Setting dequeue packet count to max of MAX_PKT_BURST
size instead of MAX_PKTS.
Dequeue from cryptodev is called with MAX_PKTS but
routing functions allocate hop/dst_ip arrays of
size MAX_PKT_BURST. This can corrupt stack causing
stack smashing error when more than MAX_PKT_BURST
packets are returne
On 2024-09-12 17:11, Jerin Jacob wrote:
On Thu, Sep 12, 2024 at 6:50 PM Mattias Rönnblom wrote:
On 2024-09-12 15:09, Jerin Jacob wrote:
On Thu, Sep 12, 2024 at 2:34 PM Mattias Rönnblom
wrote:
Add basic micro benchmark for lcore variables, in an attempt to assure
that the overhead isn't sig
From: Jun Yang
FAF parser extracts are used to identify protocol type
instead of extracts of previous protocol' type.
FAF starts from offset 2 to include user defined flags which
will be used for soft protocol distribution.
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_flow.c | 475 +
This patch adds support for customizing firmware search path for
DDP package like the kernel behavior, it will read the search path
from "/sys/module/firmware_class/parameters/path", and try to load
DDP package.
Also, updates documentation for loading the DDP package in ice.rst.
Signed-off-by: Zh
From: Jun Yang
Support multiple extractions as well as hardware descriptions
instead of hard code.
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_ethdev.h | 1 +
drivers/net/dpaa2/dpaa2_mux.c| 395 ---
drivers/net/dpaa2/dpaa2_parse_dump.h | 2 +
dri
From: Vanshika Shukla
This patch adds the support for DPDMUX_METHOD_C_VLAN_MAC method
which implements DPDMUX based on C-VLAN and MAC address.
Signed-off-by: Vanshika Shukla
---
drivers/net/dpaa2/dpaa2_mux.c | 2 +-
drivers/net/dpaa2/mc/fsl_dpdmux.h | 16
2 files changed,
From: Vanshika Shukla
This patch adds support for adding rules in DPDMUX
to split VLAN traffic based on VLAN ids.
Signed-off-by: Vanshika Shukla
---
drivers/net/dpaa2/dpaa2_mux.c | 20
1 file changed, 20 insertions(+)
diff --git a/drivers/net/dpaa2/dpaa2_mux.c b/drivers/n
From: Jun Yang
Export API in rte_pmd_dpaa2.h
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_ethdev.c | 24
drivers/net/dpaa2/dpaa2_ethdev.h | 4
drivers/net/dpaa2/rte_pmd_dpaa2.h | 3 +++
drivers/net/dpaa2/version.map | 1 +
4 files changed, 32 inser
From: Apeksha Gupta
store drop priority in mbuf from fd.
Signed-off-by: Sachin Saxena
Signed-off-by: Apeksha Gupta
---
drivers/bus/fslmc/portal/dpaa2_hw_pvt.h | 1 +
drivers/net/dpaa2/dpaa2_rxtx.c | 1 +
2 files changed, 2 insertions(+)
diff --git a/drivers/bus/fslmc/portal/dpaa2_hw
From: Sachin Saxena
compatible with MC v10.36 or later
Signed-off-by: Sachin Saxena
---
drivers/net/dpaa2/dpaa2_mux.c | 11 +--
1 file changed, 5 insertions(+), 6 deletions(-)
diff --git a/drivers/net/dpaa2/dpaa2_mux.c b/drivers/net/dpaa2/dpaa2_mux.c
index 4390be9789..3c9e155b23 10064
From: Jun Yang
Convert VA to IOVA and check IOVA before sending parameter
to MC. Invalid IOVA of parameter sent to MC will cause system
stuck and not be recovered unless power reset.
IOVA is not checked in data path because:
1) MC is not involved and error can be recovered.
2) IOVA check impacts
From: Jun Yang
Add flow supported by soft parser to verification list.
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_flow.c | 86 --
1 file changed, 52 insertions(+), 34 deletions(-)
diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_fl
From: Gagandeep Singh
Add software based taildrop support.
