Can you tell me how to set up email reply? I don’t quite understand what you
said about inline replies, and the green text and ">" are automatically
generated?
On 9/10/2021 14:35，Wang, Haiyue wrote：
Thanks for the tool.
Anyway, please setup your mail firstly.
Your mail reply is bad for commun
On Fri, Sep 10, 2021 at 4:38 AM Chenbo Xia wrote:
>
> PCI device id and address in structure rte_kni_conf are never used
> in the test and kni library. So remove the related code.
This is probably a leftover after ethtool support removal.
ea6b39b5b847 ("kni: remove ethtool support")
ee3cac929ede
On Fri, Sep 10, 2021 at 4:38 AM Chenbo Xia wrote:
>
> PCI ID and address in structure rte_kni_conf are never used and
> will be removed in kni library. So remove the setting of them
> first in the example.
>
> Signed-off-by: Chenbo Xia
(forgot to Cc: Ferruh in previous comment)
Ditto patch 5.
I
On 9/10/21 3:05 PM, Miao Li wrote:
> This patch adds two command line arguments which will be needed when using
> virtio/vhost vdev. One argument sets rx offloads capabilities
> DEV_RX_OFFLOAD_VLAN_STRIP. The other argument sets DCB, PSS and VMDQ off
> for RX side.
>
> Signed-off-by: Miao Li
>
In DPDK 21.05, iavf already became the default VF for i40e devices.
So remove i40evf due to it's no need to maintain now.
v3:
- remove VF related code in i40e_rxtx.c.
v2:
- update 21.11 release note, remove some missed documentation.
Robin Zhang (3):
net/i40e: remove i40evf
net/iavf: remove
The default VF driver for Intel 700 Series Ethernet Controller already
switch to iavf in DPDK 21.05. And i40evf is no need to maintain now,
so remove i40evf related code.
Signed-off-by: Robin Zhang
---
drivers/net/i40e/i40e_ethdev.h|1 -
drivers/net/i40e/i40e_ethdev_vf.c | 3015 -
Due to i40evf will be removed, so there's no need to keep the devargs
option "driver=i40evf" in iavf.
Signed-off-by: Robin Zhang
---
drivers/net/iavf/iavf_ethdev.c | 52 ++
1 file changed, 2 insertions(+), 50 deletions(-)
diff --git a/drivers/net/iavf/iavf_ethdev
As announced in the deprecation note, i40evf has been removed. So removing
deprecation notice and other i40evf related documentation.
Signed-off-by: Robin Zhang
---
doc/guides/howto/lm_bond_virtio_sriov.rst | 4 ++--
doc/guides/nics/intel_vf.rst | 8 +---
doc/guides/rel_notes/de
On 09/09/2021 17:40, Rahul Shah wrote:
> rte_port_eventdev_reader_ops, rte_port_eventdev_writer_nodrops_ops,
> rte_port_eventdev_writer_ops symbols promoted
>
> Signed-off-by: Rahul Shah
> ---
> lib/port/version.map | 8 +++-
> 1 file changed, 3 insertions(+), 5 deletions(-)
>
Hi Rahul,
> -Original Message-
> From: Ma, WenwuX
> Sent: Tuesday, September 7, 2021 4:49 AM
> To: dev@dpdk.org
> Cc: maxime.coque...@redhat.com; Xia, Chenbo ;
> Jiang, Cheng1 ; Hu, Jiayu ;
> Pai G, Sunil ; Yang, YvonneX
> ; Ma, WenwuX
> Subject: [PATCH 0/4] support async dequeue for split ring
> -Original Message-
> From: Ma, WenwuX
> Sent: Tuesday, September 7, 2021 4:49 AM
> To: dev@dpdk.org
> Cc: maxime.coque...@redhat.com; Xia, Chenbo ;
> Jiang, Cheng1 ; Hu, Jiayu ;
> Pai G, Sunil ; Yang, YvonneX
> ; Wang, YuanX ; Ma,
> WenwuX ; Wang, Yinan
> Subject: [PATCH 1/4] vhost:
On Fri, Sep 10, 2021 at 9:31 AM Kinsella, Ray wrote:
> On 09/09/2021 17:40, Rahul Shah wrote:
