Hi:
> -Original Message-
> From: zhouyangchao [mailto:zhouya...@gmail.com]
> Sent: Wednesday, November 29, 2017 2:12 PM
> To: dev@dpdk.org
> Cc: qi.z.zhang@intel.comqi.z.zh...@intel.com
> Subject: [PATCH] net/fm10k: remove RSS restriction with num of queues
>
> FM10K HW does not have such
FM10K HW does not have such restrictions.
Signed-off-by: Yangchao Zhou
---
drivers/net/fm10k/fm10k_ethdev.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/fm10k/fm10k_ethdev.c b/drivers/net/fm10k/fm10k_ethdev.c
index 2d05a46..403f6c5 100644
--- a/drivers/net
Hi Akhil, Radu,
Is there any update on how we should approach this?
Thanks,
Anoob
On 11/24/2017 05:52 PM, Anoob wrote:
Hi Akhil, Radu
Please see inline.
Thanks,
Anoob
On 11/24/2017 05:04 PM, Akhil Goyal wrote:
Hi Radu,
On 11/24/2017 4:47 PM, Radu Nicolau wrote:
On 11/24/2017 10:55 A
Hi Gabriel,
Comments inline
On Tue, Nov 28, 2017 at 11:40:06AM -0600, Erik Gabriel Carrillo wrote:
> +struct rte_event_timer_adapter *
> +rte_event_timer_adapter_create(const struct rte_event_timer_adapter_conf
> *conf)
> +{
> + /* default port conf values */
> + struct rte_event_port_c
Hi Akhil,
On 24-11-2017 16:19, Akhil Goyal wrote:
Hi Anoob,
On 11/24/2017 3:28 PM, Anoob wrote:
static inline void
route4_pkts(struct rt_ctx *rt_ctx, struct rte_mbuf *pkts[],
uint8_t nb_pkts)
{
uint32_t hop[MAX_PKT_BURST * 2];
uint32_t dst_ip[MAX_PKT_BURST * 2];
+ int32
-Original Message-
> Date: Thu, 9 Nov 2017 12:22:37 +0530
> From: Abhinandan Gujjar
> To: jerin.ja...@caviumnetworks.com
> CC: dev@dpdk.org, Abhinandan Gujjar , Nikhil
> Rao
> Subject: [RFC] eventdev: add caps API and PMD callback for crypto adapter
> X-Mailer: git-send-email 1.9.1
>
>
Internal variable containing the number of TX queues for a device,
was being incorrectly assigned the number of RX queues, instead of TX.
Fixes: 27b609cbd1c6 ("ethdev: move the multi-queue mode check to specific
drivers")
Signed-off-by: Yangchao Zhou
---
drivers/net/e1000/igb_ethdev.c | 2 +-
Vlan offload flag "rx_mode.hw_vlan_extend" should be considerred at
rte_eth_dev_configure stage as well as "rx_mode.hw_vlan_strip" and
"rx_mode.hw_vlan_filter".
Fixes: 4861cde46116 ("i40e: new poll mode driver")
Cc: sta...@dpdk.org
Signed-off-by: Qi Zhang
---
drivers/net/i40e/i40e_ethdev.c | 6
On 11/23/2017 8:53 AM, Bruce Richardson wrote:
> +ifeq ($(CC_AVX2_SUPPORT), 1)
> + SRCS-$(CONFIG_RTE_LIBRTE_ACL) += i40e_rxtx_vec_avx2.c
Minor issue:
s/CONFIG_RTE_LIBRTE_ACL/CONFIG_RTE_LIBRTE_I40E_INC_VECTOR/
> +endif
On 11/20/2017 3:59 AM, Shijith Thotton wrote:
> Support for re-configuration of number of queues per port and descriptor
> size. Renamed variable representing number of descriptors as nb_desc
> from max_count.
>
> Signed-off-by: Shijith Thotton
Applied to dpdk-next-net/master, thanks.
