Hi Leandro,
Thank you for the patch! Yet something to improve:
[auto build test ERROR on net/master]
url:
https://github.com/0day-ci/linux/commits/Leandro-Dorileo/net-sched-taprio-fix-picos_per_byte-miscalculation/20190329-185809
config: m68k-allmodconfig (attached as .config)
compiler
From: Daniel Borkmann
Date: Fri, 29 Mar 2019 23:50:50 +0100
> The following pull-request contains BPF updates for your *net* tree.
>
> The main changes are:
>
> 1) Bug fix in BTF deduplication that was mishandling an equivalence
>comparison, from Andrii.
>
> 2) libbpf Makefile fixes to pro
On Fri, Mar 29, 2019 at 08:18:28PM -0700, Jakub Kicinski wrote:
> On Fri, 29 Mar 2019 17:16:05 -0700, Alexei Starovoitov wrote:
> > Realize two key ideas to speed up verification speed by ~20 times
> > 1. every 'branching' instructions records all verifier states.
> >not all of them are useful
On Fri, Mar 29, 2019 at 08:12:39PM -0700, Jakub Kicinski wrote:
> On Fri, 29 Mar 2019 17:16:07 -0700, Alexei Starovoitov wrote:
> > Branch instructions, branch targets and calls in a bpf program are
> > the places where the verifier remembers states that led to successful
> > verification of the pr
On Fri, 29 Mar 2019 17:16:05 -0700, Alexei Starovoitov wrote:
> Realize two key ideas to speed up verification speed by ~20 times
> 1. every 'branching' instructions records all verifier states.
>not all of them are useful for search pruning.
>add a simple heuristic to keep states that were
On Fri, 29 Mar 2019 17:16:10 -0700, Alexei Starovoitov wrote:
> The existing 16Mbyte verifier log limit is not enough for log_level=2
> even for small programs. Increase it to 1Gbyte.
> Note it's not a kernel memory limit.
> It's an amount of memory user space provides to store
> the verifier log.
On Fri, 29 Mar 2019 17:16:08 -0700, Alexei Starovoitov wrote:
> With large verifier speed improvement brought by the previous patch
> mark_reg_read() becomes the hottest function during verification.
> On a typical program it consumes 40% of cpu.
> mark_reg_read() walks parentage chain of registers
On Fri, 29 Mar 2019 17:16:07 -0700, Alexei Starovoitov wrote:
> Branch instructions, branch targets and calls in a bpf program are
> the places where the verifier remembers states that led to successful
> verification of the program.
> These states are used to prune brute force program analysis.
>
From: Dirk van der Merwe
The NSP now provides the ability to read from the SFF module EEPROM.
Note that even if an error occurs, the NSP may still provide some of the
data.
Signed-off-by: Dirk van der Merwe
Reviewed-by: Jakub Kicinski
---
.../ethernet/netronome/nfp/nfpcore/nfp_nsp.c | 62 +++
From: Pieter Jansen van Vuuren
With the introduction of flow_action_for_each pedit actions are no
longer grouped together, instead pedit actions are broken out per
32 byte word. This results in an inefficient use of the action list
that is pushed to hardware where each 32 byte word becomes its ow
On Fri, 29 Mar 2019 19:24:41 -0700, Jakub Kicinski wrote:
> From: Pieter Jansen van Vuuren
>
> With the introduction of flow_action_for_each pedit actions are no
> longer grouped together, instead pedit actions are broken out per
> 32 byte word. This results in an inefficient use of the action li
From: Dirk van der Merwe
Now that the NSP provides the ability to read from the SFF modules'
EEPROM, we can use this interface to implement the ethtool callback.
If the NSP only provides partial data, we log the event from within
the driver but pass a success code to ethtool to prevent it from
d
Hi!
The first patch in this series from Pieter improves the
handling of mangle actions in TC flower offload. These
used to be sent down to the driver in groups, but after
Pablo N's patches they are split out causing suboptimal
expression.
