summaryrefslogtreecommitdiffstats
path: root/net (follow)
Commit message (Collapse)AuthorAgeFilesLines
* CAN: use dev_get_by_index_rcustephen hemminger2009-11-111-3/+3
| | | | | | | | | Use new function to avoid doing read_lock(). Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: Oliver Hartkopp <oliver@hartkopp.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* IPV4: use rcu to walk list of devices in IGMPstephen hemminger2009-11-111-8/+10
| | | | | | | | This also needs to be optimized for large number of devices. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* decnet: use RCU to find network devicesstephen hemminger2009-11-111-8/+6
| | | | | | | | When showing device statistics use RCU rather than read_lock(&dev_base_lock) Compile tested only. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: use rcu for network scheduler APIstephen hemminger2009-11-111-3/+4
| | | | | | | | | Use RCU to walk list of network devices in qdisc dump. This could be optimized for large number of devices. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* vlan: eliminate use of dev_base_lockstephen hemminger2009-11-111-7/+6
| | | | | | | | Do not need to use read_lock(&dev_base_lock), use RCU instead. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* IPv6: use ipv6_addr_v4mapped()Brian Haley2009-11-111-1/+1
| | | | | | | | | Change udp6_portaddr_hash() to use ipv6_addr_v4mapped() inline instead of ipv6_addr_type(). Signed-off-by: Brian Haley <brian.haley@hp.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sit: Clean up DF code by copying from IPIPHerbert Xu2009-11-111-21/+21
| | | | | | | | | | | | | This patch rearranges the SIT DF bit handling using the new IPIP DF code. The only externally visible effect should be the case where PMTU is enabled and the MTU is exactly 1280 bytes. In this case the previous code would send packets out with DF off while the new code would set the DF bit. This is inline with RFC 4213. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Thanks, Signed-off-by: David S. Miller <davem@davemloft.net>
* ipv6: Allow inet6_dump_addr() to handle more than 64 addressesEric Dumazet2009-11-111-0/+6
| | | | | | | | | Apparently, inet6_dump_addr() is not able to handle more than 64 ipv6 addresses per device. We must break from inner loops in case skb is full, or else cursor is put at the end of list. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* ipv6: speedup inet6_dump_ifinfo()Eric Dumazet2009-11-111-17/+28
| | | | | | | | | | | When handling large number of netdevice, inet6_dump_ifinfo() is very slow because it has O(N^2) complexity. Instead of scanning one single list, we can use the 256 sub lists of the dev_index hash table, and RCU lookups. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: netlink_getname, packet_getname -- use DECLARE_SOCKADDR guardCyrill Gorcunov2009-11-113-3/+3
| | | | | | | | Use guard DECLARE_SOCKADDR in a few more places which allow us to catch if the structure copied back is too big. Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* udp: bind() optimisationEric Dumazet2009-11-112-15/+72
| | | | | | | | | UDP bind() can be O(N^2) in some pathological cases. Thanks to secondary hash tables, we can make it O(N) Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* Phonet: allocate and copy for pipe TX without sock lockRémi Denis-Courmont2009-11-111-17/+12
| | | | | Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* Phonet: put sockets in a hash tableRémi Denis-Courmont2009-11-112-25/+55
| | | | | Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge branch 'master' of ↵David S. Miller2009-11-0926-699/+466
|\ | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6
| * mac80211: async station powersave handlingJohannes Berg2009-11-066-85/+168
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some devices require that all frames to a station are flushed when that station goes into powersave mode before being able to send frames to that station again when it wakes up or polls -- all in order to avoid reordering and too many or too few frames being sent to the station when it polls. Normally, this is the case unless the station goes to sleep and wakes up very quickly again. But in that case, frames for it may be pending on the hardware queues, and thus races could happen in the case of multiple hardware queues used for QoS/WMM. Normally this isn't a problem, but with the iwlwifi mechanism we need to make sure the race doesn't happen. This makes mac80211 able to cope with the race with driver help by a new WLAN_STA_PS_DRIVER per-station flag that can be controlled by the driver and tells mac80211 whether it can transmit frames or not. This flag must be set according to very specific rules outlined in the documentation for the function that controls it. When we buffer new frames for the station, we normally set the TIM bit right away, but while the driver has blocked transmission to that sta we need to avoid that as well since we cannot respond to the station if it wakes up due to the TIM bit. Once the driver unblocks, we can set the TIM bit. Similarly, when the station just wakes up, we need to wait until all other frames are flushed before we can transmit frames to that station, so the same applies here, we need to wait for the driver to give the OK. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: fix internal scan requestJohannes Berg2009-11-051-0/+1