On 11/20/2017 3:01 AM, Andrew Rybchenko wrote:
> From: Ivan Malov
>
> Falcon support has been withdrawn from libefx, however, there is still
> an obsolete Falcon-specific assertion that efx_mac_stats_upload()
> and efx_port_poll() aren't concurrent. To be consistent with an overall
> Falcon suppo
build error:
.../dpdk/build/build/lib/librte_eal/linuxapp/kni/igb_main.c:2809:2:
error: implicit declaration of function ‘setup_timer’;
did you mean ‘sk_stop_timer’? [-Werror=implicit-function-declaration]
setup_timer(&adapter->watchdog_timer, &igb_watchdog,
^~~
sk_stop_timer
cc1: a
On 11/14/2017 4:42 AM, Ilya Matveychikov wrote:
Please add maintainer to cc next time.
> Signed-off-by: Ilya V. Matveychikov
Reviewed-by: Ferruh Yigit
On 5/3/2017 4:21 AM, Hemant Agrawal wrote:
> This patch adds following:
> 1. option to configure the mac address during create
> 2. inform usespace, if mac address is being changed in linux
>
> Signed-off-by: Hemant Agrawal
Hi Hemant,
This RFC is waiting in patchwork for a while, is there plan
When a PMD finishes probing, it creates the new port by calling
the function rte_eth_dev_allocate().
A notification of the new port is sent there to the upper layer.
When a PMD finishes removal of a port, it calls the function
rte_eth_dev_release_port().
A notification of the destroyed port is sen
The pointer to the user parameter of the callback registration is
automatically pass to the callback function.
There is no point to allow changing this user parameter by a caller.
That's why this parameter is always set to NULL by PMDs and set only
in ethdev layer before calling the callback functi
From: Stephen Hemminger
The DPDK needs to stay up to date with current LTS Linux kernel support.
If the kernel is older than LTS it is likely to be insecure and buggy.
Therefore only require DPDK to work on oldest LTS kernel.
If distribution vendors want to support DPDK on older kernels, that is
On 9/7/2017 1:17 AM, Wu, Jingjing wrote:
>>>
>>> Since dealing with device is kind of new, it can be OK to create new
>>> command tree, but there are already hotplug commands per port:
>>> "port attach #PCI|#VDEV_NAME"
>>> "port detach #P"
>>>
>>
>> Those two commands deal with the etherdev hotplug
On 11/28/2017 2:17 AM, Andrew Rybchenko wrote:
> On 11/27/2017 10:58 PM, Ferruh Yigit wrote:
>> On 11/16/2017 12:04 AM, Andrew Rybchenko wrote:
>>> From: Mark Spender
>>>
>> What has been fixed here?
>> If you can provide more details I can amend the commit later.
>
> EFSYS_PROBE1 takes one typed
On Tue, Nov 28, 2017 at 5:50 PM, Walker, Benjamin wrote:
> On Tue, 2017-11-28 at 14:16 +, Alejandro Lucero wrote:
> >
> >
> > On Mon, Nov 27, 2017 at 5:58 PM, Walker, Benjamin <
> benjamin.wal...@intel.com>
> > wrote:
> > > On Sun, 2017-11-05 at 01:17 +0100, Thomas Monjalon wrote:
> > > > Hi,
On Tue, 2017-11-28 at 14:16 +, Alejandro Lucero wrote:
>
>
> On Mon, Nov 27, 2017 at 5:58 PM, Walker, Benjamin
> wrote:
> > On Sun, 2017-11-05 at 01:17 +0100, Thomas Monjalon wrote:
> > > Hi, restarting an old topic,
> > >
> > > 05/01/2017 16:52, Tan, Jianfeng:
> > > > On 1/5/2017 5:34 AM, W
This commit adds the logic that is shared by all event timer adapter
drivers; the common code handles instance allocation and some
initialization.
Signed-off-by: Erik Gabriel Carrillo
---
drivers/event/sw/sw_evdev.c| 16 +
lib/librte_eventdev/Makefile
This commit adds enough test infrastructure to exercise the
allocation/deallocation routines of the event timer adapter library
minimally.
