For net/mlx4 and net/mlx5, fix cache rebuild bug and replace SMP
barriers with atomic fence.
Feifei Wang (4):
net/mlx4: fix rebuild bug for Memory Region cache
net/mlx4: replace SMP barrier with C11 barriers
net/mlx5: fix rebuild bug for Memory Region cache
net/mlx5: replace SMP barriers w
'dev_gen' is a variable to inform other cores to flush their local cache
when global cache is rebuilt.
However, if 'dev_gen' is updated after global cache is rebuilt, other
cores may load a wrong memory region lkey value from old local cache.
Timeslotmain core worker core
Replace SMP barrier with atomic thread fence.
Signed-off-by: Feifei Wang
Reviewed-by: Ruifeng Wang
---
drivers/net/mlx4/mlx4_mr.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/mlx4/mlx4_mr.c b/drivers/net/mlx4/mlx4_mr.c
index cfd7d4a9c..503e8a7bb 100644
---
'dev_gen' is a variable to inform other cores to flush their local cache
when global cache is rebuilt.
However, if 'dev_gen' is updated after global cache is rebuilt, other
cores may load a wrong memory region lkey value from old local cache.
Timeslotmain core worker core
Replace SMP barrier with atomic thread fence.
Signed-off-by: Feifei Wang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
drivers/net/mlx5/mlx5_mr.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/mlx5/mlx5_mr.c b/drivers/net/mlx5/mlx5_mr.
This commit introduced the conntrack action and item.
Usually the HW offloading is stateless. For some stateful offloading
like a TCP connection, HW module will help provide the ability of a
full offloading w/o SW participation after the connection was
established.
The basic usage is that in the
17/03/2021 23:43, Dmitry Kozlyuk:
> 2021-03-17 00:11 (UTC+0100), Thomas Monjalon:
> [...]
> > diff --git a/drivers/bus/pci/rte_bus_pci.h b/drivers/bus/pci/rte_bus_pci.h
> > index fdda046515..3d009cc74b 100644
> > --- a/drivers/bus/pci/rte_bus_pci.h
> > +++ b/drivers/bus/pci/rte_bus_pci.h
> > @@ -52
18/03/2021 00:17, Ranjit Menon:
> Hi Thomas,
>
> On 3/16/2021 4:11 PM, Thomas Monjalon wrote:
> > In Windows probing, the value RTE_PCI_KDRV_NONE was used
> > instead of RTE_PCI_KDRV_UNKNOWN (mlx case),
> > and RTE_PCI_KDRV_NIC_UIO (FreeBSD) was re-used
> > instead of having a new RTE_PCI_KDRV_NET
18/03/2021 04:56, oulijun:
> 在 2021/3/17 20:07, Thomas Monjalon 写道:
> > 17/03/2021 12:30, oulijun:
> >> 2021/3/12 19:21, Thomas Monjalon:
> >>> 12/03/2021 11:29, oulijun:
> 2021/3/10 15:59, Thomas Monjalon:
> > 10/03/2021 02:48, oulijun:
> >> Can we add an API such as rte_eth_get_devic
On 3/18/21 1:42 AM, Thomas Monjalon wrote:
> 28/02/2021 13:53, Thomas Monjalon:
>> These patches fix some build errors/warning for Alpine Linux,
>> using musl and busybox.
>> Few improvements are added on the way.
>
> No more comment on this series? Ready to merge?
Series-acked-by: Andrew Rybchen
Hi, Thomas
> -Original Message-
> From: dev On Behalf Of Thomas Monjalon
> Sent: Wednesday, March 17, 2021 1:12
> To: dev@dpdk.org
> Cc: dmitry.kozl...@gmail.com; sta...@dpdk.org; Tal Shnaiderman
> ; Narcisa Vasile ; Ranjit
> Menon ; John Alexander
> ; Pallavi Kadam
>
> Subject: [dpdk-de
Hi Maxime,
> -Original Message-
> From: Maxime Coquelin
> Sent: Tuesday, March 16, 2021 5:38 PM
> To: dev@dpdk.org; Xia, Chenbo ; amore...@redhat.com;
> david.march...@redhat.com; olivier.m...@6wind.com; bnem...@redhat.com
> Cc: Maxime Coquelin
> Subject: [PATCH v4 3/4] net/virtio: alloc
On Wed, Mar 17, 2021 at 11:42 PM Thomas Monjalon wrote:
>
> 28/02/2021 13:53, Thomas Monjalon:
> > These patches fix some build errors/warning for Alpine Linux,
> > using musl and busybox.
