There may be race condition when accessing inet_csk(sk)->icsk_bind_hash,
which leads to a crash like this:
Unable to handle kernel NULL pointer dereference at virtual
address 0010
..
[] inet_put_port+0x80/0xb8
[] tcp_set_state+0xc4/0x110
[] tcp_done+0x60/0xb8
[] tcp_nuke_addr+0x22c/0x2a0
[
From: Lisheng
The patch adds support of pause ctrl for HNS V2, and this feature is lost
by HNS V1:
1) service ports can disable rx pause frame,
2) debug ports can open tx/rx pause frame.
And this patch updates the REGs about the pause ctrl when updated
status function called by upp
This patch moves output message before setting slave to inactive, this will
print the correct status of slave device.
Signed-off-by: Zhang Shengju
---
drivers/net/bonding/bond_main.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/dri
From: Pontus Fuchs
Needed for coming improvements. No functional changes.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/hal.h | 7 +--
drivers/net/wireless/ath/wcn36xx/smd.c | 12 +---
2 files changed, 10 insertions(+), 9 deletion
From: Pontus Fuchs
Some code used priv_vif and some used vif_priv. Convert all to vif_priv
for consistency.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/smd.c | 28 ++--
1 file changed, 14 insertions(+), 14 deletions(-
As I was debugging some issues with the wcn36xx driver I found a series of
patches that exists in Eugene's github tree but not in mainline.
Noteworthy is the long requested feature of having working ARP support.
Please apply these rebased and relocated patches.
Jason Mobarak (1):
wcn36xx: don'
From: Pontus Fuchs
Makes the code a little easier to read.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/debug.c | 12 +++-
drivers/net/wireless/ath/wcn36xx/main.c| 16 +++-
drivers/net/wireless/ath/wcn36xx/pmc.c
From: Pontus Fuchs
The wcn36xx FW expects a fixed size TIM PVM in the beacon template. If
supplied with a shorter than expected PVM it will overwrite the IE
following the TIM.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/hal.h | 3 +++
drive
From: Pontus Fuchs
Pass the multicast list to FW.
This patch also adds a way to build the smd command in place. This is
needed because the MC list command is too big for the stack.
Signed-off-by: Pontus Fuchs
[bjorn: dropped FIF_PROMISC_IN_BSS usage]
Signed-off-by: Bjorn Andersson
Signed-off
From: Pontus Fuchs
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/smd.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/wcn36xx/smd.c
b/drivers/net/wireless/ath/wcn36xx/smd.c
index ce764cce1a3e..8b30
From: Pontus Fuchs
For consistency with other code.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/main.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/wcn36xx/main.c
b/drivers/net/wireless/ath/wc
From: Pontus Fuchs
This does not work with multiple sta's in a vif.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/main.c| 3 ---
drivers/net/wireless/ath/wcn36xx/smd.c | 28 +++-
drivers/net/wireless/ath/wcn36x
From: Pontus Fuchs
When converting to version 1 of the config_sta struct not all
members where copied. This fixes the problem of multicast frames
not being delivered on an encrypted network.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/smd.c
From: Pontus Fuchs
While poking at this I also change two related things. I rename one
variable to make the names consistent. I also move one assignment of
priv_sta to the declaration to save a few lines.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/
From: Pontus Fuchs
This message does not follow the canonical format and needs it's own
parser.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/smd.c | 14 --
1 file changed, 12 insertions(+), 2 deletions(-)
diff --git a/drivers/net
From: Pontus Fuchs
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/hal.h | 2 +-
drivers/net/wireless/ath/wcn36xx/main.c | 4 ++--
drivers/net/wireless/ath/wcn36xx/smd.c | 2 +-
3 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/dri
From: Pontus Fuchs
EXIT_BMPS_REQ was using the command struct for ENTER_BMPS_REQ. I
spotted this when looking at command dumps.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/smd.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --g
From: Jason Mobarak
Patch "wcn36xx: Pad TIM PVM if needed" has caused a regression in mesh
beaconing. The field tim_off is always 0 for mesh mode, and thus
pvm_len (referring to the TIM length field) and pad are both incorrectly
calculated. Thus, msg_body.beacon_length is incorrectly calculated
From: Pontus Fuchs
Knowing the association state is needed for mc filtering.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/main.c| 2 ++
drivers/net/wireless/ath/wcn36xx/wcn36xx.h | 1 +
2 files changed, 3 insertions(+)
diff --git a/drive
From: Pontus Fuchs
This fixes a problem connecting to an open network after being
connected to an encrypted network.
