Configurable Rx CRC stripping is allowed only if
running firmware variant supports it and if NIC is
configured with single PF per port and without VFs.
When KEEP_CRC is supported, CRC will be part of
the packet payload. The packet length will also
contain CRC length. At the same time, CRC length
sh
From: Roman Zhukov
Drivers cannot determine if received packet includes the FCS or not.
Only packets with an external port have the FCS included and functions
without link control privilege cannot determine the MAC configuration.
This patch is trying to make assumptions that: if PF is the only fu
From: Roman Zhukov
A new variable was added to efx_nic_cfg_s to detect and
report if FCS is supported by FW.
Signed-off-by: Roman Zhukov
Signed-off-by: Denis Pryazhennikov
Reviewed-by: Viacheslav Galaktionov
Acked-by: Andrew Rybchenko
---
drivers/common/sfc_efx/base/ef10_nic.c | 6 ++
From: Sandilya Bhagi
NIC Partitioning mode in SFC devices means multiple PFs
per network port. When NIC Partitioning is configured,
apart from the privileged adapter(s) the other
unprivileged adapter(s) will share the same physical port.
Determining NIC Partitioning mode is required to take
neces
This patch series adds support for RTE_ETH_RX_OFFLOAD_KEEP_CRC
offload in the SFC driver.
Changes in v2:
* Applied review notes;
* Fixed stats handling;
* Added missing release notes to [4/4].
Denis Pryazhennikov (3):
common/sfc_efx/base: detect and report FCS include support
common/sfc_efx/b
> -Original Message-
> From: Thomas Monjalon
> Sent: Tuesday, June 20, 2023 3:42 PM
> To: De Lara Guarch, Pablo ; Fangming Fang
> ; Power, Ciara ;
> Honnappa Nagarahalli ; Ji, Kai
> ; Ruifeng Wang ; Mcnamara,
> John ; Akhil Goyal
> Cc: dev@dpdk.org; david.march...@redhat.com; Jerin Jac
Hello David,
What's the dpdk version used in that OS?
Have you tried manually installing latest dpdk and check?
We recently fixed some mem issue, raised by valgrind.
Regards,
Asaf Penso
From: bugzi...@dpdk.org
Sent: Tuesday, June 20, 2023 2:09:48 PM
To: dev@dpdk.o
On 6/21/2023 3:43 PM, David Marchand wrote:
> As reported by Ilya [1], unconditionally calling
> rte_flow_get_restore_info() impacts an application performance for drivers
> that do not provide this ops.
> It could also impact processing of packets that require no call to
> rte_flow_get_restore_inf
Acked-by: Jeremy Spewock
On Fri, Jun 9, 2023 at 5:46 AM Juraj Linkeš
wrote:
> Pexpect is not a dedicated SSH connection library while Fabric is. With
> Fabric, all SSH-related logic is provided and we can just focus on
> what's DTS specific.
>
> Signed-off-by: Juraj Linkeš
> ---
>
> Notes:
>
On 6/16/2023 9:57 AM, Chaoyong He wrote:
> From: Peng Zhang
>
> Now the PCIE configure bar address is always related with PF ID 0.
> For the NIC with multiple PCI device, this will lead the host crash.
> Fix this problem by make the PCIE configure bar address related with
> the PF ID.
>
> Fixes:
Acked-by: Jeremy Spweock
On Thu, May 11, 2023 at 5:14 AM Juraj Linkeš
wrote:
> WIP: only one module is reformatted to serve as a demonstration.
>
> The google format is documented here [0].
>
> [0]: https://google.github.io/styleguide/pyguide.html
>
> Signed-off-by: Juraj Linkeš
> ---
> dts/f
Acked-by: Jeremy Spweock
On Tue, May 16, 2023 at 6:32 AM Juraj Linkeš
wrote:
> On Tue, May 16, 2023 at 12:16 PM Juraj Linkeš
> wrote:
> >
> > Poetry changed the syntax of dev dependencies section in version
> > 1.2.0. Remove the scripts section that did nothing.
> > Update Pylama linters:
> >
On 6/16/2023 9:14 AM, Chaoyong He wrote:
> From: Peng Zhang
>
> The basic logic is the same for PF, VF and representor port.
> Split out the initializing device information logics into a
> helper function to reduce code duplication.
>
> Signed-off-by: Peng Zhang
> Reviewed-by: Chaoyong He
> Re
On 6/15/2023 7:51 AM, Chaoyong He wrote:
> Use the complete PCI name as the part of representor port name cause the
> app crash, because the name is too long.
>
> Fix this by reduce the length, using only the parts after the first ':'.
>
> To keep unification, also use the same approach for the n
There is currently no way to know which libraries are optional.
Introduce a enable_libs option (close to what we have for drivers) so
that packagers or projects consuming DPDK can more easily select the
optional libraries that matter to them and disable other optional
libraries.
