct iw_cm_id *cm_id,
> > + struct iw_cm_conn_param *iw_param)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + struct ib_qp *qp;
> > + unsigned long flags;
> > + int ret;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->state != IW_CM_STATE_CONN_RECV) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + return -EINVAL;
> > + }
> > + /* Get the ib_qp given the QPN */
> > + qp = cm_id->device->iwcm->get_qp(cm_id->device, iw_param->qpn);
> > + if (!qp) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + return -EINVAL;
> > + }
> > + cm_id->device->iwcm->add_ref(qp);
> > + cm_id_priv->qp = qp;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +
> > + ret = cm_id->device->iwcm->accept(cm_id, iw_param);
> > + if (ret) {
> > + /* An error on accept precludes provider events */
> > + BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_RECV);
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->qp) {
> > + cm_id->device->iwcm->rem_ref(qp);
> > + cm_id_priv->qp = NULL;
> > + }
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + printk("Accept failed, ret=%d\n", ret);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + }
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_accept);
> > +
> > +/*
> > + * Active Side: CM_ID <-- CONN_SENT
> > + *
> > + * If successful, results in the generation of a CONNECT_REPLY
> > + * event. iw_cm_disconnect and iw_cm_destroy will block until the
> > + * CONNECT_REPLY event is received from the provider.
> > + */
> > +int iw_cm_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param)
> > +{
> > + struct iwcm_id_private *cm_id_priv;
> > + int ret = 0;
> > + unsigned long flags;
> > + struct ib_qp *qp;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->state != IW_CM_STATE_IDLE) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + return -EINVAL;
> > + }
> > +
> > + /* Get the ib_qp given the QPN */
> > + qp = cm_id->device->iwcm->get_qp(cm_id->device, iw_param->qpn);
> > + if (!qp) {
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + return -EINVAL;
> > + }
> > + cm_id->device->iwcm->add_ref(qp);
> > + cm_id_priv->qp = qp;
> > + cm_id_priv->state = IW_CM_STATE_CONN_SENT;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +
> > + ret = cm_id->device->iwcm->connect(cm_id, iw_param);
> > + if (ret) {
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->qp) {
> > + cm_id->device->iwcm->rem_ref(qp);
> > + cm_id_priv->qp = NULL;
> > + }
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_SENT);
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + printk("Connect failed, ret=%d\n", ret);
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + wake_up_all(&cm_id_priv->connect_wait);
> > + }
> > +
> > + return ret;
> > +}
> > +EXPORT_SYMBOL(iw_cm_connect);
> > +
> > +/*
> > + * Passive Side: new CM_ID <-- CONN_RECV
> > + *
> > + * Handles an inbound connect request. The function creates a new
> > + * iw_cm_id to represent the new connection and inherits the client
> > + * callback function and other attributes from the listening parent.
> > + *
> > + * The work item contains a pointer to the listen_cm_id and the event. The
> > + * listen_cm_id contains the client cm_handler, context and
> > + * device. These are copied when the device is cloned. The event
> > + * contains the new four tuple.
> > + *
> > + * An error on the child should not affect the parent, so this
> > + * function does not return a value.
> > + */
> > +static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > + struct iw_cm_id *cm_id;
> > + struct iwcm_id_private *cm_id_priv;
> > + int ret;
> > +
> > + /* The provider should never generate a connection request
> > + * event with a bad status.
> > + */
> > + BUG_ON(iw_event->status);
> > +
> > + /* We could be destroying the listening id. If so, ignore this
> > + * upcall. */
> > + spin_lock_irqsave(&listen_id_priv->lock, flags);
> > + if (listen_id_priv->state != IW_CM_STATE_LISTEN) {
> > + spin_unlock_irqrestore(&listen_id_priv->lock, flags);
> > + return;
> > + }
> > + spin_unlock_irqrestore(&listen_id_priv->lock, flags);
> > +
> > + cm_id = iw_create_cm_id(listen_id_priv->id.device,
> > + listen_id_priv->id.cm_handler,
> > + listen_id_priv->id.context);
> > + /* If the cm_id could not be created, ignore the request */
> > + if (IS_ERR(cm_id))
> > + return;
> > +
> > + cm_id->provider_data = iw_event->provider_data;
> > + cm_id->local_addr = iw_event->local_addr;
> > + cm_id->remote_addr = iw_event->remote_addr;
> > +
> > + cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
> > + cm_id_priv->state = IW_CM_STATE_CONN_RECV;
> > +
> > + /* Call the client CM handler */
> > + ret = cm_id->cm_handler(cm_id, iw_event);
> > + if (ret) {
> > + printk("destroying child id %p, ret=%d\n",
> > + cm_id, ret);
>
> We probably don't always want to print a message here.
>
> > + set_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags);
> > + destroy_cm_id(cm_id);
> > + if (atomic_read(&cm_id_priv->refcount)==0)
> > + kfree(cm_id);
> > + }
> > +}
> > +
> > +/*
> > + * Passive Side: CM_ID <-- ESTABLISHED
> > + *
> > + * The provider generated an ESTABLISHED event which means that
> > + * the MPA negotion has completed successfully and we are now in MPA
> > + * FPDU mode.
