Function mlx4_conv_rss_types() contains constant arrays variables
which are recreated with every call to the function. By changing the
arrays definitions from "const" to "static const" these recreations
can be saved.
Signed-off-by: Ophir Munk
---
v1:
Initial release
v2:
Upd
check")
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_ethdev.c | 12 ++-
drivers/net/mlx4/mlx4_flow.c | 45 ++
drivers/net/mlx4/mlx4_flow.h | 35
3 files changed, 87 insertions(+), 5 deletions(-)
diff --git
Shahaf asked for it in his review
> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Sunday, May 13, 2018 7:13 PM
> To: Ophir Munk
> Cc: dev@dpdk.org; Adrien Mazarguil ; Olga
> Shern ; Shahaf Shuler
> Subject: Re: [PATCH v4 2/2
Anyway will update in v5 to be sent now
> -Original Message-
> From: Ophir Munk
> Sent: Sunday, May 13, 2018 7:14 PM
> To: Thomas Monjalon
> Cc: dev@dpdk.org; Adrien Mazarguil ; Olga
> Shern ; Shahaf Shuler
> Subject: RE: [PATCH v4 2/2] net/mlx4: advertise supported
Function mlx4_conv_rss_types() contains constant arrays variables
which are recreated with every call to the function. By changing the
arrays definitions from "const" to "static const" these recreations
can be saved.
Signed-off-by: Ophir Munk
---
v1:
Initial release
v2:
Upd
check")
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_ethdev.c | 2 ++
drivers/net/mlx4/mlx4_flow.c | 79 ++
drivers/net/mlx4/mlx4_flow.h | 1 +
3 files changed, 82 insertions(+)
diff --git a/drivers/net/mlx4/mlx4_ethdev.c b/drivers/net/mlx4/mlx
Function mlx4_conv_rss_types() contains constant arrays variables
which are recreated with every call to the function. By changing the
arrays definitions from "const" to "static const" these recreations
can be saved.
Signed-off-by: Ophir Munk
---
v1:
Initial release
v2:
Upd
check")
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_ethdev.c | 2 ++
drivers/net/mlx4/mlx4_flow.c | 80 ++
drivers/net/mlx4/mlx4_flow.h | 1 +
3 files changed, 83 insertions(+)
diff --git a/drivers/net/mlx4/mlx4_ethdev.c b/drivers/net/mlx4/mlx
a
change from "0" to "1" (or vice versa).
Fixes: f503d2694825 ("net/tap: support flow API isolated mode")
Cc: sta...@dpdk.org
Reviewed-by: Raslan Darawsheh
Signed-off-by: Ophir Munk
---
v1:
initial release
v2:
1. Updates based on Keith Wiles review
2. Do not empt
v2 has a typo (/raw/row/)
Re-sending as v3.
> -Original Message-
> From: Wiles, Keith [mailto:keith.wi...@intel.com]
> Sent: Tuesday, May 15, 2018 1:18 AM
> To: Ophir Munk
> Cc: DPDK ; Pascal Mazon ;
> Thomas Monjalon ; Olga Shern
> ; Shahaf Shuler ;
> sta..
Hi Keith,
> -Original Message-
> From: Wiles, Keith [mailto:keith.wi...@intel.com]
> Sent: Monday, May 14, 2018 3:33 PM
> To: Ophir Munk
> Cc: dev@dpdk.org; Pascal Mazon ; Thomas
> Monjalon ; Olga Shern ;
> sta...@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v1] net
a
change from "0" to "1" (or vice versa).
