2015-12-23 10:09, Yuanhan Liu:
> On Wed, Dec 23, 2015 at 09:55:54AM +0800, Peter Xu wrote:
> > On Tue, Dec 22, 2015 at 04:38:30PM +, Xie, Huawei wrote:
> > > On 12/22/2015 7:39 PM, Peter Xu wrote:
> > > > I tried to unbind one of the virtio net device, I see the PCI entry
> > > > still there.
>
2015-12-23 05:13, Xie, Huawei:
> On 12/23/2015 10:57 AM, Yuanhan Liu wrote:
> > On Wed, Dec 23, 2015 at 10:41:57AM +0800, Peter Xu wrote:
> >> On Wed, Dec 23, 2015 at 10:01:35AM +0800, Yuanhan Liu wrote:
> >>> On Tue, Dec 22, 2015 at 05:56:41PM +0800, Peter Xu wrote:
> On Tue, Dec 22, 2015 at
2015-12-23 10:44, Yuanhan Liu:
> On Tue, Dec 22, 2015 at 01:38:29AM -0800, Rich Lane wrote:
> > On Mon, Dec 21, 2015 at 9:47 PM, Yuanhan Liu > linux.intel.com>
> > wrote:
> >
> > On Mon, Dec 21, 2015 at 08:47:28PM -0800, Rich Lane wrote:
> > > The queue state change callback is the one ne
Hi,
2015-12-22 14:13, Christian Ehrhardt:
> Depending on non-doc targets being built before and the setting of DESTDIR
> the examples dir could in some cases not end up in the right target.
> Reason is just a typo variable reference in the copy target.
[...]
> - $(Q)cp -a $(RTE_SDK)/examples $
On Wed, Dec 23, 2015 at 7:09 PM, Tetsuya Mukawa wrote:
> On 2015/12/22 13:47, Rich Lane wrote:
> > On Mon, Dec 21, 2015 at 7:41 PM, Yuanhan Liu <
> yuanhan.liu at linux.intel.com>
> > wrote:
> >
> >> On Fri, Dec 18, 2015 at 10:01:25AM -0800, Rich Lane wrote:
> >>> I'm using the vhost callbacks an
>
> + if (unlikely(alloc_err)) {
> + uint16_t i = entry_success;
> +
> + m->nb_segs = seg_num;
> + for (; i < free_entries; i++)
> + rte_pktmbuf_free(pkts[entry_success]); ->
> rte_pktmbuf_free(pkts[i]);
> + }
> +
> rte_comp
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Stephen Hemminger
> Sent: Wednesday, December 23, 2015 6:38 PM
> To: Xie, Huawei
> Cc: dev at dpdk.org; dprovan at bivio.net
> Subject: Re: [dpdk-dev] [PATCH v3 1/2] mbuf: provide rte_pktmbuf_alloc_bulk
> API
Extend the existing flow_director_filter to add support for packet
filter flow. Also shows how to pass the extra behavior arguments
to rewrite fields in matched filter rules.
Signed-off-by: Rahul Lakkireddy
Signed-off-by: Kumar Sanghvi
---
v2:
1. Added new field filter-type to allow specifying
Add a new packet filter flow that allows filtering a packet based on
matching ingress port, ethertype, vlan, ip, and tcp/udp fields, i.e.
matching based on any or all fields at the same time.
Add the ability to provide masks for fields in flow to allow range of
values. Allow selection of maskfull
This RFC series of patches attempt to extend the flow director filter to
add support for Chelsio T5 hardware filtering capabilities.
Chelsio T5 supports carrying out filtering in hardware which supports 3
actions to carry out on a packet which hit a filter viz.
1. Action Pass - Packets hitting a
From: "Chen Jing D(Mark)"
Add introductions on how to enable Vector FM10K Rx/Tx functions,
the preconditions and assumptions on Rx/Tx configuration parameters.
The new content also lists the limitations of vector, so app/customer
can do better to select best Rx/Tx functions.
Signed-off-by: Chen
Add i40e_udp_tunnel_flow_configre() to implement the configuration of flow rule
with 'src IP, dst IP, src port, dst port and tunnel ID' using flow director.
