Extract data offset from individual mbuf's instead of mbuf0.
Fixes: 6771216c2f44 ("drivers: mark cnxk to support disabling IOVA as PA")
Signed-off-by: Nithin Dabilpuram
---
drivers/net/cnxk/cn10k_tx.h | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/cnxk/cn10
Add PMD API support to extract HW error code from mbuf's processed
by inline IPsec inbound.
Signed-off-by: Nithin Dabilpuram
---
drivers/net/cnxk/cn10k_ethdev_sec.c | 18 ++
drivers/net/cnxk/rte_pmd_cnxk.h | 17 +
drivers/net/cnxk/version.map| 6 +
Check for mbox error code in MAC address set.
Fixes: 313cc41830ec ("common/cnxk: support NIX MAC operations")
Signed-off-by: Nithin Dabilpuram
---
drivers/common/cnxk/roc_nix_mac.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/drivers/common/cnxk/roc_nix_mac.c
b/driver
On Mon, Oct 16, 2023 at 3:37 AM Dexia Li wrote:
>
> Thanks for your commit.
> It works for me.
>
Thanks for confirming, I marked this patch as rejected.
--
David Marchand
19/09/2023 10:12, Rongwei Liu:
> This patch adds testpmd app a runtime function to test the live
> migration API.
>
> testpmd> mlx5 set flow_engine []
> Flag is optional.
>
> Signed-off-by: Rongwei Liu
> Acked-by: Viacheslav Ovsiienko
> Acked-by: Ori Kam
[...]
> +Set Flow Engine Mode
> +~
HI
BR
Rongwei
> -Original Message-
> From: Thomas Monjalon
> Sent: Monday, October 16, 2023 16:20
> To: Rongwei Liu
> Cc: dev@dpdk.org; Matan Azrad ; Slava Ovsiienko
> ; Ori Kam ; Suanming Mou
> ; Raslan Darawsheh
> Subject: Re: [PATCH v3] net/mlx5: add test for live migration
>
> Ext
Fix the incorrect ingress and egress packet number calculation.
Fixes: 7514d76d407b ("net/idpf: add basic statistics")
Cc: sta...@dpdk.org
Signed-off-by: Simei Su
---
drivers/net/idpf/idpf_ethdev.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/net/idpf/idpf_e
Fix the incorrect ingress and egress packet number calculation.
Fixes: e3289d8fb63f ("net/cpfl: support basic statistics")
Cc: sta...@dpdk.org
Signed-off-by: Simei Su
---
drivers/net/cpfl/cpfl_ethdev.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/net/cpfl/cp
If a bonding port gets released, a link status alarm callback still
referenced the ethdev port that may be reused later.
Cancel this callback when stopping the port.
Bugzilla ID: 1301
Fixes: a45b288ef21a ("bond: support link status polling")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
---
https://bugs.dpdk.org/show_bug.cgi?id=1299
David Marchand (david.march...@redhat.com) changed:
What|Removed |Added
Resolution|--- |INVALID
On Fri, Sep 29, 2023 at 3:28 PM wrote:
>
> From: Sunil Kumar Kori
>
> It adds base framework to read a given .cli file as a command line
> parameter "-s".
>
> Example:
> # ./dpdk-graph -c 0xff -- -s ./app/graph/examples/dummy.cli
>
> Each .cli file will contain commands to configure different mo
On Fri, Sep 29, 2023 at 10:36 PM wrote:
>
> From: Sunil Kumar Kori
>
> It adds framework to initiate a telnet session with application.
>
> Some configurations and debug commands are exposed as runtime APIs.
> Those commands can be invoked using telnet session.
>
> Application initiates a telnet
On Fri, Sep 29, 2023 at 3:28 PM wrote:
>
> From: Sunil Kumar Kori
>
> It adds some helper functions to parse IPv4, IPv6 and MAC addresses
> string into respective datatype.
>
> Signed-off-by: Sunil Kumar Kori
> Signed-off-by: Rakesh Kudurumalla
Acked-by: Jerin Jacob
On Fri, Sep 29, 2023 at 3:28 PM wrote:
>
> From: Rakesh Kudurumalla
>
> It adds mempool module which will be creating mempools.
