Re: [PATCH] net: take care of bonding in build_skb_flow_key (v2)

On 01/06/2016 05:03 PM, Wengang Wang wrote: In a bonding setting, we determines fragment size according to MTU and s/determines/determine Thanks a lot. Zhu Yanjun PMTU associated to the bonding master. If the slave finds the fragment size is too big, it drops the fragment and calls

question about vrf-lite

Hi David Ahern: when I test vrf-lite, I meet a question, could you help me? the envirnment is below: N2 N1 (all configs here) +---+ +--+ | | | |

[patch net-next 2/8] switchdev: add bridge vlan_filtering attribute

From: Elad Raz Adding vlan_filtering attribute to allow hardware vendor to support vlan-aware bridges. Vlan_filtering is a per-bridge attribute. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- include/net/switchdev.h | 2 ++ 1

[patch net-next 8/8] mlxsw: Remember untagged VLANs

From: Elad Raz When a vlan is been configured, remeber the untagged mode of the vlan. When displaying the list of configured VLANs, show the untagged attribute. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko ---

[patch net-next 6/8] mlxsw: Renaming local variable names for consistency

From: Elad Raz Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git

[patch net-next 7/8] mlxsw: Disable vlan_filtering for non .1D bridge

From: Elad Raz When a port is bridged, the bridge must be vlan aware bridge (.1Q) or the bridging should be on top of VLAN interfaces (.1D bridge). Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko ---

[patch net-next 5/8] mlxsw: Fixing vlans init range

From: Elad Raz Initialize VLANs 0..4095 (Remove init for VID 4096). Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c | 3 ++- 1 file changed, 2 insertions(+), 1

[PATCH] net: take care of bonding in build_skb_flow_key (v2)

In a bonding setting, we determines fragment size according to MTU and PMTU associated to the bonding master. If the slave finds the fragment size is too big, it drops the fragment and calls ip_rt_update_pmtu(), passing _skb_ and _pmtu_, trying to update the path MTU. Problem is that the target

RE: [PATCH v2 net-next] net: Implement fast csum_partial for x86_64

From: Eric Dumazet > Sent: 05 January 2016 22:19 > To: Tom Herbert > You might add a comment telling the '4' comes from length of 'adcq > 6*8(%rdi),%rax' instruction, and that the 'nop' is to compensate that > 'adcq0*8(%rdi),%rax' is using 3 bytes instead. > > We also could use .byte 0x48,

[PATCH] Bluetooth: avoid rebuilding hci_sock all the time

From: Johannes Berg Instead, allow using string formatting with send_monitor_note() and access init_utsname(). Signed-off-by: Johannes Berg --- net/bluetooth/hci_sock.c | 28 ++-- 1 file changed, 18 insertions(+), 10

[patch net-next v2 4/8] bridge: add vlan filtering change for new bridged device

From: Elad Raz Notifying hardware about newly bridged port vlan-aware changes. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- net/bridge/br_vlan.c | 11 +++ 1 file changed, 11 insertions(+) diff --git

[patch net-next v2 2/8] switchdev: add bridge vlan_filtering attribute

From: Elad Raz Adding vlan_filtering attribute to allow hardware vendor to support vlan-aware bridges. Vlan_filtering is a per-bridge attribute. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- include/net/switchdev.h | 2 ++ 1

[patch net-next v2 0/8] mlxsw: add offload support for vlan_filtering option

From: Jiri Pirko Elad says: This patch adds SWITCHDEV_ATTR_ID_BRIDGE_VLAN_FILTERING port attribute. When a bridge is offloaded to hardware, the hardware can learn if the bridge is .1Q bridge (VLAN-aware) or not VLAN aware bridge. In order to toggle the mode a user can use

[patch -next] mlxsw: core: remove an unnecessary condition

We checked "err" on the lines before so we know it's zero here. These cause a static checker warning because checking known things can indicate a bug. Maybe there is a missing assignment or we are checking the wrong variable. Signed-off-by: Dan Carpenter diff --git