> > rte_port_eventdev_reader_ops, rte_port_eventdev_writer_nodrops_ops,
> > rte_port_eventdev_writer_ops symbols promoted
> >
> > Signed-off-by: Rahul Shah
> > ---
> > lib/port/version.map | 8 +++-
> -Original Message-
> From: Ma, WenwuX
> Sent: Tuesday, September 7, 2021 4:49 AM
> To: dev@dpdk.org
> Cc: maxime.coque...@redhat.com; Xia, Chenbo ;
> Jiang, Cheng1 ; Hu, Jiayu ;
> Pai G, Sunil ; Yang, YvonneX
> ; Ma, WenwuX
> Subject: [PATCH 2/4] examples/vhost: refactor vhost enqueu
> -Original Message-
> From: Ma, WenwuX
> Sent: Tuesday, September 7, 2021 4:49 AM
> To: dev@dpdk.org
> Cc: maxime.coque...@redhat.com; Xia, Chenbo ;
> Jiang, Cheng1 ; Hu, Jiayu ;
> Pai G, Sunil ; Yang, YvonneX
> ; Ma, WenwuX
> Subject: [PATCH 3/4] examples/vhost: use a new API to quer
> -Original Message-
> From: Ma, WenwuX
> Sent: Tuesday, September 7, 2021 4:49 AM
> To: dev@dpdk.org
> Cc: maxime.coque...@redhat.com; Xia, Chenbo ;
> Jiang, Cheng1 ; Hu, Jiayu ;
> Pai G, Sunil ; Yang, YvonneX
> ; Ma, WenwuX
> Subject: [PATCH 4/4] examples/vhost: support vhost async d
During the port probe process, there are two abnormal branches that did
not release the previously requested memory, resulting in leakage. The
patch adds an iavf_uninit_vf function, which corresponds to the
iavf_init_vf function.
Fixes: ff2d0c345c3b ("net/iavf: support generic flow API")
Cc: sta..
On Thu, Sep 09, 2021 at 06:13:07PM -0500, Honnappa Nagarahalli wrote:
> In the rte_eal_remote_launch function, the lcore function pointer is
> checked for NULL. However, the pointer is never reset to NULL. Reset the
> lcore function pointer and argument after the worker has completed
> executing th
On Fri, Sep 10, 2021 at 08:51:04AM +0200, David Marchand wrote:
> On Thu, Sep 9, 2021 at 4:38 PM Bruce Richardson
> wrote:
> >
> > On Thu, Sep 09, 2021 at 02:45:06PM +0100, David Hunt wrote:
> > > Modern processors are coming with an ever increasing number of cores,
> > > and 128 does not seem lik
2021-09-09 16:33 (UTC-0700), Stephen Hemminger:
[...]
> + prm = rte_zmalloc("bpf_filter",
> + sizeof(*prm) + ebpf_len * sizeof(*ebpf), 0);
> + if (prm == NULL) {
> + rte_errno = ENOMEM;
> + return NULL;
> + }
> +
> + /* The EPBF instruct
On Fri, Sep 10, 2021 at 9:54 AM Bruce Richardson
wrote:
>
> On Fri, Sep 10, 2021 at 08:51:04AM +0200, David Marchand wrote:
> > On Thu, Sep 9, 2021 at 4:38 PM Bruce Richardson
> > wrote:
> > >
> > > On Thu, Sep 09, 2021 at 02:45:06PM +0100, David Hunt wrote:
> > > > Modern processors are coming w
On 9/9/21 8:02 PM, Ferruh Yigit wrote:
> On 8/30/2021 11:42 AM, Thomas Monjalon wrote:
>> A lot of flags are parts of a group but are documented alone.
>> The Doxygen syntax @{ and @} for grouping is used
>> to make flags appear together and have a common description.
>>
>> Some Rx/Tx offload flags
On Fri, Sep 10, 2021 at 9:49 AM Bruce Richardson
wrote:
>
> On Thu, Sep 09, 2021 at 06:13:07PM -0500, Honnappa Nagarahalli wrote:
> > In the rte_eal_remote_launch function, the lcore function pointer is
> > checked for NULL. However, the pointer is never reset to NULL. Reset the
> > lcore function
2021-09-09 16:33 (UTC-0700), Stephen Hemminger:
> The original version of the pdump library was building on
> Windows, but it was useless since the pdump utility was not being
> built.