Signed-off-by: Erik Gabriel Carrillo
---
test/test/Makefile | 1 +
test/test/test_event_timer_adapter.c | 249 +++
If an eventdev PMD does not wish to provide event timer adapter ops
definitions, the library will fall back to a default software
implementation whose entry points are added by this commit.
Signed-off-by: Erik Gabriel Carrillo
---
lib/librte_eventdev/Makefile | 1 +
lib/librte
Signed-off-by: Erik Gabriel Carrillo
---
doc/api/doxy-api-index.md | 1 +
lib/librte_eventdev/Makefile | 1 +
lib/librte_eventdev/rte_event_timer_adapter.h | 427 ++
lib/librte_eventdev/rte_eventdev.h| 4 +-
4 files ch
This set of RFC patches contains the same changes as the previous version,
but they are now broken up into multiple patches for easier review.
This patch set produces the same checkpatch warnings as the previous version:
"macro with flow control" and "avoid externs". I have left the macros in
s
On Mon, 27 Nov 2017 13:16:03 +0530
Hemant Agrawal wrote:
> The DPDK uses the Open Source BSD-3-Clause license for the core libraries
> and drivers. The kernel components are naturally GPLv2 licensed.
>
> Many of the files in the DPDK source code contain the full text of the
> applicable license.
On 11/28/2017 1:01 AM, Aaron Conole wrote:
"Tan, Jianfeng" writes:
On 11/27/2017 10:27 PM, Yuanhan Liu wrote:
On Fri, Nov 24, 2017 at 05:59:09PM +0800, Chen Hailin wrote:
Hi Aaron Conole && Jianfeng,
The stp could not work in ovs-dpdk vhostuser.
Because the attached vhost device doesn't h
Hi Jinjing,
So to verify,
Before the VF commit ("net/i40e: fix interrupt throttling setting in PF") there
was a bug that I40E_QUEUE_ITR_INTERVAL_DEFAULT
did *not* work and PF was configured with RX_INT_INTERVAL=0 usec and now it is
32usec (taken from the define)?
Thanks,
Hanoh
From: Wu, Jingji
Hi, Hanoch
If DPDK PF, the commit affects that because it introduces an argument (itr_idx)
for i40e_vsi_enable_queues_intr. And use the default itr_idx with default value
32us.
If you'd like to get the descriptor write back immediately, you can set
"CONFIG_RTE_LIBRTE_I40E_ITR_INTERVAL=0" in co
When octeontx event dev receives a packet for the event Rx adapter, the
mbuf port id should contain the appropriate ethdev id instead of
internal channel info.
Fixes: 45a914c5bd71 ("event/octeontx: support event Rx adapter")
Signed-off-by: Pavan Nikhilesh
---
drivers/event/octeontx/ssovf_worker
The channel to port id map is used by event octeontx to map the received
wqe to the respective ethdev port.
Signed-off-by: Pavan Nikhilesh
---
drivers/net/octeontx/octeontx_ethdev.c | 3 +++
drivers/net/octeontx/octeontx_ethdev.h | 6 ++
2 files changed, 9 insertions(+)
diff --git a/drivers
On Tue, Nov 28, 2017 at 01:22:01PM +, Marko Kovacevic wrote:
> Signed-off-by: Marko Kovacevic
> ---
Please include a description of the work involved in the "cleanup".
Makes life easier for reviewers to know what to expect and check that
nothing was missed.
Thanks,
/Bruce
On Tue, Nov 28, 2017 at 01:22:00PM +, Marko Kovacevic wrote:
> Signed-off-by: Marko Kovacevic
> ---
While I know that checkpatch complains about unsigned vs unsigned int,
is there some other reason for making this change wholesale in the code?
/Bruce
On Mon, Nov 27, 2017 at 5:58 PM, Walker, Benjamin wrote:
> On Sun, 2017-11-05 at 01:17 +0100, Thomas Monjalon wrote:
> > Hi, restarting an old topic,
> >
> > 05/01/2017 16:52, Tan, Jianfeng:
> > > On 1/5/2017 5:34 AM, Walker, Benjamin wrote:
> > > > > > Note that this
> > > > > > probably means t
Add support to AES-CCM, for 128, 192 and 256-bit keys.