[patch -next] fsl/fman: fix the pause_time test

pause_time is unsigned so it can't be less than zero. The bug means that we allow invalid pause-times. Fixes: 57ba4c9b56d8 ('fsl/fman: Add FMan MAC support') Signed-off-by: Dan Carpenter diff --git a/drivers/net/ethernet/freescale/fman/fman_dtsec.c

Re: [PATCH] net/rfkill: Create "airplane mode" LED trigger

On Tue, 2016-01-05 at 22:55 -0800, Marcel Holtmann wrote: >  > so I am not convinced the kernel should have the concept of airplane > mode at all. [snip long story] This is true, but that doesn't mean the patch is bad, just the naming could be different. I think the patch could name this

[patch net-next v2 8/8] mlxsw: Remember untagged VLANs

From: Elad Raz When a vlan is been configured, remeber the untagged mode of the vlan. When displaying the list of configured VLANs, show the untagged attribute. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko ---

[patch net-next v2 6/8] mlxsw: Renaming local variable names for consistency

From: Elad Raz Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git

[patch net-next v2 3/8] bridge: add vlan filtering change notification

From: Elad Raz Notifying hardware about bridge vlan-aware changes. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- net/bridge/br_vlan.c | 18 -- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git

[patch net-next v2 7/8] mlxsw: Disable vlan_filtering for non .1D bridge

From: Elad Raz When a port is bridged, the bridge must be vlan aware bridge (.1Q) or the bridging should be on top of VLAN interfaces (.1D bridge). Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko ---

[patch net-next v2 1/8] bridge: Propagate vlan add failure to user

From: Elad Raz Disallow adding interfaces to a bridge when vlan filtering operation failed. Send the failure code to the user. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- v1->v2: set err to return value of nbp_vlan_init ---

[patch net-next v2 5/8] mlxsw: Fixing vlans init range

From: Elad Raz Initialize VLANs 0..4095 (Remove init for VID 4096). Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c | 3 ++- 1 file changed, 2 insertions(+), 1

[PATCH 1/1] net: possible use after free in dst_release

dst_release should not access dst->flags after decrementing __refcnt to 0. The dst_entry may be in dst_busy_list and dst_gc_task may dst_destroy it before dst_release gets a chance to access dst->flags. Signed-off-by: Francesco Ruggeri --- net/core/dst.c | 3 ++- 1 file

Re: [PATCH] net: take care of bonding in build_skb_flow_key

Please see if the V2 is clear. thanks, wengang 在 2016年01月06日 16:18, zhuyj 写道: On 01/06/2016 04:14 PM, Wengang Wang wrote: 在 2016年01月06日 16:00, zhuyj 写道: On 01/06/2016 03:45 PM, Wengang Wang wrote: 在 2016年01月06日 15:35, zhuyj 写道: IMHO, "The path MTU is set to the active slave device, not

[patch -next] fsl/fman: use the ALIGN() macro

The original code works fine but the issue is that static checkers complain about this: ~(u16)(OFFSET_UNITS - 1) In that expression the cast to u16 is a no-op because we cast the value 15 to a u16 but then type promotion rules automatically cast it back to an int and then we do the

[patch net-next 3/8] bridge: add vlan filtering change notification

From: Elad Raz Notifying hardware about bridge vlan-aware changes. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- net/bridge/br_vlan.c | 18 -- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git

[patch net-next 1/8] bridge: Propagate vlan add failure to user

From: Elad Raz Disallow adding interfaces to a bridge when vlan filtering operation failed. Send the failure code to the user. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- net/bridge/br_if.c | 10 +- 1 file changed,

[patch net-next 0/8] mlxsw: add offload support for vlan_filtering option

From: Jiri Pirko Elad says: This patch adds SWITCHDEV_ATTR_ID_BRIDGE_VLAN_FILTERING port attribute. When a bridge is offloaded to hardware, the hardware can learn if the bridge is .1Q bridge (VLAN-aware) or not VLAN aware bridge. In order to toggle the mode a user can use

[patch net-next 4/8] bridge: add vlan filtering change for new bridged device

From: Elad Raz Notifying hardware about newly bridged port vlan-aware changes. Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko --- net/bridge/br_vlan.c | 11 +++ 1 file changed, 11 insertions(+) diff --git

