> -Original Message-
> From: Ferruh Yigit
> Sent: Wednesday, June 27, 2018 4:59 PM
> To: Ido Goshen ; Bruce Richardson
> ; John McNamara
> ; Marko Kovacevic
>
> Cc: dev@dpdk.org
> Subject: Re: [PATCH v3] net/pcap: rx_iface_in stream type support
>
> On 6/27/2018 1:04 PM, ido goshen wro
Hi Jeff
A good advance, thank you, but as I said in previous version, this patch
inserts a bug and the next one fixes it.
Patch 9 should be before patch 8 while this patch just add 1 more option for
EAL hotplug.
Please see 1 more comment below.
From: Jeff Guo
> Use testpmd for example, to show
On 06/29/2018 05:59 PM, Tiwei Bie wrote:
@@ -888,7 +914,8 @@ vhost_user_set_mem_table(struct virtio_net **pdev, struct
VhostUserMsg *pmsg)
static int
vq_is_ready(struct vhost_virtqueue *vq)
{
- return vq && vq->desc && vq->avail && vq->used &&
+ return vq &&
+ (
From: ido g
Support rx of in direction packets only
Useful for apps that also tx to eth_pcap ports in order to not see them
echoed back in as rx when out direction is also captured
Example:
In case using rx_iface and sending *single* packet to eth1
it will loop forever as the when it is sent to
Hello,
We are in process of migrating our design from DPDK 17.05 to 17.11 and we ran
into a small problem. Within our design, we have some hash tables with 4-byte
keys. While going through the changes done in 17.11, we have found there was an
added key_size check, which now requires key_size >=
Thursday, June 28, 2018 3:40 PM, Adrien Mazarguil:
> Subject: Re: [dpdk-dev] [PATCH] net/mlx4: refinements to Rx packet type
> report
>
> On Thu, Jun 28, 2018 at 09:30:28AM +0300, Moti Haimovsky wrote:
> > This commit refines the Rx Packet type flags reported by the PMD for
> > each packet being r
From: "Kudurumalla, Rakesh"
This feature is used to offload stripping of vlan header from recevied
packets and update vlan_tci field in mbuf when
DEV_RX_OFFLOAD_VLAN_STRIP & ETH_VLAN_STRIP_MASK flag is set.
Signed-off-by: Rakesh Kudurumalla
Signed-off-by: Pavan Nikhilesh
---
drivers/net/thund
Hi, Ferruh
> -Original Message-
> From: Yigit, Ferruh
> Sent: Friday, June 29, 2018 7:04 PM
> To: Zhao1, Wei ; dev@dpdk.org
> Cc: Zhang, Qi Z
> Subject: Re: [dpdk-dev] [PATCH v6] net/fm10k: add support for check
> descriptor status APIs
>
> On 6/29/2018 2:48 AM, Wei Zhao wrote:
> > rte_e
The new WRR sequence applicable after queue add/del is set
up after setting the new queue state, so a memory allocation
failure will leave behind an incorrect state.
This change separates the memory sizing + allocation for the
Rx poll and WRR array from calculation of the WRR sequence.
If there is
This patch series adds support for interrupt driven queues to the
ethernet Rx adapter, the first 3 patches prepare the code to
handle both poll and interrupt driven Rx queues, the 4th patch
patch has code changes specific to interrupt driven queues and
the final patch has test code.
Changelog:
v3-
Add test for queue add and delete, the add/delete calls
also switch queues between poll and interrupt mode.
Signed-off-by: Nikhil Rao
---
test/test/test_event_eth_rx_adapter.c | 261 +++---
1 file changed, 242 insertions(+), 19 deletions(-)
diff --git a/test/test/tes
Add support for interrupt driven queues when eth device is
configured for rxq interrupts and servicing weight for the
queue is configured to be zero.
A interrupt driven packet received counter has been added to
rte_event_eth_rx_adapter_stats.
Signed-off-by: Nikhil Rao
---
config/rte_config.h
Add a common prefix to function names and rename
few to better match functionality
Signed-off-by: Nikhil Rao
Acked-by: Jerin Jacob
---
lib/librte_eventdev/rte_event_eth_rx_adapter.c | 167 -
1 file changed, 80 insertions(+), 87 deletions(-)
diff --git a/lib/librte_event
Create a separate function that handles eth receive and
enqueue to event buffer. This function will also be called for
interrupt driven receive queues.
