Re: [dpdk-dev] [PATCH v2 0/4] eal/common: introduce rte_memset and related test

2017-01-18 02:42, Yang, Zhiyong: > From: Thomas Monjalon [mailto:thomas.monja...@6wind.com] > > The functional autotest is not integrated in the basic test suite. > > I can run command line "memset_autotest", It seems that I leave something > out. Please check app/test/autotest_data.py

Re: [dpdk-dev] [PATCH v8 4/9] eal/bus: support for scanning of bus

2017-01-18 05:15, Shreyansh Jain: > > > + * > > > + * @return > > > + * 0 for successful scan > > > + * !0 (<0) for unsuccessful scan with error value > > > > What do you mean with !0 (<0) ? > > I guess <0 is enough. > > > > Only to highlight that only '-ve' values would be returned. I

Re: [dpdk-dev] [PATCH v8 3/9] pci: split match and probe function

2017-01-18 06:17, Shreyansh Jain: > From: Thomas Monjalon [mailto:thomas.monja...@6wind.com] > > 2017-01-17 19:07, Shreyansh Jain: > > > +static int > > > +rte_eal_pci_detach_dev(struct rte_pci_driver *dr, > > > + struct rte_pci_device *dev) > > > +{ > > > + int ret; > > > +

Re: [dpdk-dev] [PATCH v8 2/9] test: add basic bus infrastructure tests

2017-01-18 06:56, Shreyansh Jain: > > > > +/* Bus list exposed */ > > > > +extern struct rte_bus_list rte_bus_list; > > > > > > I think it should be possible to write a test without the real list > > > of registered bus. > > > > Yes, it is possible. I just modeled it on test_pci assuming that is a

[dpdk-dev] [PATCH] pdump: fix nb_rx_q may be used uninitialized warning

Fix GCC 4.8.2 20140120 (Red Hat 4.8.2-16) (RHEL 7.0) false warning when build with EXTRA_CFLAGS='--coverage'. Fixes: 278f945402c5 ("pdump: add new library for packet capture") Signed-off-by: Andrew Rybchenko --- lib/librte_pdump/rte_pdump.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)

Re: [dpdk-dev] [PATCH 0/7] net/mlx5: improve single core performance

2017-01-05 14:13, Ferruh Yigit: > On 11/24/2016 4:03 PM, Nelio Laranjeiro wrote: > > This series applies on top of > > "[PATCH] eal: define generic vector types" [1][2] > > > > Using built-in vector types forces compilers to consider SIMD instructions > > in > > specific places in order to improv

Re: [dpdk-dev] [PATCH v8 2/9] test: add basic bus infrastructure tests

> -Original Message- > From: Shreyansh Jain > Sent: Wednesday, January 18, 2017 10:42 AM > To: 'Thomas Monjalon' > Cc: dev@dpdk.org > Subject: RE: [PATCH v8 2/9] test: add basic bus infrastructure tests > > > -Original Message- > > From: Thomas Monjalon [mailto:thomas.monja...@6wi

Re: [dpdk-dev] [PATCH] app/testpmd: refine xstats show

Hi Pablo, > -Original Message- > From: De Lara Guarch, Pablo > Sent: Wednesday, January 18, 2017 12:04 AM > To: Tan, Jianfeng; dev@dpdk.org > Cc: Wu, Jingjing; Tan, Jianfeng > Subject: RE: [dpdk-dev] [PATCH] app/testpmd: refine xstats show > > Hi Jianfeng, > > > -Original Message

Re: [dpdk-dev] [PATCH v8 3/9] pci: split match and probe function

> -Original Message- > From: Thomas Monjalon [mailto:thomas.monja...@6wind.com] > Sent: Wednesday, January 18, 2017 5:02 AM > To: Shreyansh Jain > Cc: dev@dpdk.org > Subject: Re: [PATCH v8 3/9] pci: split match and probe function > > 2017-01-17 19:07, Shreyansh Jain: > > +static int > > +

Re: [dpdk-dev] [PATCH] mbuf: use pktmbuf helper to create the pool

On 1/17/2017 7:01 PM, Santosh Shukla wrote: Hi Hemant, On Wed, Jan 18, 2017 at 12:12:56AM +0530, Hemant Agrawal wrote: When possible, replace the uses of rte_mempool_create() with the helper provided in librte_mbuf: rte_pktmbuf_pool_create(). This is the preferred way to create a mbuf pool. T

Re: [dpdk-dev] [PATCH v8 4/9] eal/bus: support for scanning of bus

> -Original Message- > From: Thomas Monjalon [mailto:thomas.monja...@6wind.com] > Sent: Wednesday, January 18, 2017 5:07 AM > To: Shreyansh Jain > Cc: dev@dpdk.org > Subject: Re: [PATCH v8 4/9] eal/bus: support for scanning of bus > > 2017-01-17 19:07, Shreyansh Jain: > > + * Generic bus

