Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
David S. Miller [Tue, 17 Nov 2009 08:05:02 +0000 (00:05 -0800)]
Conflicts:
drivers/net/can/Kconfig

32 files changed:
MAINTAINERS
drivers/isdn/hardware/mISDN/hfcmulti.c
drivers/isdn/i4l/isdn_ppp.c
drivers/net/can/Kconfig
drivers/net/can/dev.c
drivers/net/can/sja1000/Kconfig [new file with mode: 0644]
drivers/net/can/usb/Kconfig [new file with mode: 0644]
drivers/net/can/usb/Makefile
drivers/net/davinci_emac.c
drivers/net/forcedeth.c
drivers/net/ixgbe/ixgbe_main.c
drivers/net/ppp_generic.c
drivers/net/r6040.c
drivers/net/r8169.c
drivers/net/s2io.c
drivers/net/smsc911x.c
drivers/net/wireless/ipw2x00/ipw2100.c
drivers/net/wireless/ipw2x00/ipw2200.c
drivers/net/wireless/ipw2x00/libipw.h
drivers/net/wireless/ipw2x00/libipw_module.c
drivers/net/wireless/p54/p54usb.c
include/linux/isdn_ppp.h
include/net/sctp/structs.h
net/bluetooth/hci_conn.c
net/bluetooth/l2cap.c
net/core/dev.c
net/ipv4/ipmr.c
net/ipv4/tcp.c
net/sctp/associola.c
net/sctp/sm_statefuns.c
net/sctp/socket.c
net/sctp/transport.c

index fe1cd41..b2e3f35 100644 (file)
@@ -4304,6 +4304,7 @@ F:        drivers/video/aty/aty128fb.c
 RALINK RT2X00 WIRELESS LAN DRIVER
 P:     rt2x00 project
 M:     Ivo van Doorn <IvDoorn@gmail.com>
+M:     Gertjan van Wingerde <gwingerde@gmail.com>
 L:     linux-wireless@vger.kernel.org
 L:     users@rt2x00.serialmonkey.com (moderated for non-subscribers)
 W:     http://rt2x00.serialmonkey.com/
@@ -4391,7 +4392,7 @@ RFKILL
 M:     Johannes Berg <johannes@sipsolutions.net>
 L:     linux-wireless@vger.kernel.org
 S:     Maintained
-F      Documentation/rfkill.txt
+F:     Documentation/rfkill.txt
 F:     net/rfkill/
 
 RISCOM8 DRIVER
index faed794..a6624ad 100644 (file)
@@ -5481,7 +5481,7 @@ HFCmulti_init(void)
                if (err) {
                        printk(KERN_ERR "error registering embedded driver: "
                                "%x\n", err);
-                       return -err;
+                       return err;
                }
                HFC_cnt++;
                printk(KERN_INFO "%d devices registered\n", HFC_cnt);
index 2d14b64..642d5aa 100644 (file)
@@ -1535,10 +1535,8 @@ static int isdn_ppp_mp_bundle_array_init(void)
        int sz = ISDN_MAX_CHANNELS*sizeof(ippp_bundle);
        if( (isdn_ppp_bundle_arr = kzalloc(sz, GFP_KERNEL)) == NULL )
                return -ENOMEM;
-       for (i = 0; i < ISDN_MAX_CHANNELS; i++) {
+       for( i = 0; i < ISDN_MAX_CHANNELS; i++ )
                spin_lock_init(&isdn_ppp_bundle_arr[i].lock);
-               skb_queue_head_init(&isdn_ppp_bundle_arr[i].frags);
-       }
        return 0;
 }
 
@@ -1571,7 +1569,7 @@ static int isdn_ppp_mp_init( isdn_net_local * lp, ippp_bundle * add_to )
                if ((lp->netdev->pb = isdn_ppp_mp_bundle_alloc()) == NULL)
                        return -ENOMEM;
                lp->next = lp->last = lp;       /* nobody else in a queue */
-               skb_queue_head_init(&lp->netdev->pb->frags);
+               lp->netdev->pb->frags = NULL;
                lp->netdev->pb->frames = 0;
                lp->netdev->pb->seq = UINT_MAX;
        }
@@ -1583,29 +1581,28 @@ static int isdn_ppp_mp_init( isdn_net_local * lp, ippp_bundle * add_to )
 
 static u32 isdn_ppp_mp_get_seq( int short_seq, 
                                        struct sk_buff * skb, u32 last_seq );
-static void isdn_ppp_mp_discard(ippp_bundle *mp, struct sk_buff *from,
-                               struct sk_buff *to);
-static void isdn_ppp_mp_reassembly(isdn_net_dev *net_dev, isdn_net_local *lp,
-                                  struct sk_buff *from, struct sk_buff *to,
-                                  u32 lastseq);
-static void isdn_ppp_mp_free_skb(ippp_bundle *mp, struct sk_buff *skb);
+static struct sk_buff * isdn_ppp_mp_discard( ippp_bundle * mp,
+                       struct sk_buff * from, struct sk_buff * to );
+static void isdn_ppp_mp_reassembly( isdn_net_dev * net_dev, isdn_net_local * lp,
+                               struct sk_buff * from, struct sk_buff * to );
+static void isdn_ppp_mp_free_skb( ippp_bundle * mp, struct sk_buff * skb );
 static void isdn_ppp_mp_print_recv_pkt( int slot, struct sk_buff * skb );
 
 static void isdn_ppp_mp_receive(isdn_net_dev * net_dev, isdn_net_local * lp, 
-                               struct sk_buff *skb)
+                                                       struct sk_buff *skb)
 {
-       struct sk_buff *newfrag, *frag, *start, *nextf;
-       u32 newseq, minseq, thisseq;
-       isdn_mppp_stats *stats;
        struct ippp_struct *is;
+       isdn_net_local * lpq;
+       ippp_bundle * mp;
+       isdn_mppp_stats * stats;
+       struct sk_buff * newfrag, * frag, * start, *nextf;
+       u32 newseq, minseq, thisseq;
        unsigned long flags;
-       isdn_net_local *lpq;
-       ippp_bundle *mp;
        int slot;
 
        spin_lock_irqsave(&net_dev->pb->lock, flags);
-       mp = net_dev->pb;
-       stats = &mp->stats;
+       mp = net_dev->pb;
+        stats = &mp->stats;
        slot = lp->ppp_slot;
        if (slot < 0 || slot >= ISDN_MAX_CHANNELS) {
                printk(KERN_ERR "%s: lp->ppp_slot(%d)\n",
@@ -1616,19 +1613,20 @@ static void isdn_ppp_mp_receive(isdn_net_dev * net_dev, isdn_net_local * lp,
                return;
        }
        is = ippp_table[slot];
-       if (++mp->frames > stats->max_queue_len)
+       if( ++mp->frames > stats->max_queue_len )
                stats->max_queue_len = mp->frames;
-
+       
        if (is->debug & 0x8)
                isdn_ppp_mp_print_recv_pkt(lp->ppp_slot, skb);
 
-       newseq = isdn_ppp_mp_get_seq(is->mpppcfg & SC_IN_SHORT_SEQ,
-                                    skb, is->last_link_seqno);
+       newseq = isdn_ppp_mp_get_seq(is->mpppcfg & SC_IN_SHORT_SEQ, 
+                                               skb, is->last_link_seqno);
+
 
        /* if this packet seq # is less than last already processed one,
         * toss it right away, but check for sequence start case first 
         */
-       if (mp->seq > MP_LONGSEQ_MAX && (newseq & MP_LONGSEQ_MAXBIT)) {
+       if( mp->seq > MP_LONGSEQ_MAX && (newseq & MP_LONGSEQ_MAXBIT) ) {
                mp->seq = newseq;       /* the first packet: required for
                                         * rfc1990 non-compliant clients --
                                         * prevents constant packet toss */
@@ -1638,7 +1636,7 @@ static void isdn_ppp_mp_receive(isdn_net_dev * net_dev, isdn_net_local * lp,
                spin_unlock_irqrestore(&mp->lock, flags);
                return;
        }
-
+       
        /* find the minimum received sequence number over all links */
        is->last_link_seqno = minseq = newseq;
        for (lpq = net_dev->queue;;) {
@@ -1659,31 +1657,22 @@ static void isdn_ppp_mp_receive(isdn_net_dev * net_dev, isdn_net_local * lp,
                                         * packets */
        newfrag = skb;
 