Signed-off-by: Jijiang Liu
---
drivers/net/i40e/i40e_ethdev.c | 41
1 files changed, 41 inserti
Using SIMD instruction to accelarate encapsulation operation.
Signed-off-by: Jijiang Liu
---
lib/librte_ether/libtunnel/rte_vxlan_opt.c | 251
1 files changed, 251 insertions(+), 0 deletions(-)
create mode 100644 lib/librte_ether/libtunnel/rte_vxlan_opt.c
diff --g
This function parameters should be the same as callback function
(rte_rx/tx_callback_fn).
But we can redefine some parameters as 'unused'.
Signed-off-by: Jijiang Liu
---
lib/librte_ether/libtunnel/rte_vxlan_opt.h | 49
1 files changed, 49 insertions(+), 0 deletio
Signed-off-by: Jijiang Liu
---
lib/librte_ether/rte_ethdev.c | 60 +
1 files changed, 60 insertions(+), 0 deletions(-)
diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index c3eed49..6725398 100644
--- a/lib/librte_ether/rte_eth
Add the struct 'rte_eth_tunnel_conf' and the tunnel configuration API.
Signed-off-by: Jijiang Liu
---
lib/librte_ether/rte_ethdev.h | 28
1 files changed, 28 insertions(+), 0 deletions(-)
diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
The purpose of extending this structure is to support more tunnel filter
conditions.
Signed-off-by: Jijiang Liu
---
lib/librte_ether/rte_eth_ctrl.h | 14 +++---
1 files changed, 11 insertions(+), 3 deletions(-)
diff --git a/lib/librte_ether/rte_eth_ctrl.h b/lib/librte_ether/rte_eth_c
I want to define a set of General tunneling APIs, which are used to accelarate
tunneling packet processing in DPDK,
In this RFC patch set, I wll explain my idea using some codes.
1. Using flow director offload to define a tunnel flow in a pair of queues.
flow rule: src IP + dst IP + src port + d
Previous l3fwd-power only processes IP and IPv6 packet, other
packet's mbuf is not released, and causes a memory leak.
This patch fixes this issue.
v4 change:
- update release note inside the patch
Signed-off-by: Shaopeng He
---
doc/guides/rel_notes/release_2_3.rst | 6 ++
examples/l3fwd-
When PF establishes a connection with Switch Manager, it receives
a logic port range from SM, and registers certain logic ports from
that range, then a default VID will be send back from SM. This whole
transaction needs to be finished in dev_init, otherwise, in dev_start
the interrupt setting will
Interrupt mode framework has enable/disable functions for individual
rx queue, this patch implements these two functions.
v2 changes:
- split one big patch into three smaller ones
Signed-off-by: Shaopeng He
---
drivers/net/fm10k/fm10k_ethdev.c | 33 +
1 file ch
Previous dev_stop function stops the rx/tx queues. This patch adds logic
to disable rx queue interrupt, clean the datapath event and queue/vec map.
v5 changes:
- remove one unnecessary NULL check for rte_free
v2 changes:
- split one big patch into three smaller ones
Signed-off-by: Shaopeng H
In interrupt mode, each rx queue can have one interrupt to notify the up
layer application when packets are available in that queue. Some queues
also can share one interrupt.
Currently, fm10k needs one separate interrupt for mailbox. So, only those
drivers which support multiple interrupt vectors e
rx_descriptor_done is used by interrupt mode example application
(l3fwd-power) to check rxd DD bit to decide the RX trend,
then l3fwd-power will adjust the cpu frequency according to
the result.
v5 change:
- fix a wrong error message
Signed-off-by: Shaopeng He
---
drivers/net/fm10k/fm10k.h
This patch series adds interrupt mode support for fm10k,
contains four major parts:
1. implement rx_descriptor_done function in fm10k
2. add rx interrupt support in fm10k PF and VF
3. make sure default VID available in dev_init in fm10k
4. fix a memory leak for non-ip packet in l3fwd-power,
whi
Handle SIGINT and SIGTERM in l3fwd.
Signed-off-by: Zhihong Wang
---
examples/l3fwd/main.c | 25 +
1 file changed, 25 insertions(+)
diff --git a/examples/l3fwd/main.c b/examples/l3fwd/main.c
index 5b0c2dd..aae16d2 100644
--- a/examples/l3fwd/main.c
+++ b/examples/l3fwd/ma
Handle SIGINT and SIGTERM in l2fwd.