Signed-off-by: Fan Zhang
---
doc/guides/cryptodevs/aesni_mb.rst | 4 +
doc/guides/cryptodevs/features/aesni_mb.ini| 3 +
doc/guides/rel_notes/release_18_02.rst | 5 +
drivers/crypto/aesni_mb/rte_aesn
Signed-off-by: Marko Kovacevic
---
lib/librte_power/power_acpi_cpufreq.c | 2 +-
lib/librte_power/{rte_power_common.h => power_common.h} | 6 +++---
lib/librte_power/rte_power.c| 2 +-
3 files changed, 5 insertions(+), 5 deletions(-)
rename lib/librt
multi segment version of tx burst function was not
returning the actual number of packets sent out
in PMD xmit function.
Fixes: 1c421f18e0 ("net/thunderx: add single and multi-segment Tx")
Cc: sta...@dpdk.org
Signed-off-by: Jerin Jacob
Signed-off-by: Sunil Kulkarni
---
drivers/net/thunderx/nic
Signed-off-by: Marko Kovacevic
---
lib/librte_power/Makefile | 2 +-
...e_power_acpi_cpufreq.c => power_acpi_cpufreq.c} | 32 +++---
...e_power_acpi_cpufreq.h => power_acpi_cpufreq.h} | 28 +--
lib/librte_power/rte_power.c
Signed-off-by: Marko Kovacevic
---
lib/librte_power/Makefile | 2 +-
.../{rte_power_kvm_vm.c => power_kvm_vm.c} | 28 +++---
.../{rte_power_kvm_vm.h => power_kvm_vm.h} | 28 +++---
lib/librte_power/rte_power.c
Signed-off-by: Marko Kovacevic
---
lib/librte_power/guest_channel.c | 6 +++---
lib/librte_power/guest_channel.h | 6 +++---
lib/librte_power/rte_power.c | 4 ++--
lib/librte_power/rte_power.h | 12 ++--
lib/librte_power/rte_power_acpi_cpufre
Hi Jingjing,
1. The issue is with DPDK PF.
2. The rate is high ~10gb, one DP core, one latency core.
3. The fix is here
/* Bind all RX queues to allocated MSIX interrupt */
for (i = 0; i < nb_queue; i++) {
val = (msix_vect <<
I4
time() is defined in time.h
Fixes: ffbe3be0d4 ("app/test: add libcrypto")
Signed-off-by: Jerin Jacob
---
test/test/test_cryptodev.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/test/test/test_cryptodev.c b/test/test/test_cryptodev.c
index 1bed65dad..7997b4882 100644
--- a/test/test/test
Hi, Hanoch
Are you talking about i40 VF's latency? And you are using DPDK PF as host
driver?
In this case, we are setting the Interrupt Throttling (ITR) to be maximum. That
is to say, if the packet rate is very slow , the receive descriptor is written
back when ITR timeout, otherwise it is wr
The previuse code took a send queue entry size for stamping from the
send queue entry pointed by completion queue entry; This 2 reads were
done per packet in completion stage.
The completion burst packets number is managed by fixed size stored in
Tx queue, so we can infer that each valid completio
This counter saved the descriptor elements which are waiting to be
completted and was used to know if completion function should be
called.
This completion check can be done by other elements management
variables and we can prevent this counter management.
Remove this counter and replace the comp
Using power of 2 descriptors number makes the ring management easier
and allows to use mask operation instead of wraparound conditions.
Adjust Tx descriptor number to be power of 2 and change calculation to
use mask accordingly.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 22 +
The Tx queue send ring was managed by Tx block head,tail,count and mask
management variables which were used for managing the send queue remain
space and next places of empty or completted work queue entries.
This method suffered from an actual addresses recalculation per packet,
an unnecessary Tx
mlx4 Tx block can handle up to 4 data segments or control segment + up
to 3 data segments. The first data segment in each not first Tx block
must validate Tx queue wraparound and must use IO memory barrier before
writing the byte count.
