On Wed, Sep 30, 2015 at 06:36:17PM +0300, Avi Kivity wrote:
> As it happens, you're removing the functionality from the users who have no
> other option. They can't use vfio because it doesn't work on virtualized
> setups.
...
> Root can already do anything.
I think there's a contradiction betw
On Wed, Sep 30, 2015 at 11:00:49PM +0300, Gleb Natapov wrote:
> > You are increasing interrupt latency by a huge factor by channeling
> > interrupts through a scheduler. Let user install an
> > interrupt handler function, and be done with it.
> >
> Interrupt latency is not always hugely important
On 09/30/15 22:39, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 10:06:52PM +0300, Vlad Zolotarov wrote:
How would iommu
virtualization change anything?
>>> Kernel can use an iommu to limit device access to memory of
>>> the controlling application.
>> Ok, this is obvious but what
On Wed, Sep 30, 2015 at 09:50:08PM +0300, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 10:43:04AM -0700, Stephen Hemminger wrote:
> > On Wed, 30 Sep 2015 20:39:43 +0300
> > "Michael S. Tsirkin" wrote:
> >
> > > On Wed, Sep 30, 2015 at 10:28:07AM -0700, Stephen Hemminger wrote:
> > > > On W
On Wed, Sep 30, 2015 at 10:06:52PM +0300, Vlad Zolotarov wrote:
> >>How would iommu
> >>virtualization change anything?
> >Kernel can use an iommu to limit device access to memory of
> >the controlling application.
>
> Ok, this is obvious but what it has to do with enabling using MSI/MSI-X
> inter
On 09/30/15 22:10, Vlad Zolotarov wrote:
>
>
> On 09/30/15 22:06, Vlad Zolotarov wrote:
>>
>>
>> On 09/30/15 21:55, Michael S. Tsirkin wrote:
>>> On Wed, Sep 30, 2015 at 09:15:56PM +0300, Vlad Zolotarov wrote:
On 09/30/15 18:26, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 03
On 09/30/15 22:06, Vlad Zolotarov wrote:
>
>
> On 09/30/15 21:55, Michael S. Tsirkin wrote:
>> On Wed, Sep 30, 2015 at 09:15:56PM +0300, Vlad Zolotarov wrote:
>>>
>>> On 09/30/15 18:26, Michael S. Tsirkin wrote:
On Wed, Sep 30, 2015 at 03:50:09PM +0300, Vlad Zolotarov wrote:
> How not vi
On 09/30/15 21:55, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 09:15:56PM +0300, Vlad Zolotarov wrote:
>>
>> On 09/30/15 18:26, Michael S. Tsirkin wrote:
>>> On Wed, Sep 30, 2015 at 03:50:09PM +0300, Vlad Zolotarov wrote:
How not virtualizing iommu forces "all or nothing" approach?
>
My bad that I said its not working, apologies.
Isn?t it correct to say that single process application do not benefit from
having backing files ? In that case can make this configurable by passing a
command line argument that will either unlink or keep the backing files,
defaulting it to keepin
On Wed, Sep 30, 2015 at 09:15:56PM +0300, Vlad Zolotarov wrote:
>
>
> On 09/30/15 18:26, Michael S. Tsirkin wrote:
> >On Wed, Sep 30, 2015 at 03:50:09PM +0300, Vlad Zolotarov wrote:
> >>How not virtualizing iommu forces "all or nothing" approach?
> >Looks like you can't limit an assigned device t
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of shesha
> Sreenivasamurthy (shesha)
> Sent: Wednesday, September 30, 2015 10:44 PM
> To: dev at dpdk.org
> Cc: Michael S. Tsirkin
> Subject: Re: [dpdk-dev] Unlinking hugepage backing file after initialiation
>
On Wed, Sep 30, 2015 at 10:43:04AM -0700, Stephen Hemminger wrote:
> On Wed, 30 Sep 2015 20:39:43 +0300
> "Michael S. Tsirkin" wrote:
>
> > On Wed, Sep 30, 2015 at 10:28:07AM -0700, Stephen Hemminger wrote:
> > > On Wed, 30 Sep 2015 13:37:22 +0300
> > > Vlad Zolotarov wrote:
> > >
> > > >
> >
What I heard is the following: A multi-process DPDK application, working either
in master-worker or master-slave fashion, can potentially benefit by keeping
the backing files in hugetlbfs. However, it is does not work today as the pages
are cleaned and added back when the application restarts. O
On 09/30/15 18:26, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 03:50:09PM +0300, Vlad Zolotarov wrote:
>> How not virtualizing iommu forces "all or nothing" approach?
