[bpf-next PATCH 3/4] bpf: sockmap, refactor sockmap routines to work with hashmap

This patch only refactors the existing sockmap code. This will allow much of the psock initialization code path and bpf helper codes to work for both sockmap bpf map types that are backed by an array, the currently supported type, and the new hash backed bpf map type sockhash. Most the fallout

[bpf-next PATCH 1/4] bpf: sockmap, free memory on sock close with cork data

If a socket with pending cork data is closed we do not return the memory to the socket until the garbage collector free's the psock structure. The garbage collector though can run after the sock has completed its close operation. If this ordering happens the sock code will through a WARN_ON

[bpf-next PATCH 2/4] bpf: sockmap, duplicates release calls may NULL sk_prot

It is possible to have multiple ULP tcp_release call paths in flight if a sock is closed and simultaneously being removed from the sockmap control path. The result would be setting the sk_prot to the saved values on the first iteration and then on the second iteration setting the value to NULL.

[bpf-next PATCH 4/4] bpf: sockmap, add hash map support

Sockmap is currently backed by an array and enforces keys to be four bytes. This works well for many use cases and was originally modeled after devmap which also uses four bytes keys. However, this has become limiting in larger use cases where a hash would be more appropriate. For example users

Re: [PATCH v4 iproute2-next 0/7] cm_id, cq, mr, and pd resource tracking

On 3/29/18 3:38 PM, Steve Wise wrote: > This series enhances the iproute2 rdma tool to include dumping of > connection manager id (cm_id), completion queue (cq), memory region (mr), > and protection domain (pd) rdma resources. It is the user-space part of > the kernel resource tracking series

Re: [RFC PATCH 2/3] netdev: kernel-only IFF_HIDDEN netdevice

On 4/1/18 3:13 AM, Si-Wei Liu wrote: > Hidden netdevice is not visible to userspace such that > typical network utilites e.g. ip, ifconfig and et al, > cannot sense its existence or configure it. Internally > hidden netdev may associate with an upper level netdev > that userspace has access to.

Re: [PATCH net] route: check sysctl_fib_multipath_use_neigh earlier than hash

On 4/1/18 8:40 AM, Xin Long wrote: > Prior to this patch, when one packet is hashed into path [1] > (hash <= nh_upper_bound) and it's neigh is dead, it will try > path [2]. However, if path [2]'s neigh is alive but it's > hash > nh_upper_bound, it will not return this alive path. > This packet

Re: [PATCH v2 net-next 10/10] mlxsw: spectrum: Don't use resource ID of 0

On 4/1/18 8:34 AM, Ido Schimmel wrote: > From: Petr Machata > > In commit 145307460ba9 ("devlink: Remove top_hierarchy arg to > devlink_resource_register"), the "top_hierarchy" parameter to > devlink_resource_register() was removed in favor of using the parameter >

include/uapi/linux/ethtool.h problem

Hi, I have the following problem: I included include/uapi/linux/mii.h from my source and when I compiled I got this error: ethtool.h:1556:35: error: comparison between signed and unsigned integer expressions [-Werro   return speed <= INT_MAX || speed == SPEED_UNKNOWN; speed declared as

Re: [PATCH net-next] net/broadcom: Fixup broken build due to function name change

Le 03/31/18 à 23:48, Tal Gilboa a écrit : > Fixes: 8c6d6895bebb ("net/dim: Rename *_get_profile() functions to > *_get_rx_moderation()") > Signed-off-by: Tal Gilboa I think David just backed out your entire patch series adding TX DIM so you would have to incorporate that

Re: [PATCH net-next] net: ipv6/gre: Add GRO support

On 04/01/2018 06:17 AM, Tariq Toukan wrote: > From: Eran Ben Elisha > > Add GRO capability for IPv6 GRE tunnel and ip6erspan tap, via gro_cells > infrastructure. > > Performance testing: 55% higher badwidth. > Measuring bandwidth of 1 thread IPv4 TCP traffic over IPv6 GRE

Re: [PATCH net-next] net/mlx4_en: CHECKSUM_COMPLETE support for fragments

On 04/01/2018 01:06 AM, Tariq Toukan wrote: > > Acked-by: Tariq Toukan Thanks Tariq and Saeed for the review !

