On Tue, Mar 3, 2020 at 1:14 PM Ori Kam wrote:
>
> Hi Xiang,
>
> May I ask when do you plan to add the Hyperscan code to the DPDK?
>
> > -Original Message-
> > From: dev On Behalf Of Wang Xiang
> > Sent: Monday, March 2, 2020 9:05 AM
> > To: Ori Kam
> > Cc: jer...@marvell.com; dev@dpdk.or
Hi Xiang,
May I ask when do you plan to add the Hyperscan code to the DPDK?
> -Original Message-
> From: dev On Behalf Of Wang Xiang
> Sent: Monday, March 2, 2020 9:05 AM
> To: Ori Kam
> Cc: jer...@marvell.com; dev@dpdk.org; pbhagavat...@marvell.com; Shahaf
> Shuler ; hemant.agra...@nxp
On 03/03, taox@intel.com wrote:
>From: Zhu Tao
>
>Macro rte_errno is not a static value, so it needs to be updated in all
>error handling code.
>
>Patch 'dc36bd5dfdeb' mistakenly consider that rte_errno is a constant,
>which causes the unrecognized flow rule to be marked as recognition
>succes
Hi
From: xiangxia.m@gmail.com
> From: Tonghao Zhang
>
> When setting the CONFIG_RTE_BUILD_SHARED_LIB to y, and build the mlx5
> pmd, there is a building error.
> To fix it, add RTE_IBVERBS_LINK_DLOPEN to include relative codes.
>
> > mlx5_common.o: In function `mlx5_glue_init':
> > driver
Hi All,
> -Original Message-
> From: Jerin Jacob
> Sent: Monday, March 2, 2020 9:19 AM
> To: Pavan Nikhilesh Bhagavatula
> Cc: Ori Kam ; Jerin Jacob Kollanukkaran
> ; xiang.w.w...@intel.com; dev@dpdk.org; Shahaf Shuler
> ; hemant.agra...@nxp.com; Opher Reviv
> ; Alex Rosenbaum ; Dovrat
>
On 03/03, Qiming Yang wrote:
>Lack of GL_SWR_PM_UP_THR value for mac type 4x10G X722 caused it can't
>reach line rate in one queue one core forwarding case. It's about 1%
>performance drop. This patch add value as:
>
> Device-ID ValueComments
> 0x37D00x03030303 4x10G FPK
>
>This val
Lack of GL_SWR_PM_UP_THR value for mac type 4x10G X722 caused it can't
reach line rate in one queue one core forwarding case. It's about 1%
performance drop. This patch add value as:
Device-ID ValueComments
0x37D00x03030303 4x10G FPK
This value is hardware specific.
Fixes: 3320
From: Zhu Tao
Macro rte_errno is not a static value, so it needs to be updated in all
error handling code.
Patch 'dc36bd5dfdeb' mistakenly consider that rte_errno is a constant,
which causes the unrecognized flow rule to be marked as recognition
success. Later, when the code tried to parse the f
From: Zhu Tao
Macro rte_errno is not a static value, so it needs to be updated in all
error handling code.
Patch 'dc36bd5dfdeb' mistakenly consider that rte_errno is a constant, which
causes the unrecognized flow rule to be marked as recognition success.
Later, when the code tried to parse the f
Hi Xiaolong,
Commit SHA length non-compliance will be modified in V2 patch.
The reason why they did not return immediately after the error is that they
need to release the allocated memory after the error, which is released
uniformly in error processing, so they did not return directly.
BR,
Zhu
From: "Dey, Souvik"
Add support the set_mc_addr_list device operation in the bnx2xvf PMD.
The configured addresses are stored in the device private area, so
they can be flushed before adding new ones.
Without this v6 multicast packets were properly forwarded to the
Guest VF.
Signed-off-by: "De
On Thu, Feb 27, 2020 at 12:02 AM Kalesh A P <
kalesh-anakkur.pura...@broadcom.com> wrote:
> From: Kalesh AP
>
> Driver destroys the vnic when the port is brought down.