> Looks like you can't limit an assigned device to only access part of
> guest memory that belongs to a given process. Ei
On Wed, Sep 30, 2015 at 08:39:43PM +0300, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 10:28:07AM -0700, Stephen Hemminger wrote:
> > On Wed, 30 Sep 2015 13:37:22 +0300
> > Vlad Zolotarov wrote:
> >
> > >
> > >
> > > On 09/30/15 00:49, Michael S. Tsirkin wrote:
> > > > On Tue, Sep 29, 20
Dear Bruce Richardson and DPDK experts.
Thank you very much for your precious answer.
I found it. It seems very short and simple.
Thank you very much.
I have another question.
I don't know how the following steps work from new_tag to match variables.
/* in dpdk library. ~/dpdk-?.?.?/lib/libr
On Wed, Sep 30, 2015 at 10:28:07AM -0700, Stephen Hemminger wrote:
> On Wed, 30 Sep 2015 13:37:22 +0300
> Vlad Zolotarov wrote:
>
> >
> >
> > On 09/30/15 00:49, Michael S. Tsirkin wrote:
> > > On Tue, Sep 29, 2015 at 02:46:16PM -0700, Stephen Hemminger wrote:
> > >> On Tue, 29 Sep 2015 23:54:54
According to XL710 datasheet:
RX QLEN restrictions: When the PXE_MODE flag in the GLLAN_RCTL_0
register is cleared, the QLEN must be whole number of 32
descriptors.
TX QLEN restrictions: When the PXE_MODE flag in the GLLAN_RCTL_0
register is cleared, the QLEN must be whole number of 32
descriptors.
Add hierarchy-file support to the DPDK libraries, modules,
binary files, nic bind files and documentation,
when invoking "make install-fhs" (filesystem hierarchy standard)
runtime files will be by default installed in:
$(DESTDIR)/$(BIN_DIR) where BIN_DIR=/usr/bin (binary files)
$(DESTDIR)/$(SBIN_DI
Add hierarchy-file support to the DPDK makefiles, scripts,
examples, tools, config files and headers,
when invoking "make install-sdk" makefiles, scripts,
examples, tools, config files will be installed in:
$(DESTDIR)/$(SDK_DIR)
and headers will be installed in:
$(DESTDIR)/$(INCLUDE_DIR)
where SDK_
Add hierarchy-file support to the DPDK nic bind files,
when invoking "make install-sbin" nic bind files will
be installed by default in: $(DESTDIR)/$(SBIN_DIR)
where SBIN_DIR=/usr/sbin/dpdk_nic_bind by default,
you can override SBIN_DIR var.
This hierarchy is based on:
http://www.freedesktop.org/so
Add hierarchy-file support to the DPDK documentation,
when invoking "make install-doc" documentation files will
be installed in: $(DESTDIR)/$(DOC_DIR) where
DOC_DIR=$(DESTDIR)/usr/share/doc/dpdk by default, you can
override DOC_DIR var.
This hierarchy is based on:
http://www.freedesktop.org/softwar
Add hierarchy-file support to the DPDK modules,
when invoking "make install-mod" modules will be
installed in: $(DESTDIR)/$(KERNEL_DIR)
if RTE_EXEC_ENV=linuxapp then
KERNEL_DIR=/lib/modules/$(uname -r)/build
else KERNEL_DIR=/boot/modules
by default, you can override KERNEL_DIR var.
This hierarchy i
Add hierarchy-file support to the DPDK libraries,
when invoking "make install-lib" libraries will
be installed in: $(DESTDIR)/$(LIB_DIR)
if architecture is 64 bits then LIB_DIR=/usr/lib64
else LIB_DIR=/usr/lib by default, you can override
LIB_DIR var.
This hierarchy is based on:
http://www.freedesk
Add hierarchy-file support to the DPDK app files,
when invoking "make install-bin" app files will
be installed in: $(DESTDIR)/$(BIN_DIR)
where BIN_DIR=/usr/bin by default, you can
override BIN_DIR var.
