There are a number of consistency errors with the DQO Tx datapath, which
this series aims to address. These changes should make the driver much
more resilient against edge case and malformed packets overall.
Joshua Washington (7):
net/gve: send whole packet when mbuf is large
net/gve: clean wh
When TSO was introduced, it became possible for two differing descriptor
formats to be written to the descriptor ring, GVE_TX_PKT_DESC_DTYPE_DQO
and GVE_TX_TSO_CTX_DESC_DTYPE_DQO. Because these descriptor types have
different formats, they end up setting different fields, which can be
misinterprete
The DQ queue format expects that any MTU-sized packet or segment will
only cross at most 10 data descriptors.
In the non-TSO case, this means that a given packet simply can have at
most 10 descriptors.
In the TSO case, things are a bit more complex. For large TSO packets,
mbufs must be parsed and
The hardware supports at most 10 data descriptors per MTU-sized segment.
GVE_TX_MAX_DATA_DESCS was defined in the initial implmenentation, but
the descriptor limit was never actually enforced.
Fixes: 4022ff56 ("net/gve: support basic Tx data path for DQO")
Cc: junfeng@intel.com
Cc: sta...@
Writing zero-length descriptors to the hardware can cause the hardware
to reject that packet and stop transmitting altogether.
Fixes: 4022ff56 ("net/gve: support basic Tx data path for DQO")
Cc: junfeng@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Joshua Washington
Reviewed-by: Ankit Gar
The hardware assumes that a mismatch between the reported packet length
and the total amount of data in the descriptors is caused by a malicious
driver, leading the hardware to disable transmission altogether. To
avoid such a scenario, use rte_mbuf_check to validate that the mbuf is
correctly forme
A single packet can technically require more than 32 (free_thresh)
descriptors to send. Count the number of descriptors needed to send out
a packet in DQO Tx, and ensure that there are enough descriptors in the
ring before writing. If there are not enough free descriptors, drop the
packet and incre
Before this patch, only one descriptor would be written per mbuf in a
packet. In cases like TSO, it is possible for a single mbuf to have more
bytes than GVE_MAX_TX_BUF_SIZE_DQO. As such, instead of simply
truncating the data down to this size, the driver should actually write
descriptors for the r
From: Vladimir Medvedkin [mailto:medvedk...@gmail.com]
Sent: Monday, 7 July 2025 22.10
Hi Morten, all,
пн, 7 июл. 2025 г. в 19:09, Morten Brørup :
> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> Sent: Friday, 4 July 2025 13.32
> Hi all,
https://bugs.dpdk.org/show_bug.cgi?id=1750
Bug ID: 1750
Summary: ixgbe: X553 link remains down (NO-CARRIER)
Product: DPDK
Version: 24.11
Hardware: x86
OS: Linux
Status: UNCONFIRMED
Severity: normal
Hi Morten, all,
пн, 7 июл. 2025 г. в 19:09, Morten Brørup :
> > From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> > Sent: Friday, 4 July 2025 13.32
>
> > Hi all,
> >
> > this email discussion comes at a bit of a fortunate time for me, as I'm
> > currently looking at our vlan tag/qinq
https://bugs.dpdk.org/show_bug.cgi?id=1749
Bug ID: 1749
Summary: Ninja Build Error on Intel CPU
Product: DPDK
Version: 25.03
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: normal
Priority: N
> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> Sent: Friday, 4 July 2025 13.32
> Hi all,
>
> this email discussion comes at a bit of a fortunate time for me, as I'm
> currently looking at our vlan tag/qinq stripping behaviour in our Intel
> NIC
> drivers, and there is some discuss
On Fri, 4 Jul 2025 19:20:31 +0530
madhuker.myt...@oracle.com wrote:
> From: Madhuker Mythri
>
> In bonding PMD, member_start() function checks whether RSS(mq_mode) is
> enabled and then calling the RSS rte_eth_dev_rss_reta_update() API, which
> is returning error in-case of device configured in
On Sat, Jul 05, 2025 at 09:54:07PM +, Dhanya Pillai wrote:
> Currently, i40e driver is missing support for RTE_ETH_RSS_IPV4
> flag. Configuring rss function to RTE_ETH_RSS_IPV4 returns
> invalid rss hash function value.
