Import 2.1.120pre1

CREDITS

@@ -1233,12 +1233,12 @@ S: 7546 JA  Enschede
 S: Netherlands
 
 N: David S. Miller
-E: davem@caip.rutgers.edu
-D: Sparc hacker
-D: New Linux-Activists maintainer
+E: davem@dm.cobaltmicro.com
+D: Sparc and blue box hacker
+D: Vger Linux mailing list co-maintainer
 D: Linux Emacs elf/qmagic support + other libc/gcc things
 D: Yee bore de yee bore! ;-)
-S: 111 Alta Tierra Court
+S: 331 Santa Rosa Drive
 S: Los Gatos, California 95032
 S: USA
 

Makefile

 VERSION = 2
 PATCHLEVEL = 1
-SUBLEVEL = 119
+SUBLEVEL = 120
 
 ARCH := $(shell uname -m | sed -e s/i.86/i386/ -e s/sun4u/sparc64/ -e s/arm.*/arm/ -e s/sa110/arm/)
 

drivers/net/mace.c

@@ -91,85 +91,90 @@ bitrev(int b)
 int
 mace_probe(struct device *dev)
 {
-    int j, rev;
-    struct mace_data *mp;
-    struct device_node *maces;
-    unsigned char *addr;
-
-    maces = find_devices("mace");
-    if (maces == 0)
-	return ENODEV;
-
-    do {
-	if (maces->n_addrs != 3 || maces->n_intrs != 3) {
-	    printk(KERN_ERR "can't use MACE %s: expect 3 addrs and 3 intrs\n",
-		   maces->full_name);
-	    continue;
+	int j, rev;
+	struct mace_data *mp;
+	struct device_node *mace;
+	unsigned char *addr;
+	static int maces_found = 0;
+	static struct device_node *next_mace;
+
+	if (!maces_found) {
+		next_mace = find_devices("mace");
+		maces_found = 1;
+	}
+	mace = next_mace;
+	if (mace == 0)
+		return -ENODEV;
+	next_mace = mace->next;
+
+	if (mace->n_addrs != 3 || mace->n_intrs != 3) {
+		printk(KERN_ERR "can't use MACE %s: expect 3 addrs and 3 intrs\n",
+		       mace->full_name);
+		return -ENODEV;
 	}
 
 	if (dev == NULL)
-	    dev = init_etherdev(0, PRIV_BYTES);
+		dev = init_etherdev(0, PRIV_BYTES);
 	else {
-	    /* XXX this doesn't look right (but it's never used :-) */
-	    dev->priv = kmalloc(PRIV_BYTES, GFP_KERNEL);
-	    if (dev->priv == 0)
-		return -ENOMEM;
+		dev->priv = kmalloc(PRIV_BYTES, GFP_KERNEL);
+		if (dev->priv == 0)
+			return -ENOMEM;
 	}
 
 	mp = (struct mace_data *) dev->priv;
-	dev->base_addr = maces->addrs[0].address;
+	dev->base_addr = mace->addrs[0].address;
 	mp->mace = (volatile struct mace *)
-		ioremap(maces->addrs[0].address, 0x1000);
-	dev->irq = maces->intrs[0].line;
+		ioremap(mace->addrs[0].address, 0x1000);
+	dev->irq = mace->intrs[0].line;
 
 	if (request_irq(dev->irq, mace_interrupt, 0, "MACE", dev)) {
-	    printk(KERN_ERR "MACE: can't get irq %d\n", dev->irq);
-	    return -EAGAIN;
+		printk(KERN_ERR "MACE: can't get irq %d\n", dev->irq);
+		return -EAGAIN;
 	}
-	if (request_irq(maces->intrs[1].line, mace_txdma_intr, 0, "MACE-txdma",
+	if (request_irq(mace->intrs[1].line, mace_txdma_intr, 0, "MACE-txdma",
 			dev)) {
-	    printk(KERN_ERR "MACE: can't get irq %d\n", maces->intrs[1].line);
-	    return -EAGAIN;
+		printk(KERN_ERR "MACE: can't get irq %d\n", mace->intrs[1].line);
+		return -EAGAIN;
 	}
-	if (request_irq(maces->intrs[2].line, mace_rxdma_intr, 0, "MACE-rxdma",
+	if (request_irq(mace->intrs[2].line, mace_rxdma_intr, 0, "MACE-rxdma",
 			dev)) {
-	    printk(KERN_ERR "MACE: can't get irq %d\n", maces->intrs[2].line);
-	    return -EAGAIN;
+		printk(KERN_ERR "MACE: can't get irq %d\n", mace->intrs[2].line);
+		return -EAGAIN;
 	}
 
-	addr = get_property(maces, "mac-address", NULL);
+	addr = get_property(mace, "mac-address", NULL);
 	if (addr == NULL) {
-	    addr = get_property(maces, "local-mac-address", NULL);
-	    if (addr == NULL) {
-		printk(KERN_ERR "Can't get mac-address for MACE at %lx\n",
-		       dev->base_addr);
-		return -EAGAIN;
-	    }
+		addr = get_property(mace, "local-mac-address", NULL);
+		if (addr == NULL) {
+			printk(KERN_ERR "Can't get mac-address for MACE at %lx\n",
+			       dev->base_addr);
+			return -EAGAIN;
+		}
 	}
 
 	printk(KERN_INFO "%s: MACE at", dev->name);
 	rev = addr[0] == 0 && addr[1] == 0xA0;
 	for (j = 0; j < 6; ++j) {
-	    dev->dev_addr[j] = rev? bitrev(addr[j]): addr[j];
-	    printk("%c%.2x", (j? ':': ' '), dev->dev_addr[j]);
+		dev->dev_addr[j] = rev? bitrev(addr[j]): addr[j];
+		printk("%c%.2x", (j? ':': ' '), dev->dev_addr[j]);
 	}
 	printk("\n");
 
 	mp = (struct mace_data *) dev->priv;
 	mp->maccc = ENXMT | ENRCV;
 	mp->tx_dma = (volatile struct dbdma_regs *)
-		ioremap(maces->addrs[1].address, 0x1000);
-	mp->tx_dma_intr = maces->intrs[1].line;
+		ioremap(mace->addrs[1].address, 0x1000);
+	mp->tx_dma_intr = mace->intrs[1].line;
 	mp->rx_dma = (volatile struct dbdma_regs *)
-		ioremap(maces->addrs[2].address, 0x1000);
-	mp->rx_dma_intr = maces->intrs[2].line;
+		ioremap(mace->addrs[2].address, 0x1000);
+	mp->rx_dma_intr = mace->intrs[2].line;
 
 	mp->tx_cmds = (volatile struct dbdma_cmd *) DBDMA_ALIGN(mp + 1);
 	mp->rx_cmds = mp->tx_cmds + NCMDS_TX * N_TX_RING + 1;
 
 	memset(&mp->stats, 0, sizeof(mp->stats));
 	memset((char *) mp->tx_cmds, 0,
-	      (NCMDS_TX*N_TX_RING + N_RX_RING + 2) * sizeof(struct dbdma_cmd));
+	       (NCMDS_TX*N_TX_RING + N_RX_RING + 2) * sizeof(struct dbdma_cmd));
 	init_timer(&mp->tx_timeout);
 	mp->timeout_active = 0;
 
@@ -182,9 +187,7 @@ mace_probe(struct device *dev)
 
 	ether_setup(dev);
 
-    } while ((maces = maces->next) != 0);
-
-    return 0;
+	return 0;
 }
 
 static void mace_reset(struct device *dev)

drivers/net/plip.c

@@ -244,8 +244,8 @@ plip_init_dev(struct device *dev, struct parport *pb))
 					 plip_wakeup, plip_interrupt, 
 					 PARPORT_DEV_LURK, dev);
 
-	printk(version);
-	printk("%s: Parallel port at %#3lx, using IRQ %d\n", dev->name,
+	printk(KERN_INFO "%s", version);
+	printk(KERN_INFO "%s: Parallel port at %#3lx, using IRQ %d\n", dev->name,
 	       dev->base_addr, dev->irq);
 
 	/* Fill in the generic fields of the device structure. */
@@ -537,7 +537,7 @@ plip_receive_packet(struct device *dev, struct net_local *nl,
 		/* Malloc up new buffer. */
 		rcv->skb = dev_alloc_skb(rcv->length.h);
 		if (rcv->skb == NULL) {
-			printk(KERN_WARNING "%s: Memory squeeze.\n", dev->name);
+			printk(KERN_ERR "%s: Memory squeeze.\n", dev->name);
 			return ERROR;
 		}
 		skb_put(rcv->skb,rcv->length.h);
@@ -662,7 +662,7 @@ plip_send_packet(struct device *dev, struct net_local *nl,
 	unsigned int cx;
 
 	if (snd->skb == NULL || (lbuf = snd->skb->data) == NULL) {
-		printk(KERN_ERR "%s: send skb lost\n", dev->name);
+		printk(KERN_DEBUG "%s: send skb lost\n", dev->name);
 		snd->state = PLIP_PK_DONE;
 		snd->skb = NULL;
 		return ERROR;
@@ -817,7 +817,7 @@ plip_interrupt(int irq, void *dev_id, struct pt_regs * regs)
 	unsigned char c0;
 
 	if (dev == NULL) {
-		printk(KERN_ERR "plip_interrupt: irq %d for unknown device.\n", irq);
+		printk(KERN_DEBUG "plip_interrupt: irq %d for unknown device.\n", irq);
 		return;
 	}
 
@@ -861,7 +861,7 @@ plip_interrupt(int irq, void *dev_id, struct pt_regs * regs)
 
 	case PLIP_CN_ERROR:
 		spin_unlock_irq(&nl->lock);
-		printk(KERN_WARNING "%s: receive interrupt in error state\n", dev->name);
+		printk(KERN_ERR "%s: receive interrupt in error state\n", dev->name);
 		break;
 	}
 }
@@ -1083,7 +1083,8 @@ plip_get_stats(struct device *dev)
 	return r;
 }
 
-static int plip_config(struct device *dev, struct ifmap *map)
+static int
+plip_config(struct device *dev, struct ifmap *map)
 {
 	struct net_local *nl = (struct net_local *) dev->priv;
 	struct pardevice *pardev = nl->pardev;
@@ -1091,8 +1092,8 @@ static int plip_config(struct device *dev, struct ifmap *map)
 	if (dev->flags & IFF_UP)
 		return -EBUSY;
 
-	printk(KERN_INFO "plip: Warning, changing irq with ifconfig will be obsoleted.\n");
-	printk(KERN_INFO "plip: Next time, please set with /proc/parport/*/irq instead.\n");
+	printk(KERN_WARNING "plip: Warning, changing irq with ifconfig will be obsoleted.\n");
+	printk(KERN_WARNING "plip: Next time, please set with /proc/parport/*/irq instead.\n");
 
 	if (map->irq != (unsigned char)-1) {
 		pardev->port->irq = dev->irq = map->irq;
@@ -1177,7 +1178,7 @@ void plip_setup(char *str, int *ints)
 			/* disable driver on "parport=" or "parport=0" */
 			parport[0] = -2;
 		} else {
-			printk(KERN_WARNING "warning: 'plip=0x%x' ignored\n", 
+			printk(KERN_WARNINGING "warning: 'plip=0x%x' ignored\n", 
 			       ints[1]);
 		}
 	}

drivers/net/sunlance.c

-/* $Id: sunlance.c,v 1.79 1998/06/04 09:54:58 jj Exp $
+/* $Id: sunlance.c,v 1.81 1998/08/10 09:08:23 jj Exp $
  * lance.c: Linux/Sparc/Lance driver
  *
  *	Written 1995, 1996 by Miguel de Icaza
@@ -1135,9 +1135,10 @@ __initfunc(int sparc_lance_probe (struct device *dev))
 		return ENODEV;
 	called++;
 
-	if (idprom->id_machtype == (SM_SUN4|SM_4_330)) {
+	if ((idprom->id_machtype == (SM_SUN4|SM_4_330)) ||
+	    (idprom->id_machtype == (SM_SUN4|SM_4_470))) {
 		memset (&sdev, 0, sizeof(sdev));
-		sdev.reg_addrs[0].phys_addr = SUN4_300_ETH_PHYSADDR;
+		sdev.reg_addrs[0].phys_addr = sun4_eth_physaddr;
 		sdev.irqs[0] = 6;
 		return sparc_lance_init(dev, &sdev, 0, 0);
 	}

drivers/scsi/README.in2000

 
+UPDATE NEWS: version 1.33 - 26 Aug 98
+
+   Interrupt management in this driver has become, over
+   time, increasingly odd and difficult to explain - this
+   has been mostly due to my own mental inadequacies. In
+   recent kernels, it has failed to function at all when
+   compiled for SMP. I've fixed that problem, and after
+   taking a fresh look at interrupts in general, greatly
+   reduced the number of places where they're fiddled
+   with. Done some heavy testing and it looks very good.
+   The driver now makes use of the __initfunc() and
+   __initdata macros to save about 4k of kernel memory.
+   Once again, the same code works for both 2.0.xx and
+   2.1.xx kernels.
+
 UPDATE NEWS: version 1.32 - 28 Mar 98
 
    Removed the check for legal IN2000 hardware versions:

drivers/scsi/in2000.h

@@ -2,7 +2,7 @@
  *    in2000.h -  Linux device driver definitions for the
  *                Always IN2000 ISA SCSI card.
  *
- *    IMPORTANT: This file is for version 1.32 - 28/Mar/1998
+ *    IMPORTANT: This file is for version 1.33 - 26/Aug/1998
  *
  * Copyright (c) 1996 John Shifflett, GeoLog Consulting
  *    john@geolog.com
@@ -377,10 +377,29 @@ struct IN2000_hostdata {
 #define PR_STOP      1<<7
 
