On Thu, Nov 05, 2020 at 12:15:49AM +, Ananyev, Konstantin wrote:
>
> Hi Olivier,
>
> > m->nb_seg must be reset on mbuf free whatever the value of m->next,
> > because it can happen that m->nb_seg is != 1. For instance in this
> > case:
> >
> > m1 = rte_pktmbuf_alloc(mp);
> > rte_pktmbuf
05/11/2020 00:42, Stephen Hemminger:
> Thanks for the work cleaning out some of the cruft in the mbuf to make space
> for dynamic fields.
>
> Are you going to breakup the hash union in mbuf?
> There is some bits to be gained there for dynamic fields.
>
> - Flow director should be deprecated/gon
> -Original Message-
> From: Lance Richardson
> Sent: Thursday, November 5, 2020 2:30 AM
> To: Juraj Linkeš
> Cc: tho...@monjalon.net; Bruce Richardson ;
> acon...@redhat.com; maicolgabr...@hotmail.com; dev@dpdk.org; Ruifeng
> Wang
> Subject: Re: [dpdk-dev] [PATCH v10 1/5] net/bnxt: add
On Wed, Nov 4, 2020 at 2:34 PM Rong, Leyi wrote:
>
>
> > -Original Message-
> > From: David Marchand
> > Sent: Wednesday, November 4, 2020 4:43 PM
> > To: Rong, Leyi
> > Cc: Zhang, Qi Z ; dev
> > Subject: Re: [PATCH] examples/l3fwd: enable multiple Tx queues on a lcore
> >
> > On Wed, N
04/11/2020 23:25, David Christensen:
> On 11/4/20 1:02 PM, Thomas Monjalon wrote:
> > 04/11/2020 22:00, David Christensen:
> >>
> >> On 11/4/20 11:43 AM, Thomas Monjalon wrote:
> Signed-off-by: David Christensen
> Acked-by: Anatoly Burakov
> ---
> -#ifdef VFIO_IOMMU_SPAPR_INFO
> -Original Message-
> From: Olivier Matz
> Sent: Thursday, November 5, 2020 1:04 AM
> To: dev@dpdk.org
> Cc: jer...@marvell.com; Ruifeng Wang ; Jan
> Viktorin ; David Christensen
> ; Bruce Richardson ;
> Konstantin Ananyev ; Honnappa
> Nagarahalli ; Phil Yang
> ; Gavin Hu ; Steve Capper
> -Original Message-
> From: Olivier Matz
> Sent: Thursday, November 5, 2020 1:04 AM
> To: dev@dpdk.org
> Cc: jer...@marvell.com; Ruifeng Wang ; Jan
> Viktorin ; David Christensen
> ; Bruce Richardson ;
> Konstantin Ananyev ; Honnappa
> Nagarahalli ; Phil Yang
> ; Gavin Hu ; Steve Capper
Acked-by: Jiayu Hu
> -Original Message-
> From: yang_y...@163.com
> Sent: Thursday, November 5, 2020 10:55 AM
> To: dev@dpdk.org
> Cc: Hu, Jiayu ; Ananyev, Konstantin
> ; tho...@monjalon.net;
> yangy...@inspur.com; yang_y...@163.com
> Subject: [PATCH v2] gro: fix an error when packet is
In fme_nios_spi_init(), a mutex is locked for protecting nios
initialization process, the mutex is only unlocked when process
is successful, it should also be unlocked when process fail.
Fixes: e41856b515ce ("raw/ifpga/base: enhance driver reliability in
multi-process")
Coverity issue: 363751
Si
In opae_adapter_destroy(), pointer "adapter" is not validated before
passing it to opae_adapter_shm_free() and opae_adapter_mutex_close()
which dereference it.
Fixes: e41856b515ce ("raw/ifpga/base: enhance driver reliability in
multi-process")
Coverity issue: 363752
Signed-off-by: Wei Huang
---
Two issues are introduced from previous patch, one resides in
ifpga_fme.c and the other resides in opae_hw_api.c. Below two
patches are provided to fix them separately.