-       /* Insert new fragment into the proper sequence slot.  */
-       skb_queue_walk(&mp->frags, frag) {
-               if (MP_SEQ(frag) == newseq) {
-                       isdn_ppp_mp_free_skb(mp, newfrag);
-                       newfrag = NULL;
-                       break;
-               }
-               if (MP_LT(newseq, MP_SEQ(frag))) {
-                       __skb_queue_before(&mp->frags, frag, newfrag);
-                       newfrag = NULL;
-                       break;
-               }
-       }
-       if (newfrag)
-               __skb_queue_tail(&mp->frags, newfrag);
+       /* if this new fragment is before the first one, then enqueue it now. */
+       if ((frag = mp->frags) == NULL || MP_LT(newseq, MP_SEQ(frag))) {
+               newfrag->next = frag;
+               mp->frags = frag = newfrag;
+               newfrag = NULL;
+       }
 
-       frag = skb_peek(&mp->frags);
-       start = ((MP_FLAGS(frag) & MP_BEGIN_FRAG) &&
-                (MP_SEQ(frag) == mp->seq)) ? frag : NULL;
-       if (!start)
-               goto check_overflow;
+       start = MP_FLAGS(frag) & MP_BEGIN_FRAG &&
+                               MP_SEQ(frag) == mp->seq ? frag : NULL;
 
-       /* main fragment traversing loop
+       /* 
+        * main fragment traversing loop
         *
         * try to accomplish several tasks:
+        * - insert new fragment into the proper sequence slot (once that's done
+        *   newfrag will be set to NULL)
         * - reassemble any complete fragment sequence (non-null 'start'
         *   indicates there is a continguous sequence present)
         * - discard any incomplete sequences that are below minseq -- due
@@ -1692,46 +1681,71 @@ static void isdn_ppp_mp_receive(isdn_net_dev * net_dev, isdn_net_local * lp,
         *   come to complete such sequence and it should be discarded
         *
         * loop completes when we accomplished the following tasks:
+        * - new fragment is inserted in the proper sequence ('newfrag' is 
+        *   set to NULL)
         * - we hit a gap in the sequence, so no reassembly/processing is 
         *   possible ('start' would be set to NULL)
         *
         * algorithm for this code is derived from code in the book
         * 'PPP Design And Debugging' by James Carlson (Addison-Wesley)
         */
-       skb_queue_walk_safe(&mp->frags, frag, nextf) {
-               thisseq = MP_SEQ(frag);
-
-               /* check for misplaced start */
-               if (start != frag && (MP_FLAGS(frag) & MP_BEGIN_FRAG)) {
-                       printk(KERN_WARNING"isdn_mppp(seq %d): new "
-                              "BEGIN flag with no prior END", thisseq);
-                       stats->seqerrs++;
-                       stats->frame_drops++;
-                       isdn_ppp_mp_discard(mp, start, frag);
-                       start = frag;
-               } else if (MP_LE(thisseq, minseq)) {            
-                       if (MP_FLAGS(frag) & MP_BEGIN_FRAG)
+       while (start != NULL || newfrag != NULL) {
+
+               thisseq = MP_SEQ(frag);
+               nextf = frag->next;
+
+               /* drop any duplicate fragments */
+               if (newfrag != NULL && thisseq == newseq) {
+                       isdn_ppp_mp_free_skb(mp, newfrag);
+                       newfrag = NULL;
+               }
+
+               /* insert new fragment before next element if possible. */
+               if (newfrag != NULL && (nextf == NULL || 
+                                               MP_LT(newseq, MP_SEQ(nextf)))) {
+                       newfrag->next = nextf;
+                       frag->next = nextf = newfrag;
+                       newfrag = NULL;
+               }
+
+               if (start != NULL) {
+                       /* check for misplaced start */
+                       if (start != frag && (MP_FLAGS(frag) & MP_BEGIN_FRAG)) {
+                               printk(KERN_WARNING"isdn_mppp(seq %d): new "
+                                     "BEGIN flag with no prior END", thisseq);
+                               stats->seqerrs++;
+                               stats->frame_drops++;
+                               start = isdn_ppp_mp_discard(mp, start,frag);
+                               nextf = frag->next;
+                       }
+               } else if (MP_LE(thisseq, minseq)) {            
+                       if (MP_FLAGS(frag) & MP_BEGIN_FRAG)
                                start = frag;
-                       else {
+                       else {
                                if (MP_FLAGS(frag) & MP_END_FRAG)
-                                       stats->frame_drops++;
-                               __skb_unlink(skb, &mp->frags);
+                                       stats->frame_drops++;
+                               if( mp->frags == frag )
+                                       mp->frags = nextf;      
                                isdn_ppp_mp_free_skb(mp, frag);
+                               frag = nextf;
                                continue;
-                       }
+                       }
                }
-
-               /* if we have end fragment, then we have full reassembly
-                * sequence -- reassemble and process packet now
+               
+               /* if start is non-null and we have end fragment, then
+                * we have full reassembly sequence -- reassemble 
+                * and process packet now
                 */
-               if (MP_FLAGS(frag) & MP_END_FRAG) {
-                       minseq = mp->seq = (thisseq+1) & MP_LONGSEQ_MASK;
-                       /* Reassemble the packet then dispatch it */
-                       isdn_ppp_mp_reassembly(net_dev, lp, start, frag, thisseq);
+               if (start != NULL && (MP_FLAGS(frag) & MP_END_FRAG)) {
+                       minseq = mp->seq = (thisseq+1) & MP_LONGSEQ_MASK;
+                       /* Reassemble the packet then dispatch it */
+                       isdn_ppp_mp_reassembly(net_dev, lp, start, nextf);
+      
+                       start = NULL;
+                       frag = NULL;
 
-                       start = NULL;
-                       frag = NULL;
-               }
+                       mp->frags = nextf;
+               }
 
                /* check if need to update start pointer: if we just
                 * reassembled the packet and sequence is contiguous
@@ -1742,25 +1756,26 @@ static void isdn_ppp_mp_receive(isdn_net_dev * net_dev, isdn_net_local * lp,
                 * below low watermark and set start to the next frag or
                 * clear start ptr.
                 */ 
-               if (nextf != (struct sk_buff *)&mp->frags && 
+               if (nextf != NULL && 
                    ((thisseq+1) & MP_LONGSEQ_MASK) == MP_SEQ(nextf)) {
-                       /* if we just reassembled and the next one is here, 
-                        * then start another reassembly.
-                        */
-                       if (frag == NULL) {
+                       /* if we just reassembled and the next one is here, 
+                        * then start another reassembly. */
+
+                       if (frag == NULL) {
                                if (MP_FLAGS(nextf) & MP_BEGIN_FRAG)
-                                       start = nextf;
-                               else {
-                                       printk(KERN_WARNING"isdn_mppp(seq %d):"
-                                              " END flag with no following "
-                                              "BEGIN", thisseq);
+                                       start = nextf;
+                               else
+                               {
+                                       printk(KERN_WARNING"isdn_mppp(seq %d):"
+                                               " END flag with no following "
+                                               "BEGIN", thisseq);
                                        stats->seqerrs++;
                                }
                        }
-               } else {
-                       if (nextf != (struct sk_buff *)&mp->frags &&
-                           frag != NULL &&
-                           MP_LT(thisseq, minseq)) {
+
+               } else {
+                       if ( nextf != NULL && frag != NULL &&
+                                               MP_LT(thisseq, minseq)) {
                                /* we've got a break in the sequence
                                 * and we not at the end yet
                                 * and we did not just reassembled
@@ -1769,39 +1784,41 @@ static void isdn_ppp_mp_receive(isdn_net_dev * net_dev, isdn_net_local * lp,
                                 * discard all the frames below low watermark 
                                 * and start over */
                                stats->frame_drops++;
-                               isdn_ppp_mp_discard(mp, start, nextf);
+                               mp->frags = isdn_ppp_mp_discard(mp,start,nextf);
                        }
                        /* break in the sequence, no reassembly */
-                       start = NULL;
-               }
-               if (!start)
-                       break;
-       }
-
-check_overflow:
+                       start = NULL;
+               }
+                               
+               frag = nextf;
+       }       /* while -- main loop */
+       
+       if (mp->frags == NULL)
+               mp->frags = frag;
+               
        /* rather straighforward way to deal with (not very) possible 
-        * queue overflow
-        */
+        * queue overflow */
        if (mp->frames > MP_MAX_QUEUE_LEN) {
                stats->overflows++;
-               skb_queue_walk_safe(&mp->frags, frag, nextf) {
-                       if (mp->frames <= MP_MAX_QUEUE_LEN)
-                               break;
-                       __skb_unlink(frag, &mp->frags);
-                       isdn_ppp_mp_free_skb(mp, frag);
+               while (mp->frames > MP_MAX_QUEUE_LEN) {
+                       frag = mp->frags->next;
+                       isdn_ppp_mp_free_skb(mp, mp->frags);
+                       mp->frags = frag;
                }
        }
        spin_unlock_irqrestore(&mp->lock, flags);
 }
 