>
> Following commands are exposed:
> - mempool size buffers \
> cache numa
> - help mempool
>
> User will add this command in .cli file according to its
Provide driver flag which gives an option to map the cdx device
resource before probing the device driver. External driver can
use rte_cdx_map_device() and rte_cdx_unmap_device() APIs to map/
unmap device resource separately.
Signed-off-by: Abhijit Gangurde
---
v4:
- rte_cdx_map_device() and rt
This patch adds testpmd app a runtime function to test the live
migration API.
testpmd> mlx5 set flow_engine []
Flag is optional and in bitmap style.
Signed-off-by: Rongwei Liu
Acked-by: Viacheslav Ovsiienko
Acked-by: Ori Kam
---
doc/guides/nics/mlx5.rst| 14
drivers/net/ml
BR
Rongwei
> -Original Message-
> From: Rongwei Liu
> Sent: Monday, October 16, 2023 16:26
> To: Thomas Monjalon
> Cc: dev@dpdk.org; Matan Azrad ; Slava Ovsiienko
> ; Ori Kam ; Suanming Mou
> ; Raslan Darawsheh
> Subject: RE: [PATCH v3] net/mlx5: add test for live migration
>
> HI
>
16/10/2023 10:25, Rongwei Liu:
> From: Thomas Monjalon
> > 19/09/2023 10:12, Rongwei Liu:
> > > + testpmd> mlx5 set flow_engine []
> >
> > What are the flags?
> >
> The flag is optional and defined a as bitmap.
> For now, only one value is accepted: BIT(0).
> I don't have any idea to propaga
09/10/2023 13:15, Raslan Darawsheh:
> Series applied to next-net-mlx,
> Replied originally to V2
> Now to the correct version !
A release note should be added:
* **Updated NVIDIA mlx5 net driver.**
* Added support for Network Service Header (NSH) flow matching.
Hi
BR
Rongwei
> -Original Message-
> From: Thomas Monjalon
> Sent: Monday, October 16, 2023 17:27
> To: Rongwei Liu
> Cc: dev@dpdk.org; Matan Azrad ; Slava Ovsiienko
> ; Ori Kam ; Suanming Mou
> ; Raslan Darawsheh
> Subject: Re: [PATCH v3] net/mlx5: add test for live migration
>
> Ext
Improve log output by adding some newlines.
Signed-off-by: Tomasz Duszynski
Reviewed-by: Jerin Jacob Kollanukkaran
Tested-by: Jerin Jacob Kollanukkaran
---
drivers/raw/cnxk_gpio/cnxk_gpio.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/raw/cnxk_gpio/cnxk_g
GPIO PMD uses a mixture of standard sysfs attributes and custom
ioctl()s to control behaviour of respective GPIOs available in
the system.
This means that each userspace application, either running as a primary
or a secondary, should be able to control a set of distinct GPIOs.
In rare cases where
On 2023/10/16 16:47, David Marchand wrote:
> If a bonding port gets released, a link status alarm callback still
> referenced the ethdev port that may be reused later.
> Cancel this callback when stopping the port.
>
> Bugzilla ID: 1301
> Fixes: a45b288ef21a ("bond: support link status polling")
>
On 10/13/2023 9:37 PM, lon...@linuxonhyperv.com wrote:
> From: Long Li
>
> Add a missing "\n" to DP logs. This makes the logs correctly formatted.
>
> Signed-off-by: Long Li
>
Fixes: e2d3a3c060c4 ("net/mana: use datapath logging")
Cc: sta...@dpdk.org
Applied to dpdk-next-net/main, th
On 10/7/2023 9:36 AM, Jie Hai wrote:
> On 2023/9/28 21:15, Ferruh Yigit wrote:
>> On 9/28/2023 8:42 AM, Jie Hai wrote:
>>> The DPDK framework reports the queue status, which is stored in
>>> 'dev->data->tx_queue_state' and 'dev->data->rx_queue_state'.The
>>> state is currently maintained by the dri
Hi Thomas,
Gentle ping.
Could you please consider merging this series in RC1.
Thanks,
Amit Shukla
From: fengchengwen
Sent: Tuesday, October 10, 2023 7:10 AM
To: Amit Prakash Shukla ; Kevin Laatz
; Bruce Richardson
Cc: dev@dpdk.org; Jerin Jacob Kollanukkaran ;
conor.wa...@intel.com; Vamsi K
On 9/8/2023 12:28 PM, Jie Hai wrote:
> The DPDK framework reports the queue state, which is stored in
> dev->data->tx_queue_state and dev->data->rx_queue_state. The
> state is maintained by the driver. Users may determine whether
> a queue participates in packet forwarding based on the state,
> for
On 7/7/2023 7:53 PM, lon...@linuxonhyperv.com wrote:
> From: Long Li
>
> Set the queue state when queue is started/stopped
>
> Signed-off-by: Long Li
>
Applied to dpdk-next-net/main, thanks.