[PATCH net 2/2] net: bgmac: Fix endian access in bgmac_dma_tx_ring_free()

bgmac_dma_tx_ring_free() assigns the ctl1 word which is a litle endian 32-bit word without using proper accessors, fix this, and because a length cannot be negative, use unsigned int while at it. Fixes: 9cde94506eac ("bgmac: implement scatter/gather support") Signed-off-by: Florian Fainelli

[PATCH net 1/2] net: bgmac: Correctly annotate register space

All the members: base, idm_base and nicpm_base should be annotated with __iomem since they are pointers to register space. This fixes a bunch of sparse reported warnings. Fixes: f6a95a24957a ("net: ethernet: bgmac: Add platform device support") Fixes: dd5c5d037f5e ("net: ethernet: bgmac: add NS2

[PATCH net 0/2] net: bgmac: Couple of sparse warnings

Hi all, This patch series fixes a couple of warnings reported by sparse, should not cause any functional problems since bgmac is typically used on LE platforms anyway. Florian Fainelli (2): net: bgmac: Correctly annotate register space net: bgmac: Fix endian access in

Re: [PATCH net-next] net/broadcom: Fixup broken build due to function name change

Hi Tal, I love your patch! Yet something to improve: [auto build test ERROR on net-next/master] url: https://github.com/0day-ci/linux/commits/Tal-Gilboa/net-broadcom-Fixup-broken-build-due-to-function-name-change/20180401-145037 config: x86_64-randconfig-x005-201813 (attached as .config

Re: [PATCH net-next 09/11] devlink: convert occ_get op to separate registration

On Fri, Mar 30, 2018 at 08:45:50AM -0600, David Ahern wrote: > On 3/29/18 2:33 PM, Ido Schimmel wrote: > > From: Jiri Pirko > > > > This resolves race during initialization where the resources with > > ops are registered before driver and the structures used by occ_get > > op

Re: [PATCH net-next 00/11] mlxsw: Various cleanups

On Sat, Mar 31, 2018 at 10:14:51PM -0400, David Miller wrote: > From: David Miller > Date: Sat, 31 Mar 2018 21:54:50 -0400 (EDT) > > > From: Ido Schimmel > > Date: Thu, 29 Mar 2018 23:33:23 +0300 > > > >> The first 10 patches from Jiri perform small

Re: [PATCH net-next v1] ipvs: fix multiplicative hashing in sh/dh/lblc/lblcr algorithms

Hello, On Sun, 1 Apr 2018, Vincent Bernat wrote: > The sh/dh/lblc/lblcr algorithms are using Knuth's multiplicative > hashing incorrectly. This results in uneven distribution. Good catch. > To fix this, the result has to be shifted by a constant. In "Lecture > 21: Hash

Re: [net-next V8 PATCH 01/16] mlx5: basic XDP_REDIRECT forward support

On 31/03/2018 3:05 PM, Jesper Dangaard Brouer wrote: .. @@ -844,6 +851,15 @@ static inline int mlx5e_xdp_handle(struct mlx5e_rq *rq, if (unlikely(!mlx5e_xmit_xdp_frame(rq, di, ))) trace_xdp_exception(rq->netdev, prog, act); return true;

Re: WARNING: refcount bug in should_fail

syzbot wrote: > > On Sun, Mar 4, 2018 at 6:57 AM, Tetsuo Handa > > wrote: > >> Switching from mm to fsdevel, for this report says that put_net(net) in > >> rpc_kill_sb() made net->count < 0 when mount_ns() failed due to > >> register_shrinker() failure. > > >>

[PATCH net-next] net: ipv6/gre: Add GRO support

From: Eran Ben Elisha Add GRO capability for IPv6 GRE tunnel and ip6erspan tap, via gro_cells infrastructure. Performance testing: 55% higher badwidth. Measuring bandwidth of 1 thread IPv4 TCP traffic over IPv6 GRE tunnel while GRO on the physical interface is disabled.

[PATCH v2 net-next 05/10] mlxsw: remove kvd_hash_granularity from config profile struct

From: Jiri Pirko This should not be part of the struct, as the struct fields are tightly coupled with the FW command payload of the same name. Just use the "granularity" define directly, as in other places. Signed-off-by: Jiri Pirko Signed-off-by: Ido

[PATCH v2 net-next 04/10] mlxsw: spectrum: Change KVD linear parts from list to array

From: Jiri Pirko The parts info is array. The parts copy this info array, yet they are a list. So make the indexing according to the id and change the list of parts into array of parts. This helps to eliminate lookups and constructs like mlxsw_sp_kvdl_part_update() (took me

[PATCH v2 net-next 10/10] mlxsw: spectrum: Don't use resource ID of 0

From: Petr Machata In commit 145307460ba9 ("devlink: Remove top_hierarchy arg to devlink_resource_register"), the "top_hierarchy" parameter to devlink_resource_register() was removed in favor of using the parameter "parent_resource_id" exclusively to determine who the parent

[PATCH v2 net-next 08/10] mlxsw: Move "resources_query_enable" out of mlxsw_config_profile