> > Few improvements are added on the way.
>
> No more comment on this series? Ready to merge?
This will pro
Enable GTPU pattern for CVL switch filter. This patch only
supports outer l3/l4 filtering.
Signed-off-by: Yuying
---
doc/guides/rel_notes/release_21_05.rst | 3 +
drivers/net/ice/ice_switch_filter.c| 91 ++
2 files changed, 94 insertions(+)
diff --git a/doc/guides/r
Currently, the flow meter policy does not support multiple actions
per color; also the allowed action types per color are very limited.
In addition, the policy cannot be pre-defined.
Due to the growing in flow actions offload abilities there is a potential
for the user to use variety of actions pe
Currently, all the meter objects are managed by the user IDs:
meter, profile and policy.
Hence, each PMD should manage data-structure in order to
map each API ID to the private PMD management structure.
>From the application side, it has all the picture how meter
is going to be assigned to flows a
18/01/2021 16:16, Xueming Li:
> --- a/lib/librte_eal/common/eal_common_devargs.c
> +++ b/lib/librte_eal/common/eal_common_devargs.c
> + if (ret != 0) {
> + if (devargs->data && devargs->data != devstr) {
Better to make comparison explicit:
if (devargs->data != NULL
> +
From: Kalesh AP
VFs does not have the privilege to change link configuration.
But the driver silently returns success to these ethdev callbacks
without actually issuing the HWRM command to bring the link up/down.
Fixes: 5c206086feaa ("net/bnxt: add link state operations")
Cc: sta...@dpdk.org
Si
From: Kalesh AP
Fixed to return error when PTP support is not supported on the port.
Also, removed an unnecessary check inside bnxt_get_rx_ts().
Fixes: b11cceb83a34 ("net/bnxt: support timesync")
Cc: sta...@dpdk.org
Signed-off-by: Kalesh AP
Reviewed-by: Ajit Kumar Khaparde
Reviewed-by: Somnat
The driver should notify the guest for each traffic burst detected by CQ
polling.
The CQ polling trigger is defined by `event_mode` device argument,
either by busy polling on all the CQs or by blocked call to HW
completion event using DevX channel.
Also, the polling event modes can move to blocke
18/01/2021 16:16, Xueming Li:
> In current design, legacy parser rte_devargs_parse() saved scratch
> buffer to devargs.args while new parser rte_devargs_layers_parse() saved
> to devargs.data. Code using devargs had to know the difference and
> cleaned up memory accordingly - error prone.
>
> This
Adding bare minimum PMD library and doc build infrastructure
and claim the maintainership for ngbe PMD.
Signed-off-by: Jiawen Wu
---
MAINTAINERS| 6 ++
doc/guides/nics/features/ngbe.ini | 10 +
doc/guides/nics/index.rst | 1 +
doc/guide
Add device IDs for Wangxun 1Gb NICs, and register rte_ngbe_pmd.
Signed-off-by: Jiawen Wu
---
drivers/net/ngbe/base/meson.build | 18 +++
drivers/net/ngbe/base/ngbe_devids.h | 83 +
drivers/net/ngbe/meson.build| 6 +++
drivers/net/ngbe/ngbe_ethdev.c
This patch set provides a skeleton of ngbe PMD,
which adapted to Wangxun WX1860 series NICs.
Jiawen Wu (6):
net/ngbe: add build and doc infrastructure
net/ngbe: add device IDs
net/ngbe: support probe and remove
net/ngbe: add device init and uninit
net/ngbe: add log type and error type
Add basic init and uninit function.
Map device IDs and subsystem IDs to single ID for easy opearation.
Signed-off-by: Jiawen Wu
---
drivers/net/ngbe/base/meson.build | 4 +-
drivers/net/ngbe/base/ngbe.h | 11 ++
drivers/net/ngbe/base/ngbe_hw.c| 59 ++
drivers/net/ngbe/base
Add log type and error type to trace functions.