On 10/16/2023 12:51 PM, Ferruh Yigit wrote:
> On 9/8/2023 12:28 PM, Jie Hai wrote:
>> The DPDK framework reports the queue state, which is stored in
>> dev->data->tx_queue_state and dev->data->rx_queue_state. The
>> state is maintained by the driver. Users may determine whether
>> a queue participa
On 8/29/2023 7:29 PM, lon...@linuxonhyperv.com wrote:
> From: Long Li
>
> Set the queue state when queue is started/stopped.
>
> Signed-off-by: Long Li
> Acked-by: Stephen Hemminger
>
Applied to dpdk-next-net/main, thanks.
On 10/16/2023 11:33 AM, fengchengwen wrote:
> On 2023/10/16 16:47, David Marchand wrote:
>> If a bonding port gets released, a link status alarm callback still
>> referenced the ethdev port that may be reused later.
>> Cancel this callback when stopping the port.
>>
>> Bugzilla ID: 1301
>> Fixes: a
On 10/13/2023 8:46 AM, Chaoyong He wrote:
> From: Zerun Fu
>
> The former commit imports a register read operation into the data path
> logic, which will severely degrade performance.
> Fix this bug by moving the register read logic out of the data path.
> Read 'cap_extend' only once in the initi
On Fri, Sep 29, 2023 at 3:29 PM wrote:
>
> From: Sunil Kumar Kori
>
> It adds ethdev module to configure ethernet devices.
>
> Following commands are exposed:
> - ethdev rxq txq
> - ethdev mtu
> - ethdev promiscuous
> - ethdev show
> - ethdev stats
> - ethdev ip4 addr add netmas
On 10/10/2023 7:17 AM, Chaoyong He wrote:
> This patch series fix two converity issue 403098 and 403100.
>
> Long Wu (2):
> net/nfp: fix checking return value
> net/nfp: fix illegal memory accesses
>
>
Series applied to dpdk-next-net/main, thanks.
On 10/10/2023 7:09 AM, Chaoyong He wrote:
> From: James Hershaw
>
> The pointer to the beginning of the MAC stats counters for port 1 are
> incorrectly set as the pointer to the beginning of the port 0 MAC stats
> counters, plus the size of the MAC stats counters multiplied by the port
> number.
On 10/10/2023 7:06 AM, Chaoyong He wrote:
> From: Long Wu
>
> When sending a jumbo packet on NFDk the packet must be split between
> multiple descriptors. The first descriptor contains the packet header
> and is limited to NFDK_DESC_TX_DMA_LEN_HEAD bytes. If the packet is
> large, one or more pay
The DPDK commandline library is widely used by apps and examples within
DPDK, but it is not documented in our programmers guide and it requires
a lot of boilerplate code definitions in order to used. We can improve
this situation by creating a simple python script to automatically
generate the boil
The cmdline library was not documented in our programmers guide, so add
a new chapter on it. This chapter covers step-by-step how to use the
library, rather than focusing on the library internals. This complements
the existing cmdline example app document, providing more details on the
process of u
Provide a "dpdk-cmdline-gen.py" script for application developers to
quickly generate the boilerplate code necessary for using the cmdline
library.
Example of use:
The script takes an input file with a list of commands the user wants in
the app, where the parameter variables are tagged with the ty
To allow use of the DPDK python scripts (installed in $(prefix)/bin)
from within the makefiles of our examples, we need to export the PATH
variable with the location of our installed scripts from within our CI
scripts. This matches what is already done for other paths e.g. the
PKG_CONFIG_PATH varia
Use the dpdk-cmdline-gen script to autogenerate all the boilerplate
structs and defines for the commandline part of the app.
Signed-off-by: Bruce Richardson
---
examples/multi_process/simple_mp/Makefile | 12 +-
examples/multi_process/simple_mp/meson.build | 9 ++
.../multi_process/simpl
Use the dpdk-cmdline-gen script to autogenerate all the boilerplate
structs and defines for the commandline part of the app.
