This patch adds an IAVF testpmd command "set tx lldp on" which
will register an mbuf dynfield IAVF_TX_LLDP_DYNFIELD to indicate
the need to test transmit LLDP packet.
It needs to stop and restart Tx port to select correct Tx path.
Signed-off-by: Zhichao Zeng
---
doc/guides/nics/intel_vf.rst
This patch adds an avx512 ctx Tx path that supports context descriptor,
filling in the SWTCH_UPLINK bit based on mbuf
dynfield IAVF_TX_LLDP_DYNFIELD to support transmitting LLDP packet.
Signed-off-by: Zhichao Zeng
---
drivers/net/iavf/iavf_rxtx.c| 5 +
drivers/net/iavf/iavf_rxtx
This patch adds an mbuf dynfield IAVF_TX_LLDP_DYNFIELD to determine
whether or not to fill the SWTCH_UPLINK bit in the Tx context descriptor
to transmit LLDP packet.
Signed-off-by: Zhichao Zeng
---
drivers/net/iavf/iavf_ethdev.c | 5 +
drivers/net/iavf/iavf_rxtx.c | 13 +++--
driv
This patch set adds an IAVF testpmd command "set tx lldp on" which
will register an mbuf dynfield IAVF_TX_LLDP_DYNFIELD to indicate
the need to test TX LLDP packet. It needs to stop and restart Tx port
to select correct Tx path.
IAVF will fill the SWTCH_UPLINK bit in the Tx context descriptor base
Ping
> --
> This patch introduces new function to get rss device config and adds code to
> forward the RSS control command to backend through hw control queue if
> RSS feature is negotiated.
> This patch will help to negotiate VIR
If crypto device requires headroom and tailroom, the mbuf
of dst in out-of-place should reserve the headroom and
tailroom as well, otherwise there will be no enough room
for dst mbuf.
Fixes: bf9d6702eca9 ("app/crypto-perf: use single mempool")
Signed-off-by: Suanming Mou
---
app/test-crypto-per
This commit fixes the bugs in multi-segment size calculation.
*** BLURB HERE ***
Suanming Mou (2):
app/test-crypto-perf: fix copy segment size calculation
app/test-crypto-perf: fix dst_mbuf size calculation
app/test-crypto-perf/cperf_test_common.c | 5 +++--
1 file changed, 3 insertions(+),
For the case crypto device requires headroom and tailroom,
the segment_sz in options also contains the headroom_sz
and tailroom_sz, but mbuf's data_len is user's segment_sz
without headroom_sz and tailroom_sz. That means the data
size to be copied should use user's segment_sz instead
of options->se
AEAD users RTE_CRYPTO_AEAD_OP_* with aead_op and CIPHER uses
RTE_CRYPTO_CIPHER_OP_* with cipher_op in current code.
This commit aligns aead_op and cipher_op operation to fix
incorrect AEAD verification.
Fixes: df52cb3b6e13 ("app/crypto-perf: move verify as single test type")
Signed-off-by: Suanm
The function memcmp() returns an integer less than, equal to,
or greater than zero. In current code, if the first memcmp()
returns less than zero and the second memcmp() returns greater
than zero, the sum of results may still be 0 and indicates
verify succussed.
This commit converts the return val
In fill_multi_seg_mbuf(), when remaining_segments is 0,
rte_mbuf m's next should pointer to NULL instead of a
new rte_mbuf, that casues setting m->next as NULL out
of the while loop to the invalid mbuf.
This commit fixes the invalid mbuf next operation.
Fixes: bf9d6702eca9 ("app/crypto-perf: use
When the vf starts to request multiple queues, the pf sends a reset
command to the vf. During the reset process, adminq sends an abnormal
message to pf for an unknown reason, and the resource request fails
resulting in a coredump.