Signed-off-by: Gagandeep Singh
---
drivers/bus/fslmc/portal/dpaa2_hw_pvt.h | 2 +-
drivers/net/dpaa2/dpaa2_rxtx.c | 24 +++-
2 files changed, 24 insertions(+), 2 deletions(-)
diff --git a/drivers/bus/fslm
From: Gagandeep Singh
driver was reserving memory in an array for 8 queues only,
but it can support many more queues configuration.
This patch fixes the memory corruption issue by defining the
queue array with correct size.
Fixes: 72100f0dee21 ("net/dpaa2: support level 2 in traffic management"
From: Jun Yang
Access sp instruction area to check if sp is loaded.
Signed-off-by: Jun Yang
Signed-off-by: Vanshika Shukla
---
drivers/net/dpaa2/dpaa2_ethdev.c | 4 ++
drivers/net/dpaa2/dpaa2_ethdev.h | 2 +
drivers/net/dpaa2/dpaa2_flow.c | 88
3 files cha
From: Jun Yang
Support AH/ESP flow with SPI field.
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_flow.c | 528 -
1 file changed, 385 insertions(+), 143 deletions(-)
diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 7207547
From: Jun Yang
Control flow by protocols inside tunnel.
The tunnel flow items applied by application are in order from
outer to inner. The inner items start from tunnel item, something
like vxlan, GRE etc.
For example:
flow create 0 ingress pattern ipv4 / vxlan / ipv6 / end
actions pf /
From: Jun Yang
Soft parser extracts ECPRI header and message to specified
areas of parser result.
Flow is classified according to the ECPRI extracts from praser result.
This implementation supports ECPRI over ethernet/vlan/UDP and various
types/messages combinations.
Signed-off-by: Jun Yang
---
From: Jun Yang
Configure gtp flow to support RSS and FS.
Check FAF of parser result to identify GTP frame.
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_flow.c | 170 ++---
1 file changed, 137 insertions(+), 33 deletions(-)
diff --git a/drivers/net/dpaa2/dpaa
From: Jun Yang
Support combination of RAW extract and header extracts.
RAW extract can start from any absolute offset.
TBD: relative offset support.
To support relative offset of previous L3 protocol item,
extracts should be expanded to identify if the frame is:
vlan or none-vlan.
To support re
From: Jun Yang
Extracts from vxlan header for distribution.
The vxlan header is set by soft parser code in
soft parser context located from offset 43 of parser results:
vxlan protocol is identified by vxlan bit of frame attribut flags.
The parser result extracts are added for this functionalit
From: Jun Yang
export DPAA2_PRINT_RX_PARSER_RESULT=1 is used to dump
RX parser result and frame attribute flags generated by
hardware parser and soft parser.
The parser results are converted to big endian described in RM.
The areas set by soft parser are dump as well.
Signed-off-by: Jun Yang
--
From: Jun Yang
Declare rte_fslmc_vfio_mem_dmamap and rte_fslmc_vfio_mem_dmaunmap
in bus_fslmc_driver.h for external usage.
Signed-off-by: Jun Yang
---
drivers/bus/fslmc/bus_fslmc_driver.h | 7 ++-
drivers/bus/fslmc/fslmc_bus.c| 2 +-
drivers/bus/fslmc/fslmc_vfio.c
From: Rohit Raj
Remove miss flow id macro name to DPNI_FS_MISS_DROP since its
conflicting with enum. Also, set default miss flow id to 0.
Signed-off-by: Rohit Raj
---
drivers/net/dpaa2/dpaa2_flow.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/dpaa2/dpaa2
From: Rohit Raj
Change qbman_eq_desc name to avoid redefining same variable.
Signed-off-by: Rohit Raj
---
drivers/bus/fslmc/qbman/qbman_portal.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/bus/fslmc/qbman/qbman_portal.c
b/drivers/bus/fslmc/qbman/qbman_porta
From: Rohit Raj
Since error code was being set to 0 in case of error which is a valid
fd, it caused memory leak issue.
This issue have been fixed by changing zero to a valid non fd error.