Signed-off-by: Zhihong Wang
---
examples/l2fwd/main.c | 25 +
1 file changed, 25 insertions(+)
diff --git a/examples/l2fwd/main.c b/examples/l2fwd/main.c
index 720fd5a..0594037 100644
--- a/examples/l2fwd/main.c
+++ b/examples/l2fwd/ma
Handle SIGINT and SIGTERM in testpmd.
Signed-off-by: Zhihong Wang
---
app/test-pmd/testpmd.c | 23 +++
1 file changed, 23 insertions(+)
diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
index 98ae46d..c259ba3 100644
--- a/app/test-pmd/testpmd.c
+++ b/app/test-pmd/
This patch handles SIGINT and SIGTERM in testpmd, l2fwd, and l3fwd, make sure
all ports are properly stopped and closed.
For virtual ports, the stop and close function may deal with resource cleanup,
such as socket files unlinking.
Zhihong Wang (3):
app/test-pmd: Handle SIGINT and SIGTERM in t
Workaround for the issue of cannot processing adminq commands during
initialization, when 2x40G or 4x10G is receiving packets in highest
throughput. Register 0x002698a8 and 0x002698ac should be cleared at
first, and restored with the default values at the end. No more
details, as they are not expos
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Bernard Iremonger
> Sent: Wednesday, December 23, 2015 9:45 AM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH] virtio: fix crashes in virtio stats functions
>
> This initialisation of nb_rx_queues and nb_t
Le 23 d?c. 2015 10:12, "Qiu, Michael" a ?crit :
>
> Is it suitable to put so many code in commit log?
It is more explicit than a text/comment. I do not think it should be
maintained code.
>
> Thanks,
> Michael
> On 12/22/2015 5:36 PM, Didier Pallard wrote:
> > As demonstrated by the following co
User should be able to configure ethdev with zero rx/tx queues, but both should
not be zero.
After above change, rte_eth_dev_tx_queue_config, rte_eth_dev_rx_queue_config
should
allocate memory for rx/tx queues only when number of rx/tx queues are nonzero.
Signed-off-by: Reshma Pattan
---
lib/l
Macro RTE_PROC_PRIMARY_OR_ERR_RET blocking the secondary process from API usage.
API access should be given to both secondary and primary.
Signed-off-by: Reshma Pattan
---
lib/librte_cryptodev/rte_cryptodev.c | 42
1 file changed, 42 deletions(-)
diff --git
Macros RTE_PROC_PRIMARY_OR_ERR_RET and RTE_PROC_PRIMARY_OR_RET
are blocking the secondary process from using the APIs.
API access should be given to both secondary and primary.
Fix minor checkpath issues in rte_ethdev.h
Reported-by: Sean Harte
Signed-off-by: Reshma Pattan
---
lib/librte_ether/
On 12/23/2015 7:25 PM, linhaifeng wrote:
>
>>
>> +if (unlikely(alloc_err)) {
>> +uint16_t i = entry_success;
>> +
>> +m->nb_segs = seg_num;
>> +for (; i < free_entries; i++)
>> +rte_pktmbuf_free(pkts[entry_success]); ->
>> rte_pktmbuf_f
Hi Alex,
> I've re-posted the unified patch upstream and it should start showing up in
> the next linux-next build. ?I expect the dpdk code won't be merged until
> after this gets back into a proper kernel, but could we get the dpdk
> modifications posted as rfc for others looking to try it?
I ha
Hi Jijang,
I like an idea of tunnel API very much.
I have a few questions.
1. I see that you have only i40e support due to lack of HW tunneling support in
other NICs.
I don't see a way how do you want to handle tunneling requests for NICs without
HW offload.