Signed-off-by: Dav
Rework deprecated libraries selection by introducing a new configuration
option.
This breaks existing configurations that were relying on disable_libs=''
for enabling deprecated libraries.
On the other hand, it will make enabling optional libraries more
straightforward by taking the deprecated lib
kni dependency to IOVA configuration does not need to be expressed in
the top level lib/meson.build file.
Signed-off-by: David Marchand
---
lib/kni/meson.build | 5 +
lib/meson.build | 3 ---
2 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/lib/kni/meson.build b/lib/kni/mes
The enabled_libs variable is renamed and moved to the top level
meson.build file, for sake of consistency wrt to drivers and
applications similar variables.
Signed-off-by: David Marchand
---
app/test/meson.build | 2 +-
buildtools/chkincs/meson.build | 2 +-
lib/meson.build
This series is one implementation to try and please users who want to
select more easily which parts of DPDK are built.
It introduces a change in behavior for enabling deprecated libraries:
this series is aimed at the next release but sent early as a demo of
what changes are required.
If no stron
On 6/7/2023 2:02 PM, Ivan Malov wrote:
> On EF100 hardware, match-action engine (MAE) can be equipped
> with an assistance table for connection tracking (CT). In it,
> an entry key is a set of exact match fields: an EtherType, a
> pair of IP addresses, a L4 protocol ID and a pair of L4 port
> numbe
On 6/7/2023 2:02 PM, Ivan Malov wrote:
> From: Denis Pryazhennikov
>
> Pickup new FW interface definitions.
>
> Signed-off-by: Denis Pryazhennikov
> Reviewed-by: Ivan Malov
> Reviewed-by: Andy Moreton
<...>
> @@ -9127,15 +10514,13 @@
> /***/
> /* MC_CMD_DYN
On 6/7/2023 2:02 PM, Ivan Malov wrote:
> For this offload to work, the innermost pattern items must
> provide the full set of exact match criteria, which are as
> follows: EtherType, IP DST, IP SRC, TP protocol ID, TP DST
> and TP SRC, where the protocol types can be autodetected.
>
> The offload
On Wed, Jun 21, 2023 at 7:36 PM Thomas Monjalon wrote:
>
> Nobody knows how to build the feature.
> When the dependency "rxp_compiler" is found,
> the header file is not available:
>
> drivers/regex/cn9k/cn9k_regexdev_compiler.c:12:10: fatal error:
> rxp-compiler.h: No such file or directo
As reported by Ilya [1], unconditionally calling
rte_flow_get_restore_info() impacts an application performance for drivers
that do not provide this ops.
It could also impact processing of packets that require no call to
rte_flow_get_restore_info() at all.
Register a dynamic mbuf flag when an appl
On 6/15/2023 4:27 AM, Chaoyong He wrote:
> From: James Hershaw
>
> The NFP PMD driver retrieves the linkspeed by reading the
> NFP_NET_CFG_STS_LINK register. This register is not set by all versions
> of the firmware. Changes in this patch include:
>
> Addition of listener function to deal with
QAT HW generates bytes of 0x00 digest, even when a digest of len 0 is
requested for NULL. This caused test failures when the test vector had
digest len 0, as the buffer has unexpected changed bytes.
By placing the digest into the cookie for NULL authentication,
the buffer remains unchanged as expe
On 6/21/2023 12:09 PM, Ivan Malov wrote:
> Hi Ferruh,
>
> Thank you so much for your review notes. You suggested to squash
> some pairs of "common/sfc_base/efx" and "net/sfc" patches, so
> that logical changes would be unified. I see your point.
>
> On the other hand, however, doing so would be r
Nobody knows how to build the feature.
When the dependency "rxp_compiler" is found,
the header file is not available:
drivers/regex/cn9k/cn9k_regexdev_compiler.c:12:10: fatal error:
rxp-compiler.h: No such file or directory
It seems that it depends on a proprietay library.
According to wh
On 6/15/2023 2:54 AM, Chaoyong He wrote:
> From: Peng Zhang
>
> When adding support for multiple processes the number of secondary process
> on the PF was limited by the way NFP access registers on the device over
> the CPP interface.
>
> The limitation on the CPP interface appears to have been
> -Original Message-
> From: Slava Ovsiienko
> Sent: Tuesday, June 20, 2023 7:44 PM
>
> > -Original Message-
> > From: David Marchand
> > Sent: Tuesday, June 20, 2023 2:10 PM
> > To: dev@dpdk.org
> > Cc: NBU-Contact-Thomas Monjalon (EXTERNAL) ;
> > i.maxim...@ovn.org; Ali Alnub
On 6/15/2023 2:30 AM, Chaoyong He wrote:
> The VF PMD can not get a valid 'cpp' data filed in the 'struct
> nfp_net_hw', the value of which will be NULL always. This cause any
> logic which use of 'cpp' will make the app crash.
