From b9f5f52cca3e94f1e7509f366aa250ebbe1ed0b5 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Tue, 3 Jun 2008 16:03:15 -0700 Subject: net: neighbour table ABI problem The neighbor table time of last use information is returned in the incorrect unit. Kernel to user space ABI's need to use USER_HZ (or milliseconds), otherwise the application has to try and discover the real system HZ value which is problematic. Linux has standardized on keeping USER_HZ consistent (100hz) even when kernel is running internally at some other value. This change is small, but it breaks the ABI for older version of iproute2 utilities. But these utilities are already broken since they are looking at the psched_hz values which are completely different. So let's just go ahead and fix both kernel and user space. Older utilities will just print wrong values. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/core/neighbour.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'net/core') diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 5d9d7130bd6..3896de79dfb 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -2057,9 +2057,9 @@ static int neigh_fill_info(struct sk_buff *skb, struct neighbour *neigh, goto nla_put_failure; } - ci.ndm_used = now - neigh->used; - ci.ndm_confirmed = now - neigh->confirmed; - ci.ndm_updated = now - neigh->updated; + ci.ndm_used = jiffies_to_clock_t(now - neigh->used); + ci.ndm_confirmed = jiffies_to_clock_t(now - neigh->confirmed); + ci.ndm_updated = jiffies_to_clock_t(now - neigh->updated); ci.ndm_refcnt = atomic_read(&neigh->refcnt) - 1; read_unlock_bh(&neigh->lock); -- cgit v1.2.3-70-g09d2 From 7557af25155a82ac2dad73eec6b0166868bf8ea2 Mon Sep 17 00:00:00 2001 From: Brice Goglin Date: Tue, 3 Jun 2008 16:07:45 -0700 Subject: net_dma: remove duplicate assignment in dma_skb_copy_datagram_iovec No need to compute copy twice in the frags loop in dma_skb_copy_datagram_iovec(). Signed-off-by: Brice Goglin Acked-by: Shannon Nelson Signed-off-by: Maciej Sosnowski Signed-off-by: Dan Williams Signed-off-by: David S. Miller --- net/core/user_dma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/core') diff --git a/net/core/user_dma.c b/net/core/user_dma.c index 0ad1cd57bc3..c77aff9c6eb 100644 --- a/net/core/user_dma.c +++ b/net/core/user_dma.c @@ -75,7 +75,7 @@ int dma_skb_copy_datagram_iovec(struct dma_chan *chan, end = start + skb_shinfo(skb)->frags[i].size; copy = end - offset; - if ((copy = end - offset) > 0) { + if (copy > 0) { skb_frag_t *frag = &skb_shinfo(skb)->frags[i]; struct page *page = frag->page; -- cgit v1.2.3-70-g09d2 From bc3ed28caaef55e7e3a9316464256353c5f9b1df Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Tue, 3 Jun 2008 16:36:54 -0700 Subject: netlink: Improve returned error codes Make nlmsg_trim(), nlmsg_cancel(), genlmsg_cancel(), and nla_nest_cancel() void functions. Return -EMSGSIZE instead of -1 if the provided message buffer is not big enough. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- include/net/genetlink.h | 4 ++-- include/net/netlink.h | 20 +++++++++----------- net/core/neighbour.c | 3 ++- net/core/rtnetlink.c | 3 ++- net/netlink/attr.c | 12 ++++++------ net/netlink/genetlink.c | 6 ++++-- net/sched/sch_dsmark.c | 6 ++++-- net/sched/sch_gred.c | 3 ++- net/sched/sch_hfsc.c | 2 +- net/sched/sch_red.c | 3 ++- net/wireless/nl80211.c | 12 ++++++++---- 11 files changed, 42 insertions(+), 32 deletions(-) (limited to 'net/core') diff --git a/include/net/genetlink.h b/include/net/genetlink.h index decdda54682..747c255d1df 100644 --- a/include/net/genetlink.h +++ b/include/net/genetlink.h @@ -162,9 +162,9 @@ static inline int genlmsg_end(struct sk_buff *skb, void *hdr) * @skb: socket buffer the message is stored in * @hdr: generic netlink message header */ -static inline int genlmsg_cancel(struct sk_buff *skb, void *hdr) +static inline void genlmsg_cancel(struct sk_buff *skb, void *hdr) { - return nlmsg_cancel(skb, hdr - GENL_HDRLEN - NLMSG_HDRLEN); + nlmsg_cancel(skb, hdr - GENL_HDRLEN - NLMSG_HDRLEN); } /** diff --git a/include/net/netlink.h b/include/net/netlink.h index 112dcdf7e34..dfc3701dfcc 100644 --- a/include/net/netlink.h +++ b/include/net/netlink.h @@ -556,14 +556,12 @@ static inline void *nlmsg_get_pos(struct sk_buff *skb) * @skb: socket buffer the message is stored in * @mark: mark to trim to * - * Trims the message to the provided mark. Returns -1. + * Trims the message to the provided mark. */ -static inline int nlmsg_trim(struct sk_buff *skb, const void *mark) +static inline void nlmsg_trim(struct sk_buff *skb, const void *mark) { if (mark) skb_trim(skb, (unsigned char *) mark - skb->data); - - return -1; } /** @@ -572,11 +570,11 @@ static inline int nlmsg_trim(struct sk_buff *skb, const void *mark) * @nlh: netlink message header * * Removes the complete netlink message including all - * attributes from the socket buffer again. Returns -1. + * attributes from the socket buffer again. */ -static inline int nlmsg_cancel(struct sk_buff *skb, struct nlmsghdr *nlh) +static inline void nlmsg_cancel(struct sk_buff *skb, struct nlmsghdr *nlh) { - return nlmsg_trim(skb, nlh); + nlmsg_trim(skb, nlh); } /** @@ -775,7 +773,7 @@ static inline int __nla_parse_nested_compat(struct nlattr *tb[], int maxtype, int nested_len = nla_len(nla) - NLA_ALIGN(len); if (nested_len < 0) - return -1; + return -EINVAL; if (nested_len >= nla_attr_size(0)) return nla_parse(tb, maxtype, nla_data(nla) + NLA_ALIGN(len), nested_len, policy); @@ -1080,11 +1078,11 @@ static inline int nla_nest_compat_end(struct sk_buff *skb, struct nlattr *start) * @start: container attribute * * Removes the container attribute and including all nested - * attributes. Returns -1. + * attributes. Returns -EMSGSIZE */ -static inline int nla_nest_cancel(struct sk_buff *skb, struct nlattr *start) +static inline void nla_nest_cancel(struct sk_buff *skb, struct nlattr *start) { - return nlmsg_trim(skb, start); + nlmsg_trim(skb, start); } /** diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 3896de79dfb..65f01f71b3f 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -1714,7 +1714,8 @@ static int neightbl_fill_parms(struct sk_buff *skb, struct neigh_parms *parms) return nla_nest_end(skb, nest); nla_put_failure: - return nla_nest_cancel(skb, nest); + nla_nest_cancel(skb, nest); + return -EMSGSIZE; } static int neightbl_fill_info(struct sk_buff *skb, struct neigh_table *tbl, diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index cf857c4dc7b..a9a77216310 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -498,7 +498,8 @@ int rtnetlink_put_metrics(struct sk_buff *skb, u32 *metrics) return nla_nest_end(skb, mx); nla_put_failure: - return nla_nest_cancel(skb, mx); + nla_nest_cancel(skb, mx); + return -EMSGSIZE; } int rtnl_put_cacheinfo(struct sk_buff *skb, struct dst_entry *dst, u32 id, diff --git a/net/netlink/attr.c b/net/netlink/attr.c index feb326f4a75..47bbf45ae5d 100644 --- a/net/netlink/attr.c +++ b/net/netlink/attr.c @@ -400,13 +400,13 @@ void __nla_put_nohdr(struct sk_buff *skb, int attrlen, const void *data) * @attrlen: length of attribute payload * @data: head of attribute payload * - * Returns -1 if the tailroom of the skb is insufficient to store + * Returns -EMSGSIZE if the tailroom of the skb is insufficient to store * the attribute header and payload. */ int nla_put(struct