-static void isdn_ppp_mp_cleanup(isdn_net_local *lp)
+static void isdn_ppp_mp_cleanup( isdn_net_local * lp )
 {
-       struct sk_buff *skb, *tmp;
-
-       skb_queue_walk_safe(&lp->netdev->pb->frags, skb, tmp) {
-               __skb_unlink(skb, &lp->netdev->pb->frags);
-               isdn_ppp_mp_free_skb(lp->netdev->pb, skb);
-       }
+       struct sk_buff * frag = lp->netdev->pb->frags;
+       struct sk_buff * nextfrag;
+       while( frag ) {
+               nextfrag = frag->next;
+               isdn_ppp_mp_free_skb(lp->netdev->pb, frag);
+               frag = nextfrag;
+       }
+       lp->netdev->pb->frags = NULL;
 }
 
 static u32 isdn_ppp_mp_get_seq( int short_seq, 
@@ -1838,115 +1855,72 @@ static u32 isdn_ppp_mp_get_seq( int short_seq,
        return seq;
 }
 
-static void isdn_ppp_mp_discard(ippp_bundle *mp, struct sk_buff *from,
-                               struct sk_buff *to)
+struct sk_buff * isdn_ppp_mp_discard( ippp_bundle * mp,
+                       struct sk_buff * from, struct sk_buff * to )
 {
-       if (from) {
-               struct sk_buff *skb, *tmp;
-               int freeing = 0;
-
-               skb_queue_walk_safe(&mp->frags, skb, tmp) {
-                       if (skb == to)
-                               break;
-                       if (skb == from)
-                               freeing = 1;
-                       if (!freeing)
-                               continue;
-                       __skb_unlink(skb, &mp->frags);
-                       isdn_ppp_mp_free_skb(mp, skb);
+       if( from )
+               while (from != to) {
+                       struct sk_buff * next = from->next;
+                       isdn_ppp_mp_free_skb(mp, from);
+                       from = next;
                }
-       }
-}
-
-static unsigned int calc_tot_len(struct sk_buff_head *queue,
-                                struct sk_buff *from, struct sk_buff *to)
-{
-       unsigned int tot_len = 0;
-       struct sk_buff *skb;
-       int found_start = 0;
-
-       skb_queue_walk(queue, skb) {
-               if (skb == from)
-                       found_start = 1;
-               if (!found_start)
-                       continue;
-               tot_len += skb->len - MP_HEADER_LEN;
-               if (skb == to)
-                       break;
-       }
-       return tot_len;
+       return from;
 }
 
-/* Reassemble packet using fragments in the reassembly queue from
- * 'from' until 'to', inclusive.
- */
-static void isdn_ppp_mp_reassembly(isdn_net_dev *net_dev, isdn_net_local *lp,
-                                  struct sk_buff *from, struct sk_buff *to,
-                                  u32 lastseq)
+void isdn_ppp_mp_reassembly( isdn_net_dev * net_dev, isdn_net_local * lp,
+                               struct sk_buff * from, struct sk_buff * to )
 {
-       ippp_bundle *mp = net_dev->pb;
-       unsigned int tot_len;
-       struct sk_buff *skb;
+       ippp_bundle * mp = net_dev->pb;
        int proto;
+       struct sk_buff * skb;
+       unsigned int tot_len;
 
        if (lp->ppp_slot < 0 || lp->ppp_slot >= ISDN_MAX_CHANNELS) {
                printk(KERN_ERR "%s: lp->ppp_slot(%d) out of range\n",
                        __func__, lp->ppp_slot);
                return;
        }
-
-       tot_len = calc_tot_len(&mp->frags, from, to);
-
-       if (MP_FLAGS(from) == (MP_BEGIN_FRAG | MP_END_FRAG)) {
-               if (ippp_table[lp->ppp_slot]->debug & 0x40)
+       if( MP_FLAGS(from) == (MP_BEGIN_FRAG | MP_END_FRAG) ) {
+               if( ippp_table[lp->ppp_slot]->debug & 0x40 )
                        printk(KERN_DEBUG "isdn_mppp: reassembly: frame %d, "
-                              "len %d\n", MP_SEQ(from), from->len);
+                                       "len %d\n", MP_SEQ(from), from->len );
                skb = from;
                skb_pull(skb, MP_HEADER_LEN);
-               __skb_unlink(skb, &mp->frags);
                mp->frames--;   
        } else {
-               struct sk_buff *walk, *tmp;
-               int found_start = 0;
+               struct sk_buff * frag;
+               int n;
 
-               if (ippp_table[lp->ppp_slot]->debug & 0x40)
-                       printk(KERN_DEBUG"isdn_mppp: reassembling frames %d "
-                              "to %d, len %d\n", MP_SEQ(from), lastseq,
-                              tot_len);
+               for(tot_len=n=0, frag=from; frag != to; frag=frag->next, n++)
+                       tot_len += frag->len - MP_HEADER_LEN;
 
-               skb = dev_alloc_skb(tot_len);
-               if (!skb)
+               if( ippp_table[lp->ppp_slot]->debug & 0x40 )
+                       printk(KERN_DEBUG"isdn_mppp: reassembling frames %d "
+                               "to %d, len %d\n", MP_SEQ(from), 
+                               (MP_SEQ(from)+n-1) & MP_LONGSEQ_MASK, tot_len );
+               if( (skb = dev_alloc_skb(tot_len)) == NULL ) {
                        printk(KERN_ERR "isdn_mppp: cannot allocate sk buff "
-                              "of size %d\n", tot_len);
-
-               found_start = 0;
-               skb_queue_walk_safe(&mp->frags, walk, tmp) {
-                       if (walk == from)
-                               found_start = 1;
-                       if (!found_start)
-                               continue;
+                                       "of size %d\n", tot_len);
+                       isdn_ppp_mp_discard(mp, from, to);
+                       return;
+               }
 
-                       if (skb) {
-                               unsigned int len = walk->len - MP_HEADER_LEN;
-                               skb_copy_from_linear_data_offset(walk, MP_HEADER_LEN,
-                                                                skb_put(skb, len),
-                                                                len);
-                       }
-                       __skb_unlink(walk, &mp->frags);
-                       isdn_ppp_mp_free_skb(mp, walk);
+               while( from != to ) {
+                       unsigned int len = from->len - MP_HEADER_LEN;
 
-                       if (walk == to)
-                               break;
+                       skb_copy_from_linear_data_offset(from, MP_HEADER_LEN,
+                                                        skb_put(skb,len),
+                                                        len);
+                       frag = from->next;
+                       isdn_ppp_mp_free_skb(mp, from);
+                       from = frag; 
                }
        }
-       if (!skb)
-               return;
-
        proto = isdn_ppp_strip_proto(skb);
        isdn_ppp_push_higher(net_dev, lp, skb, proto);
 }
 
-static void isdn_ppp_mp_free_skb(ippp_bundle *mp, struct sk_buff *skb)
+static void isdn_ppp_mp_free_skb(ippp_bundle * mp, struct sk_buff * skb)
 {
        dev_kfree_skb(skb);
        mp->frames--;
index c16e6ff..732b093 100644 (file)
@@ -35,63 +35,9 @@ config CAN_CALC_BITTIMING
          arguments "tq", "prop_seg", "phase_seg1", "phase_seg2" and "sjw".
          If unsure, say Y.
 