Main changes from v2:
- Fix coding style issue
Wei Huang (2):
raw/ifpga/base: unlock mutex when nios init fail
raw/ifpga/ba
From: Yi Yang
For VxLAN packets, GRO will mistakenly reassemble them
if inner L3 is IPv6, inner L4 is TCP or UDP, and outer L3
is IPv4 because the value of IS_IPV4_VXLAN_TCP4/UDP4_PKT
is true for them.
This fix makes sure IS_IPV4_TCP_PKT, IS_IPV4_UDP_PKT,
IS_IPV4_VXLAN_TCP4_PKT and IS_IPV4_VXLA
In opae_adapter_destroy(), pointer "adapter" is not validated before
passing it to opae_adapter_shm_free() and opae_adapter_mutex_close()
which derefence it.
Fixes: e41856b515ce ("raw/ifpga/base: enhance driver reliability in
multi-process")
Coverity issue: 363752
Signed-off-by: Wei Huang
---
In fme_nios_spi_init(), a mutex is locked for protecting nios
initialization process, the mutex is only unlocked when process
is successful, it should also be unlocked when process fail.
Fixes: e41856b515ce ("raw/ifpga/base: enhance driver reliability in
multi-process")
Coverity issue: 363751
Si
Two issues are introduced from previous patch, one resides in
ifpga_fme.c and the other resides in opae_hw_api.c. Below two
patches are provided to fix them separately.
Wei Huang (2):
raw/ifpga/base: unlock mutex when nios init fail
raw/ifpga/base: check pointer adapter before derefence
dri
On Thursday, November 5, 2020 1:24 AM, Ferruh Yigit wrote:
> On 11/3/2020 11:08 PM, Thomas Monjalon wrote:
> > 27/10/2020 12:37, Ferruh Yigit:
> >> On 10/27/2020 2:39 AM, Jiawen Wu wrote:
> >>> On Monday, October 26, 2020 10:56 PM, Ferruh Yigit wrote:
> Series applied to dpdk-next-net/main, th
At 2020-11-05 09:41:13, "Hu, Jiayu" wrote:
>Hi Yi,
>
>> -Original Message-
>> From: yang_y...@163.com
>> Sent: Wednesday, November 4, 2020 12:23 PM
>> To: dev@dpdk.org
>> Cc: Hu, Jiayu ; Ananyev, Konstantin
>> ; tho...@monjalon.net;
>> yangy...@inspur.com; yang_y...@163.com
>> Subject: [P
Hi Yi,
> -Original Message-
> From: yang_y...@163.com
> Sent: Wednesday, November 4, 2020 12:23 PM
> To: dev@dpdk.org
> Cc: Hu, Jiayu ; Ananyev, Konstantin
> ; tho...@monjalon.net;
> yangy...@inspur.com; yang_y...@163.com
> Subject: [PATCH] gro: fix an error when packet is IPv6
>
> From:
>
> Hi,
>
> On Tue, Nov 03, 2020 at 04:03:46PM +0100, Morten Brørup wrote:
> > > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Slava Ovsiienko
> > > Sent: Tuesday, November 3, 2020 3:03 PM
> > >
> > > Hi, Morten
> > >
> > > > From: Morten Brørup
> > > > Sent: Tuesday, November 3, 2020
Hi Olivier,
> m->nb_seg must be reset on mbuf free whatever the value of m->next,
> because it can happen that m->nb_seg is != 1. For instance in this
> case:
>
> m1 = rte_pktmbuf_alloc(mp);
> rte_pktmbuf_append(m1, 500);
> m2 = rte_pktmbuf_alloc(mp);
> rte_pktmbuf_append(m2, 500);
>
Thanks for the work cleaning out some of the cruft in the mbuf to make space
for dynamic fields.
Are you going to breakup the hash union in mbuf?
There is some bits to be gained there for dynamic fields.
- Flow director should be deprecated/gone/killed so hash.fdir can disappear.
- QoS sched
On Wed, Sep 16, 2020 at 4:21 AM Juraj Linkeš wrote:
>
> From: Ruifeng Wang
>
> Expand vector PMD support to aarch32.
