summaryrefslogtreecommitdiffstats
path: root/net
Commit message (Collapse)AuthorAgeFilesLines
* tcp: do not scale TSO segment size with reordering degreeNeal Cardwell2011-11-292-2/+2
| | | | | | | | | | | | | | | | Since 2005 (c1b4a7e69576d65efc31a8cea0714173c2841244) tcp_tso_should_defer has been using tcp_max_burst() as a target limit for deciding how large to make outgoing TSO packets when not using sysctl_tcp_tso_win_divisor. But since 2008 (dd9e0dda66ba38a2ddd1405ac279894260dc5c36) tcp_max_burst() returns the reordering degree. We should not have tcp_tso_should_defer attempt to build larger segments just because there is more reordering. This commit splits the notion of deferral size used in TSO from the notion of burst size used in cwnd moderation, and returns the TSO deferral limit to its original value. Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* atm: br2684: Avoid alignment issuesPascal Hambourg2011-11-291-1/+2
| | | | | | | | Use memcmp() instead of cast to u16 when checking the PAD field. Signed-off-by: Pascal Hambourg <pascal@plouf.fr.eu.org> Signed-off-by: chas williams - CONTRACTOR <chas@cmf.nrl.navy.mil> Signed-off-by: David S. Miller <davem@davemloft.net>
* atm: br2684: Make headroom and hard_header_len depend on the payload typePascal Hambourg2011-11-291-2/+6
| | | | | | | | | | Routed payload requires less headroom than bridged payload. So do not reallocate headroom if not needed. Also, add worst case AAL5 overhead to netdev->hard_header_len. Signed-off-by: Pascal Hambourg <pascal@plouf.fr.eu.org> Signed-off-by: chas williams - CONTRACTOR <chas@cmf.nrl.navy.mil> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: optimize socket timestampingEric Dumazet2011-11-291-15/+12
| | | | | | | | We can test/set multiple bits from sk_flags at once, to shorten a bit socket setup/dismantle phase. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: dont call jump_label_dec from irq contextEric Dumazet2011-11-293-3/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Igor Maravic reported an error caused by jump_label_dec() being called from IRQ context : BUG: sleeping function called from invalid context at kernel/mutex.c:271 in_atomic(): 1, irqs_disabled(): 0, pid: 0, name: swapper 1 lock held by swapper/0: #0: (&n->timer){+.-...}, at: [<ffffffff8107ce90>] call_timer_fn+0x0/0x340 Pid: 0, comm: swapper Not tainted 3.2.0-rc2-net-next-mpls+ #1 Call Trace: <IRQ> [<ffffffff8104f417>] __might_sleep+0x137/0x1f0 [<ffffffff816b9a2f>] mutex_lock_nested+0x2f/0x370 [<ffffffff810a89fd>] ? trace_hardirqs_off+0xd/0x10 [<ffffffff8109a37f>] ? local_clock+0x6f/0x80 [<ffffffff810a90a5>] ? lock_release_holdtime.part.22+0x15/0x1a0 [<ffffffff81557929>] ? sock_def_write_space+0x59/0x160 [<ffffffff815e936e>] ? arp_error_report+0x3e/0x90 [<ffffffff810969cd>] atomic_dec_and_mutex_lock+0x5d/0x80 [<ffffffff8112fc1d>] jump_label_dec+0x1d/0x50 [<ffffffff81566525>] net_disable_timestamp+0x15/0x20 [<ffffffff81557a75>] sock_disable_timestamp+0x45/0x50 [<ffffffff81557b00>] __sk_free+0x80/0x200 [<ffffffff815578d0>] ? sk_send_sigurg+0x70/0x70 [<ffffffff815e936e>] ? arp_error_report+0x3e/0x90 [<ffffffff81557cba>] sock_wfree+0x3a/0x70 [<ffffffff8155c2b0>] skb_release_head_state+0x70/0x120 [<ffffffff8155c0b6>] __kfree_skb+0x16/0x30 [<ffffffff8155c119>] kfree_skb+0x49/0x170 [<ffffffff815e936e>] arp_error_report+0x3e/0x90 [<ffffffff81575bd9>] neigh_invalidate+0x89/0xc0 [<ffffffff81578dbe>] neigh_timer_handler+0x9e/0x2a0 [<ffffffff81578d20>] ? neigh_update+0x640/0x640 [<ffffffff81073558>] __do_softirq+0xc8/0x3a0 Since jump_label_{inc|dec} must be called from process context only, we must defer jump_label_dec() if net_disable_timestamp() is called from interrupt context. Reported-by: Igor Maravic <igorm@etf.rs> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* NET: NETROM: Fix formatting.Ralf Baechle2011-11-291-1/+2
| | | | | | | The Linux coding style wants the return statement on its own line. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* NET: NETROM: Cleanup argument SIOCADDRT ioctl argument checking.Ralf Baechle2011-11-291-4/+2
| | | | | | | | | | nr_route.ndigis is unsigned int so the nr_route.ndigis < 0 expression is never true and can be dropped. Doing the nr_ax25_dev_get call later allows the nr_route.ndigis test to bail out without having to dev_put. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Cc: Thomas Osterried <thomas@osterried.de> Signed-off-by: David S. Miller <davem@davemloft.net>
* NET: NETROM: When adding a route verify length of mnemonic string.Ralf Baechle2011-11-291-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | struct nr_route_struct's mnemonic permits a string of up to 7 bytes to be used. If userland passes a not zero terminated string to the kernel adding a node to the routing table might result in the kernel attempting to read copy a too long string. Mnemonic is part of the NET/ROM routing protocol; NET/ROM routing table updates only broadcast 6 bytes. The 7th byte in the mnemonic array exists only as a \0 termination character for the kernel code's convenience. Fixed by rejecting mnemonic strings that have no terminating \0 in the first 7 characters. Do this test only NETROM_NODE to avoid breaking NETROM_NEIGH where userland might passing an uninitialized mnemonic field. Initial patch by Dan Carpenter <dan.carpenter@oracle.com>. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Walter Harms <wharms@bfs.de> Cc: Thomas Osterried <thomas@osterried.de> Acked-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* NET: AX.25: Check ioctl arguments to avoid overflows further down the road.Ralf Baechle2011-11-291-6/+11