-config CAN_SJA1000
-       depends on CAN_DEV && HAS_IOMEM
-       tristate "Philips SJA1000"
-       ---help---
-         Driver for the SJA1000 CAN controllers from Philips or NXP
-
-config CAN_SJA1000_ISA
-       depends on CAN_SJA1000 && ISA
-       tristate "ISA Bus based legacy SJA1000 driver"
-       ---help---
-         This driver adds legacy support for SJA1000 chips connected to
-         the ISA bus using I/O port, memory mapped or indirect access.
-
-config CAN_SJA1000_PLATFORM
-       depends on CAN_SJA1000
-       tristate "Generic Platform Bus based SJA1000 driver"
-       ---help---
-         This driver adds support for the SJA1000 chips connected to
-         the "platform bus" (Linux abstraction for directly to the
-         processor attached devices).  Which can be found on various
-         boards from Phytec (http://www.phytec.de) like the PCM027,
-         PCM038.
-
-config CAN_SJA1000_OF_PLATFORM
-       depends on CAN_SJA1000 && PPC_OF
-       tristate "Generic OF Platform Bus based SJA1000 driver"
-       ---help---
-         This driver adds support for the SJA1000 chips connected to
-         the OpenFirmware "platform bus" found on embedded systems with
-         OpenFirmware bindings, e.g. if you have a PowerPC based system
-         you may want to enable this option.
-
-config CAN_EMS_PCI
-       tristate "EMS CPC-PCI, CPC-PCIe and CPC-104P Card"
-       depends on PCI && CAN_SJA1000
-       ---help---
-         This driver is for the one, two or four channel CPC-PCI,
-         CPC-PCIe and CPC-104P cards from EMS Dr. Thomas Wuensche
-         (http://www.ems-wuensche.de).
-
-config CAN_EMS_USB
-       tristate "EMS CPC-USB/ARM7 CAN/USB interface"
-       depends on USB && CAN_DEV
-       ---help---
-         This driver is for the one channel CPC-USB/ARM7 CAN/USB interface
-         from from EMS Dr. Thomas Wuensche (http://www.ems-wuensche.de).
-
-config CAN_KVASER_PCI
-       tristate "Kvaser PCIcanx and Kvaser PCIcan PCI Cards"
-       depends on PCI && CAN_SJA1000
-       ---help---
-         This driver is for the the PCIcanx and PCIcan cards (1, 2 or
-         4 channel) from Kvaser (http://www.kvaser.com).
-
 config CAN_AT91
        tristate "Atmel AT91 onchip CAN controller"
-       depends on CAN && CAN_DEV && ARCH_AT91SAM9263
+       depends on CAN_DEV && ARCH_AT91SAM9263
        ---help---
          This is a driver for the SoC CAN controller in Atmel's AT91SAM9263.
 
@@ -127,6 +73,10 @@ config CAN_MPC52XX
          This driver can also be built as a module.  If so, the module
          will be called mpc5xxx_can.
 
+source "drivers/net/can/sja1000/Kconfig"
+
+source "drivers/net/can/usb/Kconfig"
+
 config CAN_DEBUG_DEVICES
        bool "CAN devices debugging messages"
        depends on CAN
index 26c89aa..c1bb29f 100644 (file)
@@ -677,6 +677,11 @@ nla_put_failure:
        return -EMSGSIZE;
 }
 
+static size_t can_get_xstats_size(const struct net_device *dev)
+{
+       return sizeof(struct can_device_stats);
+}
+
 static int can_fill_xstats(struct sk_buff *skb, const struct net_device *dev)
 {
        struct can_priv *priv = netdev_priv(dev);
@@ -705,6 +710,7 @@ static struct rtnl_link_ops can_link_ops __read_mostly = {
        .changelink     = can_changelink,
        .get_size       = can_get_size,
        .fill_info      = can_fill_info,
+       .get_xstats_size = can_get_xstats_size,
        .fill_xstats    = can_fill_xstats,
 };
 
diff --git a/drivers/net/can/sja1000/Kconfig b/drivers/net/can/sja1000/Kconfig
new file mode 100644 (file)
index 0000000..4c67492
--- /dev/null
@@ -0,0 +1,47 @@
+menuconfig CAN_SJA1000
+       tristate "Philips/NXP SJA1000 devices"
+       depends on CAN_DEV && HAS_IOMEM
+
+if CAN_SJA1000
+
+config CAN_SJA1000_ISA
+       tristate "ISA Bus based legacy SJA1000 driver"
+       depends on ISA
+       ---help---
+         This driver adds legacy support for SJA1000 chips connected to
+         the ISA bus using I/O port, memory mapped or indirect access.
+
+config CAN_SJA1000_PLATFORM
+       tristate "Generic Platform Bus based SJA1000 driver"
+       ---help---
+         This driver adds support for the SJA1000 chips connected to
+         the "platform bus" (Linux abstraction for directly to the
+         processor attached devices).  Which can be found on various
+         boards from Phytec (http://www.phytec.de) like the PCM027,
+         PCM038.
+
+config CAN_SJA1000_OF_PLATFORM
+       tristate "Generic OF Platform Bus based SJA1000 driver"
+       depends on PPC_OF
+       ---help---
+         This driver adds support for the SJA1000 chips connected to
+         the OpenFirmware "platform bus" found on embedded systems with
+         OpenFirmware bindings, e.g. if you have a PowerPC based system
+         you may want to enable this option.
+
+config CAN_EMS_PCI
+       tristate "EMS CPC-PCI, CPC-PCIe and CPC-104P Card"
+       depends on PCI
+       ---help---
+         This driver is for the one, two or four channel CPC-PCI,
+         CPC-PCIe and CPC-104P cards from EMS Dr. Thomas Wuensche
+         (http://www.ems-wuensche.de).
+
+config CAN_KVASER_PCI
+       tristate "Kvaser PCIcanx and Kvaser PCIcan PCI Cards"
+       depends on PCI
+       ---help---
+         This driver is for the the PCIcanx and PCIcan cards (1, 2 or
+         4 channel) from Kvaser (http://www.kvaser.com).
+
+endif
diff --git a/drivers/net/can/usb/Kconfig b/drivers/net/can/usb/Kconfig
new file mode 100644 (file)
index 0000000..bbc78e0
--- /dev/null
@@ -0,0 +1,10 @@
+menu "CAN USB interfaces"
+       depends on USB && CAN_DEV
+
+config CAN_EMS_USB
+       tristate "EMS CPC-USB/ARM7 CAN/USB interface"
+       ---help---
+         This driver is for the one channel CPC-USB/ARM7 CAN/USB interface
+         from from EMS Dr. Thomas Wuensche (http://www.ems-wuensche.de).
+
+endmenu
index c3f75ba..0afd51d 100644 (file)
@@ -3,3 +3,5 @@
 #
 
 obj-$(CONFIG_CAN_EMS_USB) += ems_usb.o
+
+ccflags-$(CONFIG_CAN_DEBUG_DEVICES) := -DDEBUG
index 79ce8e8..8edac89 100644 (file)
@@ -2136,9 +2136,6 @@ static int emac_poll(struct napi_struct *napi, int budget)
        u32 status = 0;
        u32 num_pkts = 0;
 
-       if (!netif_running(ndev))
-               return 0;
-
        /* Check interrupt vectors and call packet processing */
        status = emac_read(EMAC_MACINVECTOR);
 
index 0a1c2bb..73fe977 100644 (file)
@@ -5820,10 +5820,7 @@ static int __devinit nv_probe(struct pci_dev *pci_dev, const struct pci_device_i
                        dev->dev_addr);
                dev_printk(KERN_ERR, &pci_dev->dev,
                        "Please complain to your hardware vendor. Switching to a random MAC.\n");
-               dev->dev_addr[0] = 0x00;
-               dev->dev_addr[1] = 0x00;
-               dev->dev_addr[2] = 0x6c;
-               get_random_bytes(&dev->dev_addr[3], 3);
+               random_ether_addr(dev->dev_addr);
        }
 
        dprintk(KERN_DEBUG "%s: MAC Address %pM\n",
index 884152d..dceed80 100644 (file)
@@ -5988,6 +5988,7 @@ static pci_ers_result_t ixgbe_io_slot_reset(struct pci_dev *pdev)
        } else {
                pci_set_master(pdev);
                pci_restore_state(pdev);
+               pci_save_state(pdev);
 
                pci_wake_from_d3(pdev, false);
 
index 9bf2a6b..965adb6 100644 (file)
@@ -1944,8 +1944,15 @@ ppp_receive_mp_frame(struct ppp *ppp, struct sk_buff *skb, struct channel *pch)
        }
 
