> -Original Message-
> From: Maxime Coquelin
> Sent: Tuesday, January 5, 2021 11:35 PM
> To: dev@dpdk.org; Xia, Chenbo ; amore...@redhat.com
> Cc: Maxime Coquelin
> Subject: [PATCH v2] net/virtio: improve logs in Vhost-vDPA DMA mapping
>
> This patch adds debug logs in vhost_vdpa_dma_map
>On 1/5/21 3:27 PM, Maxime Coquelin wrote:
>>
>>
>> On 1/5/21 3:16 PM, Maxime Coquelin wrote:
>>>
>>>
>>> On 11/17/20 11:06 AM, Joyce Kong wrote:
Optimize packed ring Rx batch path with NEON instructions.
Signed-off-by: Joyce Kong
Reviewed-by: Ruifeng Wang
---
drive
From: Alvin Zhang
RSS look-up table initialization is done incorrectly due to divide-by-zero
error.
Add a check to rx-queue count.
Fixes: 50370662b727 ("net/ice: support device and queue ops")
Cc: sta...@dpdk.org
Signed-off-by: Alvin Zhang
---
drivers/net/ice/ice_ethdev.c | 10 ++
1 f
Hi Maxime,
> -Original Message-
> From: Maxime Coquelin
> Sent: Tuesday, December 22, 2020 9:57 PM
> To: dev@dpdk.org; Xia, Chenbo ; amore...@redhat.com
> Cc: Maxime Coquelin
> Subject: [PATCH 3/3] vhost: optimize vhost virtqueue struct
>
> This patch moves vhost_virtuqueue struct field
ICE choices vector TX path or basic TX path by macro
'ICE_NO_VECTOR_FLAGS'.
This patch adds 'DEV_TX_OFFLOAD_IPV4_CKSUM' in 'ICE_NO_VECTOR_FLAGS'
to make IPv4 checksum offload processed by basic TX path.
Fixes: a22483208800 ("net/ice: disable TSO offload in vector path")
Signed-off-by: Murphy Yan
Hi Maxime,
> -Original Message-
> From: Maxime Coquelin
> Sent: Tuesday, December 22, 2020 9:57 PM
> To: dev@dpdk.org; Xia, Chenbo ; amore...@redhat.com
> Cc: Maxime Coquelin
> Subject: [PATCH 2/3] vhost: move dirty logging cache out of the virtqueue
>
> This patch moves the per-virtque
Hi, Pavan
> -邮件原件-
> 发件人: Pavan Nikhilesh Bhagavatula
> 发送时间: 2021年1月5日 17:29
> 收件人: Feifei Wang ; jer...@marvell.com; Harry
> van Haaren
> 抄送: dev@dpdk.org; nd ; Honnappa Nagarahalli
> ; sta...@dpdk.org; Ruifeng Wang
> ; nd
> 主题: RE: [RFC PATCH v1 4/6] app/eventdev: add relea
Implement a generic copy flow API to allow copying of an arbitrary
header field (as well as mark, metadata or tag) to another item.
This generic copy mechanism removes the necessity to implement a
separate RTE Flow action every time we need to modify a new packet
field in the future. A user-provid
Acked-by: Qiming Yang
> -Original Message-
> From: Zhang, Qi Z
> Sent: 2021年1月8日 12:35
> To: Yang, Qiming
> Cc: Wang, Haiyue ; Guo, Jia ;
> dev@dpdk.org; Yigit, Ferruh ; Zhang, Qi Z
>
> Subject: [PATCH 0/9] ice base update batch 2
>
> Summary:
>
> 1. Add GTPU inner support for FDIR.
From: Alvin Zhang
1. Delete original code.
2. Add 2 tables(One maps flow pattern and RSS type to PCTYPE,
another maps RSS type to input set).
3. Parse RSS pattern and RSS type to get PCTYPE.
4. Parse RSS action to get queues, RSS function and hash field.
5. Create and destroy RSS filters.
6. C
From: Alvin Zhang
The api should return the system error status, but it returned the
hardware error status, this is confused for the caller.
This patch adds check on hardware execution status and returns -EIO
in case of hardware execution failure.
Signed-off-by: Alvin Zhang
Fixes: 1d4b2b4966bb
From: Alvin Zhang
The command here does not create a queue region, but only sets the
lookup table, so the descriptions in the doc is not exact.