Signed-off-by: Pontus Fuchs
Signed-off-by: Bjorn Andersson
---
drivers/net/wireless/ath/wcn36xx/main.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/net/wireles
On Tue, Mar 29, 2016 at 12:47:20PM +0800, Wei-Ning Huang wrote:
> "single skb allocation failure" happens when system is under heavy
> memory pressure. Add __GFP_REPEAT to skb allocation call so kernel
> attempts to reclaim pages and retry the allocation.
Oh, that's interesting, we're back to thi
Hi Mugunthan,
On Tue, Mar 29, 2016 at 6:00 AM, Mugunthan V N wrote:
> Hi Yegor
>
> On Wednesday 16 March 2016 08:35 PM, Yegor Yefremov wrote:
>> I have an am335x based board using CPSW in Dual EMAC mode. Without
>> VLAN IDs I can receive and send multicast packets [1]. When I create
>> VLAN ID:
>
On Mon, Mar 28, 2016 at 9:42 AM, Xin Long wrote:
> There is an issue when we use mavtap over team:
> When we replug nic links from team0, the real nics's mc list will not
> include the maddr for macvtap any more. then we can't receive pkts to
> macvtap device, as they are filterred by mc list of n
On Mon, Mar 28, 2016 at 9:15 PM, Alex Duyck wrote:
> On Mon, Mar 28, 2016 at 9:01 PM, Tom Herbert wrote:
>> On Mon, Mar 28, 2016 at 8:27 PM, Alexander Duyck
>> wrote:
>>> On Mon, Mar 28, 2016 at 8:17 PM, Tom Herbert wrote:
On Mon, Mar 28, 2016 at 6:54 PM, Jesse Gross wrote:
> On Mon,
"single skb allocation failure" happens when system is under heavy
memory pressure. Add __GFP_REPEAT to skb allocation call so kernel
attempts to reclaim pages and retry the allocation.
Signed-off-by: Wei-Ning Huang
---
drivers/net/wireless/marvell/mwifiex/sdio.c | 12
1 file chang
On Mon, Jan 11, 2016 at 1:02 AM, Eugene Krasnikov wrote:
> Better late than never! Looks good to me.
>
Unfortunately I ran into an issue with ordering of operations between
the WiFi driver and the wcnss_ctrl driver. So an updated series is on
the way, but this depends on changes to the wcnss_ctrl
On Sun, Mar 27, 2016 at 10:42 PM, Martin KaFai Lau wrote:
>
> On Fri, Mar 25, 2016 at 04:05:51PM -0700, Yuchung Cheng wrote:
> > Looks like an interesting and useful patch. Since HTTP2 allows
> > multiplexing data stream frames from multiple logical streams on a
> > single socket,
> > how would yo
From: Xin Long
Date: Tue, 29 Mar 2016 00:42:31 +0800
> There is an issue when we use mavtap over team:
> When we replug nic links from team0, the real nics's mc list will not
> include the maddr for macvtap any more. then we can't receive pkts to
> macvtap device, as they are filterred by mc list
On Mon, Mar 28, 2016 at 9:01 PM, Tom Herbert wrote:
> On Mon, Mar 28, 2016 at 8:27 PM, Alexander Duyck
> wrote:
>> On Mon, Mar 28, 2016 at 8:17 PM, Tom Herbert wrote:
>>> On Mon, Mar 28, 2016 at 6:54 PM, Jesse Gross wrote:
On Mon, Mar 28, 2016 at 6:24 PM, Tom Herbert wrote:
> On Mon,
From: Michael Chan
Date: Mon, 28 Mar 2016 19:46:03 -0400
> Misc. small fixes for net.
Series applied, thanks Michael.