>
> Signed-off-by: Ruifeng Wang
> Acked-by: Ajit Khaparde
> ---
> drivers/net/bnxt/bnxt_rxq.h | 2 +-
> drivers/net/bnxt/bnxt_rxr.h | 2 +-
> drivers/net/bnxt/bnxt_txr.h | 2 +-
On 11/4/20 1:02 PM, Thomas Monjalon wrote:
04/11/2020 22:00, David Christensen:
On 11/4/20 11:43 AM, Thomas Monjalon wrote:
Signed-off-by: David Christensen
Acked-by: Anatoly Burakov
---
-#ifdef VFIO_IOMMU_SPAPR_INFO_DDW
- /* try possible page_shift and levels for workaround
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Tuesday, October 20, 2020 1:56 PM
>
> Hi Morten,
> Any update about this patch please?
Thomas,
v4 contains all the modifications suggested by Olivier:
https://patchwork.dpdk.org/patch/82198/
Except a depreciation notice about the old
A new DPDK release candidate is ready for testing:
https://git.dpdk.org/dpdk/tag/?id=v20.11-rc2
There are 628 new patches in this snapshot.
The good news is that -rc2 is smaller than -rc1 ;)
Release notes:
http://doc.dpdk.org/guides/rel_notes/release_20_11.html
Highlights of 20.1
> > >
> > > Each core already comes with its local storage for mcslock (in its
> > > stack), therefore there is no need to define an additional per-lcore
> mcslock.
> > >
> > > Fixes: 32dcb9fd2a22 ("test/mcslock: add MCS queued lock unit test")
> > >
> > > Signed-off-by: Olivier Matz
Reviewed-by
04/11/2020 18:22, Ciara Power:
> A "+" symbol was incorrectly placed at the beginning of a line,
> this is now removed.
>
> Fixes: 52af6ccb2b39 ("telemetry: add utility functions for creating JSON")
> Cc: bruce.richard...@intel.com
> Cc: sta...@dpdk.org
>
> Signed-off-by: Ciara Power
Applied, t
Hi Honnappa,
On Wed, Nov 04, 2020 at 05:57:19PM +, Honnappa Nagarahalli wrote:
>
>
> >
> > Each core already comes with its local storage for mcslock (in its stack),
> > therefore there is no need to define an additional per-lcore mcslock.
> >
> > Fixes: 32dcb9fd2a22 ("test/mcslock: add MC
04/11/2020 22:00, David Christensen:
>
> On 11/4/20 11:43 AM, Thomas Monjalon wrote:
> >> Signed-off-by: David Christensen
> >> Acked-by: Anatoly Burakov
> >> ---
> >> -#ifdef VFIO_IOMMU_SPAPR_INFO_DDW
> >> - /* try possible page_shift and levels for workaround */
> >> + /* if
On 11/4/20 11:43 AM, Thomas Monjalon wrote:
Signed-off-by: David Christensen
Acked-by: Anatoly Burakov
---
-#ifdef VFIO_IOMMU_SPAPR_INFO_DDW
- /* try possible page_shift and levels for workaround */
+ /* if at first we don't succeed, try more levels */
04/11/2020 09:04, Wisam Monther:
> Hi,
>
> You can add my ack:
>
> Acked-by: Wisam Jaddo mailto:wis...@nvidia.com>>
>
> Thomas,
> Do you have comments left here?
No comment, it looks very good.
Applied, thanks
On 11/4/2020 5:01 PM, Olivier Matz wrote:
Add a missing space.
Fixes: 869bf6d222bb ("net/ring: fix coding style")
Signed-off-by: Olivier Matz
Cc: sta...@dpdk.org
Reviewed-by: Ferruh Yigit
04/11/2020 21:19, Ferruh Yigit:
> On 11/4/2020 5:55 PM, Thomas Monjalon wrote:
> > 04/11/2020 18:07, Ferruh Yigit:
> >> On 11/4/2020 4:51 PM, Thomas Monjalon wrote:
> >>> 03/11/2020 14:29, Ferruh Yigit:
> On 11/2/2020 11:48 AM, Ferruh Yigit wrote:
> > On 11/2/2020 8:52 AM, SteveX Yang wrot
On 11/4/2020 5:55 PM, Thomas Monjalon wrote:
04/11/2020 18:07, Ferruh Yigit:
On 11/4/2020 4:51 PM, Thomas Monjalon wrote:
03/11/2020 14:29, Ferruh Yigit:
On 11/2/2020 11:48 AM, Ferruh Yigit wrote:
On 11/2/2020 8:52 AM, SteveX Yang wrote:
When the max rx packet length is smaller than the sum
04/11/2020 10:28, David Hunt:
> Remove notice announcing an already-implemented change.