CID: 26661848
Signed-off-by: Rohit Raj
---
drivers/bus/fslmc/fslmc_vfio.c | 19 ++-
1 file
From: Rohit Raj
Fix Issues reported by coverity (NXP Internal Coverity)
Signed-off-by: Rohit Raj
---
drivers/bus/fslmc/qbman/qbman_debug.c | 49 +--
1 file changed, 32 insertions(+), 17 deletions(-)
diff --git a/drivers/bus/fslmc/qbman/qbman_debug.c
b/drivers/bus/fslm
From: Jun Yang
Create dpaa2 device with object instead of object ID.
Assign each dpaa2 object with it's container.
Signed-off-by: Jun Yang
---
drivers/bus/fslmc/bus_fslmc_driver.h | 39
drivers/bus/fslmc/fslmc_vfio.c | 3 +-
drivers/bus/fslmc/portal/dpaa
From: Jun Yang
IOVA mode should not be configured with CFLAGS because
1) User can perform "--iova-mode" to configure IOVA.
2) IOVA mode is determined by negotiation between multiple devices.
Eal is in VA mode only when all devices support VA mode.
Hence:
1) Remove RTE_LIBRTE_DPAA2_USE_PHYS_IO
From: Jun Yang
Remove unused GITS translator VFIO mapping.
Signed-off-by: Jun Yang
---
drivers/bus/fslmc/fslmc_vfio.c | 50 --
1 file changed, 50 deletions(-)
diff --git a/drivers/bus/fslmc/fslmc_vfio.c b/drivers/bus/fslmc/fslmc_vfio.c
index 31011b8532..f5d398c
From: Jun Yang
MC VFIO is not registered into RTE VFIO. Primary process registers
MC vfio mp action for secondary process to request.
VFIO/Container handlers are provided via CMSG.
Primary process is responsible to connect MC VFIO group to container.
In addition, MC VFIO code is refactored accor
From: Rohit Raj
Free vfio_group_fd if add group fails to avoid ersource leak
NXP coverity-id: 26661846
Signed-off-by: Rohit Raj
---
drivers/bus/fslmc/fslmc_vfio.c | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/drivers/bus/fslmc/fslmc_vfio.c b/drivers/bus/fslmc/fslmc_v
From: Jun Yang
Get vfio group fd directly from file system instead of
from RTE API to avoid conflicting with PCIe VFIO.
FSL MC VFIO should have it's own logic which doe NOT depend on
RTE VFIO.
Signed-off-by: Jun Yang
---
drivers/bus/fslmc/fslmc_vfio.c | 88 ++
d
From: Jun Yang
Ignore reserved bits of BMan acquire response number.
Signed-off-by: Jun Yang
---
drivers/bus/fslmc/qbman/qbman_portal.c | 26 --
1 file changed, 16 insertions(+), 10 deletions(-)
diff --git a/drivers/bus/fslmc/qbman/qbman_portal.c
b/drivers/bus/fslmc/q
From: Gagandeep Singh
This patch upgrades the MC version compaitbility to 10.37
Signed-off-by: Gagandeep Singh
Signed-off-by: Apeksha Gupta
---
doc/guides/platform/dpaa2.rst | 4 +-
drivers/bus/fslmc/mc/dpio.c | 94 -
drivers/bus/fslmc/mc/fsl_dpcon.h
From: Jun Yang
This patch add support to check the DPAA platform type from
the applications.
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_ethdev.c | 16 +---
drivers/net/dpaa2/dpaa2_flow.c| 5 ++---
drivers/net/dpaa2/rte_pmd_dpaa2.h | 4
drivers/net/dpaa2/version
From: Brick Yang
If SFP module is not connected to the port and flow control is
configured using flow control API, link will show DOWN even after
connecting the SFP module and fiber cable.
This issue cannot be reproduced if only SFP module is connected and
fiber cable is disconnected before conf
From: Rohit Raj
This patch add support to update the duplex value along with
link status and link speed after setting the link UP.
Signed-off-by: Rohit Raj
---
drivers/net/dpaa2/dpaa2_ethdev.c | 15 ++-
1 file changed, 10 insertions(+), 5 deletions(-)
diff --git a/drivers/net/dpaa
From: Vanshika Shukla
This patch adds the support for DPDMUX_METHOD_C_VLAN_MAC method
which implements DPDMUX based on C-VLAN and MAC address.
