On Wed, Nov 15, 2017 at 09:09:49AM +0200, Shahaf Shuler wrote:
> Mark action flag was not set on the flow, causing the rx burst function
> after port start to ignore the flow mark.
>
> Fixes: 8086cf08b2f0 ("net/mlx5: handle RSS hash configuration in RSS flow")
> Cc: nelio.laranje...@6wind.com
>
>
Mark action flag was not set on the flow, causing the rx burst function
after port start to ignore the flow mark.
Fixes: 8086cf08b2f0 ("net/mlx5: handle RSS hash configuration in RSS flow")
Cc: nelio.laranje...@6wind.com
Signed-off-by: Shahaf Shuler
---
drivers/net/mlx5/mlx5_flow.c | 1 +
1 fil
It will clear all queue region related configuration
when dev stop even if threr is no queue region config,
so this may cause error. So add check when flush queue
region config and delete it when device stop.
Fixes: 7cbecc2f742 ("net/i40e: support queue region set and flush")
Signed-off-by: Wei Z
It will clear all queue region related configuration
when dev stop even if threr is no queue region config,
so this may cause error. So add check when flush queue
region config and delete it when device stop.
Fixes: 7cbecc2f742 ("net/i40e: support queue region set and flush")
Signed-off-by: Wei Z
Hi,
Does anyone have any inputs/suggestion on this issue?
Regards,
Nitin
-Original Message-
From: Nitin Katiyar
Sent: Wednesday, October 25, 2017 11:57 AM
To: 'Stephen Hemminger'
Cc: dev@dpdk.org
Subject: RE: [dpdk-dev] Issue with MTU/max_rx_pkt_len handling by different
NICs/PMD drive
> On Nov 9, 2017, at 5:43 AM, Adrien Mazarguil
> wrote:
>
> This patch removes all code associated with symbols not internally relied
> on by other DPDK components, makes struct cmdline opaque and then proceeds
> to re-implement the remaining functionality as a wrapper to the editline
> librar
HI Greg,
There is no restriction on sending the code to DPDK for public review
and it is up to the respective maintainers to provide comments.
However, your patches can not be merged to DPDK and it's next trees till the
Licensing issues are sorted out.
Regards,
Hemant
> -Origina
16/10/2017 16:27, Ivan Boule:
> To help administrative tasks on ports, new per-port information need
> to be added into the data structure rte_eth_dev_info supplied by the
> dev_infos_get() function exported by a Poll Mode Driver.
>
> See http://dpdk.org/ml/archives/dev/2017-September/074885.html
> > Add deprecation notice to highlight a future change in the
> > default configuration behavior for flexible payload for RSS.
> >
> > Signed-off-by: John McNamara
>
> Acked-by: Andrey Chilikin
It is a specific driver behaviour change notice.
So applied with only one ack, thanks.
14/11/2017 09:24, Olivier MATZ:
> On Tue, Nov 14, 2017 at 06:09:35AM +0100, Thomas Monjalon wrote:
> > > > >> Update types of variables to correspond to nb_segs type change from
> > > > >> uint8_t to uint16_t.
> > > > >>
> > > > >> Fixes: 97cb466d65c9 ("mbuf: use 2 bytes for port and nb segments")
The return value for rte_event_port_{link, unlink}() is defined as the
"number of {links, unlinks} actually established." However, the eventdev
layer's error checking returns negative error values. This commit aligns
the eventdev code with the API definition by having it set rte_errno and
return 0
14/11/2017 20:35, Bruce Richardson:
> On Tue, Nov 14, 2017 at 09:04:24AM +0100, maria.lingem...@ericsson.com wrote:
> > From: Maria Lingemark
> >
> > Fixes: 278f945402c5 ("pdump: add new library for packet capture")
> > Cc: reshma.pat...@intel.com
> >
> > Signed-off-by: Maria Lingemark
> > ---
> On Nov 15, 2017, at 12:32 AM, Hanoch Haim (hhaim) wrote:
>
> I would add this too
>
> - } else if (rte_atomic16_add_return(&m->refcnt_atomic, -1) == 0)
>
> Should be :
>
> + } else if (likely(rte_mbuf_refcnt_update(m, -1) == 0)) {
>
>
> Hanoh
>
Why not to send the separate patch for t
I would add this too
- } else if (rte_atomic16_add_return(&m->refcnt_atomic, -1) == 0)
Should be :
+ } else if (likely(rte_mbuf_refcnt_update(m, -1) == 0)) {
Hanoh
-Original Message-
From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Ilya Matveychikov
Sent: Tuesday, November 14,
On Tue, Nov 14, 2017 at 09:04:24AM +0100, maria.lingem...@ericsson.com wrote:
> From: Maria Lingemark
>
> Fixes: 278f945402c5 ("pdump: add new library for packet capture")
> Cc: reshma.pat...@intel.com
>
> Signed-off-by: Maria Lingemark
> ---
Acked-by: Bruce Richardson
14/11/2017 12:32, Andrew Rybchenko:
> Please, if possible, include these minor fixes in 17.11 release.
