Partial DMA unmap is not supported by VFIO type1 IOMMU
in Linux. Though the return value is zero, the returned
DMA unmap size is not same as expected size.
So add test case and fix to both heap triggered DMA
mapping and user triggered DMA mapping/unmapping.
Refer vfio_dma_do_unmap() in drivers/vfi
In order to save DMA entries limited by kernel both for externel
memory and hugepage memory, an attempt was made to map physically
contiguous memory in one go. This cannot be done as VFIO IOMMU type1
does not support partially unmapping a previously mapped memory
region while Heap can request for m
Test case alloc's system pages and tries to performs a user
DMA map and unmap both partially and fully.
Signed-off-by: Nithin Dabilpuram
Acked-by: Anatoly Burakov
---
app/test/meson.build | 1 +
app/test/test_vfio.c | 106 +++
2 files changed, 1
Currently external memory test uses 4K page size.
VFIO DMA mapping works only with system page granularity.
Earlier it was working because all the contiguous mappings
were coalesced and mapped in one-go which ended up becoming
a lot bigger page. Now that VFIO DMA mappings both in IOVA as VA
and IO
Partial unmapping is not supported for VFIO IOMMU type1
by kernel. Though kernel gives return as zero, the unmapped size
returned will not be same as expected. So check for
returned unmap size and return error.
For IOVA as PA, DMA mapping is already at memseg size
granularity. Do the same even for
Hi David,
As mentioned below in #1, I sent v5 with memory allocated from Heap which I
think
falls in initially configured DMA window 0x0 0x40 atleast in
Linux as DPDK memory starts after HEAP.
Let me know if it is ok for POWER9 system.
On Thu, Dec 03, 2020 at 12:44:06PM +0530, Nit
On 12/13/2020 8:06 PM, Liron Himi wrote:
-Original Message-
From: Ferruh Yigit
Sent: Wednesday, 9 December 2020 12:50
To: Liron Himi ; bruce.richard...@intel.com
Cc: dev@dpdk.org; Akhil Goyal
Subject: [EXT] Re: [dpdk-dev] [PATCH v1 1/2] meson: update meson build for
armada drivers
Ex
Add a test case to test scan operation post clear of half
cacheline of slabs.
Also fix meson.build to include test_bitmap.c in the compilation.
Signed-off-by: Nithin Dabilpuram
---
app/test/meson.build | 1 +
app/test/test_bitmap.c | 34 +-
2 files changed, 34
Currently bitmap line not empty check API assumes cache line
of 64B and only checks 8 slabs. Since in 128B cacheline, we
have 16 slabs per cacheline, plt_bitmap_clear() will mark
complete line as empty as soon as 8 slabs are full thereby
breaking bitmap scan functionality. Fix it by defining new
__
11/12/2020 21:07, Ferruh Yigit:
> On 12/4/2020 5:36 PM, David Marchand wrote:
> > With the recent changes in terms of free access to the Travis CI, let's
> > offer an alternative with GitHub Actions.
> > Running jobs on ARM is not supported unless using external runners, so
> > this commit only add
Hi Jiawei,
> -Original Message-
> From: dev On Behalf Of Jiawei Zhu
> Sent: Saturday, December 12, 2020 12:53 AM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v2] net/virtio-user: fix run close(0) and close
> callfd
>
> From: Jiawei Zhu
>
> When i < VIRTIO_MAX_VIRTQUEUES and j == i,
David Marchand writes:
> With the recent changes in terms of free access to the Travis CI, let's
> offer an alternative with GitHub Actions.
> Running jobs on ARM is not supported unless using external runners, so
> this commit only adds builds for x86_64 and cross compiling for i386 and
> aarch6
David Marchand writes:
> v21 ABI will be maintained until v21.11.
>
> Let's use the latest released libabigail 1.8.
>
> In GitHub Actions, libabigail binaries and the ABI reference are stored
> in two shared caches as all branches can use the same.
