> -Original Message-
> From: Thomas Monjalon
> Sent: Thursday, July 22, 2021 6:24 AM
> To: Xueming(Steven) Li
> Cc: Slava Ovsiienko ; dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v4 00/16] net/mlx5: support Sub-Function
>
> 21/07/2021 16:37, Xueming Li:
> > Sub-Function [1] is a porti
在 2021/7/21 23:29, Ferruh Yigit 写道:
On 7/19/2021 4:35 AM, Huisong Li wrote:
Hi, Ferruh
Hi Huisong,
Thanks for the review.
在 2021/7/10 1:29, Ferruh Yigit 写道:
There is a confusion on setting max Rx packet length, this patch aims to
clarify it.
'rte_eth_dev_configure()' API accepts max Rx
On 7/21/21 11:22 AM, Andrew Rybchenko wrote:
> From: Ivan Ilchenko
>
> Fix the driver to report maximum MTU obtained from config if
> VIRTIO_NET_F_MTU is supported or calculated based on maximum
> Rx packet length.
>
> Fixes: ad97ceece12c ("ethdev: add min/max MTU to device info")
> Cc: sta..
On 7/21/21 11:29 AM, Andrew Rybchenko wrote:
> On 7/20/21 7:19 PM, Maxime Coquelin wrote:
>>
>>
>> On 7/20/21 9:54 AM, Andrew Rybchenko wrote:
>>> From: Ivan Ilchenko
>>>
>>> Report Rx scatter offload capability depending on
>>> VIRTIO_NET_F_MRG_RXBUF.
>>>
>>> If Rx scatter is not requested, en
> You should implement checks for crypto doc in devtools/check-doc-vs-
> code.sh
Ok will look into it.
> 20/07/2021 14:14, David Marchand:
> > On Tue, Jul 20, 2021 at 1:59 PM Akhil Goyal wrote:
> > >
> > > Hi David,
> > > >
> > > > > > deps += ['common_octeontx', 'mempool_octeontx', 'bus_vdev',
> > > > > 'net_octeontx']
> > > > > > +deps += ['crypto_octeontx']
> > > > >
> > > > > This extra depen
Old clanng requires libatomic as well as gcc. Avoid compiler name and
version based checks. Add custom test for 16-byte atomic operations
to find out if libatomic is required to build.
Bugzilla ID: 760
Signed-off-by: Andrew Rybchenko
---
drivers/net/sfc/meson.build | 22 +++---
Old clanng requires libatomic as well as gcc. Avoid compiler name and
version based checks. Add custom test for 16-byte atomic operations
to find out if libatomic is required to build.
Fixes: 96fd2bd69b58 ("net/sfc: support flow action count in transfer rules")
Bugzilla ID: 760
Signed-off-by: And
Function check_tx_thresh is called with wrong parameter. If the
check fails, tx_queue_setup should return error not keep going.
iThis patch fixes above issues.
Fixes: 69dd4c3d0898 ("net/avf: enable queue and device")
Cc: sta...@dpdk.org
Signed-off-by: Xiaoyun Li
---
drivers/net/iavf/iavf_rxtx.c
From: Mohamad Noor Alim Hussin
Finding with "librt" keyword would give the output with full path of librt such
as /usr/lib/gcc/x86_64-linux-gnu/7/../../../x86_64-linux-gnu/librt.so
instead of -lrt in libdpdk.pc pkg-config file.
Assume find_library() will prepend "lib", thus remove "lib" from "li
Claim ownership for crypto API layer.
Have been reviewing patches from quite some time.
Signed-off-by: Akhil Goyal
---
MAINTAINERS | 1 +
1 file changed, 1 insertion(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index c68acbcd06..8b5a3f0249 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -420,6 +420,
Enabled build of Octeontx crypto PMD on non linux OS. Other Octeontx
PMDs are enabled already.
This is to avoid ABI test failure on an OS once we add dependency
between a driver which is built to another which is not.
Signed-off-by: Shijith Thotton
---
drivers/crypto/octeontx/meson.build | 4 --
On 7/20/21 7:25 PM, Ferruh Yigit wrote:
On 6/4/2021 3:42 PM, Andrew Rybchenko wrote:
From: Ivan Ilchenko
Document valid combinations of input arguments in accordance with
current implementation in ethdev.