>
> In 19.05, rte_power_set_env was changed to return -1 in cases where
> the envorinment was already set up, and for the same release, a
> deprecation notice was added for the following commit:
> 'commit 5a5f3
> > The typo "withe" should have been "with the". This is now fixed.
>
> Acked-by: John McNamara
Applied, thanks
03/11/2020 23:05, David Christensen:
> The SPAPR IOMMU requires that a DMA window size be defined before memory
> can be mapped for DMA. Current code dynamically modifies the DMA window
> size in response to every new memory allocation which is potentially
> dangerous because all existing mappings
On 04/11/2020 18:58, Dharmik Thakkar wrote:
Avoid code duplication by combining single and multi threaded tests
Also, enable support for more than 2 writers
Signed-off-by: Dharmik Thakkar
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
app/test/test_lpm_perf.c | 356
On 04/11/2020 18:58, Dharmik Thakkar wrote:
Remove redundant error checking for reader threads
since they never return error.
Fixes: eff30b59cc2e ("test/lpm: add RCU performance tests")
Cc: honnappa.nagaraha...@arm.com
Cc: sta...@dpdk.org
Signed-off-by: Dharmik Thakkar
Reviewed-by: Ruifeng
On 04/11/2020 18:58, Dharmik Thakkar wrote:
Return error if Add/Delete fail in multiwriter perf test
Return error if single or multi writer test fails
Fixes: eff30b59cc2e ("test/lpm: add RCU performance tests")
Cc: honnappa.nagaraha...@arm.com
Cc: sta...@dpdk.org
Signed-off-by: Dharmik Thak
On 04/11/2020 18:58, Dharmik Thakkar wrote:
Fix incorrect calculations for LPM adds, LPM deletes,
and average cycles in RCU QSBR perf tests
Since, rcu qsbr tests run for 'RCU_ITERATIONS' and not
'ITERATIONS', replace 'ITERATIONS' with 'RCU_ITERATIONS'
for calculating adds, deletes, and cycles
28/07/2020 07:58, Hemant Agrawal:
> This patch removes the dual keyword from dual license
> definitions to avoid confusion. As the *dual* word is
> not required to be added SPDX license.
>
> Signed-off-by: Hemant Agrawal
> ---
> Note that following licenses are not exceptions:-
> - BSD-3-C
04/11/2020 19:34, Thomas Monjalon:
> 29/10/2020 14:36, luca.bocca...@gmail.com:
> > From: Luca Boccassi
> >
> > Signed-off-by: Luca Boccassi
> > ---
> > Would be really great to hook up some spell checkers to the CI, so
> > that I'm not the only one who gets nagged by automated tools :-)
>
> Th
Remove redundant error checking for reader threads
since they never return error.
Fixes: eff30b59cc2e ("test/lpm: add RCU performance tests")
Cc: honnappa.nagaraha...@arm.com
Cc: sta...@dpdk.org
Signed-off-by: Dharmik Thakkar
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
app
Return error if Add/Delete fail in multiwriter perf test
Return error if single or multi writer test fails
Fixes: eff30b59cc2e ("test/lpm: add RCU performance tests")
Cc: honnappa.nagaraha...@arm.com
Cc: sta...@dpdk.org
Signed-off-by: Dharmik Thakkar
Reviewed-by: Ruifeng Wang
Reviewed-by: Honn
Fix incorrect calculations for LPM adds, LPM deletes,
and average cycles in RCU QSBR perf tests
Since, rcu qsbr tests run for 'RCU_ITERATIONS' and not
'ITERATIONS', replace 'ITERATIONS' with 'RCU_ITERATIONS'
for calculating adds, deletes, and cycles.
Also, for multi-writer perf test, each writer
Avoid code duplication by combining single and multi threaded tests
Also, enable support for more than 2 writers
Signed-off-by: Dharmik Thakkar
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
app/test/test_lpm_perf.c | 356 +--
1 file change
Fix LPM adds, LPM deletes, and cycle calculation.
Return error if LPM add/delete fails in multi-writer test.
Return error if single or multi writer test fails
Remove redundant error checking for readers.
Combine single and multi threaded test cases to avoid code duplication.