 
-int in2000_detect(Scsi_Host_Template *);
+#include <linux/version.h>
+
+#if LINUX_VERSION_CODE < 0x020100   /* 2.0.xx */
+# define in2000__INITFUNC(function) function
+# define in2000__INIT
+# define in2000__INITDATA
+# define CLISPIN_LOCK(flags)   do { save_flags(flags); cli(); } while(0)
+# define CLISPIN_UNLOCK(flags) restore_flags(flags)
+#else                               /* 2.1.xxx */
+# include <linux/init.h>
+# include <asm/spinlock.h>
+# define in2000__INITFUNC(function) __initfunc(function)
+# define in2000__INIT __init
+# define in2000__INITDATA __initdata
+# define CLISPIN_LOCK(flags)   spin_lock_irqsave(&io_request_lock, flags)
+# define CLISPIN_UNLOCK(flags) spin_unlock_irqrestore(&io_request_lock, flags)
+#endif
+
+
+int in2000_detect(Scsi_Host_Template *) in2000__INIT;
 int in2000_queuecommand(Scsi_Cmnd *, void (*done)(Scsi_Cmnd *));
 int in2000_abort(Scsi_Cmnd *);
-void in2000_setup(char *, int *);
+void in2000_setup(char *, int *) in2000__INIT;
 int in2000_proc_info(char *, char **, off_t, int, int, int);
 struct proc_dir_entry proc_scsi_in2000;
 int in2000_biosparam(struct scsi_disk *, kdev_t, int *);
@@ -392,6 +411,33 @@ int in2000_reset(Scsi_Cmnd *, unsigned int);
 #define IN2000_CPL      2
 #define IN2000_HOST_ID  7
 
+#if LINUX_VERSION_CODE < 0x020100   /* 2.0.xx */
+
+#define IN2000 {  NULL,                /* link pointer for modules */ \
+                  NULL,                /* usage_count for modules */ \
+                  &proc_scsi_in2000,   /* pointer to /proc/scsi directory entry */ \
+                  in2000_proc_info,    /* pointer to proc info function */ \
+                  "Always IN2000",     /* device name */ \
+                  in2000_detect,       /* returns number of in2000's found */ \
+                  NULL,                /* optional unload function for modules */ \
+                  NULL,                /* optional misc info function */ \
+                  NULL,                /* send scsi command, wait for completion */ \
+                  in2000_queuecommand, /* queue scsi command, don't wait */ \
+                  in2000_abort,        /* abort current command */ \
+                  in2000_reset,        /* reset scsi bus */ \
+                  NULL,                /* slave_attach - unused */ \
+                  in2000_biosparam,    /* figures out BIOS parameters for lilo, etc */ \
+                  IN2000_CAN_Q,        /* max commands we can queue up */ \
+                  IN2000_HOST_ID,      /* host-adapter scsi id */ \
+                  IN2000_SG,           /* scatter-gather table size */ \
+                  IN2000_CPL,          /* commands per lun */ \
+                  0,                   /* board counter */ \
+                  0,                   /* unchecked dma */ \
+                  DISABLE_CLUSTERING \
+               }
+
+#else       /* 2.1.xxx */
+
 #define IN2000 {  proc_dir:        &proc_scsi_in2000,   /* pointer to /proc/scsi directory entry */ \
                   proc_info:       in2000_proc_info,    /* pointer to proc info function */ \
                   name:            "Always IN2000",     /* device name */ \
@@ -408,5 +454,7 @@ int in2000_reset(Scsi_Cmnd *, unsigned int);
                   use_new_eh_code: 0                    /* new error code - not using it yet */ \
                 }
 
+#endif
+
 
 #endif /* IN2000_H */

fs/exec.c

@@ -569,6 +569,15 @@ int flush_old_exec(struct linux_binprm * bprm)
 	return retval;
 }
 
+/*
+ * We mustn't allow tracing of suid binaries, unless
+ * the tracer has the capability to trace anything..
+ */
+static inline int must_not_trace_exec(struct task_struct * p)
+{
+	return (p->flags & PF_PTRACED) && !cap_raised(p->p_pptr->cap_effective, CAP_SYS_PTRACE);
+}
+
 /* 
  * Fill the binprm structure from the inode. 
  * Check permissions, then read the first 512 bytes
@@ -657,15 +666,12 @@ int prepare_binprm(struct linux_binprm *bprm)
 		}
 	}
 
-
-
-
 	if (id_change || cap_raised) {
 		/* We can't suid-execute if we're sharing parts of the executable */
 		/* or if we're being traced (or if suid execs are not allowed)    */
 		/* (current->mm->count > 1 is ok, as we'll get a new mm anyway)   */
 		if (IS_NOSUID(inode)
-		    || (current->flags & PF_PTRACED)
+		    || must_not_trace_exec(current)
 		    || (atomic_read(&current->fs->count) > 1)
 		    || (atomic_read(&current->sig->count) > 1)
 		    || (atomic_read(&current->files->count) > 1)) {

include/linux/if.h

@@ -114,6 +114,7 @@ struct ifreq
 		int	ifru_mtu;
 		struct  ifmap ifru_map;
 		char	ifru_slave[IFNAMSIZ];	/* Just fits the size */
+		char	ifru_newname[IFNAMSIZ];
 		__kernel_caddr_t	ifru_data;
 	} ifr_ifru;
 };
@@ -133,6 +134,7 @@ struct ifreq
 #define ifr_ifindex	ifr_ifru.ifru_ivalue	/* interface index	*/
 #define ifr_bandwidth	ifr_ifru.ifru_ivalue    /* link bandwidth	*/
 #define ifr_qlen	ifr_ifru.ifru_ivalue	/* Queue length 	*/
+#define ifr_newname	ifr_ifru.ifru_newname	/* New name		*/
 
 /*
  * Structure used in SIOCGIFCONF request.

include/linux/if_packet.h

@@ -25,9 +25,10 @@ struct sockaddr_ll
 #define PACKET_BROADCAST	1		/* To all		*/
 #define PACKET_MULTICAST	2		/* To group		*/
 #define PACKET_OTHERHOST	3		/* To someone else 	*/
-#define PACKET_OUTGOING		4		/* Originated by us	*/
-#define PACKET_LOOPBACK		5
-#define PACKET_FASTROUTE	6
+#define PACKET_OUTGOING		4		/* Outgoing of any type */
+/* These ones are invisible by user level */
+#define PACKET_LOOPBACK		5		/* MC/BRD frame looped back */
+#define PACKET_FASTROUTE	6		/* Fastrouted frame	*/
 
 /* Packet socket options */
 

include/linux/in6.h

@@ -35,7 +35,14 @@ struct in6_addr
 		__u16		u6_addr16[8];
 		__u32		u6_addr32[4];
 #if (~0UL) > 0xffffffff
+#ifndef __RELAX_IN6_ADDR_ALIGNMENT
+		/* Alas, protocols do not respect 64bit alignmnet.
+		   rsvp/pim/... are broken. However, it is good
+		   idea to force correct alignment always, when
+		   it is possible.
+		 */
 		__u64		u6_addr64[2];
+#endif
 #endif
 	} in6_u;
 #define s6_addr			in6_u.u6_addr8
@@ -100,20 +107,35 @@ struct ipv6_mreq {
 #define IPPROTO_NONE		59	/* IPv6 no next header		*/
 #define IPPROTO_DSTOPTS		60	/* IPv6 destination options	*/
 
+/*
+ *	IPv6 TLV options.
+ */
+#define IPV6_TLV_PAD0		0
+#define IPV6_TLV_PADN		1
+#define IPV6_TLV_ROUTERALERT	20
+#define IPV6_TLV_JUMBO		194
+
 /*
  *	IPV6 socket options
  */
 
 #define IPV6_ADDRFORM		1
 #define IPV6_PKTINFO		2
-#define IPV6_RXHOPOPTS		3 /* obsolete name */
-#define IPV6_RXDSTOPTS		4 /* obsolete name */
-#define IPV6_HOPOPTS		IPV6_RXHOPOPTS  /* new name */
-#define IPV6_DSTOPTS		IPV6_RXDSTOPTS  /* new name */
-#define IPV6_RXSRCRT		5
+#define IPV6_HOPOPTS		3
+#define IPV6_DSTOPTS		4
+#define IPV6_RTHDR		5
 #define IPV6_PKTOPTIONS		6
 #define IPV6_CHECKSUM		7
 #define IPV6_HOPLIMIT		8
+#define IPV6_NEXTHOP		9
+#define IPV6_AUTHHDR		10
+
+#if 0
+/* Aliases for obsolete names */
+#define IPV6_RXHOPOPTS		IPV6_HOPOPTS
+#define IPV6_RXDSTOPTS		IPV6_DSTOPTS
+#define IPV6_RXSRCRT		IPV6_RTHDR
+#endif
 
 /*
  *	Alternative names

include/linux/ipv6.h

@@ -4,6 +4,9 @@
 #include <linux/in6.h>
 #include <asm/byteorder.h>
 
+/* The latest drafts declared increase in minimal mtu up to 1280. */
+
+#define IPV6_MIN_MTU	1280
 
 /*
  *	Advanced API
@@ -58,8 +61,6 @@ struct ipv6_opt_hdr {
 #define ipv6_optlen(p)  (((p)->hdrlen+1) << 3)
 #endif
 
-
-
 /*
  *	routing header type 0 (used in cmsghdr struct)
  */
@@ -72,10 +73,11 @@ struct rt0_hdr {
 #define rt0_type		rt_hdr.type;
 };
 
-#ifdef __KERNEL__
-
 /*
  *	IPv6 fixed header
+ *
+ *	BEWARE, it is incorrect. The first 4 bits of flow_lbl
+ *	are glued to priority now, forming "class".
  */
 
 struct ipv6hdr {
@@ -87,7 +89,7 @@ struct ipv6hdr {
 				priority:4;
 #else
 #error	"Please fix <asm/byteorder.h>"
-#endif						
+#endif
 	__u8			flow_lbl[3];
 
 	__u16			payload_len;
@@ -98,28 +100,24 @@ struct ipv6hdr {
 	struct	in6_addr	daddr;
 };
 
-/*
- *	The length of this struct cannot be greater than the length of
- *	the proto_priv field in a sk_buff which is currently
- *	defined to be 16 bytes.
- *	Pointers take upto 8 bytes (sizeof(void *) is 8 on the alpha).
- */
-struct ipv6_options 
-{
-	/* length of extension headers   */
-
-	__u16			opt_flen;	/* after fragment hdr */
-	__u16			opt_nflen;	/* before fragment hdr */
+#ifdef __KERNEL__
 
-	/* 
-	 * protocol options 
-	 * usually carried in IPv6 extension headers
-	 */
+/* 
+   This structure contains results of exthdrs parsing
+   as offsets from skb->nh.
+ */
 
-	struct ipv6_rt_hdr		*srcrt;	/* Routing Header */
+struct inet6_skb_parm
+{
+	int			iif;
+	__u16			ra;
+	__u16			hop;
+	__u16			auth;
+	__u16			dst0;
+	__u16			srcrt;
+	__u16			dst1;
 };
 
-
 #endif
 
 #endif

include/linux/ipv6_route.h

@@ -26,7 +26,6 @@ enum
 #define RTF_ALLONLINK	0x00020000	/* fallback, no routers on link	*/
 #define RTF_ADDRCONF	0x00040000	/* addrconf route - RA		*/
 
-#define RTF_LINKRT	0x00100000	/* link specific - device match	*/
 #define RTF_NONEXTHOP	0x00200000	/* route with no nexthop	*/
 #define RTF_EXPIRES	0x00400000
 

include/linux/netdevice.h

@@ -232,8 +232,7 @@ struct device
 	
 	unsigned short		flags;	/* interface flags (a la BSD)	*/
 	unsigned short		gflags;
-	unsigned short		metric;	/* routing metric (not used)	*/
-	unsigned short		mtu;	/* interface MTU value		*/
+	unsigned		mtu;	/* interface MTU value		*/
 	unsigned short		type;	/* interface hardware type	*/
 	unsigned short		hard_header_len;	/* hardware hdr length	*/
 	void			*priv;	/* pointer to private data	*/

include/linux/netlink.h

@@ -16,8 +16,8 @@ struct sockaddr_nl
 {
 	sa_family_t	nl_family;	/* AF_NETLINK	*/
 	unsigned short	nl_pad;		/* zero		*/
-	__kernel_pid_t	nl_pid;		/* process pid	*/
-	unsigned	nl_groups;	/* multicast groups mask */
+	__u32		nl_pid;		/* process pid	*/
+       	__u32		nl_groups;	/* multicast groups mask */
 };
 
 struct nlmsghdr
@@ -26,7 +26,7 @@ struct nlmsghdr
 	__u16		nlmsg_type;	/* Message content */
 	__u16		nlmsg_flags;	/* Additional flags */
 	__u32		nlmsg_seq;	/* Sequence number */
-	__kernel_pid_t	nlmsg_pid;	/* Sending process PID */
+	__u32		nlmsg_pid;	/* Sending process PID */
 };
 
 /* Flags values */
@@ -64,7 +64,7 @@ struct nlmsghdr
 #define NLMSG_DATA(nlh)  ((void*)(((char*)nlh) + NLMSG_LENGTH(0)))
 #define NLMSG_NEXT(nlh,len)	 ((len) -= NLMSG_ALIGN((nlh)->nlmsg_len), \
 				  (struct nlmsghdr*)(((char*)(nlh)) + NLMSG_ALIGN((nlh)->nlmsg_len)))
-#define NLMSG_OK(nlh,len) ((nlh)->nlmsg_len >= sizeof(struct nlmsghdr) && \
+#define NLMSG_OK(nlh,len) ((len) > 0 && (nlh)->nlmsg_len >= sizeof(struct nlmsghdr) && \
 			   (nlh)->nlmsg_len <= (len))
 #define NLMSG_PAYLOAD(nlh,len) ((nlh)->nlmsg_len - NLMSG_SPACE((len)))
 
