This patch adds functional and performance tests for membership
library.
Signed-off-by: Yipeng Wang
---
test/test/Makefile | 3 +
test/test/test_member.c | 682 +++
test/test/test_member_perf.c | 643 +++
This patch enables the Membership library.
Signed-off-by: Yipeng Wang
---
MAINTAINERS| 7 +++
config/common_base | 5 +
lib/librte_member/Makefile | 2 ++
mk/rte.app.mk | 2 ++
4 files changed, 16 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
For key search, the signatures of all entries are compared against
the signature of the key that is being looked up. Since all
signatures are contguously put in a bucket, they can be compared
with vector instructions (AVX2), achieving higher lookup performance.
This patch adds AVX2 implementation
Bloom Filter (BF) [1] is a well-known space-efficient
probabilistic data structure that answers set membership queries.
Vector of Bloom Filters (vBF) is an extension to traditional BF
that supports multi-set membership testing. Traditional BF will
return found or not-found for each key. vBF will al
One of the set-summray structure is hash-table based
set-summary (HTSS). One example is cuckoo filter [1].
Comparing to a traditional hash table, HTSS has a much more
compact structure. For each element, only one signature and
its corresponding set ID is stored. No key comparison is required
duri
Membership library is an extension and generalization of a traditional
filter (for example Bloom Filter) structure. In general, the Membership
library is a data structure that provides a "set-summary" and responds
to set-membership queries of whether a certain element belongs to a
set(s). A members
DPDK Membership Library provides an API that can be used by many DPDK
applications to conduct one or more set-membership tests (we mention some
possible use cases below, but interested readers can refer to
[1] for a wider survey of use cases).
The basic functionalities of the Membership Library in
Thanks Stephen.
Agree. Will remove dependency on PCI for now. And num_queues should have been
declared as uint16_t.
> -Original Message-
> From: Stephen Hemminger [mailto:step...@networkplumber.org]
> Sent: Friday 1 September 2017 21:04
> To: Mokhtar, Amr
> Cc: dev@dpdk.org
> Subject: Re
> +/* Forward declaration */
> +struct rte_pci_device;
> +
> +/** Device information structure used by an application to discover a devices
> + * capabilities and current configuration
> + */
> +struct rte_bbdev_info {
> + int socket_id; /**< NUMA socket that device is on */
> + const cha
Hello,
Following on the proposal of bbdev, is there any comments. You feedback is
highly appreciated.
May I propose to take this topic up to discussion by the Technical Board at
next convenient time?
Regards,
Amr
> -Original Message-
> From: Mokhtar, Amr
> Sent: Friday 25 August 2017 14
On 8/30/2017 7:17 PM, Andrew Rybchenko wrote:
> Few base driver patches have checkpatches.sh errors due to difference
> in coding conventions.
>
> Andrew Rybchenko (1):
> net/sfc: use RSS key size define from base driver
>
> Ivan Malov (3):
> net/sfc: fix unused variable warnings in RSS-agnos
Port info command prints information about all supported flow types,
including dynamically mapped ones.
Signed-off-by: Kirill Rybalchenko
---
app/test-pmd/config.c | 9 ++---
1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index 3ae
Hey Evgeny,
Please see inline.
On Thu, Aug 31, 2017 at 9:35 AM, Evgeny Agronsky
wrote:
> I'm basicly asking because of it's poor performance under high
Well, it is not the academic cuckoo hash implementation, so the
performance is not that bad and it also utilizes cache ;)
Please have a look at
Update 'ddp get info' command to display protocols defined in a profile
Signed-off-by: Kirill Rybalchenko
---
app/test-pmd/cmdline.c | 98 +-
1 file changed, 97 insertions(+), 1 deletion(-)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdli
This patch adds new package info types to get list of protocols,
pctypes and ptypes defined in a profile
Signed-off-by: Kirill Rybalchenko
---
drivers/net/i40e/rte_pmd_i40e.c | 167
drivers/net/i40e/rte_pmd_i40e.h | 25 ++
2 files changed, 192 insert
This patch adds ability to request information about protocols defined in
dynamic
device personalization profile
v2:
Some code style warnings were removed
Kirill Rybalchenko (2):
net/i40e: get information about protocols defined in ddp profile
app/testpmd: get information about protocols def
On 8/30/2017 4:59 PM, Gaetan Rivet wrote:
> This is unacceptable behavior.