Signed-off-by: Bruce Richardson
---
Note: the original help text on some of the commands in this example
were not useful "this command do not handle any arguments". Ther
Use the dpdk-cmdline-gen script to autogenerate all the boilerplate
structs and defines for the commandline part of the app.
Signed-off-by: Bruce Richardson
---
examples/multi_process/hotplug_mp/Makefile| 12 +-
examples/multi_process/hotplug_mp/commands.c | 147 ++
example
Use the dpdk-cmdline-gen script to autogenerate all the boilerplate
structs and defines for the commandline part of the app.
Signed-off-by: Bruce Richardson
---
examples/vdpa/Makefile | 12 ++--
examples/vdpa/commands.list | 5 ++
examples/vdpa/main.c| 131 ++-
On Mon, Oct 16, 2023 at 06:50:47PM +0530, Jerin Jacob wrote:
> On Fri, Sep 29, 2023 at 3:29 PM wrote:
> >
> > From: Sunil Kumar Kori
> >
> > It adds ethdev module to configure ethernet devices.
> >
> > Following commands are exposed:
> > - ethdev rxq txq
> > - ethdev mtu
> > - ethdev pr
Hi Maxime,
Just a heads up that Hernan is going on paternity leave, I will be covering for
his series pending for 23.11. Ping me if there is any update required from
review that I may have missed.
Pending series in patchwork:
* SDK + doc update https://patches.dpdk.org/project/dpdk/list/?se
13/10/2023 17:57, Ferruh Yigit:
> Drivers start/stop device queues on port start/stop, but not all drivers
> update queue state accordingly.
>
> This becomes more visible if a specific queue stopped explicitly and
> port stopped/started later, in this case although all queues are
> started, the st
From: Pavan Nikhilesh
Update fc check to count SQEs instead of SQBs while waiting for
space.
Signed-off-by: Pavan Nikhilesh
---
drivers/event/cnxk/cn10k_tx_worker.h | 33 ++--
drivers/event/cnxk/cn9k_worker.h | 33 +---
2 files changed, 51 in
On 10/16/2023 3:59 PM, Thomas Monjalon wrote:
> 13/10/2023 17:57, Ferruh Yigit:
>> Drivers start/stop device queues on port start/stop, but not all drivers
>> update queue state accordingly.
>>
>> This becomes more visible if a specific queue stopped explicitly and
>> port stopped/started later, in
After calling rte_event_dev_info_get() the ".dev" field of the info
structure should have a pointer to the underlying device, allowing the
user to e.g. get the device name using using rte_dev_name(info.dev).
The SW eventdev info structure did not return a correct device pointer,
though, instead re
16/10/2023 11:16, Abhijit Gangurde:
> Provide driver flag which gives an option to map the cdx device
> resource before probing the device driver. External driver can
> use rte_cdx_map_device() and rte_cdx_unmap_device() APIs to map/
> unmap device resource separately.
>
> Signed-off-by: Abhijit G
On Fri, Sep 29, 2023 at 8:16 PM wrote:
>
> From: Sunil Kumar Kori
>
> It adds ipv4_lookup module to configure LPM table. This LPM table
> will be used for IPv4 lookup and forwarding.
>
> Following commands are exposed:
> - ipv4_lookup route add ipv4 netmask via
> - help ipv4_lookup
>
> Signe
On 10/16/2023 3:59 PM, Thomas Monjalon wrote:
> 13/10/2023 17:57, Ferruh Yigit:
>> Drivers start/stop device queues on port start/stop, but not all drivers
>> update queue state accordingly.
>>
>> This becomes more visible if a specific queue stopped explicitly and
>> port stopped/started later, in
On Fri, Sep 29, 2023 at 3:29 PM wrote:
>
> From: Sunil Kumar Kori
>
> It adds application's command line parameter "--enable-graph-stats"
> to enable dumping graph stats on console.
>
> By default, no graph stats will be printed on console but same can
> be dumped via telnet session using "graph
On Fri, Sep 29, 2023 at 5:56 PM wrote:
>
> From: Sunil Kumar Kori
>
> In the continuation of following feedback
> https://patches.dpdk.org/project/dpdk/patch/20230425131516.3308612-5-vattun...@marvell.com/
> this patch series adds dpdk-graph application to exercise various
> usecases using graph.