Re: [dpdk-dev] [PATCH v8 2/9] test: add basic bus infrastructure tests

> -Original Message- > From: Thomas Monjalon [mailto:thomas.monja...@6wind.com] > Sent: Wednesday, January 18, 2017 4:54 AM > To: Shreyansh Jain > Cc: dev@dpdk.org > Subject: Re: [PATCH v8 2/9] test: add basic bus infrastructure tests > > 2017-01-17 19:07, Shreyansh Jain: > > +REGISTER_TE

Re: [dpdk-dev] [PATCH v8 1/9] eal/bus: introduce bus abstraction

> -Original Message- > From: Thomas Monjalon [mailto:thomas.monja...@6wind.com] > Sent: Wednesday, January 18, 2017 4:50 AM > To: Shreyansh Jain > Cc: dev@dpdk.org > Subject: Re: [PATCH v8 1/9] eal/bus: introduce bus abstraction > > 2017-01-17 19:07, Shreyansh Jain: > > +void > > +rte_bus

Re: [dpdk-dev] [PATCH 5/5] net/virtio: fix Tso when mbuf is shared

On Tue, Jan 17, 2017 at 12:18:25PM +0100, Olivier Matz wrote: > > I hope I could have time to dig this further, since, honestly, I don't > > quite like this patch: it makes things un-maintainable. > > Well, I'm not that proud of the patch, but that's the best solution > I've found. Nevertheless sa

Re: [dpdk-dev] [PATCH] devtools: check stable tag in fixes

On Tue, Jan 17, 2017 at 07:42:33PM +0100, Thomas Monjalon wrote: > 2017-01-17 18:15, Ferruh Yigit: > > On 1/17/2017 2:54 PM, Thomas Monjalon wrote: > > > The tag "Cc: sta...@dpdk.org" must be set when the commit must be > > > backported to a stable branch. > > > > > > It must be located just below

[dpdk-dev] replay MAC address in rte_eth_dev_config_restore()

Hi, I have a question on MAC address replay logic in rte_eth_dev_config_restore(): lib/librte_ether/rte_ethdev.c: code snippet of rte_eth_dev_config_restore() ~~ /* replay MAC address configuration */ for (i = 0; i < dev_info.max_mac_addrs; i++) { addr = dev->data-

Re: [dpdk-dev] net/virtio: build is broken on FreeBSD

> -Original Message- > From: Thomas Monjalon [mailto:thomas.monja...@6wind.com] > Sent: Wednesday, January 18, 2017 6:10 AM > To: Andrew Rybchenko; Yuanhan Liu; Tan, Jianfeng > Cc: dev@dpdk.org > Subject: Re: [dpdk-dev] net/virtio: build is broken on FreeBSD > > 2017-01-17 17:39, Andrew R

Re: [dpdk-dev] [PATCH v2] doc: add known uio_pci_generic issue for i40e

On 1/18/2017 1:36 AM, Ferruh Yigit wrote: On 1/17/2017 5:00 PM, Wu, Jingjing wrote: -Original Message- From: Yigit, Ferruh Sent: Tuesday, January 17, 2017 4:31 AM To: Guo, Jia ; Zhang, Helin ; Wu, Jingjing Cc: dev@dpdk.org Subject: Re: [dpdk-dev] [PATCH v2] doc: add known uio_pci_ge

Re: [dpdk-dev] [PATCH v3 1/4] vmxnet3: Avoid memory leak in vmxnet3_dev_rx_queue_setup.

> On Jan 18, 2017, at 10:05 AM, Yong Wang wrote: > > Any downside with free/reallocation now that memzone can be freed? > Allocation with max ring size should work but is kind of wasteful in terms of > memory usage and I assume this type of ring size change should not be a > frequent operat

Re: [dpdk-dev] [PATCH v2 0/4] eal/common: introduce rte_memset and related test

hi, Thomas: Thanks for your reply. > -Original Message- > From: Thomas Monjalon [mailto:thomas.monja...@6wind.com] > Sent: Wednesday, January 18, 2017 4:14 AM > To: Yang, Zhiyong > Cc: Richardson, Bruce ; Ananyev, Konstantin > ; yuanhan@linux.intel.com; De Lara > Guarch, Pablo

Re: [dpdk-dev] [PATCH] net/virtio: fix build without virtio-user

On Tue, Jan 17, 2017 at 11:22:41PM +0100, Thomas Monjalon wrote: > When CONFIG_RTE_VIRTIO_USER is disabled (default on FreeBSD), > the virtio driver cannot be compiled: > > librte_pmd_virtio.a(virtio_ethdev.o): In function `eth_virtio_dev_init': > (.text+0x1eba): undefined reference to `virtio_use