Signed-off-by: Vanshika Shukla
---
drivers/net/dpaa2/dpaa2_mux.c | 59 +--
drivers/net/dpaa2/mc/fsl_dpdmux.h | 18 +
From: Rohit Raj
Add rte_fslmc_close API to close all the DPAA2 devices while
closing the DPDK application.
Signed-off-by: Rohit Raj
---
drivers/bus/fslmc/bus_fslmc_driver.h | 3 +
drivers/bus/fslmc/fslmc_bus.c| 13
drivers/bus/fslmc/fslmc_vfio.c | 87 +++
From: Hemant Agrawal
This patch add supports to dump dpdmux counters as they are required
to identify the reasons for packet drop in dpdmux.
Signed-off-by: Hemant Agrawal
Signed-off-by: Rohit Raj
---
drivers/net/dpaa2/dpaa2_mux.c | 84 +++
drivers/net/dpaa2/rte
From: Vanshika Shukla
This patch adds PTP one-step timestamping support.
dpni_set_single_step_cfg() MC API is utilized with offset provided
to insert correction time on frame.
Signed-off-by: Vanshika Shukla
---
drivers/net/dpaa2/dpaa2_ethdev.c | 61 +++
drivers/net
From: Apeksha Gupta
This patch add proper debug info for check information of
max-pkt-len and configured params.
also store MTU
Signed-off-by: Apeksha Gupta
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_ethdev.c | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/d
From: Hemant Agrawal
This patch marks dpcon_close API as internal symbol and
also adds it into version map file
Signed-off-by: Hemant Agrawal
---
drivers/bus/fslmc/mc/fsl_dpcon.h | 3 ++-
drivers/bus/fslmc/version.map| 1 +
2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/driv
From: Jun Yang
Create TX SG pool only for primary process and lookup
this pool in secondary process.
Signed-off-by: Jun Yang
---
drivers/net/dpaa2/dpaa2_ethdev.c | 46 +++-
1 file changed, 33 insertions(+), 13 deletions(-)
diff --git a/drivers/net/dpaa2/dpaa2_ethde
From: Vanshika Shukla
This series includes:
-> Fixes and enhancements for NXP DPAA2 drivers.
-> Upgrade with MC version 10.37
-> Enhancements in DPDMUX code
-> Fixes for coverity issues reported
Apeksha Gupta (2):
net/dpaa2: add proper MTU debugging print
net/dpaa2: store drop priority in mb
On 2024/9/13 4:26, Stephen Hemminger wrote
> +
> +Finalizing
> +--
> +
> +Once the driver has been upstreamed, the author has
> +a responsibility to the community to maintain it.
> +
> +This includes the public test report. Authors must send a public
> +test report after the first upstreami
On 2024/9/12 17:18, fengchengwen wrote:
> On 2024/9/12 16:19, WanRenyong wrote:
>> On 2024/9/12 13:50, Stephen Hemminger wrote:
>>> On Thu, 12 Sep 2024 12:14:08 +0800
>>> "WanRenyong" wrote:
>>>
>
>> +};
>> +
> Does this device driver depend on some upstr
Yes, it depends o
#
September 12, 2024
Attendees
* Patrick Robb
* Paul Szczepanek
* Alex Chapman
* Luca Vizzarro
* Jeremey Spewock
* Juraj Linkeš
#
Minutes
===
> > @classmethod
> > def from_dict(cls, d: BuildTargetConfigDict) -> Self:
> > r"""A convenience method that processes the inputs before creating
> > an instance.
> >
> > -`arch`, `os`, `cpu` and `compiler` are converted to
> > :class:`Enum`\s and
> > -`name` i
On Tue, Sep 10, 2024 at 7:30 AM Juraj Linkeš wrote:
>
> The subject line should be all lowercase (except for abbreviations and
> maybe some other exceptions).
Ack.
>
> On 5. 7. 2024 19:13, Nicholas Pratte wrote:
> > Remove the list of devices from the schema, as these are unuesed.