The ramaining two patches from Dirk add support for readi
Hi,
This patch clears up some confusion around the meaning of bit 12
for FW messages related to VLAN and flower offload.
Pieter says:
It fixes issues with matching, pushing and popping vlan tags.
We replace the vlan CFI bit with a vlan present bit that
indicates the presence of a vlan tag. We als
From: Pieter Jansen van Vuuren
We no longer set CFI when pushing vlan tags, therefore we remove
the CFI bit from push vlan.
Fixes: 1a1e586f54bf ("nfp: add basic action capabilities to flower offloads")
Signed-off-by: Pieter Jansen van Vuuren
Signed-off-by: Louis Peens
Reviewed-by: Jakub Kicins
From: Pieter Jansen van Vuuren
Replace vlan CFI bit with a vlan present bit that indicates the
presence of a vlan tag. Previously the driver incorrectly assumed
that an vlan id of 0 is not matchable, therefore we indicate vlan
presence with a vlan present bit.
Fixes: 5571e8c9f241 ("nfp: extend f
From: David Ahern
Update fib_table_lookup tracepoint to take a fib_nh_common struct and
dump the v6 gateway address if the nexthop uses it.
Signed-off-by: David Ahern
---
include/trace/events/fib.h | 45 ++---
net/ipv4/fib_trie.c| 2 +-
2 files
From: David Ahern
With the exception of the nexthop weight, the nexthop attributes used by
fib_nexthop_info and fib_add_nexthop come from the fib_nh_common struct.
Update both to use it and change fib_nexthop_info to check the family
as needed.
nexthop weight comes from the common struct for exi
From: David Ahern
Most of the ipv4 code only needs data from fib_nh_common. Add
fib_nh_common selection to fib_result and update users to use it.
Right now, fib_nh_common in fib_result will point to a nexthop within
a fib_info. Later, it can point to data within a nexthop struct.
Signed-off-by:
From: David Ahern
While the cached routes, nh_pcpu_rth_output and nh_rth_input, are IPv4
specific, a later patch wants to make them accessible for IPv6
gateways with IPv4 routes but also allow IPv6 routes to have their own
cached entries. Move the IPv4 cached routes from fib_nh to fib_nh_common
a
From: David Ahern
Flip IPv4 to use nhc_exceptions for fnhe_hash_bucket and drop
nh_exceptions from fib_nh.
Simplify the code a bit by converting update_or_create_fnhe,
find_exception and rt_cache_route to take fib_nh_common as the input
over fib_nh.
Signed-off-by: David Ahern
---
include/net/
From: David Ahern
Exceptions are really per device, so move rt6i_exception_bucket to
fib_nh_common as a generic nhc_exception_bucket. Move the flushed
flag to common as well. fib_nh_common for both is a strategic choice
to reduce memory consumption. Moving to fib6_nh pushes the struct over
256 wh
From: David Ahern
Similar to ipv6, move addition of nexthop attributes to dump
message into helpers that are called for both single path and
multipath routes. Align the new helpers to the IPv6 variant
which most notably means computing the flags argument based on
settings in nh_flags.
The RTA_FL
From: David Ahern
Second set of three with the end goal of enabling IPv6 gateways with IPv4
routes.
This set moves:
- the ipv4 tracepoint to take a fib_nh_common and updates it to handle
a v6 gateway.
- cached routes and exception buckets to fib_nh_common
- consolidates route notifications to
From: David Ahern
Export fib_nexthop_info and fib_add_nexthop for use by IPv6 code.
Remove rt6_nexthop_info and rt6_add_nexthop in favor of the IPv4
versions. Update fib_nexthop_info for IPv6 linkdown check and
RTA_GATEWAY for AF_INET6.
Signed-off-by: David Ahern
---
include/net/ip_fib.h |
From: Andi Kleen
ping_v4_net_ops references init functions, so needs to be __initdata.
ping_proc_exit is then referenced from __initdata, so also needs
to be __init.