Re: [dpdk-dev] [PATCH v4 7/7] app/test: add ARMv8 crypto tests and test vectors

On Tue, Jan 17, 2017 at 04:49:03PM +0100, zbigniew.bo...@caviumnetworks.com wrote: > From: Zbigniew Bodek > > Introduce unit tests for ARMv8 crypto PMD. > Add test vectors for short cases such as 160 bytes. > These test cases are ARMv8 specific since the code provides > different processing path

Re: [dpdk-dev] [PATCH v4 1/7] lib: add cryptodev type for the upcoming ARMv8 PMD

On Tue, Jan 17, 2017 at 04:48:57PM +0100, zbigniew.bo...@caviumnetworks.com wrote: > From: Zbigniew Bodek Please change the commit subject to cryptodev: add cryptodev type for the ARMv8 PMD from lib: add cryptodev type for the upcoming ARMv8 PMD > > Add type and name for ARMv8 crypto PMD >

Re: [dpdk-dev] [PATCH v3 1/4] vmxnet3: Avoid memory leak in vmxnet3_dev_rx_queue_setup.

Any downside with free/reallocation now that memzone can be freed? Allocation with max ring size should work but is kind of wasteful in terms of memory usage and I assume this type of ring size change should not be a frequent operation. From: nickcooper-zhangtonghao [mailto:n...@opencloud.tech]

Re: [dpdk-dev] [PATCH v6 14/18] net/ixgbe: parse L2 tunnel filter

Hi, Ferruh > -Original Message- > From: Yigit, Ferruh > Sent: Tuesday, January 17, 2017 6:03 PM > To: Zhao1, Wei ; Adrien Mazarguil > > Cc: dev@dpdk.org; Lu, Wenzhuo > Subject: Re: [dpdk-dev] [PATCH v6 14/18] net/ixgbe: parse L2 tunnel filter > > On 1/17/2017 9:27 AM, Zhao1, Wei wrote:

Re: [dpdk-dev] [PATCH v3 1/4] vmxnet3: Avoid memory leak in vmxnet3_dev_rx_queue_setup.

> On Jan 18, 2017, at 4:15 AM, Yong Wang wrote: > >> -Original Message- >> From: Nick Zhang [mailto:n...@opencloud.tech ] >> Sent: Sunday, January 8, 2017 7:00 PM >> To: Yong Wang mailto:yongw...@vmware.com>> >> Cc: ferruh.yi...@intel.com

Re: [dpdk-dev] net/qede/base: build is broken on FreeBSD 11.0 with clang 3.8.0

> From: Andrew Rybchenko [mailto:arybche...@solarflare.com] > Sent: Tuesday, January 17, 2017 7:00 AM > > Hi, > > build is broken for me on FreeBSD 11.0. Just default config and gmake: > > == Build drivers/net/qede >CC base/ecore_dev.o >CC base/ecore_hw.o >CC base/ecore_cxt.o > /tmp/

[dpdk-dev] [PATCH v5 29/29] net/vmxnet3: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Yong Wang Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob --- drivers/net/vmxnet3/vmxne

[dpdk-dev] [PATCH v5 25/29] net/nfp: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Alejandro Lucero Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob Acked-by: Alejandro Luc

[dpdk-dev] [PATCH v5 26/29] net/qede: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Harish Patil CC: Rasesh Mody Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob --- drive

[dpdk-dev] [PATCH v5 28/29] net/virtio: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Huawei Xie CC: Yuanhan Liu Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob Acked-by: Yu

[dpdk-dev] [PATCH v5 27/29] net/thunderx: use eal I/O device memory read/write API

Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. Signed-off-by: Jerin Jacob --- drivers/net/thunderx/base/nicvf_plat.h | 36 -- 1 file changed

[dpdk-dev] [PATCH v5 23/29] net/i40e: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Helin Zhang CC: Jingjing Wu Signed-off-by: Santosh Shukla Signed-off-by: Satha Rao Signed-off-by

[dpdk-dev] [PATCH v5 24/29] net/ixgbe: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Helin Zhang CC: Konstantin Ananyev Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob ---

[dpdk-dev] [PATCH v5 22/29] net/fm10k: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Jing Chen Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob --- drivers/net/fm10k/base/fm

[dpdk-dev] [PATCH v5 20/29] net/enic: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: John Daley CC: Nelson Escobar Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob --- driv

[dpdk-dev] [PATCH v5 18/29] net/cxgbe: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Rahul Lakkireddy Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob --- drivers/net/cxgbe/