> Port hw filter setting such as default mac address and
> unicast mac filters will be applied when port is started.
>
> Fixed to
Remove unnecessary variable "meta" in ice_flow_create and
ice_flow_validate, it should be defined when really be needed:
its ice_parse_engine_create and ice_parse_engine_validate.
In a validate opertion, a meta is not necessary be created during
parser, so NULL will be parsed to low level engine a
On Fri, Feb 28, 2020 at 2:04 AM Kalesh A P <
kalesh-anakkur.pura...@broadcom.com> wrote:
> From: Kalesh AP
>
> There is a race condition when port is closed while error recovery
> is happening in driver. Fixed this by canceling the recovery threads
> during port close.
>
> Fixes: df6cd7c1f73a ("n
On 03/03, taox@intel.com wrote:
>From: Zhu Tao
>
>Macro rte_errno is not a static value, so it needs to be updated in all
>error handling code.
>
>Patch 'dc36bd5dfd' mistakenly consider that rte_errno is a constant, which
>causes the unrecognized flow rule to be marked as recognition success.
test_dep_obj is being formatted as string object, so can't be used as static
obj..
Amit
> -Original Message-
> From: Bruce Richardson
> Sent: Monday, March 2, 2020 4:23 PM
> To: Amit Gupta
> Cc: Harman Kalra ; dev@dpdk.org; sta...@dpdk.org
> Subject: [EXT] Re: [dpdk-dev] [PATCH v2 1/1
> -Original Message-
> From: Ye, Xiaolong
> Sent: Tuesday, March 3, 2020 10:53 AM
> To: Zhang, Xiao
> Cc: Yigit, Ferruh ; dev@dpdk.org; Xing, Beilei
> ; Zhang, Qi Z ; sta...@dpdk.org
> Subject: Re: [dpdk-dev] net/i40e: add promiscuous configure unsupported check
>
> On 03/03, Zhang, Xi
On 03/03, Zhang, Xiao wrote:
>Hi Ferruh,
>
>> -Original Message-
>> From: Yigit, Ferruh
>> Sent: Monday, March 2, 2020 5:09 PM
>> To: Zhang, Xiao ; dev@dpdk.org
>> Cc: Xing, Beilei ; Zhang, Qi Z ;
>> sta...@dpdk.org
>> Subject: Re: [dpdk-dev] net/i40e: add promiscuous configure unsupported
From: Zhu Tao
Macro rte_errno is not a static value, so it needs to be updated in all
error handling code.
Patch 'dc36bd5dfd' mistakenly consider that rte_errno is a constant, which
causes the unrecognized flow rule to be marked as recognition success.
Later, when the code tried to parse the flo
Hi Ferruh,
> -Original Message-
> From: Yigit, Ferruh
> Sent: Monday, March 2, 2020 5:09 PM
> To: Zhang, Xiao ; dev@dpdk.org
> Cc: Xing, Beilei ; Zhang, Qi Z ;
> sta...@dpdk.org
> Subject: Re: [dpdk-dev] net/i40e: add promiscuous configure unsupported check
>
> On 2/28/2020 3:22 AM, Xiao
+Xiaolong
BR,
Zhu, Tao
> -Original Message-
> From: Zhu, TaoX
> Sent: Wednesday, February 26, 2020 8:07 PM
> To: Ananyev, Konstantin ; Lu, Wenzhuo
>
> Cc: dev@dpdk.org; Zhu, TaoX ; sta...@dpdk.org
> Subject: [PATCH] net/ixgbe: fix link status inconsistencies
>
> From: Zhu Tao
>
> Sett
Hi, Qiming
On 02/26, Qiming Yang wrote:
>GL_SWR_PM_UP_THR value lack in X722 caused configure register write not be
>called. This is the root cause of performance drop. This patch add support
>for:
Could you elaborate what kind of performance drop caused by the missing
register
cfg?
Thanks,
Xi
In case of bnx2xvf pmd, tx packets can support vland id in 2 ways :
1. setting the mbuf ol_flags=PKT_TX_VLAN_PKT and passing the
vlanid in mbuf->vlan_tci.