        /* Pull completed packets off the queue and receive them. */
-       while ((skb = ppp_mp_reconstruct(ppp)))
-               ppp_receive_nonmp_frame(ppp, skb);
+       while ((skb = ppp_mp_reconstruct(ppp))) {
+               if (pskb_may_pull(skb, 2))
+                       ppp_receive_nonmp_frame(ppp, skb);
+               else {
+                       ++ppp->dev->stats.rx_length_errors;
+                       kfree_skb(skb);
+                       ppp_receive_error(ppp);
+               }
+       }
 
        return;
 
index 7dfcb58..8b14c6e 100644 (file)
@@ -1085,7 +1085,7 @@ static int __devinit r6040_init_one(struct pci_dev *pdev,
        int bar = 0;
        u16 *adrp;
 
-       printk(KERN_INFO "%s\n", version);
+       printk("%s\n", version);
 
        err = pci_enable_device(pdev);
        if (err)
index 1b0aa4c..98f6c51 100644 (file)
@@ -3243,9 +3243,9 @@ static void __devexit rtl8169_remove_one(struct pci_dev *pdev)
 static void rtl8169_set_rxbufsize(struct rtl8169_private *tp,
                                  struct net_device *dev)
 {
-       unsigned int mtu = dev->mtu;
+       unsigned int max_frame = dev->mtu + VLAN_ETH_HLEN + ETH_FCS_LEN;
 
-       tp->rx_buf_sz = (mtu > RX_BUF_SIZE) ? mtu + ETH_HLEN + 8 : RX_BUF_SIZE;
+       tp->rx_buf_sz = (max_frame > RX_BUF_SIZE) ? max_frame : RX_BUF_SIZE;
 }
 
 static int rtl8169_open(struct net_device *dev)
index ddccf5f..0dd7839 100644 (file)
@@ -3494,6 +3494,7 @@ static void s2io_reset(struct s2io_nic *sp)
 
                /* Restore the PCI state saved during initialization. */
                pci_restore_state(sp->pdev);
+               pci_save_state(sp->pdev);
                pci_read_config_word(sp->pdev, 0x2, &val16);
                if (check_pci_device_id(val16) != (u16)PCI_ANY_ID)
                        break;
index 6a9f51d..7f01e60 100644 (file)
@@ -986,7 +986,7 @@ static int smsc911x_poll(struct napi_struct *napi, int budget)
        struct net_device *dev = pdata->dev;
        int npackets = 0;
 
-       while (likely(netif_running(dev)) && (npackets < budget)) {
+       while (npackets < budget) {
                unsigned int pktlength;
                unsigned int pktwords;
                struct sk_buff *skb;
index 240cff1..6e2fc0c 100644 (file)
@@ -6029,7 +6029,7 @@ static struct net_device *ipw2100_alloc_device(struct pci_dev *pci_dev,
        struct ipw2100_priv *priv;
        struct net_device *dev;
 
-       dev = alloc_ieee80211(sizeof(struct ipw2100_priv), 0);
+       dev = alloc_ieee80211(sizeof(struct ipw2100_priv));
        if (!dev)
                return NULL;
        priv = libipw_priv(dev);
@@ -6342,7 +6342,7 @@ static int ipw2100_pci_init_one(struct pci_dev *pci_dev,
                sysfs_remove_group(&pci_dev->dev.kobj,
                                   &ipw2100_attribute_group);
 
-               free_ieee80211(dev, 0);
+               free_ieee80211(dev);
                pci_set_drvdata(pci_dev, NULL);
        }
 
@@ -6400,7 +6400,7 @@ static void __devexit ipw2100_pci_remove_one(struct pci_dev *pci_dev)
                if (dev->base_addr)
                        iounmap((void __iomem *)dev->base_addr);
 
-               free_ieee80211(dev, 0);
+               free_ieee80211(dev);
        }
 
        pci_release_regions(pci_dev);
index 61ef890..5c6ff58 100644 (file)
@@ -103,25 +103,6 @@ static int antenna = CFG_SYS_ANTENNA_BOTH;
 static int rtap_iface = 0;     /* def: 0 -- do not create rtap interface */
 #endif
 
-static struct ieee80211_rate ipw2200_rates[] = {
-       { .bitrate = 10 },
-       { .bitrate = 20, .flags = IEEE80211_RATE_SHORT_PREAMBLE },
-       { .bitrate = 55, .flags = IEEE80211_RATE_SHORT_PREAMBLE },
-       { .bitrate = 110, .flags = IEEE80211_RATE_SHORT_PREAMBLE },
-       { .bitrate = 60 },
-       { .bitrate = 90 },
-       { .bitrate = 120 },
-       { .bitrate = 180 },
-       { .bitrate = 240 },
-       { .bitrate = 360 },
-       { .bitrate = 480 },
-       { .bitrate = 540 }
-};
-
-#define ipw2200_a_rates                (ipw2200_rates + 4)
-#define ipw2200_num_a_rates    8
-#define ipw2200_bg_rates       (ipw2200_rates + 0)
-#define ipw2200_num_bg_rates   12
 
 #ifdef CONFIG_IPW2200_QOS
 static int qos_enable = 0;
@@ -8673,6 +8654,24 @@ static int ipw_sw_reset(struct ipw_priv *priv, int option)
  *
  */
 
+static int ipw_wx_get_name(struct net_device *dev,
+                          struct iw_request_info *info,
+                          union iwreq_data *wrqu, char *extra)
+{
+       struct ipw_priv *priv = libipw_priv(dev);
+       mutex_lock(&priv->mutex);
+       if (priv->status & STATUS_RF_KILL_MASK)
+               strcpy(wrqu->name, "radio off");
+       else if (!(priv->status & STATUS_ASSOCIATED))
+               strcpy(wrqu->name, "unassociated");
+       else
+               snprintf(wrqu->name, IFNAMSIZ, "IEEE 802.11%c",
+                        ipw_modes[priv->assoc_request.ieee_mode]);
+       IPW_DEBUG_WX("Name: %s\n", wrqu->name);
+       mutex_unlock(&priv->mutex);
+       return 0;
+}
+
 static int ipw_set_channel(struct ipw_priv *priv, u8 channel)
 {
        if (channel == 0) {
@@ -9972,7 +9971,7 @@ static int ipw_wx_sw_reset(struct net_device *dev,
 /* Rebase the WE IOCTLs to zero for the handler array */
 #define IW_IOCTL(x) [(x)-SIOCSIWCOMMIT]
 static iw_handler ipw_wx_handlers[] = {
-       IW_IOCTL(SIOCGIWNAME) = (iw_handler) cfg80211_wext_giwname,
+       IW_IOCTL(SIOCGIWNAME) = ipw_wx_get_name,
        IW_IOCTL(SIOCSIWFREQ) = ipw_wx_set_freq,
        IW_IOCTL(SIOCGIWFREQ) = ipw_wx_get_freq,
        IW_IOCTL(SIOCSIWMODE) = ipw_wx_set_mode,
@@ -11417,100 +11416,16 @@ static void ipw_bg_down(struct work_struct *work)
 /* Called by register_netdev() */
 static int ipw_net_init(struct net_device *dev)
 {
-       int i, rc = 0;
        struct ipw_priv *priv = libipw_priv(dev);
-       const struct libipw_geo *geo = libipw_get_geo(priv->ieee);
-       struct wireless_dev *wdev = &priv->ieee->wdev;
        mutex_lock(&priv->mutex);
 
