From: Xuan Ding
This patch adds the use case for async dequeue API. Vswitch can
leverage DMA device to accelerate vhost async dequeue path.
Signed-off-by: Wenwu Ma
Signed-off-by: Yuan Wang
Signed-off-by: Xuan Ding
---
doc/guides/sample_app_ug/vhost.rst | 9 +-
examples/vhost/main.c
From: Xuan Ding
This patch implements asynchronous dequeue data path for vhost split
ring, a new API rte_vhost_async_try_dequeue_burst() is introduced.
Signed-off-by: Xuan Ding
Signed-off-by: Yuan Wang
---
doc/guides/prog_guide/vhost_lib.rst| 7 +
doc/guides/rel_notes/release_22_07.rst
From: Xuan Ding
This patch refactors copy_desc_to_mbuf() used by the sync
path to support both sync and async descriptor to mbuf filling.
Signed-off-by: Xuan Ding
Reviewed-by: Maxime Coquelin
---
lib/vhost/vhost.h | 1 +
lib/vhost/virtio_net.c | 48 --
From: Xuan Ding
This patch extracts the descriptors to buffers filling from
copy_desc_to_mbuf() into a dedicated function. Besides, enqueue
and dequeue path are refactored to use the same function
sync_fill_seg() for preparing batch elements, which simplifies
the code without performance degradat
From: Xuan Ding
This patch refactors vhost async enqueue path and dequeue path to use
the same function async_fill_seg() for preparing batch elements,
which simplifies the code without performance degradation.
Signed-off-by: Xuan Ding
Reviewed-by: Maxime Coquelin
---
lib/vhost/virtio_net.c |
From: Xuan Ding
The presence of asynchronous path allows applications to offload memory
copies to DMA engine, so as to save CPU cycles and improve the copy
performance. This patch set implements vhost async dequeue data path
for split ring. The code is based on latest enqueue changes [1].
This p
In vhost_user_msg_handler(), if vhost message handling
failed, we should check whether the queue is locked and
release the lock before returning. Or, it will cause a
deadlock later.
Signed-off-by: Wenwu Ma
---
lib/vhost/vhost_user.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/lib/vhost
Currently the "extern C" section ends right before rte_dev_dma_unmap
and other DMA function declarations, causing some C++ compilers to
produce C++ mangled symbols to rte_dev_dma_unmap instead of C symbols.
This leads to build failures later when linking a final executable
against this object.
The
Fix merge error in initialization for rxDataRing feature.
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
Reviewed-by: Jochen Behrens
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/driver
The packet type is set even if it is a fragmented packet
Tested, using testpmd, for different hardware versions on
ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
Reviewed-by: Jochen Behrens
---
drivers/net/vmxnet3/vmxnet3_rxtx.c | 17 +
1 file changed, 17 insertions(+)
diff --
Currently the driver assumes that the size of the reta table is
four times the number of Rx queues at multiple places. However,
it sets it to four times the maximum number of queues (4 * 32 = 128)
when it first initializes the device. Change the other uses to use
the stored value, not the assumed d
vmxnet3 version 6 supports some new features, including but
not limited to:
- Increased max MTU up to 9190
- Increased max number of queues, both for Rx and Tx
- Removes power-of-two limitations
- Extended interrupt structures, required implementation for
additional number of queues
Tested, usin
Implement fw_version_get API for vmxnet3
Align addition order in vmxnet3_eth_dev_ops with eth_dev_ops
Tested, using testpmd, for different hardware version on
ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
Reviewed-by: Jochen Behrens
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 35 +
Count the number of entries in the Rx queue for debugging.
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2
Signed-off-by: Pankaj Gupta
Reviewed-by: Jochen Behrens
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 1 +
drivers/net/vmxnet3/vmxnet3_ethdev.h | 3 +++
drivers/
Add reta query and reta update support for VMXNET3
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
Reviewed-by: Jochen Behrens
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 67
1 file changed, 67 insertions(+)
d
Add VMXNET3 v5 support
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
Reviewed-by: Jochen Behrens
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 7 +--
drivers/net/vmxnet3/vmxnet3_ethdev.h | 2 ++
2 files changed, 7 insertions(+), 2 dele
Adding Reviewed-by info
Reviewed-by: Jochen Behrens
Pankaj Gupta (8):
net/vmxnet3: add V5 support
net/vmxnet3: implement reta query and reta update
net/vmxnet3: add Rx queue usage count utility
net/vmxnet3: report HW version on FW version get
net/vmxnet3: version 6
net/vmxnet3: advert
Pankaj,
Can you add reviewed-by info? Also, you need to add info of what has changed
since the last version.