@@ -86,10 +86,11 @@ struct nlmsgerr
 struct netlink_skb_parms
 {
 	struct ucred		creds;		/* Skb credentials	*/
-	pid_t			pid;
-	unsigned		groups;
-	pid_t			dst_pid;
-	unsigned		dst_groups;
+	__u32			pid;
+	__u32			groups;
+	__u32			dst_pid;
+	__u32			dst_groups;
+	kernel_cap_t		eff_cap;
 };
 
 #define NETLINK_CB(skb)		(*(struct netlink_skb_parms*)&((skb)->cb))
@@ -102,10 +103,10 @@ extern int netlink_post(int unit, struct sk_buff *skb);
 extern int init_netlink(void);
 extern struct sock *netlink_kernel_create(int unit, void (*input)(struct sock *sk, int len));
 extern void netlink_ack(struct sk_buff *in_skb, struct nlmsghdr *nlh, int err);
-extern int netlink_unicast(struct sock *ssk, struct sk_buff *skb, pid_t pid, int nonblock);
-extern void netlink_broadcast(struct sock *ssk, struct sk_buff *skb, pid_t pid,
-			      unsigned group, int allocation);
-extern void netlink_set_err(struct sock *ssk, pid_t pid, unsigned group, int code);
+extern int netlink_unicast(struct sock *ssk, struct sk_buff *skb, __u32 pid, int nonblock);
+extern void netlink_broadcast(struct sock *ssk, struct sk_buff *skb, __u32 pid,
+			      __u32 group, int allocation);
+extern void netlink_set_err(struct sock *ssk, __u32 pid, __u32 group, int code);
 
 /*
  *	skb should fit one page. This choice is good for headerless malloc.
@@ -125,28 +126,8 @@ struct netlink_callback
 	long		args[4];
 };
 
-#if 0
-
-void* nlmsg_broadcast(struct sock*, unsigned long type, int len, unsigned groups);
-struct skb_buff *nlmsg_alloc(unsigned long type, int len,
-			     unsigned long seq, unsigned long pid, int allocation);
-void __nlmsg_transmit(struct sock*, int allocation);
-
-extern __inline__ void nlmsg_release(struct sk_buff *skb)
-{
-	atomic_dec(skb->users);
-}
-
-extern __inline__ void nlmsg_transmit(struct sk_buff *sk, int allocation)
-{
-	if (sk->write_queue.qlen)
-		__nlmsg_transmit(sk, allocation);
-}
-
-#endif
-
 extern __inline__ struct nlmsghdr *
-__nlmsg_put(struct sk_buff *skb, pid_t pid, u32 seq, int type, int len)
+__nlmsg_put(struct sk_buff *skb, u32 pid, u32 seq, int type, int len)
 {
 	struct nlmsghdr *nlh;
 	int size = NLMSG_LENGTH(len);

include/linux/notifier.h

@@ -98,6 +98,7 @@ extern __inline__ int notifier_call_chain(struct notifier_block **n, unsigned lo
 #define NETDEV_CHANGEMTU	0x0007
 #define NETDEV_CHANGEADDR	0x0008
 #define NETDEV_GOING_DOWN	0x0009
+#define NETDEV_CHANGENAME	0x000A
 
 #define SYS_DOWN	0x0001	/* Notify of system down */
 #define SYS_RESTART	SYS_DOWN

include/linux/proc_fs.h

@@ -119,7 +119,7 @@ enum net_directory_inos {
 	PROC_NET_AX25_BPQETHER,
 	PROC_NET_IP_MASQ_APP,
 	PROC_NET_RT6,
-	PROC_NET_RT6_TREE,
+	PROC_NET_SNMP6,
 	PROC_NET_RT6_STATS,
 	PROC_NET_NDISC,
 	PROC_NET_STRIP_STATUS,

include/linux/rtnetlink.h

@@ -5,7 +5,6 @@
 #include <linux/netlink.h>
 
 #define RTNL_DEBUG 1
-/* #define CONFIG_RTNL_OLD_IFINFO 1 */
 
 
 /****
@@ -66,14 +65,14 @@ struct rtattr
 
 #define RTA_ALIGNTO	4
 #define RTA_ALIGN(len) ( ((len)+RTA_ALIGNTO-1) & ~(RTA_ALIGNTO-1) )
-#define RTA_OK(rta,len) ((rta)->rta_len >= sizeof(struct rtattr) && \
+#define RTA_OK(rta,len) ((len) > 0 && (rta)->rta_len >= sizeof(struct rtattr) && \
 			 (rta)->rta_len <= (len))
 #define RTA_NEXT(rta,attrlen)	((attrlen) -= RTA_ALIGN((rta)->rta_len), \
 				 (struct rtattr*)(((char*)(rta)) + RTA_ALIGN((rta)->rta_len)))
 #define RTA_LENGTH(len)	(RTA_ALIGN(sizeof(struct rtattr)) + (len))
 #define RTA_SPACE(len)	RTA_ALIGN(RTA_LENGTH(len))
 #define RTA_DATA(rta)   ((void*)(((char*)(rta)) + RTA_LENGTH(0)))
-#define RTA_PAYLOAD(rta) ((rta)->rta_len - RTA_LENGTH(0))
+#define RTA_PAYLOAD(rta) ((int)((rta)->rta_len) - RTA_LENGTH(0))
 
 
 
@@ -91,18 +90,9 @@ struct rtmsg
 
 	unsigned char		rtm_table;	/* Routing table id */
 	unsigned char		rtm_protocol;	/* Routing protocol; see below	*/
-#ifdef CONFIG_RTNL_OLD_IFINFO
-	unsigned char		rtm_nhs;	/* Number of nexthops */
-#else
 	unsigned char		rtm_scope;	/* See below */	
-#endif
 	unsigned char		rtm_type;	/* See below	*/
 
-#ifdef CONFIG_RTNL_OLD_IFINFO
-	unsigned short		rtm_optlen;	/* Byte length of rtm_opt */
-	unsigned char		rtm_scope;	/* See below */	
-	unsigned char		rtm_whatsit;	/* Unused byte */
-#endif
 	unsigned		rtm_flags;
 };
 
@@ -176,9 +166,6 @@ enum rt_scope_t
 #define RTM_F_NOTIFY		0x100	/* Notify user of route change	*/
 #define RTM_F_CLONED		0x200	/* This route is cloned		*/
 #define RTM_F_EQUALIZE		0x400	/* Multipath equalizer: NI	*/
-#ifdef CONFIG_RTNL_OLD_IFINFO
-#define RTM_F_NOPMTUDISC	0x800	/* Do not make PMTU discovery	*/
-#endif
 
 /* Reserved table identifiers */
 
@@ -206,17 +193,10 @@ enum rtattr_type_t
 	RTA_GATEWAY,
 	RTA_PRIORITY,
 	RTA_PREFSRC,
-#ifndef CONFIG_RTNL_OLD_IFINFO
 	RTA_METRICS,
 	RTA_MULTIPATH,
 	RTA_PROTOINFO,
 	RTA_FLOW,
-#else
-	RTA_WINDOW,
-	RTA_RTT,
-	RTA_MTU,
-	RTA_IFNAME,
-#endif
 	RTA_CACHEINFO
 };
 
@@ -253,18 +233,12 @@ struct rtnexthop
 #define RTNH_ALIGNTO	4
 #define RTNH_ALIGN(len) ( ((len)+RTNH_ALIGNTO-1) & ~(RTNH_ALIGNTO-1) )
 #define RTNH_OK(rtnh,len) ((rtnh)->rtnh_len >= sizeof(struct rtnexthop) && \
-			   (rtnh)->rtnh_len <= (len))
+			   ((int)(rtnh)->rtnh_len) <= (len))
 #define RTNH_NEXT(rtnh)	((struct rtnexthop*)(((char*)(rtnh)) + RTNH_ALIGN((rtnh)->rtnh_len)))
 #define RTNH_LENGTH(len) (RTNH_ALIGN(sizeof(struct rtnexthop)) + (len))
 #define RTNH_SPACE(len)	RTNH_ALIGN(RTNH_LENGTH(len))
 #define RTNH_DATA(rtnh)   ((struct rtattr*)(((char*)(rtnh)) + RTNH_LENGTH(0)))
 
-#ifdef CONFIG_RTNL_OLD_IFINFO
-#define RTM_RTNH(r) ((struct rtnexthop*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct rtmsg)) \
-					   + NLMSG_ALIGN((r)->rtm_optlen)))
-#define RTM_NHLEN(nlh,r) ((nlh)->nlmsg_len - NLMSG_SPACE(sizeof(struct rtmsg)) - NLMSG_ALIGN((r)->rtm_optlen))
-#endif
-
 /* RTM_CACHEINFO */
 
 struct rta_cacheinfo
@@ -424,35 +398,6 @@ struct rtgenmsg
  * on network protocol.
  */
 
-#ifdef CONFIG_RTNL_OLD_IFINFO
-struct ifinfomsg
-{
-	unsigned char	ifi_family;		/* Dummy	*/
-	unsigned char	ifi_addrlen;		/* Length of HW address */
-	unsigned short	ifi_pad__;
-	int		ifi_index;		/* Link index	*/
-	int		ifi_link;		/* Physical device */
-	char		ifi_name[IFNAMSIZ];
-	struct sockaddr	ifi_address;		/* HW address	*/
-	struct sockaddr	ifi_broadcast;		/* HW broadcast	*/
-	unsigned	ifi_flags;		/* IFF_* flags	*/
-	int		ifi_mtu;		/* Link mtu	*/
-	char		ifi_qdiscname[IFNAMSIZ];/* Id of packet scheduler */
-	int		ifi_qdisc;		/* Packet scheduler handle */
-};
-
-enum
-{
-	IFLA_UNSPEC,
-	IFLA_ADDRESS,
-	IFLA_BROADCAST,
-	IFLA_IFNAME,
-	IFLA_QDISC,
-	IFLA_STATS
-};
-
-#else
-
 struct ifinfomsg
 {
 	unsigned char	ifi_family;
@@ -475,8 +420,6 @@ enum
 	IFLA_STATS
 };
 
-#endif
-
 
 #define IFLA_MAX IFLA_STATS
 
@@ -588,7 +531,7 @@ struct rtnetlink_link
 
 extern struct rtnetlink_link * rtnetlink_links[NPROTO];
 extern int rtnetlink_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb);
-extern int rtnetlink_send(struct sk_buff *skb, u32 pid, unsigned group, int echo);
+extern int rtnetlink_send(struct sk_buff *skb, u32 pid, u32 group, int echo);
 
 extern void __rta_fill(struct sk_buff *skb, int attrtype, int attrlen, const void *data);
 

include/linux/skbuff.h

@@ -537,6 +537,19 @@ extern __inline__ struct sk_buff *dev_alloc_skb(unsigned int length)
 	return skb;
 }
 
+extern __inline__ struct sk_buff *
+skb_cow(struct sk_buff *skb, unsigned int headroom)
+{
+	headroom = (headroom+15)&~15;
+
+	if ((unsigned)skb_headroom(skb) < headroom || skb_cloned(skb)) {
+		struct sk_buff *skb2 = skb_realloc_headroom(skb, headroom);
+		kfree_skb(skb);
+		skb = skb2;
+	}
+	return skb;
+}
+
 extern struct sk_buff *		skb_recv_datagram(struct sock *sk,unsigned flags,int noblock, int *err);
 extern unsigned int		datagram_poll(struct file *file, struct socket *sock, struct poll_table_struct *wait);
 extern int			skb_copy_datagram(struct sk_buff *from, int offset, char *to,int size);

include/linux/socket.h

@@ -88,20 +88,27 @@ struct cmsghdr {
 
 /*
  *	Get the next cmsg header
+ *
+ *	PLEASE, do not touch this function. If you think, that it is
+ *	incorrect, grep kernel sources and think about consequences
+ *	before trying to improve it.
+ *
+ *	Now it always returns valid, not truncated ancillary object
+ *	HEADER. But caller still MUST check, that cmsg->cmsg_len is
+ *	inside range, given by msg->msg_controllen before using
+ *	ansillary object DATA.				--ANK (980731)
  */
  
 __KINLINE struct cmsghdr * __cmsg_nxthdr(void *__ctl, __kernel_size_t __size,
 					       struct cmsghdr *__cmsg)
 {
-	unsigned char * __ptr;
+	struct cmsghdr * __ptr;
 
-	if (__cmsg->cmsg_len < sizeof(struct cmsghdr))
-		return NULL;
-	__ptr = ((unsigned char *) __cmsg) +  CMSG_ALIGN(__cmsg->cmsg_len);
-	if (__ptr >= (unsigned char *) __ctl + __size)
+	__ptr = (struct cmsghdr*)(((unsigned char *) __cmsg) +  CMSG_ALIGN(__cmsg->cmsg_len));
+	if ((unsigned long)((char*)(__ptr+1) - (char *) __ctl) > __size)
 		return NULL;
 
-	return (struct cmsghdr *) __ptr;
+	return __ptr;
 }
 
 __KINLINE struct cmsghdr * cmsg_nxthdr (struct msghdr *__msg, struct cmsghdr *__cmsg)

include/linux/sockios.h

@@ -45,6 +45,7 @@
 #define SIOCSIFMEM	0x8920		/* set memory address (BSD)	*/
 #define SIOCGIFMTU	0x8921		/* get MTU size			*/
 #define SIOCSIFMTU	0x8922		/* set MTU size			*/
+#define SIOCSIFNAME	0x8923		/* set interface name */
 #define	SIOCSIFHWADDR	0x8924		/* set hardware address 	*/
 #define SIOCGIFENCAP	0x8925		/* get/set encapsulations       */
 #define SIOCSIFENCAP	0x8926		

include/net/dst.h

@@ -35,7 +35,6 @@ struct dst_entry
 	atomic_t		use;		/* client references	*/
 	struct device	        *dev;
 	int			obsolete;
-	__u32			priority;
 	unsigned long		lastuse;
 	unsigned		mxlock;
 	unsigned		window;

include/net/flow.h

@@ -21,6 +21,10 @@ struct flowi {
 			struct in6_addr *	saddr;
 		} ip6_u;
 	} nl_u;
+#define fl6_dst		nl_u.ip6_u.daddr
+#define fl6_src		nl_u.ip6_u.saddr
+#define fl4_dst		nl_u.ip4_u.daddr
+#define fl4_src		nl_u.ip4_u.saddr
 
 	int	oif;
 

include/net/ip6_fib.h

@@ -23,7 +23,8 @@
 
 struct rt6_info;
 