Signed-off-by: Jiawen Wu
---
doc/guides/nics/ngbe.rst| 41 +
drivers/net/ngbe/base/ngbe_status.h | 124
drivers/net/ngbe/base/ngbe_type.h | 1 +
drivers/net/ngbe/ngbe_ethdev.c | 20 +
Add basic PCIe ethdev probe and remove.
Signed-off-by: Jiawen Wu
---
doc/guides/nics/features/ngbe.ini | 1 +
drivers/net/ngbe/ngbe_ethdev.c| 77 +--
2 files changed, 75 insertions(+), 3 deletions(-)
diff --git a/doc/guides/nics/features/ngbe.ini
b/doc/guides/n
Define all registers that will be used.
Signed-off-by: Jiawen Wu
---
drivers/net/ngbe/base/ngbe_regs.h | 1489 +
drivers/net/ngbe/base/ngbe_type.h |2 +
2 files changed, 1491 insertions(+)
create mode 100644 drivers/net/ngbe/base/ngbe_regs.h
diff --git a/drivers
Due to kernel issue in direct MKEY creation using the DevX API for
physical memory, this patch replaces the counter MR creation to use
Verbs API.
Fixes: 3aa279157fa0 ("net/mlx5: synchronize flow counter pool creation")
Cc: sta...@dpdk.org
Signed-off-by: Michael Baum
Acked-by: Matan Azrad
---
d
Adjusting mkey creations to use Verbs instead of DevX API.
Michael Baum (2):
net/mlx5: workaround ASO memory region creation
net/mlx5: workaround counter memory region creation
drivers/common/mlx5/linux/mlx5_common_verbs.c | 1 -
drivers/common/mlx5/windows/mlx5_common_os.c | 23 +++---
Due to kernel issue in direct MKEY creation using the DevX API for
physical memory, this patch replaces the ASO MR creation to use Verbs
API.
Fixes: f935ed4b645a ("net/mlx5: support flow hit action for aging")
Cc: sta...@dpdk.org
Signed-off-by: Michael Baum
Acked-by: Matan Azrad
---
drivers/co
From: Thomas Monjalon
> 17/03/2021 18:39, Ferruh Yigit:
> > On 3/9/2021 9:48 AM, Thomas Monjalon wrote:
> > > Most debug logs are using DRV_LOG(DEBUG,) but some were using
> > > DEBUG().
> > > The macro DEBUG is doing nothing if not compiled with
> > > RTE_LIBRTE_MLX5_DEBUG.
> > >
> > > As it is
> -Original Message-
> From: dapengx...@intel.com
>
> Fixes: 6833f919f56b ("examples/packet_ordering: convert to new ethdev
> offloads API")
> Cc: sta...@dpdk.org
Also, need to add CC: i.e CC: Shahaf Shuler
Other than that , patch looks ok to me. Please include my ack in next vers
This series of patches include changes to l3fwd example application.
Some improvements are made for better usage of CPU cycles and memory.
Ruifeng Wang (4):
examples/l3fwd: tune prefetch for better performance
examples/l3fwd: eliminate unnecessary calculations
examples/l3fwd: eliminate unnec
Packet header is prefetched before packet processing for better
memory access performance. As L2 header will be updated by l3fwd,
using of prefetch for store hint will set cache line to proper
status and reduce cache maintenance overhead.
With this change, 12.9% performance uplift was measured on
Both L2 and L3 headers will be used in forward processing. And these
two headers are in the same cache line. It has the same effect for
prefetching with L2 header address and prefetching with L3 header
address.
Changed to use L2 header address for prefetching. The change showed
no measurable perfo
Number of rx queue and number of rx port in lcore config are constants
during the period of l3 forward application running. But compiler has
no this information.
Copied values from lcore config to local variables and used the local
variables for iteration. Compiler can see that the local variables
There are some holes in data struct lcore_conf. The holes are
due to alignment requirement.
For struct lcore_rx_queue, there is no need to make every element
of this type to be cache line aligned, because the data is not
shared between cores.