        if (ipw_up(priv)) {
-               rc = -EIO;
-               goto out;
-       }
-
-       memcpy(wdev->wiphy->perm_addr, priv->mac_addr, ETH_ALEN);
-
-       /* fill-out priv->ieee->bg_band */
-       if (geo->bg_channels) {
-               struct ieee80211_supported_band *bg_band = &priv->ieee->bg_band;
-
-               bg_band->band = IEEE80211_BAND_2GHZ;
-               bg_band->n_channels = geo->bg_channels;
-               bg_band->channels =
-                       kzalloc(geo->bg_channels *
-                               sizeof(struct ieee80211_channel), GFP_KERNEL);
-               /* translate geo->bg to bg_band.channels */
-               for (i = 0; i < geo->bg_channels; i++) {
-                       bg_band->channels[i].band = IEEE80211_BAND_2GHZ;
-                       bg_band->channels[i].center_freq = geo->bg[i].freq;
-                       bg_band->channels[i].hw_value = geo->bg[i].channel;
-                       bg_band->channels[i].max_power = geo->bg[i].max_power;
-                       if (geo->bg[i].flags & LIBIPW_CH_PASSIVE_ONLY)
-                               bg_band->channels[i].flags |=
-                                       IEEE80211_CHAN_PASSIVE_SCAN;
-                       if (geo->bg[i].flags & LIBIPW_CH_NO_IBSS)
-                               bg_band->channels[i].flags |=
-                                       IEEE80211_CHAN_NO_IBSS;
-                       if (geo->bg[i].flags & LIBIPW_CH_RADAR_DETECT)
-                               bg_band->channels[i].flags |=
-                                       IEEE80211_CHAN_RADAR;
-                       /* No equivalent for LIBIPW_CH_80211H_RULES,
-                          LIBIPW_CH_UNIFORM_SPREADING, or
-                          LIBIPW_CH_B_ONLY... */
-               }
-               /* point at bitrate info */
-               bg_band->bitrates = ipw2200_bg_rates;
-               bg_band->n_bitrates = ipw2200_num_bg_rates;
-
-               wdev->wiphy->bands[IEEE80211_BAND_2GHZ] = bg_band;
-       }
-
-       /* fill-out priv->ieee->a_band */
-       if (geo->a_channels) {
-               struct ieee80211_supported_band *a_band = &priv->ieee->a_band;
-
-               a_band->band = IEEE80211_BAND_5GHZ;
-               a_band->n_channels = geo->a_channels;
-               a_band->channels =
-                       kzalloc(geo->a_channels *
-                               sizeof(struct ieee80211_channel), GFP_KERNEL);
-               /* translate geo->bg to a_band.channels */
-               for (i = 0; i < geo->a_channels; i++) {
-                       a_band->channels[i].band = IEEE80211_BAND_2GHZ;
-                       a_band->channels[i].center_freq = geo->a[i].freq;
-                       a_band->channels[i].hw_value = geo->a[i].channel;
-                       a_band->channels[i].max_power = geo->a[i].max_power;
-                       if (geo->a[i].flags & LIBIPW_CH_PASSIVE_ONLY)
-                               a_band->channels[i].flags |=
-                                       IEEE80211_CHAN_PASSIVE_SCAN;
-                       if (geo->a[i].flags & LIBIPW_CH_NO_IBSS)
-                               a_band->channels[i].flags |=
-                                       IEEE80211_CHAN_NO_IBSS;
-                       if (geo->a[i].flags & LIBIPW_CH_RADAR_DETECT)
-                               a_band->channels[i].flags |=
-                                       IEEE80211_CHAN_RADAR;
-                       /* No equivalent for LIBIPW_CH_80211H_RULES,
-                          LIBIPW_CH_UNIFORM_SPREADING, or
-                          LIBIPW_CH_B_ONLY... */
-               }
-               /* point at bitrate info */
-               a_band->bitrates = ipw2200_a_rates;
-               a_band->n_bitrates = ipw2200_num_a_rates;
-
-               wdev->wiphy->bands[IEEE80211_BAND_5GHZ] = a_band;
-       }
-
-       set_wiphy_dev(wdev->wiphy, &priv->pci_dev->dev);
-
-       /* With that information in place, we can now register the wiphy... */
-       if (wiphy_register(wdev->wiphy)) {
-               rc = -EIO;
-               goto out;
+               mutex_unlock(&priv->mutex);
+               return -EIO;
        }
 
-out:
        mutex_unlock(&priv->mutex);
-       return rc;
+       return 0;
 }
 
 /* PCI driver stuff */
@@ -11641,7 +11556,7 @@ static int ipw_prom_alloc(struct ipw_priv *priv)
        if (priv->prom_net_dev)
                return -EPERM;
 
-       priv->prom_net_dev = alloc_ieee80211(sizeof(struct ipw_prom_priv), 1);
+       priv->prom_net_dev = alloc_ieee80211(sizeof(struct ipw_prom_priv));
        if (priv->prom_net_dev == NULL)
                return -ENOMEM;
 
@@ -11660,7 +11575,7 @@ static int ipw_prom_alloc(struct ipw_priv *priv)
 
        rc = register_netdev(priv->prom_net_dev);
        if (rc) {
-               free_ieee80211(priv->prom_net_dev, 1);
+               free_ieee80211(priv->prom_net_dev);
                priv->prom_net_dev = NULL;
                return rc;
        }
@@ -11674,7 +11589,7 @@ static void ipw_prom_free(struct ipw_priv *priv)
                return;
 
        unregister_netdev(priv->prom_net_dev);
-       free_ieee80211(priv->prom_net_dev, 1);
+       free_ieee80211(priv->prom_net_dev);
 
        priv->prom_net_dev = NULL;
 }
@@ -11702,7 +11617,7 @@ static int __devinit ipw_pci_probe(struct pci_dev *pdev,
        struct ipw_priv *priv;
        int i;
 
-       net_dev = alloc_ieee80211(sizeof(struct ipw_priv), 0);
+       net_dev = alloc_ieee80211(sizeof(struct ipw_priv));
        if (net_dev == NULL) {
                err = -ENOMEM;
                goto out;
@@ -11850,7 +11765,7 @@ static int __devinit ipw_pci_probe(struct pci_dev *pdev,
        pci_disable_device(pdev);
        pci_set_drvdata(pdev, NULL);
       out_free_ieee80211:
-       free_ieee80211(priv->net_dev, 0);
+       free_ieee80211(priv->net_dev);
       out:
        return err;
 }
@@ -11917,7 +11832,7 @@ static void __devexit ipw_pci_remove(struct pci_dev *pdev)
        pci_release_regions(pdev);
        pci_disable_device(pdev);
        pci_set_drvdata(pdev, NULL);
-       free_ieee80211(priv->net_dev, 0);
+       free_ieee80211(priv->net_dev);
        free_firmware();
 }
 
index bf45391..1e334ff 100644 (file)
@@ -31,7 +31,6 @@
 #include <linux/ieee80211.h>
 
 #include <net/lib80211.h>
-#include <net/cfg80211.h>
 
 #define LIBIPW_VERSION "git-1.1.13"
 
@@ -784,15 +783,12 @@ struct libipw_geo {
 
 struct libipw_device {
        struct net_device *dev;
-       struct wireless_dev wdev;
        struct libipw_security sec;
 
        /* Bookkeeping structures */
        struct libipw_stats ieee_stats;
 
        struct libipw_geo geo;
-       struct ieee80211_supported_band bg_band;
-       struct ieee80211_supported_band a_band;
 
        /* Probe / Beacon management */
        struct list_head network_free_list;
@@ -1018,8 +1014,8 @@ static inline int libipw_is_cck_rate(u8 rate)
 }
 
 /* ieee80211.c */
-extern void free_ieee80211(struct net_device *dev, int monitor);
-extern struct net_device *alloc_ieee80211(int sizeof_priv, int monitor);
+extern void free_ieee80211(struct net_device *dev);
+extern struct net_device *alloc_ieee80211(int sizeof_priv);
 extern int libipw_change_mtu(struct net_device *dev, int new_mtu);
 
 extern void libipw_networks_age(struct libipw_device *ieee,
index a0e9f6a..eb2b608 100644 (file)
@@ -62,9 +62,6 @@ MODULE_DESCRIPTION(DRV_DESCRIPTION);
 MODULE_AUTHOR(DRV_COPYRIGHT);
 MODULE_LICENSE("GPL");
 
-struct cfg80211_ops libipw_config_ops = { };
-void *libipw_wiphy_privid = &libipw_wiphy_privid;
-
 static int libipw_networks_allocate(struct libipw_device *ieee)
 {
        if (ieee->networks)
@@ -143,7 +140,7 @@ int libipw_change_mtu(struct net_device *dev, int new_mtu)
 }
 EXPORT_SYMBOL(libipw_change_mtu);
 
-struct net_device *alloc_ieee80211(int sizeof_priv, int monitor)
+struct net_device *alloc_ieee80211(int sizeof_priv)
 {
        struct libipw_device *ieee;
        struct net_device *dev;
@@ -160,31 +157,10 @@ struct net_device *alloc_ieee80211(int sizeof_priv, int monitor)
 
        ieee->dev = dev;
 