>
> While at it, we can reproduce changes from t
Currently, the maximum flow priority to user is 4, PMD will
manage 3 sub-priorities per user priority according to L2, L3
and L4. This count of priority is not enough if user want to
do some flow match by priorities, such as LPM, for one IPV4
address, we need 32 priorities for each bit of 32 mask l
Signed-off-by: Ferruh Yigit
---
content/roadmap/_index.md | 8
1 file changed, 8 insertions(+)
diff --git a/content/roadmap/_index.md b/content/roadmap/_index.md
index 456690266cfc..1eec863566c2 100644
--- a/content/roadmap/_index.md
+++ b/content/roadmap/_index.md
@@ -13,13 +13,19 @@ T
Hi Maxime and David:
Could we start to review this patch?
/Thanks, huawei
On 2020/11/10 20:42, David Marchand wrote:
On Tue, Nov 10, 2020 at 1:35 PM 谢华伟(此时此刻） wrote:
Previously there are different ways to get port address based on
different DPDK uio driver(IGB_UIO/UIO_PCI_GENERIC/VFIO), whi
Hi Yicai,
> In some situations, we would get several ip fragments, which total
> data length is less than min_ip_len(64) and padding with zeros.
> We simulated intermediate fragments by modifying the MTU.
> To illustrate the problem, we simplify the packet format and
> ignore the impact of the
On Fri, Dec 4, 2020 at 6:37 PM David Marchand wrote:
>
> With the recent changes in terms of free access to the Travis CI, let's
> offer an alternative with GitHub Actions.
> Running jobs on ARM is not supported unless using external runners, so
> this commit only adds builds for x86_64 and cross
On 12/9/2020 3:16 AM, Steve Yang wrote:
The jumbo frame used the 'RTE_ETHER_MAX_LEN' as boundary condition, this
fix will change the boundary condition with 'RTE_ETHER_MTU'.
When the MTU(1500) set, the frame type of rx packet will be different
if used different overhead, it will cause the consis
In l3fwd no of transmit queues is calculated based on no of
lcores with which it is launched. Hence maximum no of tx
queues possible per port should depend on RTE_MAX_LCORE value.
Fixes: 26b5b020 ("examples/l3fwd: modularize")
Cc: sta...@dpdk.org
Signed-off-by: Harman Kalra
---
examples/l3f
On 12/13/2020 6:16 AM, Tal Shnaiderman wrote:
On older processors, NUMA isn't bound to PCIe locality.
those cases return ERROR_NOT_FOUND in response to the
SetupDiGetDevicePropertyW call with DEVPKEY_Device_Numa_Node
attribute.
This error fails the probe process for the PCIe device.
this commit
On Thu, Sep 17, 2020 at 10:05 AM Olivier Matz wrote:
>
> Hi Lance,
>
> On Mon, Aug 24, 2020 at 04:11:45PM -0400, Lance Richardson wrote:
> > I was looking for some clarification regarding how rx checksum
> > flags should be set for tunnel packets having both inner and outer
> > IP/L4 headers.
> >
Write completion queue doorbell before receive descriptor
doorbell to avoid possiblity of completion queue overflow
when completion queue size is equal to receive descriptor
ring size. Remove unnecessary compiler barriers (db write
functions have the necessary barriers.)
Fixes: 637e34befd9c ("net/
Limit number of representor packets transferred per poll
to requested burst size.
Fixes: 6dc83230b43b ("net/bnxt: support port representor data path")
Signed-off-by: Lance Richardson
Reviewed-by: Ajit Kumar Khaparde
Reviewed-by: Somnath Kotur
Cc: sta...@dpdk.org
---
drivers/net/bnxt/bnxt_rxr.c
Write completion queue doorbell before receive descriptor
doorbell to avoid possibility of completion queue overflow
when completion queue size is equal to receive descriptor
ring size. Remove unnecessary compiler barriers (db write
functions have the necessary barriers.)