| | | | | | | | | | | | | | | | | | | | Very large, nonsenical arguments or use in very extreme conditions could result in integer overflows. Check ioctls arguments to avoid such overflows and return -EINVAL for too large arguments. To allow the use of AX.25 for even the most extreme setup (think packet radio to the Phase 5E mars probe) we make no further attempt to clamp the argument range. Originally reported by Fan Long <longfancn@gmail.com> and a first patch was sent by Xi Wang <xi.wang@gmail.com>. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Cc: Xi Wang <xi.wang@gmail.com> Cc: Joerg Reuter <jreuter@yaina.de> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Thomas Osterried <thomas@osterried.de> Signed-off-by: David S. Miller <davem@davemloft.net>
* dsa: Move switch drivers to new directory drivers/net/dsaBen Hutchings2011-11-297-1859/+1
| | | | | | | | Support for specific hardware belongs under drivers/net/ not net/. Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Acked-by: Lennert Buytenhek <buytenh@wantstofly.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* dsa: Move all definitions needed by drivers into <net/dsa.h>Ben Hutchings2011-11-295-95/+4
| | | | | | | | | | Any headers included by drivers should be under include/, and any definitions they use are not really private to the core as the name "dsa_priv.h" suggests. Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Acked-by: Lennert Buytenhek <buytenh@wantstofly.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* dsa: Remove unnecessary exportsBen Hutchings2011-11-291-2/+0
| | | | | | | | | I mistakenly exported functions from slave.c that are only called from dsa.c, part of the same module. Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Acked-by: Lennert Buytenhek <buytenh@wantstofly.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge branch 'for-davem' of ↵David S. Miller2011-11-2827-413/+785
|\ | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next
| * Merge branch 'master' of ↵John W. Linville2011-11-2827-413/+785
| |\ | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next into for-davem
| | * Merge branch 'master' of ↵John W. Linville2011-11-22209-345/+682
| | |\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux Conflicts: drivers/net/wireless/iwlegacy/iwl-debugfs.c drivers/net/wireless/iwlegacy/iwl-rx.c drivers/net/wireless/iwlegacy/iwl-scan.c drivers/net/wireless/iwlegacy/iwl-tx.c include/net/bluetooth/bluetooth.h
| | * | mac80211: fix RCU warnings in meshJohannes Berg2011-11-211-22/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Sparse RCU checking reports two warnings in the mesh path table code. These are due to questionable uses of rcu_dereference. To fix the first one, get rid of mesh_gate_add() and just make mesh_path_add_gate() do the correct deref. To fix the second one, simply remove rcu_dereference() in mesh_gate_del() -- it already gets a proper pointer as indicated by the prototype (no __rcu annotation) and confirmed by the code. Cc: Javier Cardona <javier@cozybit.com> Cc: Thomas Pedersen <thomas@cozybit.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | cfg80211: work around a sparse issueJohannes Berg2011-11-211-11/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | sparse reports: net/wireless/util.c:499:30: error: cannot size expression net/wireless/util.c:503:30: error: cannot size expression This is evidently due to the EXPORT_SYMBOL() of the bridge_tunnel_header and rfc1042 header variables. Move them to the end of the file to work around the sparse issue. The error itself from sparse can be ignored safely, but since sparse stops parsing at errors, other issues after this would go undetected. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: Support ht-cap over-rides.Ben Greear2011-11-219-18/+140
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This implements ht-cap over-rides for mac80211 drivers. HT may be disabled, making an /a/b/g/n station act like an a/b/g station. HT40 may be disabled forcing the station to be HT20 even if the AP and local hardware support HT40. MAX-AMSDU may be disabled. AMPDU-Density may be increased. AMPDU-Factor may be decreased. This has been successfully tested with ath9k using patched wpa_supplicant and iw. Signed-off-by: Ben Greear <greearb@candelatech.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | wireless: Support ht-capabilities over-rides.Ben Greear2011-11-214-7/+91