Member len of struct mbuf_table can be moved out. So d
> -Original Message-
> From: dev On Behalf Of Leyi Rong
> Sent: Wednesday, March 17, 2021 9:14 AM
> To: Zhang, Qi Z ; Lu, Wenzhuo
> Cc: dev@dpdk.org; Rong, Leyi
> Subject: [dpdk-dev] [PATCH 2/2] net/ice: add Rx AVX512 offload path
>
> Split AVX512 Rx data path into two, one is for basic
18/03/2021 09:36, Slava Ovsiienko:
> From: Thomas Monjalon
> > In Windows probing, the value RTE_PCI_KDRV_NONE was used instead of
> > RTE_PCI_KDRV_UNKNOWN (mlx case), and RTE_PCI_KDRV_NIC_UIO
> > (FreeBSD) was re-used instead of having a new RTE_PCI_KDRV_NET_UIO for
> > Windows NetUIO.
>
> As far
In Windows probing, the value RTE_PCI_KDRV_NONE was used
instead of RTE_PCI_KDRV_UNKNOWN.
This value covers the mlx case where the kernel driver is in place,
offering a bifurcated mode to the userspace driver.
When the kernel driver is listed as unknown,
there is no special treatment in DPDK probin
Fix typos in the names of kernel drivers based on UIO,
and make sure the generic term for the interface is UIO in capitals.
Fixes: 3a78b2f73206 ("doc: add virtio crypto PMD guide")
Fixes: 3cc4d996fa75 ("doc: update VFIO usage in qat crypto guide")
Fixes: 39922c470e3c ("doc: add known uio_pci_gener
Currently RSS expansion only support IPv4 as GRE payload or
delivery protocol (RFC2784). IPv6 as GRE payload or delivery protocol
(RFC7676) is not supported.
This patch add RSS expansion for RFC7676 so PMD can expand flow item
correctly.
Fixes: f4b901a46aec ("net/mlx5: add flow GRE item")
Cc: sta
If cache is enabled, objects will be retrieved/put from/to cache,
subsequently from/to the common pool. Now the debug stats caculate
the objects retrived/put from/to cache and pool together, it is
better to distinguish the data number from local cache and common
pool.
Signed-off-by: Joyce Kong
--
If cache is enabled, objects will be retrieved/put from/to cache,
subsequently from/to the common pool. Now the debug stats calculate
the objects retrieved/put from/to cache and pool together, it is
better to distinguish the data number from local cache and common
pool.
Signed-off-by: Joyce Kong
OK, but I think this bug affects all stable release about 6 years before today.
On Tue, Sep 1, 2020 at 3:23 PM Pai G, Sunil wrote:
>
> Hi,
>
> Yes , OVS was using pkg-config even before these patches were rolled out.
> But it always used to pick the DPDK shared libs by default for OVS even on
> using the -Bstatic/-Bshared flags.
> These patches from Bruce simplify the proces
'3' is not valid RTE_COLOR_ enum value.
Fixes: de3cfa2c9823 ("sched: initial import")
Cc: sta...@dpdk.org
Signed-off-by: Konstantin Ananyev
---
examples/qos_sched/app_thread.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/examples/qos_sched/app_thread.c b/examples/qos_sche
In many usage scenarios input mbufs for rte_sched_port_enqueue()
are not yet in the CPU cache(s). That causes quite significant stalls
due to memory latency. Current implementation tries to migitate it
using SW pipeline and SW prefetch techniques, but stalls are still present.
Rework rte_sched_port
> Subject: [PATCH v2] bus/pci: fix Windows kernel driver categories
>
> In Windows probing, the value RTE_PCI_KDRV_NONE was used instead of
> RTE_PCI_KDRV_UNKNOWN.
> This value covers the mlx case where the kernel driver is in place, offering a
> bifurcated mode to the userspace driver.
> When the
18/03/2021 10:32, Jiawen Wu:
> +#ifdef RTE_LIBRTE_NGBE_DEBUG_RX
> +extern int ngbe_logtype_rx;
> +#define PMD_RX_LOG(level, fmt, args...) \
> + rte_log(RTE_LOG_ ## level, ngbe_logtype_rx, \
> + "%s(): " fmt "\n", __func__, ##args)
> +#else
> +#define PMD_RX_LOG(level, fmt, args.