Signed-off-by: Alvin Zhang
Fixes: feaae285b342 ("net/i40e: support hash configuration in RSS flow")
Cc: sta...@dpdk.org
---
V2: Divide the patch into
From: Kalesh AP
1. On the new 58818 chips, the RX completion is largely the same except
for the new completion opcode and the stripped VLAN format and checksum
status. Added bnxt_parse_csum_v2(), bnxt_parse_pkt_type_v2() and
bnxt_rx_vlan_v2() to support the new RX completion logic.
2. Di
From: Kalesh AP
Newer devices like SR2 may have chip backing store and do not require
host backed memory allocation.
In these cases, HWRM_FUNC_BACKING_STORE_QCAPS will return a zero entry
size to indicate contexts for which the host should not allocate backing
store.
Selectively allocate contex
From: Kalesh AP
Add the new chip specific TPA v2 logic to bnxt_tpa_start() to fully support
TPA on the new chip.
Signed-off-by: Kalesh AP
Reviewed-by: Ajit Khaparde
---
drivers/net/bnxt/bnxt_rxr.c | 72 +++--
drivers/net/bnxt/bnxt_rxr.h | 12 ++-
2 files ch
From: Kalesh AP
Modify VNIC accounting when enabling RFS on newer chips.
Unlike legacy chips, newer chips don't need additional VNIC resources
for ntuple filter. Fix the code accordingly so that we don't reserve
and allocate additional VNICs on newer chips.
Signed-off-by: Kalesh AP
Reviewed-by:
From: Kalesh AP
The 58818 chips support two different checksum modes.
Host driver has to register with FW which checksum mode it
prefers to use. DPDK driver want to use "cs_all_ok_mode=1".
FW advertises the support of the different checksum modes
on per VNIC basis in the HWRM_VNIC_QCAPS response.
From: Kalesh AP
The new chip (Stingray 2) is part of the P5 chip family with a number
of changes:
1. Implement the epoch doorbell bit for 58818 chip. With the new doorbell
infrastructure and the unbounded index logic, now set the epoch
doorbell bit to support proper doorbell operation on t
Add support for 58818/Stingray2 chip.
The new chip Stingray 2 belongs to the P5 generation.
Some changes in the PMD are required to support this new chip.
This patchset adds the necessary changes.
Please apply.
v1->v2: rebased to latest dpdk-next-net-brcm tree.
v2->v3: rebased to latest code base.
> -Original Message-
> From: Maxime Coquelin
> Sent: Tuesday, December 22, 2020 9:57 PM
> To: dev@dpdk.org; Xia, Chenbo ; amore...@redhat.com
> Cc: Maxime Coquelin
> Subject: [PATCH 1/3] vhost: remove unused Vhost virtqueue field
>
> This patch removes the "backend" field of the
> vhost_
Updated the Copyright for 2021
Updated FreeBSD ice driver version.
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/README | 4 ++--
drivers/net/ice/base/ice_acl.c | 2 +-
drivers/net/ice/base/ice_acl.h | 2 +-
drivers/net/ice/base/ice_acl_ctrl.c | 2 +-
dri
The number of nodes added counter was updated incorrectly. This issue
was exposed when the driver tried to add more than 128 queues per TC.
Fix added to update the counter correctly.
Fixes: 93e84b1bfc92 ("net/ice/base: add basic Tx scheduler")
Cc: sta...@dpdk.org
Signed-off-by: Victor Raj
Signe
A few style issues reported by checkpatch have snuck into the code,
resolve the style issues.
PARENTHESIS_ALIGNMENT: Alignment should match open parenthesis
COMPLEX_MACRO: Macros with complex values should be enclosed in parentheses
Signed-off-by: Bruce Allan
Signed-off-by: Qi Zhang
---
driver
Add dummy packets for IPV4_GTPU with inner IPV4/UDP/TCP with all
kinds of GTPU (EH) type (i.e., IP/EH/DL/UL) for AVF FDIR.
Signed-off-by: Junfeng Guo
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_fdir.c | 355 +---
drivers/net/ice/base/ice_type.h | 18 ++
2 f
Beyond a specific version of firmware, there is no need to provide
override values to the firmware when setting PHY capabilities. In this
case, we do not need to indicate whether we're in Strict or Lenient Link
Mode.