On Mon, 16 Oct 2023 21:31:29 +0530
Jerin Jacob wrote:
> ### [PATCH] app/graph: add telnet connectivity framework
>
> WARNING:STRNCPY: Prefer strscpy, strscpy_pad, or __nonstring over
> strncpy - see: https://github.com/KSPP/linux/issues/90
> #205: FILE: app/graph/conn.c:162:
> + strncpy(co
On 10/13/2023 7:06 AM, Chaoyong He wrote:
> Make the source files follow the alphabeta sequence.
> Also update the copyright header line.
>
> Signed-off-by: Chaoyong He
> Reviewed-by: Long Wu
> Reviewed-by: Peng Zhang
> ---
> drivers/net/nfp/meson.build | 23 ---
> 1 file c
On 10/13/2023 7:06 AM, Chaoyong He wrote:
> This patch series aims to unify the coding style of NFP PMD, make the
> logics following the same rules, to make it easier to understand and
> extend.
> Also prepare for the upcoming vDPA PMD patch series.
>
> ---
> v2:
> * Add some missing modification.
From: Hamdan Igbaria
Add support for creation of packet reformat object,
via the ALLOC_PACKET_REFORMAT_CONTEXT command.
Signed-off-by: Hamdan Igbaria
---
drivers/common/mlx5/mlx5_prm.h | 39 +
drivers/net/mlx5/hws/mlx5dr_cmd.c | 60 ++
drive
From: Hamdan Igbaria
Add the ability to create forward table and FTE.
Signed-off-by: Hamdan Igbaria
---
drivers/common/mlx5/mlx5_prm.h| 4
drivers/net/mlx5/hws/mlx5dr_cmd.c | 13 +
drivers/net/mlx5/hws/mlx5dr_cmd.h | 19 +++
3 files changed, 36 insertions(
From: Hamdan Igbaria
Add mlx5dr_devx_obj struct to mlx5dr_action, so we could hold
the FT obj in dest table action.
Signed-off-by: Hamdan Igbaria
---
drivers/net/mlx5/hws/mlx5dr_action.c | 4
drivers/net/mlx5/hws/mlx5dr_action.h | 3 +++
drivers/net/mlx5/hws/mlx5dr_table.c | 1 -
3 files
From: Erez Shitrit
In FDB it will direct the packet into the hypervisor vport.
That allows the user to mirror packets into the default-miss vport.
Signed-off-by: Erez Shitrit
---
drivers/net/mlx5/hws/mlx5dr_action.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/drivers/net/m
From: Shun Hao
This patch supports mirroring by adding an dest_array action. The action
accecpts a list containing multiple destination actions, and can duplicate
packet and forward to each destination in the list.
Signed-off-by: Shun Hao
Acked-by: Alex Vesker
---
drivers/common/mlx5/mlx5_prm
From: Haifei Luo
In dest_array action, an optional reformat action can be applied to each
destination. This patch supports this by using the extended destination
entry.
Signed-off-by: Haifei Luo
Signed-off-by: Shun Hao
---
drivers/common/mlx5/mlx5_prm.h | 15 +++
drivers/net/mlx5/hw
Replace if() with switch().
Signed-off-by: Gregory Etelson
---
drivers/net/mlx5/mlx5_flow_hw.c | 70 ++---
1 file changed, 39 insertions(+), 31 deletions(-)
diff --git a/drivers/net/mlx5/mlx5_flow_hw.c b/drivers/net/mlx5/mlx5_flow_hw.c
index 6fcf654e4a..b2215fb5cf 10
HWS mirror clones original packet to one or two destinations and
proceeds with the original packet path.
The mirror has no dedicated RTE flow action type.
Mirror object is referenced by INDIRECT_LIST action.
INDIRECT_LIST for a mirror built from actions list:
SAMPLE [/ SAMPLE] / / END
Mirro
HWS mirror flow action validation rejected flows in NIC domain,
if PMD FDB mode was active.
The patch allows NIC mirror action in FDB mode.
Fixes: 0284c9b82ee8 ("net/mlx5: support HWS mirror action")
Signed-off-by: Gregory Etelson
---
drivers/net/mlx5/mlx5_flow_hw.c | 20 +---
1
Count and age actions in HWS template are translated into the same
DR5 action.
PMD maintains dedicated variable - `cnt_off`, that points
action location in DR5 array.