This hierarchy is based on:
http://www.freedesktop.org/software/systemd/man/file-hierarchy.html
Add hierarchy-file support to the DPDK headers,
when invoking "make install-headers" headers will
be installed in: $(DESTDIR)/$(INCLUDE_DIR)
where INCLUDE_DIR=/usr/include/dpdk by default,
you can override INCLUDE_DIR var.
This hierarchy is based on:
http://www.freedesktop.org/software/systemd/man/
DPDK package lacks of a mechanism to install libraries, headers
applications, kernel modules and sdk files to a file system tree.
This patch set allows to install files according to the next
proposal:
http://www.freedesktop.org/software/systemd/man/file-hierarchy.html
Using rules support is possi
> -Original Message-
> From: Singh, Jasvinder
> Sent: Wednesday, September 30, 2015 6:46 PM
> To: dev at dpdk.org
> Cc: Dumitrescu, Cristian
> Subject: [PATCH] ip_pipeline: add flow id parameter to flow classification
>
> This patch adds flow id field to the flow
> classification table e
> -Original Message-
> From: Singh, Jasvinder
> Sent: Wednesday, September 30, 2015 6:06 PM
> To: dev at dpdk.org
> Cc: Dumitrescu, Cristian
> Subject: [PATCH] ip_pipeline: add flow actions pipeline
>
>
> Signed-off-by: Jasvinder Singh
> ---
> examples/ip_pipeline/Makefile
This patch adds flow id field to the flow
classification table entries and adds table action
handlers to read flow id from table entry and
write it into the packet meta-data. The flow_id
(32-bit) parameter is also added to CLI commands
flow add, flow delete, etc.
Signed-off-by: Jasvinder Singh
--
On 09/30/2015 06:21 PM, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 05:53:54PM +0300, Avi Kivity wrote:
>> On 09/30/2015 05:39 PM, Michael S. Tsirkin wrote:
>>> On Wed, Sep 30, 2015 at 04:05:40PM +0300, Avi Kivity wrote:
On 09/30/2015 03:27 PM, Michael S. Tsirkin wrote:
> On Wed, S
On Wed, Sep 30, 2015 at 03:50:09PM +0300, Vlad Zolotarov wrote:
> How not virtualizing iommu forces "all or nothing" approach?
Looks like you can't limit an assigned device to only access part of
guest memory that belongs to a given process. Either let it access all
of guest memory ("all") or don
Up till now pipeline was bound to thread selected in the initial config.
This patch allows binding pipeline to other threads at runtime using CLI
commands.
Signed-off-by: Maciej Gajdzica
---
examples/ip_pipeline/Makefile |1 +
examples/ip_pipeline/app.h
On Wed, Sep 30, 2015 at 05:53:54PM +0300, Avi Kivity wrote:
> On 09/30/2015 05:39 PM, Michael S. Tsirkin wrote:
> >On Wed, Sep 30, 2015 at 04:05:40PM +0300, Avi Kivity wrote:
> >>
> >>On 09/30/2015 03:27 PM, Michael S. Tsirkin wrote:
> >>>On Wed, Sep 30, 2015 at 03:16:04PM +0300, Vlad Zolotarov wro
Up till now pipeline was bound to thread selected in the initial config.
This patch allows binding pipeline to other threads at runtime using CLI
commands.