[patch -next] fsl/fman: double free on probe failure

"priv" is allocated with devm_kzalloc() so freeing it here with kfree() will lead to a double free. Fixes: 3933961682a3 ('fsl/fman: Add FMan MAC driver') Signed-off-by: Dan Carpenter diff --git a/drivers/net/ethernet/freescale/fman/mac.c

Re: [patch -next] mlxsw: core: remove an unnecessary condition

Wed, Jan 06, 2016 at 10:56:30AM CET, dan.carpen...@oracle.com wrote: >We checked "err" on the lines before so we know it's zero here. > >These cause a static checker warning because checking known things can >indicate a bug. Maybe there is a missing assignment or we are checking >the wrong

Re: [PATCH 1/1] net: possible use after free in dst_release

On Wed, 2016-01-06 at 00:18 -0800, Francesco Ruggeri wrote: > dst_release should not access dst->flags after decrementing > __refcnt to 0. The dst_entry may be in dst_busy_list and > dst_gc_task may dst_destroy it before dst_release gets a chance > to access dst->flags. > > Signed-off-by:

[PATCH net-next 2/3] libcxgbi: add pdu parsing of LRO'ed skbs

The skbs could contain both paritial pdus and multiple completed pdus. Signed-off-by: Karen Xie Signed-off-by: Manoj Malviya Signed-off-by: Hariprasad Shenai --- drivers/scsi/cxgbi/libcxgbi.c | 415

[PATCH net-next 1/3] cxgb4: Add LRO infrastructure for cxgb4i driver

Signed-off-by: Hariprasad Shenai --- drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 15 +- drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 40 ++--- drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h | 5

[PATCH net-next 0/3] cxgb4/cxgb4i: add iscsi LRO support for chelsio t4/t5 adapters

Hi This patch series adds LRO iscsi support for chelsio T4/T5 adapters. This patch series has been created against net-next tree and includes patches on cxgb4 and cxgb4i driver. We have included all the maintainers of respective drivers. Kindly review the change and let us know in case of any

[PATCH net-next 3/3] cxgb4i: add iscsi LRO for chelsio t4/t5 adapters

Signed-off-by: Karen Xie Signed-off-by: Manoj Malviya Signed-off-by: Hariprasad Shenai --- drivers/scsi/cxgbi/cxgb4i/cxgb4i.c | 396 +++-- 1 file changed, 382 insertions(+), 14 deletions(-)

[PATCH net 0/2] vxlan: Set a large MTU on ovs-created vxlan devices

Prior to 4.3, openvswitch vxlan vports could transmit vxlan packets of any size, constrained only by the ability to transmit the resulting UDP packets. 4.3 introduced vxlan netdevs corresponding to vxlan vports. These netdevs have an MTU, which limits the size of a packet that can be

[PATCH net 1/2] vxlan: Relax the MTU constraint on vxlan devices

Allow the MTU of vxlan devices without an underlying device to be set to larger values (up to a maximum based on IP packet limits and vxlan overhead). Previously, their MTUs could not be set to higher than the conventional ethernet value of 1500. This is a very arbitrary value in the context of

Re: [PATCH] net: add per device sg_max_frags for skb

On Wed, 2016-01-06 at 14:16 +0100, Hans Westgaard Ry wrote: > Devices may have limits on the number of fragments in an skb they > support. Current codebase uses a constant as maximum for number of > fragments (MAX_SKB_FRAGS) one skb can hold and use. > > When enabling scatter/gather and running

Re: [RFC PATCH 00/12] drop the qdisc lock for pfifo_fast/mq

Sorry for not being as responsive as i would like to be (theman calls and i have to go). This looks like a good (tc workshop) candidate discussion, if still active by netdev11 time. On 15-12-30 12:50 PM, John Fastabend wrote: Hi, This is a first take at removing the qdisc lock on the xmit

RE: [PATCH] net: add per device sg_max_frags for skb

From: Hans Westgaard Ry > Sent: 06 January 2016 13:16 > Devices may have limits on the number of fragments in an skb they > support. Current codebase uses a constant as maximum for number of > fragments (MAX_SKB_FRAGS) one skb can hold and use. > > When enabling scatter/gather and running traffic