>
> Fixes: a286ebeb0714 ("net/i40e: add dynamic mapping of SW flow types t
On Mon, 7 Jul 2025 13:56:17 +0800
Junlong Wang wrote:
> Fix error handling when MAC configuration fails.
> This is issue was discovered during internal testing.
>
> Fixes: 1c8f68b64cfc ("net/zxdh: optimize MAC operations")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Junlong Wang
> ---
> drivers
On Mon, 7 Jul 2025 07:58:19 -0400
Jie Liu wrote:
> Add Fix compiling issues.
>
> Signed-off-by: Jie Liu
> ---
The patch series is not build in CI.
Please fix each patch so that it builds on Linux with GCC and Clang.
The overall patch set needs to build on all supported platforms before review
FYI based on conversation with Dean I pushed a fixup adding
the use_virtual_functions: false field to the test_run.example.yaml, so
that users can more easily understand how to enable the VF testruns.
https://git.dpdk.org/next/dpdk-next-dts/commit/?id=934206f3bfc10fb8dcd852479f3683c01a18e352
On Mon, 7 Jul 2025 07:58:19 -0400
Jie Liu wrote:
> Add Fix compiling issues.
>
> Signed-off-by: Jie Liu
Please integrate these into the individual patches.
Each patch in series must compile cleanly because at some later time
a developer will be needing to do "git bisect" and if one of these
p
On Mon, 7 Jul 2025 07:58:06 -0400
Jie Liu wrote:
> +Dynamic Logging Parameters
> +~~
> +
> +One may leverage EAL option "--log-level" to change default levels
> +for the log types supported by the driver. The option is used with
> +an argument typically consisting of two
On Mon, 7 Jul 2025 07:58:06 -0400
Jie Liu wrote:
> diff --git a/doc/guides/nics/index.rst b/doc/guides/nics/index.rst
> index 10a2eca3b0..67e6374c69 100644
> --- a/doc/guides/nics/index.rst
> +++ b/doc/guides/nics/index.rst
> @@ -72,3 +72,4 @@ Network Interface Controller Drivers
> vmxnet3
> diff --git a/drivers/net/sxe/Makefile b/drivers/net/sxe/Makefile
> new file mode 100644
> index 00..f75fd35e32
> --- /dev/null
> +++ b/drivers/net/sxe/Makefile
> @@ -0,0 +1,66 @@
> +# SPDX-License-Identifier: BSD-3-Clause
> +# Copyright(c) 2010-2016 Intel Corporation
> +
> +include $(RT
On Mon, Jul 7, 2025 at 3:41 PM Aaron Conole wrote:
>
> David Marchand writes:
>
> > If not passing an explicit compiler for C++, meson uses c++ which
> > defaults to /usr/bin/g++ on a Ubuntu system.
> > Explicitly choose which compiler to use for C++.
> >
> > Signed-off-by: David Marchand
> > --
David Marchand writes:
> If not passing an explicit compiler for C++, meson uses c++ which
> defaults to /usr/bin/g++ on a Ubuntu system.
> Explicitly choose which compiler to use for C++.
>
> Signed-off-by: David Marchand
> ---
> Changes since v1:
> - rebased,
> - moved CXX tweaking in the bran
> From: Mario Kuka [mailto:k...@cesnet.cz]
> Sent: Monday, 7 July 2025 09.51
>
> Introduce a new ptype for identifying slow protocol packets.
>
> Signed-off-by: Mario Kuka
> * <'ether type'=[0x8847|0x8848]>
> */
> #define RTE_PTYPE_L2_ETHER_MPLS 0x000a
> +/**
> + * Ethernet
> From: Shani Peretz [mailto:shper...@nvidia.com]
> Sent: Monday, 7 July 2025 07.45
>
> > From: Stephen Hemminger
> > Sent: Monday, 16 June 2025 18:30
> >
> > On Mon, 16 Jun 2025 10:29:05 +0300
> > Shani Peretz wrote:
> >
> > > This feature is designed to monitor the lifecycle of mempool objects
Support rss offload.