Fixes: 79c913a42f0 ("ethdev: retrieve xstats by ID")
Cc: sta...@dpdk.org
Signed-off-by:
Hi Andrew,
On Thu, Jul 22, 2021 at 9:49 AM Andrew Rybchenko
wrote:
>
> Old clanng requires libatomic as well as gcc. Avoid compiler name and
s/nn/n/
> version based checks. Add custom test for 16-byte atomic operations
> to find out if libatomic is required to build.
>
> Bugzilla ID: 760
> Fi
> Enabled build of Octeontx crypto PMD on non linux OS. Other Octeontx
> PMDs are enabled already.
>
> This is to avoid ABI test failure on an OS once we add dependency
> between a driver which is built to another which is not.
Fixes: 8dc6c2f12ecf ("crypto/octeontx: add crypto adapter framework")
Hi Yaron
Freeing mbufs from a different lcore than the original lcore allocated them
causes cache miss in the mempool cache of the original lcore per mbuf
allocation - all the time the PMD will get non-hot mbufs to work with.
It can be one of the reasons for the earlier drops you see.
Matan
On 7/20/21 7:51 PM, Ferruh Yigit wrote:
On 6/4/2021 3:42 PM, Andrew Rybchenko wrote:
From: Ivan Ilchenko
Update xstats by IDs callbacks documentation in accordance with
ethdev usage of these callbacks. Document valid combinations of
input arguments to make driver implementation simpler.
Fixes
Rx/Tx doorbells stats are essential for performance investigation.
On the way fix ethdev documenation to refine requirements on
driver callback. It allows to make these callbacks a bit simpler.
Add testpmd option to show specified xstats periodically or upon
request, for example:
* --display-xs
From: Ivan Ilchenko
Add MAC stats lock in get xstats by id callback before reading
number of supported MAC stats.
Fixes: 73280c1e4ff ("net/sfc: support xstats retrieval by ID")
Cc: sta...@dpdk.org
Signed-off-by: Ivan Ilchenko
Signed-off-by: Andrew Rybchenko
Reviewed-by: Andy Moreton
---
dri
From: Ivan Ilchenko
Update MAC stats function reads adapter state with MAC stats locking
but without adapter locking. Add adapter locking before calling this
function and remove MAC stats locking since there's no point to have
it together with adapter locking. The second place MAC stats locking
i
From: Ivan Ilchenko
Document valid combinations of input arguments in accordance with
current implementation in ethdev.
Fixes: 79c913a42f0 ("ethdev: retrieve xstats by ID")
Cc: sta...@dpdk.org
Signed-off-by: Ivan Ilchenko
Signed-off-by: Andrew Rybchenko
Reviewed-by: Andy Moreton
---
lib/eth
From: Ivan Ilchenko
Update xstats by IDs callbacks documentation in accordance with
ethdev usage of these callbacks. Document valid combinations of
input arguments to make driver implementation simpler.
Fixes: 79c913a42f0 ("ethdev: retrieve xstats by ID")
Cc: sta...@dpdk.org
Signed-off-by: Ivan
From: Ivan Ilchenko
Fix xstats by ID callbacks according to ethdev usage.
Handle combinations of input arguments that are required by ethdev
and sanity check and reject other combinations on callback entry.
Fixes: 73280c1e4ff ("net/sfc: support xstats retrieval by ID")
Cc: sta...@dpdk.org
Signe
From: Ivan Ilchenko
Device may support only some MAC stats. Add mapping from ids to subset
of supported MAC stats for each port.
Fixes: 73280c1e4ff ("net/sfc: support xstats retrieval by ID")
Cc: sta...@dpdk.org
Signed-off-by: Ivan Ilchenko
Signed-off-by: Andrew Rybchenko
Reviewed-by: Andy Mo
From: Ivan Ilchenko
Fixes: 1caab2f1e68 ("net/sfc: add basic statistics")
Cc: sta...@dpdk.org
Signed-off-by: Ivan Ilchenko
Signed-off-by: Andrew Rybchenko
Reviewed-by: Andy Moreton
---
drivers/net/sfc/sfc.h| 2 +-
drivers/net/sfc/sfc_ethdev.c | 6 +++---
drivers/net/sfc/sfc_port.c
From: Ivan Ilchenko
There is no point to recalculate number of available xstats on
each request. The number is calculated once on device start
and may be returned on subsequent calls.
