aboutsummaryrefslogtreecommitdiffstats
path: root/include
Commit message (Collapse)AuthorAge
* ipv4: Pass explicit daddr arg to ip_send_reply().David S. Miller2011-05-10
| | | | | | This eliminates an access to rt->rt_src. Signed-off-by: David S. Miller <davem@davemloft.net>
* net: add mac_pton() for parsing MAC addressAlexey Dobriyan2011-05-09
| | | | | | | | | | | | | | | | | | | | | | mac_pton() parses MAC address in form XX:XX:XX:XX:XX:XX and only in that form. mac_pton() doesn't dirty result until it's sure string representation is valid. mac_pton() doesn't care about characters _after_ last octet, it's up to caller to deal with it. mac_pton() diverges from 0/-E return value convention. Target usage: if (!mac_pton(str, whatever->mac)) return -EINVAL; /* ->mac being u8 [ETH_ALEN] is filled at this point. */ /* optionally check str[3 * ETH_ALEN - 1] for termination */ Use mac_pton() in pktgen and netconsole for start. Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* vlan: remove one synchronize_net() callEric Dumazet2011-05-09
| | | | | | | | | | | | | | | | | | At VLAN dismantle phase, unregister_vlan_dev() makes one synchronize_net() call after vlan_group_set_device(grp, vlan_id, NULL). This call can be safely removed because we are calling unregister_netdevice_queue() to queue device for deletion, and this process needs at least one rcu grace period to complete. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Cc: Ben Greear <greearb@candelatech.com> Cc: Patrick McHardy <kaber@trash.net> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Jesse Gross <jesse@nicira.com> Cc: Michał Mirosław <mirq-linux@rere.qmqm.pl> Acked-by: Jesse Gross <jesse@nicira.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* garp: remove one synchronize_rcu() callEric Dumazet2011-05-09
| | | | | | | | | | | | Speedup vlan dismantling in CONFIG_VLAN_8021Q_GVRP=y cases, by using a call_rcu() to free the memory instead of waiting with expensive synchronize_rcu() [ while RTNL is held ] Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Cc: Ben Greear <greearb@candelatech.com> Cc: Patrick McHardy <kaber@trash.net> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* ipv4: Pass flow key down into ip_append_*().David S. Miller2011-05-09
| | | | | | This way rt->rt_dst accesses are unnecessary. Signed-off-by: David S. Miller <davem@davemloft.net>
* ipv4: Pass flow keys down into datagram packet building engine.David S. Miller2011-05-09
| | | | | | | | | This way ip_output.c no longer needs rt->rt_{src,dst}. We already have these keys sitting, ready and waiting, on the stack or in a socket structure. Signed-off-by: David S. Miller <davem@davemloft.net>
* net: Allow ethtool to set interface in loopback mode.Mahesh Bandewar2011-05-08
| | | | | | | | | | | | | | | | | | | | | | | | | This patch enables ethtool to set the loopback mode on a given interface. By configuring the interface in loopback mode in conjunction with a policy route / rule, a userland application can stress the egress / ingress path exposing the flows of the change in progress and potentially help developer(s) understand the impact of those changes without even sending a packet out on the network. Following set of commands illustrates one such example - a) ip -4 addr add 192.168.1.1/24 dev eth1 b) ip -4 rule add from all iif eth1 lookup 250 c) ip -4 route add local 0/0 dev lo proto kernel scope host table 250 d) arp -Ds 192.168.1.100 eth1 e) arp -Ds 192.168.1.200 eth1 f) sysctl -w net.ipv4.ip_nonlocal_bind=1 g) sysctl -w net.ipv4.conf.all.accept_local=1 # Assuming that the machine has 8 cores h) taskset 000f netserver -L 192.168.1.200 i) taskset 00f0 netperf -t TCP_CRR -L 192.168.1.100 -H 192.168.1.200 -l 30 Signed-off-by: Mahesh Bandewar <maheshb@google.com> Acked-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* ethtool: Add 20G bit definitionsYaniv Rosner2011-05-08
| | | | | | | | | | | | Add 20G supported and advertising bit definitions. 20G will be supported with the 57840 chips. Signed-off-by: Yaniv Rosner <yanivr@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> ------ include/linux/ethtool.h | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) Signed-off-by: David S. Miller <davem@davemloft.net>
* inet: Pass flowi to ->queue_xmit().David S. Miller2011-05-08
| | | | | | | | | | | This allows us to acquire the exact route keying information from the protocol, however that might be managed. It handles all of the possibilities, from the simplest case of storing the key in inet->cork.fl to the more complex setup SCTP has where individual transports determine the flow. Signed-off-by: David S. Miller <davem@davemloft.net>
* ipv4: Create inet_csk_route_child_sock().David S. Miller2011-05-08
| | | | | | | | | | | | This is just like inet_csk_route_req() except that it operates after we've created the new child socket. In this way we can use the new socket's cork flow for proper route key storage. This will be used by DCCP and TCP child socket creation handling. Signed-off-by: David S. Miller <davem@davemloft.net>