| | | | | | | | | | | | | | | | | | | | | | The internal scan request mac80211 uses to scan for IBSS networks was set up to contain no channels at all because n_channels wasn't set after setting up the channels array. Fix this to properly scan for networks. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: Fix IBSS mergeSujith2009-11-051-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | Currently, in IBSS mode, a single creator would go into a loop trying to merge/scan. This happens because the IBSS timer is rearmed on finishing a scan and the subsequent timer invocation requests another scan immediately. This patch fixes this issue by checking if we have just completed a scan run trying to merge with other IBSS networks. Signed-off-by: Sujith <Sujith.Manoharan@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: make ieee80211_find_sta per virtual interfaceJohannes Berg2009-11-051-2/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Since we have a TODO item to make all station management dependent on virtual interfaces, I figured I'd start with pushing such a change to drivers before more drivers start using the ieee80211_find_sta() API with a hw pointer and cause us grief later on. For now continue exporting the old API in form of ieee80211_find_sta_by_hw(), but discourage its use strongly. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * cfg80211: remove dead variableJohannes Berg2009-11-051-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commit 211a4d12abf86fe0df4cd68fc6327cbb58f56f81 Author: Johannes Berg <johannes@sipsolutions.net> Date: Tue Oct 20 15:08:53 2009 +0900 cfg80211: sme: deauthenticate on assoc failure accidentally introduced a dead variable, I had changed the code to not need it while creating the patch and it looks like I forgot to remove the variable (and nobody else noticed either). Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * cfg80211: validate scan channelsJohannes Berg2009-11-023-18/+35
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently it is possible to request a scan on only disabled channels, which could be problematic for some drivers. Reject such scans, and also ignore disabled channels that are given. This resuls in the scan begin/end event only including channels that are actually used. This makes the mac80211 check for disabled channels superfluous. At the same time, remove the no-IBSS check from mac80211 -- nothing says that we should not find any networks on channels that cannot be used for an IBSS, even when operating in IBSS mode. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * cfg80211: Fix WEXT compat siwauth wpa and group cipherJouni Malinen2009-11-021-4/+2
| | | | | | | | | | | | | | | | | | Neither of these commands should clear the current configuration value if they return error. Furthermore, cfg80211_set_cipher_group() should be able to handle IW_AUTH_CIPHER_NONE without reporting error. Signed-off-by: Jouni Malinen <j@w1.fi> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: fix scan abort sanity checksJohannes Berg2009-11-021-4/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Since sometimes mac80211 queues up a scan request to only act on it later, it must be allowed to (internally) cancel a not-yet-running scan, e.g. when the interface is taken down. This condition was missing since we always checked only the local->scanning variable which isn't yet set in that situation. Reported-by: Luis R. Rodriguez <mcgrof@gmail.com> Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: make CALL_TXH a statementJohannes Berg2009-11-021-16/+18
| | | | | | | | | | | | | | | | | | The multi-line code in this macro wasn't wrapped in do {} while (0) so we cannot use it in an if() branch safely in the future -- fix that. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: remove bogus codeJohannes Berg2009-10-301-7/+0
| | | | | | | | | | | | | | | | | | It's not right to do something here when returning an error, and hostapd should never have relied on it as it only fixes up a small part of the problem anyway. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: remove sent_ps_bufferedJohannes Berg2009-10-302-6/+3
| | | | | | | | | | | | | | | | | | | | | | | | This variable is set once, and tested once. However, the code path that can set it is mutually exclusive with the code path that tests it, so the test is always true. Thus we also don't need to set it either and can just remove the variable. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: also drop qos-nullfunc frames silentlyJohannes Berg2009-10-301-5/+10