Fixes: f503d2694825 ("net/tap: support flow API isolated mode")
Cc: sta...@dpdk.org
Reviewed-by: Raslan Darawsheh
Signed-off-by: Ophir Munk
---
v1:
initial release
v2:
Please ignore (due to typo errors)
v3:
1. Updates based
rg; keith.wi...@intel.com; ferruh.yi...@intel.com; Ophir
> Munk
> Cc: Vipin Varghese
> Subject: [PATCH v2] net/tap: perform proto field update for tun only
>
> The TX function is shared between TAP and TUN PMD. Checking TUN-TAP
> type field will ensure the TAP PMD will a
number of queues -
an extra file descriptor is opened on TAP creation and is closed on
TAP closure. It's only purpose is to serve as a keep-alive mechanism
for the TAP device.
Signed-off-by: Ophir Munk
---
drivers/net/tap/rte_eth_tap.c | 31 ---
drivers/ne
number of queues -
an extra file descriptor is opened on TAP creation and is closed on
TAP closure. It's only purpose is to serve as a keep-alive mechanism
for the TAP device.
Fixes: bf7b7f437b49 ("net/tap: create netdevice during probing")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
Add an RSS hash update callback to eth_dev_ops.
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/Makefile | 1 +
drivers/net/mlx4/mlx4.c | 1 +
drivers/net/mlx4/mlx4.h | 7 ++
drivers/net/mlx4/mlx4_rss.c | 59 +
4 files changed, 68
; To: Shahaf Shuler
> Cc: dev@dpdk.org; Ophir Munk
> Subject: [PATCH 2/2] net/mlx4: refactor RSS conversion functions
>
> Since commit 97b2217ae5bc ("net/mlx4: advertise supported RSS hash
> functions"), this PMD includes two similar-looking functions that convert RSS
number of queues -
an extra file descriptor is opened on TAP creation and is closed on
TAP closure. Its only purpose is to serve as a keep-alive mechanism
for the TAP device.
Fixes: bf7b7f437b49 ("net/tap: create netdevice during probing")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
v
Hi Keith,
Please find comments inline
> -Original Message-
> From: Wiles, Keith [mailto:keith.wi...@intel.com]
> Sent: Thursday, May 17, 2018 4:00 PM
> To: Ophir Munk
> Cc: dev@dpdk.org; Pascal Mazon ; Thomas
> Monjalon ; Olga Shern ;
> Shahaf Shuler ; sta...@d
number of queues -
an extra file descriptor is opened on TAP creation and is closed on
TAP closure. Its only purpose is to serve as a keep-alive mechanism
for the TAP device.
Fixes: bf7b7f437b49 ("net/tap: create netdevice during probing")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
v
Hi,
Please find comments inline.
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Matan Azrad
> Sent: Thursday, May 17, 2018 9:52 PM
> To: Gaetan Rivet
> Cc: dev@dpdk.org; sta...@dpdk.org
> Subject: [dpdk-dev] [PATCH 1/2] net/failsafe: fix removed sub-device cle
For 18.05 this path can be withdrawn.
It has already been suppressed a few days ago.
Thanks,
Ophir
> -Original Message-
> From: Adrien Mazarguil [mailto:adrien.mazarg...@6wind.com]
> Sent: Monday, May 21, 2018 7:08 PM
> To: Ophir Munk
> Cc: dev@dpdk.org; Thomas Monjal
Hi,
Overall it looks good.
Please note a few more comments below.
> -Original Message-
> From: Vipin Varghese [mailto:vipin.vargh...@intel.com]
> Sent: Tuesday, May 22, 2018 1:47 PM
> To: dev@dpdk.org; ferruh.yi...@intel.com; keith.wi...@intel.com; Ophir
> Munk
> C
ioctl call.
The fix is to detach the keep-alive queue after its creation through
TUNSETQUEUE ioctl call.
Fixes: 3101191c63ab ("net/tap: fix device removal when no queue exist")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
drivers/net/tap/rte_eth_
d) ((__u32)(((a) & 0xff) << 24) | \
---
Regards,
Ophir
From: Sameeh Jubran [mailto:sam...@daynix.com]
Sent: Friday, May 25, 2018 11:54 AM
To: dev@dpdk.org; Ophir Munk
Cc: Yan Vugenfirer
Subject: Re: C
73a3 ("net/tap: add eBPF bytes code")
Signed-off-by: Ophir Munk
---
doc/guides/nics/tap.rst | 21 +
drivers/net/tap/tap_bpf_program.c | 5 +
2 files changed, 22 insertions(+), 4 deletions(-)
diff --git a/doc/guides/nics/tap.rst b/doc/guides/nics/t
segment large TCP
packets.