In the case of translating capabilities to the configuration structure,
the modu
Fixed memory handling when memory allocated in user space was handled
as memory allocated in kernel space within QV os_dep implementation
of the ice_memdup function.
Fixes: 93e84b1bfc92 ("net/ice/base: add basic Tx scheduler")
Cc: sta...@dpdk.org
Signed-off-by: Andrii Pypchenko
Signed-off-by: Qi
Scan the 'Marker PType TCAM' session to retrieve the Rx parser PTYPE
enable information from the current package.
Signed-off-by: Haiyue Wang
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_flex_pipe.c | 79
drivers/net/ice/base/ice_flex_pipe.h | 3 ++
drivers/
hw_vsi_id is used to replace vsi_id, so remove the deprecated vsi_id.
Signed-off-by: Haiyue Wang
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_switch.h | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/net/ice/base/ice_switch.h
b/drivers/net/ice/base/ice_switch.h
index be9b74fd
Aligned the buffer the following admin commands to their new
definitions:
* 0x210 = add_vsi
* 0x211 = update_vsi
Signed-off-by: Amir Shay
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_adminq_cmd.h | 209 +-
drivers/net/ice/ice_ethdev.c | 88 +--
Summary:
1. Add GTPU inner support for FDIR.
2. refine the VSI AQ command.
3. Add PTYPE metadata learning.
4. couple bug fix and code clean
5. update the Copyright date
Qi Zhang (9):
net/ice/base: align add VSI and update VSI AQ command buffer
net/ice/base: remove the deprecated field
net/i
> -Original Message-
> From: Maxime Coquelin
> Sent: Tuesday, January 5, 2021 8:57 PM
> To: dev@dpdk.org; Xia, Chenbo ; Ding, Xuan
>
> Cc: Maxime Coquelin
> Subject: [PATCH v2 1/3] vhost: refactor postcopy region registration
>
> This patch moves the registration of memory regions to
>
Hi Maxime,
> -Original Message-
> From: Maxime Coquelin
> Sent: Tuesday, December 22, 2020 12:15 AM
> To: dev@dpdk.org; Xia, Chenbo ; amore...@redhat.com;
> david.march...@redhat.com; olivier.m...@6wind.com
> Cc: Maxime Coquelin
> Subject: [PATCH 0/3] net/virtio: make virtqueue struct ca
Remove platform restriction when building drivers.
Signed-off-by: Dmitry Kozlyuk
---
drivers/meson.build | 26 --
1 file changed, 12 insertions(+), 14 deletions(-)
diff --git a/drivers/meson.build b/drivers/meson.build
index f49d4f79b..1dfa8738f 100644
--- a/drivers/meso
clang archiver tool is llvm-ar on Windows and ar on other platforms.
MinGW always uses ar. Replace shell script (Unix-only) that calls ar
with a Python script (OS-independent) that calls an appropriate archiver
tool selected at configuration time. Move the logic not to generate
empty sources into p
Process any number of input object files and write a unified output.
Signed-off-by: Dmitry Kozlyuk
---
buildtools/pmdinfogen.py | 13 +
1 file changed, 9 insertions(+), 4 deletions(-)
diff --git a/buildtools/pmdinfogen.py b/buildtools/pmdinfogen.py
index 474168f21..965c08945 100755
Common Object File Format (COFF) is used on Windows in place of ELF.
Add COFF parser to pmdinfogen. Also add an argument to specify input
file format, which is selected at configure time based on the target.
Signed-off-by: Dmitry Kozlyuk
---
Depends-on: series-13153 ("pmdinfogen: rewrite in Pyth
Based on Python version of pmdinfogen, deferred until 21.02:
http://patchwork.dpdk.org/project/dpdk/list/?series=13153
There are a few Python libraries for PE/COFF, none of which suits the need,
so a custom COFF parser is used. Advice is welcome, options considered:
* py-coff (https://githu
Updated the Copyright for 2021.
Updated FreeBSD IAVF driver of version.