Yong
> On May 4, 2022, at 2:23 PM, Pankaj Gupta wrote:
>
> Add VMXNET3 v5 support
>
> Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2.
>
> Signed-off-by:
04/05/2022 16:46, Tyler Retzlaff пишет:
Provide a portable type-safe thread identifier.
Provide rte_thread_self for obtaining current thread identifier.
Signed-off-by: Narcisa Vasile
Signed-off-by: Tyler Retzlaff
Acked-by: Dmitry Kozlyuk
---
lib/eal/include/rte_thread.h | 22 +++
Fix merge error in initialization for rxDataRing feature.
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/drivers/net/vmxnet3/vmxnet3_ethdev.
The packet type is set even if it is a fragmented packet
Tested, using testpmd, for different hardware versions on
ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
---
drivers/net/vmxnet3/vmxnet3_rxtx.c | 17 +
1 file changed, 17 insertions(+)
diff --git a/drivers/net/vmxnet3/vmx
Currently the driver assumes that the size of the reta table is
four times the number of Rx queues at multiple places. However,
it sets it to four times the maximum number of queues (4 * 32 = 128)
when it first initializes the device. Change the other uses to use
the stored value, not the assumed d
vmxnet3 version 6 supports some new features, including but
not limited to:
- Increased max MTU up to 9190
- Increased max number of queues, both for Rx and Tx
- Removes power-of-two limitations
- Extended interrupt structures, required implementation for
additional number of queues
Tested, usin
Implement fw_version_get API for vmxnet3
Align addition order in vmxnet3_eth_dev_ops with eth_dev_ops
Tested, using testpmd, for different hardware version on
ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 35 ++--
1 file chang
Add VMXNET3 v5 support
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 7 +--
drivers/net/vmxnet3/vmxnet3_ethdev.h | 2 ++
2 files changed, 7 insertions(+), 2 deletions(-)
diff --git a/driver
Count the number of entries in the Rx queue for debugging.
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2
Signed-off-by: Pankaj Gupta
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 1 +
drivers/net/vmxnet3/vmxnet3_ethdev.h | 3 +++
drivers/net/vmxnet3/vmxnet3_rxtx.c
Add reta query and reta update support for VMXNET3
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 67
1 file changed, 67 insertions(+)
diff --git a/drivers/net/vmxne
Pankaj Gupta (8):
net/vmxnet3: add V5 support
net/vmxnet3: implement reta query and reta update
net/vmxnet3: add Rx queue usage count utility
net/vmxnet3: report HW version on FW version get
net/vmxnet3: version 6
net/vmxnet3: advertise RETA size in device info
net/vmxnet3: set packet
Hi Andrew,
Packet type was not set for fragmented packets so we are trying to set it in
all possible scenarios.
I believe TCP packets can be fragmented.
Thanks,
Pankaj
From: Andrew Rybchenko
Date: Wednesday, May 4, 2022 at 8:08 AM
To: Pankaj Gupta , Jochen Behrens ,
Yong Wang
Cc: dev@dpdk.or
On 5/3/22 5:03 AM, Thomas Monjalon wrote:
The keyword "vector" may conflict easily.
As a rule, it is better to use the alternative keyword "__vector".
Signed-off-by: Thomas Monjalon
---
drivers/net/i40e/i40e_rxtx_vec_altivec.c | 214 ++--
drivers/net/mlx5/mlx5_rxtx_vec_altivec.h |
Hi Andrew,
Changes done in this patch was accidentally removed in commit
046f1161956777e3afb13504acbe8df2ec3a383c.
This was noticed later and we are trying to address this here.
Thanks,
Pankaj
From: Andrew Rybchenko
Date: Wednesday, May 4, 2022 at 8:09 AM
To: Pankaj Gupta , Jochen Behrens ,
> Subject: Re: [Patch v2] net/netvsc: report correct stats values
>
> On 5/3/2022 9:48 PM, Long Li wrote:
> >> Subject: Re: [Patch v2] net/netvsc: report correct stats values
> >>
> >> On 5/3/2022 8:14 PM, Long Li wrote:
> Subject: Re: [Patch v2] net/netvsc: report correct stats values
>
Hi Andrew,
Please refer struct rte_eth_dev. Following assignment already preserve the
order in struct rte_eth_dev
> diff --git a/drivers/net/vmxnet3/vmxnet3_ethdev.c
> b/drivers/net/vmxnet3/vmxnet3_ethdev.c
> index eb65499cf2..a76796716b 100644
> --- a/drivers/net/vmxnet3/vmxnet3_ethdev.c
> +++
On Tue, 03 May 2022 10:47:58 +0200
Thomas Monjalon wrote:
> 24/04/2022 07:34, Subendu Santra:
> > Hi Stephen,
> >
> > We were going through the patch set:
> > https://inbox.dpdk.org/dev/20200715212228.28010-7-step...@networkplumber.org/
> > and hoping to get clarification on the behaviour if p
On 4/13/2022 10:01 AM, Min Hu (Connor) wrote:
<...>
在 2022/4/11 13:40, Ke Zhang 写道:
when creating a bonding device, if the slave device's rss key length
= standard_rss_key length + extended_hash_key length, then bonding
device will be same as slave, in function bond_ethdev_configure(),
the def
Provide a portable type-safe thread identifier.