-       if (!monitor) {
-               ieee->wdev.wiphy = wiphy_new(&libipw_config_ops, 0);
-               if (!ieee->wdev.wiphy) {
-                       LIBIPW_ERROR("Unable to allocate wiphy.\n");
-                       goto failed_free_netdev;
-               }
-
-               ieee->dev->ieee80211_ptr = &ieee->wdev;
-               ieee->wdev.iftype = NL80211_IFTYPE_STATION;
-
-               /* Fill-out wiphy structure bits we know...  Not enough info
-                  here to call set_wiphy_dev or set MAC address or channel info
-                  -- have to do that in ->ndo_init... */
-               ieee->wdev.wiphy->privid = libipw_wiphy_privid;
-
-               ieee->wdev.wiphy->max_scan_ssids = 1;
-               ieee->wdev.wiphy->max_scan_ie_len = 0;
-               ieee->wdev.wiphy->interface_modes = BIT(NL80211_IFTYPE_STATION)
-                                               | BIT(NL80211_IFTYPE_ADHOC);
-       }
-
        err = libipw_networks_allocate(ieee);
        if (err) {
                LIBIPW_ERROR("Unable to allocate beacon storage: %d\n", err);
-               goto failed_free_wiphy;
+               goto failed_free_netdev;
        }
        libipw_networks_initialize(ieee);
 
@@ -217,31 +193,19 @@ struct net_device *alloc_ieee80211(int sizeof_priv, int monitor)
 
        return dev;
 
-failed_free_wiphy:
-       if (!monitor)
-               wiphy_free(ieee->wdev.wiphy);
 failed_free_netdev:
        free_netdev(dev);
 failed:
        return NULL;
 }
 
-void free_ieee80211(struct net_device *dev, int monitor)
+void free_ieee80211(struct net_device *dev)
 {
        struct libipw_device *ieee = netdev_priv(dev);
 
        lib80211_crypt_info_free(&ieee->crypt_info);
 
        libipw_networks_free(ieee);
-
-       /* free cfg80211 resources */
-       if (!monitor) {
-               wiphy_unregister(ieee->wdev.wiphy);
-               kfree(ieee->a_band.channels);
-               kfree(ieee->bg_band.channels);
-               wiphy_free(ieee->wdev.wiphy);
-       }
-
        free_netdev(dev);
 }
 
index 17e1995..92af9b9 100644 (file)
@@ -426,12 +426,16 @@ static const char p54u_romboot_3887[] = "~~~~";
 static int p54u_firmware_reset_3887(struct ieee80211_hw *dev)
 {
        struct p54u_priv *priv = dev->priv;
-       u8 buf[4];
+       u8 *buf;
        int ret;
 
-       memcpy(&buf, p54u_romboot_3887, sizeof(buf));
+       buf = kmalloc(4, GFP_KERNEL);
+       if (!buf)
+               return -ENOMEM;
+       memcpy(buf, p54u_romboot_3887, 4);
        ret = p54u_bulk_msg(priv, P54U_PIPE_DATA,
-                           buf, sizeof(buf));
+                           buf, 4);
+       kfree(buf);
        if (ret)
                dev_err(&priv->udev->dev, "(p54usb) unable to jump to "
                        "boot ROM (%d)!\n", ret);
index 4c218ee..8687a7d 100644 (file)
@@ -157,7 +157,7 @@ typedef struct {
 
 typedef struct {
   int mp_mrru;                        /* unused                             */
-  struct sk_buff_head frags;   /* fragments sl list */
+  struct sk_buff * frags;      /* fragments sl list -- use skb->next */
   long frames;                 /* number of frames in the frame list */
   unsigned int seq;            /* last processed packet seq #: any packets
                                 * with smaller seq # will be dropped
index 6e5f0e0..cd2e187 100644 (file)
@@ -1980,7 +1980,7 @@ void sctp_assoc_set_primary(struct sctp_association *,
 void sctp_assoc_del_nonprimary_peers(struct sctp_association *,
                                    struct sctp_transport *);
 int sctp_assoc_set_bind_addr_from_ep(struct sctp_association *,
-                                    gfp_t);
+                                    sctp_scope_t, gfp_t);
 int sctp_assoc_set_bind_addr_from_cookie(struct sctp_association *,
                                         struct sctp_cookie*,
                                         gfp_t gfp);
index a975098..b7c4224 100644 (file)
@@ -211,6 +211,7 @@ struct hci_conn *hci_conn_add(struct hci_dev *hdev, int type, bdaddr_t *dst)
        conn->type  = type;
        conn->mode  = HCI_CM_ACTIVE;
        conn->state = BT_OPEN;
+       conn->auth_type = HCI_AT_GENERAL_BONDING;
 
        conn->power_save = 1;
        conn->disc_timeout = HCI_DISCONN_TIMEOUT;
index ff0233d..80d9298 100644 (file)
@@ -2206,7 +2206,7 @@ static int l2cap_build_conf_req(struct sock *sk, void *data)
 {
        struct l2cap_pinfo *pi = l2cap_pi(sk);
        struct l2cap_conf_req *req = data;
-       struct l2cap_conf_rfc rfc = { .mode = L2CAP_MODE_ERTM };
+       struct l2cap_conf_rfc rfc = { .mode = L2CAP_MODE_BASIC };
        void *ptr = req->data;
 
        BT_DBG("sk %p", sk);
@@ -2395,6 +2395,10 @@ done:
                        rfc.monitor_timeout = L2CAP_DEFAULT_MONITOR_TO;
 
                        pi->conf_state |= L2CAP_CONF_MODE_DONE;
+
+                       l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC,
+                                       sizeof(rfc), (unsigned long) &rfc);
+
                        break;
 
                case L2CAP_MODE_STREAMING:
@@ -2402,6 +2406,10 @@ done:
                        pi->max_pdu_size = rfc.max_pdu_size;
 
                        pi->conf_state |= L2CAP_CONF_MODE_DONE;
+
+                       l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC,
+                                       sizeof(rfc), (unsigned long) &rfc);
+
                        break;
 
                default:
@@ -2411,9 +2419,6 @@ done:
                        rfc.mode = pi->mode;
                }
 
-               l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC,
-                                       sizeof(rfc), (unsigned long) &rfc);
-
                if (result == L2CAP_CONF_SUCCESS)
                        pi->conf_state |= L2CAP_CONF_OUTPUT_DONE;
        }
index 4b24d79..d867522 100644 (file)
@@ -999,14 +999,15 @@ rollback:
        ret = notifier_to_errno(ret);
 
        if (ret) {
-               if (err) {
-                       printk(KERN_ERR
-                              "%s: name change rollback failed: %d.\n",
-                              dev->name, ret);
-               } else {
+               /* err >= 0 after dev_alloc_name() or stores the first errno */
+               if (err >= 0) {
                        err = ret;
                        memcpy(dev->name, oldname, IFNAMSIZ);
                        goto rollback;
+               } else {
+                       printk(KERN_ERR
+                              "%s: name change rollback failed: %d.\n",
+                              dev->name, ret);
                }
        }
 
index ef4ee45..54596f7 100644 (file)
@@ -494,8 +494,10 @@ static int vif_add(struct net *net, struct vifctl *vifc, int mrtsock)
                return -EINVAL;
        }
 
-       if ((in_dev = __in_dev_get_rtnl(dev)) == NULL)
+       if ((in_dev = __in_dev_get_rtnl(dev)) == NULL) {
+               dev_put(dev);
                return -EADDRNOTAVAIL;
+       }
        IPV4_DEVCONF(in_dev->cnf, MC_FORWARDING)++;
        ip_rt_multicast_event(in_dev);
 
index e0cfa63..524f976 100644 (file)
@@ -1183,7 +1183,9 @@ void tcp_cleanup_rbuf(struct sock *sk, int copied)
 #if TCP_DEBUG
        struct sk_buff *skb = skb_peek(&sk->sk_receive_queue);
 
