> From: Feifei Wang [mailto:feifei.wa...@arm.com]
> Sent: Tuesday, 5 December 2023 04.13
>
> 在 2023/12/4 15:41, Morten Brørup 写道:
> >> From: Feifei Wang [mailto:feifei.wa...@arm.com]
> >> Sent: Monday, 4 December 2023 03.39
> >>
> >> For 'rte_pktmbuf_prefree_seg' function, 'rte_mbuf_refcnt_read(m)
s/robustify/enhance/ ?
On 10/23/23 11:55, David Marchand wrote:
A simple comment in vhost_user_msg_handler() is not that robust.
Add a lock_all_qps property to message handlers so that their
implementation can add a build check and assert a vq is locked.
Signed-off-by: David Marchand
---
li
Why is m->nb_segs initialized in rte_pktmbuf_prefree_seg()?
It's part of the m->rearm_data, and will be initialized on RX descriptor rearm
anyway.
Slightly related:
When built without RTE_IOVA_IN_MBUF, m->next lives in the first cache line.
Theoretically, it would improve performance to initi
On 10/23/23 11:55, David Marchand wrote:
Now that a r/w lock is used, the access_ok field should only be updated
under a write lock.
Since the datapath code only takes a read lock on the virtqueue to check
access_ok, this lock must be released and a write lock taken before
calling vring_trans
By default all packets received on ethdev_rx node
is forwarded to pkt_cls node.This patch provides
library support to add a new node as next node to
ethdev_rx node and forward packet to new node from
rx node.
Signed-off-by: Rakesh Kudurumalla
---
v4: Resolve compilation issues
lib/node/ethdev_c
Adds a txport to forward packet for every rxport
Mapping will be used to forward packets to txport
received on rxport
Following commands are exposed:
- ethdev forward "
Signed-off-by: Rakesh Kudurumalla
---
app/graph/cli.c | 1 +
app/graph/ethdev.c | 62 +
Added portforward usecase.In this usecase
packets received Rx port is forwarded to
respective Tx port.
Signed-off-by: Rakesh Kudurumalla
---
app/graph/ethdev.c | 12 ++
app/graph/ethdev.h | 1 +
app/graph/examples/l2fwd.cli |
> -Original Message-
> From: Tomasz Duszynski
> Sent: Saturday, November 4, 2023 2:40 AM
> To: dev@dpdk.org; Jakub Palider ; Tomasz Duszynski
>
> Cc: Jerin Jacob Kollanukkaran ;
> step...@networkplumber.org
> Subject: [PATCH] raw/cnxk_gpio: switch to dynamic logging
>
> Dynamically al
On 12/5/2023 1:23 AM, wangyunjian wrote:
>
>
>> -Original Message-
>> From: Ferruh Yigit [mailto:ferruh.yi...@amd.com]
>> Sent: Monday, December 4, 2023 10:10 PM
>> To: wangyunjian ; dev@dpdk.org
>> Cc: ciara.lof...@intel.com; qi.z.zh...@intel.com; xudingke
>> ; Lilijun (Jerry) ;
>> sta..
This patch dumps contents of receviced packet descriptor from CQ
for debug to file
Signed-off-by: Rakesh Kudurumalla
---
drivers/common/cnxk/roc_cpt.h | 2 +-
drivers/common/cnxk/roc_cpt_debug.c | 56 +
drivers/common/cnxk/roc_nix.h | 2 +-
drivers/commo
Now that a r/w lock is used, the access_ok field should only be updated
under a write lock.
Since the datapath code only takes a read lock on the virtqueue to check
access_ok, this lock must be released and a write lock taken before
calling vring_translate().
Fixes: 03f77d66d966 ("vhost: change v
Add support for eth_rx_descriptor_dump for cn9k and cn10k.
This patch dumps contents of receviced packet descriptor from CQ
for debug to file
Signed-off-by: Rakesh Kudurumalla
---
drivers/net/cnxk/cn10k_ethdev.c | 67 +
drivers/net/cnxk/cn9k_ethdev.c | 53 +++
Modifying vq->access_ok should be done with a write lock taken.