>
> Typo: unus
From: Nandini Persad
This document was created to assist contributors in creating DPDK drivers
and provides suggestions and guidelines on how to upstream effectively.
Co-authored-by: Ferruh Yigit
Co-authored-by: Thomas Monjalon
Signed-off-by: Nandini Persad
Reviewed-by: Stephen Hemminger
---
21/08/2024 17:02, Juraj Linkeš:
> +if 'dts' in src:
> +os.environ['DTS_BUILD'] = "y"
That's more precisely "DTS doc build".
I think the variable name DTS_BUILD may be confusing.
[...]
> --- /dev/null
> +++ b/buildtools/get-dts-runtime-deps.py
> @@ -0,0 +1,95 @@
> +#!/usr/bin/env python3
> +#
The mac address filter test suite, whose test cases are based on old
DTS's test cases, has been refactored to interface with the new DTS
framework.
In porting over this test suite into the new framework, some
adjustments were made, namely in the EAL and TestPMD parameter provided
before executing
New methods have been added to TestPMDShell in order to produce the mac
filter's individual test cases:
- set_mac_addr
- set_multicast_mac_addr
set_mac_addr and set_multicast_addr were created for the mac filter test
suite, enabling users to both add or remove mac and multicast
addresses based o
v5:
* Addressed some concerns on docstrings based on previous reviews.
Nicholas Pratte (2):
dts: add methods for setting mac and multicast addresses
dts: mac filter test suite refactored for new dts
dts/framework/config/conf_yaml_schema.json| 3 +-
dts/framework/remote_session/testpm
> >
> > For instance, in Neoverse N1 can disable the use of SLC as LLC (a BIOS
> > setting) If SLC is not used as LLC, then your script would report the
> > unified L2
> as an LLC.
>
> Does `disabling SLC as LLC` disable L3? I think not, and what you are
> implying is
> the ` ls -d /sys/bus/cpu
On 2024-09-12 15:30, Bruce Richardson wrote:
On Thu, Sep 12, 2024 at 01:59:34PM +0200, Mattias Rönnblom wrote:
On 2024-09-12 13:17, Varghese, Vipin wrote:
[AMD Official Use Only - AMD Internal Distribution Only]
Thank you Mattias for the information, as shared by in the reply
with
Anatoly w
On Mon, Sep 02, 2024 at 10:54:12AM +0100, Anatoly Burakov wrote:
> Update the i40e base driver to the latest version of the snapshot.
>
Series-acked-by: Bruce Richardson
On apply I tagged a few more patches as bugfixes (including fixes line) for
backport to the older LTS releases.
Applied to dp
On Thu, 12 Sep 2024 14:12:55 +0200
Mattias Rönnblom wrote:
> On 2024-09-12 13:23, Varghese, Vipin wrote:
> > [Public]
> >
> > Snipped
> >>
> >>
> >> To to be clear; it's something like this I think of when I say "DOM-style"
> >> API.
> >>
> >> #ifndef RTE_HWTOPO_H
> >> #define RTE_HWTOPO_H
On Thu, Sep 12, 2024 at 7:11 PM Morten Brørup
wrote:
>
> > From: Jerin Jacob [mailto:jerinjac...@gmail.com]
> > Sent: Thursday, 12 September 2024 15.17
> >
> > On Thu, Sep 12, 2024 at 2:40 PM Morten Brørup
> > wrote:
> > >
> > > > +#define LCORE_BUFFER_SIZE (RTE_MAX_LCORE_VAR * RTE_MAX_LCORE)
>
On Thu, Sep 12, 2024 at 6:50 PM Mattias Rönnblom wrote:
>
> On 2024-09-12 15:09, Jerin Jacob wrote:
> > On Thu, Sep 12, 2024 at 2:34 PM Mattias Rönnblom
> > wrote:
> >>
> >> Add basic micro benchmark for lcore variables, in an attempt to assure
> >> that the overhead isn't significantly greater t
> -Original Message-
> From: Ferruh Yigit
> Sent: Thursday, September 12, 2024 11:20 AM
> To: Tyler Retzlaff ; Stephen Hemminger
> ; Brandes, Shai
> Cc: dev@dpdk.org; Schmeilin, Evgeny ; Beider, Ron
> ; Bernstein, Amit ;
> Atrash, Wajeeh ; Chauskin, Igor
> ; Michal Krawczyk ; Artur Roje
12/09/2024 10:26, David Marchand:
> Do not stop at the first file that got warnings.