Signed-off-by: Andi Kleen
---
net/ipv4/ping.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/ipv4/p
With large verifier speed improvement brought by the previous patch
mark_reg_read() becomes the hottest function during verification.
On a typical program it consumes 40% of cpu.
mark_reg_read() walks parentage chain of registers to mark parents as LIVE_READ.
Once the register is marked there is no
Allow bpf_prog_load_xattr() to specify log_level for program loading.
Teach libbpf to accept log_level with bit 2 set.
Increase default BPF_LOG_BUF_SIZE from 256k to 16M.
There is no downside to increase it to a maximum allowed by old kernels.
Existing 256k limit caused ENOSPC errors and users we
Add 3 basic tests that stress verifier scalability.
test_verif_scale1.c calls non-inlined jhash() function 90 times on
different position in the packet.
This test simulates network packet parsing.
jhash function is ~140 instructions and main program is ~1200 insns.
test_verif_scale2.c force inlin
The existing 16Mbyte verifier log limit is not enough for log_level=2
even for small programs. Increase it to 1Gbyte.
Note it's not a kernel memory limit.
It's an amount of memory user space provides to store
the verifier log. The kernel populates it 1k at a time.
Signed-off-by: Alexei Starovoitov
Large verifier speed improvements allow to increase
verifier complexity limit.
Now regardless of the program composition and its size it takes
little time for the verifier to hit insn_processed limit.
On typical x86 machine non-debug kernel processes 1M instructions
in 1/10 of a second.
(before the
In order to understand the verifier bottlenecks add various stats
and extend log_level:
log_level 1 and 2 are kept as-is:
bit 0 - level=1 - print every insn and verifier state at branch points
bit 1 - level=2 - print every insn and verifier state at every insn
bit 2 - level=4 - print verifier error
Branch instructions, branch targets and calls in a bpf program are
the places where the verifier remembers states that led to successful
verification of the program.
These states are used to prune brute force program analysis.
For unprivileged programs there is a limit of 64 states per such
'branch
Realize two key ideas to speed up verification speed by ~20 times
1. every 'branching' instructions records all verifier states.
not all of them are useful for search pruning.
add a simple heuristic to keep states that were successful in search pruning
and remove those that were not
2. mar
Hi David,
The following pull-request contains BPF updates for your *net* tree.
The main changes are:
1) Bug fix in BTF deduplication that was mishandling an equivalence
comparison, from Andrii.
2) libbpf Makefile fixes to properly link against libelf for the shared
object and to actually
> -Original Message-
> From: netdev-ow...@vger.kernel.org On
> Behalf Of Saeed Mahameed
> Sent: Friday, March 29, 2019 5:38 PM
> To: sa...@mellanox.com; Leon Romanovsky
> Cc: netdev@vger.kernel.org; linux-r...@vger.kernel.org; Huy Nguyen
> ; Vu Pham ; Saeed
> Mahameed
> Subject: [PATC
From: Aya Levin
Expose PRM layout for handling MPEIN (Management PCIE Info). It will be
used in the downstream patch for querying MPEIN via the driver.
Signed-off-by: Aya Levin
Signed-off-by: Saeed Mahameed
---
include/linux/mlx5/driver.h | 1 +
include/linux/mlx5/mlx5_ifc.h | 51 +
From: Tariq Toukan
Fix the following warning:
drivers/net/ethernet/mellanox/mlx5/core//fs_core.c:845:5:
warning: 'err' may be used uninitialized in this function
[-Wmaybe-uninitialized]
No real issue here. This is only a false compiler warning.
The 'err' variable is guaranteed to be init by time
From: Aya Levin
Add rate limited print macros for warning and info level. This protects
the system from burst of prints depleting HW resources and spamming dmesg.
