On 05/31/2018 04:31 AM, Wei Yongjun wrote:
> Fix to return error code -EINVAL instead of 0 if optlen is invalid.
>
> Fixes: 01d2f7e2cdd3 ("net/smc: sockopts TCP_NODELAY and TCP_CORK")
> Signed-off-by: Wei Yongjun
> ---
> net/smc/af_smc.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
On 05/31/2018 06:00 PM, Eric Dumazet wrote:
> On Thu, May 31, 2018 at 7:32 PM John Fastabend
> wrote:
>>
>>
>> Hi Wei,
>>
>> Thanks for the report and fix. It would be better to fix the
>> root cause so that IPv6 works as intended.
>>
>> I'm testing the following now,
>>
>> Author: John Fastabend
In rtnl_newlink(), NULL check is performed on m_ops however member of
ops is accessed. Fixed by accessing member of m_ops instead of ops.
[ 345.432629] BUG: KASAN: null-ptr-deref in rtnl_newlink+0x400/0x1110
[ 345.432629] Read of size 4 at addr 0088 by task ip/986
[ 345.432629]
[ 3
Since commit 049c58539f5d ("devlink: mnlg: Add support for extended ack")
devlink requires NETLINK_{CAP,EXT}_ACK. This prevents devlink from
working with older kernels that don't support these features.
host # ./devlink/devlink
Failed to connect to devlink Netlink
Fixes: 049c58539f5d ("devlink: m
This fixes a crash where we assign tcp_prot to IPv6 sockets instead
of tcpv6_prot.
Previously we overwrote the sk->prot field with tcp_prot even in the
AF_INET6 case. This patch ensures the correct tcp_prot and tcpv6_prot
are used.
Fixes: 174a79ff9515 ("bpf: sockmap with sk redirect support")
Rep
On Fri, Jun 1, 2018 at 4:18 AM Prashant Bhole
wrote:
>
> In rtnl_newlink(), NULL check is performed on m_ops however member of
> ops is accessed. Fixed by accessing member of m_ops instead of ops.
>
> [ 345.432629] BUG: KASAN: null-ptr-deref in rtnl_newlink+0x400/0x1110
> [ 345.432629] Read of s
Hello!
On 5/31/2018 1:14 PM, Tonghao Zhang wrote:
Use the % instead of while, and it may simple code and improve
the calculating. The real_num_tx_queues has been checked when
allocating and setting it.
Signed-off-by: Tonghao Zhang
---
net/core/dev.c | 8 +++-
1 file changed, 3 insertio
>> +
>> +spin_lock(&aq_nic->aq_spinlock);
>> +
>> +if (netif_running(ndev))
>> +dev_close(ndev);
>
> I don't think you can hold a spinlock around dev_close()/dev_open()
> calls.
Thanks Jakub, think you are right, will consider changing this lock to mutex.
>> +if (!netif
This series is an early preview of ongoing work to support loops
inside BPF verifier. The code is rough in spots (more rough the
further you get into the series) but we have some signs of life!
The following code runs and is verified,
SEC("classifier_tc_loop1")
int _tc_loop(struct __sk_buff *ct
From: Jiong Wang
"check_subprogs" has partitioned subprogs and is doing insn scan for each
subprog already, this patch extends it to also partition basic blocks (BB)
for each subprog.
- The first insn for each subprog start a BB.
- Branch (both cond and absolute) target start a BB.
- Insn
From: Jiong Wang
- When building domination information, we need to some array data
structure, and need to index into them using basic block index.
Calculate the basic block index during adding edges.
- The Step-1 in Tarjan algorithm is to assign dfs order number for
each bb in t
From: Jiong Wang
Do unreachable basic blocks detection as a side-product of the dfs walk
when building domination information.
Signed-off-by: Jiong Wang
Signed-off-by: John Fastabend
---
kernel/bpf/cfg.c | 19 ++-
kernel/bpf/cfg.h |3 ++-
kernel/bpf/verifier.c
From: Jiong Wang
If one bb is dominating its predecessor, then there is loop.