On Mon, Mar 28, 2016 at 8:27 PM, Alexander Duyck
wrote:
> On Mon, Mar 28, 2016 at 8:17 PM, Tom Herbert wrote:
>> On Mon, Mar 28, 2016 at 6:54 PM, Jesse Gross wrote:
>>> On Mon, Mar 28, 2016 at 6:24 PM, Tom Herbert wrote:
On Mon, Mar 28, 2016 at 4:58 PM, Alexander Duyck
wrote:
>
Hi Yegor
On Wednesday 16 March 2016 08:35 PM, Yegor Yefremov wrote:
> I have an am335x based board using CPSW in Dual EMAC mode. Without
> VLAN IDs I can receive and send multicast packets [1]. When I create
> VLAN ID:
>
> ip link add link eth1 name eth1.100 type vlan id 100
> ip addr add 192.168
On Mon, Mar 28, 2016 at 8:17 PM, Tom Herbert wrote:
> On Mon, Mar 28, 2016 at 6:54 PM, Jesse Gross wrote:
>> On Mon, Mar 28, 2016 at 6:24 PM, Tom Herbert wrote:
>>> On Mon, Mar 28, 2016 at 4:58 PM, Alexander Duyck
>>> wrote:
This patch should fix the issues seen with a recent fix to preve
On Mon, Mar 28, 2016 at 6:54 PM, Jesse Gross wrote:
> On Mon, Mar 28, 2016 at 6:24 PM, Tom Herbert wrote:
>> On Mon, Mar 28, 2016 at 4:58 PM, Alexander Duyck wrote:
>>> This patch should fix the issues seen with a recent fix to prevent
>>> tunnel-in-tunnel frames from being generated with GRO.
On Mon, Mar 28, 2016 at 4:58 PM, Alexander Duyck wrote:
> This patch should fix the issues seen with a recent fix to prevent
> tunnel-in-tunnel frames from being generated with GRO. The fix itself is
> correct for now as long as we do not add any devices that support
> NETIF_F_GSO_GRE_CSUM. When
On Mon, Mar 28, 2016 at 5:50 PM, Tom Herbert wrote:
> On Mon, Mar 28, 2016 at 4:34 PM, Jesse Gross wrote:
>> * A packet is received that is encapsulated with two layers of GRE. It
>> looks like this: Eth|IP|GRE|IP|GRE|IP|TCP
>> * The packet is processed through GRO successfully. skb->encapsulatio
On Mon, Mar 28, 2016 at 6:24 PM, Tom Herbert wrote:
> On Mon, Mar 28, 2016 at 4:58 PM, Alexander Duyck wrote:
>> This patch should fix the issues seen with a recent fix to prevent
>> tunnel-in-tunnel frames from being generated with GRO. The fix itself is
>> correct for now as long as we do not
On Mon, Mar 28, 2016 at 4:58 PM, Alexander Duyck wrote:
> This patch should fix the issues seen with a recent fix to prevent
> tunnel-in-tunnel frames from being generated with GRO. The fix itself is
> correct for now as long as we do not add any devices that support
> NETIF_F_GSO_GRE_CSUM. When
On Mon, 2016-03-28 at 19:54 -0400, David Miller wrote:
> From: Eric Dumazet
> Date: Mon, 28 Mar 2016 13:51:46 -0700
>
> > On Mon, 2016-03-28 at 13:46 -0700, Eric Dumazet wrote:
> >
> >> We have at least 384 bytes of padding in skb->head (this is struct
> >> skb_shared_info).
> >>
> >> Whatever
On Mon, Mar 28, 2016 at 4:58 PM, Alexander Duyck wrote:
> This patch should fix the issues seen with a recent fix to prevent
> tunnel-in-tunnel frames from being generated with GRO. The fix itself is
> correct for now as long as we do not add any devices that support
> NETIF_F_GSO_GRE_CSUM. When
On Mon, Mar 28, 2016 at 4:34 PM, Jesse Gross wrote:
> On Mon, Mar 28, 2016 at 3:10 PM, Tom Herbert wrote:
>> On Mon, Mar 28, 2016 at 1:34 PM, Alexander Duyck
>> wrote:
>>> On Mon, Mar 28, 2016 at 1:03 PM, Tom Herbert wrote:
On Mon, Mar 28, 2016 at 12:31 PM, Alexander Duyck
wrote:
>>>
From: Bjørn Mork
Date: Mon, 28 Mar 2016 22:38:16 +0200
> Thomas reports:
> "Windows:
...