The previous multi-segment code used "for" loop to iterate ov
At error time, the first 4 bytes of each WQE Tx block still have not
writen, so no need to stamp them because they are already stamped.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 22 +-
1 file changed, 1 insertion(+), 21 deletions(-)
diff --git a/drivers/n
There is no need to check Tx queue wraparound for segments which are
not at the beginning of a Tx block. Especially relevant in a single
segment case.
Remove unnecessary aforementioned checks from Tx path.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 20 ++--
1
When invalid lkey is sent to HW, HW sends an error notification in
completion function.
The previous code wouldn't crash but doesn't add any application report
in case of completion error, so application cannot know that packet
actually was dropped in case of invalid lkey.
Return back the lkey va
This series improves mlx4 Tx performance and fix and clean some Tx code.
1. 10% MPPS improvement for 1 queue, 1 core, 64B packets, txonly mode.
2. 20% MPPS improvement for 1 queue, 1 core, 32B*4(segs) packets, txonly mode.
Matan Azrad (8):
net/mlx4: fix Tx packet drop application report
net/m
On 11/15/2017 12:41 PM, Jianfeng Tan wrote:
In a running VM, operations (like device attach/detach) will
trigger the QEMU to resend set_mem_table to vhost-user backend.
DPDK vhost-user handles this message rudely by unmap all existing
regions and map new ones. This might lead to segfault if th
On 11/27/2017 10:58 PM, Ferruh Yigit wrote:
On 11/16/2017 12:04 AM, Andrew Rybchenko wrote:
From: Andy Moreton
We don't know what mkconfig.py is and most probably not interested in :) This
looks like your internal detail, more interesting is the effect of this change,
like "add new sensors".
Testpmd should not use ethdev ports which are managed by other DPDK
entities.
Set Testpmd ownership to each port which is not used by other entity and
prevent any usage of ethdev ports which are not owned by Testpmd.
Signed-off-by: Matan Azrad
---
app/test-pmd/cmdline.c | 100 +
Fail-safe PMD sub devices management is based on ethdev port mechanism.
So, the sub-devices management structures are exposed to other DPDK
entities which may use them in parallel to fail-safe PMD.
Use the new port ownership mechanism to avoid multiple managments of
fail-safe PMD sub-devices.
Sig
Add ownership mechanism to DPDK Ethernet devices to avoid multiple
management of a device by different DPDK entities as discussed in:
http://dpdk.org/ml/archives/dev/2017-September/074656.html
Adjusts failsafe and testpmd to use it.
Matan Azrad (5):
ethdev: free a port by a dedicated API
ethd
Call dedicated ethdev API to free port in remove time as was done in
other fail-safe places.
Signed-off-by: Matan Azrad
---
drivers/net/failsafe/failsafe_ether.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/failsafe/failsafe_ether.c
b/drivers/net/failsafe/fail
The ownership of a port is implicit in DPDK.
Making it explicit is better from the next reasons:
1. It may be convenient for multi-process applications to know which
process is in charge of a port.
2. A library could work on top of a port.
3. A port can work on top of another port.
Also in the
Use a dedicated API to free port instead of changing its state
directly.
Signed-off-by: Matan Azrad
---
lib/librte_ether/rte_ethdev.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index 318af28..2d754d9 100644
--
On 11/27/2017 10:58 PM, Ferruh Yigit wrote:
On 11/16/2017 12:04 AM, Andrew Rybchenko wrote:
From: Andy Moreton
PREfast seems a static analysis tool. Can you please give more detail about
fixed issue itself. "fix PREfast issue" is not informative about what has been
fixed.
If you can send the
On Tue, Nov 28, 2017 at 04:56:54AM -0500, junjie.j.c...@intel.com wrote:
> From: Junjie Chen
>
> This patch add a restriction to multi-process support: secondary
> processes should only run alongside primary process with same DPDK
> version, so that secondary processes can use the same hugepage
Monday, November 27, 2017 8:11 PM, Ferruh Yigit:
> On 11/26/2017 11:19 PM, Shahaf Shuler wrote:
> > Thursday, November 23, 2017 11:30 AM, Nelio Laranjeiro:
> >> This drop queue can be handled efficiently by using the drop flag in
> >> the context.