>
It is not fixing getting FW version, but fixing crash by removing F
On 2/9/2023 6:36 AM, Chengwen Feng wrote:
The memarea library is an allocator of variable-size object which based
on a memory region.
This patch provides rte_memarea_create() and rte_memarea_destroy() API.
Signed-off-by: Chengwen Feng
Reviewed-by: Dongdong Liu
Acked-by: Morten Brørup
---
I haven't followed this discussion in detail, so this might have been mentioned
already...
Omitting deprecated libs when building will not omit deprecated functions in
other libs.
In other words: Omitting deprecated libs only achieves part of what we really
want to achieve (omitting everything
On 6/13/2023 5:58 PM, Viacheslav Ovsiienko wrote:
> The "save_trace" CLI command is added to trigger
> saving the trace dumps to the trace directory.
>
Hi Viacheslav,
Trace is already saved when dpdk application terminated, I guess this is
to save the trace before exiting the application, what i
Hi Ferruh,
Thank you so much for your review notes. You suggested to squash
some pairs of "common/sfc_base/efx" and "net/sfc" patches, so
that logical changes would be unified. I see your point.
On the other hand, however, doing so would be rather unusual
from our internal process standpoint. Ty
On 2/9/2023 6:36 AM, Chengwen Feng wrote:
This patch supports memarea test of rte_memarea_create() and
rte_memarea_destroy() API.
Signed-off-by: Chengwen Feng
Reviewed-by: Dongdong Liu
Acked-by: Morten Brørup
---
Hi,
+
+static int
+test_memarea(void)
+{
+ test_memarea_prepare();
+
+
On 2/9/2023 6:36 AM, Chengwen Feng wrote:
The memarea library is an allocator of variable-size object which based
on a memory region.
This patch provides rte_memarea_create() and rte_memarea_destroy() API.
Signed-off-by: Chengwen Feng
Reviewed-by: Dongdong Liu
Acked-by: Morten Brørup
---
On Wed, Jun 21, 2023 at 11:57:18AM +0200, David Marchand wrote:
> Having an explicit link to the main repository documentation breaks
> other repositories documentation (like for LTS releases).
>
> Fixes: 7fc6ae50369d ("net/af_xdp: support CNI Integration")
> Cc: sta...@dpdk.org
>
> Signed-off-by
On Wed, Jun 21, 2023 at 10:09:20AM +0200, David Marchand wrote:
> Putting links to doc.dpdk.org/guides in our documentation should be
> avoided because it makes the documentation point at the current state of
> the DPDK tree regardless of the version it is generated against.
>
> Note: it should be
On Wed, Jun 21, 2023 at 11:54:31AM +0200, David Marchand wrote:
> On Tue, Jun 20, 2023 at 5:05 PM Bruce Richardson
> wrote:
> >
> > On Tue, Jun 20, 2023 at 04:33:15PM +0200, Thomas Monjalon wrote:
> > > 20/06/2023 11:03, Bruce Richardson:
> > > > On Tue, Jun 20, 2023 at 10:48:50AM +0200, David Mar
> -Original Message-
> From: Volodymyr Fialko
> Sent: Wednesday, June 21, 2023 11:01 AM
> To: dev@dpdk.org; Dumitrescu, Cristian
> Cc: jer...@marvell.com; ano...@marvell.com; tho...@monjalon.net;
> Volodymyr Fialko
> Subject: [PATCH v3] bitmap: add scan from offset function
>
> Curre
Currently, in the case when we search for a bit set after a particular
value, the bitmap has to be scanned from the beginning and
rte_bitmap_scan() has to be called multiple times until we hit the value.
Add a new rte_bitmap_scan_from_offset() function to initialize scan
state at the given offset
Having an explicit link to the main repository documentation breaks
other repositories documentation (like for LTS releases).
Besides, the tables numerotation may change so add an explicit anchor to
those tables and refer to them.
Fixes: b845c295cd13 ("doc: update Broadcom bnxt guide")
Cc: sta...
Having an explicit link to the main repository documentation breaks
other repositories documentation (like for LTS releases).
Fixes: 7fc6ae50369d ("net/af_xdp: support CNI Integration")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
---
doc/guides/howto/af_xdp_cni.rst | 6 ++
1 file chan
On Tue, Jun 20, 2023 at 5:05 PM Bruce Richardson
wrote:
>
> On Tue, Jun 20, 2023 at 04:33:15PM +0200, Thomas Monjalon wrote:
> > 20/06/2023 11:03, Bruce Richardson:
> > > On Tue, Jun 20, 2023 at 10:48:50AM +0200, David Marchand wrote:
> > > > On Tue, Jun 20, 2023 at 10:45 AM Bruce Richardson
> > >
This implementation has some problem, the function did not work properly,
and it broke the mark action in FDir filter, so revert this commit.