---
v5:
- Update logic
Hi Conor,
Thanks for the patch.
>
> When DPDK is compiled with gcc < 9 with the optimization level set to 1 gcc
> sees zcd in test_ring.h as possibly being uninitialised. To correct this
> error zcd
> has been initialised to {0} in test_ring_mt_peek_stress_zc.c.
>
> Signed-off-by: Con
29/10/2020 14:36, luca.bocca...@gmail.com:
> From: Luca Boccassi
>
> Signed-off-by: Luca Boccassi
> ---
> Would be really great to hook up some spell checkers to the CI, so
> that I'm not the only one who gets nagged by automated tools :-)
There is a spell checker with checkpatch
and we are cat
28/10/2020 10:59, Gregory Etelson:
> Representor is a port in DPDK that is connected to a VF in such a way
> that assuming there are no offload flows, each packet that is sent
> from the VF will be received by the corresponding representor. While
> each packet that is sent to a representor will be
03/11/2020 10:15, Matan Azrad:
> From: Tal Shnaiderman
> > The current DevX implementation of the relaxed ordering feature is enabling
> > relaxed ordering usage only if both relaxed ordering read AND write are
> > supported. In that case both relaxed ordering read and write are activated.
> >
>
>
> Each core already comes with its local storage for mcslock (in its stack),
> therefore there is no need to define an additional per-lcore mcslock.
>
> Fixes: 32dcb9fd2a22 ("test/mcslock: add MCS queued lock unit test")
>
> Signed-off-by: Olivier Matz
> ---
> app/test/test_mcslock.c | 16
04/11/2020 18:07, Ferruh Yigit:
> On 11/4/2020 4:51 PM, Thomas Monjalon wrote:
> > 03/11/2020 14:29, Ferruh Yigit:
> >> On 11/2/2020 11:48 AM, Ferruh Yigit wrote:
> >>> On 11/2/2020 8:52 AM, SteveX Yang wrote:
> When the max rx packet length is smaller than the sum of mtu size and
> ether
On 11/4/20 9:03 AM, Olivier Matz wrote:
Add missing arch-specific headers in meson.build.
Fixes: ca49b92079df ("ticketlock: enable generic ticketlock on all arch")
Cc: sta...@dpdk.org
Signed-off-by: Olivier Matz
---
lib/librte_eal/arm/include/meson.build | 1 +
lib/librte_eal/ppc/include
On Wed, Nov 4, 2020 at 5:27 PM Olivier Matz wrote:
> On Wed, Nov 04, 2020 at 08:20:00AM -0800, Stephen Hemminger wrote:
> > The dynamic flag management is broken if rte_mbuf_dynflag_lookup()
> > is done in a secondary process because the local pointer to
> > the memzone is not ever initialized.
>
On Wed, Nov 4, 2020 at 6:04 PM Olivier Matz wrote:
>
> Add missing arch-specific headers in meson.build.
>
> Fixes: ca49b92079df ("ticketlock: enable generic ticketlock on all arch")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Olivier Matz
Issue can be seen with:
diff --git a/examples/helloworld/m
On 11/3/2020 11:08 PM, Thomas Monjalon wrote:
27/10/2020 12:37, Ferruh Yigit:
On 10/27/2020 2:39 AM, Jiawen Wu wrote:
On Monday, October 26, 2020 10:56 PM, Ferruh Yigit wrote:
Series applied to dpdk-next-net/main, thanks.
57/58 & 58/58 has been distributed to the relevant commits while mergin
A "+" symbol was incorrectly placed at the beginning of a line,
this is now removed.
Fixes: 52af6ccb2b39 ("telemetry: add utility functions for creating JSON")
Cc: bruce.richard...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Ciara Power
---
app/test/test_telemetry_json.c | 2 +-
1 file changed
On 11/4/2020 4:51 PM, Thomas Monjalon wrote:
03/11/2020 14:29, Ferruh Yigit:
On 11/2/2020 11:48 AM, Ferruh Yigit wrote:
On 11/2/2020 8:52 AM, SteveX Yang wrote:
When the max rx packet length is smaller than the sum of mtu size and
ether overhead size, it should be enlarged, otherwise the VLAN
Hi Conor,
On 11/4/20 6:02 PM, Conor Walsh wrote:
> When DPDK is compiled with gcc < 9 with the optimization level set to 1
> gcc sees slot_idx as possibly being uninitialised. To correct this error
> slot_idx has been initialised to 0.