Signed-off-by: Maciej Gajdzica
---
examples/ip_pipeline/Makefile |1 +
examples/ip_pipeline/app.h
Flow actions pipeline is an extension of
flow-classification pipeline. Some of the
operations of flow classification pipeline such
as traffic metering/marking(for e.g. Single
Rate Three Color Marker (srTCM), Two Rate Three
Color Marker trTCM)), policer can be performed
separately in flow action pip
On 09/30/2015 05:39 PM, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 04:05:40PM +0300, Avi Kivity wrote:
>>
>> On 09/30/2015 03:27 PM, Michael S. Tsirkin wrote:
>>> On Wed, Sep 30, 2015 at 03:16:04PM +0300, Vlad Zolotarov wrote:
On 09/30/15 15:03, Michael S. Tsirkin wrote:
> On Wed,
On Wed, Sep 30, 2015 at 04:05:40PM +0300, Avi Kivity wrote:
>
>
> On 09/30/2015 03:27 PM, Michael S. Tsirkin wrote:
> >On Wed, Sep 30, 2015 at 03:16:04PM +0300, Vlad Zolotarov wrote:
> >>
> >>On 09/30/15 15:03, Michael S. Tsirkin wrote:
> >>>On Wed, Sep 30, 2015 at 02:53:19PM +0300, Vlad Zolotaro
2015-09-30 10:52, Neil Horman:
> On Wed, Sep 30, 2015 at 10:28:53AM +0200, David Marchand wrote:
> > On Tue, Sep 29, 2015 at 9:25 PM, Stephen Hemminger <
> > stephen at networkplumber.org> wrote:
> >
> > > On Tue, 10 Mar 2015 09:14:28 -0400
> > > Neil Horman wrote:
> > > > I don't see how this wo
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Maciej Gajdzica
> Sent: Wednesday, September 30, 2015 5:26 PM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH v2 1/1] ip_pipeline: added dynamic pipeline
> reconfiguration
>
> Up till now pipeline was bound
Hi Neil
> > > > > +++ b/lib/librte_ether/rte_ethdev.h
> > > > > @@ -1635,8 +1635,23 @@ struct rte_eth_dev_data {
> > > > > all_multicast : 1, /**< RX all multicast mode ON(1) /
> OFF(0). */
> > > > > dev_started : 1, /**< Device state: STARTED(1) /
> STOPPED(0). */
>
This patch fixes bug in app_link_config.
Signed-off-by: Jasvinder Singh
---
examples/ip_pipeline/pipeline/pipeline_common_fe.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/examples/ip_pipeline/pipeline/pipeline_common_fe.c
b/examples/ip_pipeline/pipeline/pipeline_common_f
This patch adds following features to the
routing-pipeline to enable it for various NFV
use-cases;
1.Fast-path ARP table enable/disable
2.Double-tagged VLAN (Q-in-Q) packet enacapsulation
for the next-hop
3.MPLS encapsulation for the next-hop
4.Add colour (Traffic-class for QoS) to the MPLS tag
5.
Forgot to delete debug printfs, v2 incoming - Self NACK
> -Original Message-
> From: Gajdzica, MaciejX T
> Sent: Wednesday, September 30, 2015 6:11 PM
> To: dev at dpdk.org
> Cc: Gajdzica, MaciejX T
> Subject: [PATCH 1/1] ip_pipeline: added dynamic pipeline reconfiguration
>
> Up till now
On 09/30/2015 03:27 PM, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 03:16:04PM +0300, Vlad Zolotarov wrote:
>>
>> On 09/30/15 15:03, Michael S. Tsirkin wrote:
>>> On Wed, Sep 30, 2015 at 02:53:19PM +0300, Vlad Zolotarov wrote:
On 09/30/15 14:41, Michael S. Tsirkin wrote:
> On Wed
Documentation update about implementation details and requirements for
Dynamic RSS Configuration for Bonding.
Signed-off-by: Tomasz Kulasek
---
.../prog_guide/link_bonding_poll_mode_drv_lib.rst | 34 ++--
1 file changed, 32 insertions(+), 2 deletions(-)
diff --git a/doc/guide
Signed-off-by: Tomasz Kulasek
---
.../prog_guide/link_bonding_poll_mode_drv_lib.rst |8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/doc/guides/prog_guide/link_bonding_poll_mode_drv_lib.rst
b/doc/guides/prog_guide/link_bonding_poll_mode_drv_lib.rst
index 96e554f..0
This patch fills bonding port's stats with a sum of corresponding values
taken from bonded slaves, when stats are requested for bonding port.
v5 changes:
- removed queue_stats_mapping_set from eth_dev_ops of bonding device
Signed-off-by: Tomasz Kulasek
---
drivers/net/bonding/rte_eth_bond_pmd.