>
> The new version of pdump with filtering now has dependency
> on bpf. But bpf library is not available on Wind
Hi David,
> -Original Message-
> From: David Marchand
> Sent: Friday, September 10, 2021 3:18 PM
> To: Xia, Chenbo
> Cc: dev ; Dumitrescu, Cristian ;
> Yigit, Ferruh
> Subject: Re: [dpdk-dev] [PATCH 6/8] examples/ip_pipeline: remove setting of
> PCI ID and address
>
> On Fri, Sep 10, 2
In the iavf_dev_rx_queue_start function, if the iavf_switch_queue
or iavf_switch_queue_lv function fails, the previously applied mbuf
is not released, resulting in leakage. The patch fixes the problem.
Fixes: 9cf9c02bf6ee ("net/iavf: add enable/disable queues for large VF")
Cc: sta...@dpdk.org
---
Previously, we set txq affinity to 0 and let firmware
to perform round-robin when bonding. Firmware uses a
global counter to assign txq affinity to different
physical ports accord to remainder after division.
There are three dis-advantages:
1. The global counter is shared between kernel and dpdk.
Previously, we set txq affinity to 0 and let firmware
to perform round-robin when bonding. Firmware uses a
global counter to assign txq affinity to different
physical ports accord to remainder after division.
There are three dis-advantages:
1. The global counter is shared between kernel and dpdk.
10/09/2021 10:06, David Marchand:
> On Fri, Sep 10, 2021 at 9:54 AM Bruce Richardson
> wrote:
> >
> > On Fri, Sep 10, 2021 at 08:51:04AM +0200, David Marchand wrote:
> > > On Thu, Sep 9, 2021 at 4:38 PM Bruce Richardson
> > > wrote:
> > > >
> > > > On Thu, Sep 09, 2021 at 02:45:06PM +0100, David
In DPDK 21.05, iavf already became the default VF for i40e devices.
So remove i40evf due to it's no need to maintain now.
v4:
- resolve compile warning issue.
v3:
- remove VF related code in i40e_rxtx.c.
v2:
- update 21.11 release note, remove some missed documentation.
Robin Zhang (3):
net/i
The default VF driver for Intel 700 Series Ethernet Controller already
switch to iavf in DPDK 21.05. And i40evf is no need to maintain now,
so remove i40evf related code.
Signed-off-by: Robin Zhang
---
drivers/net/i40e/i40e_ethdev.h|1 -
drivers/net/i40e/i40e_ethdev_vf.c | 3015 -
Due to i40evf will be removed, so there's no need to keep the devargs
option "driver=i40evf" in iavf.
Signed-off-by: Robin Zhang
---
drivers/net/iavf/iavf_ethdev.c | 52 ++
1 file changed, 2 insertions(+), 50 deletions(-)
diff --git a/drivers/net/iavf/iavf_ethdev
As announced in the deprecation note, i40evf has been removed. So removing
deprecation notice and other i40evf related documentation.
Signed-off-by: Robin Zhang
---
doc/guides/howto/lm_bond_virtio_sriov.rst | 4 ++--
doc/guides/nics/intel_vf.rst | 8 +---
doc/guides/rel_notes/de
Below patches are target to DPDK 20.11.
They are based on the previous patch set in
http://patches.dpdk.org/project/dpdk/list/?series=18158&state=%2A&archive=both.
They are not for LTS upstream, just for users to cherrypick.
feature includes
1. flow priority for switch.
2. support maximum to 256 q
From: Yuying Zhang
This patch is not for LTS upstream, just for users to cherry-pick.
Priority 0 and 1 were supported in switch filter. However,
FDIR, ACL and RSS don't support priority 1. Add priority check
in FDIR, ACL and RSS filter parse functions.
Signed-off-by: Yuying Zhang
---
drivers/
This patch is not for LTS upstream, just for users to cherry-pick.
256 queues can be allowed now. Improve the code to support 256 queues
for per PF.
Signed-off-by: Dapeng Yu
Signed-off-by: Wenjun Wu
---
drivers/net/ice/ice_ethdev.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
dif
From: Yuying Zhang
This patch is not for LTS upstream, just for users to cherry-pick.