>
> Signed-off-by: Conor Walsh
> ---
> lib/librte_vhost/vir
When DPDK is compiled with gcc < 9 with the optimization level set to 1
gcc sees zcd in test_ring.h as possibly being uninitialised. To correct
this error zcd has been initialised to {0} in
test_ring_mt_peek_stress_zc.c.
Signed-off-by: Conor Walsh
---
app/test/test_ring_mt_peek_stress_zc.c | 4 +
Each core already comes with its local storage for mcslock (in its
stack), therefore there is no need to define an additional per-lcore
mcslock.
Fixes: 32dcb9fd2a22 ("test/mcslock: add MCS queued lock unit test")
Signed-off-by: Olivier Matz
---
app/test/test_mcslock.c | 16 ++--
1 f
Add missing arch-specific headers in meson.build.
Fixes: ca49b92079df ("ticketlock: enable generic ticketlock on all arch")
Cc: sta...@dpdk.org
Signed-off-by: Olivier Matz
---
lib/librte_eal/arm/include/meson.build | 1 +
lib/librte_eal/ppc/include/meson.build | 1 +
lib/librte_eal/x86/include/
Add missing arch-specific headers in meson.build.
Fixes: 2173fb61 ("mcslock: add MCS queued lock implementation")
Cc: sta...@dpdk.org
Signed-off-by: Olivier Matz
---
lib/librte_eal/arm/include/meson.build | 1 +
lib/librte_eal/ppc/include/meson.build | 1 +
lib/librte_eal/x86/include/meson.
Prefetching only helps performance if it is done several 100
instructions before the actual use. The purpose of the prefetch
is to read ahead, it doesn't help if the next instruction
will block.
The code in the bnxt driver was doing these unnecessary prefetches.
Signed-off-by: Stephen Hemminger
When DPDK is compiled with gcc < 9 with the optimization level set to 1
gcc sees slot_idx as possibly being uninitialised. To correct this error
slot_idx has been initialised to 0.
Signed-off-by: Conor Walsh
---
lib/librte_vhost/virtio_net.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Add a missing space.
Fixes: 869bf6d222bb ("net/ring: fix coding style")
Signed-off-by: Olivier Matz
---
drivers/net/ring/rte_eth_ring.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ring/rte_eth_ring.c b/drivers/net/ring/rte_eth_ring.c
index d6115abdd5..41402a5
m->nb_seg must be reset on mbuf free whatever the value of m->next,
because it can happen that m->nb_seg is != 1. For instance in this
case:
m1 = rte_pktmbuf_alloc(mp);
rte_pktmbuf_append(m1, 500);
m2 = rte_pktmbuf_alloc(mp);
rte_pktmbuf_append(m2, 500);
rte_pktmbuf_chain(m1, m2);
m0 =
On Tue, 3 Nov 2020 23:03:44 -0800
Stephen Hemminger wrote:
> This addresses several warnings from python3 lint.
> Split out from earlier patchset
>
> Stephen Hemminger (6):
> dpdk-devbind: use argparse instead of getopt
> dpdk-devbind: fix indentation
> dpdk-devbind: fix python lint warni
03/11/2020 14:29, Ferruh Yigit:
> On 11/2/2020 11:48 AM, Ferruh Yigit wrote:
> > On 11/2/2020 8:52 AM, SteveX Yang wrote:
> >> When the max rx packet length is smaller than the sum of mtu size and
> >> ether overhead size, it should be enlarged, otherwise the VLAN packets
> >> will be dropped.
> >>
> On Nov 4, 2020, at 9:46 AM, Medvedkin, Vladimir
> wrote:
>
> Hi Thakkar,
>
> On 03/11/2020 22:23, Dharmik Thakkar wrote:
>> Avoid code duplication by combining single and multi threaded tests
>> Also, enable support for more than 2 writers
>> Signed-off-by: Dharmik Thakkar
>> Reviewed-by:
On Wed, Nov 4, 2020 at 5:36 PM Bruce Richardson
wrote:
>
> On Wed, Nov 04, 2020 at 04:57:21PM +0100, David Marchand wrote:
> > This script inspects an ELF file (binary or shared library) and its
> > linked dependencies by following DT_NEEDED tags.