Re: [PATCH v2 net-next] net: Implement fast csum_partial for x86_64

On Wed, 2016-01-06 at 10:16 +, David Laight wrote: > From: Eric Dumazet > > Sent: 05 January 2016 22:19 > > To: Tom Herbert > > You might add a comment telling the '4' comes from length of 'adcq > > 6*8(%rdi),%rax' instruction, and that the 'nop' is to compensate that > > 'adcq

Re: [PATCH net-next 2/3] libcxgbi: add pdu parsing of LRO'ed skbs

Hi Hariprasad, [auto build test ERROR on net-next/master] url: https://github.com/0day-ci/linux/commits/Hariprasad-Shenai/cxgb4-Add-LRO-infrastructure-for-cxgb4i-driver/20160106-203040 config: i386-allmodconfig (attached as .config) reproduce: # save the attached .config to linux

Re: [PATCH net-next 3/3] cxgb4i: add iscsi LRO for chelsio t4/t5 adapters

Hi Hariprasad, [auto build test ERROR on net-next/master] url: https://github.com/0day-ci/linux/commits/Hariprasad-Shenai/cxgb4-Add-LRO-infrastructure-for-cxgb4i-driver/20160106-203040 config: i386-allmodconfig (attached as .config) reproduce: # save the attached .config to linux

[PATCH net 2/2] vxlan: Set a large MTU on ovs-created vxlan devices

Prior to 4.3, vxlan vports could transmit vxlan packets of any size, constrained only by the ability to transmit the resulting UDP packets. 4.3 introduced vxlan netdevs corresponding to vxlan vports. These netdevs have an MTU, which limits the size of a packet that can be successfully

[PATCH v2] Bluetooth: avoid rebuilding hci_sock all the time

From: Johannes Berg Instead, allow using string formatting with send_monitor_note() and access init_utsname(). Signed-off-by: Johannes Berg --- net/bluetooth/hci_sock.c | 28 ++-- 1 file changed, 18 insertions(+), 10

[PATCH] net: add per device sg_max_frags for skb

Devices may have limits on the number of fragments in an skb they support. Current codebase uses a constant as maximum for number of fragments (MAX_SKB_FRAGS) one skb can hold and use. When enabling scatter/gather and running traffic with many small messages the codebase uses the maximum number

[PATCH] net: qmi_wwan: Add WeTelecom-WPD600N

From: Kristian Evensen The WeTelecom-WPD600N is an LTE module that, in addition to supporting most "normal" bands, also supports LTE over 450MHz. Manual testing showed that only interface number three replies to QMI messages. Cc: Bjørn Mork

[PATCH v2 net-next] ethernet/atheros/alx: sanitize buffer sizing and padding

This is based on the work done by Przemek Rudy in bug 70761 at bugzilla.kernel.org, but with some work done to disentagle and clarify things a bit. Similar to Przemek's work and other drivers, we're adding a padding of 16 here, but we're also disentangling mtu size calculations from max buffer

Re: [PATCH v5 net-next 3/3] dt-bindings: net: macb: Add NP4 macb variant

On Tue, Jan 05, 2016 at 02:39:18PM +0100, Neil Armstrong wrote: > Add NP4 macb SoC variant. > > Signed-off-by: Neil Armstrong > --- > Documentation/devicetree/bindings/net/macb.txt | 1 + > 1 file changed, 1 insertion(+) Acked-by: Rob Herring -- To

Re: [PATCH] af_unix: Fix splice-bind deadlock

Hannes Frederic Sowa writes: > On Sun, Jan 3, 2016, at 19:03, Rainer Weikusat wrote: [reorder i_mutex and readlock locking] > I was concerned because of the comment in skb_socket_splice: > > /* Drop the socket lock, otherwise we have reverse > *

RE: [PATCH v2 net-next] net: Implement fast csum_partial for x86_64

From: Eric Dumazet > Sent: 06 January 2016 14:25 > On Wed, 2016-01-06 at 10:16 +, David Laight wrote: > > From: Eric Dumazet > > > Sent: 05 January 2016 22:19 > > > To: Tom Herbert > > > You might add a comment telling the '4' comes from length of 'adcq > > > 6*8(%rdi),%rax' instruction, and