Current code uses switch filter as backup of FDIR in non-pipeline
mode and Value 1 denotes higher priority. This patch refines priority
usage to match the original design of rte_flow attribute. When priority
is
This patch is not for LTS upstream, just for users to cherry-pick.
Add 6-tuple RSS support for pattern eth/vlan/ipv4, eth/vlan/ipv4/sctp,
eth/vlan/ipv4/udp,eth/vlan/ipv4/tcp.
Signed-off-by: Wenjun Wu
---
drivers/net/ice/ice_hash.c | 15 +++
1 file changed, 15 insertions(+)
diff --g
This patch is not for LTS upstream, just for users to cherry-pick.
The queue number actually applied should be the maximum integer power
of 2 less than or equal to min(vsi->nb_qps, ICE_MAX_Q_PER_TC), so we
need to get the most significant 1 bit. However the return value of
function rte_bsf32 is th
This patch is not for LTS upstream, just for users to cherry-pick.
Add L4 support for QinQ switch filter as following flow patterns:
eth / vlan / vlan / ipv4 / udp
eth / vlan / vlan / ipv4 / tcp
eth / vlan / vlan / ipv6 / udp
eth / vlan / vlan / ipv6 / tcp
Signed-off-by: Steve Yang
Signed-off-by
From: Steve Yang
This patch is not for LTS upstream, just for users to cherry-pick.
This patch adds more dummy packet types for QinQ packet,
it enables tcp/udp layer of ipv4/ipv6 for QinQ payload,
so we can use L4 dst/src port as input set for switch
filter.
For Example:
flow create 0 ingre
In the iavf_dev_rx_queue_start function, if the iavf_switch_queue
or iavf_switch_queue_lv function fails, the previously applied mbuf
is not released, resulting in leakage. The patch fixes the problem.
Fixes: 9cf9c02bf6ee ("net/iavf: add enable/disable queues for large VF")
Cc: sta...@dpdk.org
Si
In l3fwd-power, there is default port configuration which requires RSS and
IPV4/UDP/TCP checksum. Once device does not support these, the l3fwd-power will
exit and report an error.
Miao
-Original Message-
From: Maxime Coquelin
Sent: Friday, September 10, 2021 3:24 PM
To: Li, Miao ; de
On Fri, Sep 10, 2021 at 10:34 AM Li, Miao wrote:
>
> In l3fwd-power, there is default port configuration which requires RSS and
> IPV4/UDP/TCP checksum. Once device does not support these, the l3fwd-power
> will exit and report an error.
Maxime suggested to update the port configuration based o
The requirement to support filter by
dst MAC + outer VLAN id + inner VLAN id + dst IP + dst port.
Currently we only support pattern eth / vlan/ vlan/ IP,
in switch filter, the request to add l4 support as
eth / vlan / vlan / IP / udp|tcp.
Steve Yang (2):
net/ice: add L4 support for QinQ switch
Add L4 support for QinQ switch filter as following flow patterns:
eth / vlan / vlan / ipv4 / udp
eth / vlan / vlan / ipv4 / tcp
eth / vlan / vlan / ipv6 / udp
eth / vlan / vlan / ipv6 / tcp
Signed-off-by: Steve Yang
---
drivers/net/ice/ice_generic_flow.c | 4
drivers/net/ice/ice_switch_fi
This patch adds more dummy packet types for QinQ packet,
it enables tcp/udp layer of ipv4/ipv6 for QinQ payload,
so we can use L4 dst/src port as input set for switch
filter.
For Example:
flow create 0 ingress pattern eth / vlan tci is 2 / vlan tci is 1 \
/ ipv4 src is 10.1.1.1 / tcp src i
To improve performance of vhost Tx, merge repeated loop in eth_vhost_tx.
Move "vlan insert" from eth_vhost_tx to virtio_dev_rx_packed
and virtio_dev_rx_split to reduce a loop iteration.
Fixes: f63d356ee993 ("net/vhost: insert/strip VLAN header in software")
Cc: sta...@dpdk.org
Signed-off-by: Gaox
This series is mainly adding support for RSS to Virtio PMD
driver. The two last patches are fixing an issue in testpmd
that prevents to configure the RSS hash types, and fix
an issue spotted in the mlx5 driver while looking for
inspiration.