Signed-off-by: Nikhil Rao
Acked-by: Jerin Jacob
---
lib/librte_eventdev/rte_event_eth_rx_adapter.c | 67 ++
1 file chan
https://bugs.dpdk.org/show_bug.cgi?id=67
Bug ID: 67
Summary: multi_process/l2fwd_fork failed to compile
Product: DPDK
Version: 18.05
Hardware: All
OS: All
Status: CONFIRMED
Severity: normal
Prior
If an attached PCI device be attached again, it will cause
rte_pci_device->device.name be corrupted due to unexpected
rte_devargs_remove.
Fixes: 7e8b26650146 ("eal: fix hotplug add / remove")
Cc: sta...@dpdk.org
Signed-off-by: Qi Zhang
---
lib/librte_eal/common/eal_common_dev.c | 21 +++
Add missing NULL point check in ixgbe_pf_host_uninit, or it may cause
segement fault when detach a device.
Fixes: cf80ba6e2038 ("net/ixgbe: add support for representor ports")
Cc: sta...@dpdk.org
Signed-off-by: Qi Zhang
---
drivers/net/ixgbe/ixgbe_pf.c | 15 +--
1 file changed, 9 i
Starting from rdma-core v19, Mellanox OFED 4.4, the Verbs resources
cleanup is properly activated in plug-out process while setting the
MLX5_DEVICE_FATAL_CLEANUP environment variable to 1.
Set the aforementioned variable to 1.
Signed-off-by: Matan Azrad
---
drivers/net/mlx5/mlx5.c | 5 +
1
When link needs to go up, I40E_AQ_PHY_AN_ENABLED is always be set in DPDK.
So all speeds are always set. This causes speed config never works.
This patch fixes this issue and only allows to set available speeds. If
link needs to go up and speed setting is not supported, it will print
warning and s
v8:
- update rte_eal_version.map due to new API added.
- minor reword on release note.
- minor fix on commit log and code style.
NOTE:
Some issues which is not related with this patchset is expected when
play with hotplug_mp sample as belows.
- Attach a PCI device twice may cause device can't
This patch cover the multi-process hotplug case when a share device
attach/detach request be issued from secondary process
device attach on secondary:
a) seconary send sync request to primary.
b) primary receive the request and attach the new device if failed
goto i).
c) primary forward attach
Introduce new API rte_eal_register_mp_init that help to register
a callback function which will be invoked right after multi-process
channel be established (rte_mp_channel_init). Typically the API
will be used by other module that want it's mp channel action callbacks
can be registered during rte_e
Add driver API rte_eth_release_port_private to support the
requirement that an ethdev only be released on secondary process,
so only local state be set to unused, share data will not be
reset so primary process can still use it.
Signed-off-by: Qi Zhang
Reviewed-by: Andrew Rybchenko
Acked-by: Rem
When attach a private device from secondary as the first one, we need
to make sure rte_eth_dev_shared_data is initialized, the patch add
necessary IPC for secondary to inform primary to do initialization.
Signed-off-by: Qi Zhang
---
lib/librte_ethdev/ethdev_mp.c | 2 ++
lib/librte_ethdev/e
We are going to introduce the solution to handle different hotplug
cases in multi-process situation, it include below scenario:
1. Attach a share device from primary
2. Detach a share device from primary
3. Attach a share device from secondary
4. Detach a share device from secondary
5. Attach a pr
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/i40e/i40e_ethdev.c | 2 ++
1 file changed, 2 ins
Introduce API rte_eth_dev_lock and rte_eth_dev_unlock to let
application lock or unlock on specific ethdev, a locked device
can't be detached, this help applicaiton to prevent unexpected
device detaching, especially in multi-process envrionment.
Aslo introduce the new API rte_eth_dev_lock_with_cal
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/ixgbe/ixgbe_ethdev.c | 3 +++
1 file changed, 3
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/kni/rte_eth_kni.c | 11 +++
1 file chang
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/af_packet/rte_eth_af_packet.c | 11 +++
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/null/rte_eth_null.c | 16 +++-
1 fil
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/bonding/rte_eth_bond_pmd.c | 11 +++
1 f
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/octeontx/octeontx_ethdev.c | 16
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/softnic/rte_eth_softnic.c | 19 -
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/pcap/rte_eth_pcap.c | 15 ++-
1 file
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
---
drivers/net/vhost/rte_eth_vhost.c | 11 +++
1 file c
Previously, detach port on secondary process will mess primary
process and cause same device can't be attached again, by take
advantage of rte_eth_release_port_private, we can support this
with minor change.