Signed-off-by: Aya Levin
Signed-off-by: Saeed Mahameed
---
drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h | 10 ++
1
Function setup and teardown procedures are the basic procedure that
each mlx5 pci function should perform to boot up a mlx5 device function
and initialize basic communication with FW, before allocating any higher
level software/firmware resources.
This provides a better logical separation of mlx5
From: Huy Nguyen
Detach mlx5_core mdev messages from pci device mdev->pdev messages and
provide a better report/debug of different mlx5 device types.
This patch does not change any functionality.
Signed-off-by: Huy Nguyen
Signed-off-by: Vu Pham
Signed-off-by: Saeed Mahameed
---
.../ethernet
From: Huy Nguyen
Add bar_addr field to store bar-0 address to avoid calling
pci_resource_start with hard-coded bar-0 as parameter.
Also note that different mlx5 device types will have bar_addr
on different bars.
This patch does not change any functionality.
Signed-off-by: Huy Nguyen
Signed-off
From: Huy Nguyen
Replace pci dev_err/warn/info messages with mlx5_core_err/warn/info
messages to provide a better report/debug of different mlx5 device types.
This patch does not change any functionality.
Signed-off-by: Huy Nguyen
Signed-off-by: Vu Pham
Signed-off-by: Saeed Mahameed
---
dri
Software structure initialization should be in mdev_init stage.
This provides a better logical separation of mlx5 core device
initialization flow and will help to seamlessly support creating different
mlx5 device types such as PF, VF and SF mlx5 sub-function virtual device.
This patch does not ch
From: Huy Nguyen
Use mlx5_core mdev private name in message instead of using pci dev_name
to provide a better report/debug of different mlx5 device types.
This patch does not change any functionality.
Signed-off-by: Huy Nguyen
Signed-off-by: Vu Pham
Signed-off-by: Saeed Mahameed
---
drivers
This patch does not change any functionality.
Signed-off-by: Vu Pham
Signed-off-by: Saeed Mahameed
---
.../net/ethernet/mellanox/mlx5/core/main.c| 49 +--
1 file changed, 22 insertions(+), 27 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c
b/driver
Using foundation from previous patches to factor mlx5_load_one flow
into three stages:
1. mlx5_function_setup() from previous patch to setup function
2. mlx5_init_once() from previous patch to init software objects
according to hw caps
3. New mlx5_load() to load mlx5 components
This provides a bet
From: Maxim Mikityanskiy
As there is no user of mlx5_write64 that passes a spinlock to
mlx5_write64, remove this functionality and simplify the function.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: Eran Ben Elisha
Signed-off-by: Saeed Mahameed
---
drivers/infiniband/hw/mlx5/qp.c
Separate resources initialization from pci initialization.
This provides a better logical separation of mlx5 core device
initialization flow and will help to seamlessly support creating different
mlx5 device types such as PF, VF and SF mlx5 sub-function virtual device.
This patch does not change
From: Maxim Mikityanskiy
MLX5_*_DOORBELL_LOCK macros provided a way to avoid locking for
mlx5_write64 on 64-bit platforms where it's not necessary. Currently all
calls to mlx5_write64 don't use a spinlock, so the macros became unused.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: Eran Ben Elis
Hi,
This series includes some misc shared code updates for mlx5-next branch.
>From Maxim, Remove un-used macros and spinlock from mlx5 code.
>From Aya, Expose Management PCIE info register layout and add rate limit
print macros.
>From Tariq, Compilation warning fix in fs_core.c
>From Vu, Huy a
On Fri, Mar 29, 2019 at 2:21 PM Jiri Pirko wrote:
> Fri, Mar 29, 2019 at 07:59:26PM CET, jakub.kicin...@netronome.com wrote:
> >On Fri, 29 Mar 2019 07:49:05 +0100, Jiri Pirko wrote:
> >> Thu, Mar 28, 2019 at 10:40:02PM CET, jakub.kicin...@netronome.com wrote:
> >> >On Thu, 28 Mar 2019 22:12:42 +01
From: Saeed Mahameed
Date: Fri, 29 Mar 2019 13:42:08 -0700
> This series introduces some fixes to mlx5 driver.