The first motivation for this series is to eventually
sup
This patch adds RSS support to the Virtio PMD. It provides
the capability to update the hash key, hash types and reta
table on the fly (without needing to stop/start the
device). However, the key length and the number of reta
entries are fixed to 40B and 128 entries respectively. This
is done in or
port_rss_hash_key_update() initializes rss_conf with the
RSS hash type and key provided by the user, but it calls
rte_eth_dev_rss_hash_conf_get() before calling
rte_eth_dev_rss_hash_update(), which overides the parsed
config with current NIC's config.
While the RSS key value is set again after, th
This patch fixes RETA updating for entries above 64.
Without ithat, these entries are never updated as
calculated mask value will always be 0.
Fixes: 634efbc2c8c0 ("mlx5: support RETA query and update")
Cc: sta...@dpdk.org
Cc: nelio.laranje...@6wind.com
Signed-off-by: Maxime Coquelin
---
driver
> -Original Message-
> From: Medvedkin, Vladimir
> Sent: Monday, September 6, 2021 5:03 PM
> To: dev@dpdk.org
> Cc: Mcnamara, John ; sta...@dpdk.org
> Subject: [PATCH] hash: fix doxygen comments
>
The git diff makes this look like the ifdef is moving but I see that you are
moving th
On 9/10/21 12:17 PM, Maxime Coquelin wrote:
> port_rss_hash_key_update() initializes rss_conf with the
> RSS hash type and key provided by the user, but it calls
> rte_eth_dev_rss_hash_conf_get() before calling
> rte_eth_dev_rss_hash_update(), which overides the parsed
> config with current NIC's c
On 9/10/21 11:51 AM, Andrew Rybchenko wrote:
> On 9/10/21 12:17 PM, Maxime Coquelin wrote:
>> port_rss_hash_key_update() initializes rss_conf with the
>> RSS hash type and key provided by the user, but it calls
>> rte_eth_dev_rss_hash_conf_get() before calling
>> rte_eth_dev_rss_hash_update(), w
On 9/10/21 12:17 PM, Maxime Coquelin wrote:
> This patch adds RSS support to the Virtio PMD. It provides
I'd drop the first sentence since summary already says so and
suggest to start the second from "Provide ..."
> the capability to update the hash key, hash types and reta
reta -> RETA (see dev
On 9/10/21 12:57 PM, Maxime Coquelin wrote:
>
>
> On 9/10/21 11:51 AM, Andrew Rybchenko wrote:
>> On 9/10/21 12:17 PM, Maxime Coquelin wrote:
>>> port_rss_hash_key_update() initializes rss_conf with the
>>> RSS hash type and key provided by the user, but it calls
>>> rte_eth_dev_rss_hash_conf_get
Currently, there is no way to map memory for DMA in a way that allows
unmapping it partially later, because some IOMMU's do not support
partial unmapping. There is a workaround of mapping all of these
segments separately, but this is inconvenient and silly, so this
commit adds a proper API that doe
Add support for dicts of dicts to telemetry library.
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
---
app/test/test_telemetry_data.c | 29 +++
lib/telemetry/telemetry.c | 43 +++---
lib/telemetry/telemetry_data.c | 3 ++-
3 file
Add support for:
TSO, NAT-T/UDP encapsulation, ESN
AES_CCM, CHACHA20_POLY1305 and AES_GMAC
SA telemetry
mbuf offload flags
Initial SQN value
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
Radu Nicolau (10):
securit
Allow user to provision a per security session maximum segment size
(MSS) for use when Transmit Segmentation Offload (TSO) is supported.
The MSS value will be used when PKT_TX_TCP_SEG or PKT_TX_UDP_SEG
ol_flags are specified in mbuf.
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Sign
Add support for specifying UDP port params for UDP encapsulation option.
RFC3948 section-2.1 does not enforce using specific the UDP ports for
UDP-Encapsulated ESP Header
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
Update ipsec_xform definition to include ESN field.
This allows the application to control the ESN starting value.