Signed-off-by: Ivan Ilchenko
Signed-off-by: Andrew Rybchenko
Reviewed-by: Andy Moreton
---
drivers/net/sfc/s
From: Ivan Ilchenko
Move getting MAC stats code that involves locking to separate functions
to simplify addition of new xstats.
Signed-off-by: Ivan Ilchenko
Signed-off-by: Andrew Rybchenko
Reviewed-by: Andy Moreton
---
drivers/net/sfc/sfc.h| 4 ++
drivers/net/sfc/sfc_ethdev.c | 73 +
From: Ivan Ilchenko
Rx/Tx doorbells statistics are collected in software and
available per queue. These stats are useful for performance
investigation.
Signed-off-by: Ivan Ilchenko
Signed-off-by: Andrew Rybchenko
Reviewed-by: Andy Moreton
---
drivers/net/sfc/meson.build| 1 +
drive
From: Ivan Ilchenko
Add 'display-xstats' option for using in accompanying with Rx/Tx statistics
(i.e. 'stats-period' option or 'show port stats' interactive command) to
display specified list of extended statistics.
Signed-off-by: Ivan Ilchenko
Signed-off-by: Andrew Rybchenko
---
app/test-pmd
On 7/22/2021 8:21 AM, Huisong Li wrote:
>
> 在 2021/7/21 23:29, Ferruh Yigit 写道:
>> On 7/19/2021 4:35 AM, Huisong Li wrote:
>>> Hi, Ferruh
>>>
>> Hi Huisong,
>>
>> Thanks for the review.
>>
>>> 在 2021/7/10 1:29, Ferruh Yigit 写道:
There is a confusion on setting max Rx packet length, this patch
On 7/22/21 1:12 PM, Ferruh Yigit wrote:
On 7/22/2021 8:21 AM, Huisong Li wrote:
在 2021/7/21 23:29, Ferruh Yigit 写道:
On 7/19/2021 4:35 AM, Huisong Li wrote:
Hi, Ferruh
Hi Huisong,
Thanks for the review.
在 2021/7/10 1:29, Ferruh Yigit 写道:
There is a confusion on setting max Rx packet leng
On 7/22/2021 2:31 AM, Ajit Khaparde wrote:
>
>
>
> > [snip]
> >
> >> diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
> >> index faf3bd901d75..9f288f98329c 100644
> >> --- a/lib/ethdev/rte_ethdev.h
> >> +++ b/lib/ethdev/rte_ethdev.h
> >> @@ -410,7 +410,7
On 7/21/21 6:55 PM, Martin Havlik wrote:
It is now clearly stated that RTE flow rules can be
created only after the port is started.
Signed-off-by: Martin Havlik
---
doc/guides/nics/mlx5.rst | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/doc/guides/nics/mlx5.rst b/d
On 7/21/21 9:16 PM, Stephen Hemminger wrote:
On Wed, 21 Jul 2021 17:58:14 +0200
Martin Havlik wrote:
The ability to create RTE flow rules, depending on
port status, can and does differ between PMDs.
Now the doc reflects that.
Signed-off-by: Martin Havlik
---
doc/guides/prog_guide/rte_flow.
Hi Matan
We work with mbuf in all threads and lcores,
We pass them from one thread to another through the dpdk ring before releasing
them.
There are drops in 10K to 100K pps, we can't stay with these drops.
The drops are in the imissed counter from rte_eth_stats_get, so I thought that
the drops
On 7/22/21 1:27 PM, Ferruh Yigit wrote:
On 7/22/2021 2:31 AM, Ajit Khaparde wrote:
> [snip]
>
>> diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
>> index faf3bd901d75..9f288f98329c 100644
>> --- a/lib/ethdev/rte_ethdev.h
>> +++ b/lib/ethdev/rte_eth
From: Chengchang Tang
This patch add support for dump the device registers from a running
application. It can help developers locate the problem.
Signed-off-by: Chengchang Tang
Signed-off-by: Min Hu (Connor)
---
v3:
* delete memset of dev_info.
v2:
* some logs are adjusted and error string ar
Currently, the flow indexed pools are allocated per port, the allocation
was missing in Windows code.