Signed-off-by: Tomasz Kulasek
---
app/test/Makefile|8 +
app/test/test_link_bonding_rssconf.c | 679 ++
2 files changed, 687 insertions(+)
create mode 100644 app/test/test_link_bonding_rssconf.c
diff --git a/app/test/Makefile b/app/test/M
Signed-off-by: Tomasz Kulasek
---
drivers/net/null/Makefile |2 +-
drivers/net/null/rte_eth_null.c |2 +-
drivers/net/null/rte_eth_null.h | 40 +
drivers/net/null/rte_pmd_null_version.map |7 +
4 files changed, 49
This implementation allows to set and read RSS configuration for null
device, and is used to validate right values propagation over the slaves,
in test units for dynamic RSS configuration for bonding.
v5 changes:
- replaced memcpy with rte_memcpy
Signed-off-by: Tomasz Kulasek
---
drivers/net/n
This patch adds a possibility to configure more than one queue on null
device.
v5 changes:
- fixed queues number configuration (using internals->nb_*_queues instead
of dev->data->nb_*_queues)
Signed-off-by: Tomasz Kulasek
---
drivers/net/null/rte_eth_null.c | 28 +++---
This patch initializes eth_dev->link_intr_cbs queue used when null pmd is
added to the bonding.
v5 changes:
- removed unnecessary malloc for eth_driver (rte_null_pmd)
Signed-off-by: Tomasz Kulasek
---
drivers/net/null/rte_eth_null.c |2 ++
1 file changed, 2 insertions(+)
diff --git a/driv
Bonding device implements independent management of RSS settings. It
stores its own copies of settings i.e. RETA, RSS hash function and RSS
key. It?s required to ensure consistency.
1) RSS hash function set for bonding device is maximal set of RSS hash
functions supported by all bonded devices. Th
OVERVIEW
1) Setting .rxmode.mq_mode for bonding device to ETH_MQ_RX_RSS makes bonding
device fully RSS-capable, so all slaves are synchronized with its configuration.
This mode is intended to provide RSS configuration as known from "dynamic RSS
configuration for one port" and made slaves t
On 09/30/15 15:27, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 03:16:04PM +0300, Vlad Zolotarov wrote:
>>
>> On 09/30/15 15:03, Michael S. Tsirkin wrote:
>>> On Wed, Sep 30, 2015 at 02:53:19PM +0300, Vlad Zolotarov wrote:
On 09/30/15 14:41, Michael S. Tsirkin wrote:
> On Wed, Sep
Ravi Kerur writes:
> Macros RTE_MBUF_DATA_DMA_ADDR and RTE_MBUF_DATA_DMA_ADDR_DEFAULT
> are defined in each PMD driver file. Move those macros into common
> lib/librte_mbuf/rte_mbuf.h file. PMD drivers include rte_mbuf.h
> file directly/indirectly hence no additionl header file inclusion
> is nec
> -Original Message-
> From: Singh, Jasvinder
> Sent: Wednesday, September 30, 2015 4:26 PM
> To: dev at dpdk.org
> Cc: Dumitrescu, Cristian
> Subject: [PATCH] ip_pipeline: fixed bug in app_link_config
>
> This patch fixes bug in app_link_config.
>
> Signed-off-by: Jasvinder Singh
> --
> -Original Message-
> From: Singh, Jasvinder
> Sent: Wednesday, September 30, 2015 4:21 PM
> To: dev at dpdk.org
> Cc: Dumitrescu, Cristian
> Subject: [PATCH] ip_pipeline: add more functions to routing-pipeline
>
> This patch adds following features to the
> routing-pipeline to enable i
From: Ian Betts
This commit adds the build controls for the
performance-thread sample
Signed-off-by: Ian Betts
---
config/common_linuxapp | 6
config/defconfig_i686-native-linuxapp-gcc| 6
config/defconfig_i686-native-linuxapp-icc| 6
config/defc
From: Ian Betts
This commit adds a simple pthread_shim example for the
cooperative included with this patchset.
The shim demonstrates a way in which legacy code writtem for
pthreads could be adapted to lighweight threads.
Signed-off-by: Ian Betts
---
examples/performance-thread/pthread_shim/M
From: Ian Betts
This commit adds an l3fwd derivative that allows multiple
EAL threads to be run on a single physical core or multiple
lightwieght threads to be run in an EAL thread.
Its purpose is to facilitate characterization of performance
with different threading models.
It depends on a sim
From: Ian Betts
This commit adds a cooperative scheduler subsystem in the
performance-thread sample application.