Signed-off-by: Jie Liu
---
drivers/net/sxe/base/sxe_offload_common.c | 12 +-
drivers/net/sxe/pf/sxe.h | 2 +
drivers/net/sxe/pf/sxe_offload.c | 298 ++
drivers/net/sxe/pf/sxe_offload.h | 33 +++
4 files changed, 339
Add Fix compiling issues.
Signed-off-by: Jie Liu
---
drivers/net/sxe/base/sxe_compat_version.h | 1 -
drivers/net/sxe/base/sxe_hw.c | 21 ++---
drivers/net/sxe/base/sxe_offload_common.c | 1 -
drivers/net/sxe/base/sxe_queue_common.c | 3 ++-
drivers/net/sxe/base/
Add custom cmd led ctrl.
Signed-off-by: Jie Liu
---
drivers/net/sxe/Makefile| 2 +
drivers/net/sxe/include/sxe/sxe_cli.h | 218
drivers/net/sxe/include/sxe/sxe_ioctl.h | 21 +++
drivers/net/sxe/meson.build | 2 +
drivers/net/sxe/pf/sxe_
Add simd function.
Signed-off-by: Jie Liu
---
drivers/net/sxe/Makefile| 7 +
drivers/net/sxe/base/sxe_queue_common.c | 52 +-
drivers/net/sxe/base/sxe_rx_common.c| 125 -
drivers/net/sxe/meson.build | 9 +
drivers/net/sxe/pf/sxe.h| 3 +
Add xstats function.
Signed-off-by: Jie Liu
---
drivers/net/sxe/Makefile | 1 +
drivers/net/sxe/meson.build| 1 +
drivers/net/sxe/pf/sxe.h | 2 +
drivers/net/sxe/pf/sxe_main.c | 2 +
drivers/net/sxe/pf/sxe_stats.c | 577 +
drivers/net/sxe/
Add ptp module.
Signed-off-by: Jie Liu
---
drivers/net/sxe/Makefile | 1 +
drivers/net/sxe/meson.build | 1 +
drivers/net/sxe/pf/sxe.h | 9 ++
drivers/net/sxe/pf/sxe_ptp.c | 209 +++
drivers/net/sxe/pf/sxe_ptp.h | 26 +
5 files changed, 246 ins
Add dcb function.
Signed-off-by: Jie Liu i
---
drivers/net/sxe/Makefile| 2 +
drivers/net/sxe/meson.build | 1 +
drivers/net/sxe/pf/rte_pmd_sxe.h| 3 +-
drivers/net/sxe/pf/sxe.h| 2 +
drivers/net/sxe/pf/sxe_dcb.c| 944 +++
Add filter function.
Signed-off-by: Jie Liu
---
drivers/net/sxe/pf/sxe.h| 4 +
drivers/net/sxe/pf/sxe_filter.c | 191
drivers/net/sxe/pf/sxe_filter.h | 29 +
drivers/net/sxe/pf/sxe_main.c | 1 +
4 files changed, 225 insertions(+)
diff --git a/
Support vlan filter.
Signed-off-by: Jie Liu
---
drivers/net/sxe/base/sxe_offload_common.c | 4 +
drivers/net/sxe/pf/sxe.h | 5 +-
drivers/net/sxe/pf/sxe_filter.c | 277 ++
drivers/net/sxe/pf/sxe_filter.h | 13 +
drivers/net/sxe/pf/sxe
Add link, flow ctrl, mac ops, mtu ops function.
Signed-off-by: Jie Liu
---
drivers/net/sxe/Makefile | 3 +
drivers/net/sxe/meson.build| 3 +
drivers/net/sxe/pf/sxe.h | 11 +
drivers/net/sxe/pf/sxe_ethdev.c| 443 +-
drivers/net/sxe/pf/sxe_filter.c
Adding a minimum maintainable directory structure for the
network driver and request maintenance of the sxe driver.