| | | | | | | | | | | | | | | | | | We drop nullfunc frames, but not qos-nullfunc frames, even though those could be used for PS state control as well. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: unconditionally set IEEE80211_TX_CTL_SEND_AFTER_DTIMJohannes Berg2009-10-301-4/+3
| | | | | | | | | | | | | | | | | | | | | | | | When mac80211 is asked to buffer multicast frames in AP mode, it will not set the flag indicating that the frames should be sent after the DTIM beacon for those frames buffered in software. Fix this little inconsistency by always setting that flag in the buffering code path. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: deprecate qual valueJohannes Berg2009-10-303-5/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This value is unused by mac80211, because it was only be used by wireless extensions, and turned out to not be useful there because the quality value needs to be comparable between scan results and the current value which is impossible when the qual value is calculated taking into account noise, for example. Since it is unused anyway, this patch deprecates it in the hope that drivers will remove their sometimes quite expensive calculations of the value. I'm open to actual uses of the value, but the best way of using it seems to be what the Intel drivers do which should probably be generalised if we have noise values from the hardware. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: introduce ieee80211_beacon_get_tim()Johannes Berg2009-10-301-3/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | Compared to ieee80211_beacon_get(), the new function ieee80211_beacon_get_tim() returns information on the location and length of the TIM IE, which some drivers need in order to generate the TIM on the device. The old function, ieee80211_beacon_get(), becomes a small static inline wrapper around the new one to not break all drivers. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: make align adjustment code support paged SKBZhu Yi2009-10-301-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | This fixed a BUG_ON in __skb_trim() when paged rx is used in iwlwifi driver. Yes, the whole mac80211 stack doesn't support paged SKB yet. But let's start the work slowly from small code snippets. Reported-and-tested-by: Abhijeet Kolekar <abhijeet.kolekar@intel.com> Signed-off-by: Zhu Yi <yi.zhu@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: remove RX_FLAG_RADIOTAPJohannes Berg2009-10-302-25/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | While there may be a case for a driver adding its own bits of radiotap information, none currently does. Also, drivers would have to copy the code to generate the radiotap bits that now mac80211 generates. If some driver in the future needs to add some driver-specific information I'd expect that to be in a radiotap vendor namespace and we can add a different way of passing such data up and having mac80211 include it. Additionally, rename IEEE80211_CONF_RADIOTAP to IEEE80211_CONF_MONITOR since it's still used by b43(legacy) to obtain per-frame timestamps. The purpose of this patch is to simplify the RX code in mac80211 to make it easier to add paged skb support. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: fix radiotap header generationJohannes Berg2009-10-301-14/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In commit 601ae7f25aea58f208a7f640f6174aac0652403a Author: Bruno Randolf <br1@einfach.org> Date: Thu May 8 19:22:43 2008 +0200 mac80211: make rx radiotap header more flexible code was added that tried to align the radiotap header position in memory based on the radiotap header length. Quite obviously, that is completely useless. Instead of trying to do that, use unaligned accesses to generate the radiotap header. To properly do that, we also need to mark struct ieee80211_radiotap_header packed, but that is fine since it's already packed (and it should be marked packed anyway since its a wire format). Cc: Bruno Randolf <br1@einfach.org> Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: split hardware scan by bandJohannes Berg2009-10-303-32/+80
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There's currently a very odd bug in mac80211 -- a hardware scan that is done while the hardware is really operating on 2.4 GHz will include CCK rates in the probe request frame, even on 5 GHz (if the driver uses the mac80211 IEs). Vice versa, if the hardware is operating on 5 GHz the 2.4 GHz probe requests will not include CCK rates even though they should. Fix this by splitting up cfg80211 scan requests by band -- recalculating the IEs every time -- and requesting only per-band scans from the driver. Apparently this bug hasn't been a problem yet, but it is imaginable that some older access points get confused if confronted with such behaviour. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: remove outdated commentJohannes Berg2009-10-301-2/+0