It is used in the performance-thread sample application
by the l3fwd-thread application to enable multiple
lightweight threads to be run in an EAL thread.
Signed-off-by: Ian Betts
-
From: Ian Betts
This commit adds documentation for the performance-thread
sample application.
Signed-off-by: Ian Betts
---
doc/guides/rel_notes/release_2_2.rst|6 +
doc/guides/sample_app_ug/index.rst |1 +
doc/guides/sample_app_ug/performance_thread.rst | 1221
From: Ian Betts
Performance thread example application
This example comprises layer 3 forwarding derivative intended to
facilitate characterization of performance with different
threading models, specifically:-
1. EAL threads running on different physical cores
2. EAL threads running on the s
This driver allows using PCI device with Message Signalled Interrupt
from userspace. The API is similar to the igb_uio driver used by the DPDK.
Via ioctl it provides a mechanism to map MSI-X interrupts into event
file descriptors similar to VFIO.
VFIO is a better choice if IOMMU is available, but
Allow UIO device driver to provide ioctl interface.
This allow additional API's for UIO.
Signed-off-by: Stephen Hemminger
---
drivers/uio/uio.c | 15 +++
include/linux/uio_driver.h | 3 +++
2 files changed, 18 insertions(+)
diff --git a/drivers/uio/uio.c b/drivers/uio/uio.
This is a new UIO device driver to allow supporting MSI-X and MSI devices
in userspace. It has been used in environments like VMware and older versions
of QEMU/KVM where no IOMMU support is available.
Stephen Hemminger (2):
*** BLURB HERE ***
Stephen Hemminger (2):
uio: add support for ioctls
Add separate functions to configure VMDQ and RSS.
Update dglort map and logic ports accordingly.
Reset MAC/VLAN filter after VMDQ configure was changed.
Signed-off-by: Shaopeng He
---
drivers/net/fm10k/fm10k_ethdev.c | 164 +--
1 file changed, 141 insertions(+
The patch does below things for fm10k MAC/VLAN filter:
- Add separate functions to add MAC address for VMDQ and
main VSI.
- Disable modification to VLAN filter in VMDQ mode.
- In device close phase, delete logic ports to remove all
the MAC/VLAN filters.
Signed-off-by: Shaopeng He
---
drivers
Add multi-queue checking in device configure process.
Currently, VMDQ and RSS are supported.
Signed-off-by: Shaopeng He
---
drivers/net/fm10k/fm10k_ethdev.c | 44
1 file changed, 44 insertions(+)
diff --git a/drivers/net/fm10k/fm10k_ethdev.c b/drivers/ne
This patch series adds VMDQ support to fm10k.
It includes the functions to configure VMDQ mode and
add MAC address for each VMDQ queue pool.
It also includes logic to do sanity check for
multi-queue settings.
1. implement rx_descriptor_done function in fm10k
2. make sure default VID available in
On Wed, Sep 30, 2015 at 03:16:04PM +0300, Vlad Zolotarov wrote:
>
>
> On 09/30/15 15:03, Michael S. Tsirkin wrote:
> >On Wed, Sep 30, 2015 at 02:53:19PM +0300, Vlad Zolotarov wrote:
> >>
> >>On 09/30/15 14:41, Michael S. Tsirkin wrote:
> >>>On Wed, Sep 30, 2015 at 02:26:01PM +0300, Vlad Zolotarov
On 09/30/15 15:03, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 02:53:19PM +0300, Vlad Zolotarov wrote:
>>
>> On 09/30/15 14:41, Michael S. Tsirkin wrote:
>>> On Wed, Sep 30, 2015 at 02:26:01PM +0300, Vlad Zolotarov wrote:
The whole idea is to bypass kernel. Especially for networking.
On Wed, 30 Sep 2015 14:55:03 -0700
Ravi Kerur wrote:
> +static inline uint64_t rte_mbuf_data_dma_addr(struct rte_mbuf *mb)
> +{
> + return ((uint64_t)((mb)->buf_physaddr + (mb)->data_off));
> +}
> +
> +static inline uint64_t rte_mbuf_data_dma_addr_default(struct rte_mbuf *mb)
> +{
> + ret
On Wed, Sep 30, 2015 at 02:53:19PM +0300, Vlad Zolotarov wrote:
>
>
> On 09/30/15 14:41, Michael S. Tsirkin wrote:
> >On Wed, Sep 30, 2015 at 02:26:01PM +0300, Vlad Zolotarov wrote:
> >>The whole idea is to bypass kernel. Especially for networking...