Signed-off-by: Qi Zhang
---
drivers/common/iavf/README| 4 ++--
drivers/common/iavf/iavf_adminq.c | 2 +-
drivers/common/iavf/iavf_adminq.h | 2 +-
drivers/common/iavf/iavf_adminq_cmd.h | 2 +-
drivers/co
Add new opcode VIRTCHNL_OP_DCF_VLAN_OFFLOAD to set VLAN offload
by DCF, the virtchnl message includes:
1. A valid target VF
2. Type of VLAN to be supported: outer or inner
3. Ethertype of the VLAN (either 0x8100 or 0x88A8 or 0x9100)
4. VLAN insert settings
a). No insert offload, VLAN ID in the p
Currently VIRTCHNL only allows for VLAN filtering and offloads to happen
on a single 802.1Q VLAN. Add support to filter and offload on inner,
outer, and/or inner + outer VLANs.
This is done by introducing the new capability
VIRTCHNL_VF_OFFLOAD_VLAN_V2. The flow to negotiate this new capability
is
Qi Zhang (3):
common/iavf: add support for new VLAN capabilities
common/iavf: add opcode to set VLAN offload by DCF
common/iavf: update Copyright date
drivers/common/iavf/README| 4 +-
drivers/common/iavf/iavf_adminq.c | 2 +-
drivers/common/iavf/iavf_adminq.h | 2
> -Original Message-
> From: Thomas Monjalon
> Sent: Friday, January 8, 2021 12:59 AM
> To: Guo, Jia ; Zhang, Qi Z
> Cc: Wu, Jingjing ; Yang, Qiming
> ; Wang, Haiyue ;
> dev@dpdk.org; Yigit, Ferruh ;
> andrew.rybche...@oktetlabs.ru; or...@nvidia.com; getel...@nvidia.com;
> Dodji Sekete
03/01/2021 12:01, g...@marvell.com:
> From: Guy Kaneti
>
> Sentinel was missing from pci_id_ree_table[] array initialization
> which caused it to overflow.
>
> Bugzilla ID: 603
> Fixes: 4cd1c5fd9 ("regex/octeontx2: introduce REE driver")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Guy Kaneti
App
07/01/2021 21:14, Alexander Kozyrev:
> > 07/01/2021 17:57, Alexander Kozyrev:
> > > > 07/01/2021 16:22, Alexander Kozyrev:
> > > > > > 07/01/2021 16:10, Alexander Kozyrev:
> > > > > > > > > > Thursday, January 7, 2021 10:18, Thomas Monjalon
> > > > > >
> > > > > > > > > > > RTE Flows API lacks the
> 07/01/2021 17:57, Alexander Kozyrev:
> > > 07/01/2021 16:22, Alexander Kozyrev:
> > > > > 07/01/2021 16:10, Alexander Kozyrev:
> > > > > > > > > Thursday, January 7, 2021 10:18, Thomas Monjalon
> > > > >
> > > > > > > > > > RTE Flows API lacks the ability to save an arbitrary header
> > > > > >
https://bugs.dpdk.org/show_bug.cgi?id=609
Bug ID: 609
Summary: Nic stops receiving traffic and rx missed counter
starts increasing on Intel igb I350 Nics
Product: DPDK
Version: 18.11
Hardware: x86
OS: Linux
On 1/6/21 7:43 AM, Matan Azrad wrote:
> When the vDPA device is closed, the driver polling thread is canceled.
> The polling thread locks the configuration mutex while it polls the CQs.
>
> When the cancellation happens, it may terminate the thread inside the
> critical section what remains the
On 1/7/2021 6:04 PM, Ferruh Yigit wrote:
Meeting minutes of 7 January 2021
-
<...>
* rte_flow action change RFC is out, it is a design change, please review
* http://patchwork.dpdk.org/patch/85384/
Hi Andrew, Jerin, Ajit, Qi,
Can you please help on r
Meeting minutes of 7 January 2021
-
Agenda:
* Release Dates
* Subtrees
* LTS
* Opens
Participants:
* Arm
* Broadcom
* Debian/Microsoft
* Intel
* Marvell
* Nvidia
* NXP
* Red Hat
Release Dates
-
* Happy new year!
* v21.02 dates
* Proposal/V1 passe
On 1/6/21 4:06 AM, Xueming Li wrote:
> The next parameters control the HW queue moderation feature.
> This feature helps to control the traffic performance and latency
> tradeoff.