-struct fib6_node {
+struct fib6_node
+{
 	struct fib6_node	*parent;
 	struct fib6_node	*left;
 	struct fib6_node	*right;
@@ -43,12 +44,14 @@ struct fib6_node {
  *
  */
 
-struct rt6key {
+struct rt6key
+{
 	struct in6_addr	addr;
 	int		plen;
 };
 
-struct rt6_info {
+struct rt6_info
+{
 	union {
 		struct dst_entry	dst;
 		struct rt6_info		*next;
@@ -56,21 +59,16 @@ struct rt6_info {
 
 #define rt6i_dev			u.dst.dev
 #define rt6i_nexthop			u.dst.neighbour
-#define rt6i_use			u.dst.use
-#define rt6i_ref			u.dst.refcnt
-
-#define rt6i_tstamp			u.dst.lastuse
 
 	struct fib6_node		*rt6i_node;
 
 	struct in6_addr			rt6i_gateway;
 	
-	int				rt6i_keylen;
-
 	u32				rt6i_flags;
 	u32				rt6i_metric;
 	u8				rt6i_hoplimit;
 	unsigned long			rt6i_expires;
+	atomic_t			rt6i_ref;
 
 	union {
 		struct flow_rule	*rt6iu_flowr;
@@ -84,6 +82,33 @@ struct rt6_info {
 	struct rt6key			rt6i_src;
 };
 
+struct fib6_walker_t
+{
+	struct fib6_walker_t *prev, *next;
+	struct fib6_node *root, *node;
+	struct rt6_info *leaf;
+	unsigned char state;
+	unsigned char prune;
+	int (*func)(struct fib6_walker_t *);
+	void *args;
+};
+
+extern struct fib6_walker_t fib6_walker_list;
+
+extern __inline__ void fib6_walker_link(struct fib6_walker_t *w)
+{
+	w->next = fib6_walker_list.next;
+	w->prev = &fib6_walker_list;
+	w->next->prev = w;
+	w->prev->next = w;
+}
+
+extern __inline__ void fib6_walker_unlink(struct fib6_walker_t *w)
+{
+	w->next->prev = w->prev;
+	w->prev->next = w->next;
+	w->prev = w->next = w;
+}
 
 struct rt6_statistics {
 	__u32		fib_nodes;
@@ -97,8 +122,6 @@ struct rt6_statistics {
 #define RTN_ROOT	0x0002		/* tree root node		*/
 #define RTN_RTINFO	0x0004		/* node with valid routing info	*/
 
-#define RTN_TAG		0x0100
-
 /*
  *	priority levels (or metrics)
  *
@@ -128,11 +151,16 @@ extern struct fib6_node		*fib6_lookup(struct fib6_node *root,
 					     struct in6_addr *daddr,
 					     struct in6_addr *saddr);
 
-#define RT6_FILTER_RTNODES	1
+struct fib6_node		*fib6_locate(struct fib6_node *root,
+					     struct in6_addr *daddr, int dst_len,
+					     struct in6_addr *saddr, int src_len);
+
+extern void			fib6_clean_tree(struct fib6_node *root,
+						int (*func)(struct rt6_info *, void *arg),
+						int prune, void *arg);
 
-extern void			fib6_walk_tree(struct fib6_node *root,
-					       f_pnode func, void *arg,
-					       int filter);
+extern int			fib6_walk(struct fib6_walker_t *w);
+extern int			fib6_walk_continue(struct fib6_walker_t *w);
 
 extern int			fib6_add(struct fib6_node *root,
 					 struct rt6_info *rt);

include/net/ip6_route.h

@@ -12,23 +12,6 @@
 #include <net/flow.h>
 #include <net/ip6_fib.h>
 
-/*
- *	Structure for assync processing of operations on the routing
- *	table
- */
-
-struct rt6_req {
-	int			operation;
-	struct rt6_info		*ptr;
-
-	struct rt6_req		*next;
-	struct rt6_req		*prev;
-
-#define RT_OPER_ADD		1
-#define RT_OPER_DEL		2
-};
-
-
 struct pol_chain {
 	int			type;
 	int			priority;
@@ -53,8 +36,7 @@ extern void			ip6_route_cleanup(void);
 
 extern int			ipv6_route_ioctl(unsigned int cmd, void *arg);
 
-extern struct rt6_info *	ip6_route_add(struct in6_rtmsg *rtmsg,
-					      int *err);
+extern int			ip6_route_add(struct in6_rtmsg *rtmsg);
 extern int			ip6_del_rt(struct rt6_info *);
 
 extern int			ip6_rt_addr_add(struct in6_addr *addr,
@@ -85,15 +67,15 @@ extern struct rt6_info *	rt6_add_dflt_router(struct in6_addr *gwaddr,
 
 extern void			rt6_purge_dflt_routers(int lst_resort);
 
-extern struct rt6_info *	rt6_redirect(struct in6_addr *dest,
+extern void			rt6_redirect(struct in6_addr *dest,
 					     struct in6_addr *saddr,
-					     struct in6_addr *target,
-					     struct device *dev,
+					     struct neighbour *neigh,
 					     int on_link);
 
-extern void			rt6_pmtu_discovery(struct in6_addr *addr,
+extern void			rt6_pmtu_discovery(struct in6_addr *daddr,
+						   struct in6_addr *saddr,
 						   struct device *dev,
-						   int pmtu);
+						   u32 pmtu);
 
 struct nlmsghdr;
 struct netlink_callback;
@@ -103,22 +85,25 @@ extern int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *a
 extern int inet6_rtm_getroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg);
 
 extern void rt6_ifdown(struct device *dev);
+extern void rt6_mtu_change(struct device *dev, unsigned mtu);
 
 /*
  *	Store a destination cache entry in a socket
  *	For UDP/RAW sockets this is done on udp_connect.
  */
 
-extern __inline__ void ip6_dst_store(struct sock *sk, struct dst_entry *dst)
+extern __inline__ void ip6_dst_store(struct sock *sk, struct dst_entry *dst,
+				     struct in6_addr *daddr)
 {
 	struct ipv6_pinfo *np;
 	struct rt6_info *rt;
-		
+
 	np = &sk->net_pinfo.af_inet6;
 	dst_release(xchg(&sk->dst_cache,dst));
-	
+
 	rt = (struct rt6_info *) dst;
-	
+
+	np->daddr_cache = daddr;
 	np->dst_cookie = rt->rt6i_node ? rt->rt6i_node->fn_sernum : 0;
 }
 

include/net/ip_fib.h

@@ -212,7 +212,7 @@ extern int		fib_semantic_match(int type, struct fib_info *,
 extern struct fib_info	*fib_create_info(const struct rtmsg *r, struct kern_rta *rta,
 					 const struct nlmsghdr *, int *err);
 extern int fib_nh_match(struct rtmsg *r, struct nlmsghdr *, struct kern_rta *rta, struct fib_info *fi);
-extern int fib_dump_info(struct sk_buff *skb, pid_t pid, u32 seq, int event,
+extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
 			 u8 tb_id, u8 type, u8 scope, void *dst, int dst_len, u8 tos,
 			 struct fib_info *fi);
 extern int fib_sync_down(u32 local, struct device *dev, int force);

include/net/ipv6.h

@@ -4,7 +4,7 @@
  *	Authors:
  *	Pedro Roque		<roque@di.fc.ul.pt>
  *
- *	$Id: ipv6.h,v 1.12 1998/07/15 05:05:02 davem Exp $
+ *	$Id: ipv6.h,v 1.13 1998/08/26 12:02:11 davem Exp $
  *
  *	This program is free software; you can redistribute it and/or
  *      modify it under the terms of the GNU General Public License
@@ -86,53 +86,44 @@ struct frag_hdr {
 
 #include <net/sock.h>
 
-extern struct ipv6_mib	ipv6_statistics;
+extern struct ipv6_mib		ipv6_statistics;
+extern struct icmpv6_mib	icmpv6_statistics;
+extern struct udp_mib		udp_stats_in6;
 
-struct ipv6_frag {
-	__u16			offset;
-	__u16			len;
-	struct sk_buff		*skb;
-
-	struct frag_hdr		*fhdr;
-
-	struct ipv6_frag	*next;
+struct ip6_ra_chain
+{
+	struct ip6_ra_chain	*next;
+	struct sock		*sk;
+	int			sel;
+	void			(*destructor)(struct sock *);
 };
 
+extern struct ip6_ra_chain	*ip6_ra_chain;
+
 /*
- *	Equivalent of ipv4 struct ipq
+   This structure is prepared by protocol, when parsing
+   ancillary data and passed to IPv6.
  */
 
-struct frag_queue {
+struct ipv6_txoptions
+{
+	/* Length of this structure */
+	int			tot_len;
 
-	struct frag_queue	*next;
-	struct frag_queue	*prev;
+	/* length of extension headers   */
 
-	__u32			id;		/* fragment id		*/
-	struct in6_addr		saddr;
-	struct in6_addr		daddr;
-	struct timer_list	timer;		/* expire timer		*/
-	struct ipv6_frag	*fragments;
-	struct device		*dev;
-	__u8			last_in;	/* has last segment arrived? */
-	__u8			nexthdr;
-	__u8			*nhptr;
-};
+	__u16			opt_flen;	/* after fragment hdr */
+	__u16			opt_nflen;	/* before fragment hdr */
 
-struct ipv6_tlvtype
-{
-	u8 type;
-	u8 len;
-};
+	struct ipv6_opt_hdr	*hopopt;
+	struct ipv6_opt_hdr	*dst0opt;
+	struct ipv6_rt_hdr	*srcrt;	/* Routing Header */
+	struct ipv6_opt_hdr	*auth;
+	struct ipv6_opt_hdr	*dst1opt;
 
-struct ip6_ra_chain
-{
-	struct ip6_ra_chain	*next;
-	struct sock		*sk;
-	int			sel;
-	void			(*destructor)(struct sock *);
+	/* Option buffer, as read by IPV6_PKTOPTIONS, starts here. */
 };
 
-extern struct ip6_ra_chain	*ip6_ra_chain;
 
 extern int 			ip6_ra_control(struct sock *sk, int sel,
 					       void (*destructor)(struct sock *));
@@ -140,18 +131,13 @@ extern int 			ip6_ra_control(struct sock *sk, int sel,
 
 extern int			ip6_call_ra_chain(struct sk_buff *skb, int sel);
 
-extern int 			ip6_dstopt_unknown(struct sk_buff *skb,
-						   struct ipv6_tlvtype *hdr);
+extern u8 *			ipv6_reassembly(struct sk_buff **skb, u8 *nhptr);
 
-extern int			ipv6_routing_header(struct sk_buff **skb, 
-						    struct device *dev,
-						    __u8 *nhptr, 
-						    struct ipv6_options *opt);
+extern u8 *			ipv6_parse_hopopts(struct sk_buff *skb, u8 *nhptr);
 
-extern int			ipv6_reassembly(struct sk_buff **skb, 
-						struct device *dev, 
-						__u8 *nhptr,
-						struct ipv6_options *opt);
+extern u8 *			ipv6_parse_exthdrs(struct sk_buff **skb, u8 *nhptr);
+
+extern struct ipv6_txoptions *  ipv6_dup_options(struct sock *sk, struct ipv6_txoptions *opt);
 
 #define IPV6_FRAG_TIMEOUT	(60*HZ)		/* 60 seconds */
 
@@ -226,7 +212,7 @@ extern int			ipv6_rcv(struct sk_buff *skb,
 extern int			ip6_xmit(struct sock *sk,
 					 struct sk_buff *skb,
 					 struct flowi *fl,
-					 struct ipv6_options *opt);
+					 struct ipv6_txoptions *opt);
 
 extern int			ip6_nd_hdr(struct sock *sk,
 					   struct sk_buff *skb,
@@ -240,7 +226,7 @@ extern int			ip6_build_xmit(struct sock *sk,
 					       const void *data,
 					       struct flowi *fl,
 					       unsigned length,
-					       struct ipv6_options *opt,
+					       struct ipv6_txoptions *opt,
 					       int hlimit, int flags);
 
 /*
@@ -256,28 +242,27 @@ extern int			ip6_mc_input(struct sk_buff *skb);
  *	Extension header (options) processing
  */
 
-extern int			ipv6opt_bld_rthdr(struct sk_buff *skb,
-						  struct ipv6_options *opt,
-						  struct in6_addr *addr,
-						  int proto);
-
-extern int			ipv6opt_srcrt_co(struct sockaddr_in6 *sin6, 
-						 int len, 
-						 struct ipv6_options *opt);
-
-extern int			ipv6opt_srcrt_cl(struct sockaddr_in6 *sin6, 
-						 int num_addrs, 
-						 struct ipv6_options *opt);
-
-extern int			ipv6opt_srt_tosin(struct ipv6_options *opt,
-						  struct sockaddr_in6 *sin6,
-						  int len);
-
-extern void			ipv6opt_free(struct ipv6_options *opt);
-
-extern struct ipv6_opt_hdr *	ipv6_skip_exthdr(struct ipv6_opt_hdr *hdr, 
+extern u8 *			ipv6_build_nfrag_opts(struct sk_buff *skb,
+						      u8 *prev_hdr,
+						      struct ipv6_txoptions *opt,
+						      struct in6_addr *daddr,
+						      u32 jumbolen);
+extern u8 *			ipv6_build_frag_opts(struct sk_buff *skb,
+						     u8 *prev_hdr,
+						     struct ipv6_txoptions *opt);
+extern void 			ipv6_push_nfrag_opts(struct sk_buff *skb,
+						     struct ipv6_txoptions *opt,
+						     u8 *proto,
+						     struct in6_addr **daddr_p);
+extern void			ipv6_push_frag_opts(struct sk_buff *skb,
+						    struct ipv6_txoptions *opt,
+						    u8 *proto);
+
+extern u8 *			ipv6_skip_exthdr(struct ipv6_opt_hdr *hdr, 
 					         u8 *nexthdrp, int len);
 