This commits creates of a small private mbuf pool in TAP PMD required by
librte_gso.
The number of buffers will be 64 - each of 128 bytes length.
Ophir Munk (2):
net/tap: calculate checksums of multi segs packets
net/tap: support TSO (TCP Segment Offload)
drivers/net
header size equals
TCP header size).
Signed-off-by: Ophir Munk
---
drivers/net/tap/rte_eth_tap.c | 54 ---
1 file changed, 40 insertions(+), 14 deletions(-)
diff --git a/drivers/net/tap/rte_eth_tap.c b/drivers/net/tap/rte_eth_tap.c
index 61d6465..df23c4d 100644
another PCI device which
supports TSO capability in HW.
For more details on librte_gso implementation please refer to dpdk
documentation.
The number of newly generated TCP TSO segments is limited to 64.
Signed-off-by: Ophir Munk
---
drivers/net/tap/Makefile | 2 +-
drivers/net/tap
Patch sent for this release.
> -Original Message-
> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
> Sent: Monday, April 09, 2018 7:39 PM
> To: Ophir Munk ; dev@dpdk.org; Pascal Mazon
>
> Cc: Thomas Monjalon ; Olga Shern
>
> Subject: Re: [dpdk-dev] [RFC 0/2
Add DEV_TX_OFFLOAD_TCP_TSO to failsafe Tx offload default capabilities.
The net result of failsafe Tx capabilities is the logical AND of Tx
capabilities among all failsafe sub_devices and failsafe own default
capabilities.
Signed-off-by: Ophir Munk
---
drivers/net/failsafe/failsafe_ops.c | 3
s commits fixes this issue by creating tap and failsafe devices
with short names such as "tap_net_vsc0" or "net_failsafe_vsc0".
Fixes: e7dc5d7becc5 ("net/vdev_netvsc: implement core functionality")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
drivers/net/vdev_ne
> -Original Message-
> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
> Sent: Tuesday, April 10, 2018 5:31 PM
> To: Ophir Munk ; dev@dpdk.org; Pascal Mazon
>
> Cc: Thomas Monjalon ; Olga Shern
>
> Subject: Re: [dpdk-dev] [RFC 0/2] TAP TSO Implementation
&
".
This commits creates tap and failsafe devices with short names such as
"net_tap_netvsc0" or "net_failsafe_netvsc0".
> -Original Message-
> From: Matan Azrad
> Sent: Tuesday, April 10, 2018 11:04 AM
> To: Ophir Munk ; dev@dpdk.org
> Cc: Thomas Monjalon
Hi Vipin,
This patch (adding TUN to TAP) has been Acked and accepted in next-net branch.
I have some questions regarding the implementation (please find below).
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Vipin Varghese
> Sent: Tuesday, April 03, 2018 12:38
rules.
TAP eBPF requires Linux version 4.9 configured with BPF. TAP PMD will
successfully compile on systems with old or non-BPF configured kernels but
RSS rules creation on TAP devices will not be supported.
Signed-off-by: Ophir Munk
---
drivers/net/tap/Makefile | 11 +
drivers/net
the packet with the RSS queue according the resulting hash,
then reclassify the packet.
Ophir Munk (2):
net/tap: add eBPF instructions to TAP device
net/tap: implement RSS with eBPF classifier and action
drivers/net/tap/Makefile | 31 +
drivers/net/tap/rte_eth_tap.h |
ch new RSS rule. A BPF map entry contains the
RSS queues of that rule.