>
> Each packet completion report from HW to SW requires CQ processing by SW
> and triggers interrupt for the guest
Hi David,
On Thu, Jan 7, 2021 at 1:02 AM David Marchand
wrote:
> On Thu, Jan 7, 2021 at 2:42 AM Ashish Sadanandan
> wrote:
> >
> > Hi Olivier,
> >
> > On Wed, Jan 6, 2021 at 6:21 AM Olivier Matz
> wrote:
> > >
> > > Hi Ashish,
> > >
> > > Yes, it should reference the patch that introduced the
On 1/6/21 4:06 AM, Xueming Li wrote:
> This patch introduces new parameters for VirtQ CQ moderation, used for
> performance tuning.
>
> Signed-off-by: Xueming Li
> ---
> drivers/common/mlx5/mlx5_devx_cmds.c | 3 +++
> drivers/common/mlx5/mlx5_devx_cmds.h | 3 +++
> drivers/common/mlx5/mlx5_pr
On 12/21/20 4:50 PM, Joyce Kong wrote:
> Simply replace the smp barriers with atomic thread fence for vhost control
> path, if there are no synchronization points.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> lib/librte_vhost/vhost.c | 18 +-
> lib/l
The header was missing the extern "C" directive which causes name
mangling of functions by C++ compilers, leading to linker errors
complaining of undefined references to these functions.
Fixes: 4958ca3a443a ("mbuf: support dynamic fields and flags")
Cc: olivier.m...@6wind.com
Cc: sta...@dpdk.org
On 12/21/20 4:50 PM, Joyce Kong wrote:
> Simply relace smp barriers with atomic thread fence for
> virtio packed vring.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> lib/librte_vhost/virtio_net.c | 9 +
> 1 file changed, 5 insertions(+), 4 deletions(-)
>
Reviewe
07/01/2021 17:57, Alexander Kozyrev:
> > 07/01/2021 16:22, Alexander Kozyrev:
> > > > 07/01/2021 16:10, Alexander Kozyrev:
> > > > > > > > Thursday, January 7, 2021 10:18, Thomas Monjalon
> > > >
> > > > > > > > > RTE Flows API lacks the ability to save an arbitrary header
> > > > > > > > > field
07/01/2021 16:24, Zhang, Qi Z:
> From: Thomas Monjalon
> > 07/01/2021 13:47, Zhang, Qi Z:
> > > From: Thomas Monjalon
> > > > 07/01/2021 10:32, Guo, Jia:
> > > > > From: Thomas Monjalon
> > > > > > 24/12/2020 07:59, Jeff Guo:
> > > > > > > --- a/lib/librte_ethdev/rte_ethdev.h
> > > > > > > +++ b
> 07/01/2021 16:22, Alexander Kozyrev:
> > > 07/01/2021 16:10, Alexander Kozyrev:
> > > > > > > Thursday, January 7, 2021 10:18, Thomas Monjalon
> > >
> > > > > > > > RTE Flows API lacks the ability to save an arbitrary header
> > > > > > > > field in
> > > > > > > > order to use it later for adv
The RegEx engine as no limitation on number of queues.
This commits modifies the max supported queues reported to the application.
Fixes: fbc8c7003b93 ("regex/mlx5: add completion queue creation")
Cc: sta...@dpdk.org
Signed-off-by: Ori Kam
---
drivers/regex/mlx5/mlx5_rxp.c | 3 +--
1 file chan
The high priority match request flags means that the
RegEx engine should stop on the first match.
This commit add this flag check to the RegEx engine.
Signed-off-by: Ori Kam
---
drivers/regex/mlx5/mlx5_regex_fastpath.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drive
Due to Kernel requirement the memory allocated must be aligned to 2M.
Fixes: b34d816363b5 ("regex/mlx5: support rules import")
Cc: sta...@dpdk.org
Signed-off-by: Ori Kam
---
drivers/regex/mlx5/mlx5_rxp.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/regex/mlx5/mlx
In order to know which groups in the RegEx engine
should be used there is a need to check the req_flags.
This commit adds the missing check.
Cc: sta...@dpdk.org
Fixes: 4d4e245ad637 ("regex/mlx5: support enqueue")
Signed-off-by: Ori Kam
---
drivers/regex/mlx5/mlx5_regex_fastpath.c | 21 +++
The rof file holds programming instructions for
a given HW version.