>
> Fixes: a0194d828100 ("net/failsafe: add flexible device definition")
> Fixes: 35ffe4208140 ("net/failsafe: fix missing pclose after popen")
> Cc: sta...@dpdk.org
>
> Reported-by: Matan Azrad
> Signed-off-by: Gaetan Ri
On Thu, 31 Aug 2017 23:46:59 +0200
Markus Theil wrote:
> This patch substitutes the custom MSI/MSI-X mask code and uses
> already existing kernel APIs. Feedback/small corrections to the previous
> patch of this series are also incorporated.
>
> Signed-off-by: Markus Theil
> ---
> lib/librte_ea
On Fri, 1 Sep 2017 16:57:00 +0800
Xiaoyun Li wrote:
> This patch dynamically selects functions of memcpy at run-time based
> on CPU flags that current machine supports. This patch uses function
> pointers which are bind to the relative functions at constrctor time.
> In addition, AVX512 instruct
Hi,
FYI, Just to note a bug I found when using FMV with -O flag:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81128
I see the fix is on the trunk now but not sure if it's been back ported to
other point releases so dev's might need to be careful about which gcc versions
they use and be a
Hello Shachar,
Thank you for the patch.
Please could you specify patch set you used in new version, IIUC,
it's [PATCH 0/5] net/mlx5: cleanups" and [PATCH 0/5] net/mlx5:
enable/disable vPMD and some cleanups
by Nélio Laranjeiro.
Also rdma-core version, I used HEAD (commit
202b39622e67b0424
CC Joao's correct address
On Wed, Aug 30, 2017 at 06:10:28PM +, Jianfeng Tan wrote:
> Following the calls on the mailing list:
> http://dpdk.org/ml/archives/dev/2017-June/068151.html
> The Technical Board decided to drop Xen dom0 support from EAL:
> http://dpdk.org/ml/archives/dev/2017
Hello,
I'm using dpdk lib for various purposes, lately I needed to use a hashmap
for a certain needs and the one I found as part of the library was cuckoo
hashing based hashmap, I'm just wondering is there any reason for cuckoo
hashing? I'm basicly asking because of it's poor performance under hig
Remove unnecessary check for new flow type for rss hash
filter update.
Signed-off-by: Kirill Rybalchenko
---
lib/librte_ether/rte_ethdev.c | 8
1 file changed, 8 deletions(-)
diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index 0597641..c470997 100644
--- a/
Add new commands to manipulate with dynamic flow type to
pctype mapping table in i40e PMD.
Commands allow to print table, modify it and reset to default value.
Signed-off-by: Kirill Rybalchenko
---
app/test-pmd/cmdline.c | 311 +++-
doc/guides/testpmd
Add new functions which allow modify, return or reset to default
the contents of flow type to pctype dynamic mapping table.
Signed-off-by: Kirill Rybalchenko
---
drivers/net/i40e/rte_pmd_i40e.c | 98 +
drivers/net/i40e/rte_pmd_i40e.h | 61 +
Implement dynamic mapping of software flow types to hardware pctypes.
This allows to add new flow types and pctypes for DDP without changing
API of the driver. The mapping table is located in private
data area for particular network adapter and can be individually
modified with set of appropriate f
Implement dynamic mapping of software flow types to hardware pctypes.
This allows to map new flow types to pctypes without changing
API of the driver.
v2:
Remove unnecessary check for new flow types.
Re-arrange patchset to avoid compillation errors.
Remove unnecessary usage of statically defined f
On 8/30/2017 10:08 AM, Andrew Rybchenko wrote:
> From: Ivan Malov
>
> Support for the feature is added to EFX Rx datapath
>
> Signed-off-by: Ivan Malov
> Signed-off-by: Andrew Rybchenko
Series applied to dpdk-next-net/master, thanks.
On Fri, 2017-09-01 at 14:55 +0100, Bruce Richardson wrote:
> On Fri, Sep 01, 2017 at 02:32:52PM +0100, Luca Boccassi wrote:
> > On Fri, 2017-09-01 at 11:04 +0100, Bruce Richardson wrote:
> > > Support building igb_uio using meson and ninja. For this, we
> > > still
> > > use the
> > > kernel's kbui
During initialization, mac address is read from configuration bar. This is
the default option when using VFs.