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This allows users to disable features such as HT, HT40, and to modify the MCS, AMPDU, and AMSDU settings for drivers that support it. The MCS, AMPDU, and AMSDU features that may be disabled are are reported in the phy-info netlink message as a mask. Attemping to disable features that are not supported will take no affect, but will not return errors. This is to aid backwards compatibility in user-space apps that may not be clever enough to deal with parsing the the capabilities mask. This patch only enables the infrastructure. An additional patch will enable the feature in mac80211. Signed-off-by: Ben Greear <greearb@candelatech.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | cfg80211/mac80211: Revert "move information element parsing logic to cfg80211"Johannes Berg2011-11-213-168/+230
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | No other driver ever ended up using this, and the commit forgot to move the prototype so no driver could have used it. Revert it, if any driver shows up and needs it it can be moved again, but until then it's more efficient to have it in mac80211 where the only user is. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | nl80211: make get_vlan logic more commonJohannes Berg2011-11-211-23/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | get_vlan() sets the output parameter even if it returns an error, which is a bit odd. Instead, convert it to use ERR_PTR. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | cfg80211: add flags for off-channel capabilitiesJohannes Berg2011-11-212-6/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently mac80211 implements these for all devices, but given restrictions of some devices that isn't really true, so prepare for being able to remove the capability for some mac80211 devices. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Acked-by: Kalle Valo <kvalo@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: minor cleanup to mesh state lockingJohannes Berg2011-11-211-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | First time I tried smatch, and it says: mesh_hwmp.c +870 mesh_queue_preq(21) error: double lock 'bottom_half:' mesh_hwmp.c +873 mesh_queue_preq(24) error: double unlock 'bottom_half:' mesh_hwmp.c +886 mesh_queue_preq(37) error: double unlock 'bottom_half:' Which is indeed true -- there's no point in disabling BHs again if we just did that a few lines earlier, so remove. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | nl80211: add testmode to the list of supported commandsKalle Valo2011-11-211-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | User space might want to test if driver supports testmode. Adding testmode to the list of supported commands makes this easier. I omitted testmode_dump() in purpose. I assume all drivers implementing testmode_dump() will also implement testmode_cmd(). Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: remove unused ASSOC_AP flagJohannes Berg2011-11-213-5/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | WLAN_STA_ASSOC_AP indicates that the station entry is for an AP we're associated to but isn't used so remove it. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: transmit fragment list to driversJohannes Berg2011-11-213-35/+69
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Drivers can usually handle fragmented packets much easier when they get the entire list of fragments at once. The only thing they need to do is keep enough space on the queues for up to ten fragments of a single MSDU. This allows them to implement this with a new operation tx_frags. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: make TX LED handling independent of fragmentationJohannes Berg2011-11-211-12/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This just prepares for passing the entire fragment list to the driver. No significant changes, but the TX throughput is calculated slightly differently now and we blink only once for each MSDU. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: move fragment flag adjustmentJohannes Berg2011-11-211-9/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Instead of adjusting the fragment flags at TX time, adjust them at fragmentation time. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: use skb list for fragmentsJohannes Berg2011-11-215-71/+82
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We are currently linking the skbs by using skb->next directly. This works, but the preferred way is to use a struct sk_buff_head instead. That also prepares for passing that to drivers directly. While at it I noticed we calculate the duration for fragments twice -- remove one of them. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | cfg80211: pass DFS region to drivers through reg_notifier()Luis R. Rodriguez2011-11-211-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This grants drivers access to the DFS region that a regulatory domain belongs to. Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | cfg80211: process regulatory DFS region for countriesLuis R. Rodriguez2011-11-213-0/+53
