> From: Maxime Leroy
>
> vdev ethdev can not be allocated on a numa socket that is not socket 0.
> The reason comes from rte_eth_dev_allocate() which uses rte_socket_id() to
> identify the socket on which vdev driver data should be allocated.
> However, at this initialization step, rte_socket_id(
> From: Adrien Mazarguil
>
> When NUMA is enabled, the per-core fwd_lcore structure mbuf pool pointer
> is not set, causing a crash.
>
> Initialize fwd_lcore after allocating NUMA memory pools.
>
> Signed-off-by: Adrien Mazarguil
Acked-by: Thomas Monjalon
Applied for version 1.6.0r2.
Thank
If some DPDK libraries are installed on the system, the linker was trying
to use them before searching in -L path.
The obscure reason is that we were prefixing -L with -Wl, to pass it
directly to the linker.
But -L is also a gcc option. And allowing gcc to process this option fixes
the issue.
Sign
From: Didier Pallard
Currently, if there is more memory in hugepages than the amount
requested by dpdk application, the memory is allocated by taking as much
memory as possible from each socket, starting from first one.
For example if a system is configured with 8 GB in 2 sockets (4 GB per
socket
From: Didier Pallard
Increasing maximum number of lcores gives a huge place to undetected
lcores in output traces. Moreover, this output does not give any
interesting information, since list of undetected lcores can be deduced
from list of detected ones.
So remove output related to undetected cor
From: Didier Pallard
In current version, coremask can only be specified using a bitmask.
It will now be possible to specify core masks in 2 different ways:
- Using a bitmask (-c 0xnnn): bitmask must be in hex format and start with 0x
- Using a core set description in following format: -c [c1[-c2]
Add the command "port config X rss-hash-key key" in the 'testpmd'
application to configure the RSS hash key used to compute the RSS
hash of input [IP] packets received on port X.
Signed-off-by: Ivan Boule
---
app/test-pmd/cmdline.c | 96 +++-
app/tes
1) Add a new function "rss_hash_conf_get" in the PMD API to retrieve the
current configuration of the RSS functions and/or of the RSS key used
by a NIC to compute the RSS hash of input packets.
The new function uses the existing data structure "rte_eth_rss_conf" for
returning the RSS ha
The function cmd_config_rss_parsed() associated with the command
"port config rss all" required to first stop all ports, in order to
then entirely re-configure all ports with the new RSS hash computation
parameters.
Use now the new function rte_eth_dev_rss_hash_conf_update() that dynamically
only c
1) Add a new function "rss_hash_conf_update" in the PMD API to dynamically
update the RSS flags and/or the RSS key used by a NIC to compute the RSS
hash of input packets.
The new function uses the existing data structure "rte_eth_rss_conf" for
the argument that contains the new hash fla
Each entry of the RSS redirection table (RETA) of igb and ixgbe ports
contains a 4-bit RX queue index, thus imposing RSS RX queue indices to
be strictly lower than 16.
In addition, if a RETA entry is configured with a RX queue index that is
strictly lower than 16, but is greater or equal to the num
This set of patches allows to dynamically get and set the RSS configuration of a
port:
- rss functions (IP/UDP/TCP ...)
- rss hash key
--
Ivan Boule
Ivan Boule (5):
ethdev: check RX queue indices in RETA config against number of
queues
ethdev: allow to set RSS hash computation flags and/
From: Ivan Boule
Add a new specific packet processing engine in the "testpmd" application that
only replies to ARP requests and to ICMP echo requests.
For this purpose, a new "icmpecho" forwarding mode is provided that can be
dynamically selected with the following testpmd command:
set fwd i
When starting/stopping ports, a link status check on all available ports is
done. This can be annoying when cables are not plugged at the time.
Default behavior is untouched.
Signed-off-by: David Marchand
---
app/test-pmd/cmdline.c| 41 +
app/test-pm
From: Zijie Pan
Display port number and MAC address at start up.
It is useful when configuring a packet generator.
Signed-off-by: Zijie Pan
---
app/test-pmd/testpmd.c |6 ++
1 file changed, 6 insertions(+)
diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
index 9c56914..a66
From: Adrien Mazarguil
When NUMA is enabled, the per-core fwd_lcore structure mbuf pool pointer
is not set, causing a crash.
Initialize fwd_lcore after allocating NUMA memory pools.
Signed-off-by: Adrien Mazarguil
---
app/test-pmd/testpmd.c | 20 +++-
1 file changed, 11 inse
From: Maxime Leroy
vdev ethdev can not be allocated on a numa socket that is not socket 0.
The reason comes from rte_eth_dev_allocate() which uses rte_socket_id() to
identify the socket on which vdev driver data should be allocated.
However, at this initialization step, rte_socket_id() always ret
2014-04-15 15:51, David Marchand:
> From: Liu Xiaofeng
>
> Now socket id is from device's numa_node, if it is invalid, just set it to 0
> as default to avoid crash which will be caused by the reference to
> port_per_socket[socket_id].
