- 16 Nov, 2010 7 commits
-
-
Eric Dumazet authored
macvlan is a stacked device, like tunnels. We should use the lockless mechanism we are using in tunnels and loopback. This patch completely removes locking in TX path. tx stat counters are added into existing percpu stat structure, renamed from rx_stats to pcpu_stats. Note : this reverts commit 2c114553 (macvlan: add multiqueue capability) Note : rx_errors converted to a 32bit counter, like tx_dropped, since they dont need 64bit range. Signed-off-by:
Eric Dumazet <eric.dumazet@gmail.com> Cc: Patrick McHardy <kaber@trash.net> Cc: Ben Greear <greearb@candelatech.com> Cc: Ben Hutchings <bhutchings@solarflare.com> Acked-by:
Patrick McHardy <kaber@trash.net> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Neil Horman authored
packet: Enhance AF_PACKET implementation to not require high order contiguous memory allocation (v4) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Version 4 of this patch. Change notes: 1) Removed extra memset. Didn't think kcalloc added a GFP_ZERO the way kzalloc did :) Summary: It was shown to me recently that systems under high load were driven very deep into swap when tcpdump was run. The reason this happened was because the AF_PACKET protocol has a SET_RINGBUFFER socket option that allows the user space application to specify how many entries an AF_PACKET socket will have and how large each entry will be. It seems the default setting for tcpdump is to set the ring buffer to 32 entries of 64 Kb each, which implies 32 order 5 allocation. Thats difficult under good circumstances, and horrid under memory pressure. I thought it would be good to make that a bit more usable. I was going to do a simple conversion of the ring buffer from contigous pages to iovecs, but unfortunately, the metadata which AF_PACKET places in these buffers can easily span a page boundary, and given that these buffers get mapped into user space, and the data layout doesn't easily allow for a change to padding between frames to avoid that, a simple iovec change is just going to break user space ABI consistency. So I've done this, I've added a three tiered mechanism to the af_packet set_ring socket option. It attempts to allocate memory in the following order: 1) Using __get_free_pages with GFP_NORETRY set, so as to fail quickly without digging into swap 2) Using vmalloc 3) Using __get_free_pages with GFP_NORETRY clear, causing us to try as hard as needed to get the memory The effect is that we don't disturb the system as much when we're under load, while still being able to conduct tcpdumps effectively. Tested successfully by me. Signed-off-by:
Neil Horman <nhorman@tuxdriver.com> Acked-by:
Maciej Żenczykowski <zenczykowski@gmail.com> Reported-by:
-
Joe Perches authored
Using %pV reduces the number of printk calls and eliminates any possible message interleaving from other printk calls. Signed-off-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Jan Engelhardt authored
The changed functions do not modify the NL messages and/or attributes at all. They should use const (similar to strchr), so that callers which have a const nlmsg/nlattr around can make use of them without casting. While at it, constify a data array. Signed-off-by:
Jan Engelhardt <jengelh@medozas.de> Signed-off-by:
-
John Fastabend authored
Fix ref count bug introduced by commit 2de79570 Author: Lorenzo Colitti <lorenzo@google.com> Date: Wed Oct 27 18:16:49 2010 +0000 ipv6: addrconf: don't remove address state on ifdown if the address is being kept Fix logic so that addrconf_ifdown() decrements the inet6_ifaddr refcnt correctly with in6_ifa_put(). Reported-by:
Stephen Hemminger <shemminger@vyatta.com> Signed-off-by:
John Fastabend <john.r.fastabend@intel.com> Acked-by:
-
-
David S. Miller authored
ERROR: "netif_get_vlan_features" [drivers/net/xen-netfront.ko] undefined! Reported-by:
Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by:
-
- 15 Nov, 2010 33 commits
-
-
Vasanthy Kolluri authored
Fix data type of argument passed to pci_alloc_consistent and pci_free_consistent routines. Signed-off-by:
Vasanthy Kolluri <vkolluri@cisco.com> Signed-off-by:
Roopa Prabhu <roprabhu@cisco.com> Signed-off-by:
David Wang <dwang2@cisco.com> Signed-off-by:
-
Alan Cox authored
Fallout from the TIOCGICOUNT work Signed-off-by:
Alan Cox <alan@linux.intel.com> Signed-off-by:
-
Eric Dumazet authored
br_port_get() renamed to br_port_get_rtnl() to make clear RTNL is held. Signed-off-by:
-
stephen hemminger authored
The macro br_port_exists() is not enough protection when only RCU is being used. There is a tiny race where other CPU has cleared port handler hook, but is bridge port flag might still be set. Signed-off-by:
-
stephen hemminger authored
Suggested by Eric's bridge RCU changes. Signed-off-by:
-
Eric Dumazet authored
Add br_should_route_hook_t typedef, this is the only way we can get a clean RCU implementation for function pointer. Move route_hook to location where it is used. Signed-off-by:
-
Eric Dumazet authored
Add modern __rcu annotatations to bridge multicast table. Use newer hlist macros to avoid direct access to hlist internals. Signed-off-by:
-
Joe Perches authored
Signed-off-by:
-
Joe Perches authored
Signed-off-by:
Sjur Braendeland <sjur.brandeland@stericsson.com> Signed-off-by:
-
Joe Perches authored
Signed-off-by:
-
Joe Perches authored
Signed-off-by:
-
Joe Perches authored
Signed-off-by:
-
Joe Perches authored
Signed-off-by:
-
Joe Perches authored
Signed-off-by:
-
Joe Perches authored
Signed-off-by:
-
-
Tom Herbert authored
This patch move RX queue allocation to alloc_netdev_mq and freeing of the queues to free_netdev (symmetric to TX queue allocation). Each kobject RX queue takes a reference to the queue's device so that the device can't be freed before all the kobjects have been released-- this obviates the need for reference counts specific to RX queues. Signed-off-by:
Tom Herbert <therbert@google.com> Signed-off-by:
-
Tom Herbert authored
TX queues are now allocated in alloc_netdev_mq and freed in free_netdev. Signed-off-by:
-
Timo Teräs authored
The GRE Key field is intended to be used for identifying an individual traffic flow within a tunnel. It is useful to be able to have XFRM policy selector matches to have different policies for different GRE tunnels. Signed-off-by:
Timo Teräs <timo.teras@iki.fi> Signed-off-by:
-
David S. Miller authored
net/8021q/vlanproc.c: In function 'vlandev_seq_show': net/8021q/vlanproc.c:283:20: warning: unused variable 'fmt' Signed-off-by: -
Christian Lamparter authored
This patch replaces the handcrafted sign extension cruft with a generic bitop function. Signed-off-by:
Christian Lamparter <chunkeey@googlemail.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com>
-
Andreas Herrmann authored
This patch moves code out from wireless drivers where two different functions are defined in three code locations for the same purpose and provides a common function to sign extend a 32-bit value. Signed-off-by:
Andreas Herrmann <andreas.herrmann3@amd.com> Signed-off-by:
-
Grazvydas Ignotas authored
Make use the newly added method to pass platform data for wl1251 too. This allows to eliminate some redundant code. Cc: Ohad Ben-Cohen <ohad@wizery.com> Signed-off-by:
Grazvydas Ignotas <notasas@gmail.com> Acked-by:
Kalle Valo <kvalo@adurom.com> Acked-by:
Luciano Coelho <luciano.coelho@nokia.com> Acked-by:
Tony Lindgren <tony@atomide.com> Signed-off-by:
-
Grazvydas Ignotas authored
Add runtime PM support, similar to how it's done for wl1271. This allows to power down the card when the driver is loaded but network is not in use. Cc: Ohad Ben-Cohen <ohad@wizery.com> Signed-off-by:
-
Grazvydas Ignotas authored
Call interface specific power callback before calling board specific one. Also allow that callback to fail. This is how it's done for wl1271 and will be used for runtime_pm support. Signed-off-by:
-
Wey-Yi Guy authored
For 6000 series devices and up, enable automatic update MAC's register for better power usage in PSP mode Signed-off-by:
Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by:
-
Shanyu Zhao authored
Disconnected antenna algorithm is used to find out which antennas are disconnected. It should be disabled for devices that support advanced bluetooth coexist. Signed-off-by:
Shanyu Zhao <shanyu.zhao@intel.com> Signed-off-by:
-
Shanyu Zhao authored
Disconnected antenna algorithm is seperated into its own function from chain noise calibration routine for better code management. Signed-off-by:
-
Johannes Berg authored
When the HT information is changed due to BSS changes (like legacy stations joining) we need to recalculate HT RXON parameters. Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
-
Johannes Berg authored
During the RXON rewrite, this code got lost. When we've just associated, we need to enable all calibrations and see if some were already finished. Add back the missing code. Signed-off-by:
-
Johannes Berg authored
The RXON rework resulted in a massive loss of throughput because we weren't programming the device completely correctly -- the BSSID has to be programmed into the device before the AP station is uploaded. To fix this, simply always send the unassoc RXON, i.e. even when it was already unassoc so that the BSSID and some other parameters are updated properly. Signed-off-by:
-
Johannes Berg authored
Garen noticed that this was wrong. Fix the calibration -- default to multiple chains and fall back to single where possible. Signed-off-by:
-
Wey-Yi Guy authored
"mac80211: Fix WMM driver queue configuration" inadvertedly broke iwlwifi, because now mac80211 configures the QoS settings before assoc, and therefore before HT. Thus, iwlwifi no longer told the device about the HT setting, which it needs to -- and thus throughput went down a lot. Fix this by resending the QoS command to the device not only when QoS/WMM settings change, but also when HT changes. Signed-off-by:
-