Annotate vring_translate() and vring_invalidate().
Signed-off-by: David Marchand
Acked-by: Eelco Chaudron
---
Changes since v1:
- moved fixes in separate patches,
---
lib/vhost/vhost.h | 7 +--
lib/vhost/vhost_user.c | 8
vring_translate and vring_invalidate change the vq access_ok field.
The access_ok field should only be updated under a (write) lock.
Fixes: a9120db8b98b ("vhost: add VDUSE device startup")
Fixes: ad67c65efda1 ("vhost: add VDUSE device stop")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
Acke
Calling vring_invalidate must be done with a (write) lock taken on the
virtqueue.
Fixes: 72d002b3ebda ("vhost: fix vring address handling during live migration")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
Acked-by: Eelco Chaudron
---
Changes since v1:
- moved fix out of patch 3,
---
li
A simple comment in vhost_user_msg_handler() is not that robust.
Add a lock_all_qps property to message handlers so that their
implementation can add a build check and assert a vq is locked.
Signed-off-by: David Marchand
Acked-by: Eelco Chaudron
Reviewed-by: Maxime Coquelin
---
Changes since v
On Tue, Dec 05, 2023 at 09:04:08AM +0100, Morten Brørup wrote:
> > From: Feifei Wang [mailto:feifei.wa...@arm.com]
> > Sent: Tuesday, 5 December 2023 04.13
> >
> > 在 2023/12/4 15:41, Morten Brørup 写道:
> > >> From: Feifei Wang [mailto:feifei.wa...@arm.com]
> > >> Sent: Monday, 4 December 2023 03.39
On Wed, Nov 8, 2023 at 9:03 AM wrote:
>
> From: Satheesh Paul
>
> Added ROC API to get MKEX capability.
>
> Signed-off-by: Satheesh Paul
> Reviewed-by: Kiran Kumar K
Applied to dpdk-next-net-mrvl/for-main. Thanks
> ---
> drivers/common/cnxk/roc_npc.c | 24
> drive
On Tue, Dec 05, 2023 at 10:07:56AM +0100, Morten Brørup wrote:
> Why is m->nb_segs initialized in rte_pktmbuf_prefree_seg()?
>
> It's part of the m->rearm_data, and will be initialized on RX descriptor
> rearm anyway.
>
Presumably this is to have a sane default for apps that allocate buffers
dir
On 12/5/23 10:45, David Marchand wrote:
vring_translate and vring_invalidate change the vq access_ok field.
The access_ok field should only be updated under a (write) lock.
Fixes: a9120db8b98b ("vhost: add VDUSE device startup")
Fixes: ad67c65efda1 ("vhost: add VDUSE device stop")
Cc: sta...@
On 12/5/23 10:45, David Marchand wrote:
Calling vring_invalidate must be done with a (write) lock taken on the
virtqueue.
Fixes: 72d002b3ebda ("vhost: fix vring address handling during live migration")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
Acked-by: Eelco Chaudron
---
Changes s
API mbox_alloc_msg_npa_aq_enq() mbox is called if SQ needs
to be updated from mbox response else mbox call to kernel
is bypassed reducing the time taken to complete
roc_nix_tm_sq_aura_fc() function.This reduces ethdev
teardown time by 20%.
Signed-off-by: Rakesh Kudurumalla
---
drivers/common/cnx
Added a new roc API to disable SQB aura FC
and update SQ state to disabled state in TX queue
stop.The same SQ status is verified during sq flush
to enable or disable SQB aura FC during ethdev
teardown.This fix reduces teardown time by 90%.
Signed-off-by: Rakesh Kudurumalla
---
drivers/common/cnx
Invoked newly added roc API to disable
SQB aura FC during TX queue start and TX queue
stop. This fix reduces ethdev teardown time
Signed-off-by: Rakesh Kudurumalla
---
drivers/net/cnxk/cnxk_ethdev.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/cnxk/cnxk_eth
> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> Sent: Tuesday, 5 December 2023 10.56
>
> On Tue, Dec 05, 2023 at 10:07:56AM +0100, Morten Brørup wrote:
> > Why is m->nb_segs initialized in rte_pktmbuf_prefree_seg()?