This patch fixes the issue by checking the reset state before reset
> -Original Message-
> From: Su, Simei
> Sent: Thursday, December 28, 2023 4:38 PM
> To: Zhang, Qi Z
> Cc: dev@dpdk.org; Yang, Qiming ; Su, Simei
>
> Subject: [PATCH] net/ice: updated 23.11 recommended matching list
>
> Signed-off-by: Simei Su
> ---
> doc/guides/nics/ice.rst | 2 ++
> -Original Message-
> From: Mingjin Ye
> Sent: Tuesday, January 2, 2024 6:52 PM
> To: dev@dpdk.org
> Cc: Yang, Qiming ; Ye, MingjinX
> ; Wu, Wenjun1 ; Zhang,
> Yuying ; Xing, Beilei ; Su,
> Simei ; Wu, Jingjing
> Subject: [PATCH v7 2/2] net/iavf: add diagnostic support in TX path
>
>
Hi Chuanyu,
> -Original Message-
> From: Chuanyu Xue
> Sent: Saturday, December 30, 2023 5:30 AM
> To: Su, Simei
> Cc: Xing, Beilei ; chuanyu@uconn.edu;
> dev@dpdk.org; Zhang, Qi Z ; Lu, Wenzhuo
>
> Subject: RE: [PATCH] net/e1000: support launchtime feature
>
> >>
> >> >> +static i
> -Original Message-
> From: Mingjin Ye
> Sent: Tuesday, January 2, 2024 6:52 PM
> To: dev@dpdk.org
> Cc: Yang, Qiming ; Ye, MingjinX
> ; sta...@dpdk.org; Wu, Jingjing
> ; Xing, Beilei
> Subject: [PATCH v7 1/2] net/iavf: fix Rx/Tx burst in multi-process
>
> In a multi-process environm
Txonly forwarding engine does not call the Tx preparation API
before transmitting packets. This may cause some problems.
TSO breaks when MSS spans more than 8 data fragments. Those
packets will be dropped by Tx preparation API, but it will cause
MDD event if txonly forwarding engine does not call
Stop writing RTE_TOOLCHAIN_XXX macros to rte_build_config.h. When an
application builds it doesn't necessarily use the same toolchain that
DPDK was built with.
Instead evaluate toolchain predefined macros and define
RTE_TOOLCHAIN_XXX macros as appropriate each time rte_config.h is
preprocessed.
S
When EAL is built with MSVC it is possible to dynamically load plugins
on Windows. Hook eal_plugins_init into rte_eal_init if built with MSVC
and provide code to load plugins on Windows.
Signed-off-by: Tyler Retzlaff
---
lib/eal/common/eal_common_options.c | 90 ++
When EAL is built with MSVC it is possible to dynamically load plugins
on Windows. Hook eal_plugins_init into rte_eal_init if built with MSVC
and provide code to load plugins on Windows.
Tyler Retzlaff (1):
eal: initialize shared plugins on Windows
lib/eal/common/eal_common_options.c | 90
The design of the macros requires a type to be provided to the macro.
By expanding the type parameter inside of typeof it also inadvertently
allows an expression to be used which appears not to have been intended
after evaluating the parameter name and existing macro use.
Technically this is an A
The design of the macros requires a type to be provided to the macro.
By expanding the type parameter inside of typeof it also inadvertently
allows an expression to be used which appears not to have been intended
after evaluating the parameter name and existing macro use.
Technically this is an A
Adjust the use of per lcore macro to provide type as the first argument
as to not require __typeof__ during expansion.
Signed-off-by: Tyler Retzlaff
---
lib/eal/common/eal_common_errno.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/eal/common/eal_common_errno.c
b/lib/
Adjust the use of per lcore macro to provide type as the first
argument as to not require __typeof__ during expansion.
Signed-off-by: Tyler Retzlaff
---
drivers/crypto/ipsec_mb/pmd_aesni_gcm_priv.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/crypto/ipsec_mb/pmd_ae
On Tue, 2 Jan 2024 07:24:24 -0500
Qi Zhang wrote:
> Added Testpmd CLI support for dumping Tx scheduling tree.
>
> Usage:
> testpmd>txsched dump
>
> The output file is in "dot" format, which can be converted
> into an image file using Graphviz.
>
> - In "brief" mode, all scheduling nodes
Stop writing RTE_TOOLCHAIN_XXX macros to rte_build_config.h. When an
application builds it doesn't necessarily use the same toolchain that
DPDK was built with.