sk_buff *skb, int attrtype, int attrlen, const void *data) { if (unlikely(skb_tailroom(skb) < nla_total_size(attrlen))) - return -1; + return -EMSGSIZE; __nla_put(skb, attrtype, attrlen, data); return 0; @@ -418,13 +418,13 @@ int nla_put(struct sk_buff *skb, int attrtype, int attrlen, const void *data) * @attrlen: length of attribute payload * @data: head of attribute payload * - * Returns -1 if the tailroom of the skb is insufficient to store + * Returns -EMSGSIZE if the tailroom of the skb is insufficient to store * the attribute payload. */ int nla_put_nohdr(struct sk_buff *skb, int attrlen, const void *data) { if (unlikely(skb_tailroom(skb) < NLA_ALIGN(attrlen))) - return -1; + return -EMSGSIZE; __nla_put_nohdr(skb, attrlen, data); return 0; @@ -436,13 +436,13 @@ int nla_put_nohdr(struct sk_buff *skb, int attrlen, const void *data) * @attrlen: length of attribute payload * @data: head of attribute payload * - * Returns -1 if the tailroom of the skb is insufficient to store + * Returns -EMSGSIZE if the tailroom of the skb is insufficient to store * the attribute payload. */ int nla_append(struct sk_buff *skb, int attrlen, const void *data) { if (unlikely(skb_tailroom(skb) < NLA_ALIGN(attrlen))) - return -1; + return -EMSGSIZE; memcpy(skb_put(skb, attrlen), data, attrlen); return 0; diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c index d16929c9b4b..f5aa23c3e88 100644 --- a/net/netlink/genetlink.c +++ b/net/netlink/genetlink.c @@ -554,7 +554,8 @@ static int ctrl_fill_info(struct genl_family *family, u32 pid, u32 seq, return genlmsg_end(skb, hdr); nla_put_failure: - return genlmsg_cancel(skb, hdr); + genlmsg_cancel(skb, hdr); + return -EMSGSIZE; } static int ctrl_fill_mcgrp_info(struct genl_multicast_group *grp, u32 pid, @@ -590,7 +591,8 @@ static int ctrl_fill_mcgrp_info(struct genl_multicast_group *grp, u32 pid, return genlmsg_end(skb, hdr); nla_put_failure: - return genlmsg_cancel(skb, hdr); + genlmsg_cancel(skb, hdr); + return -EMSGSIZE; } static int ctrl_dumpfamily(struct sk_buff *skb, struct netlink_callback *cb) diff --git a/net/sched/sch_dsmark.c b/net/sched/sch_dsmark.c index 0df911fd67b..64465bacbe7 100644 --- a/net/sched/sch_dsmark.c +++ b/net/sched/sch_dsmark.c @@ -444,7 +444,8 @@ static int dsmark_dump_class(struct Qdisc *sch, unsigned long cl, return nla_nest_end(skb, opts); nla_put_failure: - return nla_nest_cancel(skb, opts); + nla_nest_cancel(skb, opts); + return -EMSGSIZE; } static int dsmark_dump(struct Qdisc *sch, struct sk_buff *skb) @@ -466,7 +467,8 @@ static int dsmark_dump(struct Qdisc *sch, struct sk_buff *skb) return nla_nest_end(skb, opts); nla_put_failure: - return nla_nest_cancel(skb, opts); + nla_nest_cancel(skb, opts); + return -EMSGSIZE; } static const struct Qdisc_class_ops dsmark_class_ops = { diff --git a/net/sched/sch_gred.c b/net/sched/sch_gred.c index 3a9d226ff1e..c89fba56db5 100644 --- a/net/sched/sch_gred.c +++ b/net/sched/sch_gred.c @@ -582,7 +582,8 @@ append_opt: return nla_nest_end(skb, opts); nla_put_failure: - return nla_nest_cancel(skb, opts); + nla_nest_cancel(skb, opts); + return -EMSGSIZE; } static void gred_destroy(struct Qdisc *sch) diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c index 87293d0db1d..fdfaa3fcc16 100644 --- a/net/sched/sch_hfsc.c +++ b/net/sched/sch_hfsc.c @@ -1360,7 +1360,7 @@ hfsc_dump_class(struct Qdisc *sch, unsigned long arg, struct sk_buff *skb, nla_put_failure: nla_nest_cancel(skb, nest); - return -1; + return -EMSGSIZE; } static int diff --git a/net/sched/sch_red.c b/net/sched/sch_red.c index 3dcd493f4f4..5c569853b9c 100644 --- a/net/sched/sch_red.c +++ b/net/sched/sch_red.c @@ -281,7 +281,8 @@ static int