Add BPF action for each new RSS rule in which the Toeplitz RSS hash is
calculated based on L3 addresses and L4 ports. Mark the packet with the RSS
queue according the resulting RSS hash, then reclassify the packet.
Signed-off
Hi Pascal,
Please see inline
> -Original Message-
> From: Pascal Mazon [mailto:pascal.ma...@6wind.com]
> Sent: Wednesday, January 10, 2018 11:35 AM
> To: Ophir Munk ; dev@dpdk.org
> Cc: Thomas Monjalon ; Olga Shern
>
> Subject: Re: [PATCH v3 1/2] net/tap: add eBP
Hi Jason,
Thanks for your inputs.
Please see inline
> -Original Message-
> From: Jason Wang [mailto:jasow...@redhat.com]
> Sent: Wednesday, January 10, 2018 12:11 PM
> To: Ophir Munk ; dev@dpdk.org; Pascal Mazon
>
> Cc: Thomas Monjalon ; Olga Shern
>
> Subject:
Hi Pascal,
Please see inline
> -Original Message-
> From: Pascal Mazon [mailto:pascal.ma...@6wind.com]
> Sent: Wednesday, January 10, 2018 12:20 PM
> To: Ophir Munk ; dev@dpdk.org
> Cc: Thomas Monjalon ; Olga Shern
>
> Subject: Re: [PATCH v3 2/2] net/tap: im
. Each C array (corresponding to BPF
bytes code) is downloaded to the kernel using BPF systm call.
The C program and the external tool source files will be added in
separate commits.
In this commit the bytes code is not part of DPDK compilation
Signed-off-by: Ophir Munk
---
drivers/net/tap
inux headers are searched for.
If they are not present missing definitions are locally added
(tap_autoconf.h).
If the kernel cannot support a BPF operation - at runtime it will
gracefully reject the netlink message (with BPF) sent to it.
Ophir Munk (5):
net/tap: support actions for diffe
Add a generic TC actions handling for TC actions: "mirred",
"gact", "skbedit". This will be useful when introducing
BPF actions, as it uses TCA_BPF_ACT instead of TCA_FLOWER_ACT
Signed-off-by: Ophir Munk
---
drivers/net/tap/Makefile | 8 ++
drivers/net/tap/r
ter an RSS hash was calculated on the packet
and saved in skb->cb[1]
Section l3_l4 - BPF action to calculate RSS hash on packet layers 3
and 4
This file is not part of DPDK tree compilation.
Signed-off-by: Ophir Munk
---
drivers/net/tap/tap_bpf_program.c | 221 +
nloaded to the kernel for each new RSS rule.
TAP eBPF requires Linux version 4.9 configured with BPF. TAP PMD will
successfully compile on systems with old or non-BPF configured kernels but
RSS rules creation on TAP devices will not be successful
Signed-off-by: Ophir Munk
---
drivers/net/ta
() - create BPF RSS map for storing RSS
parameters per RSS rule
tap_flow_bpf_update_rss_elem() - update BPF map entry with RSS rule
parameters
Signed-off-by: Ophir Munk
---
drivers/net/tap/Makefile | 6 ++
drivers/net/tap/tap_bpf.h | 112 +
drivers/net/tap
Add a generic TC actions handling for TC actions: "mirred",
"gact", "skbedit". This will be useful when introducing
BPF actions, as it uses TCA_BPF_ACT instead of TCA_FLOWER_ACT
Signed-off-by: Ophir Munk
---
drivers/net/tap/Makefile | 8 ++
drivers/net/tap/r
ject the netlink message (with BPF) sent to it.
Ophir Munk (6):
net/tap: support actions for different classifiers
net/tap: add eBPF bytes code
net/tap: add eBPF program file
net/tap: add eBPF API
net/tap: implement TAP RSS using eBPF
doc: detail new tap RSS feature in guides
doc/gui
. Each C array (corresponding to BPF
bytes code) is downloaded to the kernel using BPF systm call.