Signed-off-by: Qi Zhang
Acked-by: Keith Wiles
---
drivers/net/tap/rte_eth_tap.c | 17 ++
The sample code demonstrate device (ethdev only) management
at multi-process envrionment. User can attach/detach a device
on primary process and see it is synced on secondary process
automatically, also user can lock a device to prevent it be
detached or unlock it to go back to default behaviour.
Update release notes for the new multi process hotplug feature.
Signed-off-by: Qi Zhang
---
doc/guides/rel_notes/release_18_08.rst | 20
1 file changed, 20 insertions(+)
diff --git a/doc/guides/rel_notes/release_18_08.rst
b/doc/guides/rel_notes/release_18_08.rst
index bc01
If devices always use descriptors in the same order in which they have
been made available. These devices can offer the VIRTIO_F_IN_ORDER
feature. If negotiated, this knowledge allows devices to notify the use
of a batch of buffers to virtio driver by only writing used ring index.
Vhost user devic
In latest virtio-spec, new feature bit VIRTIO_F_IN_ORDER was introduced.
When this feature has been negotiated, virtio driver will use
descriptors in ring order: starting from offset 0 in the table, and
wrapping around at the end of the table. Vhost devices will always use
descriptors in the same o
Add parameters for configuring VIRTIO_NET_F_MRG_RXBUF and
VIRTIO_F_IN_ORDER feature bits. If feature is disabled, also update
corresponding unsupported feature bit.
Signed-off-by: Marvin Liu
Reviewed-by: Maxime Coquelin
diff --git a/doc/guides/nics/virtio.rst b/doc/guides/nics/virtio.rst
index
If VIRTIO_F_IN_ORDER has been negotiated, driver will use descriptors in
ring order: starting from offset 0 in the table, and wrapping around at
the end of the table. Also introduce use_inorder_[rt]x flag for
selection of IN_ORDER [RT]x handlers.
Signed-off-by: Marvin Liu
Reviewed-by: Maxime Coqu
Add new function for freeing IN_ORDER descriptors. As descriptors will
be allocated and freed sequentially when IN_ORDER feature was
negotiated. There will be no need to utilize chain for freed descriptors
management, only index update is enough.
Signed-off-by: Marvin Liu
Reviewed-by: Maxime Coqu
This patch introduces unsupported features mask for virtio-user device.
For virtio-user server mode, when reconnecting virtio-user will retrieve
vhost device features as base and then unmask unsupported features.
Signed-off-by: Marvin Liu
Reviewed-by: Maxime Coquelin
diff --git a/drivers/net/vi
IN_ORDER virtio-user Tx function support Tx checksum offloading and
TSO which also support on normal Tx function. So extracts common part
into separated function for reuse.
Signed-off-by: Marvin Liu
Reviewed-by: Maxime Coquelin
diff --git a/drivers/net/virtio/virtio_rxtx.c b/drivers/net/virtio/
IN_ORDER Rx function depends on merge-able feature. Descriptors
allocation and free will be done in bulk.
Virtio dequeue logic:
dequeue_burst_rx(burst mbufs)
for (each mbuf b) {
if (b need merge) {
merge remained mbufs
add merged mbuf to
After IN_ORDER Rx/Tx paths added, need to update Rx/Tx path selection
logic.
Rx path select logic: If IN_ORDER and merge-able are enabled will select
IN_ORDER Rx path. If IN_ORDER is enabled, Rx offload and merge-able are
disabled will select simple Rx path. Otherwise will select normal Rx
path.
Signed-off-by: Marvin Liu
Reviewed-by: Maxime Coquelin
diff --git a/drivers/net/virtio/virtio_ethdev.h
b/drivers/net/virtio/virtio_ethdev.h
index cd8070248..350e9ce73 100644
--- a/drivers/net/virtio/virtio_ethdev.h
+++ b/drivers/net/virtio/virtio_ethdev.h
@@ -36,6 +36,7 @@
1ULL << VIRT
HW needs each pool to be mapped to an aura set of 16 auras.
Previously, pool to aura mapping was considered to be 1:1.
Fixes: 02fd6c744350 ("mempool/octeontx: support allocation")
Cc: sta...@dpdk.org
Signed-off-by: Pavan Nikhilesh
Acked-by: Santosh Shukla
---
v2 Changes:
- use macro to avoid
2018-06-28 15:15 GMT+02:00 Rafal Kozik :
>
> Write combining (WC) increases NIC performance by making better
> utilization of PCI bus. ENA PMD may make usage of this feature.
>
> To enable it load igb_uio driver with wc_activate set to 1.
>
> Signed-off-by: Rafal Kozik
> Acked-by: Bruce Richardson
51 matches
Mail list logo