> Linux:
...
> Reported-by: Thomas Schäfer
> Signed-off-by: Bjørn Mork
Applied and queued up for -stable, thanks!
This patch should fix the issues seen with a recent fix to prevent
tunnel-in-tunnel frames from being generated with GRO. The fix itself is
correct for now as long as we do not add any devices that support
NETIF_F_GSO_GRE_CSUM. When such a device is added it could have the
potential to mess thing
The typo caused the wrong flow control bit to be set.
Reported by: Ajit Khaparde
Signed-off-by: Michael Chan
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
b/drivers/net/ethernet/broadc
The size of every padded firmware message is specified in the first
HWRM_VER_GET response message. Use this value to pad every message
after that.
Signed-off-by: Michael Chan
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 6 +-
drivers/net/ethernet/broadcom/bnxt/bnxt.h | 2 ++
2 files chan
Misc. small fixes for net.
Michael Chan (4):
bnxt_en: Initialize CP doorbell value before ring allocation
bnxt_en: Implement proper firmware message padding.
bnxt_en: Fix typo in bnxt_hwrm_set_pause_common().
bnxt_en: Fix ethtool -a reporting.
drivers/net/ethernet/broadcom/bnxt/bnxt.c
To report flow control tx/rx settings accurately regardless of autoneg
setting, we should use link_info->req_flow_ctrl. Before this patch,
the reported settings were only correct when autoneg was on.
Signed-off-by: Michael Chan
---
drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 6 ++
1
From: Prashant Sreedharan
The existing code does the following:
allocate completion ring
initialize completion ring doorbell
disable interrupts on this completion ring by writing to the doorbell
We can have a race where firmware sends an asynchronous event to the host
after completio
From: Eric Dumazet
Date: Mon, 28 Mar 2016 13:51:46 -0700
> On Mon, 2016-03-28 at 13:46 -0700, Eric Dumazet wrote:
>
>> We have at least 384 bytes of padding in skb->head (this is struct
>> skb_shared_info).
>>
>> Whatever garbage we might read, current code is fine.
>>
>> We have to deal with
From: Jan Engelhardt
Date: Mon, 28 Mar 2016 22:20:39 +0200 (CEST)
>
> On Monday 2016-03-28 21:29, David Miller wrote:
> > @@ -3716,6 +3716,8 @@ void tcp_parse_options(const struct sk_buff *skb,
> > length--;
> > continue;
> > default:
>>>
On Mon, Mar 28, 2016 at 03:39:42PM -0700, Cong Wang wrote:
> On Fri, Mar 25, 2016 at 5:16 PM, Martin KaFai Lau wrote:
> > On Fri, Mar 25, 2016 at 04:55:27PM -0700, Martin KaFai Lau wrote:
> >> void ip6_sk_update_pmtu(struct sk_buff *skb, struct sock *sk, __be32 mtu)
> >> {
> >> + struct dst_
On Mon, Mar 28, 2016 at 3:10 PM, Tom Herbert wrote:
> On Mon, Mar 28, 2016 at 1:34 PM, Alexander Duyck
> wrote:
>> On Mon, Mar 28, 2016 at 1:03 PM, Tom Herbert wrote:
>>> On Mon, Mar 28, 2016 at 12:31 PM, Alexander Duyck
>>> wrote:
On Mon, Mar 28, 2016 at 11:47 AM, Tom Herbert wrote:
On Fri, Mar 25, 2016 at 5:16 PM, Martin KaFai Lau wrote:
> On Fri, Mar 25, 2016 at 04:55:27PM -0700, Martin KaFai Lau wrote:
>> void ip6_sk_update_pmtu(struct sk_buff *skb, struct sock *sk, __be32 mtu)
>> {
>> + struct dst_entry *odst;
>> +
>> + odst = sk_dst_get(sk);
>> +
>> ip6_u
On Fri, Mar 25, 2016 at 11:11 AM, Eric Dumazet wrote:
> On Fri, 2016-03-25 at 10:17 -0700, Cong Wang wrote:
>
>> 1) sock lock protects the whole update: the whole check, update, recheck,
>> set logic, to make sure another CPU will not do the same to the same socket
>> at the same time.