>
> Please pull and let me know if there is any problem.
Pulled.
> For -stable v4.11
> ('net/mlx5: Decrease default mr cache size')
>
> For -stable v4.12
> ('net/mlx5e: Add a lock on
On 3/29/19 2:57 PM, Jakub Kicinski wrote:
> On Fri, 29 Mar 2019 14:29:11 -0700, Florian Fainelli wrote:
>>> Out of curiosity, what are the shared features? It seems mlx5 drives
>>> a lot of our API design, it'd be good if the community had a better
>>> understanding of it.
>>>
>>> The situation w
On 3/28/19 2:12 PM, Jiri Pirko wrote:
> From: Jiri Pirko
>
> Pass the switch ID down the to devlink through devlink_port_attrs_set()
> so it can be used by devlink_compat_switch_id_get(). Leave
> ndo_get_port_parent_id implementation only for legacy.
Nit you are passing the switch fabric id (dst
On 29 Mar 2019, at 14:24, Jakub Kicinski wrote:
> On Fri, 29 Mar 2019 16:14:57 +, Jonathan Lemon wrote:
>> Background:
>>
>> With AF_XDP, packets are placed on queues which are serviced directly
>> from
>> an application running in user space, which promises greatly
>> increased
>> perform
On Fri, 29 Mar 2019 14:29:11 -0700, Florian Fainelli wrote:
> > Out of curiosity, what are the shared features? It seems mlx5 drives
> > a lot of our API design, it'd be good if the community had a better
> > understanding of it.
> >
> > The situation with pipelined devices is somewhat murky. D
On 3/29/19 11:59 AM, Jakub Kicinski wrote:
> On Fri, 29 Mar 2019 07:49:05 +0100, Jiri Pirko wrote:
>> Thu, Mar 28, 2019 at 10:40:02PM CET, jakub.kicin...@netronome.com wrote:
>>> On Thu, 28 Mar 2019 22:12:42 +0100, Jiri Pirko wrote:
From: Jiri Pirko
To provide visibility of the po
On 3/29/19 12:17 PM, Heiner Kallweit wrote:
> Add description of new PHY tunable Fast Link Down to ethtool man page.
>
> v2:
> - improve wording
>
> Signed-off-by: Heiner Kallweit
Reviewed-by: Florian Fainelli
--
Florian
On 3/29/19 12:15 PM, Heiner Kallweit wrote:
> This patch adds support for PHY tunable Fast Link Down. Like downshift
> it uses an u8 parameter.
>
> Signed-off-by: Heiner Kallweit
Reviewed-by: Florian Fainelli
--
Florian
On Fri, 29 Mar 2019 16:14:57 +, Jonathan Lemon wrote:
> Background:
>
> With AF_XDP, packets are placed on queues which are serviced directly from
> an application running in user space, which promises greatly increased
> performance. However, not all incoming packets should be delivered to t
On 3/29/19 12:15 PM, Heiner Kallweit wrote:
> In preparation of adding support for Fast Link Down as PHY tunable
> let's simplify the handling of PHY tunable downshift a little.
>
> Signed-off-by: Heiner Kallweit
Reviewed-by: Florian Fainelli
--
Florian
On 3/29/19 12:14 PM, Heiner Kallweit wrote:
> Sync ethtool-copy.h with linux-next from 03/28/2019. This provides
> access to the new PHY tunable for Fast Link Down support.