Provide rte_thread_self for obtaining current thread identifier.
Signed-off-by: Narcisa Vasile
Signed-off-by: Tyler Retzlaff
Acked-by: Dmitry Kozlyuk
---
lib/eal/include/rte_thread.h | 22 ++
lib/eal/unix/rte_thread.c| 11
Implement functions for getting/setting thread affinity.
Threads can be pinned to specific cores by setting their
affinity attribute.
Windows error codes are translated to errno-style error codes.
The possible return values are chosen so that we have as
much semantic compatibility between platform
Establish unit test for testing thread api. Initial unit tests
for rte_thread_{get,set}_affinity_by_id().
Signed-off-by: Narcisa Vasile
Signed-off-by: Tyler Retzlaff
---
app/test/meson.build| 2 ++
app/test/test_threads.c | 81 +
2 files chan
this series provides basic dependencies for additional eal thread api
additions. series includes
* basic platform error number conversion.
* function to get current thread identifier.
* functions to get and set affinity with platform agnostic thread
identifier.
* minimal unit test of get and set
On 5/3/22 07:22, Pankaj Gupta wrote:
Fix merge error in initialization for rxDataRing feature.
Is it a bug fix? If so, it requires corresponding "Fixes:" tag and Cc to
sta...@dpdk.org in order to be backported to stable releases.
Tested, using testpmd, for different hardware version on ESXi
On 5/3/22 07:22, Pankaj Gupta wrote:
The packet type is set even if it is a fragmented packet
I'm wondering if is really IPv4/IPv6 fragmented packets or just
scattered on Rx across many Rx buffers.
I'm asking since fragmented sounds weird with TCP, since TCP
spec forbids fragmentation.
Teste
Consider:
net/vmxnet3: advertise RETA size in device info
On 5/3/22 07:22, Pankaj Gupta wrote:
Currently the driver assumes that the size of the reta table is
reta -> RETA (see devtools/words-case.txt)
four times the number of rx queues at multiple places. However,
rx -> Rx
it sets it t
Summary should say what is done. Consider:
net/vmxnet3: support virtual HW version 6
On 5/3/22 07:22, Pankaj Gupta wrote:
vmxnet3 version 6 supports some new features, including but
not limited to:
- Increased max MTU up to 9190
- Increased max number of queues, both for RX and TX
RX -> Rx, T
> From: Andrew Rybchenko [mailto:andrew.rybche...@oktetlabs.ru]
> Sent: Wednesday, 4 May 2022 16.28
>
> rx -> Rx
>
> in summary in accordance with spelling in DPDK
>
> On 5/3/22 07:22, Pankaj Gupta wrote:
> > Count the number of entries in the rx queue for debugging.
>
> rx -> Rx
>
> As I unde
hw -> HW
However, the API says that it is a firmware version get
Consier:
net/vmxnet3: report HW version on FW version get
On 5/3/22 07:22, Pankaj Gupta wrote:
Implement fw_version_get API for vmxnet3
fw -> FW
"for vmxnet3" is a duplicate, since summary already says so.
Tested, using tes
Currently, when we free previously allocated memory, we mark the area as
"freed" for ASan purposes (flag 0xfd). However, sometimes, freeing a
malloc element will cause pages to be unmapped from memory and re-backed
with anonymous memory again. This may cause ASan's "use-after-free"
error down the l
On 5/3/22 07:22, Pankaj Gupta wrote:
Pankaj Gupta (8):
vmxnet3: Added V5 support
vmxnet3: implement reta query and reta update
vmxnet3: add rx queue usage count utility
vmxnet3: add get hw version api
vmxnet3, version 6
vmxnet3: set reta size
vmxnet3: Set packet for fragmente
rx -> Rx
in summary in accordance with spelling in DPDK
On 5/3/22 07:22, Pankaj Gupta wrote:
Count the number of entries in the rx queue for debugging.
rx -> Rx
As I understand debugging is not the only purpose of the API.
Tested, using testpmd, for different hardware version on ESXi 7.0
On 5/3/22 07:22, Pankaj Gupta wrote:
Added reta query and reta update support for VMXNET3
Tested, using testpmd, for different hardware version on ESXi 7.0 Update 2.