+extern struct ipv6_txoptions *	ipv6_invert_rthdr(struct sock *sk,
+						  struct ipv6_rt_hdr *hdr);
 
 
 /*

include/net/ndisc.h

@@ -60,12 +60,7 @@ extern int			ndisc_init(struct net_proto_family *ops);
 
 extern void			ndisc_cleanup(void);
 
-extern int			ndisc_rcv(struct sk_buff *skb,
-					  struct device *dev,
-					  struct in6_addr *saddr,
-					  struct in6_addr *daddr,
-					  struct ipv6_options *opt,
-					  unsigned short len);
+extern int			ndisc_rcv(struct sk_buff *skb, unsigned long len);
 
 extern void			ndisc_send_ns(struct device *dev,
 					      struct neighbour *neigh,

include/net/pkt_sched.h

@@ -147,12 +147,12 @@ extern psched_time_t	psched_time_base;
 
 #if PSCHED_CLOCK_SOURCE == PSCHED_JIFFIES
 
-#define PSCHED_WATCHER
+#define PSCHED_WATCHER unsigned long
 
-extern unsigned long psched_time_mark;
+extern PSCHED_WATCHER psched_time_mark;
 
 #if HZ == 100
-#define PSCHED_JSCALE 7
+#define PSCHED_JSCALE 13
 #elif HZ == 1024
 #define PSCHED_JSCALE 10
 #else
@@ -179,9 +179,9 @@ extern int psched_clock_scale;
 
 #elif defined (__alpha__)
 
-#define PSCHED_WATCHER
+#define PSCHED_WATCHER u32
 
-extern u32 psched_time_mark;
+extern PSCHED_WATCHER psched_time_mark;
 
 #define PSCHED_GET_TIME(stamp) \
 ({ u32 __res; \

include/net/protocol.h

@@ -48,17 +48,13 @@ struct inet_protocol
 #if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
 struct inet6_protocol 
 {
-	int	(*handler)(struct sk_buff *skb, struct device *dev,
-			struct in6_addr *saddr,
-			struct in6_addr *daddr,
-			struct ipv6_options *opt, 
-			unsigned short len,
-			int redo, struct inet6_protocol *protocol);
+	int	(*handler)(struct sk_buff *skb,
+			unsigned long len);
 
-	void	(*err_handler)(struct sk_buff *skb, int type, int code, unsigned char *buff,
-			__u32 info, struct in6_addr *saddr,
-			struct in6_addr *daddr,
-			struct inet6_protocol *protocol);
+	void	(*err_handler)(struct sk_buff *skb, struct ipv6hdr *hdr,
+			       struct inet6_skb_parm *opt,
+			       int type, int code, unsigned char *buff,
+			       __u32 info);
 	struct inet6_protocol *next;
 	unsigned char	protocol;
 	unsigned char	copy:1;

include/net/rawv6.h

@@ -10,19 +10,17 @@ extern struct sock *raw_v6_htable[RAWV6_HTABLE_SIZE];
 extern struct sock *raw_v6_lookup(struct sock *sk, unsigned short num,
 				  struct in6_addr *loc_addr, struct in6_addr *rmt_addr);
 
-extern int			rawv6_rcv(struct sk_buff *skb, 
-					  struct device *dev,
-					  struct in6_addr *saddr, 
-					  struct in6_addr *daddr,
-					  struct ipv6_options *opt, 
-					  unsigned short len);
+extern int			rawv6_rcv(struct sock *sk,
+					  struct sk_buff *skb, 
+					  unsigned long len);
 
 
 extern void			rawv6_err(struct sock *sk,
+					  struct sk_buff *skb,
+					  struct ipv6hdr *hdr,
+					  struct inet6_skb_parm *opt,
 					  int type, int code, 
-					  unsigned char *buff,
-					  struct in6_addr *saddr,
-					  struct in6_addr *daddr);
+					  unsigned char *buff, u32 info);
 
 #endif
 

include/net/route.h

@@ -27,6 +27,11 @@
 #include <net/dst.h>
 #include <linux/in_route.h>
 #include <linux/rtnetlink.h>
+#include <linux/route.h>
+
+#ifndef __KERNEL__
+#warning This file is not supposed to be used outside of kernel.
+#endif
 
 #define RT_HASH_DIVISOR	    	256
 
@@ -36,7 +41,6 @@
  */
 #define RT_CACHE_BUBBLE_THRESHOLD	(5*HZ)
 
-#include <linux/route.h>
 
 #define RTO_ONLINK	0x01
 #define RTO_TPROXY	0x80000000
@@ -87,7 +91,8 @@ struct rtable
 #endif
 };
 
-#ifdef __KERNEL__
+extern struct rtable 	*rt_hash_table[RT_HASH_DIVISOR];
+
 extern void		ip_rt_init(void);
 extern void		ip_rt_redirect(u32 old_gw, u32 dst, u32 new_gw,
 				       u32 src, u8 tos, struct device *dev);
@@ -131,7 +136,4 @@ extern __inline__ int ip_route_connect(struct rtable **rp, u32 dst, u32 src, u32
 	return ip_route_output(rp, dst, src, tos, oif);
 }
 
-#endif
-
-
 #endif	/* _ROUTE_H */

include/net/snmp.h

@@ -52,11 +52,14 @@ struct ipv6_mib
 {
 	unsigned long	Ip6InReceives;
  	unsigned long	Ip6InHdrErrors;
+ 	unsigned long	Ip6InTooBigErrors;
+ 	unsigned long	Ip6InNoRoutes;
  	unsigned long	Ip6InAddrErrors;
- 	unsigned long	Ip6ForwDatagrams;
  	unsigned long	Ip6InUnknownProtos;
+ 	unsigned long	Ip6InTruncatedPkts;
  	unsigned long	Ip6InDiscards;
  	unsigned long	Ip6InDelivers;
+ 	unsigned long	Ip6OutForwDatagrams;
  	unsigned long	Ip6OutRequests;
  	unsigned long	Ip6OutDiscards;
  	unsigned long	Ip6OutNoRoutes;
@@ -67,6 +70,8 @@ struct ipv6_mib
  	unsigned long	Ip6FragOKs;
  	unsigned long	Ip6FragFails;
  	unsigned long	Ip6FragCreates;
+ 	unsigned long	Ip6InMcastPkts;
+ 	unsigned long	Ip6OutMcastPkts;
 };
  
 struct icmp_mib
@@ -98,6 +103,43 @@ struct icmp_mib
  	unsigned long	IcmpOutAddrMasks;
  	unsigned long	IcmpOutAddrMaskReps;
 };
+
+struct icmpv6_mib
+{
+	unsigned long	Icmp6InMsgs;
+	unsigned long	Icmp6InErrors;
+
+	unsigned long	Icmp6InDestUnreachs;
+	unsigned long	Icmp6InPktTooBigs;
+	unsigned long	Icmp6InTimeExcds;
+	unsigned long	Icmp6InParmProblems;
+
+	unsigned long	Icmp6InEchos;
+	unsigned long	Icmp6InEchoReplies;
+	unsigned long	Icmp6InGroupMembQueries;
+	unsigned long	Icmp6InGroupMembResponses;
+	unsigned long	Icmp6InGroupMembReductions;
+	unsigned long	Icmp6InRouterSolicits;
+	unsigned long	Icmp6InRouterAdvertisements;
+	unsigned long	Icmp6InNeighborSolicits;
+	unsigned long	Icmp6InNeighborAdvertisements;
+	unsigned long	Icmp6InRedirects;
+
+	unsigned long	Icmp6OutMsgs;
+
+	unsigned long	Icmp6OutDestUnreachs;
+	unsigned long	Icmp6OutPktTooBigs;
+	unsigned long	Icmp6OutTimeExcds;
+	unsigned long	Icmp6OutParmProblems;
+
+	unsigned long	Icmp6OutEchoReplies;
+	unsigned long	Icmp6OutRouterSolicits;
+	unsigned long	Icmp6OutNeighborSolicits;
+	unsigned long	Icmp6OutNeighborAdvertisements;
+	unsigned long	Icmp6OutRedirects;
+	unsigned long	Icmp6OutGroupMembResponses;
+	unsigned long	Icmp6OutGroupMembReductions;
+};
  
 struct tcp_mib
 {
@@ -131,6 +173,9 @@ struct linux_mib
 	unsigned long	SyncookiesRecv;
 	unsigned long	SyncookiesFailed;
 	unsigned long	EmbryonicRsts;
+	unsigned long	PruneCalled; 
+	unsigned long	RcvPruned;
+	unsigned long	OfoPruned;
 };
  	
 #endif

include/net/sock.h

@@ -87,6 +87,8 @@
 
 #include <asm/atomic.h>
 
+#define MIN_WRITE_SPACE	2048
+
 /* The AF_UNIX specific socket options */
 struct unix_opt {
 	int 			family;
@@ -134,6 +136,7 @@ struct ipv6_pinfo {
 	struct in6_addr 	saddr;
 	struct in6_addr 	rcv_saddr;
 	struct in6_addr		daddr;
+	struct in6_addr		*daddr_cache;
 
 	__u32			flow_lbl;
 	int			hop_limit;
@@ -141,21 +144,28 @@ struct ipv6_pinfo {
 	int			mcast_oif;
 	__u8			priority;
 
-
-	/* sockopt flags */
-
-	__u8			recvsrcrt:1,
-	                        rxinfo:1,
+	/* pktoption flags */
+	union {
+		struct {
+			__u8	srcrt:2,
+			        rxinfo:1,
 				rxhlim:1,
 				hopopts:1,
 				dstopts:1,
-				mc_loop:1,
-                                unused:2;
+                                authhdr:1,
+                                unused:1;
+		} bits;
+		__u8		all;
+	} rxopt;
+
+	/* sockopt flags */
+	__u8			mc_loop:1;
 
 	struct ipv6_mc_socklist	*ipv6_mc_list;
 	__u32			dst_cookie;
 
-	struct ipv6_options	*opt;
+	struct ipv6_txoptions	*opt;
+	struct sk_buff		*pktoptions;
 };
 
 struct raw6_opt {
@@ -207,6 +217,10 @@ struct tcp_opt {
 	__u32	snd_wl2;	/* Ack sequence for update		*/
 	__u32	snd_wnd;	/* The window we expect to receive	*/
 	__u32	max_window;
+	__u32	pmtu_cookie;	/* Last pmtu seen by socket		*/
+	__u16	mss_cache;	/* Cached effective mss, not including SACKS */
+	__u16	mss_clamp;	/* Maximal mss, negotiated at connection setup */
+	__u16	ext_header_len;	/* Dave, do you allow mw to use this hole? 8) --ANK */
 	__u8	pending;	/* pending events			*/
 	__u8	retransmits;
 	__u32	last_ack_sent;	/* last ack we sent			*/
@@ -226,6 +240,7 @@ struct tcp_opt {
  	__u32	snd_ssthresh;	/* Slow start size threshold		*/
 	__u8	dup_acks;	/* Consequetive duplicate acks seen from other end */
 	__u8	delayed_acks;
+	__u16	user_mss;  	/* mss requested by user in ioctl */
 
 	/* Two commonly used timers in both sender and receiver paths. */
  	struct timer_list	retransmit_timer;	/* Resend (no ack)	*/
@@ -252,7 +267,6 @@ struct tcp_opt {
 		wscale_ok,	/* Wscale seen on SYN packet		*/
 		sack_ok;	/* SACK seen on SYN packet		*/
 	char	saw_tstamp;	/* Saw TIMESTAMP on last packet		*/
-        __u16	in_mss;		/* MSS option received from sender	*/
         __u8	snd_wscale;	/* Window scaling received from sender	*/
         __u8	rcv_wscale;	/* Window scaling to send to receiver	*/
         __u32	rcv_tsval;	/* Time stamp value             	*/
@@ -270,6 +284,9 @@ struct tcp_opt {
 	__u32	urg_seq;
 	__u32	urg_data;
 
+	__u32	last_seg_size;	/* Size of last incoming segment */
+	__u32	rcv_mss;	/* MSS used for delayed ACK decisions */ 
+
 	struct open_request	*syn_wait_queue;
 	struct open_request	**syn_wait_last;
 
@@ -390,12 +407,6 @@ struct sock {
 
 	struct proto		*prot;
 
-	/* mss is min(mtu, max_window)
-	 * XXX Fix this, mtu only used in one TCP place and that is it -DaveM
-	 */
-	unsigned short		mtu;       /* mss negotiated in the syn's */
-	unsigned short		mss;       /* current eff. mss - can change */
-	unsigned short		user_mss;  /* mss requested by user in ioctl */
 	unsigned short		shutdown;
 
 #if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
@@ -868,6 +879,26 @@ extern __inline__ int sock_error(struct sock *sk)
 	return -err;
 }
 
+extern __inline__ unsigned long sock_wspace(struct sock *sk)
+{
+	int amt = 0;
+
+	if (!(sk->shutdown & SEND_SHUTDOWN)) {
+		amt = sk->sndbuf - atomic_read(&sk->wmem_alloc);
+		if (amt < 0) 
+			amt = 0;
+	}
+	return amt;
+}
+
+/*
+ *	Default write policy as shown to user space via poll/select/SIGIO
+ *	Kernel internally doesn't use the MIN_WRITE_SPACE threshold.
+ */
+extern __inline__ int sock_writeable(struct sock *sk) 
+{
+	return sock_wspace(sk) >= MIN_WRITE_SPACE;
+}
 