2. the tx packet itself has the vlan id included in the packet.
The first case is working as expected but the second case where
the vlan id is i
If a vhost device is closed before eth_dev_configure is done
to the device, internal resources allocated to the device
does not freed. This patch fixes it.
Fixes: 3d01b759d267 ("net/vhost: delay driver setup")
Signed-off-by: Itsuro Oda
---
drivers/net/vhost/rte_eth_vhost.c | 16 +++-
> -Original Message-
> From: Ye, Xiaolong
> Sent: Monday, March 2, 2020 4:20 PM
> To: Zhang, Qi Z
> Cc: Xing, Beilei ; Cao, Yahui ;
> dev@dpdk.org; sta...@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] net/ice: remove unnecessary variable
>
> Hi, Qi
>
> On 02/25, Qi Zhang wrote:
> >Remov
If running DPDK as non-root, some extra capabilities may be required.
The Mellanox devices, using a bifurcated model with Linux drivers,
have some specific requirements summarized in mlx5 PMD guide.
Signed-off-by: Thomas Monjalon
---
doc/guides/nics/mlx5.rst | 26 ++
1 fi
- When running the dedault configuration of autotest of the make test
it'd take 900 seconds (15 minutes) for the script TIMEOUT and marks
Failed for a specific device test that is not supported on the
system under test.
- Adding the checking for those tests, print out as "Skipped [Not Run]"
> > > +/**
> > > + * @internal Reserve ring elements to enqueue several objects on the
> > > +ring
> > > + *
> > > + * @param r
> > > + * A pointer to the ring structure.
> > > + * @param esize
> > > + * The size of ring element, in bytes. It must be a multiple of 4.
> > > + * This must be
Jan and Rastislav are leaving Netcope. Martin and Jakub
will replace them as Netcope PMD maintainers.
Signed-off-by: Jan Remes
---
MAINTAINERS | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index f4e0ed8e0..69ce71ffb 100644
--- a/MAINTAINERS
02/03/2020 17:34, Jan Remes:
> Jan and Rastislav are leaving Netcope. Martin and Jakub
> will replace them as Netcope PMD maintainers.
>
> Signed-off-by: Jan Remes
Applied
Add an new device argument 'no-rx', which will prevent PMD receiving
packets.
This is useful for testing when a PMD is needed only to send packets to.
Signed-off-by: Ferruh Yigit
---
doc/guides/nics/null.rst| 4 +++
drivers/net/null/rte_eth_null.c | 55 -
Secondary process uses the primary process device and while setting the
Rx/Tx functions it uses the device arguments from the secondary process
instead of the primary ones.
This may cause primary and secondary process use different Rx/Tx
functions unintentionally.
Fixes: bccc77a6a74a ("net/null:
Net null PMD was missing documentation, adding it.
Signed-off-by: Ferruh Yigit
---
doc/guides/nics/index.rst | 1 +
doc/guides/nics/null.rst | 39 +++
2 files changed, 40 insertions(+)
create mode 100644 doc/guides/nics/null.rst
diff --git a/doc/guides/nic
Group device argument to the struct, to increase code readability.
Signed-off-by: Ferruh Yigit
---
drivers/net/null/rte_eth_null.c | 31 ++-
1 file changed, 18 insertions(+), 13 deletions(-)
diff --git a/drivers/net/null/rte_eth_null.c b/drivers/net/null/rte_eth_null
There is no need to check if the argument exist or not,
`rte_kvargs_process` returns success if the argument is not provided at
all.
Fixes: c743e50c475f ("null: new poll mode driver")
Cc: sta...@dpdk.org
Signed-off-by: Ferruh Yigit
---
Cc: muk...@igel.co.jp
---
drivers/net/null/rte_eth_null.c |
Prefer 'unsigned int' storage type keyword against 'unsigned', this also
silence the checkpatch warnings.