>
> Signed-off-by: Heiner Kallweit
Reviewed-by: Florian Fainelli
--
Florian
Fri, Mar 29, 2019 at 07:59:26PM CET, jakub.kicin...@netronome.com wrote:
>On Fri, 29 Mar 2019 07:49:05 +0100, Jiri Pirko wrote:
>> Thu, Mar 28, 2019 at 10:40:02PM CET, jakub.kicin...@netronome.com wrote:
>> >On Thu, 28 Mar 2019 22:12:42 +0100, Jiri Pirko wrote:
>> >> From: Jiri Pirko
>> >>
>> >
On 3/29/19 1:09 PM, Heiner Kallweit wrote:
> Add suspend / resume callbacks for AQR107 family. Suspend powers down
> the complete chip except MDIO and internal CPU.
>
> Signed-off-by: Heiner Kallweit
Reviewed-by: Florian Fainelli
--
Florian
On Thu, 28 Mar 2019 02:35:38 -0700
Eric Dumazet wrote:
> > obj-$(CONFIG_NET_SCH_ETF) += sch_etf.o
> > obj-$(CONFIG_NET_SCH_TAPRIO) += sch_taprio.o
> > +obj-$(CONFIG_NET_SCH_DUALPI2) += sch_dualpi2.o
> >
> > obj-$(CONFIG_NET_CLS_U32) += cls_u32.o
> > obj-$(CONFIG_NET_CLS_ROUTE4)
Hi Leandro,
Thank you for the patch! Yet something to improve:
[auto build test ERROR on net/master]
url:
https://github.com/0day-ci/linux/commits/Leandro-Dorileo/net-sched-taprio-fix-picos_per_byte-miscalculation/20190329-185809
config: i386-randconfig-l0-03292131 (attached as .config
Hi Cong,
OK, so I've renamed conndscp to conntrack and hopefully this are
flexible enough for future conntrack->skb operations to be added in the
future. How does this one fly?
Cheers,
Kevin
Kevin Darbyshire-Bryant (1):
net: sched: Introduce conntrack action
include/net/tc_act/tc_conntrack
conntrack is a new tc filter action module. It is designed to restore
DSCPs stored in conntrack marks
The feature is intended for use and has been found useful for restoring
ingress classifications based on egress classifications across links
that bleach or otherwise change DSCP, typically home I
From: Gavi Teitz
Previously, a false positive would be caught if the TIRs list is
empty, since the err value was initialized to -ENOMEM, and was only
updated if a TIR is refreshed. This is resolved by initializing the
err value to zero.
Fixes: b676f653896a ("net/mlx5e: Refactor refresh TIRs")
Si
From: Omri Kahalon
Traditionally, the PF (Physical Function) which resides on vport 0 was
the E-switch manager. Since the ECPF (Embedded CPU Physical Function),
which resides on vport 0xfffe, was introduced as the E-Switch manager,
the assumption that the E-switch manager is on vport 0 is incorre
From: Artemy Kovalyov
Delete initialization of high order entries in mr cache to decrease initial
memory footprint. When required, the administrator can populate the
entries with memory keys via the /sys interface.
This approach is very helpful to significantly reduce the per HW function
memory
From: Huy Nguyen
Make sure the struct mlx5_flow_destination is zero before
filling in the field.
Fixes: 8da202b24913 ("net/mlx5: E-Switch, Add support for VEPA in legacy mode.")
Signed-off-by: Huy Nguyen
Reviewed-by: Daniel Jurgens
Signed-off-by: Saeed Mahameed
---
drivers/net/ethernet/mella
From: Huy Nguyen
Set xon = xoff - netdev's max_mtu.
netdev's max_mtu will give enough time for the pause frame to
arrive at the sender.
Fixes: 0696d60853d5 ("net/mlx5e: Receive buffer configuration")
Signed-off-by: Huy Nguyen
Signed-off-by: Saeed Mahameed
---
.../mellanox/mlx5/core/en/port_bu
From: Aditya Pakki
idr_find() can return a NULL value to 'flow' which is used without a
check. The patch adds a check to avoid potential NULL pointer dereference.