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The wireless-regdb now has support for mapping a country to one DFS region. CRDA sends this to us now so process it so we can provide that hint to drivers. This will later be used by code for processing DFS in a way that meets the criteria for the DFS region the country belongs to. Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: call ieee80211_recalc_idle() after sending packetsFelix Fietkau2011-11-172-8/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some drivers (e.g. ath9k) assume that it's safe to go into low-power mode immediately after the idle state changes. To support that, mac80211 even calls drv_flush() before that happens. In some instances, mac80211 sent a packet right after recalculating the idle state, this patch fixes that. Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | minstrel_ht: Remove unused function parametersPatrick Kelle2011-11-171-14/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove unused function parameters in the following functions: minstrel_calc_rate_ewma() minstrel_ht_calc_tp() minstrel_aggr_check() minstrel_ht_set_rate() Signed-off-by: Patrick Kelle <patrick.kelle81@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: memory leak in mesh_queue_preq()Dan Carpenter2011-11-171-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We recently introduced a return here, but we need to call kfree first. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: use kfree_skb() instead of kfree()Dan Carpenter2011-11-171-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | sk_buff structs should be freed using kfree_skb(). This was introduced recently in 029458212 "mac80211: Save probe response data for bss". Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: Get rid of search loop for rate group indexHelmut Schaa2011-11-171-16/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Finding the group index for a specific rate is done by looping through all groups and returning if the correct one is found. This code is called for each tx'ed frame and thus it makes sense to reduce its runtime. Do this by calculating the group index by this formula based on the SGI and HT40 flags as well as the stream number: idx = (HT40 * 2 * MINSTREL_MAX_STREAMS) + (SGI * MINSTREL_MAX_STREAMS) + (streams - 1) Hence, the groups are ordered by th HT40 flag first, then by the SGI flag and afterwards by the number of used streams. This should reduce the runtime of minstrel_ht_get_group_idx considerable. Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: Check rate->idx before rate->countHelmut Schaa2011-11-171-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The drivers are not required to fill in rate->count if rate->idx is set to -1. Hence, we should first check rate->idx before accessing rate->count. Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: remove crypto special case for auth framesJohannes Berg2011-11-171-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The shared key authentication frame that needs to be encrypted (the third one in the shared key handshake) is directly encrypted in ieee80211_send_auth and the IEEE80211_TX_INTFL_DONT_ENCRYPT is set. All others are not encrypted, so the only way to get to this is erroneously on no-monitor AP side. Remove the special case for authentication frames to fix the AP shared key side when operating without cooked monitor interfaces -- with cooked monitor the IEEE80211_TX_INTFL_DONT_ENCRYPT also gets set, so we never get here -- an AP never encrypts auth frames. Without this patch, an AP operating in WEP mode with my no-monitor patches would erroneously encrypt all authentication frames, instead of none. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
* | | | sch_sfb: use skb_flow_dissect()Eric Dumazet2011-11-281-3/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Current SFB double hashing is not fulfilling SFB theory, if two flows share same rxhash value. Using skb_flow_dissect() permits to really have better hash dispersion, and get tunnelling support as well. Double hashing point was mentioned by Florian Westphal Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | cls_flow: use skb_flow_dissect()Eric Dumazet2011-11-281-132/+48
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Instead of using a custom flow dissector, use skb_flow_dissect() and benefit from tunnelling support. This lack of tunnelling support was mentioned by Dan Siemon. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | net: use skb_flow_dissect() in __skb_get_rxhash()Eric Dumazet2011-11-281-111/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | No functional changes. This uses the code we factorized in skb_flow_dissect() Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | net: introduce skb_flow_dissect()Eric Dumazet2011-11-282-1/+135
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We use at least two flow dissectors in network stack, with known limitations and code duplication. Introduce skb_flow_dissect() to factorize this, highly inspired from existing dissector from __skb_get_rxhash() Note : We extensively use skb_header_pointer(), this permits us to not touch skb at all. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | tcp: tcp_sendmsg() wrong access to sk_route_capsEric Dumazet2011-11-281-4/+4