>>
>> 2) th
On Mon, Mar 28, 2016 at 1:34 PM, Alexander Duyck
wrote:
> On Mon, Mar 28, 2016 at 1:03 PM, Tom Herbert wrote:
>> On Mon, Mar 28, 2016 at 12:31 PM, Alexander Duyck
>> wrote:
>>> On Mon, Mar 28, 2016 at 11:47 AM, Tom Herbert wrote:
On Mon, Mar 28, 2016 at 10:37 AM, Alexander Duyck
wrot
On 03/25/2016 11:10 PM, Andrew Collins wrote:
> On 03/25/2016 02:43 PM, Matthias Schiffer wrote:
>> We've tried your patch, and it changes the symptoms a bit, but doesn't fix
>> the panic. I've attached kernel logs of the crash both before and after
>> applying the patch.
>>
>> One note: I did not
On Mon, 2016-03-28 at 23:11 +0200, Jozsef Kadlecsik wrote:
> In net/netfilter/nf_conntrack_proto_tcp.c we copy the options into a
> buffer with skb_header_pointer(), so it's not a false positive there and
> the KASAN report referred to that part.
>
Although the out of bound could be one extra
On Mon, 28 Mar 2016, Eric Dumazet wrote:
> On Mon, 2016-03-28 at 22:20 +0200, Jan Engelhardt wrote:
> > On Monday 2016-03-28 21:29, David Miller wrote:
> > >>> > > @@ -3716,6 +3716,8 @@ void tcp_parse_options(const struct sk_buff
> > >>> > > *skb,
> > >>> > > length--;
> > >>> > >
On Fri, 2016-03-25 at 17:08 -0700, Tom Herbert wrote:
> On Fri, Mar 25, 2016 at 3:29 PM, Eric Dumazet wrote:
> > +/* Must be called under rcu_read_lock().
>
>
> It might be just as easy to do the rcu_read_lock() within the
> function. That way we don't need to require callers to do it now.
>
>
On Mon, 2016-03-28 at 13:46 -0700, Eric Dumazet wrote:
> We have at least 384 bytes of padding in skb->head (this is struct
> skb_shared_info).
>
> Whatever garbage we might read, current code is fine.
>
> We have to deal with a false positive here.
Very similar to the one fixed in
https://git
On Mon, 2016-03-28 at 22:20 +0200, Jan Engelhardt wrote:
> On Monday 2016-03-28 21:29, David Miller wrote:
> >>> > > @@ -3716,6 +3716,8 @@ void tcp_parse_options(const struct sk_buff
> >>> > > *skb,
> >>> > > length--;
> >>> > > continue;
> >>> > > default:
>
This commit adds the support for netpoll, which is used
to implement netconsole.
Signed-off-by: Ezequiel Garcia
---
Tested on Armada 370 Mirabox and Armada XP Openblocks AX3-4
with netconsole.