> > So far a simple librte_pmd prefix was used as
>
> 22/10/2020 14:24, Andrew Rybchenko:
> > Intel SSE has __m128i, but ARMv8 has __uint128_t. So, add compat
> > efsys_uint128_t to be used in driver source and have either __u128i or
> > __uint128_t behind.
>
> Could we do something in EAL for this common need?
rte_int128_t is defined in EAL.
On Wed, Nov 04, 2020 at 04:57:21PM +0100, David Marchand wrote:
> This script inspects an ELF file (binary or shared library) and its
> linked dependencies by following DT_NEEDED tags.
> So far a simple librte_pmd prefix was used as a filter to only parse
> DPDK drivers dependencies.
> While the re
22/10/2020 14:24, Andrew Rybchenko:
> Intel SSE has __m128i, but ARMv8 has __uint128_t. So, add compat
> efsys_uint128_t to be used in driver source and have either __u128i
> or __uint128_t behind.
Could we do something in EAL for this common need?
On Wed, Nov 04, 2020 at 08:20:00AM -0800, Stephen Hemminger wrote:
> The dynamic flag management is broken if rte_mbuf_dynflag_lookup()
> is done in a secondary process because the local pointer to
> the memzone is not ever initialized.
>
> Fix it by using the same checks as dynfield_register().
>
On Wed, 4 Nov 2020 09:21:05 +
Bruce Richardson wrote:
> On Tue, Nov 03, 2020 at 10:53:04PM -0800, Stephen Hemminger wrote:
> > Rearrange code to make it pass python lint totally clean! This includes
> > add a main function, docstring, and some variable name changes.
> >
> > Signed-off-by: S
The dynamic flag management is broken if rte_mbuf_dynflag_lookup()
is done in a secondary process because the local pointer to
the memzone is not ever initialized.
Fix it by using the same checks as dynfield_register().
I.e if shared memory zone has not been looked up already,
then discover it.
F
04/11/2020 16:52, Hemant Agrawal:
> Hi Thomas,
>Can you apply the following trivial patch to 20.11?
>
> https://patches.dpdk.org/patch/74896/
Yes, added in my list for -rc2.
This script inspects an ELF file (binary or shared library) and its
linked dependencies by following DT_NEEDED tags.
So far a simple librte_pmd prefix was used as a filter to only parse
DPDK drivers dependencies.
While the reason is not clear from the commitlog of the patch that
introduced this fil
Hi Thomas,
Can you apply the following trivial patch to 20.11?
https://patches.dpdk.org/patch/74896/
Regards,
Hemant
Hi Thakkar,
On 03/11/2020 22:23, Dharmik Thakkar wrote:
Avoid code duplication by combining single and multi threaded tests
Also, enable support for more than 2 writers
Signed-off-by: Dharmik Thakkar
Reviewed-by: Ruifeng Wang
---
app/test/test_lpm_perf.c | 359 +
Hi,
On Tue, Nov 03, 2020 at 04:03:46PM +0100, Morten Brørup wrote:
> > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Slava Ovsiienko
> > Sent: Tuesday, November 3, 2020 3:03 PM
> >
> > Hi, Morten
> >
> > > From: Morten Brørup
> > > Sent: Tuesday, November 3, 2020 14:10
> > >
> > > > From
Recent patch uses a local string array as input for function
rte_flow_error_set().
This stack memory may be later used by other code sections,
overwriting the desired error string.
This patch impelemnets an error string for the specific case
requested, of ICMP item not supported in Verbs flow engi
On 11/4/2020 2:03 AM, Guoyang Zhou wrote:
This patch fixes can support two or more aeqs from chip
config file and also support four aeqs before, and fix log
info for pf command channel when something is error.
--
v1:
- support two or more aeqs for chip
- fix log info for pf command channel
On 11/4/2020 11:46 AM, Ferruh Yigit wrote:
On 11/4/2020 3:26 AM, Ajit Khaparde wrote:
On Tue, Nov 3, 2020 at 6:58 PM wangyunjian wrote:
From: Yunjian Wang
The ethdev port id is 16 bits now. This patch fixes the data type
of the variable for 'pid', which changing from uint32_t to uint16_t.