Fixes: 637e34befd9c ("net
> -Original Message-
> From: dev On Behalf Of Andrew Boyer
> Sent: Friday, December 11, 2020 5:06 PM
> To: dev@dpdk.org
> Cc: Andrew Boyer ; Burakov, Anatoly
> ; sta...@dpdk.org
> Subject: [dpdk-dev] [PATCH] eal: don't log uninitialized string
>
> Our Coverity scan pointed out that path[]
Add some missing interrupt implementations on Windows.
Also add respective functions to export file.
Signed-off-by: Tal Shnaiderman
Signed-off-by: Pallavi Kadam
Reviewed-by: Ranjit Menon
Acked-by: Narcisa Vasile
v2 changes:
Add remaining interrupt APIs (Dmitry Kozlyuk)
---
lib/librte
Hi Dmitry,
On 12/11/2020 1:07 PM, Dmitry Kozlyuk wrote:
On Thu, 10 Dec 2020 16:26:07 -0800, Pallavi Kadam wrote:
Add some missing interrupt implementations on Windows.
Also add respestive functions to export file.
Signed-off-by: Tal Shnaiderman
Signed-off-by: Pallavi Kadam
Reviewed-by: Ranji
From: Souvik Dey
Sent: Saturday, December 12, 2020 9:05 PM
To: Xing, Beilei ; Guo, Jia ; Zhang,
Qi Z
Cc: dev@dpdk.org; Souvik Dey
Subject: [PATCH v2] net/i40e: issue with ADD VLAN from Guest
Reset the configuration of vlan strip that would be change
by the pf kernel driver when adding vlan fr
This patchset aims to support AVX512 vPMD on i40e.
And the changes are only target to AVX512 vector path.
Leyi Rong (3):
net/i40e: remove devarg use-latest-supported-vec
net/i40e: add AVX512 vector path
net/i40e: optimize Tx by using AVX512
doc/guides/nics/i40e.rst|9 -
As eal parameter --force-max-simd-bitwidth is already introduced,
to make it more clear when setting rx/tx function, remove
devarg use-latest-supported-vec support.
Signed-off-by: Leyi Rong
---
doc/guides/nics/i40e.rst | 9 ---
drivers/net/i40e/i40e_ethdev.c | 63 +--
dr
Add AVX512 support for i40e PMD. This patch adds i40e_rxtx_vec_avx512.c
to support i40e AVX512 vPMD.
This patch aims to enable AVX512 on i40e vPMD. Main changes are focus
on Rx path compared with AVX2 vPMD.
Signed-off-by: Leyi Rong
Signed-off-by: Bruce Richardson
---
drivers/net/i40e/i40e_rxtx
Optimize Tx path by using AVX512 instructions and vectorize the
tx free bufs process.
Signed-off-by: Leyi Rong
Signed-off-by: Bruce Richardson
---
drivers/net/i40e/i40e_rxtx.c| 19 +++
drivers/net/i40e/i40e_rxtx.h| 4 +
drivers/net/i40e/i40e_rxtx_vec_avx512.c | 146 ++
Tested-by: Zhou, Jun
-Original Message-
From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of dapengx...@intel.com
Sent: Friday, December 11, 2020 9:35 AM
To: Guo, Jia
Cc: dev@dpdk.org; Yu, DapengX ; Zhang, Qi Z
Subject: [dpdk-dev] [PATCH] net/ixgbe: fix fdirctrl register setting
From:
> -Original Message-
> From: Ananyev, Konstantin [mailto:konstantin.anan...@intel.com]
> Sent: Monday, December 14, 2020 10:45 PM
> To: luyicai ; dev@dpdk.org
> Cc: Zhoujingbin (Robin, Russell Lab) ; chenchanghu
> ; Lilijun (Jerry) ;
> Linhaifeng ; Guohongzhi (Russell Lab)
> ; wangyun
In order to allow mbuf mark ID update in Rx data-path, there is a
mechanism in the PMD to enable it according to the rte_flows.