The C program and the external tool source files will be added in
separate commits.
In this commit the bytes code is not part of DPDK compilation
Signed-off-by: Ophir Munk
---
drivers/net/tap
ter an RSS hash was calculated on the packet
and saved in skb->cb[1]
Section l3_l4 - BPF action to calculate RSS hash on packet layers 3
and 4
This file is not part of DPDK tree compilation.
Signed-off-by: Ophir Munk
---
drivers/net/tap/tap_bpf_program.c | 221 +
() - create BPF RSS map for storing RSS
parameters per RSS rule
tap_flow_bpf_update_rss_elem() - update BPF map entry with RSS rule
parameters
Signed-off-by: Ophir Munk
---
drivers/net/tap/Makefile | 6 ++
drivers/net/tap/tap_bpf.h | 112 +
drivers/net/tap
nloaded to the kernel for each new RSS rule.
TAP eBPF requires Linux version 4.9 configured with BPF. TAP PMD will
successfully compile on systems with old or non-BPF configured kernels but
RSS rules creation on TAP devices will not be successful
Signed-off-by: Ophir Munk
---
drivers/net/ta
Signed-off-by: Ophir Munk
---
doc/guides/nics/tap.rst | 60 +
1 file changed, 60 insertions(+)
diff --git a/doc/guides/nics/tap.rst b/doc/guides/nics/tap.rst
index 04086b1..dc6f834 100644
--- a/doc/guides/nics/tap.rst
+++ b/doc/guides/nics/tap.rst
ring compilation time the required Linux headers are searched for.
If they are not present missing definitions are locally added
(tap_autoconf.h).
If the kernel cannot support a BPF operation - at runtime it will
gracefully reject the netlink message (with BPF) sent to it.
Ophir Munk (6):
Add a generic TC actions handling for TC actions: "mirred",
"gact", "skbedit". This will be useful when introducing
BPF actions, as it uses TCA_BPF_ACT instead of TCA_FLOWER_ACT
Signed-off-by: Ophir Munk
Acked-by: Pascal Mazon
---
drivers/net/tap/Makefile
n separate commits.
Signed-off-by: Ophir Munk
Acked-by: Pascal Mazon
---
drivers/net/tap/tap_bpf_insns.h | 1693 +++
1 file changed, 1693 insertions(+)
create mode 100644 drivers/net/tap/tap_bpf_insns.h
diff --git a/drivers/net/tap/tap_bpf_insns.h b/drive
ate RSS hash on packet
layers 3 and 4
This file is not part of DPDK tree compilation.
Signed-off-by: Ophir Munk
Acked-by: Pascal Mazon
---
drivers/net/tap/tap_bpf_program.c | 221 ++
1 file changed, 221 insertions(+)
create mode 100644 drivers/net/
Signed-off-by: Ophir Munk
Acked-by: Pascal Mazon
---
doc/guides/nics/tap.rst | 60 +
1 file changed, 60 insertions(+)
diff --git a/doc/guides/nics/tap.rst b/doc/guides/nics/tap.rst
index 04086b1..dc6f834 100644
--- a/doc/guides/nics/tap.rst
+++ b
() - create BPF RSS map for storing RSS
parameters per RSS rule
tap_flow_bpf_update_rss_elem() - update BPF map entry with RSS rule
parameters
Signed-off-by: Ophir Munk
Acked-by: Pascal Mazon
---
drivers/net/tap/Makefile | 6 ++
drivers/net/tap/tap_bpf.h | 112
nloaded to the kernel for each new RSS rule.
TAP eBPF requires Linux version 4.9 configured with BPF. TAP PMD will
successfully compile on systems with old or non-BPF configured kernels but
RSS rules creation on TAP devices will not be successful
Signed-off-by: Ophir Munk
Acked-by: Pa
e v5 as is or replace it with v6.