RE: [Intel-wired-lan] [PATCH 2/2] ixgbe: restrict synchronization of link_up and speed

>-Original Message- >From: zhuyj [mailto:zyjzyj2...@gmail.com] >Sent: Tuesday, January 05, 2016 9:42 PM >To: Tantilov, Emil S; Kirsher, Jeffrey T; Brandeburg, Jesse; Nelson, >Shannon; Wyborny, Carolyn; Skidmore, Donald C; Allan, Bruce W; Ronciak, >John; Williams, Mitch A;

Re: [PATCH net-next 2/5] sctp: apply rhashtable api to send/recv path

On Wed, Jan 6, 2016 at 3:07 AM, Vlad Yasevich wrote: > On 12/30/2015 10:50 AM, Xin Long wrote: >> apply lookup apis to two functions, for __sctp_endpoint_lookup_assoc >> and __sctp_lookup_association, it's invoked in the protection of sock >> lock, it will be safe, but

RE: [RFC PATCH net-next 10/24] mdio: Move allocation of interrupts into core

> Have mdio_alloc() create the array of interrupt numbers, and > initialize it to POLLING. This is what most MDIO drivers want, so > allowing code to be removed from the drivers. > [S.H] Snip. > diff --git a/drivers/net/phy/mdio_bus.c b/drivers/net/phy/mdio_bus.c > index

[PATCH] mkiss: fix scribble on freed memory

commit d79f16c046086f4fe0d42184a458e187464eb83e fixed a user triggerable scribble on free memory but added a new one which allows the user to scribble even more and user controlled data into freed space. As with 6pack we need to halt the queue before we free the buffers, because the transmit

Re: [PATCH v2 net-next] net: Implement fast csum_partial for x86_64

On Wed, 2016-01-06 at 14:49 +, David Laight wrote: > Someone also pointed out that the code is memory limited (dual add > chains making no difference), so why is it unrolled at all? Because it matters if the data is already present in CPU caches. So why not unrolling if it helps in some

Re: [PATCH v2] Bluetooth: avoid rebuilding hci_sock all the time

Hi Johannes, > Instead, allow using string formatting with send_monitor_note() > and access init_utsname(). > > Signed-off-by: Johannes Berg > --- > net/bluetooth/hci_sock.c | 28 ++-- > 1 file changed, 18 insertions(+), 10 deletions(-) patch has

[PATCH net-next 0/2] net: break dependency of drivers on geneve and vxlan

Device drivers which support geneve or vxlan offloading have a dependency on the correlating kernel modules. Thus those drivers automatically loaded the geneve or vxlan modules. Break this dependency with this small series. Hannes Frederic Sowa (2): vxlan: break dependency to network drivers

[PATCH net-next 2/2] geneve: break dependency to network drivers

Signed-off-by: Hannes Frederic Sowa --- drivers/net/geneve.c | 30 +++--- include/linux/netdevice.h | 1 + include/net/geneve.h | 7 +++ 3 files changed, 31 insertions(+), 7 deletions(-) diff --git a/drivers/net/geneve.c

[PATCH net-next 1/2] vxlan: break dependency to network drivers

Signed-off-by: Hannes Frederic Sowa --- drivers/net/vxlan.c | 17 +++-- include/linux/netdevice.h | 1 + include/net/vxlan.h | 5 + 3 files changed, 13 insertions(+), 10 deletions(-) diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c

Re: [PATCH] net: qmi_wwan: Add WeTelecom-WPD600N

Kristian Evensen writes: > The WeTelecom-WPD600N is an LTE module that, in addition to supporting most > "normal" bands, also supports LTE over 450MHz. Manual testing showed that > only interface number three replies to QMI messages. > > Cc: Bjørn Mork

Re: question about vrf-lite

On 1/6/16 2:53 AM, roy.qing...@gmail.com wrote: Hi David Ahern: when I test vrf-lite, I meet a question, could you help me? the envirnment is below: N2 N1 (all configs here) +---+ +--+

[PATCH net-next] net: move ndo_features_check() close to ndo_start_xmit()