On 11/4/2020 1:28 PM, Matan Azrad wrote:
From: Ferruh Yigit
On 11/3/2020 7:33 AM, Matan Azrad wrote:
Hi Ferruh
Thank you for the fast review.
Please see inline
From: Ferruh Yigit
On 11/1/2020 5:48 PM, Matan Azrad wrote:
When an age action becomes aged-out the rte_flow_get_aged_flows
sho
mlx5 PMD splited the sampling flow into prefix flow and suffix
flow. On the sample action translation function, the scaled
group value of suffix flow be attached into sample object and
saved into sample resource.
mlx5 PMD fetched the group value from the sample resource to
create the suffix flow.
From: Ferruh Yigit
> On 11/3/2020 7:33 AM, Matan Azrad wrote:
> > Hi Ferruh
> >
> > Thank you for the fast review.
> > Please see inline
> >
> > From: Ferruh Yigit
> >> On 11/1/2020 5:48 PM, Matan Azrad wrote:
> >>> When an age action becomes aged-out the rte_flow_get_aged_flows
> >>> should ret
From: Andrey Vesnovaty
Attach SFT flow context to packet with SFT action.
Match on SFT flow context (attached to packet),
with SFT item.
Signed-off-by: Andrey Vesnovaty
---
lib/librte_ethdev/rte_flow.h | 75
1 file changed, 75 insertions(+)
diff --git a/li
The RFC introduces Stateful Flow Table (SFT) API and changes needed in
both ethdev an RTE flow to support SFT functionality.
SFT library provides a framework for applications that need to maintain
context across different packets of the connection.
The goals of the SFT library:
- Accelerate flow
Defines RTE SFT (Stateful Flow Table) APIs for Stateful Flow Table library.
Currently, DPDK enables only stateless offloading, using the rte_flow.
stateless means that each packet is handled without any knowledge of
previous or future packets.
As we look at the industry, there is much demand to s
Defines RTE SFT (Stateful Flow Table) APIs for Stateful Flow Table library.
Currently, DPDK enables only stateless offloading, using the rte_flow.
stateless means that each packet is handled without any knowledge of
privious or future packets.
As we look at the industry, there is much demand to s
The RFC introduces Stateful Flow Table (SFT) API and changes needed in
both ethdev an RTE flow to support SFT functionality.
SFT library provides a framework for applications that need to maintain
context across different packets of the connection.
The goals of the SFT library:
- Accelerate flow
From: Andrey Vesnovaty
Attach SFT flow context to packet with SFT action.
Match on SFT flow context (attached to packet),
with SFT item.
Signed-off-by: Andrey Vesnovaty
---
lib/librte_ethdev/rte_flow.h | 75
1 file changed, 75 insertions(+)
diff --git a/li
On 11/3/2020 7:33 AM, Matan Azrad wrote:
Hi Ferruh
Thank you for the fast review.
Please see inline
From: Ferruh Yigit
On 11/1/2020 5:48 PM, Matan Azrad wrote:
When an age action becomes aged-out the rte_flow_get_aged_flows should
return the action context supplied by the configuration struct
On 11/4/2020 3:26 AM, Ajit Khaparde wrote:
On Tue, Nov 3, 2020 at 6:58 PM wangyunjian wrote:
From: Yunjian Wang
The ethdev port id is 16 bits now. This patch fixes the data type
of the variable for 'pid', which changing from uint32_t to uint16_t.
RTE_MAX_ETHPORTS is the maximum number of po
I think Thomas can add it if he want to merge it in this version.
Moreover I think you need to keep the ack between versions and the person who
acked before should comment if newer version not ok with him, otherwise he is
ok with it. “Using git commit —amend”.
Thomas, correct me if I’m wrong, A
On 11/4/2020 11:11 AM, Ferruh Yigit wrote:
On 11/4/2020 1:24 AM, wangyunjian wrote:
From: Yunjian Wang
Coverity flags that 'rx_conf' variable is used before
it's checked for NULL. This patch fixes this issue.
Coverity issue: 363570
Fixes: 4ff702b5dfa9 ("ethdev: introduce Rx buffer split")
Si
1 - 100 of 134 matches
Mail list logo