In case of mlx5_fpga_sbu_conn_sendmsg() failure, free buf allocated
using kzalloc.
Fixes: ab412e1dd7db ("net/mlx5: Accel, add TLS rx
From: Roi Dayan
The esw fdb table has a union of legacy and offloads members.
So if we were in a certain esw mode we could set some memebers and not
set null which is fine as on destroy path and don't care.
But then moving from legacy to switchdev a second time, the cleanup flow
of legacy mode ch
From: Aya Levin
Ethtool option set_link_ksettings allows setting of legacy link-modes
or extended link-modes. Refine the decision of which type of link-modes
is set.
Fixes: 6a897372417e ("net/mlx5: ethtool, Add ethtool support for 50Gbps per
lane link modes")
Signed-off-by: Aya Levin
Signed-of
From: Eli Britstein
The driver allocates an encap context based on the tunnel properties,
and reuse that context for all flows using the same tunnel properties.
Commit df2ef3bff193 ("net/mlx5e: Add GRE protocol offloading")
introduced another tunnel protocol other than the single VXLAN
previously
From: Roi Dayan
The esw offloads structures share a union with the legacy mode structs.
Reset the offloads struct to zero in init to protect from null
assumptions made by the legacy mode code.
Signed-off-by: Roi Dayan
Reviewed-by: Or Gerlitz
Signed-off-by: Saeed Mahameed
---
drivers/net/ethe
From: Huy Nguyen
Set minimum speed in xoff threshold formula to 40Gbps
Fixes: 0696d60853d5 ("net/mlx5e: Receive buffer configuration")
Signed-off-by: Huy Nguyen
Signed-off-by: Saeed Mahameed
---
.../net/ethernet/mellanox/mlx5/core/en/port_buffer.c | 11 ++-
1 file changed, 6 insertio
From: Aya Levin
Allow configuration of legacy link-modes even when extended link-modes
are supported. This requires reading of legacy advertisement even when
extended link-modes are supported. Since legacy and extended
advertisement are mutually excluded, wait for empty reply from extended
advert
From: Tonghao Zhang
The capacity of FDB offloading and NIC offloading table are
different, and when allocating the pedit actions, we should
use the correct namespace type.
Fixes: c500c86b0c75d ("net/mlx5e: support for two independent packet edit
actions")
Cc: Pablo Neira Ayuso
Signed-off-by: T
From: Dmytro Linkin
For some protocols we are not allowing IP header rewrite offload, since
the HW is not capable to properly adjust the l4 checksum. However, TTL
& HOPLIMIT modification can be done for all IP protocols, because they
are not part of the pseudo header taken into account for checks
From: Yuval Avnery
Refresh tirs is looping over a global list of tirs while netdevs are
adding and removing tirs from that list. That is why a lock is
required.
Fixes: 724b2aa15126 ("net/mlx5e: TIRs management refactoring")
Signed-off-by: Yuval Avnery
Signed-off-by: Saeed Mahameed
---
drivers
Hi Dave,
This series introduces some fixes to mlx5 driver.
Please pull and let me know if there is any problem.
For -stable v4.11
('net/mlx5: Decrease default mr cache size')
For -stable v4.12
('net/mlx5e: Add a lock on tir list')
For -stable v4.13
('net/mlx5e: Fix error handling when refreshi
after previous changes, xfrm_mode contains no function pointers anymore
and all modules defining such struct contain no code except an init/exit
functions to register the xfrm_mode struct with the xfrm core.
Just place the xfrm modes core and remove the modules,
the run-time xfrm_mode register/unr
Now that we have the family available directly in the
xfrm_mode struct, we can use that and avoid one extra dereference.
Signed-off-by: Florian Westphal
---
net/ipv4/ip_vti.c | 2 +-
net/ipv6/ip6_vti.c| 2 +-
net/xfrm/xfrm_input.c | 4 ++--
net/xfrm/xfrm_interface.c | 2 +-
4
There are only two versions (tunnel and transport). The ip/ipv6 versions
are only differ in sizeof(iphdr) vs ipv6hdr.