This patch adds support for reading the mac address using the NSPU
interface when PMD works with the PF.
Signed-off-by: Alejandro Lucero
---
drivers/net/nfp/nfp_net.c | 59 +
NFP PMD implement now PF and VF drivers. Although the driver
functionality is the same by now, except for initialization, it
will change with future PF additions.
A new feature is required for describing the firmware upload
capability coming with the NFP PF now, so the PF file will be
updated soon
Signed-off-by: Alejandro Lucero
---
doc/guides/rel_notes/release_17_11.rst | 9 +
1 file changed, 9 insertions(+)
diff --git a/doc/guides/rel_notes/release_17_11.rst
b/doc/guides/rel_notes/release_17_11.rst
index 170f4f9..8773b08 100644
--- a/doc/guides/rel_notes/release_17_11.rst
+++ b
NFP PMD has now support for both, PF and VFs. This patch updates
the guide and give some information about implications.
Signed-off-by: Alejandro Lucero
---
doc/guides/nics/nfp.rst | 77 +
1 file changed, 58 insertions(+), 19 deletions(-)
diff --g
PMD has to configure the hardware port: link up when port started and
link down when port stopped. This is not required for VFs but it is
for PF ports.
A minor refactoring in PMD stop and close functions is done because the
Link down needs to happen just when device is stopped.
Signed-off-by: Ale
Adding a new NSPU command for being able to read and write the ethernet
port table from/to the NFP. This will allow the PMD to put the Link up
or down when a port is started or stopped. Until now, this was performed
by the firmware independently of PMD functionality.
The ethernet port table has al
NFP can support several physical ports per PF device. Depending on
firmware info, one or more eth_dev objects will need to be created.
This patch adds the call to create just one eth_dev by now with future
commits supporting the multiport option. Once the eth_dev has been
created, probe function i
A NFP PF PCI device can have several physical ports, up to 8. Because
DPDK core creates one eth_dev per PCI device, nfp pf probe function
is used. Number of PF ports is obtained from firmware symbol using
NSPU API. Inside PF probe function an eth_dev per port is created and
nfp_net_init invoked for
nfp_net_init is where a dpdk port related to a eth_dev is initialized.
NFP VF vNICs use VF PCI BARs as they come after SRIOV is enabled. But for
NFP PF vNIC just a subset of PF PCI BARs are used.
This patch adds support for mapping the right PCI BAR subsets for the PF
vNIC. It uses the NSPU API fu
NFP vNICs use a subset of PCI device BARs. vNIC config bar depends on
firmware symbol defining how to map it through a NFP expansion bar.
This patch adds a NSPU API function for getting a vNIC config bar
mapped through a expansion bar giving a firmware symbol. The PMD will
use the PCI bar offset r
NFP vNICs use a subset of PCI device BARs. vNIC rx/tx bars point to
NFP hardware queues unit. Unline vNIC config bar, the NFP address is
always the same so the NFP expansion bar configuration always uses
the same hardcoded physical address.
This patch adds a NSPU API function for getting vNIC rx/t
PMD will use this function for uploading the firmware. First, a
symbol resolution is done for finding out if there is a firmware
already there. If not, a NFP reset is called before using NSPU
fw upload code.
PMD PF probe function is now using this logic.
Signed-off-by: Alejandro Lucero
---
driv
Firmware has symbols helping to configure things like number of
PF ports, vNIC BARs addresses inside NFP memories, or ethernet
link state. Different firmware apps have different things to map
and likely different internal NFP addresses to use.
Host drivers can use the NSPU interface for getting sy
Using NSPU interface for fw upload. Firmware file needs to be
installed in specific path inside system firmware directory.
NSPU buffer is used for writing the firmware before sending the
command.
Signed-off-by: Alejandro Lucero
---
drivers/net/nfp/nfp_nspu.c | 66 +++
NSPU interface declares a buffer controlled by the NFP NSP service
processor. It is possible to send commands to the NSP using the NSPU
and this buffer for data related to the command. A command can imply
buffer read, buffer write, both or none.