Allocate indexed pool for the Windows case too.
Fixes: b4edeaf3efd5 ("net/mlx5: replace flow list with indexed pool")
Signed-off-by: Suanming Mou
Acked-by: Tal Shnaiderman
Acked-by: Matan Azra
On 7/19/21 7:18 PM, Ferruh Yigit wrote:
On 7/19/2021 10:55 AM, Wang, Jie1X wrote:
-Original Message-
From: Yigit, Ferruh
Sent: Friday, July 16, 2021 4:52 PM
To: Li, Xiaoyun ; Wang, Jie1X ;
dev@dpdk.org
Cc: andrew.rybche...@oktetlabs.ru; sta...@dpdk.org
Subject: Re: [dpdk-stable] [PAT
2021-07-22 13:32 (UTC+0300), Andrew Rybchenko:
> On 7/21/21 6:55 PM, Martin Havlik wrote:
> > It is now clearly stated that RTE flow rules can be
> > created only after the port is started.
> >
> > Signed-off-by: Martin Havlik
> > ---
> > doc/guides/nics/mlx5.rst | 6 +-
> > 1 file changed
21/07/2021 10:54, Bing Zhao:
> When creating a meter policy, the actions for yellow color can be
> specified together with green color. The mlx5 PMD now supports to
> set the policy actions for yellow color.
>
> The actions list that is supported for yellow is the same as that
> for green.
>
> Ac
For CX6 in FDB domain, pop and push VLAN on both ingress and
egress directions are supported.
For CX6 in NIC domain, and CX5 in both FWD add NIC domain, pop
VLAN is only supported on ingress direction, push VLAN is only
supported on egress direction.
Signed-off-by: Dong Zhou
Acked-by: Matan Azra
01/07/2021 08:39, Michael Baum:
> The constant representing the size of the metadata is defined as a
> unsigned int variable with 32-bit.
> Similarly the constant representing the maximal output is also defined
> as a unsigned int variable with 32-bit.
>
> There is potentially overflowing expressi
> Michael Baum (3):
> regex/mlx5: fix memory region unregistration
> regex/mlx5: fix leak in PCI remove function
> regex/mlx5: fix redundancy in PCI remove function
Applied, thanks
22/07/2021 09:48, Dong Zhou:
> For CX6 in FDB domain, pop and push VLAN on both ingress and
> egress directions are supported.
>
> For CX6 in NIC domain, and CX5 in both FWD add NIC domain, pop
> VLAN is only supported on ingress direction, push VLAN is only
> supported on egress direction.
>
> S
Hi, all
I notice ethdev support dev_reset ops, which could be used to recover from
errors, and only 13+ drivers support this function.
And also there is event for reset: RTE_ETH_EVENT_INTR_RESET, and only 6
drivers support it (most of them are VF).
This provides users with two ways to
20/07/2021 09:53, Dmitry Kozlyuk:
> When device configuration was interrupted by a signal,
> mlx5_rxq/txq_release() could access yet unitinialized array
> and crash the application. Add checks whether queue array
> is initialized.
>
> Fixes: a1366b1a2be3 ("net/mlx5: add reference counter on DPDK R
21/07/2021 14:51, Dmitry Kozlyuk:
> mlx5_ind_table_obj_modify() first references queues from the new list,
> then applies the new list to HW. In case of apply failure the function
> dereferenced queues from the old list, while it should be the new list.
>
> Fixes: fa7ad49e96b5 ("net/mlx5: fix shar
10/07/2021 12:35, wangyunjian:
> From: Yunjian Wang
>
> This patch fixes the use-after-free bug which was reported by Coverity
> Scan in the mlx5_dma_unmap function.
>
> Coverity issue: 371679
> Fixes: 992e6df3dafe ("common/mlx5: free MR resource on device DMA unmap")
> Cc: sta...@dpdk.org
>
>
21/07/2021 10:43, Matan Azrad:
> Better title:
> net/mlx5/windows: fix indexed pools allocation
even better: keep the "on Windows" at the end.
22/07/2021 08:59, Suanming Mou:
> Currently, the flow indexed pools are allocated per port, the allocation
> was missing in Windows code.
>
> Allocate indexed pool for the Windows case too.