Signed-off-by: Jiong Wang
Signed-off-by: John Fastabend
---
kernel/bpf/cfg.c | 22 ++
kernel/bpf/cfg.h |1 +
kernel/bpf/verifier.c |8
3 files changed, 31 insertions(
From: Jiong Wang
This patch add edges between basic blocks. Both edges for predecessors and
successors are added.
Signed-off-by: Jiong Wang
Signed-off-by: John Fastabend
---
kernel/bpf/cfg.c | 129 -
kernel/bpf/cfg.h |1
kernel/b
From: Jiong Wang
As we have detected loop and unreachable insns based on domination
information and call graph, there is no need of check_cfg.
This patch removes check_cfg and it's associated push_insn.
state prune heuristic marking as moved to check_subprog.
Signed-off-by: Jiong Wang
---
ke
From: Jiong Wang
This patch build call graph during insn scan inside check_subprogs.
Then do recursive and unreachable subprog detection using call graph.
Signed-off-by: Jiong Wang
Signed-off-by: John Fastabend
---
include/linux/bpf_verifier.h |1
kernel/bpf/cfg.c | 133
From: Jiong Wang
During building control flow graph, we need to build basic block nodes
and edge nodes etc. These nodes needs to allocated dynamically as we
don't have pre-scan pass to know how many nodes we need accurately.
It is better to manage their allocation using memory pool which could
re
From: Jiong Wang
This patch centre find_subprog and add_subprog to cfg.c.
Signed-off-by: Jiong Wang
Signed-off-by: John Fastabend
---
kernel/bpf/cfg.c | 41 +
kernel/bpf/cfg.h |2 ++
kernel/bpf/verifier.c | 42 -
From: Jiong Wang
Because there are going to be quite a few nodes (bb, edge etc) for a
reasonable sized program, the size of cfg nodes matters.
The cfg traversal used at the moment are mostly via edge which contains
pointers to both src and dst basic block. The traversal is not via links
between
At the moment any BPF_ADD or BPF_NEG with a pointer type will create
a new pointer and destroy the register range. Then any memory access
after this will fail because it looks like no bounds checking has
been done, even if it was previously done on the other pointer. So
patterns like this fail,
From: Jiong Wang
We don't want to do any further loop bounds analysis for irreducible loop,
so just reject programs containing it.
The current DOM based loop detection can't detect irreducible loop. We'd
use the algorithm given by Eric Stoltz to detect it. The algorithm requires
DOM info.
Algor
Add functions to pretty print the CFG and DOM table. We can also
add contrib scripts to print this dot format so tools can visualize
them. I at least found this helpful. Will add scripts in tools/bpf
follow up patch.
For development we are always printing these but should put these
noisy routines
Allow bounded loops if we can become convinced they will in fact
terminate.
At a high level this is done in steps the first two steps are done
outside of the do_check routine which "runs" the instructions.
1. Use dom tree to find loops.
2. Find loop induction variable in the loop. (I use loop
Add some simple loop examples. For now I just load these
using bpftool but eventually they should have a loader
simliar to test_verifier except for C snippets.
We want to use C files here instead of BPF because most
users will be using C clang/llvm and want to test how
well this works with the cod
While developing this I found it useful to always have the log
output when dealing with bpftool. This is a quick hack to enable
it but we should add a '-v' option shortly.
Then add a set of good and bad loop examples to test with. These
are all very basic at the moment but will get better soon.
S
On 05/31/2018 06:14 AM, Tonghao Zhang wrote:
> Use the % instead of while, and it may simple code and improve
> the calculating. The real_num_tx_queues has been checked when
> allocating and setting it.
>
> Signed-off-by: Tonghao Zhang
> ---
> net/core/dev.c | 8 +++-
> 1 file changed, 3
On 05/31/2018 09:51 PM, Qing Huang wrote:
>
> It would be great if you could share the test case that triggered the KASAN
> report in your
> environment. Our QA has been running intensive tests using 8KB or 4KB chunk
> size configuration
> for some time, no one has reported memory corruption
RFC4821 say: The value for this timer MUST NOT be less than
5 minutes and is recommended to be 10 minutes, per RFC 1981.