Initial command for resetting the firmware is added
A NFP PF PCI devices can have PCI ID 4000 or 6000.
Signed-off-by: Alejandro Lucero
---
drivers/net/nfp/nfp_net.c | 4
drivers/net/nfp/nfp_net_pmd.h | 1 +
2 files changed, 5 insertions(+)
diff --git a/drivers/net/nfp/nfp_net.c b/drivers/net/nfp/nfp_net.c
index cb48fdc..ca481e4 100644
-
Configuring the NFP PMD for using the PF requires access through the
NSPU interface for device configuration. This patch adds a specific probe
function for the PF which uses the NSPU interface. Just basic NSPU access
is done by now reading the NSPU ABI version.
NSPU ABI version needs to be greater
NFP PMD has just had support for SRIOV VFs until now. This patch set adds
support for the PF, but just for being used as another DPDK port. No VF
management is added by now.
NFP is a programmable device and it supports virtual NICs (vNICs) through
firmware implementation. Different firmware applic
Working with the PF requires access to the NFP for basic configuration.
NSP is the NFP Service Processor helping with hardware and firmware
configuration. NSPU is the NSP user space interface for working with the
NSP.
Configuration through NSPU allows to create PCI BAR windows for accessing
differ
On Fri, Sep 01, 2017 at 02:32:52PM +0100, Luca Boccassi wrote:
> On Fri, 2017-09-01 at 11:04 +0100, Bruce Richardson wrote:
> > Support building igb_uio using meson and ninja. For this, we still
> > use the
> > kernel's kbuild system, by calling out to make, since it's safer and
> > easier
> > than
From: Alan Dewar
The RED code stores the maximum threshold is a 32-bit integer as a
pseudo fixed-point floating number with 10 fractional bits. Twelve
other bits are used to encode the filter weight, leaving just 10 bits
for the queue length. This limits the maximum queue length supported
by RE
On Fri, 2017-09-01 at 11:04 +0100, Bruce Richardson wrote:
> Support building igb_uio using meson and ninja. For this, we still
> use the
> kernel's kbuild system, by calling out to make, since it's safer and
> easier
> than trying to reproduce that in meson. A list of suitable file
> dependencies
> -Original Message-
> From: Hemant Agrawal [mailto:hemant.agra...@nxp.com]
> Sent: Friday, September 01, 2017 3:41 AM
> To: David Harton (dharton) ; tho...@monjalon.net;
> ferruh.yi...@intel.com; ajit.khapa...@broadcom.com; John Daley (johndale)
> ; konstantin.anan...@intel.com; jingjing
On Fri, Sep 01, 2017 at 11:52:17AM +0200, Olivier MATZ wrote:
> > > @@ -1534,7 +1554,6 @@ eth_virtio_dev_init(struct rte_eth_dev *eth_dev)
> > > RTE_BUILD_BUG_ON(RTE_PKTMBUF_HEADROOM < sizeof(struct
> > > virtio_net_hdr_mrg_rxbuf));
> > >
> > > eth_dev->dev_ops = &virtio_eth_dev_ops;
> > > -
On Fri, Sep 01, 2017 at 11:58:07AM +0200, Olivier MATZ wrote:
> On Fri, Sep 01, 2017 at 05:25:38PM +0800, Yuanhan Liu wrote:
> > On Thu, Aug 31, 2017 at 03:40:14PM +0200, Olivier Matz wrote:
> > > Split use_simple_rxtx into use_simple_rx and use_simple_tx,
> > > and ensure that only use_simple_tx i
On Fri, Sep 01, 2017 at 12:24:40PM +0100, Ferruh Yigit wrote:
> Hi Adrien,
>
> On 9/1/2017 9:06 AM, Adrien Mazarguil wrote:
> > The main purpose of this large series is to relieve the mlx4 PMD from its
> > dependency on Mellanox OFED to instead rely on the standard rdma-core
> > package provided b
Hi Adrien,
On 9/1/2017 9:06 AM, Adrien Mazarguil wrote:
> The main purpose of this large series is to relieve the mlx4 PMD from its
> dependency on Mellanox OFED to instead rely on the standard rdma-core
> package provided by Linux distributions.