>
> Fixes: b4edeaf3efd5 ("net/mlx5: replace flow list with indexed pool")
>
> Signed-off-by: Suanming Mou
20/07/2021 09:51, Alexander Kozyrev:
> Register C is used in the extensive metadata mode number 1 and its
> width can vary from 0 to 32 bits depending on the kernel usage of it.
>
> There are several issues associated with this mode (dv_xmeta_en=1):
> 1. The metadata setting assumes that the width
16/07/2021 12:47, Slava Ovsiienko:
> > -Original Message-
> > From: Alexander Kozyrev
> > Sent: Friday, July 16, 2021 11:43
> > To: dev@dpdk.org
> > Cc: Raslan Darawsheh ; Matan Azrad
> > ; Slava Ovsiienko
> > Subject: [PATCH v2] net/mlx5: do not allow copy to mark via modify field
> >
>
On Thu, 22 Jul 2021 13:15:04 +0300
Andrew Rybchenko wrote:
> > I don't think we care about type of transmission in this level, I assume we
> > define min MTU mainly for the HW limitation and configuration. That is why
> > it
> > makes sense to me to use Ethernet frame lenght limitation (not IPv4
21/07/2021 10:31, Viacheslav Ovsiienko:
> The ROCE LAG bond device requires neither E-Switch nor SR-IOV
> configurations. It means the ROCE LAG bond device might be
> presented as a single port Infiniband device.
>
> The mlx5 PMD wrongly recognized standalone ROCE LAG bond device
> as E-Switch con
> > Representor failed to probe in isolated mode due to callback of retrieving
> > representor info missing. This patch adds it back.
> >
> > Fixes: cb95feefdd03 ("net/mlx5: support sub-function representor")
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: Xueming Li
> Acked-by: Viacheslav Ovsiie
18/07/2021 13:15, Lior Margalit:
> The flow did not expand correctly when it included a GTP item.
>
> Added GTP node to the expansion graph as possible next node
> after IPv4/IPv6 UDP node.
>
> Fixes: 592f05b29a25 ("net/mlx5: add RSS flow action")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Lior Ma
20/07/2021 17:17, Lior Margalit:
> The user is able to create a flow rule pattern with ETH after GTP
> although it is not supported by the flex-parser configuration.
>
> Failed the rule validation in such case with proper error message.
>
> Fixes: 23c1d42c7138 ("net/mlx5: split flow validation to
18/07/2021 12:29, Liang Ma:
> From: Liang Ma
>
> rte prefix header should be exported in meson.build
>
> Fixes: 23f627e0ed28 (net/mlx5: add flow sync API)
Fixes: efa79e68c8cd ("net/mlx5: support fine grain dynamic flag")
Cc: sta...@dpdk.org
> Signed-off-by: Liang Ma
Applied with some rewords
22/07/2021 15:50, fengchengwen:
> Hi, all
>
> I notice ethdev support dev_reset ops, which could be used to recover from
> errors, and only 13+ drivers support this function.
> And also there is event for reset: RTE_ETH_EVENT_INTR_RESET, and only 6
> drivers support it (most of them are VF
22/07/2021 07:07, Xia, Chenbo:
> From: Jiang, Cheng1
> > When the guest memory is hotplugged, the vhost application which
> > enables DMA acceleration must stop DMA transfers before the vhost
> > re-maps the guest memory.
> >
> > This patch set is to provide an unsafe API to drain inflight pkts
>
Setting MTU bigger than RTE_ETHER_MTU requires the jumbo frame support,
and application should enable the jumbo frame offload support for it.
When jumbo frame offload is not enabled by application, but MTU bigger
than RTE_ETHER_MTU is requested there are two options, either fail or
enable jumbo fr
Move requested MTU value check to the API to prevent the duplicated
code.
Signed-off-by: Ferruh Yigit
Reviewed-by: Andrew Rybchenko
Reviewed-by: Rosen Xu
---
drivers/net/axgbe/axgbe_ethdev.c| 15 ---
drivers/net/bnxt/bnxt_ethdev.c | 2 +-
drivers/net/cxgbe/cxgbe_e
Removing 'DEV_RX_OFFLOAD_JUMBO_FRAME' offload flag.
