- 24 Feb, 2014 7 commits
-
-
David S. Miller authored
Amir Vadai says: ==================== net/mlx4: Mellanox driver update 01-01-2014 This small patchset has a fix to a bogus usage of netif_get_num_default_rss_queues() in mlx4_en driver. Changes from V1: - Removed affinity_hint patch, to make it a generic instead of mlx specific Changes from V0: - Instead of reverting the netif_get_num_default_rss_queues() in mlx4_en, fixing it to limit the actual number of receive queues instead of limiting the number of IRQ's. Patchset was applied and tested against commit: cb6e926 "ipv6:fix checkpatch errors with assignment in if condition" ==================== Signed-off-by:David S. Miller <davem@davemloft.net>
-
Ido Shamay authored
This fix a performance bug introduced by commit 90b1ebe7 "mlx4: set maximal number of default RSS queues", which limits the numbers of IRQs opened by core module. The limit should be on the number of queues in the indirection table - rx_rings, and not on the number of IRQ's. Also, limiting on mlx4_core initialization instead of in mlx4_en, prevented using "ethtool -L" to utilize all the CPU's, when performance mode is prefered, since limiting this number to 8 reduces overall packet rate by 15%-50% in multiple TCP streams applications. For example, after running ethtool -L <ethx> rx 16 Packet rate Before the fix 897799 After the fix 1142070 Results were obtained using netperf: S=200 ; ( for i in $(seq 1 $S) ; do ( \ netperf -H 11.7.13.55 -t TCP_RR -l 30 &) ; \ wait ; done | grep "1 1" | awk '{SUM+=$6} END {print SUM}' ) CC: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by:
Ido Shamay <idos@mellanox.com> Signed-off-by:
Amir Vadai <amirv@mellanox.com> Signed-off-by:
-
Ido Shamay authored
mlx4_en_add() is too long. Moving set number of RX rings to a utiltity function to improve readability and modulization of the code. Signed-off-by:
-
dingtianhong authored
The bond_xxx_info_query() was already in RTNL, so no need to use bond lock to protect the bond slave list, so remove it. Cc: Jay Vosburgh <fubar@us.ibm.com> Cc: Veaceslav Falico <vfalico@redhat.com> Cc: Andy Gospodarek <andy@greyhouse.net> Signed-off-by:
Ding Tianhong <dingtianhong@huawei.com> Signed-off-by:
-
dingtianhong authored
The bond_info_show_master already in RCU read-side critical section, and the we access curr_active_slave without the curr_slave_lock, we could not sure whether the curr_active_slave will be changed during the processing, so use RCU to protected the pointer. Cc: Jay Vosburgh <fubar@us.ibm.com> Cc: Veaceslav Falico <vfalico@redhat.com> Cc: Andy Gospodarek <andy@greyhouse.net> Signed-off-by:
-
dingtianhong authored
The __netpoll_setup() will check the slave's flag and ndo_poll_controller just like the slave_dev_support_netpoll() does, and slave_dev_support_netpoll() was not used by any place, so remove it. Cc: Jay Vosburgh <fubar@us.ibm.com> Cc: Veaceslav Falico <vfalico@redhat.com> Cc: Andy Gospodarek <andy@greyhouse.net> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-nextDavid S. Miller authored
Steffen Klassert says: ==================== 1) Introduce skb_to_sgvec_nomark function to add further data to the sg list without calling sg_unmark_end first. Needed to add extended sequence number informations. From Fan Du. 2) Add IPsec extended sequence numbers support to the Authentication Header protocol for ipv4 and ipv6. From Fan Du. 3) Make the IPsec flowcache namespace aware, from Fan Du. 4) Avoid creating temporary SA for every packet when no key manager is registered. From Horia Geanta. 5) Support filtering of SA dumps to show only the SAs that match a given filter. From Nicolas Dichtel. 6) Remove caching of xfrm_policy_sk_bundles. The cached socket policy bundles are never used, instead we create a new cache entry whenever xfrm_lookup() is called on a socket policy. Most protocols cache the used routes to the socket, so this caching is not needed. 7) Fix a forgotten SADB_X_EXT_FILTER length check in pfkey, from Nicolas Dichtel. 8) Cleanup error handling of xfrm_state_clone. ==================== Signed-off-by:
-
- 21 Feb, 2014 17 commits
-
-
David S. Miller authored
Aaron Brown says: ==================== Intel Wired LAN Driver Updates This series contains updates to i40e and (mostly to) i40evf. Mitch provides most the work for this series. For the vf driver he requests a reset on a tx hang, removes vlan filtes on close since we already remove the MAC filters, fixes some crashes, gets rid of PCI DAC as it does not mean much on virtualized PCIe parts, skips assigning the device name that just gets renamed anyway, stores the descriptor ring size in a manner that allows the use of common tx and rx code with the PF driver and makes a handful of cosmetic fixes. For i40e he removes a delay left over from debugging and changes a do/while loop to a for loop to avoid hitting another delay each time. Catherine fixes inconsistent MSI and MSI-X messages and bumps the driver version. v2: Removed unnecessary periods and redundant OOM message. ==================== Signed-off-by: -
Catherine Sullivan authored
Update the driver versions. Change-ID: I3fe23024d17da0e614ce126edb365bb2c428d482 Signed-off-by:
Catherine Sullivan <catherine.sullivan@intel.com> Signed-off-by:
Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by:
Kavindya Deegala <kavindya.s.deegala@intel.com> Signed-off-by:
Aaron Brown <aaron.f.brown@intel.com> Signed-off-by:
-
Catherine Sullivan authored
Fix inconsistent use of MSIX and MSI-X in messages. Change-ID: Iae9ffb42819677c34544719044ed77632e06147d Signed-off-by:
-
Mitch Williams authored
Change the do/while to a for loop, so we don't hit the delay each time, even when the register is ready for action. Don't bother to set or clear the QENA_STAT bit as it is read-only. Change-ID: Ie464718804dd79f6d726f291caa9b0c872b49978 Signed-off-by:
Mitch Williams <mitch.a.williams@intel.com> Signed-off-by:
-
Mitch Williams authored
Ain't nothing gonna break my stride, nobody's gonna slow me down, oh no. I got to keep on moving. This was originally put in for debugging just-in-case purposes and never removed. Change-ID: Ic12c2e179c3923f54e6ba0a9e4ab05d25c3bab29 Signed-off-by:
-
Mitch A Williams authored
Remove a bogus space. Signed-off-by:
Sibai Li <sibai.li@intel.com> Signed-off-by:
-
Mitch Williams authored
A bunch of changes merit a new version number, and since these were made in the new year, update the copyright date. Change-ID: Ic3f282bf0c20679b9fb06860211afa7c78055bc2 Signed-off-by:
-
Mitch Williams authored
Keep the descriptor ring size in the actual ring structs instead of in the adapter struct. This enables us to use common tx and rx code with the i40e PF driver. Also update copyrights. Change-ID: I2861e599b2b4c76441c062ea14400f4750f54d0e Signed-off-by:
-
Mitch Williams authored
We don't need to set an interface name here; the net core will do that, and then it will get renamed by udev anyway. Change-ID: I839a17837d19bedd1f490bff32ac5b85b4bfd97f Signed-off-by:
-
Mitch Williams authored
This comment is simply not true. Change-ID: If006b02b60984601a24257a951ae873dff568008 Signed-off-by:
-
Mitch Williams authored
Make sure errors are reported at the correct log level, quit printing the function name every time, and make the messages more consistent in format. v2: Removed unnecessary periods and redundant OOM message. Change-ID: I50e443467519ad3850def131d84626c50612c611 Signed-off-by:
-
Mitch Williams authored
PCI DAC doesn't really mean much on a virtualized PCI Express part, so get rid of that check and just always set the HIGHDMA flag in the net device. Change-ID: I2040272be0e7934323f470c2bc73fbdd4f93e2b6 Signed-off-by:
-
Mitch Williams authored