> >... on dumb hardware that doesn't support d
On Wed, Sep 30, 2015 at 12:41 PM, Aaron Conole wrote:
> Ravi Kerur writes:
>
> > Macros RTE_MBUF_DATA_DMA_ADDR and RTE_MBUF_DATA_DMA_ADDR_DEFAULT
> > are defined in each PMD driver file. Move those macros into common
> > lib/librte_mbuf/rte_mbuf.h file. PMD drivers include rte_mbuf.h
> > file di
Macros RTE_MBUF_DATA_DMA_ADDR and RTE_MBUF_DATA_DMA_ADDR_DEFAULT
are defined in each PMD driver file. Convert macros to inline
functions and move them to common lib/librte_mbuf/rte_mbuf.h file.
PMD drivers include rte_mbuf.h file directly/indirectly hence no
additionl header file inclusion is neces
On 09/30/15 14:41, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 02:26:01PM +0300, Vlad Zolotarov wrote:
>> The whole idea is to bypass kernel. Especially for networking...
> ... on dumb hardware that doesn't support doing that securely.
On a very capable HW that supports whatever security
Dear DPDK experts.
I am Ick-Sung Choi living in South Korea.
I have a question about DPDK? packet distributor library.
Is there any example application to used DPDK packet distributor library?
I am trying to experiment simple function using DPDK packet distributor library.
If I can study an ex
Handle only supported flow director behaviors
Signed-off-by: Andrey Chilikin
---
drivers/net/ixgbe/ixgbe_fdir.c |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_fdir.c b/drivers/net/ixgbe/ixgbe_fdir.c
index 5c8b833..cf0e8be 100644
--- a/drivers/n
Add support for RTE_ETH_FDIR_PASSTHRU flow director behavior so output queue is
assigned by other filters
Signed-off-by: Andrey Chilikin
---
drivers/net/i40e/i40e_fdir.c | 12 ++--
1 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/drivers/net/i40e/i40e_fdir.c b/drivers/n
Add new flow director behavior RTE_ETH_FDIR_PASSTHRU to assign a queue by other
filters
Signed-off-by: Andrey Chilikin
---
lib/librte_ether/rte_eth_ctrl.h |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/lib/librte_ether/rte_eth_ctrl.h b/lib/librte_ether/rte_eth_ctrl.h
i
This patch set adds new flow director behavior "passthru" on Intel X(L)710 NIC.
When this mode is selected flow director will direct packet to LAN while the
queue is defined by other filters. This can be used to extract flexible payload
to RX desriptor with the flow director filter while targeted
On Wed, Sep 30, 2015 at 02:26:01PM +0300, Vlad Zolotarov wrote:
> The whole idea is to bypass kernel. Especially for networking...
... on dumb hardware that doesn't support doing that securely.
Colour me unimpressed.
--
MST
On Wed, 30 Sep 2015 23:09:33 +0300
Vlad Zolotarov wrote:
>
>
> On 09/30/15 22:39, Michael S. Tsirkin wrote:
> > On Wed, Sep 30, 2015 at 10:06:52PM +0300, Vlad Zolotarov wrote:
> How would iommu
> virtualization change anything?