Instead evaluate toolchain predefined macros and define
RTE_TOOLCHAIN_XXX macros as appropriate each time rte_config.h is
preprocessed.
S
> On x86 platforms, max lcores are limited to 128 by default.
>
> On AMD EPYC processors, this limit was adjusted for native
> builds in the previous patch.
> https://patches.dpdk.org/project/dpdk/patch/
> 20230925151027.558546-1-sivaprasad.tumm...@amd.com/
>
> As agreed earlier in mailing list
> Use rte_pktmbuf_free_bulk() API instead of looping through the packets
> and freeing individually.
>
> Signed-off-by: Anoob Joseph
> Suggested-by: Stephen Hemminger
> ---
> examples/ipsec-secgw/ipsec-secgw.h | 5 +
> 1 file changed, 1 insertion(+), 4 deletions(-)
>
> diff --git a/examp
Currently Tx hierarchy commit only take effect if device
already be started, as after a dev start / stop cycle, queues
has been removed and added back which cause the Tx scheduler
tree return to original topo.
In this patch, the hierarchy commit function will simply return
if device has not be sta
1. Always reset all Tx scheduler at the beginning of a commit action.
This prevent unexpected remains from previous commit.
2. Reset all Tx scheduler nodes if a commit failed.
For leaf node, stop queues which will remove sched node from
scheduler tree, then start queues which will add sched nod
Enable the configuration of weight for Tx scheduler node at
the queue group level. This patch also consolidate weight
configuration across various levels by exposing the base
code API 'ice_sched_cfg_node_bw_alloc'.
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_sched.c | 2 +-
drivers/net
Consolidate Tx scheduler node configuration into a function:
'ice_cfg_hw_node", where rate limit, weight, priority will be
configured for queue group level and queue level.
Signed-off-by: Qi Zhang
---
drivers/net/ice/ice_tm.c | 97
1 file changed, 49 inse
Enable the configuration of peak and committed rates for a Tx scheduler
node at the VSI level. This patch also consolidate rate configuration
across various levels into a single function 'ice_set_node_rate.'
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_sched.c | 2 +-
drivers/net/ice/b
The committed flag for tx schedular configuration is not used
in PF only mode, remove the redundent code.
Signed-off-by: Qi Zhang
---
drivers/net/ice/ice_tm.c | 14 --
1 file changed, 14 deletions(-)
diff --git a/drivers/net/ice/ice_tm.c b/drivers/net/ice/ice_tm.c
index f5ea47ae83..
The patchset enhanced ice rte_tm implemenations
Qi Zhang (6):
net/ice: remove redundent code
net/ice: support VSI level bandwidth config
net/ice: support queue group weight configure
net/ice: refactor hardware Tx sched node config
net/ice: reset Tx sched node during commit
net/ice: sup
The only way to enable diagnostics for TX paths is to modify the
application source code. Making it difficult to diagnose faults.
In this patch, the devarg option "mbuf_check" is introduced and the
parameters are configured to enable the corresponding diagnostics.
supported cases: mbuf, size, seg
In a multi-process environment, a secondary process operates on shared
memory and changes the function pointer of the primary process, resulting
in a crash when the primary process cannot find the function address
during an Rx/Tx burst.
Fixes: 5b3124a0a6ef ("net/iavf: support no polling when link
Fixed Rx/Tx crash in multi-process environment and
added Tx diagnostic feature.
Mingjin Ye (2):
net/iavf: fix Rx/Tx burst in multi-process
net/iavf: add diagnostic support in TX path
doc/guides/nics/intel_vf.rst | 4 +
drivers/net/iavf/iavf.h| 54 ++-
drivers/net/iavf/iavf_e
> -Original Message-
> From: Rakesh Kudurumalla
> Sent: Tuesday, January 2, 2024 1:01 PM
> To: Nithin Kumar Dabilpuram ; Pavan
> Nikhilesh Bhagavatula
> Cc: dev@dpdk.org; Jerin Jacob Kollanukkaran ;
> Rakesh Kudurumalla
> Subject: [EXT] [PATCH v10 1/3] node: support to add next node to e
39 matches
Mail list logo