When a flow with mark ID and RSS/QUEUE action exists, all the relevant
Rx queues will be enabled to report the mark ID.
When shared RSS action is combined with mark actio
Hi, dapeng
> -Original Message-
> From: dapengx...@intel.com
> Sent: Friday, December 11, 2020 9:35 AM
> To: Guo, Jia
> Cc: dev@dpdk.org; Yu, DapengX ; Zhang, Qi Z
>
> Subject: [PATCH] net/ixgbe: fix fdirctrl register setting
>
> From: YU DAPENG
>
> The function ixgbe_fdir_set_flexby
> -Original Message-
> From: Wang, Haiyue
> Sent: Monday, December 14, 2020 15:12
> To: dev@dpdk.org
> Cc: Yang, Qiming ; Wu, Jingjing
> ; Zhang, Qi Z ; Wang, Haiyue
> ; Xing, Beilei
> Subject: [PATCH v2 5/5] net/iavf: support new VLAN virtchnl opcodes
>
> The new VLAN virtchnl opcode
Hi, Haiyue
Will this patch set cover DCF QinQ enable in the future?
Qiming
> -Original Message-
> From: Wang, Haiyue
> Sent: Monday, December 14, 2020 15:12
> To: dev@dpdk.org
> Cc: Yang, Qiming ; Wu, Jingjing
> ; Zhang, Qi Z ; Wang, Haiyue
>
> Subject: [PATCH v2 0/5] Add AVF & DCF VLAN
> -Original Message-
> From: Yang, Qiming
> Sent: Tuesday, December 15, 2020 13:22
> To: Wang, Haiyue ; dev@dpdk.org
> Cc: Wu, Jingjing ; Zhang, Qi Z
> Subject: RE: [PATCH v2 0/5] Add AVF & DCF VLAN feaure
>
> Hi, Haiyue
> Will this patch set cover DCF QinQ enable in the future?
Yes.
>
> -Original Message-
> From: Yang, Qiming
> Sent: Tuesday, December 15, 2020 13:20
> To: Wang, Haiyue ; dev@dpdk.org
> Cc: Wu, Jingjing ; Zhang, Qi Z ;
> Xing, Beilei
>
> Subject: RE: [PATCH v2 5/5] net/iavf: support new VLAN virtchnl opcodes
>
>
>
> > -Original Message-
> > F
Change Summary:
1. Support double vlan mode.
2. Support eCPRI msg type 0 filter on fdir.
3. Support ecPRI udp tunnel create/destroy.
4. Support vxlan vni filter on fdir.
5. Support GTP filter on switch.
6. Couple bugs fixes and code clean.
Qi Zhang (27):
net/ice/base: modify ptype map for UDP
Add below ptypes into ice_ptypes_udp_il.
MAC_IPV4_GTPC_TEID
MAC_IPV6_GTPC_TEID
MAC_IPV4_GTPC
MAC_IPV6_GTPC
MAC_IPV4_GTPU
MAC_IPV6_GTPU
So outer UDP can also be selected as input set for GTPC and
GTPU non IP packet.
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_flow.c | 2 +-
1 file chan
250 msec timeout is insufficient for some AQ commands. Advice from FW
team was to increase the timeout. Increased to 1 second.
Signed-off-by: Fabio Pricoco
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_controlq.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/dri
Added read of security revision disabled and updated disabled.
Signed-off-by: Fabio Pricoco
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_common.c | 17 +
drivers/net/ice/base/ice_type.h | 4
2 files changed, 21 insertions(+)
diff --git a/drivers/net/ice/base/ice
The ice_read_flash_module interface for reading from the various NVM
modules was introduced.