Fixes: e24f43f2c824 ("net/ice: support flow mark action in switch")
Signed-off-by: Zhichao Zeng
---
drivers/net/ice/ice_switch_filter.c | 9 -
1 f
Hi Jiayu,
Thanks for the comments. I have replied inline below. Will address the
review comments, but I think this would be a good patch to have in general.
Please let me know your thoughts.
On Tue, Jun 20, 2023 at 1:06 PM Hu, Jiayu wrote:
> Hi Kumara,
>
> Please see replies inline.
>
> Thanks,
Add support for IPv6 GRO for TCP packets
Signed-off-by: Kumara Parameshwaran
---
v1:
* Changes to support GRO for TCP/ipv6 packets. This does not include
vxlan changes.
* The GRO is performed only for ipv6 packets that does not contain
extension headers.
Refactor the existing code to reuse data strutures and functions
to add support for IPv6 GRO
Signed-off-by: Kumara Parameshwaran
---
v1:
* Changes to support GRO for TCP/ipv6 packets. This does not include
vxlan changes.
* The GRO is performed only for ipv6 packets that
On Thu, Jun 15, 2023 at 9:08 AM Fengnan Chang
wrote:
>
> When there is not enough space to memsegs, we should prompt
> which configuration should be modified instead of printing
> some numbers.
>
> Signed-off-by: Fengnan Chang
This lgtm, but I have a question:
$ git grep 'RTE_STR(RTE_MAX'
app/t
Signed-off-by: Kumara Parameshwaran
---
v1:
* Changes to support GRO for TCP/ipv6 packets. This does not include
vxlan changes.
* The GRO is performed only for ipv6 packets that does not contain
extension headers.
* The logic for the TCP coalescing rem
Refactor the existing code to reuse data strutures and functions
to add support for IPv6 GRO
Signed-off-by: Kumara Parameshwaran
---
v1:
* Changes to support GRO for TCP/ipv6 packets. This does not include
vxlan changes.
* The GRO is performed only for ipv6 packets that
On Fri, May 12, 2023 at 4:30 PM Arnaud Fiorini
wrote:
>
> The tracepoints added are used to track lcore role and status,
> as well as service mapping and service runstates. These
> tracepoints are then used in analyses in Trace Compass.
>
> Signed-off-by: Arnaud Fiorini
Acked-by: Harry van Haaren
Hi,
When I create bond4 with ConnectX-4 , mlx_flow_create return error;
When bond_port is opened, slave_port will be closed to sync bond_port
config
??bond_ethdev_start -> slave_configure->rte_eth_dev_stop??
?6?7slave_configure(struct rte_eth_dev *bonded_eth_dev,
Putting links to doc.dpdk.org/guides in our documentation should be
avoided because it makes the documentation point at the current state of
the DPDK tree regardless of the version it is generated against.
Note: it should be acceptable to mention previous versions of DPDK (like
in old release note
This patch fixes duplicate VF reset done check in dev_reset with large VF,
which cause some errors when starting testpmd with large VF.
Fixes: 7a93cd3575eb ("net/iavf: add VF reset check")
Signed-off-by: Zhichao Zeng
---
drivers/net/iavf/iavf_ethdev.c | 20 +++-
1 file changed, 1
> -Original Message-
> From: David Marchand
> Sent: Tuesday, June 20, 2023 2:10 PM
> To: dev@dpdk.org
> Cc: NBU-Contact-Thomas Monjalon (EXTERNAL) ;
> i.maxim...@ovn.org; Ali Alnubani ; Aman Singh
> ; Yuying Zhang ;
> Matan Azrad ; Slava Ovsiienko
> ; Ori Kam ; Suanming Mou
> ; David Chris
Hello,
On Tue, Jun 20, 2023 at 6:44 PM Slava Ovsiienko wrote:
> > As reported by Ilya [1], unconditionally calling
> > rte_flow_get_restore_info() impacts an application performance for drivers
> > that do not provide this ops.
> > It could also impact processing of packets that require no call t
When a Mellanox NIC is attached to a bus on a numa
socket, it tries to allocate rte memory in that socket.
If hugepages are not configured/available on that rte socket
mlx5_common_pci_probe fails with ENOMEM.
In this patch, a memflag MLX5_MEM_FALLBACK_ANY_SOCKET is
introduced which when set on mlx
21/06/2023 08:44, Ruifeng Wang:
> + * @note
> + * This function is deprecated. It adds complexity to the memory model
> + * used by this project. C11 memory model should always be used.
> + * rte_atomic_thread_fence() should be used instead.
> */
> static inline void rte_smp_mb(void);
I thin
60 matches
Mail list logo