| | | | | | | | | | | | | | | | This comment hasn't been a real TODO item for a long time now since we fixed that quite a while ago. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: fix dynamic power save for devices with nullfunc support in hwKalle Valo2009-10-301-1/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In TX path it was assumed that dynamic power save works only if IEEE80211_HW_PS_NULLFUNC_STACK is set. But is not the case, there are devices which have nullfunc support in hardware but need mac80211 to handle dynamic power save timers, TI's wl1251 is one of them. The fix is to not check for IEEE80211_HW_PS_NULLFUNC_STACK in is_dynamic_ps_enabled(), instead check IEEE80211_HW_SUPPORTS_PS and IEEE80211_HW_SUPPORTS_DYNAMIC_PS flags and act accordingly. Tested with wl1251. Signed-off-by: Kalle Valo <kalle.valo@nokia.com> Acked-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: refactor dynamic power save checkKalle Valo2009-10-301-3/+21
| | | | | | | | | | | | | | | | | | Refactor dynamic power save checks to a function of it's own for better readibility. No functional changes. Signed-off-by: Kalle Valo <kalle.valo@nokia.com> Reviewed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * cfg80211/mac80211: use debugfs_remove_recursiveJohannes Berg2009-10-3014-462/+51
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We can save a lot of code and pointers in the structs by using debugfs_remove_recursive(). First, change cfg80211 to use debugfs_remove_recursive() so that drivers do not need to clean up any files they added to the per-wiphy debugfs (if and only if they are ok to be accessed until after wiphy_unregister!). Then also make mac80211 use debugfs_remove_recursive() where necessary -- it need not remove per-wiphy files as cfg80211 now removes those, but netdev etc. files still need to be handled but can now be removed without needing struct dentry pointers to all of them. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
* | Merge branch 'master' of ↵David S. Miller2009-11-093-28/+33
|\ \ | | | | | | | | | | | | | | | | | | master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 Conflicts: drivers/net/can/usb/ems_usb.c
| * | can: should not use __dev_get_by_index() without locksEric Dumazet2009-11-081-7/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bcm_proc_getifname() is called with RTNL and dev_base_lock not held. It calls __dev_get_by_index() without locks, and this is illegal (might crash) Close the race by holding dev_base_lock and copying dev->name in the protected section. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: Oliver Hartkopp <oliver@hartkopp.net> Signed-off-by: David S. Miller <davem@davemloft.net>
| * | ipip: Fix handling of DF packets when pmtudisc is OFFHerbert Xu2009-11-071-15/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | RFC 2003 requires the outer header to have DF set if DF is set on the inner header, even when PMTU discovery is off for the tunnel. Our implementation does exactly that. For this to work properly the IPIP gateway also needs to engate in PMTU when the inner DF bit is set. As otherwise the original host would not be able to carry out its PMTU successfully since part of the path is only visible to the gateway. Unfortunately when the tunnel PMTU discovery setting is off, we do not collect the necessary soft state, resulting in blackholes when the original host tries to perform PMTU discovery. This problem is not reproducible on the IPIP gateway itself as the inner packet usually has skb->local_df set. This is not correctly cleared (an unrelated bug) when the packet passes through the tunnel, which allows fragmentation to occur. For hosts behind the IPIP gateway it is readily visible with a simple ping. This patch fixes the problem by performing PMTU discovery for all packets with the inner DF bit set, regardless of the PMTU discovery setting on the tunnel itself. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
| * | netfilter: xt_connlimit: fix regression caused by zero family valueJan Engelhardt2009-11-071-6/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit v2.6.28-rc1~717^2~109^2~2 was slightly incomplete; not all instances of par->match->family were changed to par->family. References: http://bugzilla.netfilter.org/show_bug.cgi?id=610 Signed-off-by: Jan Engelhardt <jengelh@medozas.de> Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | xfrm: SAD entries do not expire correctly after suspend-resumeYury Polyanskiy2009-11-091-13/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fixes the following bug in the current implementation of net/xfrm: SAD entries timeouts do not count the time spent by the machine in the suspended state. This leads to the connectivity problems because after resuming local machine thinks that the SAD entry is still valid, while it has already been expired on the remote