It's purpose is two-fold. First, it enables reading data from the CSS
header, used to allow accessing the image security revisions. Second, it
allowed reading from either the 1st or the 2nd NVM bank. This
Modify ice_get_nvm_srev and ice_get_orom_srev to take the
ice_flash_bank enumeration that specifies whether to read from the
active or the inactive flash module. Rename and refactor the
ice_read_active_nvm_module and ice_read_active_orom_module functions to
take the bank enum value as well.
With t
Refactor ice_read_flash_module so that it takes a size and a length
value, rather than always reading in 2-byte increments. The
ice_read_nvm_module and ice_read_orom_module wrapper functions will
still read a u16 with the byte-swapping enabled.
This will be used in a future change to implement rea
The driver currently reads the combo image version data from within the
Boot Configuration TLV block of the PFA area of the NVM. This allows
access to the active Option ROM version data, assuming that it has been
properly copied into this section.
There is no equivalent method for reading the Opti
Similar to ice_get_inactive_orom_ver, add a function to read the NVM
version data from the inactive section of flash. The primary motivation
of this function is to allow the driver to report the version of
a pending update that has not yet been activated.
To do this, refactor ice_get_nvm_ver_info
Add support to enable GTP filtering using advanced switch filter mechanism.
Example of GTP filter criteria:
Outer UDP + GTP (Optional) + Inner IPv4[6] + Inner TCP/UDP
Signed-off-by: Kiran Patil
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_flex_type.h | 1 +
drivers/net/ice/base/ice_
Replaced the netlist related macro and removed some
unused macro
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_adminq_cmd.h | 26
drivers/net/ice/base/ice_type.h | 43 ++-
2 files changed, 35 insertions(+), 34 deletions(-)
diff --git a/drive
Add definitions for the various health status codes that can be reported
by firmware.
Signed-off-by: Nick Nunley
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_adminq_cmd.h | 29 +++
1 file changed, 29 insertions(+)
diff --git a/drivers/net/ice/base/ice_adminq_cmd
AQ 0x0706 lets user to verify and recalculate NVM checksum. The usage
depends on command flags.
Signed-off-by: Milena Olech
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_nvm.c | 28
drivers/net/ice/base/ice_nvm.h | 1 +
2 files changed, 29 insertions(+)
dif
There is added condition,to copy information about used module into
memory
Signed-off-by: Szymon T Cudzilo
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_common.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/drivers/net/ice/base/ice_common.c
b/drivers/net/ice/base/ice_common.
The VLAN mode of the device has to be configured while the global
configuration lock is held while downloading the DDP, specifically after
the DDP has been downloaded. In order to support this a VLAN mode
interface was added. By default the device will stay in single VLAN
mode (SVM), which is the c
Enable VXLAN VNI field in FDIR.
Signed-off-by: Zhirun Yan
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_fdir.c | 43 +
drivers/net/ice/base/ice_fdir.h | 8 ++
drivers/net/ice/base/ice_flow.c | 21
drivers/net/ice/base/ice_flow.h | 6
The tcam information in AQ command buffer is not correct when destroy
tunnel entries. The tcam count was always ONE even destroy multiple
entries, and the offset of tcam memory was also incorrect. This patch is
to fix this issue.
Fixes: 884efe3ca1f3 ("net/ice/base: add VXLAN/generic tunnel managem
If a driver wants to configure double VLAN mode (DVM) it needs to
first check if the DDP supports DVM. To do this the driver needs to read
the package metadata section via the upload section AQ (0x04C1).
If the DDP doesn't support configuring double VLAN mode (DVM), then
there is nothing to do reg
Retry sending some AQ commands, as result of EBUSY AQ error.
This change follows the latest guidelines from HW. It is better
to retry the same AQ command several times, as the result of
EBUSY, instead of returning error to the caller right away.
