Hi Ariel,
So how are you making sure you are not getting more packets in while you are
waiting for dequeue() to return zero for some time?
As stated, I don?t like this approach, firstly because it forces the on-the-fly
BW configuration changes to result in dropping packets for a considerable
a
There is a tricky type below (leave of -> live off), correcting ...
-Original Message-
From: Dumitrescu, Cristian
Sent: Thursday, May 29, 2014 11:28 PM
To: 'Ivan Boule'; dev at dpdk.org
Subject: RE: [dpdk-dev] [PATCH 04/29] mbuf: added offset of packet meta-data in
the packet buffer just
Hi Ivan,
I hate to disagree with you :), but no, we do not break the scatter-gather
feature. We actually implemented the Packet Framework IPv4 fragmentation and
reassembly ports to validate exactly this.
Maybe the confusion comes from the RTE_MBUF_TO_BADDR macro. This macro only
works (and it
-- Forwarded message --
From: "Ariel Rodriguez"
Date: May 29, 2014 8:59 PM
Subject: RE: [dpdk-dev] Please any one who can help me with librte_sched
To: "Dumitrescu, Cristian"
Cc:
Hi christian. I am aware of that... maybe i wold have to explain better my
approach.
1) The client a
Hi Neil,
Packet Framework does not compete against OVS. OVS is an application (for
virtual switching), while Packet Framework is a toolbox to build applications.
Can somebody pick OVS building blocks and reuse them to build other
applications that use the OpenFlow design principles (port, table
There is a compilation error using latest NIC share codes when
RTE_NIC_BYPASS=y,
the root cause is that the setup_link API have already changed in the share
codes, so change ixgbe_setup_mac_link_multispeed_fixed_fiber()
for eliminating the compilation error.
Signed-off-by: jijiangl
Tested-by:
Bug: when "core id" is greater than 9, the cpu_layout.py output doesn't align.
Socket 0Socket 1
- -
Core 9 [4, 16] [10, 22]
Core 10 [5, 17] [11, 23]
Solution: adjust output format to align based on the maximum length of the
Hi Thomas,
Thanks for your advice.
I'll modify the patch and send it as 'patch v3' later.
On 2014/5/29 15:22, Thomas Monjalon wrote:
> Hi Shannon,
>
> I feel this version is better but it's really complicated to read.
>
>> +max_processor_len=len(str(len(cores)*len(sockets)*2-1))
>> +max_core_ma
This patch supports multiple queues feature in DPDK based virtio-net frontend.
It firstly gets max queue number of virtio-net from virtio PCI configuration and
then send command to negotiate the queue number with backend; When receiving and
transmitting packets, it negotiates multiple virtio-net qu
This patch cleanups some coding style issue, and fixes some errors and warnings
reported by checkpatch.pl.
Signed-off-by: Ouyang Changchun
Tested-by: Waterman Cao
This patch passed Testpmd testing base on commit:
57f0ba5f8b8588dfa6ffcd001447ef6337afa6cd.
See test environment information as the
This v4 patch series replace previous v1, v2, v3 patch for virtio-net multiple
queues feature.
Please apply this v4 patch series and ignore previous patches.
It splits previous one patch into the following 2 patches for easy to review:
Cleanup the existing codes in virtio-net PMD;
Support multipl
Bug: when "core id" is greater than 9, the cpu_layout.py output doesn't align.
Socket 0Socket 1
- -
Core 9 [4, 16] [10, 22]
Core 10 [5, 17] [11, 23]
Solution: adjust output format to align based on the maximum length of the
> -Original Message-
> From: Shaw, Jeffrey B
> Sent: Wednesday, May 28, 2014 5:55 PM
> To: Doherty, Declan; dev at dpdk.org
> Subject: RE: [dpdk-dev] [PATCH 1/4] Link Bonding Library
>
> Hi Declan,
> I'm worried about one thing in "bond_ethdev_tx_broadcast()" related to
> freeing of the
Jijiang, I'm not sure to understand.
Could you send a proper patch for this issue?
Thanks
--
Thomas
2014-05-17 07:48, Liu, Jijiang:
> In Linux kernel versions < 3.14 , the vstart is kernel virtual address, the
> API prototype below. int xen_create_contiguous_region(unsigned long vstart,
> unsig
> The patch changes the way of reserving memory in Dom0 driver. It will
> reserve memory at installing rte_dom0_mm.ko kernel module instead of
> requesting memory dynamically during DPDK application startup. Meanwhile,
> now driver requests memory size of 4M once first, if it failed, and request
>
2014-05-29 15:18, Ouyang Changchun:
> This v4 patch series replace previous v1, v2, v3 patch for virtio-net
> multiple queues feature. Please apply this v4 patch series and ignore
> previous patches.
>
> It splits previous one patch into the following 2 patches for easy to
> review: Cleanup the ex
Add entries into the API documentation for the new distributor library.
Signed-off-by: Bruce Richardson
---
doc/doxy-api-index.md | 1 +
doc/doxy-api.conf | 1 +
2 files changed, 2 insertions(+)
diff --git a/doc/doxy-api-index.md b/doc/doxy-api-index.md
index 2825c08..6e75a6e 100644
--- a/d
Add a set of unit tests and some basic performance test for the
distributor library. These tests cover all the major functionality of
the library on both distributor and worker sides.