> >
> > It's part of the m->rearm_data, and will be initialized on
Hi Maryam,
Apologies for making it confusing.
As per the afxdp-plugins-for-kubernetes code it looks like the udsPodPath or
udsSockDir[1] and udsPodSock[1] are constant and not changing. Only the
interface name is changing and we are already passing the interface name
through the command line.
Hi Burakov,
The patch has been sent for your review for quite some time. Can you please
help to take a look and give some feedback? Your prompt response will be
appreciated a lot and very helpful for our next step moving forward.
-Mingjin
> -Original Message-
> From: Ye, MingjinX
>
PMD does not switch to no polling mode when the PF triggers a reset event
or the watchdog detects a reset event. In this scenario, data path will
access the freed resources and cause a core dump.
This patch fixes this issue by automatically switching modes on VF reset.
Fixes: 5b3124a0a6ef ("net/i
From 4cf72842a07b2270876939fd2bb2367efaad95f4 Mon Sep 17 00:00:00 2001
From: Yuan Zhiyuan
Date: Fri, 1 Dec 2023 11:27:51 +
Subject: [PATCH] vhost: optimize vhost user get protocol features
variable features is unused in vhost_user_get_protocol_features.
Signed-off-by: Yuan Zhiyuan
---
lib/
On 12/5/23 10:45, David Marchand wrote:
Modifying vq->access_ok should be done with a write lock taken.
Annotate vring_translate() and vring_invalidate().
Signed-off-by: David Marchand
Acked-by: Eelco Chaudron
---
Changes since v1:
- moved fixes in separate patches,
---
lib/vhost/vhost.h
On Tue, Nov 28, 2023 at 10:17 PM Srikanth Yalavarthi
wrote:
>
> From: Anup Prabhu
>
> Exclude the hardware and firmware latency of model data
> caching run from xstats calculation.
>
> Fixes: 9cfad6c334f2 ("ml/cnxk: update device and model xstats functions")
>
> Signed-off-by: Anup Prabhu
> Acke
Hi Shibin
As I've already explained in my previous email, they are constant ATM,
however they will become configurable. I am implementing the Operator
and it will make a lot of these "fixed" params configurable. My
recommendation is not to try to generate the path in DPDK - as it's
likely to
On 04/12/2023 17:18, Koikkara Reeny, Shibin wrote:
Prerequisites
@@ -223,8 +224,7 @@ Howto run dpdk-testpmd with CNI plugin:
securityContext:
capabilities:
add:
- - CAP_NET_RAW
- - CAP_BPF
+ - NET_RAW
Need to updat
The C code generated for the tokens for matching IP addresses in
commandlines was missing the "static" prefix present in the output for
the other data-types.
Fixes: 3791e9ed ("buildtools: add a tool to generate cmdline boilerplate")
Cc: sta...@dpdk.org
Reported-by: Sunil Kumar Kori
Signed-of
https://bugs.dpdk.org/show_bug.cgi?id=1333
Bug ID: 1333
Summary: DMA Perf test unable to parse multiple DMA devices
from configuration file
Product: DPDK
Version: 23.11
Hardware: x86
OS: Linux
St
From: Maryam Tahhan
Sent: Tuesday, December 5, 2023 11:31 AM
To: Koikkara Reeny, Shibin ;
ferruh.yi...@amd.com; step...@networkplumber.org; lihuis...@huawei.com;
fengcheng...@huawei.com; liuyongl...@huawei.com
Cc: dev@dpdk.org
Subject: Re: [v2] net/af_xdp: enable a sock path alongside use_cni
Hi Jochen,
The VMXNET3 driver supports two different max MTU sizes, depending on VMware
guest version:
https://elixir.bootlin.com/dpdk/latest/source/drivers/net/vmxnet3/vmxnet3_ethdev.c#L1650
I think the reported max MTU size should reflect this too, e.g.:
https://elixir.bootlin.com/dpdk/latest/
From: Satheesh Paul
Adding support for represented port flow item for cnxk device.