18/03/2021 10:32, Jiawen Wu:
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -886,6 +886,12 @@ F: drivers/net/txgbe/
> F: doc/guides/nics/txgbe.rst
> F: doc/guides/nics/features/txgbe.ini
>
> +Wangxun txgbe
> +M: Jiawen Wu
> +F: drivers/net/ngbe/
> +F: doc/guides/nics/ngbe.rst
> +F: doc/guides/ni
Currently, upper-layer application could get queue state only
through pointers such as dev->data->tx_queue_state[queue_id],
this is not the recommended way to access it. So this patch
add get queue state when call rte_eth_rx_queue_info_get and
rte_eth_tx_queue_info_get API.
Note: The hairpin queue
On Wed, Mar 17, 2021 at 11:01:25AM +0100, Thomas Monjalon wrote:
> 17/03/2021 10:48, David Marchand:
> > On Wed, Mar 17, 2021 at 10:31 AM Thomas Monjalon
> > wrote:
> > >
> > > The macro RTE_VERSION is broken since updated with function calls.
> > > It is a build-time version number, and must be
On Wed, Mar 17, 2021 at 04:15:35PM +0100, Thomas Monjalon wrote:
> Some functions were introduced in DPDK 21.05 to query the version parts
> (prefix, year, month, minor, suffix, release) at runtime.
> Per guidelines, these new public functions must be marked with
> __rte_experimental and ABI versio
On Wed, Mar 17, 2021 at 9:04 PM Maxime Coquelin
wrote:
> > diff --git a/drivers/net/virtio/virtio_user/vhost_user.c
> > b/drivers/net/virtio/virtio_user/vhost_user.c
> > index ec2c53c8fb..18ae29eed2 100644
> > --- a/drivers/net/virtio/virtio_user/vhost_user.c
> > +++ b/drivers/net/virtio/virtio_u
This commit fixes problem with to small offset when both offsets
(auth, cipher) are non zero in digest encrypt case,
when using out-of-place and sgl.
Fixes: 40002f6c2a24 ("crypto/qat: extend support for digest-encrypted
auth-cipher")
Cc: sta...@dpdk.org
Signed-off-by: Arek Kusztal
---
drivers/
Hey Christian,
> back in 19.11.4 these DPDK changes were not picked up as they have broken
> builds as discussed here.
> Later on the communication was that all this works fine now and thereby
> Luca has "reverted the reverts" in 19.11.6 [1].
>
> But today we were made aware that still no OVS
From: Jack Min
> Currently RSS expansion only support IPv4 as GRE payload or delivery protocol
> (RFC2784). IPv6 as GRE payload or delivery protocol
> (RFC7676) is not supported.
>
> This patch add RSS expansion for RFC7676 so PMD can expand flow item
> correctly.
>
> Fixes: f4b901a46aec ("n
18/03/2021 13:28, Bruce Richardson:
> On Wed, Mar 17, 2021 at 11:01:25AM +0100, Thomas Monjalon wrote:
> > 17/03/2021 10:48, David Marchand:
> > > On Wed, Mar 17, 2021 at 10:31 AM Thomas Monjalon
> > > wrote:
> > > >
> > > > The macro RTE_VERSION is broken since updated with function calls.
> > >
> Subject: [dpdk-dev] [PATCH 1/3] Add EAL threads API
>
> From: Narcisa Vasile
>
> EAL must hide the environment specifics from apps and libraries.
> Add an EAL API for managing threads.
>
> Signed-off-by: Narcisa Vasile
> Signed-off-by: Dmitry Malloy
> ---
Hi Naty, Dmitry,
Thank you for a
On 3/18/21 2:36 PM, Pai G, Sunil wrote:
> Hey Christian,
>
>
>
>> back in 19.11.4 these DPDK changes were not picked up as they have broken
>> builds as discussed here.
>> Later on the communication was that all this works fine now and thereby
>> Luca has "reverted the reverts" in 19.11.6 [1].
On Thu, Mar 18, 2021 at 7:38 AM Hemant Agrawal wrote:
>
> This patch introduce the baseband device drivers for NXP's
> LA1200 series software defined baseband modem.