Currnet PMD did not initialize the `cnt_off` variable during
shared counter / age actions initialization.
Fixes: feb1f2fe2b76 ("ne
Flow actions in HWS actions template must be arranged according to
pre-defined order.
MLX5 PMD handles RAW_DECAP / RAW_ENCAP actions sequence as a single
RAW ENCAP or DECAP flow action.
When the PMD scanned flow actions to locate position where
MODIFY_FIELD action could be inserted it processed t
1. In `rte_flow_actions_template`, rename `actions_off` into `dr_off`.
2. Remove duplicated code in template table creation.
Signed-off-by: Gregory Etelson
---
drivers/net/mlx5/mlx5_flow.h| 2 +-
drivers/net/mlx5/mlx5_flow_hw.c | 117 ++--
2 files changed, 52 in
PMD can implicitly add flow actions to application actions template.
If PMD added actions to a template is must track location of the
original application actions in modified template.
The patch adds tracking ability for the orignal acton in a template.
Fixes: ddb68e4 ("net/mlx5: add extended met
PMD used a buffer located on expired stack to store
mirror reformat data.
The patch moves reformat buffer to the same context as the mirror
action creation.
Fixes: 0284c9b82ee8 ("net/mlx5: support HWS mirror action")
Signed-off-by: Gregory Etelson
---
drivers/net/mlx5/mlx5_flow_hw.c | 17 ++
Signed-off-by: Gregory Etelson
---
drivers/net/mlx5/mlx5_flow.c| 69 +-
drivers/net/mlx5/mlx5_flow.h| 67 -
drivers/net/mlx5/mlx5_flow_hw.c | 427 +++-
3 files changed, 482 insertions(+), 81 deletions(-)
diff --git a/drivers/net/mlx5/mlx5_flow.c b/dr
Indirect action list METER_MARK handle and configuration parameters
can be independently masked or non-masked in actions template.
Non-masked configuration state is saved in the
`mlx5_action_construct_data::shared_meter` object.
The patch moves indirect action list callback from
`mlx5_action_const
We are temporarily exposed while the CI is enabled that tests
CC=clang && -Denable_stdatomic.
In the meantime two mistakes have been found so fix them to allow
the CI to be enabled.
Tyler Retzlaff (2):
power: fix use of rte stdatomic
event/cnxk: remove single use of rte stdatomic
drivers/ev
The variable operated on by the single use of rte stdatomic was not
RTE_ATOMIC(T) specified. Remove the use of stdatomic for now to fix
LLVM build with enable_stdatomic=true. event/cnxk will be converted
to rte stdatomic in a later series.
Fixes: 14a4aa9eae71 ("event/cnxk: support get remaining ti
* rte stdatomic functions operate on RTE_ATOMIC(T) specified types not
regular T add missing specifier for amd_pstate_power_info state field
* use rte_memory_order_xxx instead of __ATOMIC_XXX
Fixes: 1ed04d33cf19 ("power: support amd-pstate cpufreq driver")
Cc: sivaprasad.tumm...@amd.com
Signed-
On Mon, Oct 16, 2023 at 8:53 PM Tyler Retzlaff
wrote:
>
> We are temporarily exposed while the CI is enabled that tests
> CC=clang && -Denable_stdatomic.
>
> In the meantime two mistakes have been found so fix them to allow
> the CI to be enabled.
>
> Tyler Retzlaff (2):
> power: fix use of rte
A new API to allow power monitoring condition on event port to
optimize power when no events are arriving on an event port for
the worker core to process in an eventdev based pipelined application.
Signed-off-by: Sivaprasad Tummala
---
lib/eventdev/eventdev_pmd.h | 22 ++
lib
Currently sw eventdev pmd does not support ``rte_power_monitor`` api.
This patch adds support by adding monitor callback that is called
whenever we enter sleep state and need to check if it is time to
wake up.
Signed-off-by: Sivaprasad Tummala
---
drivers/event/sw/sw_evdev.c| 1 +
drive
This patch adds support for an event dequeue callbacks when configured
to allow additional processing, power management (frequency scaling,
monitor/sleep etc.,) and soon.