In order to support future generation of HW it
was decided that the rof file will hold number
of rule configurations, and the driver will use
the one that matches the HW version.
In current code we force sync after each write bloc
This series adds a few fixes and improvements to the Nvidia RegEx PMD.
V2:
* Fix small issue in combined rule patch.
* Add new patch for number of queues.
Ori Kam (5):
regex/mlx5: fix memory rule alignment
regex/mlx5: add support for combined rule file
regex/mlx5: fix support for group id
07/01/2021 16:22, Alexander Kozyrev:
> > 07/01/2021 16:10, Alexander Kozyrev:
> > > > > > Thursday, January 7, 2021 10:18, Thomas Monjalon
> >
> > > > > > > RTE Flows API lacks the ability to save an arbitrary header field
> > > > > > > in
> > > > > > > order to use it later for advanced packet m
On 12/21/20 4:50 PM, Joyce Kong wrote:
> Used idx can be synchronized by one-way barrier instead of full
> write barrier for split vring.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> lib/librte_vhost/vdpa.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Rev
On 12/21/20 4:50 PM, Joyce Kong wrote:
> Relax the full read barrier to one-way barrier for desc flags in
> packed vring.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> lib/librte_vhost/virtio_net.c | 5 +++--
> 1 file changed, 3 insertions(+), 2 deletions(-)
>
Reviewed-
On 12/21/20 4:50 PM, Joyce Kong wrote:
> The ordering between avail index and desc reads has been enforced
> by load-acquire for split vring, so smp_rmb barrier is not needed
> behind it.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> lib/librte_vhost/virtio_net.c | 7 ++--
On 12/21/20 4:50 PM, Joyce Kong wrote:
> As function desc_is_avail performs a load-acquire barrier to
> enforce the ordering between desc flags and desc content, it is
> unnecessary to add a rte_smp_rmb barrier around the trace which
> follows desc_is_avail.
>
> Signed-off-by: Joyce Kong
> Rev
On 12/21/20 4:50 PM, Joyce Kong wrote:
> Simply replace the rte_smp_mb barriers with SEQ_CST atomic thread fence,
> if there is no load/store operations.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> examples/vhost_blk/vhost_blk.c | 8
> 1 file changed, 4 inserti
On 12/21/20 4:50 PM, Joyce Kong wrote:
> Use C11 atomic APIs with one-way barriers to replace two-way
> barriers when operating enqueue/dequeue. Used->idx and avail->idx
> are the synchronization points for split vring.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> exampl
On Thu, 7 Jan 2021 15:59:59 +
Ferruh Yigit wrote:
> >>
> >> int main(void)
> >> {
> >>pkt_burst_flow_gen();
> >>return 0;
> >> }
> >
> > If I change your code like this to use union, Gcc 10 is still broken.
>
> This worked fine for me: https://godbolt.org/z/vdsxh9
I was looking
On 12/21/20 3:23 PM, Joyce Kong wrote:
> Replace the smp barriers with atomic thread fence for synchronization
> between different threads, if there are no load/store operations.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> drivers/net/virtio/virtqueue.h | 8
>
On 1/7/2021 3:50 PM, Stephen Hemminger wrote:
On Wed, 6 Jan 2021 23:39:39 -0600
George Prekas wrote:
On 1/6/2021 12:02 PM, Ferruh Yigit wrote:
On 12/5/2020 5:42 AM, George Prekas wrote:
Strict-aliasing rules are violated by cast to uint16_t* in flowgen.c
and the calculated IP checksum is wro
On Wed, 6 Jan 2021 23:39:39 -0600
George Prekas wrote:
> On 1/6/2021 12:02 PM, Ferruh Yigit wrote:
> > On 12/5/2020 5:42 AM, George Prekas wrote:
> >> Strict-aliasing rules are violated by cast to uint16_t* in flowgen.c
> >> and the calculated IP checksum is wrong on GCC 9 and GCC 10.