Place this in the core and use x->outer_mode->encap type to call the
correct adjustment helper.
Before:
text databss dec filename
15730311 6937008 4046908 26714
Same is input indirection. Only exception: we need to export
xfrm_outer_mode_output for pktgen.
Increases size of vmlinux by about 163 byte:
Before:
textdata bss dec filename
15730208 6936948 4046908 26714064 vmlinux
After:
15730311 6937008 4046908 26714227 vmlinux
xfr
This will be useful to know if we're supposed to decode ipv4 or ipv6.
While at it, make the unregister function return void, all module_exit
functions did just BUG(); there is never a point in doing error checks
if there is no way to handle such error.
Signed-off-by: Florian Westphal
---
includ
No need for any indirection or abstraction here, both functions
are pretty much the same and quite small, they also have no external
dependencies.
xfrm_prepare_input can then be made static.
With allmodconfig build, size increase of vmlinux is 25 byte:
Before:
text data bss dec
This series removes all function pointers from the xfrm_mode struct.
They are either replaced by direct calls or a static helper.
This reduces code size: even merging all xfrm{4,6}_,mode* modules
into the core reduces total size, as the infrastructure to maintain
dynamic (un)register can be remove
This structure is now only 4 bytes, so its more efficient
to cache a copy rather than its address.
No significant size difference in allmodconfig vmlinux.
With non-modular kernel that has all XFRM options enabled, this
series reduces vmlinux image size by ~11kb. All xfrm_mode
indirections are gon
Adds an EXPORT_SYMBOL for afinfo_get_rcu, as it will now be called from
ipv6 in case of CONFIG_IPV6=m.
This change has virtually no effect on vmlinux size, but it reduces
afinfo size and allows followup patch to make xfrm modes const.
v2: mark if (afinfo) tests as likely (Sabrina)
re-fetch af
similar to previous patch: no external module dependencies,
so we can avoid the indirection by placing this in the core.
This change removes the last indirection from xfrm_mode and the
xfrm4|6_mode_{beet,tunnel}.c modules contain (almost) no code anymore.
Before:
textdata bss dec
No external dependencies on any module, place this in the core.
Increase is about 1800 byte for xfrm_input.o.
The beet helpers get added to internal header, as they can be reused
from xfrm_output.c in the next patch (kernel contains several
copies of them in the xfrm{4,6}_mode_beet.c files).
Befo
These functions are small and we only have versions for tunnel
and transport mode for ipv4 and ipv6 respectively.
Just place the 'transport or tunnel' conditional in the protocol
specific function instead of using an indirection.
Before:
3226 12 0 3238 net/ipv4/esp4_offload.
Add suspend / resume callbacks for AQR107 family. Suspend powers down
the complete chip except MDIO and internal CPU.
Signed-off-by: Heiner Kallweit
---
drivers/net/phy/aquantia_main.c | 16
1 file changed, 16 insertions(+)
diff --git a/drivers/net/phy/aquantia_main.c b/drivers
Hi Olga,
Thank you for the patch! Yet something to improve:
[auto build test ERROR on net-next/master]
url:
https://github.com/0day-ci/linux/commits/Olga-Albisser/sched-add-dualpi2-scheduler-module/20190329-235019
config: arm-allmodconfig (attached as .config)
compiler: arm-linux-gnueabi
On Fri, Mar 29, 2019 at 3:46 PM Eric Dumazet wrote:
>
> TCP stack relies on the fact that a freshly allocated skb
> has skb->cb[] and skb_shinfo(skb)->tx_flags cleared.
>
> When recycling tx skb, we must ensure these fields are cleared.
>
> Fixes: 472c2e07eef0 ("tcp: add one skb cache for tx")
> S
1 - 100 of 132 matches
Mail list logo