server. The cause of this is very simple: the timeouts in the net/xfrm are bound to the old mod_timer() timers. This patch reassigns them to the CLOCK_REALTIME hrtimer. I have been using this version of the patch for a few months on my machines without any problems. Also run a few stress tests w/o any issues. This version of the patch uses tasklet_hrtimer by Peter Zijlstra (commit 9ba5f0). This patch is against 2.6.31.4. Please CC me. Signed-off-by: Yury Polyanskiy <polyanskiy@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | net/compat_ioctl: support SIOCWANDEVArnd Bergmann2009-11-091-0/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This adds compat_ioctl support for SIOCWANDEV, which has always been missing. The definition of struct compat_ifreq was missing an ifru_settings fields that is needed to support SIOCWANDEV, so add that and clean up the whitespace damage in the struct definition. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | net, compat_ioctl: fix SIOCGMII ioctlsArnd Bergmann2009-11-091-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | SIOCGMIIPHY and SIOCGMIIREG return data through ifreq, so it needs to be converted on the way out as well. SIOCGIFPFLAGS is unused, but has the same problem in theory. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | udp: multicast RX should increment SNMP/sk_drops counter in allocation failuresEric Dumazet2009-11-092-2/+18
| | | | | | | | | | | | | | | | | | | | | When skb_clone() fails, we should increment sk_drops and SNMP counters. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | ipv6: udp: Optimise multicast receptionEric Dumazet2009-11-091-24/+47
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | IPV6 UDP multicast rx path is a bit complex and can hold a spinlock for a long time. Using a small (32 or 64 entries) stack of socket pointers can help to perform expensive operations (skb_clone(), udp_queue_rcv_skb()) outside of the lock, in most cases. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | ipv4: udp: Optimise multicast receptionEric Dumazet2009-11-091-26/+50
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | UDP multicast rx path is a bit complex and can hold a spinlock for a long time. Using a small (32 or 64 entries) stack of socket pointers can help to perform expensive operations (skb_clone(), udp_queue_rcv_skb()) outside of the lock, in most cases. It's also a base for a future RCU conversion of multicast recption. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: Lucian Adrian Grijincu <lgrijincu@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | ipv6: udp: optimize unicast RX pathEric Dumazet2009-11-091-3/+109
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We first locate the (local port) hash chain head If few sockets are in this chain, we proceed with previous lookup algo. If too many sockets are listed, we take a look at the secondary (port, address) hash chain. We choose the shortest chain and proceed with a RCU lookup on the elected chain. But, if we chose (port, address) chain, and fail to find a socket on given address, we must try another lookup on (port, in6addr_any) chain to find sockets not bound to a particular IP. -> No extra cost for typical setups, where the first lookup will probabbly be performed. RCU lookups everywhere, we dont acquire spinlock. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | ipv4: udp: optimize unicast RX pathEric Dumazet2009-11-091-3/+112
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We first locate the (local port) hash chain head If few sockets are in this chain, we proceed with previous lookup algo. If too many sockets are listed, we take a look at the secondary (port, address) hash chain we added in previous patch. We choose the shortest chain and proceed with a RCU lookup on the elected chain. But, if we chose (port, address) chain, and fail to find a socket on given address, we must try another lookup on (port, INADDR_ANY) chain to find socket not bound to a particular IP. -> No extra cost for typical setups, where the first lookup will probabbly be performed. RCU lookups everywhere, we dont acquire spinlock. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | udp: secondary hash on (local port, local address)Eric Dumazet2009-11-091-5/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Extends udp_table to contain a secondary hash table. socket anchor for this second hash is free, because UDP doesnt use skc_bind_node : We define an union to hold both skc_bind_node & a new hlist_nulls_node udp_portaddr_node udp_lib_get_port() inserts sockets into second hash chain (additional cost of one atomic op) udp_lib_unhash() deletes socket from second hash chain (additional cost of one atomic op) Note : No spinlock lockdep annotation is needed, because lock for the secondary hash chain is always get after lock for primary hash chain. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-18 14:51:21 +0100