>
> Signed-off-by: David Marchand
It is always better to get all issues in one pass.
Acked-by: Thomas Monjalon
12/09/2024 10:26, David Marchand:
> --- a/devtools/check-forbidden-tokens.awk
> +++ b/devtools/check-forbidden-tokens.awk
> @@ -32,14 +32,11 @@ BEGIN {
> for (i in deny_expr) {
> forbidden_added = "^\\+.*" deny_expr[i];
> forbidden_removed="
On 9/12/24 15:16, Ferruh Yigit wrote:
On 9/12/2024 1:08 PM, Maxime Coquelin wrote:
Hi Ferruh,
On 9/12/24 10:30, Ferruh Yigit wrote:
On 9/11/2024 8:32 PM, Maxime Coquelin wrote:
This series enables importing Linux Kernel uAPI headers
into the DPDK repository. It aims at solving alignment
is
> From: Jerin Jacob [mailto:jerinjac...@gmail.com]
> Sent: Thursday, 12 September 2024 15.17
>
> On Thu, Sep 12, 2024 at 2:40 PM Morten Brørup
> wrote:
> >
> > > +#define LCORE_BUFFER_SIZE (RTE_MAX_LCORE_VAR * RTE_MAX_LCORE)
> >
> > Considering hugepages...
> >
> > Lcore variables may be allocate
Excellent. Will send the update in accordingly.
From: Ferruh Yigit
Sent: Thursday, September 12, 2024 6:37:35 AM
To: Nandini Persad
Cc: dev@dpdk.org ; Thomas Mojalon ; Stephen
Hemminger
Subject: Re: [PATCH v3] doc: add new driver guidelines
On 9/12/2024 2:18 PM
On 9/12/2024 2:18 PM, Nandini Persad wrote:
> I like the separation. I can include it in V4 and see, it would be
> helpful to know if it’s more or less confusing that way.
>
> For the prompt before each list, can we say something like “Avoid doing
> the following” and “Suggested actions” or somet
On Thu, Sep 12, 2024 at 01:59:34PM +0200, Mattias Rönnblom wrote:
> On 2024-09-12 13:17, Varghese, Vipin wrote:
> > [AMD Official Use Only - AMD Internal Distribution Only]
> >
> >
> > > Thank you Mattias for the information, as shared by in the reply
> > > with
> > > >> Anatoly we want
https://bugs.dpdk.org/show_bug.cgi?id=1541
Bug ID: 1541
Summary: Add support for VM SR-IOV testing
Product: DPDK
Version: unspecified
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: normal
Pr
On 2024-09-12 15:09, Jerin Jacob wrote:
On Thu, Sep 12, 2024 at 2:34 PM Mattias Rönnblom
wrote:
Add basic micro benchmark for lcore variables, in an attempt to assure
that the overhead isn't significantly greater than alternative
approaches, in scenarios where the benefits aren't expected to s
On 2024-09-12 11:17, Bruce Richardson wrote:
On Thu, Sep 12, 2024 at 02:19:07AM +, Varghese, Vipin wrote:
[Public]
> > > >
> > > >
> > > >>>
> > > >>>
> > > >>> Thank you Mattias for the comments and question, please let
me
> > > >>> try to exp
I like the separation. I can include it in V4 and see, it would be helpful to
know if it’s more or less confusing that way.
For the prompt before each list, can we say something like “Avoid doing the
following” and “Suggested actions” or something a little better grammatically.
We could also ju
On Thu, Sep 12, 2024 at 2:40 PM Morten Brørup
wrote:
>
> > +#define LCORE_BUFFER_SIZE (RTE_MAX_LCORE_VAR * RTE_MAX_LCORE)
>
> Considering hugepages...