Such a series deserves a cover letter.
You should copy bbdev maintainer and cryptodev subtree maintainer.
Quickly looked at the se
> -Original Message-
> From: Xiaozhen Ban
> Sent: Thursday, March 18, 2021 19:44
> To: Guo, Jia ; Wang, Haiyue
> Cc: dev@dpdk.org; sta...@dpdk.org
> Subject: RE: RE: [PATCH] net/e1000: fix filter control return value
>
> OK, but I think this bug affects all stable release about 6 years b
On Thu, Mar 18, 2021 at 03:41:35PM +0100, Thomas Monjalon wrote:
> 18/03/2021 13:28, Bruce Richardson:
> > On Wed, Mar 17, 2021 at 11:01:25AM +0100, Thomas Monjalon wrote:
> > > 17/03/2021 10:48, David Marchand:
> > > > On Wed, Mar 17, 2021 at 10:31 AM Thomas Monjalon
> > > > wrote:
> > > > >
> >
On Thu, Mar 18, 2021 at 2:01 AM Narcisa Ana Maria Vasile
wrote:
> diff --git a/lib/librte_eal/common/eal_common_thread.c
> b/lib/librte_eal/common/eal_common_thread.c
> index 73a055902..5219e783e 100644
> --- a/lib/librte_eal/common/eal_common_thread.c
> +++ b/lib/librte_eal/common/eal_common_thr
Fixes: b33ccdb17f55 ("net/ark: provide API for hardware modules MPU RQP and
pktdir")
Cc: sta...@dpdk.org
Signed-off-by: Ed Czeck
---
drivers/net/ark/ark_ethdev.c | 1 +
drivers/net/ark/ark_pktdir.c | 2 +-
drivers/net/ark/ark_pktdir.h | 2 +-
3 files changed, 3 insertions(+), 2 deletions(-)
di
Allocate mbufs for Rx path in bulk of at least 64 buffers
to improve performance. Allow recovery even without
a Rx operation to support lack of buffers in pool.
Fixes: be410a861598 ("net/ark: add recovery for lack of mbufs")
Cc: sta...@dpdk.org
Signed-off-by: Ed Czeck
---
drivers/net/ark/ark_et
- New PCIe IDs using net/ark driver
- Update Version IDs and structures specified by hardware
- New internal descriptor status for TX
- Adjust data placement in RX operations, headroom in retained
for segmented mbufs
Signed-off-by: Ed Czeck
---
doc/guides/nics/ark.rst | 5 ++
drivers/n
- Rename extension functions with rte_pmd_ark prefix
- Update local function documentation
Signed-off-by: Ed Czeck
---
v3:
- split function rename from previous commit
v4:
- reorder patches renaming before adding
v5:
- Keep the extension function changes in ark_ext.h
---
drivers/net/ark/ark_ethd
In this commit we generalize the movement of user-specified
meta data between mbufs and FPGA AXIS tuser fields using
user-defined hook functions.
- Previous use of PMD dynfields are removed
- Remove emptied rte_pmd_ark.h
- Hook function added to ark_user_ext
- Add hook function calls in Rx and Tx
remove unnecessary includes
no functional changes
Signed-off-by: Ed Czeck
---
drivers/net/ark/ark_ethdev.c | 17 ++---
drivers/net/ark/ark_pktchkr.c | 4
drivers/net/ark/ark_pktgen.c | 20 ++--
drivers/net/ark/ark_pktgen.h | 1 +
4 files changed, 17 insertio
On Wed, Mar 17, 2021 at 09:25:26PM +0100, Ilya Maximets wrote:
Hi,
Some questions to understand the problems that SocketPair Broker solves:
> Even more configuration tricks required in order to share some sockets
> between different containers and not only with the host, e.g. to
> create service c
In case an event from a previous stage is required to be forwarded
to a crypto adapter and PMD supports internal event port in crypto
adapter, exposed via capability
RTE_EVENT_CRYPTO_ADAPTER_CAP_INTERNAL_PORT_OP_FWD, we do not have
a way to check in the API rte_event_enqueue_burst(), whether it is
This set contains a series of updates to the ioat driver, described in each of
the individual patchsets.