Signed-off-by: Kiran Kumar K
Signed-off-by: Satheesh Paul
---
v2:
* Fix issue with flow key alg for cnxk.
doc/guides/nics/features/cnxk.ini| 1 +
doc/guides/nics/features/cnxk_vf.ini | 1 +
drivers/com
Hi Maryam,
I was suggesting with reference to the constant ATM and the implementation in
afxdp_plugin. If you think they will also be changing in the future then please
go with what you think will be the best.
Regards,
Shibin
From: Maryam Tahhan
Sent: Tuesday, December 5, 2023 11:29 AM
To: Ko
In xdp_umem_configure() allocated memzone for the 'umem', we should
free it when xsk_umem__create() call fails, otherwise it will lead to
memory zone leak. To fix it move 'umem->mz = mz;' assignment after
'mz == NULL' check.
Fixes: f1debd77efaf ("net/af_xdp: introduce AF_XDP PMD")
Cc: sta...@dpdk.
> -Original Message-
> From: Ferruh Yigit [mailto:ferruh.yi...@amd.com]
> Sent: Tuesday, December 5, 2023 5:42 PM
> To: wangyunjian ; dev@dpdk.org
> Cc: ciara.lof...@intel.com; qi.z.zh...@intel.com; xudingke
> ; Lilijun (Jerry) ;
> sta...@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] net/af_
>
> In xdp_umem_configure() allocated memzone for the 'umem', we should
> free it when xsk_umem__create() call fails, otherwise it will lead to
> memory zone leak. To fix it move 'umem->mz = mz;' assignment after
> 'mz == NULL' check.
>
> Fixes: f1debd77efaf ("net/af_xdp: introduce AF_XDP PMD")
>
> -Original Message-
> From: Maryam Tahhan
> Sent: Monday, December 4, 2023 10:31 AM
> To: ferruh.yi...@amd.com; step...@networkplumber.org;
> lihuis...@huawei.com; fengcheng...@huawei.com;
> liuyongl...@huawei.com; Koikkara Reeny, Shibin
>
> Cc: dev@dpdk.org; Tahhan, Maryam
> Subject
On 05/12/2023 13:43, Loftus, Ciara wrote:
also be provided\n",
Thanks for the patch Maryam.
Do we really need the use_cni devarg anymore if we must also always pair it
with a uds_path string?
I am in favour of removing it if both yourself and Shibin think that makes
sense too.
Ciara
Hey Cia
Hi Ciara,
I agree.
Regards,
Shibin
From: Maryam Tahhan
Sent: Tuesday, December 5, 2023 2:38 PM
To: Loftus, Ciara ; ferruh.yi...@amd.com;
step...@networkplumber.org; lihuis...@huawei.com; fengcheng...@huawei.com;
liuyongl...@huawei.com; Koikkara Reeny, Shibin
Cc: dev@dpdk.org
Subject: Re: [v2
This set contains some small enhancements for the cmdline generation
script introduced in the last release. Specifically:
* Add support for commands with an optional variable parameter. This
is needed to support command pairs like testpmd's "start tx_first"
and "start tx_first 128" (to send 12
Sometimes a user may want to have a command which takes an optional
parameter. For commands with an optional constant string, this is no
issue, as it can be configured as two separate commands, e.g.
start
start tx_first.
However, if we want to have a variable parameter on the command, we hit
issu
The IP address type should be generic for both IPv4 and IPv6 and so use
the cmdline lib's TOKEN_IPADDR_INITIALIZER rather than
TOKEN_IPV4_INITIALIZER.
Fixes: 3791e9ed ("buildtools: add a tool to generate cmdline boilerplate")
Cc: sta...@dpdk.org
Signed-off-by: Bruce Richardson
---
buildtool
Add support for generating cmdline lib code to just match IPv4 addresses
or IPv6 addresses, rather than IP addresses in general.