* sctp: Store a flowi in transports to provide persistent keying.David S. Miller2011-05-08
| | | | | | | | | Several future simplifications are possible now because of this. For example, the sctp_addr unions can simply refer directly to the flowi information. Signed-off-by: David S. Miller <davem@davemloft.net>
* ethtool: remove phys_id from ethtool_opsStephen Hemminger2011-05-08
| | | | | | | | After that all the upstream kernel drivers now use phys_id, and the old ethtool_ops interface (phys_id) can be removed. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* inet: Decrease overhead of on-stack inet_cork.David S. Miller2011-05-06
| | | | | | | | | | | | | | | | When we fast path datagram sends to avoid locking by putting the inet_cork on the stack we use up lots of space that isn't necessary. This is because inet_cork contains a "struct flowi" which isn't used in these code paths. Split inet_cork to two parts, "inet_cork" and "inet_cork_full". Only the latter of which has the "struct flowi" and is what is stored in inet_sock. Signed-off-by: David S. Miller <davem@davemloft.net> Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
* Merge branch 'master' of ↵David S. Miller2011-05-05
|\ | | | | | | | | | | | | master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 Conflicts: drivers/net/tg3.c
| * usbnet: Resubmit interrupt URB if device is openPaul Stewart2011-04-28
| | | | | | | | | | | | | | | | | | | | | | Resubmit interrupt URB if device is open. Use a flag set in usbnet_open() to determine this state. Also kill and free interrupt URB in usbnet_disconnect(). [Rebased off git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git] Signed-off-by: Paul Stewart <pstew@chromium.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* | Merge branch 'for-davem' of ↵David S. Miller2011-05-05
|\ \ | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6
| * \ Merge branch 'master' of ↵John W. Linville2011-05-05
| |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6 into for-davem Conflicts: drivers/net/wireless/libertas/if_cs.c drivers/net/wireless/rtlwifi/pci.c net/bluetooth/l2cap_sock.c
| | * \ Merge branch 'master' of ↵John W. Linville2011-04-29
| | |\ \ | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/padovan/bluetooth-next-2.6
| | | * | Bluetooth: Add store_hint parameter to mgmt_new_keyJohan Hedberg2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Even for keys that shouldn't be stored some use cases require the knowledge of a new key having been created so that the conclusion of a successful pairing can be made. Therefore, always send the mgmt_new_key event but add a store_hint parameter to it to indicate to user space whether the key should be stored or not. Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Remove old_key_type from mgmt_ev_new_keyJohan Hedberg2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | User space shouldn't have any need for the old key type so remove it from the corresponding Management interface event. Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Fix link key persistent storage criteriaJohan Hedberg2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Link keys should only be stored if very specific criteria of the authentication process are fulfilled. This patch essentially copies the criteria that user space has so far been using to the kernel side so that the management interface works properly. Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Add confirm_hint parameter to user confirmation requestsJohan Hedberg2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When accepting a pairing request which fulfills the SSP auto-accept criteria we need to push the request all the way to the user for confirmation. This patch adds a new hint to the user_confirm_request management event so user space can know when to show a numeric comparison dialog and when to show a simple yes/no confirmation dialog. Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Add variable SSP auto-accept delay supportJohan Hedberg2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some test systems require an arbitrary delay to the auto-accept test cases for Secure Simple Pairing in order for the tests to pass. Previously when this was handled in user space it was worked around by code modifications and recompilation, but now that it's on the kernel side it's more convenient if there's a debugfs interface for it. Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Add secure flag for mgmt_pin_code_reqWaldemar Rymarkiewicz2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Extend the mgmt_pin_code_request interface to require secure pin code (16 digit) for authentication. This is a kernel part of the secure pin code requirement notification to user space agent. Code styling fix by Johan Hedberg. Signed-off-by: Waldemar Rymarkiewicz <waldemar.rymarkiewicz@tieto.com> Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Map sec_level to link key requirementsWaldemar Rymarkiewicz2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Keep the link key type together with connection and use it to map security level to link key requirements. Authenticate and/or encrypt connection if the link is insufficiently secure. Signed-off-by: Waldemar Rymarkiewicz <waldemar.rymarkiewicz@tieto.