Instead of drivers announce this capability, application can deduct the
capability by checking reported 'dev_info.max_mtu' or
'dev_info.max_rx_pktlen'.
And instead of application explicitly set this flag to enable jumbo
frames, this can be deduc
Both 'rte_eth_dev_configure()' & 'rte_eth_dev_set_mtu()' sets MTU but
have slightly different checks. Like one checks min MTU against
RTE_ETHER_MIN_MTU and other RTE_ETHER_MIN_LEN.
Checks moved into common function to unify the checks. Also this has
benefit to have common error logs.
Suggested-by
Remove 'max-pkt-len' parameter.
Signed-off-by: Ferruh Yigit
---
examples/ip_reassembly/main.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/examples/ip_reassembly/main.c b/examples/ip_reassembly/main.c
index b92f0e460178..5c1b951c0d80 100644
--- a/examples/ip_reassembly/main.c
+++ b/examp
20/07/2021 15:36, Liang Ma:
> From: Liang Ma
>
> GCC 6.3.0 has a known bug which related to _mm512_extracti64x4_epi64.
> Please reference https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82887
>
> Some DPDK PMD avx512 version heavily use _mm512_extracti64x4_epi6,
> which cause building failure with
> -Original Message-
> From: Xueming(Steven) Li
> Sent: Wednesday, July 7, 2021 14:53
> Cc: dev@dpdk.org; Xueming(Steven) Li ;
> sta...@dpdk.org; Matan Azrad ; Shahaf Shuler
> ; Slava Ovsiienko
> Subject: [PATCH] net/mlx5: fix SubFunction representor probe in isolate
> mode
>
> Represent
22/07/2021 11:17, Akhil Goyal:
> > Enabled build of Octeontx crypto PMD on non linux OS. Other Octeontx
> > PMDs are enabled already.
> >
> > This is to avoid ABI test failure on an OS once we add dependency
> > between a driver which is built to another which is not.
>
> Fixes: 8dc6c2f12ecf ("cr
22/07/2021 21:06, Thomas Monjalon:
> 22/07/2021 11:17, Akhil Goyal:
> > > Enabled build of Octeontx crypto PMD on non linux OS. Other Octeontx
> > > PMDs are enabled already.
> > >
> > > This is to avoid ABI test failure on an OS once we add dependency
> > > between a driver which is built to anot
22/07/2021 10:37, Akhil Goyal:
> Claim ownership for crypto API layer.
> Have been reviewing patches from quite some time.
>
> Signed-off-by: Akhil Goyal
> ---
> Crypto API
> +M: Akhil Goyal
> M: Declan Doherty
> T: git://dpdk.org/next/dpdk-next-crypto
> F: lib/cryptodev/
Applied, thanks,
21/07/2021 02:17, Long Li:
> From: Stephen Hemminger
> >
> > Looks good, minor comment. You don't have to check for NULL before calling
> > rte_free().
> > Rte_free(NULL) is a NOP like free(NULL).
> >
> > Sorry for top posting; but if you send to my Microsoft account you are stuck
> > with what
21/07/2021 16:39, David Hunt:
> On 21/7/2021 3:26 PM, Anatoly Burakov wrote:
> > Currently in scale mode, multi-queue initialization will attempt to
> > initialize and de-initialize the per-lcore power library structures
> > multiple times. Fix it to only do this whenever we either enabling
> > fir
21/07/2021 11:27, Richael Zhuang:
> The freqs array size is RTE_MAX_LCORE_FREQS. Before filling the
> array with num_freqs elements, restrict the total num to
> RTE_MAX_LCORE_FREQS. This fix aims to fix the coverity scan issue
> like:
> Overrunning array "pi->freqs" of 256 bytes by passing it to a
On Thu, Jul 22, 2021 at 11:55 AM Andrew Rybchenko
wrote:
>
> From: Ivan Ilchenko
>
> Add 'display-xstats' option for using in accompanying with Rx/Tx statistics
> (i.e. 'stats-period' option or 'show port stats' interactive command) to
> display specified list of extended statistics.
>
> Signed-o
On Thu, Jul 22, 2021 at 11:12 AM David Marchand
wrote:
> On Thu, Jul 22, 2021 at 9:49 AM Andrew Rybchenko
> wrote:
> >
> > Old clang requires libatomic as well as gcc. Avoid compiler name and
> > version based checks. Add custom test for 16-byte atomic operations
> > to find out if libatomic is r
09/07/2021 17:19, Kevin Traynor:
> rte_stats_bitrate_reg() API states it returns 'Zero on success'.