From: Eric Dumazet TX fast path uses ndo_start_xmit(), ndo_features_check() and ndo_select_queue(). Move ndo_features_check() close to ndo_start_xmit() to increase data locality. All "struct net_device_ops" should now be using C99 initializers. Signed-off-by: Eric Dumazet

Re: [PATCH] net: hns: avoid uninitialized variable warning:

On 1/5/2016 9:43 PM, David Miller wrote: From: Arnd Bergmann Date: Fri, 01 Jan 2016 23:27:57 +0100 gcc fails to see that the use of the 'last_offset' variable in hns_nic_reuse_page() is used correctly and issues a bogus warning: drivers/net/ethernet/hisilicon/hns/hns_enet.c:

Re: [REGRESSION] tcp/ipv4: kernel panic because of (possible) division by zero

On Wed, Jan 6, 2016 at 8:50 AM, Oleksandr Natalenko wrote: > > Unfortunately, the patch didn't help -- I've got the same stacktrace with > slightly different offset (+3) within the function. > > Now trying to get full stacktrace via netconsole. Need more time. > >

Re: [PATCH net-next 1/5] sctp: add the rhashtable apis for sctp global transport hashtable

On Thu, Jan 07, 2016 at 01:01:11AM +0800, Xin Long wrote: > On Wed, Jan 6, 2016 at 2:38 AM, Vlad Yasevich wrote: > > On 12/30/2015 10:50 AM, Xin Long wrote: > >> tranport hashtbale will replace the association hashtable to do the > >> lookup for transport, and then get

Re: [REGRESSION] tcp/ipv4: kernel panic because of (possible) division by zero

On Wed, Jan 6, 2016 at 10:19 AM, Yuchung Cheng wrote: > On Wed, Jan 6, 2016 at 8:50 AM, Oleksandr Natalenko > wrote: >> >> Unfortunately, the patch didn't help -- I've got the same stacktrace with >> slightly different offset (+3) within the

[PATCH net] Driver: Vmxnet3: Fix regression caused by 5738a09

Reported-by: Bingkuo Liu Signed-off-by: Shrikrishna Khare --- drivers/net/vmxnet3/vmxnet3_drv.c | 8 drivers/net/vmxnet3/vmxnet3_int.h | 4 ++-- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/net/vmxnet3/vmxnet3_drv.c

Re: [PATCH net-next 2/2] geneve: break dependency to network drivers

On 06.01.2016 19:00, Jesse Gross wrote: On Wed, Jan 6, 2016 at 7:41 AM, Hannes Frederic Sowa wrote: diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c index 24b077a32c1c9c..548925d1571cb1 100644 --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c +static int

Re: [REGRESSION] tcp/ipv4: kernel panic because of (possible) division by zero

Sure, but after catching the stacktrace. On середа, 6 січня 2016 р. 10:43:45 EET Yuchung Cheng wrote: > Could you turn off ecn (sysctl net.ipv4.tcp_ecn=0) to see if this still > happen? > >> On December 22, 2015 4:10:32 AM EET, Yuchung Cheng wrote: > >> >On Mon, Dec 21,

Re: [RFC v5 1/6] Timekeeping cross timestamp interface for device drivers

On Mon, Jan 4, 2016 at 4:45 AM, Christopher S. Hall wrote: > ACKNOWLEDGMENT: The original correlated clock source and cross > timestamp code was developed by Thomas Gleixner > . It has changed considerably and any mistakes are > mine. > > The

[PATCHv2 net-next 10/22] mdio: Move allocation of interrupts into core

Have mdio_alloc() create the array of interrupt numbers, and initialize it to POLLING. This is what most MDIO drivers want, so allowing code to be removed from the drivers. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- v2: Remove

[PATCHv2 net-next 08/22] phy: Centralise print about attached phy

Many Ethernet drivers contain the same netdev_info() print statement about the attached phy. Move it into the phy device code. Additionally add a varargs function which can be used to append additional information. Signed-off-by: Andrew Lunn --- v2: Add a function to do the