>
> While compatibility with Mellanox OFED is pres
On 8/30/2017 8:11 AM, Shijith Thotton wrote:
> Support MTU change in the range ETHER_MIN_MTU to PF_MTU. A drop in PF
> MTU lowers VF MTU if it goes out of range.
>
> Signed-off-by: Shijith Thotton
Applied to dpdk-next-net/master, thanks.
> -Original Message-
> From: Adrien Mazarguil [mailto:adrien.mazarg...@6wind.com]
> Sent: Friday, September 01, 2017 4:06 AM
> To: dev@dpdk.org
> Cc: Gaëtan Rivet; Legacy, Allain
> Subject: [PATCH v2 03/51] net/mlx4: check max number of ports dynamically
>
> Use maximum number reported by
On 8/23/2017 10:44 AM, Nélio Laranjeiro wrote:
> On Sun, Aug 13, 2017 at 03:25:12PM +0300, Matan Azrad wrote:
>> The corrupted code doesn't return error when probe function
>> fails due to error in device mac address getting.
>> By this way, the probe function may return success even if the
>> ETH
> -Original Message-
> From: Li, Xiaoyun
> Sent: Friday, September 1, 2017 10:29 AM
> To: Ananyev, Konstantin ; Richardson, Bruce
>
> Cc: dev@dpdk.org; Wang, Zhihong ; Zhang, Qi Z
> ; Lu, Wenzhuo
>
> Subject: RE: [dpdk-dev] [PATCH v2 1/3] eal/x86: run-time dispatch over memcpy
>
> Hi
On Fri, Sep 01, 2017 at 11:03:59AM +0100, Bruce Richardson wrote:
> Following on from the two previous RFCs [1] [2], here is a cleaned up
> patchset to serve as a start-point for getting all of DPDK building with
> meson and ninja.
>
Just for reference, I've test-compiled this on, Fedora 26 with
On Fri, Sep 01, 2017 at 11:28:54AM +0100, James Poole wrote:
> Fixes: dc39e2f359b5 ("eventdev: add ring structure for events")
> Cc: bruce.richard...@intel.com
>
> Signed-off-by: James Poole
Acked-by: Bruce Richardson
On Fri, Sep 01, 2017 at 11:45:42AM +0200, Maxime Coquelin wrote:
> On 08/24/2017 04:19 AM, Tiwei Bie wrote:
> > This patch adaptively batches the small guest memory copies.
> > By batching the small copies, the efficiency of executing the
> > memory LOAD instructions can be improved greatly, becaus
Hi Ferruh,
On Fri, Aug 18, 2017 at 02:39:18PM +0100, Ferruh Yigit wrote:
> On 8/1/2017 5:53 PM, Adrien Mazarguil wrote:
> > Verbs support for RSS, inline receive and extended device query calls has
> > not been optional for a while. Their absence is untested and is therefore
> > unsupported.
> >
Add to the contributors guide details on how to add libraries and drivers
and integrate them with the DPDK build system(s).
Signed-off-by: Bruce Richardson
---
doc/guides/contributing/coding_style.rst | 214 +++
1 file changed, 214 insertions(+)
diff --git a/doc/guid
Normally, each library has it's own version number based on the ABI.
Add an option to have all libs just use the DPDK version number as the
.so version.
Signed-off-by: Bruce Richardson
---
drivers/meson.build | 8 +++-
lib/meson.build | 8 +++-
meson_options.txt | 1 +
3 files chan
Have the "usertools" scripts installed when doing a DPDK install using
ninja. They will be copies to $prefix/bin, generally /usr/local/bin,
alongside testpmd.
Signed-off-by: Bruce Richardson
---
meson.build | 3 ++-
usertools/meson.build | 32
2 files
Signed-off-by: Bruce Richardson
---
app/meson.build | 32 +
app/test-pmd/meson.build | 74
meson.build | 3 +-
3 files changed, 108 insertions(+), 1 deletion(-)
create mode 100644 app/meson.build
create
Add the af_packet, null, pcap and ring PMDs to the meson build.
Signed-off-by: Bruce Richardson
---
config/rte_config.h | 4
drivers/net/af_packet/meson.build | 35 +
drivers/net/meson.build | 2 +-
drivers/net/null/meson.build
Add the e1000, fm10k, i40e and ixgbe drivers to the meson & ninja build.