>
> However, the implementation directly returns the return of
> rte_metrics_reg_names() which may be zero or positive on success,
> with a positive value also indicating the index.
>
> The user of
07/07/2021 11:14, Mohsin Kazmi:
> On Wed, Jun 30, 2021 at 3:09 PM Olivier Matz wrote:
> > > + if (ol_flags & (PKT_TX_OUTER_IPV4 | PKT_TX_OUTER_IPV6)) {
> > > inner_l3_offset += m->outer_l2_len + m->outer_l3_len;
> > > + /*
> > > + * prepare outer ipv4 hea
+Cc more people for reviews.
07/07/2021 11:40, Mohsin Kazmi:
> Preparation the headers for the hardware offload
> misses the outer ipv4 checksum offload.
> It results in bad checksum computed by hardware NIC.
>
> This patch fixes the issue by setting the outer ipv4
> checksum field to 0.
nit: pl
21/07/2021 17:28, Dmitry Kozlyuk:
> 2021-07-07 13:25 (UTC-0700), Jie Zhou:
> > eal/windows alarm APIs rte_eal_alarm_set and rte_eal_alarm_cancel
> > did not check parameters to fail fast for invalid parameters, which
> > caught by DPDK UT alarm_autotest.
> >
> > Enforce eal/windows alarm APIs para
On Thu, Jul 22, 2021 at 11:55 AM Andrew Rybchenko
wrote:
>
> diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
> index 8468018cf3..baffef1642 100644
> --- a/app/test-pmd/cmdline.c
> +++ b/app/test-pmd/cmdline.c
> @@ -3609,6 +3609,62 @@ cmdline_parse_inst_t cmd_stop = {
>
> /* *** SET C
On Wed, Jul 21, 2021 at 11:14 PM Thinh Tran wrote:
>
> A older version of complier would fail to generate code for new Power
compiler
> CPUs when it uses "-mcpu=native" argument.
> This patch will test if the compiler supports the current Power CPU type
> then proceeds with "-mcpu=native" argume
On Thu, Jul 22, 2021 at 3:08 PM Thomas Monjalon wrote:
>
> 22/07/2021 21:06, Thomas Monjalon:
> > 22/07/2021 11:17, Akhil Goyal:
> > > > Enabled build of Octeontx crypto PMD on non linux OS. Other Octeontx
> > > > PMDs are enabled already.
> > > >
> > > > This is to avoid ABI test failure on an OS
Release Dates
-
* v21.08
- Proposal/V1:Wednesday, 2 June (completed)
- rc1:Saturday, 10 July (completed)
- rc2:Friday,23 July
- rc3:Thursday, 29 July
- rc4:Wednesday, 4 August
- Release:Friday, 6 August
On 22/07/2021 20:46, Thomas Monjalon wrote:
> 09/07/2021 17:19, Kevin Traynor:
>> rte_stats_bitrate_reg() API states it returns 'Zero on success'.
>>
>> However, the implementation directly returns the return of
>> rte_metrics_reg_names() which may be zero or positive on success,
>> with a positive
15/07/2021 09:01, Asaf Penso:
> Hello DPDK community,
>
> I would like to bring up a discussion about a way to have code snippets as an
> example for proper usage.
> The DPDK tree is filled with great pieces of code that are well documented
> and maintained in high quality.
> I feel we are a bit
On Thu, Jul 8, 2021 at 1:00 PM John Levon wrote:
>
> get_hugepage_dir() was implemented in such a way that a --huge-dir
> option had to exactly match the mountpoint, but there's no reason for
> this restriction. Fix the implementation to allow a sub-directory within
> a suitable hugetlbfs mountpoi
22/07/2021 22:20, Brandon Lo:
> On Thu, Jul 22, 2021 at 3:08 PM Thomas Monjalon wrote:
> >
> > 22/07/2021 21:06, Thomas Monjalon:
> > > 22/07/2021 11:17, Akhil Goyal:
> > > > > Enabled build of Octeontx crypto PMD on non linux OS. Other Octeontx
> > > > > PMDs are enabled already.