Signed-off-by: Pankaj Gupta
---
drivers/net/vmxnet3/vmxnet3_ethdev.c | 67
1 file changed, 67 in
Currently, when we free previously allocated memory, we mark the area as
"freed" for ASan purposes (flag 0xfd). However, sometimes, freeing a
malloc element will cause pages to be unmapped from memory and re-backed
with anonymous memory again. This may cause ASan's "use-after-free"
error down the l
On 5/3/2022 9:48 PM, Long Li wrote:
Subject: Re: [Patch v2] net/netvsc: report correct stats values
On 5/3/2022 8:14 PM, Long Li wrote:
Subject: Re: [Patch v2] net/netvsc: report correct stats values
On 5/3/2022 7:18 PM, Long Li wrote:
Subject: Re: [Patch v2] net/netvsc: report correct stats
On 4/21/2022 12:14 PM, Niklas Soderlund wrote:
Hi David,
Thanks for your work.
On 2022-04-08 11:41:16 +0200, David Marchand wrote:
Looking at this driver history, there was never a need for including
execinfo.h.
Signed-off-by: David Marchand
Reviewed-by: Niklas Söderlund
Fixes: bdb
On 5/03/2022 9:05 PM, Ferruh Yigit wrote:
> On 4/20/2022 2:46 PM, Walter Heymans wrote:
> > The 'max_rx_pktlen' value was previously read from hardware, which was
> > set by the running firmware. This caused confusion due to different
> > meanings of 'MAX_MTU'. This patch updates the 'max_rx_pktlen
Add example for hash function operation.
Signed-off-by: Cristian Dumitrescu
---
Change log:
V4:
-Removed redundant line in the CLI file.
-Added comment on the crc32 hash function in the .spec file.
examples/pipeline/examples/hash_func.cli | 34 +++
examples/pipeline/examples/hash_func.sp
Add support for hash functions that compute a signature for an array
of bytes read from a packet header or meta-data. Useful for flow
affinity-based load balancing.
Signed-off-by: Cristian Dumitrescu
---
Depends-on: series-22635 ("[V2,1/3] table: improve learner table timers")
Change log:
V4:
-
On Tue, May 3, 2022 at 8:23 PM Quentin Armitage wrote:
>
> rte_pmd_tun/tap_probe() allocates pmd->intr_handle in eth_dev_tap_create()
> and it should not be freed until rte_pmd_tap_remove() is called.
>
> Inspection of tap_rx_intr_vec_set() shows that the call to
> tap_tx_intr_vec_uninstall() was
On 4/28/22 16:15, Chengwen Feng wrote:
Currently the value returned when xstats is NULL of rte_eth_xstats_get()
is not specified, some PMDs (eg. hns3/ipn3ke/mvpp2/axgbe) return zero
while others return the required number of elements.
This patch defines that the return value should be the requir
On 4/26/22 05:43, Robin Zhang wrote:
Add a new telemetry command /ethdev/module_eeprom to dump the module
EEPROM of each port. The format of module EEPROM information follows
the SFF(Small Form Factor) Committee specifications.
Current the format support SFP(Small Formfactor Pluggable)/SFP+/
QSF
Le 04/05/2022 à 08:34, David Marchand a écrit :
Hello Édouard,
On Tue, May 3, 2022 at 9:05 PM Edouard Gaulué wrote:
I just received a brand new server with an Intel X710 T4L inside. I
hoped I could use it at 5GBaseT speed as claimed by Intel, but I couldn't.
In fact, I run ESXi 6.7U3 on this
Pavan Nikhilesh writes:
> Add function to quiesce any core specific resources consumed by
> the event port.
>
> When the application decides to migrate the event port to another lcore
> or teardown the current lcore it may to call `rte_event_port_quiesce`
> to make sure that all the data associ
jer...@marvell.com writes:
> From: Jerin Jacob
>
> Currently, meter object supports only DSCP based on input color table,
> The patch enhance that to support VLAN based input color table,
> color table based on inner field for the tunnel use case, and
> support for fallback color per meter if p
kirankum...@marvell.com writes:
> From: Kiran Kumar K
>
> Adding cnxk device driver support to configure custom sa index.
> Custom sa index can be configured as part of the session create
> as SPI, and later original SPI can be updated using session update.
>
> Signed-off-by: Kiran Kumar K
> -
On 4/26/22 05:43, Robin Zhang wrote:
Introduce a new telemetry command /ethdev/module_eeprom to show module
EEPROM for each port. The format of module EEPROM information follows
the SFF(Small Form Factor) Committee specifications.
Current the format support SFP(Small Formfactor Pluggable)/SFP+/
63 matches
Mail list logo