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Add definitions for link key typesWaldemar Rymarkiewicz2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Introduce the link key types defs and use them instead of magic numbers. Signed-off-by: Waldemar Rymarkiewicz <waldemar.rymarkiewicz@tieto.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Add discovering event to the Management interfaceJohan Hedberg2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds a new event to the Management interface to track when local adapters are discovering remote devices. For now this only tracks BR/EDR discovery procedures. Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com> Signed-off-by: Anderson Briglia <anderson.briglia@openbossa.org> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Add basic discovery commands to the management interfaceJohan Hedberg2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds start_discovery and stop_discovery commands to the management interface. Right now their implementation is fairly simplistic and the parameters are fixed to what user space has defaulted to so far. This is the very initial phase for discovery implementation into the kernel. Next steps include name resolution, LE scanning and bdaddr type handling. Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com> Signed-off-by: Anderson Briglia <anderson.briglia@openbossa.org> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Don't export l2cap_sock_opsGustavo F. Padovan2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | l2cap_sk_ops can be static, it's not used outside l2cap_sock.c Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Fix memory leak with L2CAP channelsGustavo F. Padovan2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A new l2cap_chan_free() is added to free the channels. Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Move conn to struct l2cap_chanGustavo F. Padovan2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There is no need to the socket deal directly with the channel, most of the time it cares about the channel only. Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Move more vars to struct l2cap_chanGustavo F. Padovan2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In this commit, psm, scid and dcid. Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Move more channel info to struct l2cap_chanGustavo F. Padovan2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In this commit, omtu, imtu, flush_to, mode and sport. It also remove the pi var from l2cap_sock_sendmsg(). Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Move more vars to struct l2cap_chanGustavo F. Padovan2011-04-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In this commit all ERTM and Streaming Mode specific vars. Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Move some more elements to struct l2cap_chanGustavo F. Padovan2011-04-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In this commit sec_level, force_reliable, role_switch and flushable. Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Rename l2cap_do_connect() to l2cap_chan_connect()Gustavo F. Padovan2011-04-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | l2cap_chan_connect() is a much better name and reflects what this functions is doing (or will do once socket dependence is removed from the core). Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Move conf_state to struct l2cap_chanGustavo F. Padovan2011-04-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | First move of elements depending on user data. Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Bluetooth: Refactor L2CAP channel allocationGustavo F. Padovan2011-04-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If the allocation happens at l2cap_sock_create() will be able to use the struct l2cap_chan to store channel info that comes from the user via setsockopt. Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
| | | * | Merge master.kernel.org:/pub/scm/linux/kernel/git/padovan/bluetooth-2.6Gustavo F. Padovan2011-04-20
| | | |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Conflicts: net/bluetooth/l2cap_core.c net/bluetooth/l2cap_sock.c net/bluetooth/mgmt.c