Signed-off-by: Li RongQing
---
net/ipv4/sysctl_net_ipv4.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_n
On 31/05/18 08:38 AM, Vlad Buslov wrote:
Hi Jamal,
On current net-next I still have action with single reference after last
step:
~$ sudo $TC -s actions ls action skbedit
total acts 1
action order 0: skbedit mark 1 pipe
On Fri, Jun 01, 2018 at 05:16:58PM +0900, Prashant Bhole wrote:
> In rtnl_newlink(), NULL check is performed on m_ops however member of
> ops is accessed. Fixed by accessing member of m_ops instead of ops.
>
> [ 345.432629] BUG: KASAN: null-ptr-deref in rtnl_newlink+0x400/0x1110
> [ 345.432629]
On 05/30/2018 09:13 PM, Prashant Bhole wrote:
>
>
> On 5/31/2018 4:59 AM, John Fastabend wrote:
>> On 05/30/2018 12:29 PM, Alexei Starovoitov wrote:
>>> On Wed, May 30, 2018 at 02:56:09PM +0900, Prashant Bhole wrote:
In order to reduce runtime of tests, recently timout for select() call
On 05/30/2018 09:42 PM, Prashant Bhole wrote:
> Test failures are not identified because exit code of RX/TX threads
> is not checked. Also threads are not returning correct exit code.
>
> - Return exit code from threads depending on test execution status
> - In main thread, check the exit code of
On receiving a IGMPv2/v3 query, based on max_delay set in the header a
timer is started to send out a response after a random time within
max_delay. If the system then moves into suspend state, Report is
delayed until system wakes up.
In one reported scenario, on arm64 devices, max_delay was set t
On 05/30/2018 09:42 PM, Prashant Bhole wrote:
> Currently 10us delay is too low for many tests to succeed. It needs to
> be increased. Also, many corked tests are expected to hit rx timeout
> irrespective of timeout value.
>
> - This patch sets 1000usec timeout value for corked tests because less
From: Prashant Bhole
Date: Fri, 1 Jun 2018 17:16:58 +0900
> In rtnl_newlink(), NULL check is performed on m_ops however member of
> ops is accessed. Fixed by accessing member of m_ops instead of ops.
>
> [ 345.432629] BUG: KASAN: null-ptr-deref in rtnl_newlink+0x400/0x1110
> [ 345.432629] Rea
On 06/01/2018 07:05 AM, Tejaswi Tanikella wrote:
> On receiving a IGMPv2/v3 query, based on max_delay set in the header a
> timer is started to send out a response after a random time within
> max_delay. If the system then moves into suspend state, Report is
> delayed until system wakes up.
>
>
Fri, Jun 01, 2018 at 10:18:49AM CEST, ivec...@redhat.com wrote:
>Since commit 049c58539f5d ("devlink: mnlg: Add support for extended ack")
>devlink requires NETLINK_{CAP,EXT}_ACK. This prevents devlink from
>working with older kernels that don't support these features.
>
>host # ./devlink/devlink
>
From: Ursula Braun
smc_setsockopt contains a check for the length specified on the
setsockopt socket call. If checked, it is supposed to return with
EINVAL. But the equivalent check is already contained in the preceding
setsockopt call on the internal TCP socket. That means, the extra
check can b
Le 31/05/2018 à 17:51, Nicolas Dichtel a écrit :
> Le 31/05/2018 à 17:46, Stephen Hemminger a écrit :
>> On Thu, 31 May 2018 16:28:48 +0200
> [snip]
>> This makes sense. All of linkinfo that is present should be displayed.