 /* 
  *	Declarations from timer.c 

include/net/tcp.h

@@ -78,6 +78,7 @@ struct tcp_bind_bucket {
 	unsigned short		flags;
 #define TCPB_FLAG_LOCKED	0x0001
 #define TCPB_FLAG_FASTREUSE	0x0002
+#define TCPB_FLAG_GOODSOCKNUM	0x0004
 
 	struct tcp_bind_bucket	*next;
 	struct sock		*owners;
@@ -230,11 +231,8 @@ static __inline__ int tcp_sk_listen_hashfn(struct sock *sk)
 	return tcp_lhashfn(sk->num);
 }
 
-#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
-#define NETHDR_SIZE	sizeof(struct ipv6hdr)
-#else
-#define NETHDR_SIZE	sizeof(struct iphdr) + 40
-#endif
+/* Note, that it is > than ipv6 header */
+#define NETHDR_SIZE	(sizeof(struct iphdr) + 40)
 
 /*
  * 40 is maximal IP options size
@@ -257,7 +255,6 @@ static __inline__ int tcp_sk_listen_hashfn(struct sock *sk)
 #define MIN_WINDOW	2048
 #define MAX_ACK_BACKLOG	2
 #define MAX_DELAY_ACK	2
-#define MIN_WRITE_SPACE	2048
 #define TCP_WINDOW_DIFF	2048
 
 /* urg_data states */
@@ -354,7 +351,7 @@ struct tcp_v4_open_req {
 struct tcp_v6_open_req {
 	struct in6_addr		loc_addr;
 	struct in6_addr		rmt_addr;
-	struct ipv6_options	*opt;
+	struct sk_buff		*pktopts;
 	int			iif;
 };
 #endif
@@ -400,6 +397,13 @@ extern kmem_cache_t *tcp_openreq_cachep;
 /*
  *	Pointers to address related TCP functions
  *	(i.e. things that depend on the address family)
+ *
+ * 	BUGGG_FUTURE: all the idea behind this struct is wrong.
+ *	It mixes socket frontend with transport function.
+ *	With port sharing between IPv6/v4 it gives the only advantage,
+ *	only poor IPv6 needs to permanently recheck, that it
+ *	is still IPv6 8)8) It must be cleaned up as soon as possible.
+ *						--ANK (980802)
  */
 
 struct tcp_func {
@@ -414,7 +418,7 @@ struct tcp_func {
 
 	int			(*conn_request)		(struct sock *sk,
 							 struct sk_buff *skb,
-							 void *opt, __u32 isn);
+							 __u32 isn);
 
 	struct sock *		(*syn_recv_sock)	(struct sock *sk,
 							 struct sk_buff *skb,
@@ -424,6 +428,10 @@ struct tcp_func {
 	struct sock *		(*get_sock)		(struct sk_buff *skb,
 							 struct tcphdr *th);
 
+	__u16			net_header_len;
+
+
+
 	int			(*setsockopt)		(struct sock *sk, 
 							 int level, 
 							 int optname, 
@@ -490,22 +498,24 @@ extern int			tcp_ioctl(struct sock *sk,
 extern int			tcp_rcv_state_process(struct sock *sk, 
 						      struct sk_buff *skb,
 						      struct tcphdr *th,
-						      void *opt, __u16 len);
+						      unsigned len);
 
 extern int			tcp_rcv_established(struct sock *sk, 
 						    struct sk_buff *skb,
 						    struct tcphdr *th, 
-						    __u16 len);
+						    unsigned len);
 
 extern int			tcp_timewait_state_process(struct tcp_tw_bucket *tw,
 							   struct sk_buff *skb,
 							   struct tcphdr *th,
-							   void *opt, __u16 len);
+							   unsigned len);
 
 extern void			tcp_close(struct sock *sk, 
 					  unsigned long timeout);
 extern struct sock *		tcp_accept(struct sock *sk, int flags);
 extern unsigned int		tcp_poll(struct file * file, struct socket *sock, struct poll_table_struct *wait);
+extern void			tcp_write_space(struct sock *sk); 
+
 extern int			tcp_getsockopt(struct sock *sk, int level, 
 					       int optname, char *optval, 
 					       int *optlen);
@@ -536,12 +546,11 @@ extern void		       	tcp_v4_send_check(struct sock *sk,
 
 extern int			tcp_v4_conn_request(struct sock *sk,
 						    struct sk_buff *skb,
-						    void *ptr, __u32 isn);
+						    __u32 isn);
 
 extern struct sock *		tcp_create_openreq_child(struct sock *sk,
 							 struct open_request *req,
-							 struct sk_buff *skb,
-							 int mss);
+							 struct sk_buff *skb);
 
 extern struct sock *		tcp_v4_syn_recv_sock(struct sock *sk,
 						     struct sk_buff *skb,
@@ -628,30 +637,25 @@ struct tcp_sl_timer {
 
 extern struct tcp_sl_timer tcp_slt_array[TCP_SLT_MAX];
  
+extern int tcp_sync_mss(struct sock *sk, u32 pmtu);
+
 /* Compute the current effective MSS, taking SACKs and IP options,
  * and even PMTU discovery events into account.
  */
+
 static __inline__ unsigned int tcp_current_mss(struct sock *sk)
 {
 	struct tcp_opt *tp = &sk->tp_pinfo.af_tcp;
 	struct dst_entry *dst = sk->dst_cache;
-	unsigned int mss_now = sk->mss; 
+	int mss_now = tp->mss_cache; 
 
-	if(dst && (sk->mtu < dst->pmtu)) {
-		unsigned int mss_distance = (sk->mtu - sk->mss);
-
-		/* PMTU discovery event has occurred. */
-		sk->mtu = dst->pmtu;
-		mss_now = sk->mss = sk->mtu - mss_distance;
-	}
+	if (dst && dst->pmtu != tp->pmtu_cookie)
+		mss_now = tcp_sync_mss(sk, dst->pmtu);
 
 	if(tp->sack_ok && tp->num_sacks)
 		mss_now -= (TCPOLEN_SACK_BASE_ALIGNED +
 			    (tp->num_sacks * TCPOLEN_SACK_PERBLOCK));
-	if(sk->opt)
-		mss_now -= sk->opt->optlen;
-
-	return mss_now; 
+	return mss_now > 8 ? mss_now : 8;
 }
 
 /* Compute the actual receive window we are currently advertising.
@@ -715,7 +719,12 @@ extern __inline__ int tcp_raise_window(struct sock *sk)
  * skbuff.h:skbuff->cb[xxx] size appropriately.
  */
 struct tcp_skb_cb {
-	struct inet_skb_parm	header;	/* For incoming frames		*/
+	union {
+		struct inet_skb_parm	h4;
+#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
+		struct inet6_skb_parm	h6;
+#endif
+	} header;	/* For incoming frames		*/
 	__u32		seq;		/* Starting sequence number	*/
 	__u32		end_seq;	/* SEQ + FIN + SYN + datalen	*/
 	unsigned long	when;		/* used to compute rtt's	*/
@@ -787,7 +796,7 @@ static __inline__ int tcp_snd_test(struct sock *sk, struct sk_buff *skb)
 	 *
 	 * 	Don't use the nagle rule for urgent data.
 	 */
-	if (!sk->nonagle && skb->len < (sk->mss >> 1) && tp->packets_out &&
+	if (!sk->nonagle && skb->len < (tp->mss_cache >> 1) && tp->packets_out &&
 	    !(TCP_SKB_CB(skb)->flags & TCPCB_FLAG_URG))
 		nagle_check = 0;
 
@@ -913,8 +922,6 @@ extern __inline__ void tcp_syn_build_options(__u32 *ptr, int mss, int ts, int sa
 	 * SACKs don't matter, we never delay an ACK when we
 	 * have any of those going out.
 	 */
-	if(ts)
-		mss += TCPOLEN_TSTAMP_ALIGNED;
 	*ptr++ = htonl((TCPOPT_MSS << 24) | (TCPOLEN_MSS << 16) | mss);
 	if (ts) {
 		if(sack)

include/net/transp_v6.h

@@ -28,7 +28,7 @@ extern int			datagram_recv_ctl(struct sock *sk,
 extern int			datagram_send_ctl(struct msghdr *msg,
 						  int *oif,
 						  struct in6_addr **src_addr,
-						  struct ipv6_options *opt,
+						  struct ipv6_txoptions *opt,
 						  int *hlimit);
 
 #define		LOOPBACK4_IPV6		__constant_htonl(0x7f000006)
@@ -38,6 +38,8 @@ extern int			datagram_send_ctl(struct msghdr *msg,
  */
 extern struct tcp_func	ipv4_specific;
 
+extern int inet6_destroy_sock(struct sock *sk);
+
 #endif
 
 #endif

init/main.c

@@ -1129,7 +1129,6 @@ asmlinkage void __init start_kernel(void)
 	mtrr_init ();
 #endif
 
-	sock_init();
 #ifdef CONFIG_SYSCTL
 	sysctl_init();
 #endif
@@ -1200,6 +1199,9 @@ static int init(void * unused)
 	int real_root_mountflags;
 #endif
 
+	/* Networking initialization needs a process context */ 
+	sock_init();
+
 	/* Launch bdflush from here, instead of the old syscall way. */
 	kernel_thread(bdflush, NULL, CLONE_FS | CLONE_FILES | CLONE_SIGHAND);
 	/* Start the background pageout daemon. */

net/Config.in

@@ -17,7 +17,9 @@ bool 'TCP/IP networking' CONFIG_INET
 if [ "$CONFIG_INET" = "y" ]; then
   source net/ipv4/Config.in
   if [ "$CONFIG_EXPERIMENTAL" = "y" ]; then
-    tristate 'The IPv6 protocol (EXPERIMENTAL)' CONFIG_IPV6
+#   Sorry, but IPv6 as module is still invalid.
+#   tristate 'The IPv6 protocol (EXPERIMENTAL)' CONFIG_IPV6
+    bool 'The IPv6 protocol (EXPERIMENTAL)' CONFIG_IPV6
     if [ "$CONFIG_IPV6" != "n" ]; then
 	    source net/ipv6/Config.in
     fi

net/appletalk/ddp.c

@@ -1017,7 +1017,6 @@ static int atalk_create(struct socket *sock, int protocol)
 
 	sk->destruct = NULL;
 	/* Checksums on by default */
-	sk->mtu = DDP_MAXSZ;
 	sk->zapped = 1;
 
 	return (0);

net/ax25/af_ax25.c

@@ -849,7 +849,6 @@ int ax25_create(struct socket *sock, int protocol)
 	sk->destruct = ax25_free_sock;
 	sock->ops    = &ax25_proto_ops;
 	sk->protocol = protocol;
-	sk->mtu      = AX25_MTU;	/* 256 */
 
 	ax25->sk          = sk;
 	sk->protinfo.ax25 = ax25;
@@ -892,7 +891,6 @@ struct sock *ax25_make_new(struct sock *osk, struct ax25_dev *ax25_dev)
 	sk->sndbuf   = osk->sndbuf;
 	sk->debug    = osk->debug;
 	sk->state    = TCP_ESTABLISHED;
-	sk->mtu      = osk->mtu;
 	sk->sleep    = osk->sleep;
 	sk->zapped   = osk->zapped;
 

net/core/datagram.c

@@ -54,15 +54,16 @@
 
 static inline void wait_for_packet(struct sock * sk)
 {
-	unsigned long flags;
+	struct wait_queue wait = { current, NULL };
+
+	add_wait_queue(sk->sleep, &wait);
+	current->state = TASK_INTERRUPTIBLE;
 
-	release_sock(sk);
-	save_flags(flags);
-	cli();
 	if (skb_peek(&sk->receive_queue) == NULL)
-		interruptible_sleep_on(sk->sleep);
-	restore_flags(flags);
-	lock_sock(sk);
+		schedule();
+
+	current->state = TASK_RUNNING;
+	remove_wait_queue(sk->sleep, &wait);
 }
 
 /*
@@ -84,6 +85,14 @@ static inline int connection_based(struct sock *sk)
  *	This function will lock the socket if a skb is returned, so the caller
  *	needs to unlock the socket in that case (usually by calling skb_free_datagram)
  *
+ *	* It does not lock socket since today. This function is
+ *	* free of race conditions. This measure should/can improve
+ *	* significantly datagram socket latencies at high loads,
+ *	* when data copying to user space takes lots of time.
+ *	* (BTW I've just killed the last cli() in IP/IPv6/core/netlink/packet
+ *	*  8) Great win.)
+ *	*			                    --ANK (980729)
+ *
  *	The order of the tests when we find no data waiting are specified
  *	quite explicitly by POSIX 1003.1g, don't change them without having
  *	the standard around please.
@@ -94,7 +103,6 @@ struct sk_buff *skb_recv_datagram(struct sock *sk, unsigned flags, int noblock,
 	int error;
 	struct sk_buff *skb;
 
-	lock_sock(sk);
 restart:
 	while(skb_queue_empty(&sk->receive_queue))	/* No data */
 	{
@@ -129,13 +137,24 @@ struct sk_buff *skb_recv_datagram(struct sock *sk, unsigned flags, int noblock,
 	   will suddenly eat the receive_queue */
 	if (flags & MSG_PEEK)
 	{
-		unsigned long flags;
-		save_flags(flags);
-		cli();
+		unsigned long cpu_flags;
+
+		/* It is the only POTENTIAL race condition
+		   in this function. skb may be stolen by
+		   another receiver after peek, but before
+		   incrementing use count, provided kernel
+		   is reentearble (it is not) or this function
+		   is called by interrupts.
+
+		   Protect it with global skb spinlock,
+		   though for now even this is overkill.
+		                                --ANK (980728)
+		 */
+		spin_lock_irqsave(&skb_queue_lock, cpu_flags);
 		skb = skb_peek(&sk->receive_queue);
 		if(skb!=NULL)
 			atomic_inc(&skb->users);
-		restore_flags(flags);
+		spin_unlock_irqrestore(&skb_queue_lock, cpu_flags);
 	} else
 		skb = skb_dequeue(&sk->receive_queue);
 