Changes in V2:
* Updated tests to work with APIs in distributor patch V2
* Stylistic changes to clear checkpatch.p
add new configuration settings to enable/disable the distributor library
and add makefile entry to compile it once enabled.
Changes in V2:
* Patch updated to use new common config files
Signed-off-by: Bruce Richardson
---
config/common_bsdapp | 6 ++
config/common_linuxapp | 5 +
lib/
This adds the code for a new Intel DPDK library for packet distribution.
The distributor is a component which is designed to pass packets
one-at-a-time to workers, with dynamic load balancing. Using the RSS
field in the mbuf as a tag, the distributor tracks what packet tag is
being processed by wha
add new tailq to the EAL for new distributor library component.
Signed-off-by: Bruce Richardson
---
lib/librte_eal/common/include/rte_tailq_elem.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/lib/librte_eal/common/include/rte_tailq_elem.h
b/lib/librte_eal/common/include/rte_tailq_elem.
This adds a new library to the Intel DPDK whereby a set of packets can be
distributed one-at-a-time to a set of worker cores, with dynamic load balancing
being done between those workers. Flows are identified by a tag within the mbuf
(currently the RSS hash field, 32-bit value), which is used to
> From: Cao, Waterman
> Sent: Thursday, May 29, 2014 4:23 AM
> To: Doherty, Declan; dev at dpdk.org; dev at dpdk.org
> Cc: Cao, Waterman
> Subject: RE: [dpdk-dev] [PATCH 0/4] Link Bonding Library
>
> Hi declan,
>
> Do you send out Patch 1,2, 3 for link bonding?
> Only see patch 0 and 4.
>
> Th
-Original Message-
> From: Neil Horman [mailto:nhorman at tuxdriver.com]
> Sent: Wednesday, May 28, 2014 6:49 PM
> To: Doherty, Declan
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 0/4] Link Bonding Library
>
> On Wed, May 28, 2014 at 04:32:00PM +0100, declan.doherty at intel.com
Ok, I will send a patch for fixing this.
-Original Message-
From: Thomas Monjalon [mailto:thomas.monja...@6wind.com]
Sent: Thursday, May 29, 2014 5:59 PM
To: Liu, Jijiang
Cc: dev at dpdk.org; SMonderer at advaoptical.com; xen-devel at lists.xen.org
Subject: Re: [dpdk-dev] could not l2fwd
> +
> +/* flush the distributor, so that there are no outstanding packets in flight
> or
> + * queued up. */
Its not clear to me that this is a distributor only function. You modified the
comments to indicate that lcores can't preform double duty as both a worker and
a distributor, which is fine,
Hi Shannon,
I feel this version is better but it's really complicated to read.
> +max_processor_len=len(str(len(cores)*len(sockets)*2-1))
> +max_core_map_len = max_processor_len*2+4
> +if max_core_map_len < 12:
> +max_core_map_len = 12
This line is not indented as the other ones.
> +max
Hi Olivier,
2014-05-28 13:47, Olivier MATZ:
> On 05/27/2014 02:55 PM, Thomas Monjalon wrote:
> > Some linker options were not prefixed by -Wl, when using gcc:
> > -z muldefs
> > -melf_i386 (32-bit config)
> >
> > Using macro linkerprefix is fixing it.
> >
> > Signed-off-by: Thomas Monjal
Hi Neil,
2014-05-28 10:17, Neil Horman:
> On Tue, May 27, 2014 at 02:55:16PM +0200, Thomas Monjalon wrote:
> > Some linker options were not prefixed by -Wl, when using gcc:
> > -z muldefs
> > -melf_i386 (32-bit config)
> >
> > Using macro linkerprefix is fixing it.
> >
> > Signed-off-by:
On Thu, May 29, 2014 at 10:33:00AM +, Doherty, Declan wrote:
> -Original Message-
> > From: Neil Horman [mailto:nhorman at tuxdriver.com]
> > Sent: Wednesday, May 28, 2014 6:49 PM
> > To: Doherty, Declan
> > Cc: dev at dpdk.org
> > Subject: Re: [dpdk-dev] [PATCH 0/4] Link Bonding Libra
On Thu, May 29, 2014 at 08:24:56AM +0200, Thomas Monjalon wrote:
> Hi Neil,
>
> 2014-05-28 10:17, Neil Horman:
> > On Tue, May 27, 2014 at 02:55:16PM +0200, Thomas Monjalon wrote:
> > > Some linker options were not prefixed by -Wl, when using gcc:
> > > -z muldefs
> > > -melf_i386 (32-bit conf
Thank you Anatoly !
I?ll take a look at it and get back in case of questions.
-Sujith
On 28/05/14 2:22 pm, "Burakov, Anatoly" wrote:
>Hi Sujith,
>
>> One aspect of ENIC PMD is that it works with VFIO-PCI and not UIO. Hope
>> this is acceptable. The following thread in dpdk-dev influenced this
Hi declan,
Do you send out Patch 1,2, 3 for link bonding?
Only see patch 0 and 4.
Thanks
Waterman
-Original Message-
From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of declan.dohe...@intel.com
Sent: Wednesday, May 28, 2014 11:32 PM
To: dev at dpdk.org; dev at dpdk.org
Subject:
33 matches
Mail list logo