From: Jiri Pirko As struct mlxsw_config_profile is mapped to the payload of the FW command of the same name, resources_query_enable flag does not belong there. Move it to struct mlxsw_driver. Signed-off-by: Jiri Pirko Signed-off-by: Ido Schimmel

[PATCH v2 net-next 06/10] mlxsw: core: Fix arg name of MLXSW_CORE_RES_VALID and MLXSW_CORE_RES_GET

From: Jiri Pirko First arg of these helpers should be "mlxsw_core". Signed-off-by: Jiri Pirko Signed-off-by: Ido Schimmel --- drivers/net/ethernet/mellanox/mlxsw/core.h | 8 1 file changed, 4 insertions(+), 4 deletions(-)

[PATCH v2 net-next 02/10] mlxsw: spectrum_kvdl: Fix handling of resource_size_param

From: Jiri Pirko Current code uses global variables, adjusts them and passes pointer down to devlink. With every other mlxsw_core instance, the previously passed pointer values are rewritten. Fix this by de-globalize the variables. Fixes: 7f47b19bd744 ("mlxsw: spectrum_kvdl:

[PATCH v2 net-next 09/10] mlxsw: spectrum: Pass mlxsw_core as arg of mlxsw_sp_kvdl_resources_register()

From: Jiri Pirko Pass struct mlxsw_core instead of devlink since it is nicer within mlxsw code and we need both structs in mlxsw_sp_kvdl_resources_register() anyway. Signed-off-by: Jiri Pirko Signed-off-by: Ido Schimmel ---

[PATCH v2 net-next 03/10] mlxsw: Constify devlink_resource_ops

From: Jiri Pirko devlink_resource_ops should be const as the arg of register function is also const. Signed-off-by: Jiri Pirko Signed-off-by: Ido Schimmel --- drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 2 +-

[PATCH v2 net-next 07/10] mlxsw: Move "used_kvd_sizes" check to mlxsw_pci_config_profile

From: Jiri Pirko The check should be done directly in mlxsw_pci_config_profile, as for other profile items. Also, be consistent in naming with the rest and rename to "used_kvd_sizes". Signed-off-by: Jiri Pirko Signed-off-by: Ido Schimmel

[PATCH v2 net-next 01/10] mlxsw: spectrum_acl: Fix flex actions header ifndef define construct

From: Jiri Pirko Fix copy error in flex actions header ifndef define construct Signed-off-by: Jiri Pirko Signed-off-by: Ido Schimmel --- drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.h | 4 ++-- 1 file changed, 2

[PATCH v2 net-next 00/10] mlxsw: Various cleanups

Hi, The first nine patches from Jiri perform small and unrelated cleanups. The largest being the conversion of the KVD linear partitions from a list to an array, which simplifies the code. The last patch from Petr is a bug fix for a recent net-next commit that prevented the "kvd" resource from

Re: [PATCH v2 net-next 00/10] mlxsw: Various cleanups

From: Ido Schimmel Date: Sun, 1 Apr 2018 17:34:49 +0300 > The first nine patches from Jiri perform small and unrelated cleanups. > The largest being the conversion of the KVD linear partitions from a > list to an array, which simplifies the code. > > The last patch from

Re: [PATCH v15 net-next 00/12] Chelsio Inline TLS

Hi, On 4/1/2018 6:37 AM, David Miller wrote: From: Atul Gupta Date: Sat, 31 Mar 2018 21:41:51 +0530 Series for Chelsio Inline TLS driver (chtls) Series applied, thank you. Sorry for being late to the party, could you please help answer a few questions to help me

Re: [PATCH v15 net-next 01/12] tls: support for Inline tls record

On 3/31/2018 7:11 PM, Atul Gupta wrote: Facility to register Inline TLS drivers to net/tls. Setup TLS_HW_RECORD prot to listen on offload device. Cases handled - Inline TLS device exists, setup prot for TLS_HW_RECORD - Atleast one Inline TLS exists, sets TLS_HW_RECORD. - If non-inline device

[RFC v2] virtio: support packed ring

Hello everyone, This RFC implements packed ring support for virtio driver. The code was tested with DPDK vhost (testpmd/vhost-PMD) implemented by Jens at http://dpdk.org/ml/archives/dev/2018-January/089417.html Minor changes are needed for the vhost code, e.g. to kick the guest. TODO: -

[PATCH net] route: check sysctl_fib_multipath_use_neigh earlier than hash

Prior to this patch, when one packet is hashed into path [1] (hash <= nh_upper_bound) and it's neigh is dead, it will try path [2]. However, if path [2]'s neigh is alive but it's hash > nh_upper_bound, it will not return this alive path. This packet will never be sent even if path [2] is alive.