Regards,
Ophir
> -Original Message-
> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
> Sent: Saturday, January 20, 2018 6:16 PM
> To: Pascal Mazon ; Ophir Munk
> ; dev@dpdk.org
> Cc: Thomas Monjalon ; Olga Shern
>
> Subject:
definitions required for TAP
compilation.
In order to guarantee a successful eBPF compilation everywhere all the
required definitions for TAP have been locally added instead of including
file
Signed-off-by: Ophir Munk
---
drivers/net/tap/Makefile | 5 -
drivers/net/tap/tap_bpf.h | 9 +
2
wer of 2, as follows:
uint32_t ports = rte_log2_u32(conf->ports.present + 1);
If 1 was not added, in the case of one port, the number of ports would
be falsely calculated as 0.
Fixes: 82642799 ("net/mlx4: check max number of ports dynamically")
Cc: sta...@dpdk.org
Signed-off-by: Ophir
In order to guarantee a successful vdev_netvsc compilation on
old Linux distributions remove CFLAGS -std=c11 and -pedantic
Otherwise old GCC compilers may complain as follows:
cc1: error: unrecognized command line option -std=c11
Signed-off-by: Ophir Munk
---
drivers/net/vdev_netvsc/Makefile
tform driver")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
drivers/net/vdev_netvsc/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/vdev_netvsc/Makefile b/drivers/net/vdev_netvsc/Makefile
index f2b2ac5..45351b8 100644
--- a/drivers/net/vdev_netvs
")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
drivers/net/failsafe/failsafe_ether.c | 1 +
drivers/net/failsafe/failsafe_ops.c | 52 +--
2 files changed, 45 insertions(+), 8 deletions(-)
diff --git a/drivers/net/failsafe/failsafe_ether.c
b/d
of layers 2, 3 and 4 (where layer 4 header size is taken as
that of TCP).
Signed-off-by: Ophir Munk
---
drivers/net/tap/rte_eth_tap.c | 42 --
1 file changed, 32 insertions(+), 10 deletions(-)
diff --git a/drivers/net/tap/rte_eth_tap.c b/drivers/net/tap
mbuf pool in TAP required by
librte_gso.
The number of buffers will be 64 - each of 128 bytes length.
Ophir Munk (2):
net/tap: calculate checksum for multi segs packets
net/tap: implement TAP TSO
drivers/net/tap/Makefile | 2 +-
drivers/net/tap/rte_eth_tap.c | 183
device which
supports TSO capability in HW.
For more details on dpdk librte_gso implementation please refer to dpdk
documentation.
The number of newly generated TSO segments is limited to 64.
Signed-off-by: Ophir Munk
---
drivers/net/tap/Makefile | 2 +-
drivers/net/tap/rte_eth_tap.c | 157
commit uses these new APIs to allow disabling CRC stripping through
rte configuration.
A user can specify --disable-crc-stipping in testpmd command line or
'port config all crc-strip off' in testpmd interactive shell.
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4.c | 4
d
Previous to this commit mlx4 CRC stripping was executed by default and
there was no verbs API to disable it.
Current support in MLNX_OFED_4.3-1.5.0.0 and above
Signed-off-by: Ophir Munk
---
v1: initial version
v2: following internal reviews
drivers/net/mlx4/mlx4.c | 4
drivers/net
Hi Adrien,
v3 is ready and will be sent soon.
Please see inline more comments.
> -Original Message-
> From: Adrien Mazarguil [mailto:adrien.mazarg...@6wind.com]
> Sent: Friday, March 23, 2018 5:53 PM
> To: Ophir Munk
> Cc: dev@dpdk.org; Thomas Monjalon ; Olga Shern
&g
Previous to this commit mlx4 CRC stripping was executed by default and
there was no verbs API to disable it.