>>
>> Both netns and ifindex are really unsigned values. Use __u32 and print
On Thu, 2018-05-31 at 15:36 -0400, David Miller wrote:
> From: Saeed Mahameed
> Date: Wed, 30 May 2018 10:59:48 -0700
>
> > The following series is for mlx5-next tree [1], it adds the support of two
> > new device events, from Ilan Tayari:
> >
> > 1. High temperature warnings.
> > 2. FPGA QP err
From: Vlad Buslov
Date: Thu, 31 May 2018 09:52:53 +0300
> tc_ctl_tfilter handles three netlink message types: RTM_NEWTFILTER,
> RTM_DELTFILTER, RTM_GETTFILTER. However, implementation of this function
> involves a lot of branching on specific message type because most of the
> code is message-spe
From: Damien Thébault
Date: Thu, 31 May 2018 07:04:01 +
> This patch adds support for the BCM5389 switch connected through MDIO.
>
> Signed-off-by: Damien Thébault
Applied, thank you.
On Fri 01 Jun 2018 at 12:24, Jamal Hadi Salim wrote:
> On 31/05/18 08:38 AM, Vlad Buslov wrote:
>
>> Hi Jamal,
>>
>> On current net-next I still have action with single reference after last
>> step:
>> ~$ sudo $TC -s actions ls action skbedit
>> total acts 1
>>
Implement functions to atomically update and free action cookie
using rcu mechanism.
Reviewed-by: Marcelo Ricardo Leitner
Signed-off-by: Vlad Buslov
Signed-off-by: Jiri Pirko
---
include/net/act_api.h | 2 +-
include/net/pkt_cls.h | 1 +
net/sched/act_api.c | 44 +++
Extend action ops with 'delete' function. Each action type to implements
its own delete function that doesn't depend on rtnl lock.
Implement delete function that is required to delete actions without
holding rtnl lock. Use action API function that atomically deletes action
only if it is still in a
Currently, all netlink protocol handlers for updating rules, actions and
qdiscs are protected with single global rtnl lock which removes any
possibility for parallelism. This patch set is a first step to remove
rtnl lock dependency from TC rules update path.
Recently, new rtnl registration flag RT
Implement new action API function that atomically finds and deletes action
from idr by index. Intended to be used by lockless actions that do not rely
on rtnl lock.
Reviewed-by: Marcelo Ricardo Leitner
Signed-off-by: Vlad Buslov
Signed-off-by: Jiri Pirko
---
Changes from V1 to V2:
- Rename tcf_
Implement helper delete function that uses new action ops 'delete', instead
of destroying action directly. This is required so act API could delete
actions by index, without holding any references to action that is being
deleted.
Implement function __tcf_action_put() that releases reference to act
Implement function that atomically checks if action exists and either takes
reference to it, or allocates idr slot for action index to prevent
concurrent allocations of actions with same index. Use EBUSY error pointer
to indicate that idr slot is reserved.
Implement cleanup helper function that re
Return from action init function with reference to action taken,
even when overwriting existing action.
Action init API initializes its fourth argument (pointer to pointer to tc
action) to either existing action with same index or newly created action.
In case of existing index(and bind argument i
Change action API to assume that action init function always takes
reference to action, even when overwriting existing action. This is
necessary because action API continues to use action pointer after init
function is done. At this point action becomes accessible for concurrent
modifications, so u
Act API used linked list to pass set of actions to functions. It is
intrusive data structure that stores list nodes inside action structure
itself, which means it is not safe to modify such list concurrently.
However, action API doesn't use any linked list specific operations on this
set of actions
Without rtnl lock protection it is no longer safe to use pointer to tc
action without holding reference to it. (it can be destroyed concurrently)
Remove unsafe action idr lookup function. Instead of it, implement safe tcf
idr check function that atomically looks up action in idr and increments
its
Add additional 'rtnl_held' argument to act API init functions. It is
required to implement actions that need to release rtnl lock before loading
kernel module and reacquire if afterwards.
Reviewed-by: Marcelo Ricardo Leitner
Signed-off-by: Vlad Buslov
Signed-off-by: Jiri Pirko
---
Changes from
Change type of action reference counter to refcount_t.