[bpf-next PATCH 0/4] Sockmap Updates

Some additions for sockmap. First a couple fixes from the last round to ensure (a) we account for corked memory when a socket is closing and (b) do not set a null sk_prot value when tcp_close and sockmap destroy race. Then the new addition is a hash map type. As the sockmap use cases become more

Re: [PATCH iproute2-next 1/1] tc: support oneline mode in action generic printer functions

On 3/30/18 10:16 PM, Roman Mashak wrote: > Signed-off-by: Roman Mashak > --- > tc/m_action.c | 12 > 1 file changed, 8 insertions(+), 4 deletions(-) > applied to iproute2-next

Re: [PATCH iproute2-next 1/1] tc: jsonify sample action

On 3/30/18 10:20 PM, Roman Mashak wrote: > Signed-off-by: Roman Mashak > --- > tc/m_sample.c | 22 +- > 1 file changed, 13 insertions(+), 9 deletions(-) > applied to iproute2-next

[RFC PATCH 3/3] virtio_net: make lower netdevs for virtio_bypass hidden

We should move virtio_bypass to a 1-upper-with-2-hidden-lower driver model for greater compatibility with regard to preserving userpsace API and ABI. On the other hand, technically virtio_bypass should make stricter check before automatically enslaving the corresponding virtual function or

[RFC PATCH 2/3] netdev: kernel-only IFF_HIDDEN netdevice

Hidden netdevice is not visible to userspace such that typical network utilites e.g. ip, ifconfig and et al, cannot sense its existence or configure it. Internally hidden netdev may associate with an upper level netdev that userspace has access to. Although userspace cannot manipulate the lower

Re: BUG: corrupted list in sctp_association_free

On Tue, Mar 13, 2018 at 12:12 PM, Neil Horman wrote: > On Tue, Mar 13, 2018 at 02:09:09PM +0300, Dmitry Vyukov wrote: >> On Tue, Mar 13, 2018 at 1:44 PM, Xin Long wrote: >> > On Tue, Mar 13, 2018 at 3:34 PM, syzbot >> >

Re: INFO: rcu detected stall in vprintk_default

On Sun, Apr 1, 2018 at 12:50 PM, syzbot wrote: > Hello, > > syzbot hit the following crash on upstream commit > 10b84daddbec72c6b440216a69de9a9605127f7a (Sat Mar 31 17:59:00 2018 +) > Merge branch 'perf-urgent-for-linus' of >

Re: INFO: rcu detected stall in vprintk_func

On Sun, Apr 1, 2018 at 12:49 PM, syzbot wrote: > Hello, > > syzbot hit the following crash on upstream commit > 3eb2ce825ea1ad89d20f7a3b5780df850e4be274 (Sun Mar 25 22:44:30 2018 +) > Linux 4.16-rc7 > syzbot dashboard link: >

[PATCH net-next v2] ipvs: fix multiplicative hashing in sh/dh/lblc/lblcr algorithms

The sh/dh/lblc/lblcr algorithms are using Knuth's multiplicative hashing incorrectly. Replace its use by the hash_32() macro, which correctly implements this algorithm. It doesn't use the same constant, but it shouldn't matter. Signed-off-by: Vincent Bernat ---

Re: WARNING: refcount bug in should_fail

On Sun, Apr 1, 2018 at 1:11 PM, Tetsuo Handa wrote: > Dmitry Vyukov wrote: >> On Sun, Apr 1, 2018 at 12:32 PM, Dmitry Vyukov wrote: >> > On Sun, Mar 4, 2018 at 6:57 AM, Tetsuo Handa >> > wrote: >> >>

Re: WARNING: refcount bug in should_fail

On Sun, Apr 1, 2018 at 1:46 PM, Dmitry Vyukov wrote: > On Sun, Apr 1, 2018 at 1:30 PM, Dmitry Vyukov wrote: >> On Sun, Apr 1, 2018 at 1:11 PM, Tetsuo Handa >> wrote: >>> Dmitry Vyukov wrote: On Sun, Apr 1, 2018 at

Re: [PATCH net-next 0/4] Introduce adaptive TX interrupt moderation to net DIM

On 4/1/2018 6:19 AM, David Miller wrote: From: David Miller Date: Sat, 31 Mar 2018 22:02:55 -0400 (EDT) From: Tal Gilboa Date: Fri, 30 Mar 2018 09:37:29 +0300 Net DIM is a library designed for dynamic interrupt moderation. It was implemented and