>
> Lcore variables may be allocated before DPDK's memory allocator
> (rte_malloc()) is ready, so rte_malloc() cannot be used for lcore variables
On 9/12/2024 1:08 PM, Maxime Coquelin wrote:
> Hi Ferruh,
>
> On 9/12/24 10:30, Ferruh Yigit wrote:
>> On 9/11/2024 8:32 PM, Maxime Coquelin wrote:
>>> This series enables importing Linux Kernel uAPI headers
>>> into the DPDK repository. It aims at solving alignment
>>> issues between the build sy
On Thu, Sep 12, 2024 at 2:34 PM Mattias Rönnblom
wrote:
>
> Add basic micro benchmark for lcore variables, in an attempt to assure
> that the overhead isn't significantly greater than alternative
> approaches, in scenarios where the benefits aren't expected to show up
> (i.e., when plenty of cache
On 2024-09-12 11:39, Morten Brørup wrote:
+struct lcore_state {
+ uint64_t a;
+ uint64_t b;
+ uint64_t sum;
+};
+
+static __rte_always_inline void
+update(struct lcore_state *state)
+{
+ state->sum += state->a * state->b;
+}
+
+static RTE_DEFINE_PER_LCORE(struct lcore_stat
Looks good to me, and I don't understand it as triggering a second
compile on the TG node.
I wonder whether the docstring for set_up_build_target() makes it
sound like it does though?
Reviewed-by: Patrick Robb
This looks good, except for the fact that a method for sending a list
of packets has been added since you submitted this patch (which does
return the packet list, not that it matters for your application).
So, this will need to be resubmitted with your patch reformatted for
_adjust_addresses() to
On 2024-09-12 13:23, Varghese, Vipin wrote:
[Public]
Snipped
To to be clear; it's something like this I think of when I say "DOM-style" API.
#ifndef RTE_HWTOPO_H
#define RTE_HWTOPO_H
struct rte_hwtopo_node;
enum rte_hwtopo_node_type {
RTE_HWTOPO_NODE_TYPE_CPU_CORE,
RTE_HWTOPO_N
https://bugs.dpdk.org/show_bug.cgi?id=1540
Bug ID: 1540
Summary: update pw_maintainers_cli.py to apply some series to
next-dts
Product: DPDK
Version: unspecified
Hardware: All
OS: All
Status: UNC
Hi Ferruh,
On 9/12/24 10:30, Ferruh Yigit wrote:
On 9/11/2024 8:32 PM, Maxime Coquelin wrote:
This series enables importing Linux Kernel uAPI headers
into the DPDK repository. It aims at solving alignment
issues between the build system and the system applications
linked ot DPDK libraries are r
On 2024-09-12 13:17, Varghese, Vipin wrote:
[AMD Official Use Only - AMD Internal Distribution Only]
Thank you Mattias for the information, as shared by in the reply
with
>> Anatoly we want expose a new API `rte_get_next_lcore_ex` which
>> intakes a extra argument `u32 flags`.
T
[Public]
Snipped
> >
> >
> >Based on the discussions we agreed on sharing version-2 FRC for
> >extending API as `rte_get_next_lcore_extnd` with extra argument as
> >`flags`.
> >
> >As per my ideation, for the API ` rte_get_next_sibling_core`, the above
> >API can easily with f
[Public]
Snipped
>
>
> To to be clear; it's something like this I think of when I say "DOM-style"
> API.
>
> #ifndef RTE_HWTOPO_H
> #define RTE_HWTOPO_H
>
> struct rte_hwtopo_node;
>
> enum rte_hwtopo_node_type {
> RTE_HWTOPO_NODE_TYPE_CPU_CORE,
> RTE_HWTOPO_NODE_TYPE_CACHE,
> RT
[AMD Official Use Only - AMD Internal Distribution Only]
Hi Huisong,
Please find my response inline.