-       WARN_ON(skb && !before(tp->copied_seq, TCP_SKB_CB(skb)->end_seq));
+       WARN(skb && !before(tp->copied_seq, TCP_SKB_CB(skb)->end_seq),
+            KERN_INFO "cleanup rbuf bug: copied %X seq %X rcvnxt %X\n",
+            tp->copied_seq, TCP_SKB_CB(skb)->end_seq, tp->rcv_nxt);
 #endif
 
        if (inet_csk_ack_scheduled(sk)) {
@@ -1430,11 +1432,13 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
                        /* Now that we have two receive queues this
                         * shouldn't happen.
                         */
-                       if (before(*seq, TCP_SKB_CB(skb)->seq)) {
-                               printk(KERN_INFO "recvmsg bug: copied %X "
-                                      "seq %X\n", *seq, TCP_SKB_CB(skb)->seq);
+                       if (WARN(before(*seq, TCP_SKB_CB(skb)->seq),
+                            KERN_INFO "recvmsg bug: copied %X "
+                                      "seq %X rcvnxt %X fl %X\n", *seq,
+                                      TCP_SKB_CB(skb)->seq, tp->rcv_nxt,
+                                      flags))
                                break;
-                       }
+
                        offset = *seq - TCP_SKB_CB(skb)->seq;
                        if (tcp_hdr(skb)->syn)
                                offset--;
@@ -1443,8 +1447,9 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
                        if (tcp_hdr(skb)->fin)
                                goto found_fin_ok;
                        WARN(!(flags & MSG_PEEK), KERN_INFO "recvmsg bug 2: "
-                                       "copied %X seq %X\n", *seq,
-                                       TCP_SKB_CB(skb)->seq);
+                                       "copied %X seq %X rcvnxt %X fl %X\n",
+                                       *seq, TCP_SKB_CB(skb)->seq,
+                                       tp->rcv_nxt, flags);
                }
 
                /* Well, if we have backlog, try to process it now yet. */
index 8450960..7eed77a 100644 (file)
@@ -1485,15 +1485,13 @@ void sctp_assoc_rwnd_decrease(struct sctp_association *asoc, unsigned len)
  * local endpoint and the remote peer.
  */
 int sctp_assoc_set_bind_addr_from_ep(struct sctp_association *asoc,
-                                    gfp_t gfp)
+                                    sctp_scope_t scope, gfp_t gfp)
 {
-       sctp_scope_t scope;
        int flags;
 
        /* Use scoping rules to determine the subset of addresses from
         * the endpoint.
         */
-       scope = sctp_scope(&asoc->peer.active_path->ipaddr);
        flags = (PF_INET6 == asoc->base.sk->sk_family) ? SCTP_ADDR6_ALLOWED : 0;
        if (asoc->peer.ipv4_address)
                flags |= SCTP_ADDR4_PEERSUPP;
index c8fae19..d4df450 100644 (file)
@@ -384,6 +384,11 @@ sctp_disposition_t sctp_sf_do_5_1B_init(const struct sctp_endpoint *ep,
        if (!new_asoc)
                goto nomem;
 
+       if (sctp_assoc_set_bind_addr_from_ep(new_asoc,
+                                            sctp_scope(sctp_source(chunk)),
+                                            GFP_ATOMIC) < 0)
+               goto nomem_init;
+
        /* The call, sctp_process_init(), can fail on memory allocation.  */
        if (!sctp_process_init(new_asoc, chunk->chunk_hdr->type,
                               sctp_source(chunk),
@@ -401,9 +406,6 @@ sctp_disposition_t sctp_sf_do_5_1B_init(const struct sctp_endpoint *ep,
                len = ntohs(err_chunk->chunk_hdr->length) -
                        sizeof(sctp_chunkhdr_t);
 
-       if (sctp_assoc_set_bind_addr_from_ep(new_asoc, GFP_ATOMIC) < 0)
-               goto nomem_init;
-
        repl = sctp_make_init_ack(new_asoc, chunk, GFP_ATOMIC, len);
        if (!repl)
                goto nomem_init;
@@ -1452,6 +1454,10 @@ static sctp_disposition_t sctp_sf_do_unexpected_init(
        if (!new_asoc)
                goto nomem;
 
+       if (sctp_assoc_set_bind_addr_from_ep(new_asoc,
+                               sctp_scope(sctp_source(chunk)), GFP_ATOMIC) < 0)
+               goto nomem;
+
        /* In the outbound INIT ACK the endpoint MUST copy its current
         * Verification Tag and Peers Verification tag into a reserved
         * place (local tie-tag and per tie-tag) within the state cookie.
@@ -1488,9 +1494,6 @@ static sctp_disposition_t sctp_sf_do_unexpected_init(
                        sizeof(sctp_chunkhdr_t);
        }
 
-       if (sctp_assoc_set_bind_addr_from_ep(new_asoc, GFP_ATOMIC) < 0)
-               goto nomem;
-
        repl = sctp_make_init_ack(new_asoc, chunk, GFP_ATOMIC, len);
        if (!repl)
                goto nomem;
index 4085db9..66b1f02 100644 (file)
@@ -1080,6 +1080,13 @@ static int __sctp_connect(struct sock* sk,
                                err = -ENOMEM;
                                goto out_free;
                        }
+
+                       err = sctp_assoc_set_bind_addr_from_ep(asoc, scope,
+                                                             GFP_KERNEL);
+                       if (err < 0) {
+                               goto out_free;
+                       }
+
                }
 
                /* Prime the peer's transport structures.  */
@@ -1095,11 +1102,6 @@ static int __sctp_connect(struct sock* sk,
                walk_size += af->sockaddr_len;
        }
 
-       err = sctp_assoc_set_bind_addr_from_ep(asoc, GFP_KERNEL);
-       if (err < 0) {
-               goto out_free;
-       }
-
        /* In case the user of sctp_connectx() wants an association
         * id back, assign one now.
         */
@@ -1274,22 +1276,30 @@ SCTP_STATIC int sctp_setsockopt_connectx(struct sock* sk,
 }
 
 /*
- * New (hopefully final) interface for the API.  The option buffer is used
- * both for the returned association id and the addresses.
+ * New (hopefully final) interface for the API.
+ * We use the sctp_getaddrs_old structure so that use-space library
+ * can avoid any unnecessary allocations.   The only defferent part
+ * is that we store the actual length of the address buffer into the
+ * addrs_num structure member.  That way we can re-use the existing
+ * code.
  */
 SCTP_STATIC int sctp_getsockopt_connectx3(struct sock* sk, int len,
                                        char __user *optval,
                                        int __user *optlen)
 {
+       struct sctp_getaddrs_old param;
        sctp_assoc_t assoc_id = 0;
        int err = 0;
 
-       if (len < sizeof(assoc_id))
+       if (len < sizeof(param))
                return -EINVAL;
 
+       if (copy_from_user(&param, optval, sizeof(param)))
+               return -EFAULT;
+
        err = __sctp_setsockopt_connectx(sk,
-                       (struct sockaddr __user *)(optval + sizeof(assoc_id)),
-                       len - sizeof(assoc_id), &assoc_id);
+                       (struct sockaddr __user *)param.addrs,
+                       param.addr_num, &assoc_id);
 
        if (err == 0 || err == -EINPROGRESS) {
                if (copy_to_user(optval, &assoc_id, sizeof(assoc_id)))
@@ -1689,6 +1699,11 @@ SCTP_STATIC int sctp_sendmsg(struct kiocb *iocb, struct sock *sk,
                        goto out_unlock;
                }
                asoc = new_asoc;
+               err = sctp_assoc_set_bind_addr_from_ep(asoc, scope, GFP_KERNEL);
+               if (err < 0) {
+                       err = -ENOMEM;
+                       goto out_free;
+               }
 
                /* If the SCTP_INIT ancillary data is specified, set all
                 * the association init values accordingly.
@@ -1718,11 +1733,6 @@ SCTP_STATIC int sctp_sendmsg(struct kiocb *iocb, struct sock *sk,
                        err = -ENOMEM;
                        goto out_free;
                }
-               err = sctp_assoc_set_bind_addr_from_ep(asoc, GFP_KERNEL);
-               if (err < 0) {
-                       err = -ENOMEM;
-                       goto out_free;
-               }
        }
 
        /* ASSERT: we have a valid association at this point.  */
index c256e48..3b141bb 100644 (file)
@@ -308,7 +308,8 @@ void sctp_transport_route(struct sctp_transport *transport,
                /* Initialize sk->sk_rcv_saddr, if the transport is the
                 * association's active path for getsockname().
                 */
-               if (asoc && (transport == asoc->peer.active_path))
+               if (asoc && (!asoc->peer.primary_path ||
+                               (transport == asoc->peer.active_path)))
                        opt->pf->af->to_sk_saddr(&transport->saddr,
                                                 asoc->base.sk);
        } else