Depending upon the state of the driver, there are several potential pitfalls on remove. Kill the watchdog task so rmmod doesn't hang. Check the adapter->msix_entries field, not the num_msix_vectors field, which is never cleared. Change-ID: I0546048477f09fc19e481bd37efa30daae4faa88 Signed-off-by:
-
Mitch Williams authored
We remove all the MAC filters, so remove the VLAN filters, too. Change-ID: I4f7559acdf005dc3f359bf6460ce32d183c8878b Signed-off-by:
-
Mitch Williams authored
If the kernel watchdog bites us, ask the PF to reset us and attempt to reinit the driver. Change-ID: Ic97665aeeed71ce712b9c4f057e78ff8372522b9 Signed-off-by:
-
Steffen Klassert authored
The error pointer passed to xfrm_state_clone() is unchecked, so remove it and indicate an error by returning a null pointer. Signed-off-by:Steffen Klassert <steffen.klassert@secunet.com>
-
Nicolas Dichtel authored
This patch fixes commit d3623099 ("ipsec: add support of limited SA dump"). sadb_ext_min_len array should be updated with the new type (SADB_X_EXT_FILTER). Reported-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by:
-
- 20 Feb, 2014 4 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-nextDavid S. Miller authored
John W. Linville says: ==================== Please pull this batch of wireless updates intended for the 3.15 stream! For the mac80211 bits, Johannes says: "We have some cleanups and minor fixes as well as userspace API improvements from a lot of people, extended VHT support for radiotap from Emmanuel, CSA improvements from Andrei, Luca and Michal. I've also included my work on hwsim to make dynamic registration of radios possible." Along with that, we get the usual round of updates to ath9k, brcmfmac, mwifiex, wcn36xx, and the ti drivers -- nothing particularly noteworthy, mostly just random updates and refactoring. Also included is a pull of the wireless tree, intended to resolve some potential merge issues. ==================== Signed-off-by: -
John W. Linville authored
Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next into for-davem
-
Veaceslav Falico authored
bond->curr_active_slave can be changed between its deferences, even to NULL, and thus we might panic. We're always holding the rcu (rx_handler->bond_handle_frame()->bond_arp_rcv()) so fix this by rcu_dereferencing() it and using the saved. Reported-by:
Veaceslav Falico <vfalico@redhat.com> Acked-by:
-
Haiyang Zhang authored
It auto negotiates the highest NetVSP version supported by both guest and host. Signed-off-by:
Haiyang Zhang <haiyangz@microsoft.com> Reviewed-by:
K. Y. Srinivasan <kys@microsoft.com> Signed-off-by:
-
- 19 Feb, 2014 12 commits
-
-
Florian Westphal authored
Currently the kernel tries to announce a zero window when free_space is below the current receiver mss estimate. When a sender is transmitting small packets and reader consumes data slowly (or not at all), receiver might be unable to shrink the receive win because a) we cannot withdraw already-commited receive window, and, b) we have to round the current rwin up to a multiple of the wscale factor, else we would shrink the current window. This causes the receive buffer to fill up until the rmem limit is hit. When this happens, we start dropping packets. Moreover, tcp_clamp_window may continue to grow sk_rcvbuf towards rmem[2] even if socket is not being read from. As we cannot avoid the "current_win is rounded up to multiple of mss" issue [we would violate a) above] at least try to prevent the receive buf growth towards tcp_rmem[2] limit by attempting to move to zero-window announcement when free_space becomes less than 1/16 of the current allowed receive buffer maximum. If tcp_rmem[2] is large, this will increase our chances to get a zero-window announcement out in time. Reproducer: On server: $ nc -l -p 12345 <suspend it: CTRL-Z> Client: #!/usr/bin/env python import socket import time sock = socket.socket() sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1) sock.connect(("192.168.4.1", 12345)); while True: sock.send('A' * 23) time.sleep(0.005) socket buffer on server-side will grow until tcp_rmem[2] is hit, at which point the client rexmits data until -EDTIMEOUT: tcp_data_queue invokes tcp_try_rmem_schedule which will call tcp_prune_queue which calls tcp_clamp_window(). And that function will grow sk->sk_rcvbuf up until it eventually hits tcp_rmem[2]. Thanks to Eric Dumazet for running regression tests. Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by:Eric Dumazet <edumazet@google.com> Tested-by:
Florian Westphal <fw@strlen.de> Signed-off-by:
-
Erik Hugne authored
When a message could not be sent out because the destination node or link could not be found, the full message size is returned from sendmsg() as if it had been sent successfully. An application will then get a false indication that it's making forward progress. This problem has existed since the initial commit in 2.6.16. We change this to return -ENETUNREACH if the message cannot be delivered due to the destination node/link being unavailable. We also get rid of the redundant tipc_reject_msg call since freeing the buffer and doing a tipc_port_iovec_reject accomplishes exactly the same thing. Signed-off-by:
Erik Hugne <erik.hugne@ericsson.com> Reviewed-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
-
Daeseok Youn authored
sparse says: drivers/atm/solos-pci.c:763:6: warning: symbol 'solos_bh' was not declared. Should it be static? Signed-off-by:
Daeseok Youn <daeseok.youn@gmail.com> Acked-by:
Chas Williams <chas@cmf.nrl.navy.mil> Signed-off-by:
-
Daeseok Youn authored
sparse says: drivers/atm/nicstar.c:642:27: warning: Using plain integer as NULL pointer drivers/atm/nicstar.c:644:27: warning: Using plain integer as NULL pointer drivers/atm/nicstar.c:982:51: warning: Using plain integer as NULL pointer drivers/atm/nicstar.c:996:51: warning: Using plain integer as NULL pointer Signed-off-by:
-
Daeseok Youn authored
sparse says: drivers/atm/ambassador.c:1928:24: warning: Using plain integer as NULL pointer Signed-off-by:
-
Hannes Frederic Sowa authored
In case we decide in udp6_sendmsg to send the packet down the ipv4 udp_sendmsg path because the destination is either of family AF_INET or the destination is an ipv4 mapped ipv6 address, we don't honor the maybe specified ipv4 mapped ipv6 address in IPV6_PKTINFO. We simply can check for this option in ip_cmsg_send because no calls to ipv6 module functions are needed to do so. Reported-by:
Gert Doering <gert@space.net> Cc: Tore Anderson <tore@fud.no> Signed-off-by:
Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by:
-
Joe Perches authored
Make the error case return early. Make the normal return at the bottom of the function. Reduces indent for readability. Signed-off-by:
Joe Perches <joe@perches.com> Reviewed-by:
-
Joe Perches authored
It's unnecessary and less readable after a clause ending in a goto. Signed-off-by:
-
Joe Perches authored
It's smaller and faster for some architectures. Signed-off-by:
-
Don Fry authored
The pci_map_single calls never checked for error. Add error checking as requested by someone last year. Tested on 79c972. Signed-off-by:
Don Fry <pcnet32@frontier.com> Signed-off-by:
-
Don Fry authored
pcnet32_realloc_rx_ring() only worked on the first log2 number of entries in the receive ring instead of the all the entries. Replaced "1 << size" with more descriptive variable. This is my original bug from 2006. Found while testing another problem. Tested on 79C972 and 79C976. Signed-off-by:
-
Steffen Klassert authored
We currently cache socket policy bundles at xfrm_policy_sk_bundles. These cached bundles are never used. Instead we create and cache a new one whenever xfrm_lookup() is called on a socket policy. Most protocols cache the used routes to the socket, so let's remove the unused caching of socket policy bundles in xfrm. Signed-off-by:
-