| | * | | | mac80211: Add new API for rate selectionSujith Manoharan2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds a new API for setting a TX rate mask in drivers that have rate control in either the firmware or hardware. This can be used for various purposes, for example, masking out the 11b rates in P2P operation. Signed-off-by: Sujith Manoharan <Sujith.Manoharan@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | | | mac80211: allow low level drivers to report packet lossArik Nemtsov2011-04-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add API that allows low level drivers to notify mac80211 about TX packet loss. This is useful when there are FW triggers to notify the low level driver about these events. Signed-off-by: Arik Nemtsov <arik@wizery.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | | | ssb: update reject bit for Target State LowRafał Miłecki2011-04-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | My 14e4:4315 is SSB_IDLOW_SSBREV_26: read32 0xfaafcff8 -> 0x600422d5 My 14e4:4328 is SSB_IDLOW_SSBREV_24: read32 0xfaafcff8 -> 0x400422c5 My 14e4:432b is SSB_IDLOW_SSBREV_26 again: read32 0xfaafcff8 -> 0x600422d5 For all of them wl driver is using 0x2 reject bit: write32(0xf98) <- 0x00010002 So it seems SSB 2.3 is the exception using another bit. Signed-off-by: Rafał Miłecki <zajec5@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | | | ssb: cc: clear GPIOPULL registers on initRafał Miłecki2011-04-25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Signed-off-by: Rafał Miłecki <zajec5@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | | | ssb: cc: add & fix definesRafał Miłecki2011-04-25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We probably got false positive results for checking PLL being down. Signed-off-by: Rafał Miłecki <zajec5@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | | | ath9k_hw: Get AHB clock information from ath9k_platform_dataVasanthakumar Thiagarajan2011-04-25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add a bool in ath9k_platform_data to pass AHB clock speed information. Driver needs this to configure PLL on some SOCs. Signed-off-by: Vasanthakumar Thiagarajan <vasanth@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | | | Merge branch 'for-linville' of ↵John W. Linville2011-04-25
| | |\ \ \ \ | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/luca/wl12xx
| | | * | | | wl12xx: Handle platforms without level trigger interruptsIdo Yariv2011-04-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some platforms are incapable of triggering on level interrupts. Add a platform quirks member in the platform data structure, as well as an edge interrupt quirk which can be set on such platforms. When the interrupt is requested with IRQF_TRIGGER_RISING, IRQF_ONESHOT cannot be used, as we might miss interrupts that occur after the FW status is cleared and before the threaded interrupt handler exits. Moreover, when IRQF_ONESHOT is not set, iterating more than once in the threaded interrupt handler introduces a few race conditions between this handler and the hardirq handler. Currently this is worked around by limiting the loop to one iteration only. This workaround has an impact on performance. To remove to this restriction, the race conditions will need to be addressed. Signed-off-by: Ido Yariv <ido@wizery.com> Signed-off-by: Luciano Coelho <coelho@ti.com>
| | | * | | | wl12xx: Clean up and fix the 128x boot sequenceIdo Yariv2011-04-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Clean up the boot sequence code & fix the following issues: 1. Always read the registers' values and set the relevant bits instead of zeroing all other bits 2. Handle cases where wl1271_top_reg_read returns an error 3. Verify that the HW can detect the selected clock source 4. Remove 128x PG10 initialization code 5. Configure the MCS PLL to work in HP mode Signed-off-by: Ido Yariv <ido@wizery.com> Reviewed-by: Luciano Coelho <coelho@ti.com> Signed-off-by: Luciano Coelho <coelho@ti.com>
| | | * | | | wl12xx: add new board_tcxo_clock element to the platform dataLuciano Coelho2011-04-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This new value is a new type of clock setting that is used by wl128x chipsets. Signed-off-by: Luciano Coelho <coelho@ti.com>
* | | | | | | net: Add sendmmsg socket system callAnton Blanchard2011-05-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds a multiple message send syscall and is the send version of the existing recvmmsg syscall. This is heavily based on the patch by Arnaldo that added recvmmsg. I wrote a microbenchmark to test the performance gains of using this new syscall: http://ozlabs.org/~anton/junkcode/sendmmsg_test.c The test was run on a ppc64 box with a 10 Gbit network card. The benchmark can send both UDP and RAW ethernet packets. 64B UDP batch pkts/sec 1 804570 2 872800 (+ 8 %) 4 916556 (+14 %) 8 939712 (+17 %) 16 952688 (+18 %) 32 956448 (+19 %) 64 964800 (+20 %) 64B raw socket batch pkts/sec 1 1201449 2 1350028 (+12 %) 4 1461416 (+22 %) 8 1513080 (+26 %) 16 1541216 (+28 %) 32 1553440 (+29 %) 64 1557888 (+30 %) We see a 20% improvement in throughput on UDP send and 30% on raw socket send. [ Add sparc syscall entries. -DaveM ] Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: David S. Miller <davem@davemloft.net>
generated by cgit v1.2.2 (git 2.25.0) at 2025-01-09 06:22:50 -0500