@@ -144,7 +163,6 @@ struct sk_buff *skb_recv_datagram(struct sock *sk, unsigned flags, int noblock,
 	return skb;
 
 no_packet:
-	release_sock(sk);
 	*err = error;
 	return NULL;
 }
@@ -152,7 +170,6 @@ struct sk_buff *skb_recv_datagram(struct sock *sk, unsigned flags, int noblock,
 void skb_free_datagram(struct sock * sk, struct sk_buff *skb)
 {
 	kfree_skb(skb);
-	release_sock(sk);
 }
 
 /*
@@ -184,6 +201,10 @@ int skb_copy_datagram_iovec(struct sk_buff *skb, int offset, struct iovec *to,
  *	Datagram poll: Again totally generic. This also handles
  *	sequenced packet sockets providing the socket receive queue
  *	is only ever holding data ready to receive.
+ *
+ *	Note: when you _don't_ use this routine for this protocol,
+ *	and you use a different write policy from sock_writeable()
+ *	then please supply your own write_space callback.
  */
 
 unsigned int datagram_poll(struct file * file, struct socket *sock, poll_table *wait)
@@ -199,7 +220,7 @@ unsigned int datagram_poll(struct file * file, struct socket *sock, poll_table *
 		mask |= POLLERR;
 	if (sk->shutdown & RCV_SHUTDOWN)
 		mask |= POLLHUP;
-
+	
 	/* readable? */
 	if (!skb_queue_empty(&sk->receive_queue))
 		mask |= POLLIN | POLLRDNORM;
@@ -214,15 +235,8 @@ unsigned int datagram_poll(struct file * file, struct socket *sock, poll_table *
 	}
 
 	/* writable? */
-	if (!(sk->shutdown & SEND_SHUTDOWN)) {
-		if (sk->prot) {
-			if (sock_wspace(sk) >= MIN_WRITE_SPACE)
-				mask |= POLLOUT | POLLWRNORM | POLLWRBAND;
-		} else {
-			if (sk->sndbuf - atomic_read(&sk->wmem_alloc) >= MIN_WRITE_SPACE)
-				mask |= POLLOUT | POLLWRNORM | POLLWRBAND;
-		}
-	}
+	if (sock_writeable(sk))
+		mask |= POLLOUT | POLLWRNORM | POLLWRBAND;
 
 	return mask;
 }

net/core/dev.c

@@ -16,6 +16,7 @@
  *		Alan Cox <gw4pts@gw4pts.ampr.org>
  *		David Hinds <dhinds@allegro.stanford.edu>
  *		Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
+ *		Adam Sulmicki <adam@cfar.umd.edu>
  *
  *	Changes:
  *		Alan Cox	:	device private ioctl copies fields back.
@@ -51,7 +52,10 @@
  *		Andi Kleen	:	Fix error reporting for SIOCGIFCONF
  *	    Michael Chastain	:	Fix signed/unsigned for SIOCGIFCONF
  *		Cyrus Durgin	:	Cleaned for KMOD
- *
+ *		Adam Sulmicki   :	Bug Fix : Network Device Unload
+ *					A network device unload needs to purge
+ *					the backlog queue.
+ *	Paul Rusty Russel	:	SIOCSIFNAME
  */
 
 #include <asm/uaccess.h>
@@ -154,6 +158,8 @@ int netdev_fastroute_obstacles;
 struct net_fastroute_stats dev_fastroute_stat;
 #endif
 
+static void dev_clear_backlog(struct device *dev);
+
 
 /******************************************************************************************
 
@@ -171,6 +177,16 @@ int netdev_nit=0;
  *	Add a protocol ID to the list. Now that the input handler is
  *	smarter we can dispense with all the messy stuff that used to be
  *	here.
+ *
+ *	BEWARE!!! Protocol handlers, mangling input packets,
+ *	MUST BE last in hash buckets and checking protocol handlers
+ *	MUST start from promiscous ptype_all chain in net_bh.
+ *	It is true now, do not change it.
+ *	Explantion follows: if protocol handler, mangling packet, will
+ *	be the first on list, it is not able to sense, that packet
+ *	is cloned and should be copied-on-write, so that it will
+ *	change it and subsequent readers will get broken packet.
+ *							--ANK (980803)
  */
  
 void dev_add_pack(struct packet_type *pt)
@@ -448,7 +464,8 @@ int dev_close(struct device *dev)
 	/*
 	 *	Device is now down.
 	 */
-	 
+	dev_clear_backlog(dev);
+
 	dev->flags&=~(IFF_UP|IFF_RUNNING);
 #ifdef CONFIG_NET_FASTROUTE
 	dev_clear_fastroute(dev);
@@ -457,7 +474,6 @@ int dev_close(struct device *dev)
 	/*
 	 *	Tell people we are going down
 	 */
-	 
 	notifier_call_chain(&netdev_chain, NETDEV_DOWN, dev);
 
 	return(0);
@@ -685,6 +701,45 @@ static void netdev_wakeup(void)
 }
 #endif
 
+static void dev_clear_backlog(struct device *dev)
+{
+	struct sk_buff *prev, *curr;
+
+	/*
+	 *
+	 *  Let now clear backlog queue. -AS
+	 *
+	 *  We are competing here both with netif_rx() and net_bh().
+	 *  We don't want either of those to mess with skb ptrs
+	 *  while we work on them, thus cli()/sti().
+	 *
+	 *  It looks better to use net_bh trick, at least
+	 *  to be sure, that we keep interrupt latency really low. --ANK (980727)
+	 */ 
+
+	if (backlog.qlen) {
+		start_bh_atomic();
+		curr = backlog.next;
+		while ( curr != (struct sk_buff *)(&backlog) ) {
+			unsigned long flags;
+			curr=curr->next;
+			if ( curr->prev->dev == dev ) {
+				prev = curr->prev;
+				spin_lock_irqsave(&skb_queue_lock, flags);
+				__skb_unlink(prev, &backlog);
+				spin_unlock_irqrestore(&skb_queue_lock, flags);
+				kfree_skb(prev);
+			}
+		}
+		end_bh_atomic();
+#ifdef CONFIG_NET_HW_FLOWCONTROL
+		if (netdev_dropping)
+			netdev_wakeup();
+#else
+		netdev_dropping = 0;
+#endif
+	}
+}
 
 /*
  *	Receive a packet from a device driver and queue it for the upper
@@ -1320,7 +1375,7 @@ int dev_change_flags(struct device *dev, unsigned flags)
 	 */
 
 	dev->flags = (flags & (IFF_DEBUG|IFF_NOTRAILERS|IFF_RUNNING|IFF_NOARP|
-			       IFF_SLAVE|IFF_MASTER|
+			       IFF_NODYNARP|IFF_SLAVE|IFF_MASTER|
 			       IFF_MULTICAST|IFF_PORTSEL|IFF_AUTOMEDIA)) |
 				       (dev->flags & (IFF_UP|IFF_VOLATILE|IFF_PROMISC|IFF_ALLMULTI));
 
@@ -1391,12 +1446,11 @@ static int dev_ifsioc(struct ifreq *ifr, unsigned int cmd)
 			return dev_change_flags(dev, ifr->ifr_flags);
 		
 		case SIOCGIFMETRIC:	/* Get the metric on the interface (currently unused) */
-			ifr->ifr_metric = dev->metric;
+			ifr->ifr_metric = 0;
 			return 0;
 			
 		case SIOCSIFMETRIC:	/* Set the metric on the interface (currently unused) */
-			dev->metric = ifr->ifr_metric;
-			return 0;
+			return -EOPNOTSUPP;
 	
 		case SIOCGIFMTU:	/* Get the MTU of a device */
 			ifr->ifr_mtu = dev->mtu;
@@ -1419,10 +1473,8 @@ static int dev_ifsioc(struct ifreq *ifr, unsigned int cmd)
 				dev->mtu = ifr->ifr_mtu;
 				err = 0;
 			}
-			if (!err && dev->flags&IFF_UP) {
-				printk(KERN_DEBUG "SIFMTU %s(%s)\n", dev->name, current->comm);
+			if (!err && dev->flags&IFF_UP)
 				notifier_call_chain(&netdev_chain, NETDEV_CHANGEMTU, dev);
-			}
 			return err;
 
 		case SIOCGIFHWADDR:
@@ -1484,11 +1536,22 @@ static int dev_ifsioc(struct ifreq *ifr, unsigned int cmd)
 			return 0;
 
 		case SIOCSIFTXQLEN:
-			if(ifr->ifr_qlen<2 || ifr->ifr_qlen>1024)
+			/* Why <2? 0 and 1 are valid values. --ANK (980807) */
+			if(/*ifr->ifr_qlen<2 ||*/ ifr->ifr_qlen>1024)
 				return -EINVAL;
 			dev->tx_queue_len = ifr->ifr_qlen;
 			return 0;
 
+		case SIOCSIFNAME:
+			if (dev->flags&IFF_UP)
+				return -EBUSY;
+			if (dev_get(ifr->ifr_newname))
+				return -EEXIST;
+			memcpy(dev->name, ifr->ifr_newname, IFNAMSIZ);
+			dev->name[IFNAMSIZ-1] = 0;
+			notifier_call_chain(&netdev_chain, NETDEV_CHANGENAME, dev);
+			return 0;
+
 		/*
 		 *	Unknown or private ioctl
 		 */
@@ -1597,6 +1660,7 @@ int dev_ioctl(unsigned int cmd, void *arg)
 		case SIOCDELMULTI:
 		case SIOCSIFHWBROADCAST:
 		case SIOCSIFTXQLEN:
+		case SIOCSIFNAME:
 			if (!capable(CAP_NET_ADMIN))
 				return -EPERM;
 			dev_load(ifr.ifr_name);
@@ -1668,6 +1732,17 @@ int register_netdevice(struct device *dev)
 	struct device *d, **dp;
 
 	if (dev_boot_phase) {
+		/* This is NOT bug, but I am not sure, that all the
+		   devices, initialized before netdev module is started
+		   are sane. 
+
+		   Now they are chained to device boot list
+		   and probed later. If a module is initialized
+		   before netdev, but assumes that dev->init
+		   is really called by register_netdev(), it will fail.
+
+		   So that this message should be printed for a while.
+		 */
 		printk(KERN_INFO "early initialization of device %s is deferred\n", dev->name);
 
 		/* Check for existence, and append to tail of chain */

net/core/iovec.c

@@ -215,7 +215,7 @@ int csum_partial_copy_fromiovecend(unsigned char *kdata, struct iovec *iov,
 			partial_cnt = 0;
 		}
 
-		if (len - copy > 0)
+		if (len > copy)
 		{
 			partial_cnt = copy % 4;
 			if (partial_cnt)

net/core/neighbour.c

@@ -9,6 +9,9 @@
  *      modify it under the terms of the GNU General Public License
  *      as published by the Free Software Foundation; either version
  *      2 of the License, or (at your option) any later version.
+ *
+ *	Fixes:
+ *	Vitaly E. Lavrov	releasing NULL neighbor in neigh_add.
  */
 
 #include <linux/config.h>
@@ -1033,7 +1036,8 @@ int neigh_add(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 					   ndm->ndm_state,
 					   nlh->nlmsg_flags&NLM_F_REPLACE, 0);
 		}
-		neigh_release(n);
+		if (n)
+			neigh_release(n);
 		end_bh_atomic();
 		return err;
 	}
@@ -1043,7 +1047,7 @@ int neigh_add(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 
 
 static int neigh_fill_info(struct sk_buff *skb, struct neighbour *n,
-			    pid_t pid, u32 seq, int event)
+			   u32 pid, u32 seq, int event)
 {
 	unsigned long now = jiffies;
 	struct ndmsg *ndm;

net/core/rtnetlink.c

@@ -12,6 +12,8 @@
  *		as published by the Free Software Foundation; either version
  *		2 of the License, or (at your option) any later version.
  *
+ *	Fixes:
+ *	Vitaly E. Lavrov		RTA_OK arithmetics was wrong.
  */
 
 #include <linux/config.h>
@@ -29,6 +31,7 @@
 #include <linux/mm.h>
 #include <linux/slab.h>
 #include <linux/interrupt.h>
+#include <linux/capability.h>
 #include <linux/skbuff.h>
 #include <linux/init.h>
 
@@ -135,47 +138,8 @@ int rtnetlink_send(struct sk_buff *skb, u32 pid, unsigned group, int echo)
 	return err;
 }
 