> >>> Kernel can use an iommu to limit device access to m
On 09/30/15 13:58, Michael S. Tsirkin wrote:
> On Wed, Sep 30, 2015 at 01:37:22PM +0300, Vlad Zolotarov wrote:
>>
>> On 09/30/15 00:49, Michael S. Tsirkin wrote:
>>> On Tue, Sep 29, 2015 at 02:46:16PM -0700, Stephen Hemminger wrote:
On Tue, 29 Sep 2015 23:54:54 +0300
"Michael S. Tsirkin
On Wed, Sep 30, 2015 at 09:18:53AM -0400, Neil Horman wrote:
> > +}
> > diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
> > index fa06554..9cd262b 100644
> > --- a/lib/librte_ether/rte_ethdev.h
> > +++ b/lib/librte_ether/rte_ethdev.h
> > @@ -1635,8 +1635,23 @@ struct rte_
On Wed, Sep 30, 2015 at 09:14:48AM -0400, Neil Horman wrote:
> On Wed, Sep 30, 2015 at 10:56:04AM +0100, Bruce Richardson wrote:
> > On Tue, Sep 29, 2015 at 03:08:12PM -0400, Neil Horman wrote:
> > > On Mon, Sep 28, 2015 at 02:03:20PM +0100, Bernard Iremonger wrote:
> > > > add dev_flags to rte_eth
On Wed, Sep 30, 2015 at 01:37:22PM +0300, Vlad Zolotarov wrote:
>
>
> On 09/30/15 00:49, Michael S. Tsirkin wrote:
> >On Tue, Sep 29, 2015 at 02:46:16PM -0700, Stephen Hemminger wrote:
> >>On Tue, 29 Sep 2015 23:54:54 +0300
> >>"Michael S. Tsirkin" wrote:
> >>
> >>>On Tue, Sep 29, 2015 at 07:41:
On 09/30/15 00:49, Michael S. Tsirkin wrote:
> On Tue, Sep 29, 2015 at 02:46:16PM -0700, Stephen Hemminger wrote:
>> On Tue, 29 Sep 2015 23:54:54 +0300
>> "Michael S. Tsirkin" wrote:
>>
>>> On Tue, Sep 29, 2015 at 07:41:09PM +0300, Vlad Zolotarov wrote:
The security breach motivation u brou
Hi, Bruce,
> -Original Message-
> From: Richardson, Bruce
> Sent: Tuesday, September 29, 2015 10:23 PM
> To: Ananyev, Konstantin
> Cc: Chen, Jing D; dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 06/14] fm10k: add Vector RX function
>
> On Tue, Sep 29, 2015 at 01:14:26PM +, Ananyev,
> -Original Message-
> From: Ananyev, Konstantin
> Sent: Tuesday, September 29, 2015 9:14 PM
> To: Chen, Jing D; dev at dpdk.org
> Subject: RE: [dpdk-dev] [PATCH 06/14] fm10k: add Vector RX function
>
>
> > + /* A. load 4 packet in one loop
> > +* [A*. mask out 4 unused dirty fiel
Add a performance test for ring pmd, comparing performance of the pmd
compared to the basic rte_ring APIs.
Signed-off-by: Bruce Richardson
---
app/test/Makefile | 1 +
app/test/test_pmd_ring_perf.c | 188 ++
2 files changed, 189 insertions(+)
Add a one-parameter function to take an existing rte_ring and wrap it as
an ethdev, returning the port id of the new ethdev instance.
Signed-off-by: Bruce Richardson
---
drivers/net/ring/rte_eth_ring.c | 8
drivers/net/ring/rte_eth_ring.h | 14 ++
driver
Add a new field to the rte_ring structure to store the memzone pointer which
contains the ring. For rings created using rte_ring_create(), the field will
be set automatically.
This new field will allow users of the ring to query the numa node a ring is
allocated on, or to get the physical address
The ring ethdev creation function creates an ethdev, but does not
actually set it up for use. Even if it's just a single ring, the user
still needs to create a mempool, call rte_eth_dev_configure, then call
rx and tx setup functions before the ethdev can be used.
This patch changes things so that
This patchset makes it easier to create ring pmd instances from code, by
providing a simple ring->ethdev wrapper function and also ensuring that
any created rings are ready for use immediately, without having to call
configure and rx/tx queue setup.
This set also contains a set of unit tests to co
On Wed, Sep 30, 2015 at 08:41:04PM +0900, ??? wrote:
> Dear Bruce Richardson and DPDK experts.
>
> Thank you very much for your precious answer.
>
> I found it. It seems very short and simple.
>
> Thank you very much.
>
> I have another question.
>
> I don't know how the following steps
Macros RTE_MBUF_DATA_DMA_ADDR and RTE_MBUF_DATA_DMA_ADDR_DEFAULT
are defined in each PMD driver file. Move those macros into common
lib/librte_mbuf/rte_mbuf.h file. PMD drivers include rte_mbuf.h
file directly/indirectly hence no additionl header file inclusion
is necessary.
v2:
> Changed both
1 - 100 of 148 matches
Mail list logo