[PATCH net-next] net/broadcom: Fixup broken build due to function name change

Fixes: 8c6d6895bebb ("net/dim: Rename *_get_profile() functions to *_get_rx_moderation()") Signed-off-by: Tal Gilboa --- drivers/net/ethernet/broadcom/bcmsysport.c | 4 ++-- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 4 ++-- 2 files changed, 4 insertions(+), 4

Re: WARNING in refcount_sub_and_test (2)

On Fri, Mar 30, 2018 at 12:01 AM, syzbot wrote: > Hello, > > syzbot hit the following crash on bpf-next commit > 22527437e0a0c96ee3153e9d0382942b0fd4f9dd (Thu Mar 29 02:36:15 2018 +) > Merge branch 'nfp-bpf-updates' > syzbot dashboard

Re: [PATCH net-next] net/mlx4_en: CHECKSUM_COMPLETE support for fragments

On 30/03/2018 3:34 AM, Saeed Mahameed wrote: On Tue, Mar 27, 2018 at 2:21 PM, Eric Dumazet wrote: Refine the RX check summing handling to propagate the hardware provided checksum so that we do not have to compute it later in software. Signed-off-by: Eric Dumazet

[RFC PATCH 1/3] qemu: virtio-bypass should explicitly bind to a passthrough device

The new backup option allows guest virtio-bypass driver to explicitly bind to a corresponding passthrough instance, which is identifiable by the :. notation. MAC address is still validated in the guest but not the only criteria for pairing two devices. MAC address is more a matter of network

[RFC PATCH 0/3] Userspace compatible driver model for virtio_bypass

This RFC patch series attempts to hide the lower netdevs for virtio_bypass from userspace visibility, and tighten up the association between virtio_bypass and the lower passthrough netdev to be enslaved by binding to a specific device identifier explicitly. This in turn has the benefits of taking

Re: Re: WARNING: refcount bug in should_fail

On Sun, Mar 4, 2018 at 6:57 AM, Tetsuo Handa wrote: Switching from mm to fsdevel, for this report says that put_net(net) in rpc_kill_sb() made net->count < 0 when mount_ns() failed due to register_shrinker() failure. Relevant commits will be commit

Re: WARNING: refcount bug in should_fail

On Sun, Mar 4, 2018 at 6:57 AM, Tetsuo Handa wrote: > Switching from mm to fsdevel, for this report says that put_net(net) in > rpc_kill_sb() made net->count < 0 when mount_ns() failed due to > register_shrinker() failure. > > Relevant commits will be > commit

Re: WARNING: refcount bug in should_fail

Dmitry Vyukov wrote: > On Sun, Apr 1, 2018 at 12:32 PM, Dmitry Vyukov wrote: > > On Sun, Mar 4, 2018 at 6:57 AM, Tetsuo Handa > > wrote: > >> Switching from mm to fsdevel, for this report says that put_net(net) in > >> rpc_kill_sb() made

Re: [PATCH net-next v2] ipvs: fix multiplicative hashing in sh/dh/lblc/lblcr algorithms

Hello, On Sun, 1 Apr 2018, Vincent Bernat wrote: > The sh/dh/lblc/lblcr algorithms are using Knuth's multiplicative > hashing incorrectly. Replace its use by the hash_32() macro, which > correctly implements this algorithm. It doesn't use the same constant, > but it shouldn't matter. >

Re: [PATCH] net: implement IP_RECVHDRS option to get full headers through recvmsg cmsg.

On 03/31/2018 10:43 PM, Maciej Żenczykowski wrote: > From: Luigi Rizzo > > We have all sorts of different ways to fetch pre-UDP payload metadata: > IP_RECVTOS > IP_RECVTTL > IP_RECVOPTS > IP_RETOPTS > > But nothing generic which simply allows you to receive the

Re: WARNING in refcount_dec

On Thu, Mar 29, 2018 at 1:16 AM, Cong Wang wrote: > (Cc'ing netdev and Willem) > > On Wed, Mar 28, 2018 at 12:03 PM, Byoungyoung Lee > wrote: >> Another crash patterns observed: race between (setsockopt$packet_int) >> and (bind$packet). >> >>

INFO: rcu detected stall in sch_direct_xmit

Hello, syzbot hit the following crash on net-next commit 18845557fd6fc1998f2d0d8c30467f86db587529 (Thu Mar 29 20:24:06 2018 +) Merge tag 'wireless-drivers-next-for-davem-2018-03-29' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers-next syzbot dashboard link:

Re: [PATCH net-next] net/broadcom: Fixup broken build due to function name change

Hi Tal, I love your patch! Yet something to improve: [auto build test ERROR on net-next/master] url: https://github.com/0day-ci/linux/commits/Tal-Gilboa/net-broadcom-Fixup-broken-build-due-to-function-name-change/20180401-145037 config: xtensa-allmodconfig (attached as .config) compiler