> >>
> >>
On Thu, 7 Jan 2021 13:06:35 +0500
Sarosh Arif wrote:
> On Thu, Dec 17, 2020 at 11:19 PM Stephen Hemminger
> wrote:
> >
> > On Thu, 17 Dec 2020 16:16:16 +0500
> > Sarosh Arif wrote:
> >
> > > +if get_hugepages(path) != pages:
> > > +print("Unable to reserve required pages. The page
On 1/6/2021 8:54 PM, Lance Richardson wrote:
On Wed, Jan 6, 2021 at 4:27 AM Ferruh Yigit wrote:
On 12/14/2020 6:53 PM, Lance Richardson wrote:
Limit number of representor packets transferred per poll
to requested burst size.
Hi Lance,
Can you please describe the impact of the change? Sinc
> -Original Message-
> From: Thomas Monjalon
> Sent: Thursday, January 7, 2021 9:34 PM
> To: Guo, Jia ; Zhang, Qi Z
> Cc: Wu, Jingjing ; Yang, Qiming
> ; Wang, Haiyue ;
> dev@dpdk.org; Yigit, Ferruh ;
> andrew.rybche...@oktetlabs.ru; or...@nvidia.com; getel...@nvidia.com;
> Dodji Seket
On 1/7/2021 2:20 PM, George Prekas wrote:
On 1/7/2021 5:32 AM, Ferruh Yigit wrote:
On 1/7/2021 5:39 AM, George Prekas wrote:
On 1/6/2021 12:02 PM, Ferruh Yigit wrote:
On 12/5/2020 5:42 AM, George Prekas wrote:
Strict-aliasing rules are violated by cast to uint16_t* in flowgen.c
and the calcul
> 07/01/2021 16:10, Alexander Kozyrev:
> > > > > Thursday, January 7, 2021 10:18, Thomas Monjalon
>
> > > > > > RTE Flows API lacks the ability to save an arbitrary header field in
> > > > > > order to use it later for advanced packet manipulations. Examples
> > > > > > include the usage of VxLAN
07/01/2021 16:10, Alexander Kozyrev:
> > > > Thursday, January 7, 2021 10:07, Thomas Monjalon
> > > > > RTE Flows API lacks the ability to save an arbitrary header field in
> > > > > order to use it later for advanced packet manipulations. Examples
> > > > > include the usage of VxLAN ID after the
Hi
I would very much like a review for this patch.
Thank you,
Ori
> -Original Message-
> From: Ori Kam
> Sent: Thursday, December 17, 2020 12:38 PM
> Subject: [PATCH 1/2] regexdev: add resource limit reached rsp flag
>
> When scanning a buffer it is possible that the scan will abort
>
> > > Thursday, January 7, 2021 10:07, Thomas Monjalon
> > > > RTE Flows API lacks the ability to save an arbitrary header field in
> > > > order to use it later for advanced packet manipulations. Examples
> > > > include the usage of VxLAN ID after the packet is decapsulated or
> > > > storing th
07/01/2021 15:17, Alexander Kozyrev:
> > Tuesday, January 5, 2021 17:17, Thomas Monjalon
> > > RTE Flows API lacks the ability to save an arbitrary header field in
> > > order to use it later for advanced packet manipulations. Examples
> > > include the usage of VxLAN ID after the packet is decaps
26/11/2020 12:15, Wisam Jaddo:
> After this series the application will start supporting testing
> multiple threaded insertion and deletion rates.
>
> Also it will provide the latency & throughput rates of all threads.
>
>
> Wisam Jaddo (4):
> app/flow-perf: refactor flows handler
> app/flow
26/11/2020 12:15, Wisam Jaddo:
> The clock() function is not good practice to use for multiple
> cores/threads, since it measures the CPU time used by the process
> and not the wall clock time, while when running through multiple
> cores/threads simultaneously, we can burn through CPU time much
> f
On 12/21/20 3:23 PM, Joyce Kong wrote:
> Relax the full write barriers to one-way barriers for virtio
> control path for Arm platform
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> drivers/net/virtio/virtio_ethdev.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(
On Wed, 6 Jan 2021 22:35:53 +0200, Tal Shnaiderman wrote:
[...]
> +int
> +rte_thread_tls_key_create(rte_tls_key *key, void (*destructor)(void *))
> +{
> + int err;
> +
> + *key = malloc(sizeof(**key));
> + if ((*key) == NULL) {
> + RTE_LOG(DEBUG, EAL, "Cannot allocate TLS k
On 12/21/20 3:23 PM, Joyce Kong wrote:
> Replace rte_smp_wmb/rmb with rte_io_wmb/rmb as they are the same on x86
> and ppc platforms. Then, for function virtqueue_fetch_flags_packed/
> virtqueue_store_flags_packed, the if and else branch are still identical
> for the platforms except Arm.