Signed-off-by: Ferruh Yigit
---
drivers/net/null/rte_eth_null.c | 46 -
1 file changed, 23 insertions(+), 23 deletions(-)
diff --git a/drivers/net/null/rte_e
25/02/2020 19:01, Maxime Coquelin:
> On 2/21/20 2:35 PM, Xiaolong Ye wrote:
> > I volunteer to be co-maintainer for Vhost-user/Vhost PMD/Virtio PMD.
> >
>
> Thanks for offering your help, I really appreciate it!
>
> Acked-by: Maxime Coquelin
Applied, thanks
On 3/2/2020 4:43 PM, Stephen Hemminger wrote:
> On Mon, 2 Mar 2020 16:24:23 +
> Ferruh Yigit wrote:
>
>> On 2/27/2020 8:56 PM, John W. Linville wrote:
>>> On Thu, Feb 27, 2020 at 12:00:03PM -0800, Stephen Hemminger wrote:
Since qpairs is part of the vdev arguments, there is no need to
25/02/2020 18:51, Maxime Coquelin:
> On 2/21/20 2:28 PM, Tiwei Bie wrote:
> > I'm leaving Intel, and I'm not sure when I could dedicate enough
> > time to DPDK in the future. So I'm removing my name from the
> > MAINTAINERS file.
> >
> > Signed-off-by: Tiwei Bie
> > ---
> > Xiaolong has been an a
On Mon, 2 Mar 2020 16:24:23 +
Ferruh Yigit wrote:
> On 2/27/2020 8:56 PM, John W. Linville wrote:
> > On Thu, Feb 27, 2020 at 12:00:03PM -0800, Stephen Hemminger wrote:
> >> Since qpairs is part of the vdev arguments, there is no need to
> >> limit it to 16. The queue arrays can be dynamica
On 2/27/2020 8:56 PM, John W. Linville wrote:
> On Thu, Feb 27, 2020 at 12:00:03PM -0800, Stephen Hemminger wrote:
>> Since qpairs is part of the vdev arguments, there is no need to
>> limit it to 16. The queue arrays can be dynamically sized based
>> on the requested parameters.
>>
>> Signed-off-b
On 2/28/2020 4:52 PM, John W. Linville wrote:
> On Fri, Feb 28, 2020 at 10:08:43AM +, Ferruh Yigit wrote:
>> On 2/27/2020 8:00 PM, Stephen Hemminger wrote:
>>> Since qpairs is part of the vdev arguments, there is no need to
>>> limit it to 16. The queue arrays can be dynamically sized based
>>>
On Mon, 2 Mar 2020 10:40:17 -0500
Min Tang wrote:
> Hi Stephen:
>
> If there is no intention to process the response message of NVS_TYPE_RNDIS,
> would it be better to not set the flags to VMBUS_CHANPKT_FLAG_RC so that it
> won't receive any response message?
>
> Best Regards,
> Min Tang
>
> O
Hi Stephen:
If there is no intention to process the response message of NVS_TYPE_RNDIS,
would it be better to not set the flags to VMBUS_CHANPKT_FLAG_RC so that it
won't receive any response message?
Best Regards,
Min Tang
On Sun, Mar 1, 2020 at 12:54 PM Stephen Hemminger <
step...@networkplumbe
Hi Savinay,
On 02/03/2020 11:43, Savinay Dharmappa wrote:
test app to measures the performance of libipsec
api rte_ipsec_pkt_crypto and rte_ipsec_pkt_process.
Signed-off-by: Savinay Dharmappa
---
app/test/Makefile | 2 +-
app/test/meson.build | 2 +
app/test/test_ipsec_p
On Mon, Mar 2, 2020 at 7:27 AM wrote:
>
> From: Tonghao Zhang
>
> The order of mempool initiation affects mempool index in the
> rte_mempool_ops_table. For example, when building APPs with:
>
> $ gcc -lrte_mempool_bucket -lrte_mempool_ring ...
>
> The "bucket" mempool will be registered firstly,
On Mon, Mar 2, 2020 at 10:26 AM Hemant Agrawal wrote:
>
> This patch series add various patches for enhancing and fixing NXP
> fslmc bus, dpaa bus, and dpaax.