Re: [PATCH net-next v1] ipvs: fix multiplicative hashing in sh/dh/lblc/lblcr algorithms

❦ 1 avril 2018 11:11 +0300, Julian Anastasov  : >> -return (ntohl(addr_fold)*2654435761UL) & IP_VS_DH_TAB_MASK; >> +return ((ntohl(addr_fold)*2654435761U) >> >> +(32 - IP_VS_DH_TAB_BITS)) & >> +IP_VS_DH_TAB_MASK; > > Looks like the '& mask'

Re: WARNING: refcount bug in should_fail

On Sun, Apr 1, 2018 at 12:32 PM, Dmitry Vyukov wrote: > On Sun, Mar 4, 2018 at 6:57 AM, Tetsuo Handa > wrote: >> Switching from mm to fsdevel, for this report says that put_net(net) in >> rpc_kill_sb() made net->count < 0 when mount_ns()

Re: [PATCH v4 iproute2-next 1/7] rdma: update rdma_netlink.h

On Thu, Mar 29, 2018 at 09:10:30AM -0700, Steve Wise wrote: > From: Steve Wise > > Pull in the latest rdma_netlink.h which has support for > the rdma nldev resource tracking objects being added > with this patch series. > > Signed-off-by: Steve Wise

Re: WARNING: refcount bug in should_fail

On Sun, Apr 1, 2018 at 1:30 PM, Dmitry Vyukov wrote: > On Sun, Apr 1, 2018 at 1:11 PM, Tetsuo Handa > wrote: >> Dmitry Vyukov wrote: >>> On Sun, Apr 1, 2018 at 12:32 PM, Dmitry Vyukov wrote: >>> > On Sun, Mar 4, 2018 at

Re: [PATCH net] ip6_gre: remove redundant 'tunnel' setting in ip6erspan_tap_init()

From: Alexey Kodanev Date: Fri, 30 Mar 2018 20:34:33 +0300 > 'tunnel' was already set at the start of ip6erspan_tap_init(). > > Fixes: 5a963eb61b7c ("ip6_gre: Add ERSPAN native tunnel support") > Signed-off-by: Alexey Kodanev Applied to

[PATCH] net: improve ipv4 performances

As the Linux networking stack is growing, more and more protocols are added, increasing the complexity of stack itself. Modern processors, contrary to common belief, are very bad in branch prediction, so it's our task to give hints to the compiler when possible. After a few profiling and

Re: [PATCH v4 iproute2-next 0/7] cm_id, cq, mr, and pd resource tracking

On Sun, Apr 01, 2018 at 09:22:14AM -0600, David Ahern wrote: > On 3/29/18 3:38 PM, Steve Wise wrote: > > This series enhances the iproute2 rdma tool to include dumping of > > connection manager id (cm_id), completion queue (cq), memory region (mr), > > and protection domain (pd) rdma resources.

[PATCH net-next V1 4/4] net/mlx5e: Enable adaptive-TX moderation

Add support for adaptive TX moderation. This greatly reduces TX interrupt rate and increases bandwidth, mostly for TCP bandwidth over ARM architecture (below). There is a slight single stream TCP with very large message sizes degradation (x86). In this case if there's any moderation on transmitted

[PATCH net-next V1 1/4] net/dim: Rename *_get_profile() functions to *_get_rx_moderation()

Preparation for introducing adaptive TX to net DIM. Signed-off-by: Tal Gilboa --- drivers/net/ethernet/broadcom/bcmsysport.c| 6 +++--- drivers/net/ethernet/broadcom/bnxt/bnxt_dim.c | 8 drivers/net/ethernet/broadcom/genet/bcmgenet.c| 6 +++---

[PATCH net-next V1 0/4] Introduce adaptive TX interrupt moderation to net DIM

Net DIM is a library designed for dynamic interrupt moderation. It was implemented and optimized with receive side interrupts in mind, since these are usually the CPU expensive ones. This patch-set introduces adaptive transmit interrupt moderation to net DIM, complete with a usage in the mlx5e

[PATCH net-next V1 2/4] net/dim: Add "enabled" field to net_dim struct

Preparation for introducing adaptive TX to net DIM. Signed-off-by: Tal Gilboa --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 1 - drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c | 10 +++--- drivers/net/ethernet/mellanox/mlx5/core/en_main.c| 6

[PATCH net-next V1 3/4] net/dim: Support adaptive TX moderation

Interrupt moderation for TX traffic requires different profiles than RX interrupt moderation. The main goal here is to reduce interrupt rate and allow better payload aggregation by keeping SKBs in the TX queue a bit longer. Ping-pong behavior would get a profile with a short timer, so latency

Re: [PATCH net] e1000e: Remove Other from EIAC.