I think that we should have on
On Wed, Dec 23, 2015 at 10:41:57AM +0800, Peter Xu wrote:
> On Wed, Dec 23, 2015 at 10:01:35AM +0800, Yuanhan Liu wrote:
> > On Tue, Dec 22, 2015 at 05:56:41PM +0800, Peter Xu wrote:
> > > On Tue, Dec 22, 2015 at 04:32:46PM +0800, Yuanhan Liu wrote:
> > > > Actually, you are right. I mentioned in t
Hi Stephen,
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Stephen
> Hemminger
> Sent: Friday, December 4, 2015 5:14 PM
> To: Doherty, Declan
> Cc: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH 0/8] bonding: fixes and enhancements
>
> These are bug fixes an
On Tue, Dec 22, 2015 at 01:38:29AM -0800, Rich Lane wrote:
> On Mon, Dec 21, 2015 at 9:47 PM, Yuanhan Liu
> wrote:
>
> On Mon, Dec 21, 2015 at 08:47:28PM -0800, Rich Lane wrote:
> > The queue state change callback is the one new API that needs to be
> > added because
> > normal NI
On Wed, Dec 23, 2015 at 10:01:35AM +0800, Yuanhan Liu wrote:
> On Tue, Dec 22, 2015 at 05:56:41PM +0800, Peter Xu wrote:
> > On Tue, Dec 22, 2015 at 04:32:46PM +0800, Yuanhan Liu wrote:
> > > Actually, you are right. I mentioned in the last email that this is
> > > for configuration part. To answer
On Wed, Dec 23, 2015 at 10:09:49AM +0800, Yuanhan Liu wrote:
> Why can't we simply quit at pci_scan_one, once finding that it's not
> bond to uio (or similar stuff)? That would be generic enough, that we
> don't have to do similar checks for each new pmd driver.
>
> Or, am I missing something?
It
On Wed, 23 Dec 2015 00:17:53 +0800
Huawei Xie wrote:
> +
> + rc = rte_mempool_get_bulk(pool, (void **)mbufs, count);
> + if (unlikely(rc))
> + return rc;
> +
> + switch (count % 4) {
> + case 0: while (idx != count) {
> + RTE_MBUF_ASSERT(rte_mbuf_re
On Wed, 23 Dec 2015 16:49:51 +0800
Jijiang Liu wrote:
> +
> +#ifndef __INTEL_COMPILER
> +#pragma GCC diagnostic ignored "-Wcast-qual"
> +#endif
> +
> +#pragma GCC diagnostic ignored "-Wstrict-aliasing"
> +
Since this is new code, can't you please fix it to be warning safe?
On Wed, 23 Dec 2015 16:49:46 +0800
Jijiang Liu wrote:
> 1)at config phase
>
> dev_config(port, ...);
> tunnel_config(port,...);
> ...
> dev_start(port);
> ...
> rx_burst(port, rxq,... );
> tx_burst(port, txq,...);
What about dynamically adding and deleting multiple tunnels after
device has star
On Wed, Dec 23, 2015 at 09:55:54AM +0800, Peter Xu wrote:
> On Tue, Dec 22, 2015 at 04:38:30PM +, Xie, Huawei wrote:
> > On 12/22/2015 7:39 PM, Peter Xu wrote:
> > > I tried to unbind one of the virtio net device, I see the PCI entry
> > > still there.
> > >
> > > Before unbind:
> > >
> > > [ro
On Tue, Dec 22, 2015 at 05:56:41PM +0800, Peter Xu wrote:
> On Tue, Dec 22, 2015 at 04:32:46PM +0800, Yuanhan Liu wrote:
> > Actually, you are right. I mentioned in the last email that this is
> > for configuration part. To answer your question in this email, you
> > will not be able to go that fur
On Tue, Dec 22, 2015 at 04:38:30PM +, Xie, Huawei wrote:
> On 12/22/2015 7:39 PM, Peter Xu wrote:
> > I tried to unbind one of the virtio net device, I see the PCI entry
> > still there.
> >
> > Before unbind:
> >
> > [root at vm proc]# lspci -k -s 00:03.0
> > 00:03.0 Ethernet controller: Red H
This initialisation of nb_rx_queues and nb_tx_queues has been removed
from eth_virtio_dev_init.
The nb_rx_queues and nb_tx_queues were being initialised in eth_virtio_dev_init
before the tx_queues and rx_queues arrays were allocated.
The arrays are allocated when the ethdev port is configured and
Is it suitable to put so many code in commit log?