>
> - the main change is support to allow thread migration across lcores
> - improving the multi-process support
This series triggers an A
On Mon, Mar 2, 2020 at 12:43 PM Savinay Dharmappa
wrote:
>
> test app to measures the performance of libipsec
> api rte_ipsec_pkt_crypto and rte_ipsec_pkt_process.
>
> Signed-off-by: Savinay Dharmappa
Missing MAINTAINERS update.
> ---
> app/test/Makefile | 2 +-
> app/test/meson.bu
On 3/1/2020 7:56 AM, Ori Kam wrote:
>
>
>> -Original Message-
>> From: Xiao Zhang
>> Subject: [v2] app/testpmd: parse flow command line for AH
>>
>> Add AH and AH SPI to testpmd rte flow command line.
>> Add note and sample AH rules in testpmd guide.
>>
>
> Acked-by: Ori Kam
>
Applie
test app to measures the performance of libipsec
api rte_ipsec_pkt_crypto and rte_ipsec_pkt_process.
Signed-off-by: Savinay Dharmappa
---
app/test/Makefile | 2 +-
app/test/meson.build | 2 +
app/test/test_ipsec_perf.c | 637 +
3 files chang
Hi all,
Here is a list of patches targeted for stable release 19.11.1.
The planned date for the final release is March the 16th.
Please help with testing and validation of your use cases and report
any issues/results with reply-all to this mail. For the final release
the fixes and reported valid
On Mon, Mar 2, 2020 at 5:47 PM Ali Alnubani wrote:
>
> > > Do you mind sharing which OS are you using, which gcc/clang version,
> > whether you have MLNX_OFED or rdma-core installed, and which version?
> > 1. CentOS Linux release 7.6.1810 (Core)
> > 2. kernel 3.10.0-957.1.3.el7.x86_64
> > 3. OFED
From: Tonghao Zhang
When setting the CONFIG_RTE_BUILD_SHARED_LIB to y,
and build the mlx5 pmd, there is a building error.
To fix it, add RTE_IBVERBS_LINK_DLOPEN to include
relative codes.
> mlx5_common.o: In function `mlx5_glue_init':
> drivers/common/mlx5/mlx5_common.c:324: undefined reference
test app to measures the performance of libipsec
api rte_ipsec_pkt_crypto and rte_ipsec_pkt_process.
Signed-off-by: Savinay Dharmappa
---
app/test/Makefile | 2 +-
app/test/meson.build | 2 +
app/test/test_ipsec_perf.c | 637 +
3 files chang
On Mon, Mar 02, 2020 at 12:01:55PM +0530, agup...@marvell.com wrote:
> From: Amit Gupta
>
> Add a additional condition to check if all required internal
> dependencies are met before building octeontx drivers using meson.
>
> Bugzilla ID: 387
>
> Fixes: 7f615033d64f ("drivers/net: build Cavium
> > Do you mind sharing which OS are you using, which gcc/clang version,
> whether you have MLNX_OFED or rdma-core installed, and which version?
> 1. CentOS Linux release 7.6.1810 (Core)
> 2. kernel 3.10.0-957.1.3.el7.x86_64
> 3. OFED 4.7-3.2.9
> 4. FW 16.26.4012
> 5. gcc (GCC) 4.8.5 20150623 (Red
From: Nipun Gupta
When IOVA is physical address do not prefetch the annotation
of the next frame, as there is a cost involved there to convert
the physical address to virtual address.
Signed-off-by: Nipun Gupta
---
drivers/bus/fslmc/portal/dpaa2_hw_pvt.h | 6 ++--
drivers/net/dpaa2/dpaa2_rxtx
From: Rohit Raj
Enabled set link status API to start/stop phy
device from application.
Signed-off-by: Rohit Raj
---
drivers/bus/dpaa/base/qbman/process.c | 35 ---
drivers/bus/dpaa/include/process.h| 3 ++
drivers/bus/dpaa/rte_bus_dpaa_version.map | 1 +
drive
From: Rohit Raj
Enable/disable link state interrupt and get link state api is
defined using IOCTL calls.