On 2018-01-31 08:26, Benjamin Poirier wrote: > It was reported that emulated e1000e devices in vmware esxi 6.5 Build > 7526125 do not link up after commit 4aea7a5c5e94 ("e1000e: Avoid receiver > overrun interrupt bursts", v4.15-rc1). Some tracing shows that after > e1000e_trigger_lsc() is called,

Re: [PATCH net-next] net/mlx4_en: CHECKSUM_COMPLETE support for fragments

From: Eric Dumazet Date: Sun, 1 Apr 2018 09:39:38 -0700 > On 04/01/2018 01:06 AM, Tariq Toukan wrote: > >> >> Acked-by: Tariq Toukan > > Thanks Tariq and Saeed for the review ! Applied, thanks everyone.

Re: [PATCH V2] treewide/net: Rename eth_stp_addr to ether_stp_addr

From: Joe Perches Date: Sat, 31 Mar 2018 11:34:06 -0700 > eth_stp_addr is not in the same form as the other global ether__addr > addresses like ether_broadcast_addr. > > Convert it treewide. > > Miscellanea: > > o Add comment to the ether_stp_addr define to show it's for

Re: [PATCH v2 2/2] net: usb: asix88179_178a: de-duplicate code

From: Marc Zyngier Date: Sat, 31 Mar 2018 12:16:05 + > I'd suggest that this patch is reverted until we understand what is > actually going on here. Yep, makes sense, reverted.

Re: [PATCH] net: improve ipv4 performances

On Sun, 1 Apr 2018 20:31:21 +0200 Anton Gary Ceph wrote: > As the Linux networking stack is growing, more and more protocols are > added, increasing the complexity of stack itself. > Modern processors, contrary to common belief, are very bad in branch > prediction, so it's

Re: [PATCH net-next] net: tcp: update mib LISTENOVERFLOWS if syn flood

From: yuan linyu Date: Thu, 29 Mar 2018 19:20:09 +0800 > From: yuan linyu > > it's better to update mib LISTENOVERFLOWS even if packet is flood > > Signed-off-by: yuan linyu When we drop because of synflood we

Re: [PATCH 0/2] sh_eth: remove SH_ETH_OFFSET_INVALID abuses

From: Sergei Shtylyov Date: Sun, 1 Apr 2018 00:15:04 +0300 > Hello! > > Here's a set of 2 patches against DaveM's 'net-next.git' repo. They get rid of > the abuse of SH_ETH_OFFSET_INVALID for the register existence checks, so that > only its necessary uses

Re: [PATCH net-next 0/2] inet: factorize sk_wmem_alloc updates

From: Eric Dumazet Date: Sat, 31 Mar 2018 13:16:24 -0700 > While testing my inet defrag changes, I found that senders > could spend ~20% of cpu cycles in skb_set_owner_w() updating > sk->sk_wmem_alloc for every fragment they cook, competing > with TX completion of prior skbs

pull request: bluetooth-next 2018-04-01

Hi Dave, Here's (most likely) the last bluetooth-next pull request for the 4.17 kernel: - Remove unused btuart_cs driver (replaced by serial_cs + hci_uart) - New USB ID for Edimax EW-7611ULB controller - Cleanups & fixes to hci_bcm driver - Clenups to btmrvl driver Please let me know if

Re: [PATCH net-next] net/broadcom: Fixup broken build due to function name change

On 4/1/2018 7:33 PM, Florian Fainelli wrote: Le 03/31/18 à 23:48, Tal Gilboa a écrit : Fixes: 8c6d6895bebb ("net/dim: Rename *_get_profile() functions to *_get_rx_moderation()") Signed-off-by: Tal Gilboa I think David just backed out your entire patch series adding TX

Re: [PATCH v5 03/14] PCI: Add pcie_bandwidth_capable() to compute max supported link bandwidth

On 3/31/2018 12:05 AM, Bjorn Helgaas wrote: From: Tal Gilboa Add pcie_bandwidth_capable() to compute the max link bandwidth supported by a device, based on the max link speed and width, adjusted by the encoding overhead. The maximum bandwidth of the link is computed as:

Re: [PATCH v5 04/14] PCI: Add pcie_bandwidth_available() to compute bandwidth available to device

On 3/31/2018 12:05 AM, Bjorn Helgaas wrote: From: Tal Gilboa Add pcie_bandwidth_available() to compute the bandwidth available to a device. This may be limited by the device itself or by a slower upstream link leading to the device. The available bandwidth at each link

[PATCH] [PATCH] fix typo in command value in drivers/net/phy/mdio-bitbang.