Thanks,
Michael
On 12/22/2015 5:36 PM, Didier Pallard wrote:
> As demonstrated by the following code, CRC32c computation is not valid
> when buffer length is not a multiple of 4 bytes:
> (Output obtained by code below)
>
> CRC of 1 NULL bytes expec
> -Original Message-
> From: Rich Lane [mailto:rich.lane at bigswitch.com]
> Sent: Wednesday, December 23, 2015 4:08 PM
> To: dev at dpdk.org
> Cc: Zhang, Helin
> Subject: [PATCH] i40e: fix inverted check for ETH_TXQ_FLAGS_NOREFCOUNT
>
> The no-refcount path was being taken without the a
Hi Vithal,
The number of VF queues is decided by PF. Suppose you use kernel driver for PF.
So the queue number is decided by PF kernel driver.
I have a 82599ES, and find no matter ixgbevf or dpdk igb_uio is used, the rx
queue number is 2. Frankly, I believe 2 is the expected number. Surprised tha
v4 changes:
fix a silly typo in error handling when rte_pktmbuf_alloc fails
reported by haifeng
pre-allocate a bulk of mbufs instead of allocating one mbuf a time on demand
Signed-off-by: Gerald Rogers
Signed-off-by: Huawei Xie
Acked-by: Konstantin Ananyev
Acked-by: Yuanhan Liu
Tested-by: Yu
v3 changes:
move while after case 0
add context about duff's device and why we use while loop in the commit
message
v2 changes:
unroll the loop a bit to help the performance
rte_pktmbuf_alloc_bulk allocates a bulk of packet mbufs.
There is related thread about this bulk API.
http://dpdk.org/d
v4 changes:
fix a silly typo in error handling when rte_pktmbuf_alloc fails
v3 changes:
move while after case 0
add context about duff's device and why we use while loop in the commit
message
v2 changes:
unroll the loop in rte_pktmbuf_alloc_bulk to help the performance
For symmetric rte_pktm
On 12/23/2015 10:57 AM, Yuanhan Liu wrote:
> On Wed, Dec 23, 2015 at 10:41:57AM +0800, Peter Xu wrote:
>> On Wed, Dec 23, 2015 at 10:01:35AM +0800, Yuanhan Liu wrote:
>>> On Tue, Dec 22, 2015 at 05:56:41PM +0800, Peter Xu wrote:
On Tue, Dec 22, 2015 at 04:32:46PM +0800, Yuanhan Liu wrote:
Hi,
While initializing pci port (VF) DPDK is crashing while configuring the device.
Reason/location:
PMD: rte_eth_dev_configure: ethdev port_id=1 nb_rx_queues=8 > 2
EAL: Error - exiting with code: 1
System info:
DPDK version: 2.0
NIC: 82599EB, sr-iov enabled.
SR-IOV config at ES
> -Original Message-
> From: Qiu, Michael
> Sent: Tuesday, December 22, 2015 3:28 PM
> To: He, Shaopeng; dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v4 2/6] fm10k: setup rx queue interrupts for
> PF and VF
>
> On 12/21/2015 6:20 PM, Shaopeng He wrote:
> > In interrupt mode, each rx q
pre-allocate a bulk of mbufs instead of allocating one mbuf a time on demand
Signed-off-by: Gerald Rogers
Signed-off-by: Huawei Xie
Acked-by: Konstantin Ananyev
---
lib/librte_vhost/vhost_rxtx.c | 35 ++-
1 file changed, 22 insertions(+), 13 deletions(-)
diff -
v3 changes:
move while after case 0
add context about duff's device and why we use while loop in the commit
message
v2 changes:
unroll the loop a bit to help the performance
rte_pktmbuf_alloc_bulk allocates a bulk of packet mbufs.
There is related thread about this bulk API.
http://dpdk.org/d
v3 changes:
move while after case 0
add context about duff's device and why we use while loop in the commit
message
v2 changes:
unroll the loop in rte_pktmbuf_alloc_bulk to help the performance
For symmetric rte_pktmbuf_free_bulk, if the app knows in its scenarios
their mbufs are all simple mb
The no-refcount path was being taken without the application opting in to it.
Reported-by: Mike Stolarchuk
Signed-off-by: Rich Lane
---
drivers/net/i40e/i40e_rxtx.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/i40e/i40e_rxtx.c b/drivers/net/i40e/i40e_rxtx.c
in
63 matches
Mail list logo