drivers/net/ethernet/marvell/mvneta.c | 21 +
1 file changed, 21 insertions(+)
di
Thomas reports:
"Windows:
00 diagnostics
01 modem
02 at-port
03 nmea
04 nic
Linux:
T: Bus=02 Lev=01 Prnt=01 Port=03 Cnt=01 Dev#= 4 Spd=480 MxCh= 0
D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1
P: Vendor=2001 ProdID=7e19 Rev=02.32
S: Manufacturer=Mobile Connect
S: Product=Mobi
On Mon, Mar 28, 2016 at 1:03 PM, Tom Herbert wrote:
> On Mon, Mar 28, 2016 at 12:31 PM, Alexander Duyck
> wrote:
>> On Mon, Mar 28, 2016 at 11:47 AM, Tom Herbert wrote:
>>> On Mon, Mar 28, 2016 at 10:37 AM, Alexander Duyck
>>> wrote:
On Mon, Mar 28, 2016 at 9:31 AM, Tom Herbert wrote:
>>>
On Monday 2016-03-28 21:29, David Miller wrote:
>>> > > @@ -3716,6 +3716,8 @@ void tcp_parse_options(const struct sk_buff *skb,
>>> > > length--;
>>> > > continue;
>>> > > default:
>>> > > +if (length < 2)
>>> > > +return;
>>> > >
On 03/28/2016 01:01 PM, Eric Dumazet wrote:
Note : file structures got RCU freeing back in 2.6.14, and I do not
think named users ever complained about added cost ;)
Couldn't see the tree for the forest I guess :)
rick
On Mon, 2016-03-28 at 15:29 -0400, David Miller wrote:
> From: Jozsef Kadlecsik
> Date: Mon, 28 Mar 2016 18:48:51 +0200 (CEST)
>
> >> > > @@ -3716,6 +3716,8 @@ void tcp_parse_options(const struct sk_buff *skb,
> >> > > length--;
> >> > > continue;
> >> > > de
On Mon, Mar 28, 2016 at 12:31 PM, Alexander Duyck
wrote:
> On Mon, Mar 28, 2016 at 11:47 AM, Tom Herbert wrote:
>> On Mon, Mar 28, 2016 at 10:37 AM, Alexander Duyck
>> wrote:
>>> On Mon, Mar 28, 2016 at 9:31 AM, Tom Herbert wrote:
On Sun, Mar 27, 2016 at 9:38 PM, Jesse Gross wrote:
>
On Mon, 2016-03-28 at 12:11 -0700, Rick Jones wrote:
> I was under the impression that individual DNS queries were supposed to
> have not only random DNS query IDs but also originate from random UDP
> source ports. https://tools.ietf.org/html/rfc5452 4.5 at least touches
> on the topic but I d
From: Pablo Neira Ayuso
Date: Mon, 28 Mar 2016 19:57:53 +0200
> The following patchset contains Netfilter fixes for you net tree,
> they are:
...
> This batch comes with four patches to validate x_tables blobs coming
> from userspace. CONFIG_USERNS exposes the x_tables interface to
> unpriviledg
From: Jesse Gross
Date: Sun, 27 Mar 2016 21:38:34 -0700
> Generally speaking, multiple levels of encapsulation offload are not
> valid.
That's very much not true Jesse.
Please fix the regression you introduced or I will have no choice but to
revert your entire set of changes.
On Mon, Mar 28, 2016 at 11:47 AM, Tom Herbert wrote:
> On Mon, Mar 28, 2016 at 10:37 AM, Alexander Duyck
> wrote:
>> On Mon, Mar 28, 2016 at 9:31 AM, Tom Herbert wrote:
>>> On Sun, Mar 27, 2016 at 9:38 PM, Jesse Gross wrote:
On Sat, Mar 26, 2016 at 12:41 PM, Tom Herbert wrote:
> On Sa
From: Jozsef Kadlecsik
Date: Mon, 28 Mar 2016 18:48:51 +0200 (CEST)
>> > > @@ -3716,6 +3716,8 @@ void tcp_parse_options(const struct sk_buff *skb,
>> > > length--;
>> > > continue;
>> > > default:
>> > > +if (length < 2)
>> > > +re
On 03/28/2016 11:55 AM, Eric Dumazet wrote:
On Mon, 2016-03-28 at 11:44 -0700, Rick Jones wrote:
On 03/28/2016 10:00 AM, Eric Dumazet wrote:
If you mean that a busy DNS resolver spends _most_ of its time doing :
fd = socket()
bind(fd port=0)
< send and receive one frame >
close(fd)
Yes.
On Mon, 2016-03-28 at 11:44 -0700, Rick Jones wrote:
> On 03/28/2016 10:00 AM, Eric Dumazet wrote:
> > On Mon, 2016-03-28 at 09:15 -0700, Rick Jones wrote:
> >> On 03/25/2016 03:29 PM, Eric Dumazet wrote:
> >>> UDP sockets are not short lived in the high usage case, so the added
> >>> cost of call_
On Mon, Mar 28, 2016 at 10:37 AM, Alexander Duyck
wrote:
> On Mon, Mar 28, 2016 at 9:31 AM, Tom Herbert wrote:
>> On Sun, Mar 27, 2016 at 9:38 PM, Jesse Gross wrote:
>>> On Sat, Mar 26, 2016 at 12:41 PM, Tom Herbert wrote:
On Sat, Mar 19, 2016 at 9:32 AM, Jesse Gross wrote:
> When dri
On 03/28/2016 10:00 AM, Eric Dumazet wrote:
On Mon, 2016-03-28 at 09:15 -0700, Rick Jones wrote:
On 03/25/2016 03:29 PM, Eric Dumazet wrote:
UDP sockets are not short lived in the high usage case, so the added
cost of call_rcu() should not be a concern.