> >>
> >> Signed-off-by: Nelio Laranjeiro
> >> Ack
Signed-off-by: Marko Kovacevic
---
V1: Clean of of kvm power files
---
lib/librte_power/Makefile | 2 +-
.../{rte_power_kvm_vm.c => power_kvm_vm.c} | 28 +++---
.../{rte_power_kvm_vm.h => power_kvm_vm.h} | 28 +++---
lib
Signed-off-by: Marko Kovacevic
---
V1: Clean up of power common header
---
lib/librte_power/power_acpi_cpufreq.c | 2 +-
lib/librte_power/{rte_power_common.h => power_common.h} | 6 +++---
lib/librte_power/rte_power.c| 2 +-
3 files changed, 5 inse
Signed-off-by: Marko Kovacevic
---
V1: Clean up of acpi files
---
lib/librte_power/Makefile | 2 +-
...e_power_acpi_cpufreq.c => power_acpi_cpufreq.c} | 32 +++---
...e_power_acpi_cpufreq.h => power_acpi_cpufreq.h} | 28 +--
lib/librte
Signed-off-by: Marko Kovacevic
---
V1: Changed unsigned to unsigned int
---
lib/librte_power/guest_channel.c | 6 +++---
lib/librte_power/guest_channel.h | 6 +++---
lib/librte_power/rte_power.c | 4 ++--
lib/librte_power/rte_power.h | 12 ++--
On 11/27/2017 10:57 PM, Ferruh Yigit wrote:
On 11/16/2017 12:04 AM, Andrew Rybchenko wrote:
From: Mark Spender
For encapsulated packets, the firmware gives info about the inner frame
fields by default. When not using encapsulation offload, ask for info
about the outer frame instead.
Unless bu
On 11/27/2017 10:58 PM, Ferruh Yigit wrote:
On 11/16/2017 12:04 AM, Andrew Rybchenko wrote:
From: Mark Spender
What has been fixed here?
If you can provide more details I can amend the commit later.
EFSYS_PROBE1 takes one typed value (in addition to the probe name),
whereas EFSYS_PROBE has
Add support of PPPoE and L2TP in rte_net_get_ptype().
Signed-off-by: Beilei Xing
---
lib/librte_mbuf/rte_mbuf_ptype.c | 2 ++
lib/librte_mbuf/rte_mbuf_ptype.h | 26 ++
2 files changed, 28 insertions(+)
diff --git a/lib/librte_mbuf/rte_mbuf_ptype.c b/lib/librte_mbuf/rte_
Beilei Xing (2):
mbuf: support PPPoE and L2TP in software packet type parser
net/i40e: support PPPoE and L2TP in SW ptype parser function
drivers/net/i40e/i40e_ethdev.c | 72
lib/librte_mbuf/rte_mbuf_ptype.c | 2 ++
lib/librte_mbuf/rte_mbuf_ptype.h
This patch adds support for PPPoE and L2TP in software packet
type parser function.
Signed-off-by: Beilei Xing
---
drivers/net/i40e/i40e_ethdev.c | 72 +-
1 file changed, 50 insertions(+), 22 deletions(-)
diff --git a/drivers/net/i40e/i40e_ethdev.c b/driv
On 11/24/2017 07:08 PM, Maxime Coquelin wrote:
When performing live-migration with multiple queue pairs,
VHOST_USER_SET_LOG_BASE request is sent multiple times.
If packets are being processed by the PMD threads, it is
possible that they are setting bits in the dirty log map while
its region is
On 11/27/2017 10:59 PM, Ferruh Yigit wrote:
On 11/16/2017 12:04 AM, Andrew Rybchenko wrote:
From: Andrew Rybchenko
Since sign-off by is from @solarflare.com, I have updated author to:
"Andrew Rybchenko " for all patches that author and
sign-off is different, assuming both are you.
Thanks a l
75 matches
Mail list logo