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
Acked-by: Fan Zhang
Acked-by: Anoob Joseph
---
lib/s
Add tunnel type for IPsec ESP tunnels
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
Acked-by: Fan Zhang
Acked-by: Akhil Goyal
---
lib/mbuf/rte_mbuf_core.h | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/mb
Add support for AES_CCM, CHACHA20_POLY1305 and AES_GMAC.
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
Acked-by: Fan Zhang
---
lib/ipsec/crypto.h | 137 +++
lib/ipsec/esp_i
Add support for transmit segmentation offload to inline crypto processing
mode. This offload is not supported by other offload modes, as at a
minimum it requires inline crypto for IPsec to be supported on the
network interface.
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off
Add support for the IPsec NAT-Traversal use case for Tunnel mode
packets.
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
Acked-by: Fan Zhang
---
lib/ipsec/iph.h | 17 +
lib/ipsec/rte_ipsec_s
Add telemetry support for ipsec SAs
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
Acked-by: Fan Zhang
---
lib/ipsec/esp_inb.c | 1 +
lib/ipsec/esp_outb.c | 12 +-
lib/ipsec/meson.build | 2 +-
lib/ipsec/rte
Update IPsec library to support initial SQN value.
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
Acked-by: Fan Zhang
---
lib/ipsec/esp_outb.c | 19 ---
lib/ipsec/sa.c | 29
Update the IPsec library to set mbuff->ol_flags and use the configured
L3 header length when setting the mbuff->tx_offload fields
Signed-off-by: Declan Doherty
Signed-off-by: Radu Nicolau
Signed-off-by: Abhijit Sinha
Signed-off-by: Daniel Martin Buckley
---
lib/ipsec/esp_inb.c | 17 +
Thanks for the review!
On 9/10/21 12:06 PM, Andrew Rybchenko wrote:
> On 9/10/21 12:17 PM, Maxime Coquelin wrote:
>> This patch adds RSS support to the Virtio PMD. It provides
>
> I'd drop the first sentence since summary already says so and
> suggest to start the second from "Provide ..."
>
>>
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 130 -
lib/pipeline/rte_swx_pipeline_internal.h | 178 +++
2 files changed, 203 insertions(+), 105 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_sw
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 12 ++---
lib/pipeline/rte_swx_pipeline_internal.h | 32
2 files changed, 34 insertions(+), 10 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_swx_pipeline.
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 80 ++
lib/pipeline/rte_swx_pipeline_internal.h | 100 +++
2 files changed, 123 insertions(+), 57 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_sw
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 26 +++-
lib/pipeline/rte_swx_pipeline_internal.h | 53
2 files changed, 58 insertions(+), 21 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_swx_pipeli
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 22 +++-
lib/pipeline/rte_swx_pipeline_internal.h | 43
2 files changed, 47 insertions(+), 18 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_swx_pipeli
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 38 --
lib/pipeline/rte_swx_pipeline_internal.h | 51
2 files changed, 51 insertions(+), 38 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_swx_
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 348 ++---
lib/pipeline/rte_swx_pipeline_internal.h | 616 +++
2 files changed, 660 insertions(+), 304 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_swx_pi
Start to consolidate the data structures and inline functions required
by the pipeline instructions into an internal header file.
Signed-off-by: Cristian Dumitrescu
---
Depends-on: series-18297 ("[V4,1/4] table: add support learner tables")
lib/pipeline/meson.build |4 +
lib
Move the thread inline functions to the internal header file.
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 56 --
lib/pipeline/rte_swx_pipeline_internal.h | 59
2 files changed, 59 insertions(+), 56 deletions(-)
d
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 162 -
lib/pipeline/rte_swx_pipeline_internal.h | 170 +++
2 files changed, 228 insertions(+), 104 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rt
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 86 +-
lib/pipeline/rte_swx_pipeline_internal.h | 90
2 files changed, 92 insertions(+), 84 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 320 ++-
lib/pipeline/rte_swx_pipeline_internal.h | 475 +++
2 files changed, 502 insertions(+), 293 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_swx_
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 457 +--
lib/pipeline/rte_swx_pipeline_internal.h | 541 +++
2 files changed, 558 insertions(+), 440 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 36 ++--
lib/pipeline/rte_swx_pipeline_internal.h | 55
2 files changed, 58 insertions(+), 33 deletions(-)
diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_swx_pi
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline_internal.h | 29
1 file changed, 29 insertions(+)
diff --git a/lib/pipeline/rte_swx_pipeline_internal.h
b/lib/pipeline/rte_swx_pipeline_internal.h
index 791adfb471..efd136196f 100644
--- a/lib/pipeline
As per ABI policy, move the formerly experimental API's to the stable
section.