Signed-off-by: Chinh T Cao
Signed-off-by: Qi Zhang
As the user may be expected to take action on this issue, change the
message to a warning so that the message is more easily accessible than
a debug. Also, add the error code to further aide in identifying the
problem.
Signed-off-by: Tony Nguyen
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/
Remove the recursive way of adding the nodes to the layer in order
to reduce the stack usage. Instead the algorithm is modified to use
a while loop.
The previous code was scanning recursively the nodes horizontally.
The total stack consumption will be based on number of nodes present
on that layer
Added handling of allocation fault for ice_vsi_list_map_info
Should also check dereference of NULL pointer to filters VSI list
information for FWD_TO_VSI_LISt type only, otherwise, the FWD_TO_VSI type
filters by the given VSI can't be located.
Also the point *pi should not be NULL pointer, it is a
In case of new link establishment flow we should use
Report Default Configuration if FW AQ API version
supports it. This patch adds check function for Report
Default Configuration support and updates ice_set_fc(),
ice_cfg_phy_fec() and ice_aq_get_phy_caps() accordingly.
Signed-off-by: Mateusz Pacu
For get PHY abilities AQ, the specification defines "report modes"
as "with media", "without media" and "active configuration". For
clarity, rename macros to align with the specification.
Signed-off-by: Anirudh Venkataramanan
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_adminq_cmd.h |
This patch adds extracting field for flow type eCPRI over mac message
type 0 and field PC_ID which is used for FDIR/RSS packet steering.
This patch also removes redundant hdr bitmask ops for PPPOE hdr.
Signed-off-by: Yahui Cao
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_fdir.c
This patch adds extracting field for flow type eCPRI over udp message
type 0 and field PC_ID which is used for FDIR/RSS packet steering.
Signed-off-by: Yahui Cao
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_fdir.c | 57 +++-
drivers/net/ice/base/ice_fdir.h
The only time you can ever have a rq_last_status is if
a firmware event was somehow reporting a status on the receive
queue, which are generally firmware initiated events or
mailbox messages from a VF. Mostly this struct member was unused.
Fix this problem by still printing the value of the field
The change is for kernel driver RDMA usage, DPDK driver does not
enable RDMA, just for code sync.
Signed-off-by: Qi Zhang
---
drivers/net/ice/base/ice_adminq_cmd.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ice/base/ice_adminq_cmd.h
b/drivers/net/ice/base/
When testpmd launched, the default RSS hash function will only activate
'ETH_RSS_IP', it isn't consistent with the command result of
'port config all rss default'.
Removed the default 'rss_hf' limited in 'init_port_config', that will
use the default supported RSS hash function for NIC.
Fixes: 908
Li Zhang (8):
net/mlx5: use mask for meter register setting
common/mlx5: add definitions for ASO flow meter
common/mlx5: add read ASO flow meter HCA capability
common/mlx5: add DevX API to create ASO flow meter object
net/mlx5: flow meter pool to manage meter object
net/mlx5: init/unini
ASO meter feature may require to locate the flow
context tag action after the ASO action.
The register used for the meter tag is also used for
the meter color:
Bits[0-7] A meter color value set by the HW.
Bits[8-31] A flow meter context set by SW.
Currently the tag action for meter writes all the
This patch adds different PRM definitions, related to ASO flow meter
feature, in MLX5 PMD code.
Signed-off-by: Li Zhang
---
drivers/common/mlx5/mlx5_prm.h | 75 --
1 file changed, 71 insertions(+), 4 deletions(-)
diff --git a/drivers/common/mlx5/mlx5_prm.h b/driv
Add DevX API to create ASO flow meter object.