[dpdk-dev] [PATCH v5 19/29] net/e1000: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Wenzhuo Lu Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob --- drivers/net/e1000/base/e

[dpdk-dev] [PATCH v5 21/29] net/ena: use eal I/O device memory read/write API

Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. Suggested-by: Jan Medala Signed-off-by: Jerin Jacob Acked-by: Jan Medala --- drivers/net/ena/base/ena_plat_dpdk.h | 15 +++-

[dpdk-dev] [PATCH v5 16/29] net/bnxt: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Stephen Hurd CC: Ajit Khaparde Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob --- dri

[dpdk-dev] [PATCH v5 17/29] net/bnx2x: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: Harish Patil CC: Rasesh Mody Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob --- drive

[dpdk-dev] [PATCH v5 15/29] crypto/qat: use eal I/O device memory read/write API

From: Santosh Shukla Replace the raw I/O device memory read/write access with eal abstraction for I/O device memory read/write access to fix portability issues across different architectures. CC: John Griffin CC: Fiona Trahe CC: Deepak Kumar Jain Signed-off-by: Santosh Shukla Signed-off-by:

[dpdk-dev] [PATCH v5 10/29] eal: introduce I/O device memory read/write operations

This commit introduces 8-bit, 16-bit, 32bit, 64bit I/O device memory read/write operations along with the relaxed versions. The weakly-ordered machine like ARM needs additional I/O barrier for device memory read/write access over PCI bus. By introducing the eal abstraction for I/O device memory re

[dpdk-dev] [PATCH v5 14/29] eal/arm64: change barrier definitions to macros

Change rte_*wb definitions to macros in order to keep consistent with other barrier definitions in the file. Suggested-by: Jianbo Liu Signed-off-by: Jerin Jacob --- .../common/include/arch/arm/rte_atomic_64.h| 36 ++ 1 file changed, 3 insertions(+), 33 deletions(-)

[dpdk-dev] [PATCH v5 11/29] eal: generic implementation for I/O device read/write access

This patch implements the generic version of rte_read[b/w/l/q]_[relaxed] and rte_write[b/w/l/q]_[relaxed] using rte_io_wmb() and rte_io_rmb() Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/generic/rte_io.h | 118 + 1 file changed, 118 insertions(+) diff --g

[dpdk-dev] [PATCH v5 13/29] eal/arm64: override I/O device read/write access for arm64

Override the generic I/O device memory read/write access and implement it using armv8 instructions for arm64. Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/arm/rte_io.h| 4 + lib/librte_eal/common/include/arch/arm/rte_io_64.h | 199 + 2 files changed

[dpdk-dev] [PATCH v5 12/29] eal: let all architectures use generic I/O implementation

Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/arm/rte_io.h| 47 ++ lib/librte_eal/common/include/arch/ppc_64/rte_io.h | 47 ++ lib/librte_eal/common/include/arch/tile/rte_io.h | 47 ++ lib/librte_eal/common/incl

[dpdk-dev] [PATCH v5 09/29] eal/arm64: define I/O device memory barriers for arm64

CC: Jianbo Liu Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/arm/rte_atomic_64.h | 6 ++ 1 file changed, 6 insertions(+) diff --git a/lib/librte_eal/common/include/arch/arm/rte_atomic_64.h b/lib/librte_eal/common/include/arch/arm/rte_atomic_64.h index 78ebea2..ef0efc7 1

[dpdk-dev] [PATCH v5 08/29] eal/arm64: define smp barrier definition for arm64

dmb instruction based barrier is used for smp version of memory barrier. Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/arm/rte_atomic_64.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/lib/librte_eal/common/include/arch/arm/rte_atomic_64.h b/lib/l

[dpdk-dev] [PATCH v5 07/29] eal/arm64: fix memory barrier definition for arm64

dsb instruction based barrier is used for non smp version of memory barrier. Fixes: d708f01b7102 ("eal/arm: add atomic operations for ARMv8") CC: Jianbo Liu CC: sta...@dpdk.org Signed-off-by: Jerin Jacob Acked-by: Jianbo Liu --- lib/librte_eal/common/include/arch/arm/rte_atomic_64.h | 9 +

[dpdk-dev] [PATCH v5 06/29] eal/armv7: define I/O device memory barriers for ARMv7

The patch does not provide any functional change for ARMv7. I/O barriers are mapped to existing smp barriers. CC: Jan Viktorin CC: Jianbo Liu Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/arm/rte_atomic_32.h | 6 ++ 1 file changed, 6 insertions(+) diff --git a/lib/libr

[dpdk-dev] [PATCH v5 05/29] eal/arm: separate smp barrier definition for ARMv7 and ARMv8