Even a busy DNS resolver?
If you mean
Mon, Mar 28, 2016 at 06:42:31PM CEST, lucien@gmail.com wrote:
>There is an issue when we use mavtap over team:
>When we replug nic links from team0, the real nics's mc list will not
>include the maddr for macvtap any more. then we can't receive pkts to
>macvtap device, as they are filterred by
From: Vishwanath Pai
This fix adds a new reference counter (ref_netlink) for the struct ip_set.
The other reference counter (ref) can be swapped out by ip_set_swap and we
need a separate counter to keep track of references for netlink events
like dump. Using the same ref counter for dump causes a
Hi David,
The following patchset contains Netfilter fixes for you net tree,
they are:
1) There was a race condition between parallel save/swap and delete,
which resulted a kernel crash due to the increase ref for save, swap,
wrong ref decrease operations. Reported and fixed by Vishwanath Pa
When netlink unicast fails to deliver the message to userspace, we
should also check if the NFQA_CFG_F_FAIL_OPEN flag is set so we reinject
the packet back to the stack.
I think the user expects no packet drops when this flag is set due to
queueing to userspace errors, no matter if related to the
From: Arnd Bergmann
The openvswitch code has gained support for calling into the
nf-nat-ipv4/ipv6 modules, however those can be loadable modules
in a configuration in which openvswitch is built-in, leading
to link errors:
net/built-in.o: In function `__ovs_ct_lookup':
:(.text+0x2cc2c8): undefine
From: Florian Westphal
We should check that e->target_offset is sane before
mark_source_chains gets called since it will fetch the target entry
for loop detection.
Signed-off-by: Florian Westphal
Signed-off-by: Pablo Neira Ayuso
---
net/ipv4/netfilter/arp_tables.c | 17 -
net/
From: Florian Westphal
Otherwise this function may read data beyond the ruleset blob.
Signed-off-by: Florian Westphal
Signed-off-by: Pablo Neira Ayuso
---
net/ipv4/netfilter/arp_tables.c | 6 --
net/ipv4/netfilter/ip_tables.c | 6 --
net/ipv6/netfilter/ip6_tables.c | 6 --
3 file
From: Liping Zhang
Commit fa50d974d104 ("ipv4: Namespaceify ip_default_ttl sysctl knob")
use sock_net(skb->sk) to get the net namespace, but we can't assume
that sk_buff->sk is always exist, so when it is NULL, oops will happen.
Signed-off-by: Liping Zhang
Reviewed-by: Nikolay Borisov
Signed-o
Make sure the table names via getsockopt GET_ENTRIES is nul-terminated
in ebtables and all the x_tables variants and their respective compat
code. Uncovered by KASAN.
Reported-by: Baozeng Ding
Signed-off-by: Pablo Neira Ayuso
---
net/bridge/netfilter/ebtables.c | 4
net/ipv4/netfilter/arp_
From: Jarno Rajahalme
OVS should call into CT NAT for packets of new expected connections only
when the conntrack state is persisted with the 'commit' option to the
OVS CT action. The test for this condition is doubly wrong, as the CT
status field is ANDed with the bit number (IPS_EXPECTED_BIT)
From: Florian Westphal
Ben Hawkes says:
In the mark_source_chains function (net/ipv4/netfilter/ip_tables.c) it
is possible for a user-supplied ipt_entry structure to have a large
next_offset field. This field is not bounds checked prior to writing a
counter value at the supplied offset.