Signed-off-by: Li Zhang
---
drivers/common/mlx5/mlx5_devx_cmds.c | 54
drivers/common/mlx5/mlx5_devx_cmds.h | 18 +-
drivers/common/mlx5/version.map | 3 +-
3 files changed, 73 insertions(+), 2 deletions(-)
Init/uninit flow meter SQ for WQE
Signed-off-by: Li Zhang
---
drivers/net/mlx5/linux/mlx5_os.c | 17
drivers/net/mlx5/meson.build | 2 +-
drivers/net/mlx5/mlx5.c | 67 -
drivers/net/mlx5/mlx5.h | 20 ++--
Add ASO flow meter pool to manage meter object
Signed-off-by: Li Zhang
---
drivers/net/mlx5/mlx5.c| 2 +-
drivers/net/mlx5/mlx5.h| 190 -
drivers/net/mlx5/mlx5_flow.c | 44 ++-
drivers/net/mlx5/mlx5_flow.h | 186 -
drivers/net/mlx5/m
ASO flow meter send WQE and CQE handle functions
Signed-off-by: Li Zhang
---
drivers/net/mlx5/mlx5.h| 4 +
drivers/net/mlx5/mlx5_flow.h | 2 +-
drivers/net/mlx5/mlx5_flow_aso.c | 182 +
drivers/net/mlx5/mlx5_flow_dv.c| 5 +-
drivers/net/m
When ASO action is available, use it as the meter action
Signed-off-by: Shun Hao
Signed-off-by: Li Zhang
---
drivers/net/mlx5/linux/mlx5_os.c | 4 +-
drivers/net/mlx5/mlx5.c| 8 ++
drivers/net/mlx5/mlx5.h| 18 ++---
drivers/net/mlx5/mlx5_flow.c | 66 +++
Read and store the device capability of FLOW_METER_ASO general object,
using the DevX API.
Signed-off-by: Li Zhang
---
drivers/common/mlx5/mlx5_devx_cmds.c | 14 ++
drivers/common/mlx5/mlx5_devx_cmds.h | 8
2 files changed, 22 insertions(+)
diff --git a/drivers/common/mlx5
Update function mlx5_devx_cmd_query_hca_attr() to add the
reg_c_preserve bit query.
The stored metadata in register C may be lost in NIC Tx and
FDB egress while doing one one of the following operations:
- packet encapsulation.
- packet mirroring (multiple processing paths).
- packet sampling (
mlx5 E-Switch mirroring is implemented as multiple destination array in
one steering table. The array currently supports only port ID as
destination actions.
This patch adds the jump action support to the array as one of destination.
The packets can be mirrored to the port and jump to next table i
MLX5 E-Switch mirroring is implemented as multiple destination array in
one steering table. The array currently supports only port ID as
destination actions.
This RFC adds the below supports for MLX5 PMD driver:
- Supports the metadata register Cx preserve capability query.
- Supports the jump
The sampling feature introduces the scale flow group with factor,
then the scaled table value can be used for the normal path table
due to this table be created implicitly.
But if the input group value already be scaled, for example the
group value of sampling suffix flow, then use 'skip_scale" fl
While there's the modify action and sample action with ratio=1
in the E-Switch flow, and modify action is after the sample
action, means that the modify should only impact on after sample.
MLX5 PMD will monitor the above case and split the E-Switch flow
into two sub flows, smiliar as sample flow di
Support rss action in the sample sub-actions list.
The examples for the sample flow use case and result as below:
set sample_actions 0 mark id 0x12 / rss queues 0 1 2 3 end / end
flow create 0 ingress group 1 pattern eth / end actions
sample ratio 1 index 0 / jump group 2 / end
This flow will
PMD validates the rss action in the sample sub-actions list,
then translate into rdma-core action and it will be used for sample
path destination.
If the RSS action both in sample sub-actions list and original flow,
the rss level and rss type in the sample sub-actions list should be
consistent wit
Currently the sample flow only supports Queue action in NIC-Rx domain.
This RFC adds the RSS action support in the sample sub-actions list.
The examples for the sample flow with RSS action and result as below:
set sample_actions 0 mark id 0x12 / rss queues 0 1 2 3 end / end
flow create
87 matches
Mail list logo