Separate the smp barrier definition for arm and arm64 for fine control on smp barrier definition for each architecture. Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/arm/rte_atomic.h| 6 -- lib/librte_eal/common/include/arch/arm/rte_atomic_32.h | 6 ++ lib/librte_

[dpdk-dev] [PATCH v5 01/29] eal: introduce I/O device memory barriers

This commit introduce rte_io_mb(), rte_io_wmb() and rte_io_rmb(), in order to enable memory barriers between I/O device and CPU. Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/generic/rte_atomic.h | 27 ++ 1 file changed, 27 insertions(+) diff --git a/lib/librt

[dpdk-dev] [PATCH v5 04/29] eal/ppc64: define I/O device memory barriers for ppc64

The patch does not provide any functional change for ppc_64. I/O barriers are mapped to existing smp barriers. CC: Chao Zhu Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/ppc_64/rte_atomic.h | 6 ++ 1 file changed, 6 insertions(+) diff --git a/lib/librte_eal/common/inclu

[dpdk-dev] [PATCH v5 02/29] eal/x86: define I/O device memory barriers for IA

The patch does not provide any functional change for IA. I/O barriers are mapped to existing smp barriers. CC: Bruce Richardson CC: Konstantin Ananyev Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/x86/rte_atomic.h | 6 ++ 1 file changed, 6 insertions(+) diff --git a/li

[dpdk-dev] [PATCH v5 03/29] eal/tile: define I/O device memory barriers for tile

The patch does not provide any functional change for tile. I/O barriers are mapped to existing smp barriers. CC: Zhigang Lu Signed-off-by: Jerin Jacob --- lib/librte_eal/common/include/arch/tile/rte_atomic.h | 6 ++ 1 file changed, 6 insertions(+) diff --git a/lib/librte_eal/common/include

[dpdk-dev] [PATCH v5 00/29] introduce I/O device memory read/write operations

v5: 1) net/ena: removed intermediate readl/writel implementation and implemented ENA_REG_READ32/ENA_REG_WRITE32 as rte_read_relaxed/ rte_write32_relaxed(Jan Medala) 2) Rebased to dpdk.org master 3) Created a temporary branch in public repo to share changes for easy testing repo: https://github.com

Re: [dpdk-dev] [PATCH v12 00/26] Support VFD on i40e - EXPERIMENTAL

Hi Ferruh, > -Original Message- > From: Yigit, Ferruh > Sent: Tuesday, January 17, 2017 9:10 PM > To: Lu, Wenzhuo; dev@dpdk.org > Subject: Re: [dpdk-dev] [PATCH v12 00/26] Support VFD on i40e - > EXPERIMENTAL > > On 1/17/2017 8:45 AM, Wenzhuo Lu wrote: > <...> > > Bernard Iremonger (7): >

[dpdk-dev] [PATCH] net/mlx5: fix queue wrap-around of multi-packet send

When the WQ is wrapped around, it wrongly checks the condition when resetting the pointer. It should be compared against the end of the queue, not the beginning of the queue. And this isn't even needed when the length of the copying data crosses the boundary. Fixes: e5291c280c08 ("net/mlx5: use wo

[dpdk-dev] [PATCH v2] net/mlx5: lift Rx RSS indirection table size limit

The size of Rx RSS indirection table was limited by 256, but it is not required anymore for all Mellanox NICs. However, the librte_ether still limits the size by 512. Signed-off-by: Yongseok Koh --- drivers/net/mlx5/mlx5.c | 5 +++-- drivers/net/mlx5/mlx5_defs.h | 3 --- 2 files changed, 3

Re: [dpdk-dev] [PATCH v2 0/4] eal/common: introduce rte_memset and related test

Le 17/01/2017 à 21:14, Thomas Monjalon a écrit : By the way, have you tried to work on glibc, as I had suggested? Nothing here: https://sourceware.org/cgi-bin/search.cgi?wm=wrd&form=extended&m=all&s=D&ul=%2Fml%2Flibc-alpha%2F%25&q=memset

Re: [dpdk-dev] [PATCH] config: build nfp pmd support by default

On 1/17/2017 11:20 PM, Ferruh Yigit wrote: > On 1/13/2017 11:50 AM, Alejandro Lucero wrote: >> Because using a NFP PMD requires specific BSP installed, the PMD >> support was not the default option before. This was just for making >> people aware of such dependency, since there is no need for such

Re: [dpdk-dev] [PATCH v8 8/9] eal: enable PCI bus

2017-01-17 19:07, Shreyansh Jain: > - /* Probe & Initialize PCI devices */ > - if (rte_eal_pci_probe()) > - rte_panic("Cannot probe PCI\n"); > - > /* Probe all the buses and devices/drivers on them */ > if (rte_bus_probe()) > rte_panic("Cann