[PATCHv2 net-next 16/22] phy: Move PHY PM operations into phy_device

The MDIO PM operations are really PHY device PM operations. So move them into phy_device. This will be needed when we support devices on the mdio bus which are not PHYs. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli ---

[PATCHv2 net-next 17/22] phy: Centralize setting driver module owner

Rather than have each driver set the driver owner field, do it once in the core code. This will also help with later changes, when the device structure will move. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- drivers/net/phy/amd.c|

[PATCHv2 net-next 04/22] phy: Add phydev_err() and phydev_dbg() macros

In preparation for moving some of the phy_device structure members, add macros for printing errors and debug information. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- drivers/net/phy/at803x.c | 4 ++-- drivers/net/phy/bcm87xx.c | 5 +++--

[PATCHv2 net-next 02/22] mdio: Move mdiobus_read/write operatings into mdio.h

These are logically MDIO operations, not phy operations, so move them into the mdio header. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- include/linux/mdio.h | 6 ++ include/linux/phy.h | 6 +- 2 files changed, 7 insertions(+), 5

[PATCHv2 net-next 18/22] phy: Move phy specific bus match into phy_device

Matching a driver to a device has both generic parts, and parts which are specific to PHY devices. Move the PHY specific parts into phy_device. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- drivers/net/phy/mdio_bus.c | 38

[PATCHv2 net-next 12/22] of: phy: Only register a phy device for phys

We will soon support devices other than phys on the mdio bus. Look at a child's compatibility string to determine if it is a phy, before registering a phy device. Signed-off-by: Andrew Lunn --- v2: Also accept "ethernet-phy-ieee802.3-c22" as indicating a PHY. ---

[PATCHv2 net-next 19/22] mdio_bus: Generalise of_mdiobus_link_phydev()

This function should work with any sort of MDIO device which can be probed on the bus, not just PHY devices. So generalise it. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- drivers/net/phy/mdio_bus.c | 29 +++-- 1

[PATCHv2 net-next 06/22] net: dnet: Use phy_find_first() helper

Replace the open coded search for the first phy with a call to the existing helper function. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- drivers/net/ethernet/dnet.c | 8 +--- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git

[PATCHv2 net-next 03/22] phy: Use phy_read() instead of mdiobus_read()

Since we have a phydev, make use of it and the phy_read() function. This will help with later refactoring. Signed-off-by: Andrew Lunn --- v2: Revert hunks in IOCTL, which prevented access to arbitrary MDIO addresses. --- drivers/net/ethernet/hisilicon/hns/hns_ethtool.c | 104

[PATCHv2 net-next 11/22] phy: Add an mdio_device structure

Not all devices attached to an MDIO bus are phys. So add an mdio_device structure to represent the generic parts of an mdio device, and place this structure into the phy_device. Signed-off-by: Andrew Lunn --- drivers/net/ethernet/agere/et131x.c | 30 ++--

[PATCHv2 net-next 13/22] phy: Add API for {un}registering an mdio device to a bus.

Rather than have drivers directly manipulate the mii_bus structure, provide and API for registering and unregistering devices on an MDIO bus, and performing lookups. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli ---

[PATCHv2 net-next 21/22] mdio: Add support for mdio drivers.

Not all devices on an MDIO bus are PHYs. Meaning not all MDIO drivers are PHY drivers. Add support for generic MDIO drivers. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- drivers/net/phy/Makefile | 2 +- drivers/net/phy/mdio_bus.c

[PATCHv2 net-next 15/22] dsa: Register netdev before phy

When the phy is connected, an info message is printed. If the netdev it is attached to has not been registered yet, the name 'uninitialised' in the output. By registering the netdev first, then connecting they phy, we can avoid this. Signed-off-by: Andrew Lunn --- v2: call

[PATCHv2 net-next 09/22] phy: mdio-octeon: Use devm_mdiobus_alloc_size()

Rather than use devm_kzalloc(), use the mdio helper function. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- drivers/net/phy/mdio-octeon.c | 12 +--- 1 file changed, 5 insertions(+), 7 deletions(-) diff --git