Signed-off-by: Rohit Raj
---
drivers/bus/dpaa/base/fman/fman.c | 4 +-
drivers/bus/dpaa/base/qbman/process.c | 68 ++-
drivers/bus/dpaa/dpaa_bus.c
From: Nipun Gupta
This patch changes the qbman I/O function names as they are
only reading from cinh register, but writing to cena registers.
This gives way to add functions which purely work in cinh mode
Signed-off-by: Nipun Gupta
---
drivers/bus/fslmc/qbman/qbman_portal.c | 34 +
From: Sachin Saxena
Handle 2.5Gbps ethernet ports as well.
Signed-off-by: Sachin Saxena
Signed-off-by: Gagandeep Singh
---
drivers/bus/dpaa/base/fman/fman.c | 6 --
drivers/bus/dpaa/base/fman/netcfg_layer.c | 3 ++-
drivers/bus/dpaa/include/fman.h | 1 +
drivers/net/dpaa
From: Gagandeep Singh
Enable congestion handling/tail drop for TX queues.
Signed-off-by: Gagandeep Singh
---
drivers/bus/dpaa/base/qbman/qman.c| 43 +
drivers/bus/dpaa/include/fsl_qman.h | 17
drivers/bus/dpaa/rte_bus_dpaa_version.map | 7 ++
drivers/net/dpaa/dpa
From: Nipun Gupta
multiple buffer pools are not supported on a single device.
Signed-off-by: Nipun Gupta
---
drivers/mempool/dpaa/dpaa_mempool.c | 1 +
drivers/net/dpaa/dpaa_ethdev.c | 6 ++
2 files changed, 7 insertions(+)
diff --git a/drivers/mempool/dpaa/dpaa_mempool.c
b/drivers/
From: Nipun Gupta
For DPAA devices the memory maps stored in the FMAN interface
information is per process. Store them in the device process specific
area.
Signed-off-by: Nipun Gupta
---
drivers/net/dpaa/dpaa_ethdev.c | 207 -
drivers/net/dpaa/dpaa_ethdev.h |
From: Nipun Gupta
The change adds printing the thread id when portal allocation
failure occurs
Signed-off-by: Nipun Gupta
---
drivers/crypto/dpaa2_sec/dpaa2_sec_dpseci.c | 8 ++--
drivers/event/dpaa2/dpaa2_eventdev.c| 8 ++--
drivers/mempool/dpaa2/dpaa2_hw_mempool.c| 12 +
Add the error handling on failure.
Signed-off-by: Nipun Gupta
Signed-off-by: Hemant Agrawal
---
drivers/bus/fslmc/portal/dpaa2_hw_dpio.c | 28 ++--
1 file changed, 16 insertions(+), 12 deletions(-)
diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
b/drivers/bus/fslmc/p
From: Nipun Gupta
The patch reworks the portal allocation which was previously
being done on per lcore basis to a per thread basis.
Now user can also create its own threads and use DPAA2 portals
for packet I/O.
Signed-off-by: Nipun Gupta
---
drivers/bus/fslmc/Makefile | 1 +
dr
From: Nipun Gupta
The patch adds support for portal migration by disabling stashing
for the portals which is used in the non-affined threads, or on
threads affined to multiple cores
Signed-off-by: Nipun Gupta
---
drivers/bus/fslmc/portal/dpaa2_hw_dpio.c | 83 +--
.../fslmc/qbman/include/
From: Nipun Gupta
The destructor was being called for non-dpaa2 as well
Signed-off-by: Nipun Gupta
---
drivers/bus/fslmc/portal/dpaa2_hw_dpio.c | 11 ++-
1 file changed, 10 insertions(+), 1 deletion(-)
diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
b/drivers/bus/fslmc/portal/d
This patch series add various patches for enhancing and fixing NXP
fslmc bus, dpaa bus, and dpaax.