Re: [dpdk-dev] [PATCH v8 7/9] test: update bus and pci unit test cases

2017-01-17 19:07, Shreyansh Jain: > Signed-off-by: Shreyansh Jain > Reviewed-by: Ferruh Yigit > --- > app/test/test_bus.c | 152 > app/test/test_pci.c | 164 > > 2 files changed, 266 insertions

Re: [dpdk-dev] [PATCH v8 6/9] eal: integrate bus scan and probe with EAL

2017-01-17 19:07, Shreyansh Jain: > Each bus implementation defines their own callbacks for scanning its bus > and probing devices available on the bus. Enable EAL to call bus specific > scan and probe functions during DPDK initialization. > > Existing PCI scan/probe continues to exist. It will re

Re: [dpdk-dev] [PATCH v8 5/9] eal/bus: introduce support for bus probing

2017-01-17 19:07, Shreyansh Jain: > + * This is called while iterating over each registered bus. Bus object is > + * passed along assuming this is wrapped around (embedded) by Implementation > + * specific bus object. > + * > + * @param bus > + * Generic bus object which was registered with EAL

Re: [dpdk-dev] [PATCH v8 4/9] eal/bus: support for scanning of bus

2017-01-17 19:07, Shreyansh Jain: > + * Generic bus object passed only as a helper for implementation to find > + * their respective registered bus object. Implementations can choose not > + * to use this variable. I think this comment is obsolete. > + * > + * @return > + * 0 for successful s

Re: [dpdk-dev] [PATCH v8 3/9] pci: split match and probe function

2017-01-17 19:07, Shreyansh Jain: > +static int > +rte_eal_pci_detach_dev(struct rte_pci_driver *dr, > + struct rte_pci_device *dev) > +{ > + int ret; > + struct rte_pci_addr *loc; > + > + if ((dr == NULL) || (dev == NULL)) > + return -EINVAL; > + > +

[dpdk-dev] [PATCH v8 6/7] lib: added new library for latency stats

From: Harry van Haaren Signed-off-by: Reshma Pattan Signed-off-by: Remy Horton Signed-off-by: Harry van Haaren --- v7 to v8 changes: - CYCLES_PER_NS now a static inline function - "hidden" pthread creation now has polling API instead. The application must call rte_latencystats_update() to r

[dpdk-dev] [PATCH v8 7/7] app/test-pmd: add latency statistics calculation

From: Harry van Haaren This patch adds latency stats commandline argument to testpmd, allowing to specify the lcore to use for latencystats updates. Signed-off-by: Reshma Pattan Signed-off-by: Harry van Haaren Signed-off-by: Remy Horton --- v7 -> v8 changes: - add optarg to enable latencysta

[dpdk-dev] [PATCH v8 5/7] mbuf: add a timestamp to the mbuf for latencystats

From: Harry van Haaren This commit adds a uint64_t to the mbuf struct, allowing collection of latency and jitter statistics by measuring packet I/O timestamps. This change is required by the latencystats library. Signed-off-by: Reshma Pattan Signed-off-by: Harry van Haaren --- lib/librte_mbuf

[dpdk-dev] [PATCH v8 3/7] lib: add bitrate statistics library

This patch adds a library that calculates peak and average data-rate statistics. For ethernet devices. These statistics are reported using the metrics library. Signed-off-by: Remy Horton --- MAINTAINERS| 4 + config/common_base

[dpdk-dev] [PATCH v8 4/7] app/test-pmd: add bitrate statistics calculation

Calculate bitrate statistics using the bitrate stats library. The resulting statistics can be viewed via proc_info. Signed-off-by: Remy Horton --- app/test-pmd/testpmd.c | 38 +- 1 file changed, 37 insertions(+), 1 deletion(-) diff --git a/app/test-pmd/testpm

[dpdk-dev] [PATCH v8 1/7] lib: add information metrics library

This patch adds a new information metric library that allows other modules to register named metrics and update their values. It is intended to be independent of ethdev, rather than mixing ethdev and non-ethdev information in xstats. Signed-off-by: Remy Horton --- MAINTAINERS

Re: [dpdk-dev] [PATCH v8 2/9] test: add basic bus infrastructure tests

2017-01-17 19:07, Shreyansh Jain: > +REGISTER_TEST_COMMAND(bus_autotest, test_bus); You should add it to app/test/autotest_data.py > +/* Bus list exposed */ > +extern struct rte_bus_list rte_bus_list; I think it should be possible to write a test without the real list of registered bus.