|/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | Now sk_route_caps is u64, its dangerous to use an integer to store result of an AND operator. It wont work if NETIF_F_SG is moved on the upper part of u64. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> CC: Michał Mirosław <mirq-linux@rere.qmqm.pl> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | tcp: skip cwnd moderation in TCP_CA_Open in tcp_try_to_openNeal Cardwell2011-11-271-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The problem: Senders were overriding cwnd values picked during an undo by calling tcp_moderate_cwnd() in tcp_try_to_open(). The fix: Don't moderate cwnd in tcp_try_to_open() if we're in TCP_CA_Open, since doing so is generally unnecessary and specifically would override a DSACK-based undo of a cwnd reduction made in fast recovery. Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | tcp: allow undo from reordered DSACKsNeal Cardwell2011-11-271-13/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously, SACK-enabled connections hung around in TCP_CA_Disorder state while snd_una==high_seq, just waiting to accumulate DSACKs and hopefully undo a cwnd reduction. This could and did lead to the following unfortunate scenario: if some incoming ACKs advance snd_una beyond high_seq then we were setting undo_marker to 0 and moving to TCP_CA_Open, so if (due to reordering in the ACK return path) we shortly thereafter received a DSACK then we were no longer able to undo the cwnd reduction. The change: Simplify the congestion avoidance state machine by removing the behavior where SACK-enabled connections hung around in the TCP_CA_Disorder state just waiting for DSACKs. Instead, when snd_una advances to high_seq or beyond we typically move to TCP_CA_Open immediately and allow an undo in either TCP_CA_Open or TCP_CA_Disorder if we later receive enough DSACKs. Other patches in this series will provide other changes that are necessary to fully fix this problem. Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | tcp: use SACKs and DSACKs that arrive on ACKs below snd_unaNeal Cardwell2011-11-271-3/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The bug: When the ACK field is below snd_una (which can happen when ACKs are reordered), senders ignored DSACKs (preventing undo) and did not call tcp_fastretrans_alert, so they did not increment prr_delivered to reflect newly-SACKed sequence ranges, and did not call tcp_xmit_retransmit_queue, thus passing up chances to send out more retransmitted and new packets based on any newly-SACKed packets. The change: When the ACK field is below snd_una (the "old_ack" goto label), call tcp_fastretrans_alert to allow undo based on any newly-arrived DSACKs and try to send out more packets based on newly-SACKed packets. Other patches in this series will provide other changes that are necessary to fully fix this problem. Signed-off-by: Neal Cardwell <ncardwell@google.com> Acked-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | tcp: use DSACKs that arrive when packets_out is 0Neal Cardwell2011-11-271-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The bug: Senders ignored DSACKs after recovery when there were no outstanding packets (a common scenario for HTTP servers). The change: when there are no outstanding packets (the "no_queue" goto label), call tcp_fastretrans_alert() in order to use DSACKs to undo congestion window reductions. Other patches in this series will provide other changes that are necessary to fully fix this problem. Signed-off-by: Neal Cardwell <ncardwell@google.com> Acked-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | tcp: make is_dupack a parameter to tcp_fastretrans_alert()Neal Cardwell2011-11-271-4/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow callers to decide whether an ACK is a duplicate ACK. This is a prerequisite to allowing fastretrans_alert to be called from new contexts, such as the no_queue and old_ack code paths, from which we have extra info that tells us whether an ACK is a dupack. Signed-off-by: Neal Cardwell <ncardwell@google.com> Acked-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | atm: eliminate atm_guess_pdu2truesize()chas williams - CONTRACTOR2011-11-261-1/+1
| | | | | | | | | | | | | | | Signed-off-by: Chas Williams - CONTRACTOR <chas@cmf.nrl.navy.mil> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | dsa: Allow core and drivers to be built as modulesBen Hutchings2011-11-262-7/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Change the kconfig types to tristate and adjust the condition for declaring net_device::dsa_ptr to allow for this. Adjust the makefile so that if NET_DSA_MV88E6123_61_65=y and NET_DSA_MV88E6131=m or vice versa then both drivers are built-in. We could leave these options as bool and make NET_DSA_MV88E6XXX a user-selected option, but that would break existing configurations. Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net>
OpenPOWER on IntegriCloud