[PATCHv2 net-next 00/22] Support MDIO devices

The discussions about changing the way DSA probes switches resulted in the wish to have switches attached to an MDIO bus to be represented as an MDIO device. However the current code only supports PHYs on MDIO busses. This patchset remedies this problem. It consists of a number of cleanups,

[PATCHv2 net-next 14/22] phy_device: Move phy attributes into phy_device

The mdio_bus exports three attributes: - PHY ID is the unique 32-bits identifier for a MDIO device implementing standard MII registers MII_PHYSID1/2, which is not guaranteed to be the case for non-standard compliant devices (e.g: Ethernet switches) - PHY interface describes the data-path of

[PATCHv2 net-next 01/22] phy: Consistently use addr for address on an MII bus

Within phy.h, an address on an MII bus has been called both addr and phy_id. phy_id is particularly confusion, since it also means the ID found in register 3, if the device on the bus is a phy. Consistently use addr. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli

[PATCHv2 net-next 05/22] phy: add phydev_name() wrapper

Add a phydev_name() function, to help with moving some structure members from phy_device. Signed-off-by: Andrew Lunn --- v2: Use inline function rather than macro --- drivers/net/ethernet/8390/ax88796.c | 2 +- drivers/net/ethernet/adi/bfin_mac.c | 4

[PATCHv2 net-next 20/22] mdio_bus: Add comment to mdiobus_scan() and __mdiobus_register()

Make it clear that mdiobus_scan () will only find devices which have a vendor/product ID in registers 2 and 3. These are typically PHY devices. Other sort of MDIO devices, such as switches, are not expected to be found during the scan. Similarly, __mdiobus_register(), which calls mdiobus_scan()

[PATCHv2 net-next 07/22] phy: phy_{read|write}_mmd_indirect: get addr from phydev

The address of the device can be determined from the phydev structure, rather than passing it as a parameter. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli --- drivers/net/phy/bcm-phy-lib.c | 8 drivers/net/phy/dp83867.c | 6

[PATCHv2 net-next 22/22] mdio: Abstract device_remove() and device_free()

Make device_free and device_remove operations in the mdio device structure, so the core code does not need to differentiate between phy devices and generic mdio devices. Signed-off-by: Andrew Lunn Reviewed-by: Florian Fainelli ---

[BUG] skb corruption and kernel panic at forwarding with fragmentation

I've got some of these: [84408.314676] BUG: unable to handle kernel NULL pointer dereference at (null) [84408.317324] IP: [] put_page+0x5/0x50 [84408.319985] PGD 0 [84408.322583] Oops: [#1] SMP [84408.325156] Modules linked in: ppp_mppe ppp_async ppp_generic slhc 8021q fuse nfsd

Re: [PATCH net-next 0/2] Local checksum offload for VXLAN

On 18/12/15 19:41, Tom Herbert wrote: > Drivers indicate that can do NETIF_F_IP[V6]_CSUM for encapsulation by > setting enc_features. This is checked in validate_xmit_skb so that if > drive can't handle encapsulated checksum skb_checksum_help is called > there. In that case, why is udp_set_csum

Re: [PATCH net-next 2/2] geneve: break dependency to network drivers

On Wed, Jan 6, 2016 at 7:41 AM, Hannes Frederic Sowa wrote: > diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c > index 24b077a32c1c9c..548925d1571cb1 100644 > --- a/drivers/net/geneve.c > +++ b/drivers/net/geneve.c > +static int geneve_notifier(struct

Re: [PATCH net-next 2/5] sctp: apply rhashtable api to send/recv path

On Tue, Jan 05, 2016 at 02:07:11PM -0500, Vlad Yasevich wrote: > On 12/30/2015 10:50 AM, Xin Long wrote: > > apply lookup apis to two functions, for __sctp_endpoint_lookup_assoc > > and __sctp_lookup_association, it's invoked in the protection of sock > > lock, it will be safe, but

header conflict introduced by change to netfilter_ipv4/ip_tables.h

This commit breaks compilation of iproute2 with net-next. commit 1ffad83dffd675cd742286ae82dca7d746cb0da8 Author: Mikko Rapeli Date: Thu Oct 15 07:56:30 2015 +0200 netfilter: fix include files for compilation Add missing header dependencies and other small

  1   2   3   >