Signed-off-by: Anatoly Burakov
---
lib/eal/include/rte_memory.h | 16
lib/eal/version.map | 10 --
2 files changed, 4 insertions(+), 22 deletions(-)
diff --git a/lib/eal/include/rte_
As per ABI policy, move the formerly experimental API's to the stable
section.
Signed-off-by: Anatoly Burakov
---
lib/eal/include/rte_eal.h | 24
lib/eal/version.map | 14 ++
2 files changed, 6 insertions(+), 32 deletions(-)
diff --git a/lib/eal/includ
Save the instruction meta-data for later use instead of freeing it up
once the instruction translation is completed.
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 9 ++---
lib/pipeline/rte_swx_pipeline_internal.h | 2 ++
2 files changed, 8 insertions(+), 3
As per ABI policy, move the formerly experimental API's to the stable
section.
Signed-off-by: Anatoly Burakov
---
lib/eal/include/rte_eal_memconfig.h | 12
lib/eal/version.map | 8 +++-
2 files changed, 3 insertions(+), 17 deletions(-)
diff --git a/lib/eal/incl
As per ABI policy, move the formerly experimental API's to the stable
section.
Signed-off-by: Anatoly Burakov
---
lib/eal/include/rte_memory.h | 12
lib/eal/version.map | 6 +++---
2 files changed, 3 insertions(+), 15 deletions(-)
diff --git a/lib/eal/include/rte_memory.h
As per ABI policy, move the formerly experimental API's to the stable
section.
Signed-off-by: Anatoly Burakov
---
lib/eal/include/rte_fbarray.h | 26 --
lib/eal/version.map | 52 +--
2 files changed, 26 insertions(+), 52 deletions(-)
dif
As per ABI policy, move the formerly experimental API's to the stable
section.
Signed-off-by: Anatoly Burakov
---
lib/eal/include/rte_memory.h | 17 -
lib/eal/version.map | 34 +-
2 files changed, 17 insertions(+), 34 deletions(-)
diff --
For better performance, the option to run a single function per action
is now provided, which requires a single function call per action that
can be better optimized by the C compiler, as opposed to one function
call per instruction. Special table lookup instructions are added to
to support this fe
As per ABI policy, move the formerly experimental API's to the stable
section.
Signed-off-by: Anatoly Burakov
---
lib/eal/include/rte_malloc.h | 10 --
lib/eal/version.map | 20 ++--
2 files changed, 10 insertions(+), 20 deletions(-)
diff --git a/lib/eal/include
For better performance, the option to create custom instructions when
the program is translated and add them on-the-fly to the pipeline is
now provided. Multiple regular instructions can now be consolidated
into a single C function optimized by the C compiler directly.
Signed-off-by: Cristian Dumi
Lay the foundation to generate C code for the pipeline: C functions
for actions and custom instructions are generated, built as shared
object library and loaded into the pipeline.
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 44 +
1 fil
Export the array of translated instructions to a C file. There is one
such array per action and one for the pipeline.
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 1097 +++
1 file changed, 1097 insertions(+)
diff --git a/lib/pipeline/rte_s
Generate a C function for each action. For most instructions, the
associated inline function is called directly. Special care is taken
for TX, jump and return instructions.
Signed-off-by: Cristian Dumitrescu
---
lib/pipeline/rte_swx_pipeline.c | 662
1 file chang
Generate a C function for each custom instruction, which essentially
consolidate multiple regular instructions into a single function call.
The pipeline program is split into groups of instructions, and a
custom instruction is generated for each group that has more than one
instruction. Special car
Build the generated C file into a shared object library.
Signed-off-by: Cristian Dumitrescu
Signed-off-by: Cunming Liang
---
lib/pipeline/rte_swx_pipeline.c | 131 +++
lib/pipeline/rte_swx_pipeline_internal.h | 1 +
2 files changed, 132 insertions(+)
diff --git a
1 - 100 of 175 matches
Mail list logo