Signed-off-by: Ophir Munk
---
v1: initial version
v2: following internal reviews
v3: following dpdk.org mailing list reviews
drivers/net/mlx4/mlx4.c | 6 +-
drivers/net/mlx4
ot;)
Fixes: de96fe68ae95 ("net/tap: add basic flow API patterns and actions")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
Pascal - I would appreciate your feedback on the following discussion:
> > /* Immediately create the netdevice (this will create the 1st queue). */
> &g
ot;)
Fixes: de96fe68ae95 ("net/tap: add basic flow API patterns and actions")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
Pascal - I would appreciate your feedback on the follwoing review exchange:
> > /* Immediately create the netdevice (this will create the 1st queue). */
&g
ot;)
Fixes: de96fe68ae95 ("net/tap: add basic flow API patterns and actions")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
Comment 1:
Reason for V2: fixing coding style. Previous versions of this patch are
supersseded
Comment 2:
Pascal - I would appreciate your feedback on the fol
ot;)
Fixes: de96fe68ae95 ("net/tap: add basic flow API patterns and actions")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
In reply to Pascal Mazon review meesage-id:
<0e18a611-3b35-6624-013b-59c72f713...@6wind.com>
drivers/net/tap/rte_eth_tap.c | 139 +++
fail-safe PMD")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
This is V2 patch is in reply to <20170911083117.gm21...@bidouze.vm.6wind.com>
drivers/net/failsafe/failsafe_ether.c | 1 +
drivers/net/failsafe/failsafe_ops.c | 31 +++
driv
From: Moti Haimovsky
Modify PMD to send single-buffer packets directly to the device
bypassing the ibv Tx post and poll routines.
Signed-off-by: Moti Haimovsky
---
drivers/net/mlx4/mlx4_prm.h | 108 +
drivers/net/mlx4/mlx4_rxtx.c | 354 ---
From: Moti Haimovsky
This patch adds support for accessing the hardware directly when
handling Rx packets eliminating the need to use verbs in the Rx
datapath.
The number of scatters is limited to one.
Signed-off-by: Vasily Philipov
Signed-off-by: Moti Haimovsky
Signed-off-by: Ophir Munk
Changes from v2:
* Split "net/mlx4: support multi-segments Rx" commit from "net/mlx4: get back
Rx flow functionality" commit
* Semantics, code styling
* Fix check-git-log warnings
* Fix checkpatches warnings
Next (currently not included) changes:
* Replacing MLX4_TRANSPOSE() macro (Generic macro
From: Vasily Philipov
Getting hw directly on RX fast path without verbs call.
Now the number of scatters is calculated on the fly according to the
maximum expected packet size.
Signed-off-by: Vasily Philipov
Signed-off-by: Ophir Munk
---
This commit is a split from a previous commit
&quo
From: Moti Haimovsky
This patch adds support for transmitting packets spanning over
multiple buffers.
In this patch we also take into consideration the amount of entries
a packet occupies in the TxQ when setting the report-completion flag
of the chip.
Signed-off-by: Moti Haimovsky
---
drivers/
From: Moti Haimovsky
This patch adds hardware offloading support for IPV4, UDP and TCP
checksum calculation.
This commit also includes support for offloading IPV4, UDP and TCP
tunnel checksum calculation to the hardware.
Signed-off-by: Moti Haimovsky
---
drivers/net/mlx4/mlx4.c| 9 +++
From: Moti Haimovsky
This patch adds hardware offloading support for IPV4, UDP and TCP
checksum verification.
This commit also includes support for offloading IPV4, UDP and TCP tunnel
checksum verification to the hardware.
Signed-off-by: Vasily Philipov
---
drivers/net/mlx4/mlx4.c| 2
From: Moti Haimovsky
This patch adds loopback functionality used when the chip is a VF
in order to enable packet transmission between VFs and between VFs and PF.
Signed-off-by: Moti Haimovsky
---
drivers/net/mlx4/mlx4_rxtx.c | 38 ++
drivers/net/mlx4/mlx4_rx
From: Moti Haimovsky
This patch adds support for accessing the hardware directly when handling
Rx packets eliminating the need to use Verbs in the Rx data path.