[dpdk-dev] [PATCH v8 2/7] app/proc_info: add metrics displaying

From: Reshma Pattan Modify the dpdk-procinfo process to display the newly added metrics. Added new command line option "--metrics" to display metrics. Signed-off-by: Reshma Pattan Signed-off-by: Remy Horton --- app/proc_info/main.c | 74 +++- 1

[dpdk-dev] [PATCH v8 0/7] Expanded statistics reporting

This patchset extends statistics reporting to include peak and average data-rate metrics. It comes in two parts: a statistics reporting library, and a bitrate calculation library that uses it. This structure is intended to seperate statistic reporting from ethdev and allow more flexible metric regi

Re: [dpdk-dev] [PATCH] config: build nfp pmd support by default

On 1/13/2017 11:50 AM, Alejandro Lucero wrote: > Because using a NFP PMD requires specific BSP installed, the PMD > support was not the default option before. This was just for making > people aware of such dependency, since there is no need for such a > BSP for just compiling DPDK with NFP PMD sup

Re: [dpdk-dev] [PATCH v8 1/9] eal/bus: introduce bus abstraction

2017-01-17 19:07, Shreyansh Jain: > +void > +rte_bus_register(struct rte_bus *bus) > +{ > + RTE_VERIFY(bus); > + RTE_VERIFY(bus->name && strlen(bus->name)); > + > + TAILQ_INSERT_TAIL(&rte_bus_list, bus, next); > + RTE_LOG(INFO, EAL, "Registered [%s] bus.\n", bus->name); I think it

Re: [dpdk-dev] [PATCH v6 5/8] eal: introduce bus scan and probe in EAL

2017-01-17 10:33, Shreyansh Jain: > On Tuesday 17 January 2017 01:28 AM, Ferruh Yigit wrote: > > On 1/16/2017 3:38 PM, Shreyansh Jain wrote: > >> +/* Add a PCI device to PCI Bus */ > >> +void > >> +rte_eal_pci_add_device(struct rte_pci_bus *pci_bus, > >> + struct rte_pci_device *p

Re: [dpdk-dev] [PATCH] mk: fix build of assembly files for ARM64

2017-01-17 15:35, zbigniew.bo...@caviumnetworks.com: > -# for now, we don't use as but nasm. > -# AS = $(CROSS)as It looks to be a very very old comment. > +ifeq ($(CONFIG_RTE_ARCH_X86),y) > AS= nasm There is no asm file in DPDK as far as I know. So why would we have a dependency o

Re: [dpdk-dev] [PATCH] net/bnxt: make eth dev ops struct const

On Tue, 17 Jan 2017 21:57:33 +0100 Thomas Monjalon wrote: > 2017-01-16 13:44, Ferruh Yigit: > > On 1/16/2017 1:27 PM, Ferruh Yigit wrote: > > > Reported-by: Stephen Hemminger > > > Signed-off-by: Ferruh Yigit > > > > Both applied to dpdk-next-net/master, thanks. > > I must admit it is t

[dpdk-dev] [PATCH v8 4/6] examples/flow_distributor: sample app to demonstrate EFD usage

This new sample app, based on the client/server sample app, shows the user an scenario using the EFD library. It consists of: - A front-end server which has an EFD table that stores the node id for each flow key, which will distribute the incoming packets to the different nodes - A back-end n

[dpdk-dev] [PATCH v8 6/6] doc: add flow distributor guide

Signed-off-by: Sameh Gobriel Signed-off-by: Pablo de Lara Acked-by: Christian Maciocco --- MAINTAINERS |1 + doc/guides/sample_app_ug/flow_distributor.rst | 494 doc/guides/sample_app_ug/img/flow_distributor.svg | 1254

[dpdk-dev] [PATCH] net/virtio: fix build without virtio-user

When CONFIG_RTE_VIRTIO_USER is disabled (default on FreeBSD), the virtio driver cannot be compiled: librte_pmd_virtio.a(virtio_ethdev.o): In function `eth_virtio_dev_init': (.text+0x1eba): undefined reference to `virtio_user_ops' Signed-off-by: Thomas Monjalon --- Already applied on mainline git

[dpdk-dev] [PATCH v8 1/6] efd: new Elastic Flow Distributor library

Elastic Flow Distributor (EFD) is a distributor library that uses perfect hashing to determine a target/value for a given incoming flow key. It has the following advantages: - First, because it uses perfect hashing, it does not store the key itself and hence lookup performance is not dependent

[dpdk-dev] [PATCH v8 3/6] app/test: add EFD functional and perf tests

Signed-off-by: Byron Marohn Signed-off-by: Karla Saur Signed-off-by: Saikrishna Edupuganti Signed-off-by: Pablo de Lara Acked-by: Christian Maciocco --- MAINTAINERS | 1 + app/test/Makefile| 5 +- app/test/test_efd.c | 494