- the main change is support to allow thread migration across lcores
- improving the multi-process support
Apeksha Gupta (1):
bus/fslmc: fix dereferencing null pointer
Gagandeep Singh (2):
bus
From: Gagandeep Singh
This is to reduce the thread local storage
Signed-off-by: Gagandeep Singh
---
drivers/bus/fslmc/fslmc_bus.c| 2 --
drivers/bus/fslmc/portal/dpaa2_hw_dpio.h | 7 +++
drivers/bus/fslmc/portal/dpaa2_hw_pvt.h | 8
drivers/bus/fslmc/rte_fslmc.h
From: Rohit Raj
Fixed 10g port negotiation issue with another 10G/non 10G port.
Initialize the port link speed.
Fixes: c5acbb5ea20e ("net/dpaa2: support link status event")
Cc: sta...@dpdk.org
Signed-off-by: Rohit Raj
---
drivers/net/dpaa2/dpaa2_ethdev.c | 4 +---
1 file changed, 1 insertion(
From: Apeksha Gupta
This patch fixees the nxp internal coverity reported
null pointer dereferncing issue.
Fixes: 6fef517e17cf ("bus/fslmc: add qman HW fq query count API")
Cc: sta...@dpdk.org
Signed-off-by: Apeksha Gupta
---
drivers/bus/fslmc/qbman/qbman_debug.c | 9 +
1 file changed,
On 2/28/2020 3:22 AM, Xiao Zhang wrote:
> Return ENOTSUP error code when configuring i40evf promiscuous mode to
> fix port start hang issue on platforms which are unsupported to configure
> promiscuous mode.
Hi Xiao,
What is the cause of the hang, was the application keep trying because of the
"-
From: Alex Marginean
Move from doing batch refill of Rx ring from bundles of 8 to once per
enetc_clean_rx_ring call. One benefit is that we're cleaning up all the
BDs that we just processed, which should still be cached. The other is
that hardware Rx index stays a little back and doesn't cause
From: Alex Marginean
Use rte_pktmbuf_free_bulk to release all mbufs at once. This is flagged
as obsolete/not yet stable in DPDK but seems to be functional.
Don't count the released frames, it's no longer needed in the caller.
Signed-off-by: Alex Marginean
---
drivers/net/enetc/Makefile |
From: Alex Marginean
This was left to its default value. With the patch transactions are:
- coherent,
- do not allocate in downstream cache (there is none on LS1028a),
- merge surrounding data for BD writes,
- overwrite surrounding data for frame data writes.
Signed-off-by: Alex Marginean
---
From: Alex Marginean
LS1028A does not have platform cache so any reads following a hardware
write will go directly to DDR. Latency of such a read is in excess of 100
core cycles, so try to prefetch more in advance to mitigate this.
How much is worth prefetching really depends on traffic conditio
From: Alex Marginean
Since we know in advance that we're going to fill in multiple descriptors
it's convenient to allocate the buffers in batches.
Signed-off-by: Alex Marginean
---
drivers/net/enetc/enetc_rxtx.c | 21 +
1 file changed, 17 insertions(+), 4 deletions(-)
diff
From: Alex Marginean
Reorder the members of the structure so that the ones used on datapath fit
in a single cache line, to slightly reduce pressure on cache and miss rate.
Signed-off-by: Alex Marginean
---
drivers/net/enetc/enetc.h | 10 +-
1 file changed, 5 insertions(+), 5 deletions(
From: Alex Marginean
Each call to enetc_clean_tx_ring will cost at least 150-200 CPU cycles
even if no clean-up is done, due to the CI register read.
We're only calling it once at the end of the function, on the assumption
that software is slower than hardware and hardware completed sending older
From: Alex Marginean
The default value in hardware for the Rx MAC FIFO (@) is higher than it
should be and can lead to Rx lock-up under traffic.
Set it to the value recommended by hardware team, 1.
Signed-off-by: Alex Marginean
Signed-off-by: Gagandeep Singh
---
drivers/net/enetc/base/enetc_h
From: Alex Marginean
Don't read the hardware CI register in a loop, read it once, clean up and
exit.