Pro
On Mon, Mar 28, 2016 at 9:31 AM, Tom Herbert wrote:
> On Sun, Mar 27, 2016 at 9:38 PM, Jesse Gross wrote:
>> On Sat, Mar 26, 2016 at 12:41 PM, Tom Herbert wrote:
>>> On Sat, Mar 19, 2016 at 9:32 AM, Jesse Gross wrote:
When drivers express support for TSO of encapsulated packets, they
Hello.
On 03/28/2016 03:47 PM, Charles Keepax wrote:
GPIOlib will print warning messages if we call GPIO functions without a
valid GPIO. Change the code to avoid doing so.
Signed-off-by: Charles Keepax
---
drivers/net/ethernet/cadence/macb.c | 8 +---
1 file changed, 5 insertions(+), 3
On Mon, Mar 28, 2016 at 06:48:51PM +0200, Jozsef Kadlecsik wrote:
> Hi David, Pablo,
>
> David, do you agree with the patch for net/ipv4/tcp_input.c? If yes, how
> should I proceed? Should I send the whole patch to you or is it OK to send
> to Pablo?
Submit a formal patch and Cc: netdev@vger.ke
On Mon, 2016-03-28 at 09:15 -0700, Rick Jones wrote:
> On 03/25/2016 03:29 PM, Eric Dumazet wrote:
> > UDP sockets are not short lived in the high usage case, so the added
> > cost of call_rcu() should not be a concern.
>
> Even a busy DNS resolver?
If you mean that a busy DNS resolver spends _mo
On Mon, Mar 28, 2016 at 9:15 AM, Rick Jones wrote:
> On 03/25/2016 03:29 PM, Eric Dumazet wrote:
>>
>> UDP sockets are not short lived in the high usage case, so the added
>> cost of call_rcu() should not be a concern.
>
>
> Even a busy DNS resolver?
>
Should use connectionless UDP sockets.
> ric
On Tue, Mar 29, 2016 at 12:42:31AM +0800, Xin Long wrote:
> There is an issue when we use mavtap over team:
> When we replug nic links from team0, the real nics's mc list will not
> include the maddr for macvtap any more. then we can't receive pkts to
> macvtap device, as they are filterred by mc l
Hi David, Pablo,
David, do you agree with the patch for net/ipv4/tcp_input.c? If yes, how
should I proceed? Should I send the whole patch to you or is it OK to send
to Pablo?
Best regards,
Jozsef
On Mon, 28 Mar 2016, Baozeng Ding wrote:
>
>
> On 2016/3/28 10:35, Baozeng Ding wrote:
> >
> >
There is an issue when we use mavtap over team:
When we replug nic links from team0, the real nics's mc list will not
include the maddr for macvtap any more. then we can't receive pkts to
macvtap device, as they are filterred by mc list of nic.
In Bonding Driver, it syncs the uc/mc addrs in bond_e
On Sun, Mar 27, 2016 at 9:38 PM, Jesse Gross wrote:
> On Sat, Mar 26, 2016 at 12:41 PM, Tom Herbert wrote:
>> On Sat, Mar 19, 2016 at 9:32 AM, Jesse Gross wrote:
>>> When drivers express support for TSO of encapsulated packets, they
>>> only mean that they can do it for one layer of encapsulatio
On Sun, Mar 27, 2016 at 10:36 PM, Jesse Gross wrote:
> On Fri, Mar 18, 2016 at 4:25 PM, Alexander Duyck wrote:
>> diff --git a/net/core/dev.c b/net/core/dev.c
>> index edb7179bc051..666cf427898b 100644
>> --- a/net/core/dev.c
>> +++ b/net/core/dev.c
>> @@ -2711,6 +2711,19 @@ struct sk_buff *__skb
On 03/25/2016 03:29 PM, Eric Dumazet wrote:
UDP sockets are not short lived in the high usage case, so the added
cost of call_rcu() should not be a concern.
Even a busy DNS resolver?
rick jones
Hi Stephen
> There already is LED control via ethtool.
> It is more important that the existing API (ethtool) still work.
I'm having trouble finding this. The man page for ethtool only
mentions -p --identify with respect to LEDs.
I also don't see anything in include/uapi/linux/ethtool.h
Please
1 - 100 of 127 matches
Mail list logo