-#ifdef CONFIG_RTNL_OLD_IFINFO
 static int rtnetlink_fill_ifinfo(struct sk_buff *skb, struct device *dev,
-				 int type, pid_t pid, u32 seq)
-{
-	struct ifinfomsg *r;
-	struct nlmsghdr  *nlh;
-	unsigned char	 *b = skb->tail;
-
-	nlh = NLMSG_PUT(skb, pid, seq, type, sizeof(*r));
-	if (pid) nlh->nlmsg_flags |= NLM_F_MULTI;
-	r = NLMSG_DATA(nlh);
-	r->ifi_addrlen = dev->addr_len;
-	r->ifi_address.sa_family = dev->type;
-	memcpy(&r->ifi_address.sa_data, dev->dev_addr, dev->addr_len);
-	r->ifi_broadcast.sa_family = dev->type;
-	memcpy(&r->ifi_broadcast.sa_data, dev->broadcast, dev->addr_len);
-	r->ifi_flags = dev->flags;
-	r->ifi_mtu = dev->mtu;
-	r->ifi_index = dev->ifindex;
-	r->ifi_link = dev->iflink;
-	strncpy(r->ifi_name, dev->name, IFNAMSIZ-1);
-	r->ifi_qdiscname[0] = 0;
-	r->ifi_qdisc = dev->qdisc_sleeping->handle;
-	if (dev->qdisc_sleeping)
-		strcpy(r->ifi_qdiscname, dev->qdisc_sleeping->ops->id);
-	if (dev->get_stats) {
-		struct net_device_stats *stats = dev->get_stats(dev);
-		if (stats)
-			RTA_PUT(skb, IFLA_STATS, sizeof(*stats), stats);
-	}
-	nlh->nlmsg_len = skb->tail - b;
-	return skb->len;
-
-nlmsg_failure:
-rtattr_failure:
-	skb_trim(skb, b - skb->data);
-	return -1;
-}
-#else
-static int rtnetlink_fill_ifinfo(struct sk_buff *skb, struct device *dev,
-				 int type, pid_t pid, u32 seq)
+				 int type, u32 pid, u32 seq)
 {
 	struct ifinfomsg *r;
 	struct nlmsghdr  *nlh;
@@ -218,7 +182,6 @@ static int rtnetlink_fill_ifinfo(struct sk_buff *skb, struct device *dev,
 	skb_trim(skb, b - skb->data);
 	return -1;
 }
-#endif
 
 int rtnetlink_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
 {
@@ -266,12 +229,7 @@ int rtnetlink_dump_all(struct sk_buff *skb, struct netlink_callback *cb)
 void rtmsg_ifinfo(int type, struct device *dev)
 {
 	struct sk_buff *skb;
-#ifdef CONFIG_RTNL_OLD_IFINFO
-	int size = NLMSG_SPACE(sizeof(struct ifinfomsg)+
-			       RTA_LENGTH(sizeof(struct net_device_stats)));
-#else
 	int size = NLMSG_GOODSIZE;
-#endif
 
 	skb = alloc_skb(size, GFP_KERNEL);
 	if (!skb)
@@ -287,7 +245,7 @@ void rtmsg_ifinfo(int type, struct device *dev)
 
 static int rtnetlink_done(struct netlink_callback *cb)
 {
-	if (NETLINK_CREDS(cb->skb)->uid == 0 && cb->nlh->nlmsg_flags&NLM_F_ATOMIC)
+	if (cap_raised(NETLINK_CB(cb->skb).eff_cap, CAP_NET_ADMIN) && cb->nlh->nlmsg_flags&NLM_F_ATOMIC)
 		rtnl_shunlock();
 	return 0;
 }
@@ -342,13 +300,13 @@ rtnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *errp)
 	sz_idx = type>>2;
 	kind = type&3;
 
-	if (kind != 2 && NETLINK_CREDS(skb)->uid) {
+	if (kind != 2 && !cap_raised(NETLINK_CB(skb).eff_cap, CAP_NET_ADMIN)) {
 		*errp = -EPERM;
 		return -1;
 	}
 
 	if (kind == 2 && nlh->nlmsg_flags&NLM_F_DUMP) {
-		int rlen;
+		u32 rlen;
 
 		if (link->dumpit == NULL)
 			link = &(rtnetlink_links[PF_UNSPEC][type]);
@@ -357,12 +315,13 @@ rtnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *errp)
 			goto err_inval;
 
 		/* Super-user locks all the tables to get atomic snapshot */
-		if (NETLINK_CREDS(skb)->uid == 0 && nlh->nlmsg_flags&NLM_F_ATOMIC)
+		if (cap_raised(NETLINK_CB(skb).eff_cap, CAP_NET_ADMIN)
+		    && nlh->nlmsg_flags&NLM_F_ATOMIC)
 			atomic_inc(&rtnl_rlockct);
 		if ((*errp = netlink_dump_start(rtnl, skb, nlh,
 						link->dumpit,
 						rtnetlink_done)) != 0) {
-			if (NETLINK_CREDS(skb)->uid == 0 && nlh->nlmsg_flags&NLM_F_ATOMIC)
+			if (cap_raised(NETLINK_CB(skb).eff_cap, CAP_NET_ADMIN) && nlh->nlmsg_flags&NLM_F_ATOMIC)
 				atomic_dec(&rtnl_rlockct);
 			return -1;
 		}
@@ -431,7 +390,7 @@ extern __inline__ int rtnetlink_rcv_skb(struct sk_buff *skb)
 	struct nlmsghdr * nlh;
 
 	while (skb->len >= NLMSG_SPACE(0)) {
-		int rlen;
+		u32 rlen;
 
 		nlh = (struct nlmsghdr *)skb->data;
 		if (nlh->nlmsg_len < sizeof(*nlh) || skb->len < nlh->nlmsg_len)

net/core/scm.c

@@ -138,11 +138,15 @@ int __scm_send(struct socket *sock, struct msghdr *msg, struct scm_cookie *p)
 
 	for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg))
 	{
+		err = -EINVAL;
+
+		if ((unsigned long)(((char*)cmsg - (char*)msg->msg_control)
+				    + cmsg->cmsg_len) > msg->msg_controllen)
+			goto error;
+
 		if (cmsg->cmsg_level != SOL_SOCKET)
 			continue;
 
-		err = -EINVAL;
-
 		switch (cmsg->cmsg_type)
 		{
 		case SCM_RIGHTS:

net/core/skbuff.c

@@ -4,6 +4,8 @@
  *	Authors:	Alan Cox <iiitac@pyr.swan.ac.uk>
  *			Florian La Roche <rzsfl@rz.uni-sb.de>
  *
+ *	Version:	$Id: skbuff.c,v 1.53 1998/08/19 13:32:44 freitag Exp $
+ *
  *	Fixes:	
  *		Alan Cox	:	Fixed the worst of the load balancer bugs.
  *		Dave Platt	:	Interrupt stacking fix.
@@ -96,14 +98,14 @@ void skb_under_panic(struct sk_buff *skb, int sz, void *here)
 
 void show_net_buffers(void)
 {
-	printk(KERN_INFO "Networking buffers in use          : %u\n",
+	printk("Networking buffers in use          : %u\n",
 	       atomic_read(&net_skbcount));
-	printk(KERN_INFO "Total network buffer allocations   : %u\n",
+	printk("Total network buffer allocations   : %u\n",
 	       atomic_read(&net_allocs));
-	printk(KERN_INFO "Total failed network buffer allocs : %u\n",
+	printk("Total failed network buffer allocs : %u\n",
 	       atomic_read(&net_fails));
 #ifdef CONFIG_INET
-	printk(KERN_INFO "IP fragment buffer size            : %u\n",
+	printk("IP fragment buffer size            : %u\n",
 	       atomic_read(&ip_frag_mem));
 #endif	
 }
@@ -365,7 +367,7 @@ void skb_add_mtu(int mtu)
 }
 #endif
 
-__initfunc(void skb_init(void))
+void __init skb_init(void)
 {
 	skbuff_head_cache = kmem_cache_create("skbuff_head_cache",
 					      sizeof(struct sk_buff),

net/core/sock.c

@@ -7,7 +7,7 @@
  *		handler for protocols to use and generic option handler.
  *
  *
- * Version:	@(#)sock.c	1.0.17	06/02/93
+ * Version:	$Id: sock.c,v 1.70 1998/08/26 12:03:07 davem Exp $
  *
  * Authors:	Ross Biro, <bir7@leland.Stanford.Edu>
  *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
@@ -78,6 +78,7 @@
  *              Chris Evans     :       Call suser() check last on F_SETOWN
  *		Jay Schulist	:	Added SO_ATTACH_FILTER and SO_DETACH_FILTER.
  *		Andi Kleen	:	Add sock_kmalloc()/sock_kfree_s()
+ *		Andi Kleen	:	Fix write_space callback
  *
  * To Fix:
  *
@@ -445,6 +446,7 @@ int sock_getsockopt(struct socket *sock, int level, int optname,
 		case SO_RCVLOWAT:
 		case SO_SNDLOWAT:
 			v.val=1;
+			break; 
 
 		case SO_PASSCRED:
 			v.val = sock->passcred;
@@ -615,19 +617,6 @@ unsigned long sock_rspace(struct sock *sk)
 }
 
 
-/* FIXME: this is also insane. See above comment */
-unsigned long sock_wspace(struct sock *sk)
-{
-	int amt = 0;
-
-	if (sk != NULL && !(sk->shutdown & SEND_SHUTDOWN)) {
-		amt = sk->sndbuf - atomic_read(&sk->wmem_alloc);
-		if (amt < 0) 
-			amt = 0;
-	}
-	return amt;
-}
-
 /* It is almost wait_for_tcp_memory minus release_sock/lock_sock.
    I think, these locks should be removed for datagram sockets.
  */
@@ -746,17 +735,15 @@ void __release_sock(struct sock *sk)
  
 void sklist_remove_socket(struct sock **list, struct sock *sk)
 {
-	unsigned long flags;
 	struct sock *s;
 
-	save_flags(flags);
-	cli();
+	start_bh_atomic();
 
 	s= *list;
 	if(s==sk)
 	{
 		*list = s->next;
-		restore_flags(flags);
+		end_bh_atomic();
 		return;
 	}
 	while(s && s->next)
@@ -764,22 +751,19 @@ void sklist_remove_socket(struct sock **list, struct sock *sk)
 		if(s->next==sk)
 		{
 			s->next=sk->next;
-			restore_flags(flags);
-			return;
+			break;
 		}
 		s=s->next;
 	}
-	restore_flags(flags);
+	end_bh_atomic();
 }
 
 void sklist_insert_socket(struct sock **list, struct sock *sk)
 {
-	unsigned long flags;
-	save_flags(flags);
-	cli();
+	start_bh_atomic();
 	sk->next= *list;
 	*list=sk;
-	restore_flags(flags);
+	end_bh_atomic();
 }
 
 /*
@@ -914,6 +898,10 @@ int sock_no_getsockopt(struct socket *sock, int level, int optname,
 	return -EOPNOTSUPP;
 }
 
+/* 
+ * Note: if you add something that sleeps here then change sock_fcntl()
+ *       to do proper fd locking.
+ */
 int sock_no_fcntl(struct socket *sock, unsigned int cmd, unsigned long arg)
 {
 	struct sock *sk = sock->sk;
@@ -971,12 +959,15 @@ void sock_def_callback2(struct sock *sk, int len)
 	}
 }
 
-void sock_def_callback3(struct sock *sk)
+void sock_def_write_space(struct sock *sk)
 {
 	if(!sk->dead)
 	{
 		wake_up_interruptible(sk->sleep);
-		sock_wake_async(sk->socket, 2);
+
+		/* Should agree with poll, otherwise some programs break */ 
+		if (sock_writeable(sk))
+			sock_wake_async(sk->socket, 2);
 	}
 }
 
@@ -1011,7 +1002,7 @@ void sock_init_data(struct socket *sock, struct sock *sk)
 
 	sk->state_change	=	sock_def_callback1;
 	sk->data_ready		=	sock_def_callback2;
-	sk->write_space		=	sock_def_callback3;
+	sk->write_space		=	sock_def_write_space;
 	sk->error_report	=	sock_def_callback1;
 	sk->destruct            =       sock_def_destruct;
 

net/ipv4/devinet.c

 /*
  *	NET3	IP device support routines.
  *
- *	Version: $Id: devinet.c,v 1.22 1998/05/08 21:06:26 davem Exp $
+ *	Version: $Id: devinet.c,v 1.23 1998/08/26 12:03:21 davem Exp $
  *
  *		This program is free software; you can redistribute it and/or
  *		modify it under the terms of the GNU General Public License
@@ -533,8 +533,6 @@ int devinet_ioctl(unsigned int cmd, void *arg)
 				inet_del_ifa(in_dev, ifap, 0);
 				ifa->ifa_broadcast = 0;
 				ifa->ifa_anycast = 0;
-				ifa->ifa_prefixlen = 32;
-				ifa->ifa_mask = inet_make_mask(32);
 			}
 
 			ifa->ifa_address =
@@ -545,6 +543,9 @@ int devinet_ioctl(unsigned int cmd, void *arg)
 				ifa->ifa_mask = inet_make_mask(ifa->ifa_prefixlen);
 				if ((dev->flags&IFF_BROADCAST) && ifa->ifa_prefixlen < 31)
 					ifa->ifa_broadcast = ifa->ifa_address|~ifa->ifa_mask;
+			} else {
+				ifa->ifa_prefixlen = 32;
+				ifa->ifa_mask = inet_make_mask(32);
 			}
 			ret = inet_set_ifa(dev, ifa);
 			break;
@@ -702,6 +703,16 @@ static int inetdev_event(struct notifier_block *this, unsigned long event, void
 	case NETDEV_UNREGISTER:
 		inetdev_destroy(in_dev);
 		break;
+	case NETDEV_CHANGENAME:
+		if (in_dev->ifa_list) {
+			struct in_ifaddr *ifa;
+			for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next)
+				memcpy(ifa->ifa_label, dev->name, IFNAMSIZ);
+			/* Do not notify about label change, this event is
+			   not interesting to applications using netlink.
+			 */
+		}
+		break;
 	}
 
 	return NOTIFY_DONE;
@@ -716,7 +727,7 @@ struct notifier_block ip_netdev_notifier={
 #ifdef CONFIG_RTNETLINK
 
 static int inet_fill_ifaddr(struct sk_buff *skb, struct in_ifaddr *ifa,
-			    pid_t pid, u32 seq, int event)
+			    u32 pid, u32 seq, int event)
 {
 	struct ifaddrmsg *ifm;
 	struct nlmsghdr  *nlh;
@@ -729,7 +740,7 @@ static int inet_fill_ifaddr(struct sk_buff *skb, struct in_ifaddr *ifa,
 	ifm->ifa_flags = ifa->ifa_flags|IFA_F_PERMANENT;
 	ifm->ifa_scope = ifa->ifa_scope;
 	ifm->ifa_index = ifa->ifa_dev->dev->ifindex;
-	if (ifa->ifa_prefixlen)
+	if (ifa->ifa_address)
 		RTA_PUT(skb, IFA_ADDRESS, 4, &ifa->ifa_address);
 	if (ifa->ifa_local)
 		RTA_PUT(skb, IFA_LOCAL, 4, &ifa->ifa_local);