Change type of action bind counter to atomic_t.
This type is used to allow decrementing bind counter without testing
for 0 result.
Reviewed-by: Marcelo Ricardo Leitner
Signed-off-by: Vlad Buslov
Signed-off-by: Jiri Pirko
---
include/net/a
Use the right device to determine if redirect should be sent especially
when using vrf. Same as well as when sending the redirect.
Signed-off-by: Stephen Suryaputra
---
net/ipv6/ip6_output.c | 3 ++-
net/ipv6/ndisc.c | 6 ++
2 files changed, 8 insertions(+), 1 deletion(-)
diff --git a/
On 5/30/18 11:06 AM, Roopa Prabhu wrote:
> From: Roopa Prabhu
>
> Signed-off-by: Roopa Prabhu
> ---
> ip/iproute.c | 26 +-
> man/man8/ip-route.8.in | 20 +++-
> 2 files changed, 44 insertions(+), 2 deletions(-)
applied to iproute2-next. Thanks
From: Doug Ledford
Date: Fri, 01 Jun 2018 11:08:24 -0400
> On Thu, 2018-05-31 at 15:36 -0400, David Miller wrote:
>> From: Saeed Mahameed
>> Date: Wed, 30 May 2018 10:59:48 -0700
>>
>> > The following series is for mlx5-next tree [1], it adds the support of two
>> > new device events, from Ilan
From: David Ahern
A recent commit changed rtnl_talk_* to return the response message in
allocated memory so callers need to free it. The change to name_is_vrf
did not save the device index which is pointing to a struct inside the
now allocated and freed memory resulting in garbage getting returne
Hi netdev community,
I am again on bug fixing in 6lowpan branch and thought that my
needed_headroom and needed_tailroom of net_device are available inside
my ndo_start_xmit() callback.
In case of a UDP socket, it was not the case. I send a fix now to use
skb_expand_copy() to make sure this space
On Fri, Jun 01, 2018 at 11:45:58AM -0400, David Miller wrote:
> From: Doug Ledford
> Date: Fri, 01 Jun 2018 11:08:24 -0400
>
> > On Thu, 2018-05-31 at 15:36 -0400, David Miller wrote:
> >> From: Saeed Mahameed
> >> Date: Wed, 30 May 2018 10:59:48 -0700
> >>
> >> > The following series is for mlx5
syzbot was able to trick af_packet again [1]
Various commits tried to address the problem in the past,
but failed to take into account V3 header size.
[1]
tpacket_rcv: packet too big, clamped from 72 to 4294967224. macoff=96
BUG: KASAN: use-after-free in prb_run_all_ft_ops net/packet/af_packet.c
On Fri, 2018-06-01 at 19:21 +0300, Leon Romanovsky wrote:
> On Fri, Jun 01, 2018 at 11:45:58AM -0400, David Miller wrote:
> > From: Doug Ledford
> > Date: Fri, 01 Jun 2018 11:08:24 -0400
> >
> > > On Thu, 2018-05-31 at 15:36 -0400, David Miller wrote:
> > > > From: Saeed Mahameed
> > > > Date: W
On Fri, 2018-06-01 at 11:45 -0400, David Miller wrote:
> From: Doug Ledford
> Date: Fri, 01 Jun 2018 11:08:24 -0400
>
> > On Thu, 2018-05-31 at 15:36 -0400, David Miller wrote:
> >> From: Saeed Mahameed
> >> Date: Wed, 30 May 2018 10:59:48 -0700
> >>
> >> > The following series is for mlx5-next
On Fri, Jun 01, 2018 at 08:50:16AM -0700, dsah...@kernel.org wrote:
> From: David Ahern
>
> A recent commit changed rtnl_talk_* to return the response message in
> allocated memory so callers need to free it. The change to name_is_vrf
> did not save the device index which is pointing to a struct
From: Leon Romanovsky
Date: Fri, 1 Jun 2018 19:21:26 +0300
> Of course, it is harmless for both of you to pull, but it looks like
> extra work which is not needed for you.