Comments would be especially appreciated for the last patch in this set, which
converts the existing idxd vdev driver to a bus driver so that probing and
scanning can be done automatically. Thi
When using a full device instance via vfio, allow the user to specify a
maximum number of queues to configure rather than always using the max
number of supported queues.
Signed-off-by: Bruce Richardson
---
drivers/raw/ioat/idxd_pci.c | 28 ++--
1 file changed, 26 inserti
Add the driver prefix "IOAT" to log messages for the driver.
Signed-off-by: Bruce Richardson
---
drivers/raw/ioat/ioat_private.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/raw/ioat/ioat_private.h b/drivers/raw/ioat/ioat_private.h
index 6c423811ec..f032d5fe3d 1006
Add an explicit padding field to the end of the descriptor structure so
that when the batch descriptor is defined on the stack for perform-ops, the
unused space is all zeroed appropriately.
Signed-off-by: Bruce Richardson
---
drivers/raw/ioat/rte_ioat_rawdev_fns.h | 5 -
1 file changed, 4 in
The ring management in the idxd part of the driver is more complex than
it needs to be, tracking individual batches in a ring and having null
descriptors as padding to avoid having single-operation batches. This can
be simplified by using a regular ring-based layout, with additional
overflow at the
From: Kevin Laatz
Add a new API to query remaining descriptor ring capacity. This API is
useful, for example, when an application needs to enqueue a fragmented
packet and wants to ensure that all segments of the packet will be enqueued
together.
Signed-off-by: Kevin Laatz
---
drivers/raw/ioat/
Rather than using a vdev with args, DPDK can scan and initialize the
devices automatically using a bus-type driver. This bus does not need to
worry about registering device drivers, rather it can initialize the
devices directly on probe.
The device instances (queues) to use are detected from /dev
Hi Christian, Ilya
> -Original Message-
> From: Ilya Maximets
> Sent: Thursday, March 18, 2021 8:18 PM
> To: Pai G, Sunil ; Christian Ehrhardt
> ; Stokes, Ian ;
> Ilya Maximets ; Govindharajan, Hariprasad
>
> Cc: Richardson, Bruce ; Luca Boccassi
> ; sta...@dpdk.org; dev ; James Page
>
Release status meeting minutes {Date}
=
:Date: 18 March 2021
:toc:
.Agenda:
* Release Dates
* Subtrees
* LTS
* Conferences
* Opens
.Participants:
* Arm
* Debian/Microsoft
* Intel
* Marvell
* Nvidia
* Red Hat
Release Dates
-
* `v21.05` dates
- P
On Thu, Mar 18, 2021 at 04:48:49PM +0100, David Marchand wrote:
> On Thu, Mar 18, 2021 at 2:01 AM Narcisa Ana Maria Vasile
> wrote:
> > diff --git a/lib/librte_eal/common/eal_common_thread.c
> > b/lib/librte_eal/common/eal_common_thread.c
> > index 73a055902..5219e783e 100644
> > --- a/lib/librte
On Thu, Mar 18, 2021 at 02:48:01PM +, Tal Shnaiderman wrote:
> > Subject: [dpdk-dev] [PATCH 1/3] Add EAL threads API
> >
> > From: Narcisa Vasile
> >
> > EAL must hide the environment specifics from apps and libraries.
> > Add an EAL API for managing threads.
> >
> > Signed-off-by: Narcisa
> -Original Message-
> From: dev On Behalf Of Conor Walsh
> Sent: Monday, March 15, 2021 11:35 AM
> To: jer...@marvell.com; step...@networkplumber.org; Iremonger, Bernard
> ; Ananyev, Konstantin
> ; Medvedkin, Vladimir
> ; Burakov, Anatoly
>
> Cc: dev@dpdk.org; Walsh, Conor
> Subject:
On 3/18/21 6:52 PM, Stefan Hajnoczi wrote:
> On Wed, Mar 17, 2021 at 09:25:26PM +0100, Ilya Maximets wrote:
> Hi,
> Some questions to understand the problems that SocketPair Broker solves:
>
>> Even more configuration tricks required in order to share some sockets
>> between different containers a
On Thu, Mar 18, 2021 at 02:48:01PM +, Tal Shnaiderman wrote:
>
> I don't know if this table is needed, the approach should be to have the
> return value/rte_errno identical between the OSs.