> -Original Message-
> From: lihuisong (C)
> Sent: Tuesday, August 27, 2024 1:51 PM
> To: Tummala, Sivaprasad
> Cc: dev@dpdk.org; david.h...@intel.com; anatoly.bura...@intel.com;
> radu
[AMD Official Use Only - AMD Internal Distribution Only]
> Thank you Mattias for the information, as shared by in the reply
> with
> >> Anatoly we want expose a new API `rte_get_next_lcore_ex` which
> >> intakes a extra argument `u32 flags`.
> The flags can be RTE_GET_LCORE_L1 (S
> +struct lcore_state {
> + uint64_t a;
> + uint64_t b;
> + uint64_t sum;
> +};
> +
> +static __rte_always_inline void
> +update(struct lcore_state *state)
> +{
> + state->sum += state->a * state->b;
> +}
> +
> +static RTE_DEFINE_PER_LCORE(struct lcore_state, tls_lcore_state);
> +
>
On 2024/9/12 16:19, WanRenyong wrote:
> On 2024/9/12 13:50, Stephen Hemminger wrote:
>> On Thu, 12 Sep 2024 12:14:08 +0800
>> "WanRenyong" wrote:
>>
> +};
> +
Does this device driver depend on some upstr
>>> Yes, it depends on linux kernel driver of the device.
>>>
>>> Hello,
> From: Mattias Rönnblom
> Sent: Monday, September 9, 2024 8:11 PM
> To: dev@dpdk.org
> Cc: hof...@lysator.liu.se ; Van Haaren, Harry
> ; Stefan Sundkvist
> ; Mattias Rönnblom
>
> Subject: [PATCH v2] service: extend service function call statistics
>
> Add two new per-service counters.
>
> RT
On Thu, Sep 12, 2024 at 02:19:07AM +, Varghese, Vipin wrote:
>[Public]
>
>
>
>
>
>> > > >
>
>> > > >
>
>> > > >>>
>
>> > > >>>
>
>> > > >>> Thank you Mattias for the comments and question, please let
>me
>
>> > > >>> try to explain the same below
>
> +#define LCORE_BUFFER_SIZE (RTE_MAX_LCORE_VAR * RTE_MAX_LCORE)
Considering hugepages...
Lcore variables may be allocated before DPDK's memory allocator (rte_malloc())
is ready, so rte_malloc() cannot be used for lcore variables.
And lcore variables are not usable (shared) for DPDK multi-proce
On 2024/9/12 14:54, Chaoyong He wrote:
> From: James Hershaw
>
> There is currently no means to test the .set_eeprom function callback of
> a given PMD in drivers/net/. This patch adds functionality to allow a
> user to set device eeprom from the testpmd cmdline.
>
> usage:
> testpmd> set port
On 2024-09-12 09:35, Jerin Jacob wrote:
On Wed, Sep 11, 2024 at 11:08 PM Mattias Rönnblom
wrote:
Add test suite to exercise the API.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
--
RFC v5:
* Adapt tests to reflect the removal of the GET() and SET() macros.
RFC v4:
* Check
This patch set introduces a new API for static
per-lcore id data allocation.
Please refer to the API documentation for both a
rationale for this new API, and a comparison to the alternatives
available.
The adoption of this API would affect many different DPDK modules, but
the author updated onl
Add functional test suite to exercise the API.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
--
RFC v5:
* Adapt tests to reflect the removal of the GET() and SET() macros.
RFC v4:
* Check all lcore id's values for all variables in the many variables
test case.
* Introduce tes
Replace static array of cache-aligned structs with an lcore variable,
to slightly benefit code simplicity and performance.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
--
RFC v6:
* Remove a now-redundant lcore variable value memset().
RFC v5:
* Fix lcore value pointer bug introdu
Replace static array of cache-aligned structs with an lcore variable,
to slightly benefit code simplicity and performance.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
--
RFC v3:
* Replace for loop with FOREACH macro.
---
lib/power/rte_power_pmd_mgmt.c | 34 ---
Add basic micro benchmark for lcore variables, in an attempt to assure
that the overhead isn't significantly greater than alternative
approaches, in scenarios where the benefits aren't expected to show up
(i.e., when plenty of cache is available compared to the working set
size of the per-lcore dat
1 - 100 of 127 matches
Mail list logo