>
> Also one warning is displayed to user that port-numa-conf
2014-04-30 06:52, Neil Horman:
> On Wed, Apr 30, 2014 at 02:46:41AM +0200, Thomas Monjalon wrote:
> > The goal of this patch serie is to be able to package DPDK
> > for RPM-based distributions.
>
> You should merge these into a single spec file so that you only have to
> build once. That also cle
Hi Konstantin,
The content's of your patch is OK.
However, the following fix in the function `eth_igbvf_rx_init`:
- if (hw->mac.type == e1000_82576) {
+ if (hw->mac.type == e1000_vfadapt) {
should be sent in a different patch (without the spurious blank
character )
2014-04-18 09:09, Neil Horman:
> On Fri, Apr 18, 2014 at 02:56:18PM +0200, David Marchand wrote:
> > There should be no real need for this initialised field as the whole
> > structure is set to 0 in rte_config_init() by primary process, and
> > secondary processes wait for this to happen before any
2014-04-18 09:08, Neil Horman:
> On Fri, Apr 18, 2014 at 02:56:17PM +0200, David Marchand wrote:
> > We don't really need this field as it is only used when creating the
> > memzone object associated to this heap.
> >
> > Signed-off-by: David Marchand
>
> Acked-by: Neil Horman
Applied for vers
On Wed, Apr 30, 2014 at 01:09:38PM +0200, Thomas Monjalon wrote:
> 2014-04-30 06:52, Neil Horman:
> > On Wed, Apr 30, 2014 at 02:46:41AM +0200, Thomas Monjalon wrote:
> > > The goal of this patch serie is to be able to package DPDK
> > > for RPM-based distributions.
> >
> > You should merge these
Hi,
I tried a lot (more than a week) to solve the problem myself and not
to bother the list, but I didn't succeed. Maybe other people have the
same problem.
I have a simple program attached. It is intended for simple packet
capturing; captures from interface and writes to memory, and frees the
mem
On Wed, Apr 30, 2014 at 02:46:41AM +0200, Thomas Monjalon wrote:
> The goal of this patch serie is to be able to package DPDK
> for RPM-based distributions.
>
> The file naming currently doesn't allow to install different DPDK versions.
> But the packaging naming should be ready to manage differen
Hi,
Please find the attached paper http://kfall.net/ucbpage/papers/snc.pdf
Figures 4 and 5 refers about the degradation when the # of queues are
increased.
It refers sweet spot as 2 to 4 queues.
Have you please verified with smaller # of queues please?
Thanks,
-Original Message-
Fr
Package can be built with:
rpmbuild -ta virtio-net-pmd-1.1.tar.gz
Signed-off-by: Thomas Monjalon
---
pkg/dpdk-virtio.spec | 73
1 file changed, 73 insertions(+)
create mode 100644 pkg/dpdk-virtio.spec
diff --git a/pkg/dpdk-virtio.spe
Package can be built with:
rpmbuild -ta vmxnet3-usermap-1.1.tar.gz
Signed-off-by: Thomas Monjalon
Acked-by: Chris Wright
---
pkg/dpdk-vmxnet3.spec | 84 +++
1 file changed, 84 insertions(+)
create mode 100644 pkg/dpdk-vmxnet3.spec
diff -
Packages can be built with:
rpmbuild -ta memnic-1.0.tar.gz
There are packages for kernel module and PMD versions.
Signed-off-by: Thomas Monjalon
---
pkg/dpdk-memnic.spec | 107 +++
1 file changed, 107 insertions(+)
create mode 100644 pkg/
Packages can be built with:
RPM_BUILD_NCPUS=8 rpmbuild -ta dpdk-1.6.0r1.tar.gz
There are packages for runtime and development.
Once devel package is installed, it can be used like this:
make -C /usr/share/dpdk/examples/helloworld RTE_SDK=/usr/share/dpdk
Signed-off-by: Thomas Monja
The goal of this patch serie is to be able to package DPDK
for RPM-based distributions.
The file naming currently doesn't allow to install different DPDK versions.
But the packaging naming should be ready to manage different DPDK versions
having different API/ABI for different applications:
2014-04-16 09:51, Neil Horman:
> The shared libraries built with the current makefile set produce static
> libraries rather than actual shared objects. This is due to several missing
> options that are required to correctly build shared objects using ld, as
> well as a mis-specified -share option
2014-04-29 10:23, Neil Horman:
> On Tue, Apr 29, 2014 at 04:16:03PM +0200, David Marchand wrote:
> > Following debian kernel headers upgrade to 3.2.57-3, pci capability
> > accessors have been backported (upstream commit
> > 8c0d3a02c1309eb6112d2e7c8172e8ceb26ecfca, ("PCI: Add accessors for PCI
> >
2014-04-25 09:27, Neil Horman:
> On Fri, Apr 25, 2014 at 01:59:38PM +0200, Olivier Matz wrote:
> > This patch series fix compilation of bsdapp environment
> > that was broken by previous commits.
>
> Series, Minus the Makefile enhancement
> Acked-by: Neil Horman
The whole serie is applied for ve
34 matches
Mail list logo