>
> Andrew Rybchenko (2):
> net/sfc: avoid Rx queue setup failure if thresholds are set
> net/sfc: avoid Tx queue setup failure if thresholds are set
Applied, thanks
On Tue, 2017-11-14 at 16:04 +, alangordonde...@gmail.com wrote:
> From: Alan Dewar
>
> The WRR code calculates the lowest common denominator between the
> four
> WRR weights as a uint32_t value and divides the LCD by each of the
> WRR
> weights and casts the results as a uint8_t. This castin
14/11/2017 14:52, Nelio Laranjeiro:
> From: Thierry Herbelot
>
> total_length is only visible when SOFT_COUNTERS are enabled
>
> Fixes: 3f13f8c23a7c ("net/mlx5: support hardware TSO")
> Cc: shah...@mellanox.com
> Cc: sta...@dpdk.org
>
> Signed-off-by: Thierry Herbelot
> Acked-by: Nelio Laranje
Hi Tiwei,
I'm doing some regression tests with v17.11-rc4. I ran
into a hitch with testpmd running into a guest VM. It happens
that no packet gets forwarded by testpmd.
The issue seems to appear after this patch was upstreamed.
I saw there's a way to make it work, ie by avoiding to
increment t
Hi All,
I've found a minor issue in the function "rte_pktmbuf_prefree_seg"
This code is located here :
/dpdk/lib/librte_mbuf/rte_mbuf.h
- } else if (rte_atomic16_add_return(&m->refcnt_atomic, -1) == 0)
Should be :
+ } else if (likely(rte_mbuf_refcnt_update(m, -1) == 0)) {
This depends on
On 11/14/2017 3:37 PM, Anoob Joseph wrote:
When security offload is enabled, the packet should be forwarded on the
port configured in the SA. Security session will be configured on that
port only, and sending the packet on other ports could result in
unencrypted packets being sent out.
This wo
From: Alan Dewar
The WRR code calculates the lowest common denominator between the four
WRR weights as a uint32_t value and divides the LCD by each of the WRR
weights and casts the results as a uint8_t. This casting can cause
the ratios of the computed wrr costs to be wrong. For example with
WR
When security offload is enabled, the packet should be forwarded on the
port configured in the SA. Security session will be configured on that
port only, and sending the packet on other ports could result in
unencrypted packets being sent out.
This would have performance improvements too, as the p
From: Thierry Herbelot
total_length is only visible when SOFT_COUNTERS are enabled
Fixes: 3f13f8c23a7c ("net/mlx5: support hardware TSO")
Cc: shah...@mellanox.com
Cc: sta...@dpdk.org
Signed-off-by: Thierry Herbelot
Acked-by: Nelio Laranjeiro
---
drivers/net/mlx5/mlx5_rxtx.c | 2 ++
1 file ch
Hi Nelio,
On 11/14/2017 11:34 AM, Nelio Laranjeiro wrote:
Hi Olivier,
On Tue, Nov 14, 2017 at 10:58:02AM +0100, Olivier Gournet wrote:
Hi,
I can't get TX on secondary process to works with the lastest dpdk, it was
running fine with dpdk-16.11.