Just put net-next or rdma-next in the subject line(s) and that will
make it very clear what it expected to happen.
From: Steffen Klassert
Date: Thu, 31 May 2018 12:23:24 +0200
> 1) Avoid possible overflow of the offset variable
>in _decode_session6(), this fixes an infinite
>lookp there. From Eric Dumazet.
>
> 2) We may use an error pointer in the error path of
>xfrm_bundle_create(). Fix this by
From: Tonghao Zhang
Date: Thu, 31 May 2018 03:14:01 -0700
> Use the % instead of while, and it may simple code and improve
> the calculating. The real_num_tx_queues has been checked when
> allocating and setting it.
>
> Signed-off-by: Tonghao Zhang
The loop is there to avoid the expensive modu
From: Nicolas Dichtel
Date: Thu, 31 May 2018 10:59:31 +0200
> The first patch restores the possibility to bind an ip4 tunnel to an
> interface whith a large mtu.
> The second patch was spotted after the first fix. I also target it to net
> because it fixes the max mtu value that can be used for i
On Thu, 31 May 2018, Neil Horman wrote:
> On Thu, May 31, 2018 at 11:02:34AM -0700, Ronak Doshi wrote:
> >
> > On Wed, 30 May 2018, Paolo Abeni wrote:
> >
> > > Hi,
> > >
> > > On Thu, 2018-05-24 at 21:48 +, Guolin Yang wrote:
> > > > Yes, that code is not correct, we should fix that co
From: Petr Machata
Date: Thu, 31 May 2018 19:51:56 +0200
> This patchset adds more tests to the mirror-to-gretap suite where bridge
> is present in the underlay. Specifically it adds tests for bridge VLAN
> handling, FDB, and bridge port STP status.
>
> In patches #1-#3, the codebase is refactor
On 06/01/2018 12:56 AM, John Fastabend wrote:
> On 05/31/2018 06:00 PM, Eric Dumazet wrote:
>> On Thu, May 31, 2018 at 7:32 PM John Fastabend
>> wrote:
>>>
>>>
>>> Hi Wei,
>>>
>>> Thanks for the report and fix. It would be better to fix the
>>> root cause so that IPv6 works as intended.
>>>
>>> I
On Fri, Jun 01, 2018 at 01:08:18PM -0400, David Miller wrote:
> From: Leon Romanovsky
> Date: Fri, 1 Jun 2018 19:21:26 +0300
>
> > Of course, it is harmless for both of you to pull, but it looks like
> > extra work which is not needed for you.
>
> Just put net-next or rdma-next in the subject line
This fixes a crash where we assign tcp_prot to IPv6 sockets instead
of tcpv6_prot.
Previously we overwrote the sk->prot field with tcp_prot even in the
AF_INET6 case. This patch ensures the correct tcp_prot and tcpv6_prot
are used. Further, only allow ESTABLISHED connections to join the
map per no
On Tue, 29 May 2018 16:57:07 +0200
Patrick Talbert wrote:
> As mentioned in the ip-address man page, an address label must
> be equal to the device name or prefixed by the device name
> followed by a colon. Currently the only check on this input is
> to see if the device name appears at the begin
On 06/01/2018 03:46 PM, John Fastabend wrote:
> This fixes a crash where we assign tcp_prot to IPv6 sockets instead
> of tcpv6_prot.
...
> + /* ULPs are currently supported only for TCP sockets in ESTABLISHED
> + * state. Supporting sockets in LISTEN state will require us to
> +
On Wed, 30 May 2018 08:30:09 -0700
dsah...@kernel.org wrote:
> From: David Ahern
>
> RTA_CACHEINFO can be sent for non-cloned routes. If the attribute is
> present print it. Allows route dumps to print expires times for example
> which can exist on FIB entries.