Signed-off-by: Bruce Richardson
---
buildtools/dpdk-cmdline-gen.py| 12
doc/guides/prog_guide/cmdline.rst | 4
2 files changed, 16 insertions(+)
On Mon, 4 Dec 2023 05:31:01 -0500
Maryam Tahhan wrote:
> With the original 'use_cni' implementation, (using a
> hardcoded socket rather than a configurable one),
> if a single pod is requesting multiple net devices
> and these devices are from different pools, then
> the container attempts to mo
Reviewed-by: Jeremy Spewock
On Mon, Dec 4, 2023 at 5:24 AM Juraj Linkeš
wrote:
> Format according to the Google format and PEP257, with slight
> deviations.
>
> Signed-off-by: Juraj Linkeš
> ---
> dts/tests/TestSuite_hello_world.py | 16 +---
> dts/tests/TestSuite_os_udp.py | 20
On Mon, 4 Dec 2023 11:07:08 +
Konstantin Ananyev wrote:
> > > 2.25.1
> >
> > NAK.
> >
> > This patch is not race safe.
>
> +1, It is a bad idea.
The patch does raise a couple of issues that could be addressed by
rearranging. There is duplicate code, and there are no comments
to expla
Refactor to support for 5th Gen Intel Xeon Scalable processors. Note that
NTB devices within the 3rd, 4th and 5th Gen Intel Xeon Scalable processors
share the same device id, and compliant to PCIe 4.0 spec.
It is more reasonable to distinguish NTB devices by their PCIe generation.
Signed-off-by: J
> -Original Message-
> From: Stephen Hemminger
> Sent: Wednesday, December 6, 2023 2:34 AM
> To: Joyce Kong
> Cc: Jakub Grajciar ; Morten Brørup
> ; Ruifeng Wang ;
> dev@dpdk.org; nd ; sta...@dpdk.org; Liangxing Wang
>
> Subject: Re: [PATCH v1] net/memif: fix segfault with Tx burst la
This patch adds a testpmd command "set tx lldp on" which will register
an mbuf dynflag at the application level, currently only supported on.
The IAVF will fills the SWTCH_UPLINK bit in the Tx context descriptor
based on the mbuf dynflag to send the LLDP packets.
For avx512, need to close the Tx
> -Original Message-
> From: Bruce Richardson
> Sent: Tuesday, December 5, 2023 5:01 PM
> To: dev@dpdk.org
> Cc: Bruce Richardson ; sta...@dpdk.org;
> Sunil Kumar Kori
> Subject: [EXT] [PATCH] buildtools/dpdk-cmdline-gen: fix code gen for IP
> addresses
>
> External Email
>
> --
> -Original Message-
> From: Bruce Richardson
> Sent: Tuesday, December 5, 2023 8:21 PM
> To: dev@dpdk.org
> Cc: Sunil Kumar Kori ; david.march...@redhat.com;
> Bruce Richardson
> Subject: [EXT] [PATCH 1/3] buildtools/dpdk-cmdline-gen: support optional
> parameters
>
> External Email
>
> -Original Message-
> From: Bruce Richardson
> Sent: Tuesday, December 5, 2023 8:21 PM
> To: dev@dpdk.org
> Cc: Sunil Kumar Kori ; david.march...@redhat.com;
> Bruce Richardson ; sta...@dpdk.org
> Subject: [EXT] [PATCH 2/3] buildtools/dpdk-cmdline-gen: fix IP address
> initializer
>
> Ex
> -Original Message-
> From: Bruce Richardson
> Sent: Tuesday, December 5, 2023 8:21 PM
> To: dev@dpdk.org
> Cc: Sunil Kumar Kori ; david.march...@redhat.com;
> Bruce Richardson
> Subject: [EXT] [PATCH 3/3] buildtools/dpdk-cmdline-gen: add explicit IPv4
> and v6 types
>
> External Email
59 matches
Mail list logo