mdio-bitbang mentioned 10 for both read and write. However mdio read opcode is 10 and write opcode is 01 Fixed comment. Signed-off-by: Frans Meulenbroeks --- drivers/net/phy/mdio-bitbang.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git

Re: [PATCH net] vlan: also check phy_driver ts_info for vlan's real device

From: Hangbin Liu Date: Fri, 30 Mar 2018 09:44:00 +0800 > Just like function ethtool_get_ts_info(), we should also consider the > phy_driver ts_info call back. For example, driver dp83640. > > Fixes: 37dd9255b2f6 ("vlan: Pass ethtool get_ts_info queries to real device.") >

Re: [PATCH] net: sched: do not emit messages while holding spinlock

From: Li RongQing Date: Fri, 30 Mar 2018 10:11:21 +0800 > move messages emitting out of sch_tree_lock to avoid holding > this lock too long. > > Signed-off-by: Li RongQing Applied.

Re: [PATCH net 1/1] net/mlx5e: Set EQE based as default TX interrupt moderation mode

From: Saeed Mahameed Date: Fri, 30 Mar 2018 15:50:08 -0700 > From: Tal Gilboa > > The default TX moderation mode was mistakenly set to CQE based. The > intention was to add a control ability in order to improve some specific > use-cases. In general, we

Re: [PATCH net] ibmvnic: Disable irqs before exiting reset from closed state

From: John Allen Date: Fri, 30 Mar 2018 13:44:21 -0500 > When the driver is closed, all the associated irqs are disabled. In the > event that the driver exits a reset in the closed state, we should be > consistent with the state we are in directly after a close. So

Re: [PATCH net] route: check sysctl_fib_multipath_use_neigh earlier than hash

From: Xin Long Date: Sun, 1 Apr 2018 22:40:35 +0800 > Prior to this patch, when one packet is hashed into path [1] > (hash <= nh_upper_bound) and it's neigh is dead, it will try > path [2]. However, if path [2]'s neigh is alive but it's > hash > nh_upper_bound, it will not

Re: [PATCH] [PATCH] fix typo in command value in drivers/net/phy/mdio-bitbang.

From: Frans Meulenbroeks Date: Sun, 1 Apr 2018 22:52:55 +0200 > mdio-bitbang mentioned 10 for both read and write. > However mdio read opcode is 10 and write opcode is 01 > Fixed comment. > > Signed-off-by: Frans Meulenbroeks Applied.

Re: [PATCH] sky2: Increase D3 delay to sky2 stops working after suspend

From: Kai-Heng Feng Date: Sat, 31 Mar 2018 23:42:03 +0800 > The sky2 ethernet stops working after system resume from suspend: > [ 582.852065] sky2 :04:00.0: Refused to change power state, currently in > D3 > > The current 150ms delay is not enough, change it to

RE: [PATCH v4 iproute2-next 0/7] cm_id, cq, mr, and pd resource tracking

> > On Sun, Apr 01, 2018 at 09:22:14AM -0600, David Ahern wrote: > > On 3/29/18 3:38 PM, Steve Wise wrote: > > > This series enhances the iproute2 rdma tool to include dumping of > > > connection manager id (cm_id), completion queue (cq), memory region > (mr), > > > and protection domain (pd)

Re: INFO: rcu detected stall in vprintk_func

On (04/01/18 12:51), Dmitry Vyukov wrote: [..] > > IMPORTANT: if you fix the bug, please add the following tag to the commit: > > Reported-by: syzbot+3f28bd18291266ec8...@syzkaller.appspotmail.com > > It will help syzbot understand when the bug is fixed. See footer for > > details. > > If you

Re: INFO: rcu detected stall in vprintk_default

On (04/01/18 12:52), Dmitry Vyukov wrote: > > Seems to be the same as: > > #syz dup: INFO: rcu detected stall in vprintk_func > > +nfc maintainers Yes, seems to be the same issue. -ss

Re: INFO: rcu detected stall in vprintk_func

On (04/02/18 10:54), Sergey Senozhatsky wrote: > > > If you forward the report, please keep this part and the footer. > > > > > > llcp: nfc_llcp_send_ui_frame: Could not allocate PDU > > > llcp: nfc_llcp_send_ui_frame: Could not allocate PDU > > > llcp: nfc_llcp_send_ui_frame: Could not allocate

Re: [PATCH net 0/2] net: bgmac: Couple of sparse warnings

From: Florian Fainelli Date: Sun, 1 Apr 2018 10:26:28 -0700 > This patch series fixes a couple of warnings reported by sparse, should not > cause any functional problems since bgmac is typically used on LE platforms > anyway. Series applied, thanks Florian.

  1   2   >