>
> Signed-off-by: David Ahern
On Fri, 1 Jun 2018 17:02:18 +0200
Nicolas Dichtel wrote:
> Le 31/05/2018 à 17:51, Nicolas Dichtel a écrit :
> > Le 31/05/2018 à 17:46, Stephen Hemminger a écrit :
> >> On Thu, 31 May 2018 16:28:48 +0200
> > [snip]
> >> This makes sense. All of linkinfo that is present should be displayed.
>
On Wed, 30 May 2018 12:11:32 -0700
Nathan Harold wrote:
> Currently, calling 'ip xfrm monitor all' will
> actually invoke the 'all-nsid' command because the
> soft-match for 'all-nsid' occurs before the precise
> match for 'all'. This patch rearranges the checks
> so that the 'all' command, itsel
> diff --git a/drivers/infiniband/core/uverbs_std_types_counters.c
> b/drivers/infiniband/core/uverbs_std_types_counters.c
> new file mode 100644
> index ..a5bc50ceee13
> +++ b/drivers/infiniband/core/uverbs_std_types_counters.c
> @@ -0,0 +1,100 @@
> +/* SPDX-License-Identifier: ((GPL-
On Thu, May 31, 2018 at 04:43:27PM +0300, Leon Romanovsky wrote:
> From: Leon Romanovsky
>
> Changelog:
> v2->v3:
> * Change function mlx5_fc_query signature to hide the details of
>internal core driver struct mlx5_fc
> * Add commen to data[] field at struct mlx5_ib_flow_counters_data
> (m
Hi Ilias,
On 05/24/2018 01:56 AM, Ilias Apalodimas wrote:
> Hello,
>
> This is adding a new mode on the cpsw driver based around switchdev.
> In order to enable this you need to enable CONFIG_NET_SWITCHDEV,
> CONFIG_BRIDGE_VLAN_FILTERING, CONFIG_TI_CPSW_SWITCHDEV
> and add to udev config:
>
> SU
On 05/24/2018 09:56 PM, Ilias Apalodimas wrote:
> On Thu, May 24, 2018 at 06:39:04PM +0200, Andrew Lunn wrote:
>> On Thu, May 24, 2018 at 04:32:34PM +0300, Ilias Apalodimas wrote:
>>> On Thu, May 24, 2018 at 03:12:29PM +0200, Andrew Lunn wrote:
Device tree is supposed to describe the hardwa
On Thu, May 31, 2018 at 10:07 PM, kbuild test robot wrote:
> tree: https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git
> master
> head: 4b8e6ac41a594ea67ded6af6af5935f03221ea4c
> commit: ccf8dbcd062a930e64741c939ca784d15316aa0c [375/376] rtnetlink: Remove
> VLA usage
> config:
On Fri, Jun 1, 2018 at 1:26 AM, Eric Dumazet wrote:
> On Fri, Jun 1, 2018 at 4:18 AM Prashant Bhole
> wrote:
>>
>> In rtnl_newlink(), NULL check is performed on m_ops however member of
>> ops is accessed. Fixed by accessing member of m_ops instead of ops.
>>
>> [ 345.432629] BUG: KASAN: null-ptr
On Fri, Jun 01, 2018 at 11:10:01AM -0700, Ronak Doshi wrote:
>
>
> On Thu, 31 May 2018, Neil Horman wrote:
>
> > On Thu, May 31, 2018 at 11:02:34AM -0700, Ronak Doshi wrote:
> > >
> > > On Wed, 30 May 2018, Paolo Abeni wrote:
> > >
> > > > Hi,
> > > >
> > > > On Thu, 2018-05-24 at 21:48 +
On Thu, 2018-05-31 at 15:36 -0400, David Miller wrote:
> From: Saeed Mahameed
> Date: Wed, 30 May 2018 10:59:48 -0700
>
> > The following series is for mlx5-next tree [1], it adds the support
> > of two
> > new device events, from Ilan Tayari:
> >
> > 1. High temperature warnings.