Signed-off-by: Jie Liu
---
MAINTAINERS | 6 +++
doc/guides/nics/features/sxe.ini| 80 +
doc/guides/nics/features/sxe_vf.ini
On 2025-07-07 11:00, Naga Harish K, S V wrote:
-Original Message-
From: Mattias Rönnblom
Sent: Thursday, July 3, 2025 1:50 AM
To: Naga Harish K, S V ; Mattias Rönnblom
; dev@dpdk.org
Cc: Jerin Jacob ; Peter Nilsson
Subject: Re: [PATCH] event/eth_tx: prefetch mbuf headers
On 2025-05-
On 2025-07-05 06:11, Pathak, Pravin wrote:
-Original Message-
From: Mattias Rönnblom
Sent: Thursday, July 3, 2025 5:25 AM
To: Pathak, Pravin ; Jerin Jacob
Cc: dev@dpdk.org; jer...@marvell.com; Chen, Mike Ximing
; Richardson, Bruce
; tho...@monjalon.net; Marchand, David
; nipun.gu...@
Two small improvements for the cmdline file processing in testpmd.
* Now that we support multiple files, change the prompt to indicate what
file is currently being processed, and just print an EOF message when
done.
* When not echoing, the "Read" verb in the message "Read CLI commands..."
is
While testpmd allows a set of pre-prepared commands to be passed into it
at startup via the "cmdline-file" (and cmdline-file-noecho) parameters,
this is currently limited to a single file. By extending this support
to allow the parameter to be passed multiple (up to 16) times, we enable
users to ha
Some small improvements to cmdline file handling testpmd, inspired by the
desire to have support for multiple cmdline files passed on the commandline
of a testpmd run.
The implementation is somewhat complicated by the setting for echo/noecho
of the commands, because the current implementation uses
The echoing of commands contained in a file loaded via the "load"
command on testpmd CLI is governed by whether or not a cmdline-file or
cmdline-file-noecho had been passed to testpmd on the commandline.
Remove the use of a global setting for this, and explicitly add a
"load_echo" command to match
DPDK auxiliary bus first creates a list of known devices in the
`auxiliary_bus.device_list` and then tries to attach a driver to each
device on that list.
If the driver attachment has failed, a device will remain on the list.
The device destruction did not validate driver existence and
crashed.
https://bugs.dpdk.org/show_bug.cgi?id=1748
Bug ID: 1748
Summary: Clarification on Possible Buffer Overflow Issues
Reported by Fortify Tool
Product: DPDK
Version: 23.11
Hardware: All
OS: All
Statu
> -Original Message-
> From: Mattias Rönnblom
> Sent: Thursday, July 3, 2025 1:50 AM
> To: Naga Harish K, S V ; Mattias Rönnblom
> ; dev@dpdk.org
> Cc: Jerin Jacob ; Peter Nilsson
>
> Subject: Re: [PATCH] event/eth_tx: prefetch mbuf headers
>
> On 2025-05-27 12:55, Naga Harish K, S V w
Introduce a new ptype for identifying slow protocol packets.
Signed-off-by: Mario Kuka
---
* Added 'destination mac'=01-80-C2-00-00-02 required by slow protocols.
* Change ether type value from optional to mandatory.
dts/framework/remote_session/testpmd_shell.py | 2 ++
lib/mbuf/rte_mbuf_ptype.
-邮件原件-
发件人: Stephen Hemminger
发送时间: 2025年7月7日 11:33
收件人: wangfeifei (J)
抄送: Feifei Wang ; dev@dpdk.org; chenyi (CY)
主题: Re: 回复: [V5 00/18] add hinic3 pmd driver
On Mon, 7 Jul 2025 03:27:20 +
"wangfeifei (J)" wrote:
> -邮件原件-
> 发件人: Stephen Hemminger
> 发送时间: 2025年7月2日 2
48 matches
Mail list logo