RX/TX is ok on primary processs, and I'm not int
Fixes: af75078fece3 ("first public release")
Cc: intel.com
Signed-off-by: Ilya V. Matveychikov
---
There is no reason to have local variable m2 or am I wrong?
lib/librte_mbuf/rte_mbuf.h | 8 +++-
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/lib/librte_mbuf/rte_mbuf.h b/lib
Signed-off-by: Ilya V. Matveychikov
---
drivers/net/bnxt/bnxt_rxq.h | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/net/bnxt/bnxt_rxq.h b/drivers/net/bnxt/bnxt_rxq.h
index 508731eec..be190195a 100644
--- a/drivers/net/bnxt/bnxt_rxq.h
+++ b/drivers/net/bnxt/bnxt_rxq.h
@@ -59,8 +59,6 @
Hi,
Please send a v2 with the doc update that describes the new behavior and I will
ack it.
Regards,
Radu
> -Original Message-
> From: Anoob Joseph [mailto:ajos...@caviumnetworks.com]
> Sent: Monday, November 13, 2017 7:25 PM
> To: Nicolau, Radu ; Anoob Joseph
> ; Akhil Goyal ;
> Dohert
Please, if possible, include these minor fixes in 17.11 release.
Andrew Rybchenko (2):
net/sfc: avoid Rx queue setup failure if thresholds are set
net/sfc: avoid Tx queue setup failure if thresholds are set
drivers/net/sfc/sfc_rx.c | 3 +--
drivers/net/sfc/sfc_tx.c | 3 +--
2 files changed,
Rx queue prefetch, host and writeback thresholds are used for
performance fine-tuning and not applicable to Solarflare NICs.
It is safe to just log warning and do not fail Rx queue setup
to be more friendly to DPDK applications which hardcode it.
Fixes: ce35b05c635e ("net/sfc: implement Rx queue s
Tx queue prefetch, host and writeback thresholds are used for
performance fine-tuning and not applicable to Solarflare NICs.
It is safe to just log warning and do not fail Tx queue setup
to be more friendly to DPDK applications which hardcode it.
Fixes: b1b7ad933b39 ("net/sfc: set up and release T
Report error message if clearing O_NONBLOCK flag will fail,
then return from function.
Coverity issue: 143439
Fixes: ef53b6030039 ("net/virtio-user: support LSC")
Cc: jianfeng@intel.com
Cc: y...@fridaylinux.org
cc: dev@dpdk.org
Signed-off-by: SebastianX Basierski
Acked-by: Jianfeng Tan
Ack
Hi Olivier,
On Tue, Nov 14, 2017 at 10:58:02AM +0100, Olivier Gournet wrote:
> Hi,
>
> I can't get TX on secondary process to works with the lastest dpdk, it was
> running fine with dpdk-16.11.
> RX/TX is ok on primary processs, and I'm not interested in RX on secondary
> process. Each process
Hi,
I can't get TX on secondary process to works with the lastest dpdk, it was
running fine with dpdk-16.11.
RX/TX is ok on primary processs, and I'm not interested in RX on secondary
process. Each process has its
owns TX queues.
I upgraded everything to:
linux 4.14
dpdk 17.11-rc4
rdma-c
Hi Sebastian,
On 11/14/2017 10:27 AM, SebastianX Basierski wrote:
Report error message if clearing O_NONBLOCK flag will fail,
then return from function.
---
v2:
Patch title changed.
---
The above note should be placed just before diffstats to be removed
at apply time.
Coverity i
Report error message if clearing O_NONBLOCK flag will fail,
then return from function.
---
v2:
Patch title changed.
---
Coverity issue: 143439
Fixes: ef53b6030039 ("net/virtio-user: support LSC")
Cc: jianfeng@intel.com
Cc: y...@fridaylinux.org
cc: dev@dpdk.org
Signed-off-by: Sebasti
On Tue, Nov 14, 2017 at 06:09:35AM +0100, Thomas Monjalon wrote:
> > > >> Update types of variables to correspond to nb_segs type change from
> > > >> uint8_t to uint16_t.
> > > >>
> > > >> Fixes: 97cb466d65c9 ("mbuf: use 2 bytes for port and nb segments")
> > > >> Cc: olivier.m...@6wind.com
> > >
From: Maria Lingemark
Fixes: 278f945402c5 ("pdump: add new library for packet capture")
Cc: reshma.pat...@intel.com
Signed-off-by: Maria Lingemark
---
lib/librte_pdump/rte_pdump.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/librte_pdump/rte_pdump.c b/lib/librte_
38 matches
Mail list logo