> > 2. FPGA QP
On Fri, 2018-06-01 at 17:13 -0700, sae...@mellanox.com wrote:
> On Thu, 2018-05-31 at 15:36 -0400, David Miller wrote:
> > From: Saeed Mahameed
> > Date: Wed, 30 May 2018 10:59:48 -0700
> >
> > > The following series is for mlx5-next tree [1], it adds the
> > > support
> > > of two
> > > new devi
From: Eran Ben Elisha
Now, when all channels stats are saved regardless of the channel's state
{open, closed}, we can safely remove this indication and the stats spin
lock which protects it.
Fixes: 76c3810bade3 ("net/mlx5e: Avoid reset netdev stats on configuration
changes")
Signed-off-by: Eran
From: Florian Westphal
synchronize_rcu() is expensive.
The commit phase currently enforces an unconditional
synchronize_rcu() after incrementing the generation counter.
This is to make sure that a packet always sees a consistent chain, either
nft_do_chain is still using old generation (it will
From: Maor Gottlieb
Increase the aRFS flow table size to 64k so it could contain up to 64k
different streams.
Signed-off-by: Maor Gottlieb
Signed-off-by: Saeed Mahameed
---
drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/dr
Hi dave,
Sorry for the extra 2 patches in this series, but mostly the series
contains small patches and some fixes to previous patches in this
submission window, with one main patch from Tariq to improve legacy
RQ buffer management, for more information please refer to that tag
log below.
Please
From: Tariq Toukan
IPoIB WQE size is larger than a single WQEBB. Must not fetch the WQE,
and surely not memset it, until it is guaranteed that there are enough
WQEBBs available before getting to SQ/frag edge.
Fixes: 043dc78ecf07 ("net/mlx5e: TX, Use actual WQE size for SQ edge fill")
Signed-off
From: Tariq Toukan
Prefer the linear SKB configuration of Legacy RQ over the
non-linear one of Striding RQ.
This implies that ConnectX-4 LX now uses legacy RQ by default,
as it does not support the linear configuration of Striding RQ.
Signed-off-by: Tariq Toukan
Signed-off-by: Saeed Mahameed
From: Tariq Toukan
Avoid false sharing of cachelines by separating the cachelines of
TX stats that are dertied in xmit flow and in completion flow.
Signed-off-by: Tariq Toukan
Signed-off-by: Saeed Mahameed
---
drivers/net/ethernet/mellanox/mlx5/core/en_stats.c | 8
drivers/net/ethern
From: Tariq Toukan
Enhance the memory scheme of the legacy RQ, such that
only order-0 pages are used.
Whenever possible, prefer using a linear SKB, and build it
wrapping the WQE buffer.
Otherwise (for example, jumbo frames on x86), use non-linear SKB,
with as many frags as needed. In this case,
From: Tariq Toukan
Current LRO implementation in Legacy RQ uses high-order pages.
In downstream patches of this series we complete the transition
to using only order-0 pages in RX datapath (which was already done
in Striding RQ).
Unlike the more advanced Striding RQ, Legacy RQ does not make reus
From: Tariq Toukan
A call to mlx5e_tx_skb_pull_inline was mistakenly dropped
in the cited patch. Get it back.
Fixes: 043dc78ecf07 ("net/mlx5e: TX, Use actual WQE size for SQ edge fill")
Signed-off-by: Tariq Toukan
Signed-off-by: Saeed Mahameed
---
drivers/net/ethernet/mellanox/mlx5/core/en_tx
From: Eran Ben Elisha
The driver can present all SW stats even when the state not opened.
Fixed get strings, count and stats to support it.
In addition, fix tc2txq to hold a static mapping which doesn't depend on
the amount of open channels, and cannot have the same value on two
different cells
From: Adi Nissim
The representor MTU was hard coded to 1500 bytes.
Allow setting arbitrary MTU values up to the max supported by the FW.
Signed-off-by: Adi Nissim
Reviewed-by: Or Gerlitz
Signed-off-by: Saeed Mahameed
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 4
drivers/ne
1 - 100 of 117 matches
Mail list logo