Signed-off-by: Sivaprasad Tummala
---
config/rte_config.h| 3 +++
drivers/event/sw/sw_evdev_worker.c | 5
Add optional support for inline event processing within pmd dequeue
call. For a dequeue callback, events dequeued from the event port
were passed them to a callback function if configured, to allow
additional processing. e.g. unpack batch of packets from each event
on dequeue, before passing back t
Add eventdev support to enable power saving when no events
are arriving. It is based on counting the number of empty
polls and, when the number reaches a certain threshold, entering
an architecture-defined optimized power state that will either wait
until a TSC timestamp expires, or when events arr
Add power management feature support to eventdev_pipeline sample app.
A new command option "--pmd-mgmt" was added to select power management
mode on worker cores. Options currently supported are "pause/monitor".
Default, no power management features are enabled on the worker ports.
Signed-off-by:
Before this patch, max_rx_pktlen was always set to UINT16_MAX. This, in
conjunction with the MTU fix, causes problems with testpmd, as setting the
packet length with the --max-pkt-len flag causes the MTU to be set
higher than possible due to underflow.
As an example, setting --max-pkt-len=1460 (th
On Mon, 16 Oct 2023 13:59:48 -0700
Joshua Washington wrote:
> Before this patch, max_rx_pktlen was always set to UINT16_MAX. This, in
> conjunction with the MTU fix, causes problems with testpmd, as setting the
> packet length with the --max-pkt-len flag causes the MTU to be set
> higher than pos
On Mon, 16 Oct 2023 13:59:48 -0700
Joshua Washington wrote:
> conjunction with the MTU fix, causes problems with testpmd, as setting the
> packet length with the --max-pkt-len flag causes the MTU to be set
> higher than possible due to underflow.
>
> As an example, setting --max-pkt-len=1460 (th
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API.
Note the first 2 patches are redundant and will be removed before merge.
The series is a work in progress and is being submitted to get initial CI
results.
Tyler Retzlaff (21):
power
The variable operated on by the single use of rte stdatomic was not
RTE_ATOMIC(T) specified. Remove the use of stdatomic for now to fix
LLVM build with enable_stdatomic=true. event/cnxk will be converted
to rte stdatomic in a later series.
Fixes: 14a4aa9eae71 ("event/cnxk: support get remaining ti
* rte stdatomic functions operate on RTE_ATOMIC(T) specified types not
regular T add missing specifier for amd_pstate_power_info state field
* use rte_memory_order_xxx instead of __ATOMIC_XXX
Fixes: 1ed04d33cf19 ("power: support amd-pstate cpufreq driver")
Cc: sivaprasad.tumm...@amd.com
Signed-
Replace the use of gcc builtin __atomic_xxx intrinsics with corresponding
rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/power/power_acpi_cpufreq.c | 33 +
lib/power/power_cppc_cpufreq.c | 25 +
lib/power/po
Replace the use of gcc builtin __atomic_xxx intrinsics with corresponding
rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/bbdev/rte_bbdev.c | 6 +++---
lib/bbdev/rte_bbdev.h | 2 +-
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/lib/bbdev/rte_bbdev.c b
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
drivers/event/cnxk/cnxk_tim_worker.h | 4 +--
lib/eventdev/rte_event_timer_adapter.c | 66 +-
lib/eventdev/rte_even
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/ipsec/ipsec_sqn.h | 2 +-
lib/ipsec/sa.h| 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/ipsec/ipsec_sqn.h b/lib
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/mempool/rte_mempool.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/mempool/rte_mempool.h b/lib/mempool/rte_mempool.
Replace the use of gcc builtin __atomic_xxx intrinsics with corresponding
rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/eal/common/eal_common_launch.c| 10 +--
lib/eal/common/eal_common_mcfg.c | 2 +-
lib/eal/common/eal_common_proc.c | 14 ++--
lib
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/mbuf/rte_mbuf.h | 20 ++--
lib/mbuf/rte_mbuf_core.h | 4 ++--
2 files changed, 12 insertions(+), 12 deletions(-)
diff --git
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/cryptodev/rte_cryptodev.c | 22 --
lib/cryptodev/rte_cryptodev.h | 16
2 files changed, 20 insertions(+), 18
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/gpudev/gpudev.c| 6 +++---
lib/gpudev/gpudev_driver.h | 2 +-
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/lib/gpudev/gp
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/vhost/vdpa.c| 3 ++-
lib/vhost/vhost.c | 42
lib/vhost/vhost.h | 39
1 - 100 of 156 matches
Mail list logo