Signed-off-by: Bruce Richardson
---
config/rte_config.h| 12 +++
drivers/net/e1000/base/meson.build | 65 ++
drivers/net/e1000/meson.build | 44 ++
Signed-off-by: Bruce Richardson
---
config/rte_config.h| 8
drivers/crypto/meson.build | 2 +-
drivers/crypto/qat/meson.build | 41 +
3 files changed, 50 insertions(+), 1 deletion(-)
create mode 100644 drivers/crypto/qat/meson.bu
Signed-off-by: Bruce Richardson
---
drivers/meson.build | 2 +-
drivers/net/meson.build | 35 +++
2 files changed, 36 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/meson.build
diff --git a/drivers/meson.build b/drivers/meson.build
index 515e3cb
Signed-off-by: Bruce Richardson
---
drivers/crypto/meson.build | 2 +-
drivers/crypto/openssl/meson.build | 38 ++
2 files changed, 39 insertions(+), 1 deletion(-)
create mode 100644 drivers/crypto/openssl/meson.build
diff --git a/drivers/crypto/meso
Add in a crypto driver class, and implement compiling the null crypto
instance.
Signed-off-by: Bruce Richardson
---
drivers/crypto/meson.build | 35 +++
drivers/crypto/null/meson.build | 33 +
drivers/meson.build |
Add non-EAL libraries to DPDK build. The compat lib is a special case,
along with the previously-added EAL, but all other libs can be build using
the same set of commands, where the individual meson.build files only need
to specify their dependencies, source files, header files and ABI versions.
S
With build infrastructure in place, add in mempool drivers to the build.
Signed-off-by: Bruce Richardson
---
drivers/mempool/meson.build | 35 +++
drivers/mempool/ring/meson.build | 32
drivers/mempool/stack/meson.build | 32
Add in the top-level drivers meson.build file to start adding in building
of PMDs. Since all PMDs are built in pretty much the same way, the logic
for building each PMD, including extracting the pmdinfo, is included in the
top level build file. This means that each individual driver class only
need
Add the buildtools folder, and more specifically the pmdinfogen binary to
the meson and ninja build. This will be needed for building the PMDs in the
driver folder later, as the pmd info output from the tool needs to be
included in those libs.
Signed-off-by: Bruce Richardson
---
buildtools/gen-p
Support building the EAL with meson and ninja. This involves a number of
different meson.build files for iterating through all the different
subdirectories in the EAL. The library itself will be compiled on build but
the header files are only copied from their initial location once "ninja
install"
Support building igb_uio using meson and ninja. For this, we still use the
kernel's kbuild system, by calling out to make, since it's safer and easier
than trying to reproduce that in meson. A list of suitable file
dependencies is given so that we have a reasonable chance of a rebuild when
necessar
To build with meson and ninja, we need some initial infrastructure in
place. The build files for meson always need to be called "meson.build",
and options get placed in meson_options.txt
This commit adds a top-level meson.build file, which sets up the global
variables for tracking drivers, librari
Following on from the two previous RFCs [1] [2], here is a cleaned up
patchset to serve as a start-point for getting all of DPDK building with
meson and ninja.
What's covered:
* Basic infrastructure for feature detection and DPDK compilation
* Building of all DPDK libraries - as either static or s
On Fri, Sep 01, 2017 at 05:25:38PM +0800, Yuanhan Liu wrote:
> On Thu, Aug 31, 2017 at 03:40:14PM +0200, Olivier Matz wrote:
> > Split use_simple_rxtx into use_simple_rx and use_simple_tx,
> > and ensure that only use_simple_tx is updated when txq flags
> > forces to use the standard Tx handler.
>
Hi Yuanhan,
On Fri, Sep 01, 2017 at 05:19:16PM +0800, Yuanhan Liu wrote:
> On Thu, Aug 31, 2017 at 03:40:13PM +0200, Olivier Matz wrote:
> > The selection of Rx/Tx handlers is done at several places,
> > group them in one function set_rxtx_funcs().