>
> S
David Marchand writes:
> On Thu, Jan 7, 2021 at 2:33 PM Thomas Monjalon wrote:
>> > Yes that may break the ABI but fortunately the checking-abi-compatibility
>> > tool shows negative :) , thanks Ferruh' s guide.
>> > https://github.com/ferruhy/dpdk/actions/runs/468859673
>>
>> That's very stran
> Tuesday, January 5, 2021 17:17, Thomas Monjalon
> > RTE Flows API lacks the ability to save an arbitrary header field in
> > order to use it later for advanced packet manipulations. Examples
> > include the usage of VxLAN ID after the packet is decapsulated or
> > storing this ID inside the pack
On 12/21/20 3:23 PM, Joyce Kong wrote:
> As desc_is_used has a load-acquire or rte_io_rmb inside
> and wait for used desc in virtqueue, it is ok to remove
> virtio_rmb behind it.
>
> Signed-off-by: Joyce Kong
> Reviewed-by: Ruifeng Wang
> ---
> drivers/net/virtio/virtio_ethdev.c | 6 +++---
>
Latest VIC adapters support 64B CQ (completion queue) entries as well
as 16B entries available on all VIC models. 64B entries can greatly
reduce cache contention (CPU stall cycles) between DMA writes (Rx
packet descriptors) and polling CPU. The effect is very noticeable on
Intel platforms with DDIO
On Thu, Jan 7, 2021 at 2:33 PM Thomas Monjalon wrote:
> > Yes that may break the ABI but fortunately the checking-abi-compatibility
> > tool shows negative :) , thanks Ferruh' s guide.
> > https://github.com/ferruhy/dpdk/actions/runs/468859673
>
> That's very strange. An enum value is changed.
>
07/01/2021 13:47, Zhang, Qi Z:
>
> > -Original Message-
> > From: Thomas Monjalon
> > Sent: Thursday, January 7, 2021 6:12 PM
> > To: Guo, Jia
> > Cc: Zhang, Qi Z ; Wu, Jingjing
> > ;
> > Yang, Qiming ; Wang, Haiyue
> > ; dev@dpdk.org; Yigit, Ferruh
> > ; andrew.rybche...@oktetlabs.ru;
multi-threaded flows feature uses pthread function pthread_key_create
but for Windows the destruction option in the function is unimplemented.
to resolve it Windows will implement destruction mechanism to cleanup
mlx5_flow_workspace object for each terminated thread.
Linux flow will keep the curr
On 1/7/2021 11:32 AM, Ferruh Yigit wrote:
On 1/7/2021 5:39 AM, George Prekas wrote:
On 1/6/2021 12:02 PM, Ferruh Yigit wrote:
On 12/5/2020 5:42 AM, George Prekas wrote:
Strict-aliasing rules are violated by cast to uint16_t* in flowgen.c
and the calculated IP checksum is wrong on GCC 9 and G
From: Tal Shnaiderman
> While compiling with clang 11 the callers of the __mlx5_bit_off macro warns
> on the cast of pointers to unsigned long which is a smaller int type in
> Windows.
>
> warning: cast to smaller integer type 'unsigned long'
> from 'u8 (*)[16]' [-Wpointer-to-int-cast]
>
> To
> -Original Message-
> From: Thomas Monjalon
> Sent: Thursday, January 7, 2021 6:12 PM
> To: Guo, Jia
> Cc: Zhang, Qi Z ; Wu, Jingjing ;
> Yang, Qiming ; Wang, Haiyue
> ; dev@dpdk.org; Yigit, Ferruh
> ; andrew.rybche...@oktetlabs.ru; or...@nvidia.com;
> getel...@nvidia.com
> Subject: R
On Thu, Jan 07, 2021 at 05:53:12PM +0530, Kumar Amber wrote:
> Add support for secondary processes in ioat devices. The update
> allocates a memzone for a primary process or returns it in a
> secondary process.
>
> Signed-off-by: Kumar Amber
>
> ---
> v5
> * add error check for memzone lookup
>
1 - 100 of 148 matches
Mail list logo