red_dump(struct Qdisc *sch, struct sk_buff *skb) return nla_nest_end(skb, opts); nla_put_failure: - return nla_nest_cancel(skb, opts); + nla_nest_cancel(skb, opts); + return -EMSGSIZE; } static int red_dump_stats(struct Qdisc *sch, struct gnet_dump *d) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 2bdd4dddc0e..fb75f265b39 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -187,7 +187,8 @@ static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags, return genlmsg_end(msg, hdr); nla_put_failure: - return genlmsg_cancel(msg, hdr); + genlmsg_cancel(msg, hdr); + return -EMSGSIZE; } static int nl80211_dump_wiphy(struct sk_buff *skb, struct netlink_callback *cb) @@ -273,7 +274,8 @@ static int nl80211_send_iface(struct sk_buff *msg, u32 pid, u32 seq, int flags, return genlmsg_end(msg, hdr); nla_put_failure: - return genlmsg_cancel(msg, hdr); + genlmsg_cancel(msg, hdr); + return -EMSGSIZE; } static int nl80211_dump_interface(struct sk_buff *skb, struct netlink_callback *cb) @@ -928,7 +930,8 @@ static int nl80211_send_station(struct sk_buff *msg, u32 pid, u32 seq, return genlmsg_end(msg, hdr); nla_put_failure: - return genlmsg_cancel(msg, hdr); + genlmsg_cancel(msg, hdr); + return -EMSGSIZE; } static int nl80211_dump_station(struct sk_buff *skb, @@ -1267,7 +1270,8 @@ static int nl80211_send_mpath(struct sk_buff *msg, u32 pid, u32 seq, return genlmsg_end(msg, hdr); nla_put_failure: - return genlmsg_cancel(msg, hdr); + genlmsg_cancel(msg, hdr); + return -EMSGSIZE; } static int nl80211_dump_mpath(struct sk_buff *skb, -- cgit v1.2.3-70-g09d2 From 293ad60401da621b8b329abbe8c388edb25f658a Mon Sep 17 00:00:00 2001 From: Octavian Purdila Date: Wed, 4 Jun 2008 15:45:58 -0700 Subject: tcp: Fix for race due to temporary drop of the socket lock in skb_splice_bits. skb_splice_bits temporary drops the socket lock while iterating over the socket queue in order to break a reverse locking condition which happens with sendfile. This, however, opens a window of opportunity for tcp_collapse() to aggregate skbs and thus potentially free the current skb used in skb_splice_bits and tcp_read_sock. This patch fixes the problem by (re-)getting the same "logical skb" after the lock has been temporary dropped. Based on idea and initial patch from Evgeniy Polyakov. Signed-off-by: Octavian Purdila Acked-by: Evgeniy Polyakov Signed-off-by: David S. Miller --- net/core/skbuff.c | 5 +++-- net/ipv4/tcp.c | 9 ++++++++- 2 files changed, 11 insertions(+), 3 deletions(-) (limited to 'net/core') diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 5c459f2b798..1e556d31211 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -1445,6 +1445,7 @@ done: if (spd.nr_pages) { int ret; + struct sock *sk = __skb->sk; /* * Drop the socket lock, otherwise we have reverse @@ -1455,9 +1456,9 @@ done: * we call into ->sendpage() with the i_mutex lock held * and networking will grab the socket lock. */ - release_sock(__skb->sk); + release_sock(sk); ret = splice_to_pipe(pipe, &spd); - lock_sock(__skb->sk); + lock_sock(sk); return ret; } diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index f8865313862..ab66683b804 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1227,7 +1227,14 @@ int tcp_read_sock(struct sock *sk, read_descriptor_t *desc, copied += used; offset += used; } - if (offset != skb->len) + /* + * If recv_actor drops the lock (e.g. TCP splice + * receive) the skb pointer might be invalid when + * getting here: tcp_collapse might have deleted it + * while aggregating skbs from the socket queue. + */ + skb = tcp_recv_skb(sk, seq-1, &offset); + if (!skb || (offset+1 != skb->len)) break; } if (tcp_hdr(skb)->fin) { -- cgit v1.2.3-70-g09d2