> >
> > The update of hw->use_simple_rxtx is als
Hi Pierre,
On 9/1/2017 10:36 AM, Pierre wrote:
<..>
>
>
> On 01/09/17 09:58, dev-requ...@dpdk.org wrote:
>> Send dev mailing list submissions to
>> dev@dpdk.org
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>> http://dpdk.org/ml/listinfo/dev
>> or, via email, send a m
On 08/24/2017 04:19 AM, Tiwei Bie wrote:
This patch adaptively batches the small guest memory copies.
By batching the small copies, the efficiency of executing the
memory LOAD instructions can be improved greatly, because the
memory LOAD latency can be effectively hidden by the pipeline.
We saw
Hi
This might not be a good idea. With these modifications, the functions
are not inlined any more (attribute inline), and not post-optimized
either (-f lto)
As per ABI, most of the registers must be saved on the stack before
invoking a function. This is not noticeable in isolated test/perf
Hi
I send an email just now.
The original codes have so many functions because maybe other places (right now
nowhere else) would need functions like rte_mov16.
Then they can directly use it. So I am not sure if I need modify it.
The performance haven't been tested. But it should be the same as be
On Thu, Aug 31, 2017 at 03:40:14PM +0200, Olivier Matz wrote:
> Split use_simple_rxtx into use_simple_rx and use_simple_tx,
> and ensure that only use_simple_tx is updated when txq flags
> forces to use the standard Tx handler.
I think it's a good idea to split it.
> This change is also useful fo
On Thu, Aug 31, 2017 at 03:40:13PM +0200, Olivier Matz wrote:
> The selection of Rx/Tx handlers is done at several places,
> group them in one function set_rxtx_funcs().
>
> The update of hw->use_simple_rxtx is also rationalized:
> - initialized to 1 (prefer simple path)
> - in dev configure or rx
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Xiaoyun Li
> Sent: Friday, September 1, 2017 9:57 AM
> To: Richardson, Bruce
> Cc: dev@dpdk.org; Wang, Zhihong ; Zhang, Qi Z
> ; Lu, Wenzhuo
> ; Li, Xiaoyun
> Subject: [dpdk-dev] [PATCH v2 1/3] eal/x86: run-tim
This patch dynamically selects x86 EFD functions at run-time.
This patch uses function pointer and binds it to the relative
function based on CPU flags at constructor time.
Signed-off-by: Xiaoyun Li
---
lib/librte_efd/rte_efd_x86.h | 41 ++---
1 file changed,
This patch dynamically selects functions of memcpy at run-time based
on CPU flags that current machine supports. This patch uses function
pointers which are bind to the relative functions at constrctor time.
In addition, AVX512 instructions set would be compiled only if users
config it enabled and
This patchset dynamically selects functions at run-time based on CPU flags
that current machine supports. This patchset modifies mempcy, memcpy perf
test and x86 EFD, using function pointers and bind them at constructor time.
Then in the cloud environment, users can compiler once for the minimum ta
This patch modifies assignment of alignment unit from build-time
to run-time based on CPU flags that machine supports.
Signed-off-by: Xiaoyun Li
---
test/test/test_memcpy_perf.c | 40 +++-
1 file changed, 27 insertions(+), 13 deletions(-)
diff --git a/test/te
On 8/26/2017 5:36 AM, Cristian Dumitrescu wrote:
Signed-off-by: Cristian Dumitrescu
---
Changes in v1 (from RFC [1]):
- Added EXPERIMENTAL tag in rte_mtr.h and MANTAINERS for this new API
(input from Thomas)
- Added more Doxygen comments to re-inforce relationship between MTR and
flow (input
Allocation and management of Tx/Rx queue arrays is done by wrappers at the
ethdev level. The resulting information is copied to the private structure
while configuring the device, where it is managed separately by the PMD.
This is redundant and consumes space in the private structure.
Relying mor
Considering the remaining functionality, the only difference between
isolated and non-isolated mode is that a default MAC flow rule is present
with the latter.
The restriction on enabling isolated mode before creating any queues can
therefore be lifted.
Signed-off-by: Adrien Mazarguil
---
drive
While internal static functions do not cause link time conflicts, this
differentiates them from their mlx5 PMD counterparts while debugging.
No impact on functionality.
Signed-off-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4.c | 4 ++--
drivers/net/mlx4/mlx4_flow.c | 30 +++-
1 - 100 of 153 matches
Mail list logo