> > + *
> > + * This event can only be received in the CONN_RECV state. If the
> > + * remote peer closed, the ESTABLISHED event would be received followed
> > + * by the CLOSE event. If the app closes, it will block until we wake
> > + * it up after processing this event.
> > + */
> > +static int cm_conn_est_handler(struct iwcm_id_private *cm_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > + int ret = 0;
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > +
> > + /* We clear the CONNECT_WAIT bit here to allow the callback
> > + * function to call iw_cm_disconnect. Calling iw_destroy_cm_id
> > + * from a callback handler is not allowed */
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_CONN_RECV:
> > + cm_id_priv->state = IW_CM_STATE_ESTABLISHED;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = cm_id_priv->id.cm_handler(&cm_id_priv->id, iw_event);
> > + break;
> > + default:
> > + BUG_ON(1);
>
> Can just BUG_ON the state and avoid the switch. Same comment applies below.
>
> > + }
> > + wake_up_all(&cm_id_priv->connect_wait);
> > +
> > + return ret;
> > +}
> > +
> > +/*
> > + * Active Side: CM_ID <-- ESTABLISHED
> > + *
> > + * The app has called connect and is waiting for the established event to
> > + * post it's requests to the server. This event will wake up anyone
> > + * blocked in iw_cm_disconnect or iw_destroy_id.
> > + */
> > +static int cm_conn_rep_handler(struct iwcm_id_private *cm_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > + int ret = 0;
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + /* Clear the connect wait bit so a callback function calling
> > + * iw_cm_disconnect will not wait and deadlock this thread */
> > + clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
> > + switch (cm_id_priv->state) {
> > + case IW_CM_STATE_CONN_SENT:
> > + if (iw_event->status == IW_CM_EVENT_STATUS_ACCEPTED) {
> > + cm_id_priv->id.local_addr = iw_event->local_addr;
> > + cm_id_priv->id.remote_addr = iw_event->remote_addr;
> > + cm_id_priv->state = IW_CM_STATE_ESTABLISHED;
> > + } else {
> > + /* REJECTED or RESET */
> > + cm_id_priv->id.device->iwcm->rem_ref(cm_id_priv->qp);
> > + cm_id_priv->qp = NULL;
> > + cm_id_priv->state = IW_CM_STATE_IDLE;
> > + }
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > + ret = cm_id_priv->id.cm_handler(&cm_id_priv->id, iw_event);
> > + break;
> > + default:
> > + BUG_ON(1);
> > + }
> > + /* Wake up waiters on connect complete */
> > + wake_up_all(&cm_id_priv->connect_wait);
> > +
> > + return ret;
> > +}
> > +
> > +/*
> > + * CM_ID <-- CLOSING
> > + *
> > + * If in the ESTABLISHED state, move to CLOSING.
> > + */
> > +static void cm_disconnect_handler(struct iwcm_id_private *cm_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > +
> > + spin_lock_irqsave(&cm_id_priv->lock, flags);
> > + if (cm_id_priv->state == IW_CM_STATE_ESTABLISHED)
> > + cm_id_priv->state = IW_CM_STATE_CLOSING;
> > + spin_unlock_irqrestore(&cm_id_priv->lock, flags);
> > +}
> > +
> > +/*
> > + * CM_ID <-- IDLE
> > + *
> > + * If in the ESTBLISHED or CLOSING states, the QP will have have been
> > + * moved by the provider to the ERR state. Disassociate the CM_ID from
> > + * the QP, move to IDLE, and remove the 'connected' reference.
> > + *
> > + * If in some other state, the cm_id was destroyed asynchronously.
> > + * This is the last reference that will result in waking up
> > + * the app thread blocked in iw_destroy_cm_id.
> > + */
> > +static int cm_close_handler(struct iwcm_id_private *cm_id_priv,
> > + struct iw_cm_event *iw_event)
> > +{
> > + unsigned long flags;
> > + int ret = 0;
> > + /* TT */printk("%s:%d cm_id_priv=%p, state=%d\n",
> > + __FUNCTION__, __LINE__,
> > + cm_id_priv,cm_id_priv->state);
>
> Will want to remove this.
>
> - Sean
> _______________________________________________
> openib-general mailing list
> openib-general@openib.org
> http://openib.org/mailman/listinfo/openib-general
>
> To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
^ permalink raw reply
* Re: 2.6.17-rc4: netfilter LOG messages truncated via NETCONSOLE
From: Patrick McHardy @ 2006-06-01 17:34 UTC (permalink / raw)
To: Frank van Maarseveen; +Cc: linux-kernel, Kernel Netdev Mailing List
In-Reply-To: <20060601091124.GA31642@janus>
Frank van Maarseveen wrote:
> ok, now "tc -s -d qdisc show" says (after noticing missing netconsole
> packets):
>
> qdisc pfifo_fast 0: dev eth0 bands 3 priomap 1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
> Sent 155031 bytes 2067 pkt (dropped 0, overlimits 0 requeues 0)
> backlog 0b 0p requeues 0
Mhh no dropped packets. I tried to reproduce the problem by changing
netconsole to always use the dev_queue_xmit path, but works flawlessly
for me. Please try to find out if the packets are lost before or after
the qdisc by looking at the packet counter.
BTW: You still haven't sent me the packet dump (from the originating
machine).
^ 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