[dpdk-dev] [PATCH v8 2/6] efd: add AVX2 vect lookup function

Signed-off-by: Byron Marohn Signed-off-by: Pablo de Lara Signed-off-by: Saikrishna Edupuganti Acked-by: Christian Maciocco --- lib/librte_efd/rte_efd.c | 22 +++- lib/librte_efd/rte_efd_x86.h | 86 2 files changed, 107 insertions(+), 1 d

[dpdk-dev] [PATCH v8 0/6] Elastic Flow Distributor

EFD is a distributor library that uses perfect hashing to determine a target/value for a given incoming flow key. It has the following advantages: first, because it uses perfect hashing it does not store the key itself and hence lookup performance is not dependent on the key size. Second, the targe

Re: [dpdk-dev] [PATCH v7 0/6] Elastic Flow Distributor

> -Original Message- > From: De Lara Guarch, Pablo > Sent: Tuesday, January 17, 2017 10:11 PM > To: dev@dpdk.org > Cc: De Lara Guarch, Pablo > Subject: [PATCH v7 0/6] Elastic Flow Distributor > NACK, added .rej file by mistake.

Re: [dpdk-dev] net/virtio: build is broken on FreeBSD

2017-01-17 17:39, Andrew Rybchenko: > dpdk-next-net build is broken for me on FreeBSD 10.3 and 11.0 regardless > clang/gcc: Thanks for reporting. It is due to the pull of next-virtio tree. It is broken also on Linux when virtio-user is disabled. Unfortunately the compilation test in CI failed t

[dpdk-dev] [PATCH v7 6/6] doc: add flow distributor guide

Signed-off-by: Sameh Gobriel Signed-off-by: Pablo de Lara Acked-by: Christian Maciocco --- MAINTAINERS |1 + doc/guides/sample_app_ug/flow_distributor.rst | 494 doc/guides/sample_app_ug/img/flow_distributor.svg | 1254

[dpdk-dev] [PATCH v7 4/6] examples/flow_distributor: sample app to demonstrate EFD usage

This new sample app, based on the client/server sample app, shows the user an scenario using the EFD library. It consists of: - A front-end server which has an EFD table that stores the node id for each flow key, which will distribute the incoming packets to the different nodes - A back-end n

[dpdk-dev] [PATCH v7 3/6] app/test: add EFD functional and perf tests

Signed-off-by: Byron Marohn Signed-off-by: Karla Saur Signed-off-by: Saikrishna Edupuganti Signed-off-by: Pablo de Lara Acked-by: Christian Maciocco --- MAINTAINERS | 1 + app/test/Makefile| 5 +- app/test/test_efd.c | 494

[dpdk-dev] [PATCH v7 1/6] efd: new Elastic Flow Distributor library

Elastic Flow Distributor (EFD) is a distributor library that uses perfect hashing to determine a target/value for a given incoming flow key. It has the following advantages: - First, because it uses perfect hashing, it does not store the key itself and hence lookup performance is not dependent

[dpdk-dev] [PATCH v7 2/6] efd: add AVX2 vect lookup function

Signed-off-by: Byron Marohn Signed-off-by: Pablo de Lara Signed-off-by: Saikrishna Edupuganti Acked-by: Christian Maciocco --- lib/librte_efd/rte_efd.c | 22 +++- lib/librte_efd/rte_efd_x86.h | 86 2 files changed, 107 insertions(+), 1 d

[dpdk-dev] [PATCH v7 0/6] Elastic Flow Distributor

EFD is a distributor library that uses perfect hashing to determine a target/value for a given incoming flow key. It has the following advantages: first, because it uses perfect hashing it does not store the key itself and hence lookup performance is not dependent on the key size. Second, the targe

Re: [dpdk-dev] net/qede/base: build is broken on FreeBSD 11.0 with clang 3.8.0

2017-01-17 23:04, Thomas Monjalon: > 2017-01-17 17:59, Andrew Rybchenko: > > build is broken for me on FreeBSD 11.0. Just default config and gmake: > > Thanks for reporting. > It is due to the pull of next-virtio tree. > It is broken also on Linux when virtio-user is disabled. Sorry, wrong thread

Re: [dpdk-dev] net/qede/base: build is broken on FreeBSD 11.0 with clang 3.8.0

2017-01-17 17:59, Andrew Rybchenko: > build is broken for me on FreeBSD 11.0. Just default config and gmake: Thanks for reporting. It is due to the pull of next-virtio tree. It is broken also on Linux when virtio-user is disabled.

Re: [dpdk-dev] [PATCH v10 0/5] new API 'rte_eth_dev_fw_version_get'

> Qiming Yang (5): > ethdev: add firmware version get > net/e1000: add firmware version get > net/ixgbe: add firmware version get > net/i40e: add firmware version get > ethtool: display firmware version Applied, thanks

  1   2   3   4   >