> > > > >
> > >
On Thu, Jul 22, 2021 at 10:29:45PM +0200, David Marchand wrote:
> On Thu, Jul 8, 2021 at 1:00 PM John Levon wrote:
> >
> > get_hugepage_dir() was implemented in such a way that a --huge-dir
> > option had to exactly match the mountpoint, but there's no reason for
> > this restriction. Fix the imp
Can you tell what's the mbuf size? Max packet Len? MTU size?
Regards,
Asaf Penso
From: Balbeer Tiwari
Sent: Tuesday, July 20, 2021 6:17:50 PM
To: Asaf Penso ; dev@dpdk.org
Subject: RE: [dpdk-dev] DPDK Packet drop/Out of sequence issue with Jumbo
frames on MLX Co
The first argument to rte_bsf32_safe was incorrectly declared as
a 64 bit value. The code only works on 32 bit values and the underlying
function rte_bsf32 only accepts 32 bit values. This was a mistake
introduced when the safe version was added and probaly cause
by copy/paste from the 64 bit versi
> -Original Message-
> From: Thomas Monjalon
> Sent: Friday, July 23, 2021 3:43 AM
> To: Richael Zhuang
> Cc: dev@dpdk.org; David Hunt
> Subject: Re: [dpdk-dev] [PATCH v1 1/1] power: check freq count before filling
> the freqs array
>
> 21/07/2021 11:27, Richael Zhuang:
> > The freqs
v1:
add check for freq count
v2:
add "Fixes" tag in commit message
Richael Zhuang (1):
power: check freq count before filling the freqs array
lib/power/power_cppc_cpufreq.c | 5 +
lib/power/power_pstate_cpufreq.c | 5 +
2 files changed, 10 insertions(+)
--
2.20.1
The freqs array size is RTE_MAX_LCORE_FREQS. Before filling the
array with num_freqs elements, restrict the total num to
RTE_MAX_LCORE_FREQS. This fix aims to fix the coverity scan issue
like:
Overrunning array "pi->freqs" of 256 bytes by passing it to a
function which accesses it at byte offset 46
On 2021/7/22 23:46, Thomas Monjalon wrote:
> 22/07/2021 15:50, fengchengwen:
>> Hi, all
>>
>> I notice ethdev support dev_reset ops, which could be used to recover
>> from
>> errors, and only 13+ drivers support this function.
>> And also there is event for reset: RTE_ETH_EVENT_INTR_RESET,
v1:
add check for freq count
v2:
add "Fixes" tag in commit message
v3:
update commit message
Richael Zhuang (1):
power: check freq count before filling the freqs array
lib/power/power_cppc_cpufreq.c | 5 +
lib/power/power_pstate_cpufreq.c | 5 +
2 files changed, 10 insertions(+)
--
The freqs array size is RTE_MAX_LCORE_FREQS. Before filling the
array with num_freqs elements, restrict the total num to
RTE_MAX_LCORE_FREQS. This fix aims to fix the coverity scan issue
like:
Overrunning array "pi->freqs" of 256 bytes by passing it to a
function which accesses it at byte offset 46
> -Original Message-
> From: Li, Xiaoyun
> Sent: Thursday, July 22, 2021 3:56 PM
> To: dev@dpdk.org; Wu, Jingjing ; Xing, Beilei
>
> Cc: Li, Xiaoyun ; sta...@dpdk.org
> Subject: [PATCH] net/iavf: fix tx thresh check issue
>
> Function check_tx_thresh is called with wrong parameter. If
Fix drivers/net note error and do some optimization for
i40e NEON path.
Feifei Wang (4):
drivers/net: remove redundant phrases
drivers/net: fix note error for Rx vector
net/i40e: reorder Rx NEON code for better readability
net/i40e: change code order to reduce L1 cache misses
drivers/net
For the note of Rx vec path,when extract and record EOP bit, the code
note should be "as the count of dd bits doesn't care", remove the
redundant "count".
fm10k:
Fixes: 7092be8437bd ("fm10k: add vector Rx")
Cc: jing.d.c...@intel.com
i40e-altive:
Fixes: c3def6a8724c ("net/i40e: implement vector PM
1 - 100 of 106 matches
Mail list logo