> And having the specific OS errno printed.
the underlying problem here is that dpdk is adopting linux
On 3/18/21 8:47 PM, Ilya Maximets wrote:
> On 3/18/21 6:52 PM, Stefan Hajnoczi wrote:
>> On Wed, Mar 17, 2021 at 09:25:26PM +0100, Ilya Maximets wrote:
>> Hi,
>> Some questions to understand the problems that SocketPair Broker solves:
>>
>>> Even more configuration tricks required in order to share
(Bcc: Arm internal stake holders)
Hello,
Following are the work items planned for 21.05:
1) Performance improvements in L3fwd example.
2) Use C11 atomic built-ins in EAL, Service Core library and MLX4/MLX5 PMDs.
3) Enhance mempool library with additional debug counters.
4) Meson build imp
On 3/18/2021 12:49 AM, Thomas Monjalon wrote:
18/03/2021 00:17, Ranjit Menon:
Hi Thomas,
On 3/16/2021 4:11 PM, Thomas Monjalon wrote:
In Windows probing, the value RTE_PCI_KDRV_NONE was used
instead of RTE_PCI_KDRV_UNKNOWN (mlx case),
and RTE_PCI_KDRV_NIC_UIO (FreeBSD) was re-used
instead of
On 3/18/2021 3:48 AM, Thomas Monjalon wrote:
In Windows probing, the value RTE_PCI_KDRV_NONE was used
instead of RTE_PCI_KDRV_UNKNOWN.
This value covers the mlx case where the kernel driver is in place,
offering a bifurcated mode to the userspace driver.
When the kernel driver is listed as unknow
This patch adds vDPA device config space requests support.
For now, it only adds MAC address get and set. It may be
extended in next revision to support other configs like
link state.
Regarding the MAC selection strategy, if devargs MAC address
is set by the user and valid, the driver tries to sto
This patch is preliminary rework to add support for getting
and setting device's config space.
In order to get or set a device config such as its MAC address,
we need to know whether the device itself support the feature,
or if it is emulated by the frontend.
Signed-off-by: Maxime Coquelin
---
This patch introduces two virtio-user callbacks to get
and set device's config, and implements it for vDPA
backends.
Signed-off-by: Maxime Coquelin
---
drivers/net/virtio/virtio_user/vhost.h | 3 +
drivers/net/virtio/virtio_user/vhost_vdpa.c | 69 +
2 files changed, 72
This patch uses the new device config ops to get and set
the MAC address if supported.
If a valid MAC address is passed as devarg of the
Virtio-user PMD, the driver will try to store it in the
device config space. Otherwise the one provided in
the device config space will be used, if available.
S
Use correct threshold when selecting "greater than or equal to
2K" length hint.
Fixes: 6eb3cc2294fd ("net/bnxt: add initial Tx code")
Cc: sta...@dpdk.org
Signed-off-by: Lance Richardson
Reviewed-by: Ajit Kumar Khaparde
Reviewed-by: Somnath Kotur
---
drivers/net/bnxt/bnxt_txr.c | 2 +-
1 file
Remove early buffer posting logic from burst receive loop to address
several issues:
- Posting receive descriptors without first posting completion
entries risks overflowing the completion queue.
- Posting receive descriptors without updating rx_raw_prod
creates the possibility that
When the mask field of an rte_flow pattern item is NULL,
the default mask for that item type should be used.
Fixes: 5ef3b79fdfe6 ("net/bnxt: support flow filter ops")
Cc: sta...@dpdk.org
Signed-off-by: Lance Richardson
---
drivers/net/bnxt/bnxt_flow.c | 47 +++-
This series add four features according to the 21.05 roadmap, and
also two cleanups added.
Chengchang Tang (1):
net/hns3: support for outer UDP cksum
Chengwen Feng (1):
net/hns3: support runtime config to select IO burst func
Hongbo Zheng (4):
net/hns3: adjust the format of RAS related str
1 - 100 of 118 matches
Mail list logo