The number of scatters is limited to one.
Signed-off-by: Vasily Philipov
Signed-off-by: Moti Haimovsky
Signed-off-by: Ophir Munk
From: Moti Haimovsky
Modify PMD to send single-buffer packets directly to the device bypassing
the Verbs Tx post and poll routines.
Signed-off-by: Moti Haimovsky
Acked-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_prm.h | 120 +++
drivers/net/mlx4/mlx4_rxtx.c | 337 +
Calculate the number of scatters on the fly according to
the maximum expected packet size.
Signed-off-by: Vasily Philipov
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_rxq.c | 64 +---
drivers/net/mlx4/mlx4_rxtx.c | 11 +---
drivers/net/mlx4
ues.
Moti Haimovsky (6):
net/mlx4: add simple Tx bypassing Verbs
net/mlx4: restore full Rx support bypassing Verbs
net/mlx4: restore Tx gather support
net/mlx4: restore Tx checksum offloads
net/mlx4: restore Rx offloads
net/mlx4: add loopback Tx from VF
Ophir Munk (1):
net/mlx4: res
From: Moti Haimovsky
This patch adds support for transmitting packets spanning over multiple
buffers.
In this patch we also take into consideration the amount of entries a
packet occupies in the TxQ when setting the report-completion flag of the
chip.
Signed-off-by: Moti Haimovsky
Acked-by: Ad
From: Moti Haimovsky
This patch adds loopback functionality used when the chip is a VF in order
to enable packet transmission between VFs and PF.
Signed-off-by: Moti Haimovsky
Acked-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_rxtx.c | 33 +
drivers/net/mlx4/
From: Moti Haimovsky
This patch adds hardware offloading support for IPv4, UDP and TCP checksum
calculation, including inner/outer checksums on supported tunnel types.
Signed-off-by: Moti Haimovsky
Acked-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4.c| 11 +++
drivers/net/mlx
From: Moti Haimovsky
This patch adds hardware offloading support for IPV4, UDP and TCP checksum
verification, including inner/outer checksums on supported tunnel types.
It also restores packet type recognition support.
Signed-off-by: Vasily Philipov
Signed-off-by: Moti Haimovsky
Acked-by: Adr
fail-safe PMD")
Cc: sta...@dpdk.org
Signed-off-by: Ophir Munk
---
v3:
1. Rebase v2
2. Please ignore checkpatch checks on arguments re-usage - they are confirmed.
CHECK:MACRO_ARG_REUSE: Macro argument reuse ... possible side-effects?
#217: FILE: drivers/net/failsafe/failsafe_pri
Hi Gaetan,
Thanks for your quick reply. Please see comments inline.
Regards,
Ophir
> -Original Message-
> From: Gaëtan Rivet [mailto:gaetan.ri...@6wind.com]
> Sent: Friday, October 20, 2017 1:35 PM
> To: Ophir Munk
> Cc: dev@dpdk.org; Thomas Monjalon ; Olga Shern
>
/mlx4: improve performance of one Tx segment
Ophir Munk (3):
net/mlx4: inline more Tx functions
net/mlx4: save lkey in big-endian format
net/mlx4: separate Tx for multi-segments
drivers/net/mlx4/mlx4_rxtx.c | 345 +++
drivers/net/mlx4/mlx4_rxtx.h
Change functions to inline on Tx fast path to improve performance
Inside the inline function call other functions to handle "unlikely"
cases such that the inline function code footprint is small.
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_r
From: Matan Azrad
Move unnecessary error flows to DEBUG mode for
performance improvements.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 16 ++--
1 file changed, 6 insertions(+), 10 deletions(-)
diff --git a/drivers/net/mlx4/mlx4_rxtx.c b/drivers/net/mlx4/mlx4_rxtx
201 - 300 of 567 matches
Mail list logo