The issue with reading the register in a loop is that we're stalling here
trying to catch up with hardware which keeps sending traffic as long as it
has traffic to send, so in effect we could be w
This patch series includes patches to optimize and clean
the network driver for ENETC
Alex Marginean (10):
net/enetc: do not stall in clean Tx ring
net/enetc: use relaxed read for Tx CI in clean Tx
net/enetc: batch process enetc clean Tx ring calls
net/enetc: erratum wa for Rx lock-up issu
From: Alex Marginean
We don't need barriers here since this read doesn't have to be strictly
serialized in relation to other surrounding memory/register accesses.
We only want a reasonably recent value out of hardware so we know how much
we can clean.
Signed-off-by: Alex Marginean
---
drivers/
Hi Anoob,
I will merge this series in this week.
Regards,
Akhil
> -Original Message-
> From: Anoob Joseph
> Sent: Monday, March 2, 2020 2:17 PM
> To: Akhil Goyal ; Konstantin Ananyev
>
> Cc: Jerin Jacob Kollanukkaran ; Narayana Prasad Raju
> Athreya ; Ankur Dwivedi ;
> Archana Munigant
Hi Akhil, Konstantin,
Are there any more comments? Or can we have the patches merged?
Thanks,
Anoob
> -Original Message-
> From: Lukasz Bartosik
> Sent: Thursday, February 27, 2020 9:48 PM
> To: Akhil Goyal ; Radu Nicolau
> ; Thomas Monjalon
> Cc: Jerin Jacob Kollanukkaran ; Narayana P
Hi, Qi
On 02/25, Qi Zhang wrote:
>Remove unnecessary variable "meta" in ice_flow_create and
>ice_flow_validate, it should be defined when really be needed:
>its ice_parse_engine_create and ice_parse_engine_validate.
>
>The patch also move the meta's memory free from each filter
s/move/moves
>eng
From: Amaranath Somalapuram
Adding api for eth_dev_supported_ptypes_get_t
Signed-off-by: Amaranath Somalapuram
---
drivers/net/axgbe/axgbe_ethdev.c | 37
1 file changed, 37 insertions(+)
diff --git a/drivers/net/axgbe/axgbe_ethdev.c b/drivers/net/axgbe/axgbe_e
From: Amaranath Somalapuram
Adding API for axgbe_rxq_info_get, axgbe_txq_info_get
Signed-off-by: Amaranath Somalapuram
---
drivers/net/axgbe/axgbe_ethdev.c | 30 ++
1 file changed, 30 insertions(+)
diff --git a/drivers/net/axgbe/axgbe_ethdev.c b/drivers/net/axgbe/a
From: Amaranath Somalapuram
Adding API axgbe_dev_rx_descriptor_status, axgbe_dev_tx_descriptor_status
Signed-off-by: Amaranath Somalapuram
---
drivers/net/axgbe/axgbe_ethdev.c | 2 ++
drivers/net/axgbe/axgbe_rxtx.c | 46
drivers/net/axgbe/axgbe_rxtx.h | 2
From: Amaranath Somalapuram
Adding api for flow_ctrl_get and flow_ctrl_set.
By default axgbe driver flow control is disabled.
Adding dpdk flow control to set water high and low.
Signed-off-by: Amaranath Somalapuram
---
drivers/net/axgbe/axgbe_ethdev.c | 84
dri
From: Amaranath Somalapuram
Adding API for priority_flow_ctrl_set.
Priority flow control to set water high and low,
pause_time and priority.
Signed-off-by: Amaranath Somalapuram
---
drivers/net/axgbe/axgbe_ethdev.c | 94
1 file changed, 94 insertions(+)
diff
On Mon, Mar 2, 2020 at 3:35 PM Ali Alnubani wrote:
>
> > -Original Message-
> > From: Tonghao Zhang
> > Sent: Monday, March 2, 2020 2:23 AM
> > To: Ali Alnubani
> > Cc: sta...@dpdk.org; Matan Azrad ; dev@dpdk.org
> > Subject: Re: [dpdk-dev] [dpdk v1] net/mlx5: fix possible building error
94 matches
Mail list logo
