>
>
> > > >
> > > > > > +
> > > > > > +/* Add a reader thread, running on an lcore, to the list of
> > > > > > +threads
> > > > > > + * reporting their quiescent state on a TQS variable.
> > > > > > + */
> > > > > > +int __rte_experimental
> > > > > > +rte_tqs_register_lcore(struct rte_tqs *v, un
On Tuesday 11 December 2018 06:44 PM, Harman Kalra wrote:
> segmentation fault ocured as vdev->device.driver->name did not
> return correct value.
> Test2 failed as dummy_value was freed before it was used.
>
> Signed-off-by: Kallio Marko
> Signed-off-by: Harman Kalra
> ---
> drivers/raw/skele
Hi Vipin,
> -Original Message-
> From: Varghese, Vipin
> Sent: Thursday, December 13, 2018 2:02 PM
> To: Lu, Wenzhuo ; dev@dpdk.org
> Cc: Lu, Wenzhuo
> Subject: RE: [dpdk-dev] [PATCH v3 00/34] A new net PMD - ice
>
> Hi,
>
> > -Original Message-
> > From: dev On Behalf Of Wenz
Before this patch, VF gets a default number of queues from the PF.
This patch enables VF to request a different number. When VF configures
more queues, it will send VIRTCHNL_OP_REQUEST_QUEUES to PF to request
more queues, if success, PF will reset the VF.
User can run "port stop all", "port config
This patch respond the VIRTCHNL_OP_REQUEST_QUEUES msg from VF, and
process to allocated more queues for the requested VF. If successful,
PF will notify VF to reset. If unsuccessful, PF will send message to
inform VF.
Signed-off-by: Zhirun Yan
Signed-off-by: Haiyue Wang
---
drivers/net/i40e/i40e
DPDK VF send VIRTCHNL_OP_REQUEST_QUEUES to kernel PF or DPDK VF for
requesting more queues, then PF will allocate more queues.
V3:
- remove vf->reset and vf->pend_msg flag
Zhirun Yan (2):
net/i40e: support VF request more queues
net/i40e: support PF respond VF request more queues
drivers/n
Hi,
> -Original Message-
> From: dev On Behalf Of Wenzhuo Lu
> Sent: Wednesday, December 12, 2018 12:30 PM
> To: dev@dpdk.org
> Cc: Lu, Wenzhuo
> Subject: [dpdk-dev] [PATCH v3 00/34] A new net PMD - ice
>
> This patch set adds the support of a new net PMD, Intel® Ethernet Network
> Adap
On Thu, 13 Dec 2018 11:37:43 +0800
Joyce Kong wrote:
> v1: reimplement rwlock with __atomic builtins, and add a rwlock perf test
> on all available cores to benchmark the improvement.
>
> We tested the patches on three arm64 platforms, ThundeX2 gained 20%
> performance,
> Qualcomm gained 36
Hi Ferruh,
> -Original Message-
> From: Yigit, Ferruh
> Sent: Wednesday, December 12, 2018 11:19 PM
> To: Lu, Wenzhuo ; dev@dpdk.org
> Cc: Stillwell Jr, Paul M
> Subject: Re: [dpdk-dev] [PATCH v3 02/34] net/ice: Add basic structures
>
> On 12/12/2018 6:59 AM, Wenzhuo Lu wrote:
> > From:
Adding to mailing list for the last update
> -Original Message-
> From: Lu, Wenzhuo
> Sent: Thursday, December 6, 2018 1:19 PM
> To: Zhang, Qi Z ; Varghese, Vipin
>
> Cc: Yigit, Ferruh ; Zhang, Helin
> ; Xu, Qian Q
> Subject: RE: [dpdk-dev] [PATCH v2 02/20] net/ice: support device
> ini
Before this patch, there are two functions will call
i40e_dev_free_queues to free queues. For rte_eth_dev_close(), its
redundant because of duplication. For rte_eth_dev_reset() its
redundant because of not necessary, since following dev_configure
is required after dev_reset and it will be updated c
Function show_mempool is used for displaying valid MEMPOOL.
Function iter_mempool is used for iterating mempool elements for
a mempool for max of 256 bytes.
In case of show_mempool for invalid name, whole list is dump.
Signed-off-by: Vipin Varghese
---
V6:
- split iter mempool - Vipin Varghese
Function show_ring is used for displaying the RING of the
primary process.
Signed-off-by: Vipin Varghese
---
V3:
- replace space to tab in printf - Reshma Pathan
- change ring display information - Vipin Varghese
---
app/proc-info/main.c | 34 +-
1 file changed
Document update for debug options and information for PMD instances
like port, traffic manager, crypto, mempool and ring instances.
Signed-off-by: Vipin Varghese
Acked-by: John McNamara
---
V6:
- add ack for v5 - John Mcnamara
- update entry for mempool iter - Vipin Varghese
V3:
- update d
Function show_port is used for displaying the port PMD information under
under primary process. The information shows basic, per queue and security.
Signed-off-by: Vipin Varghese
---
V5:
- convert 100 to MAX_STRING_LEN - Reshma Pattan
- RSS made for port - Reshma Pattan
- update bdr_str to st
Function show_tm is used for displaying the tm PMD under the
primary process. This covers basic and per node|level details with
stats.
Signed-off-by: Vipin Varghese
---
V6:
- replace or with logical or - Reshma Pathan & Konstantin Ananyev
V3:
- memset for struct elements - Vipin Varghese
- c
Function show_crypto is used for displaying the crypto PMD under
the primary process.
Signed-off-by: Vipin Varghese
---
v4:
- add space to compare - Vipin Varghese
V3:
- replace MACRO to function - Reshma Pathan & Stephen Hemminger
- add memset for struct elements - Reshma Pathan
- change d
Update the file with MACRO for stats border, usage text information
and string comparision.
Signed-off-by: Vipin Varghese
---
V7:
- update usage text with token ":" - Reshma Pattan
- reword usage for 'iter-mempool' - Reshma Pattan
V6:
- add mempool iterate elements option - Vipin Varghese
V3
Add code for new debug options to compare usage strings and set
enable flag.
Signed-off-by: Vipin Varghese
---
V6:
- check for iter mempool - Vipin Varghese
V3:
- variables from debug to show - Vipin Varghese
V2:
- compare string from dbg to show - Stephen Hemminger
---
app/proc-info/main.
Add prototype function calls for the show functions.
Signed-off-by: Vipin Varghese
---
V6:
- add call to iter mempool - Vipin Varghese
V3:
- update function names from debug to show - Vipin Varghese
V2:
- removed if else ladder - Vipin Varghese
---
app/proc-info/main.c | 50 +++
This patch adds new debug functions to existing proc-info tool.
Motivation
==
DPDK proc-info tool is been widely used as secondary process to collect
stats for any primary process. But these are limited to DPDK NIC ports and
basic memory usage.
Motivation
==
This patch tries to
Before this patch, there are two functions will call
i40e_dev_free_queues to free queues. For rte_eth_dev_close(), its
redundant because of duplication. For rte_eth_dev_reset() its
redundant because of not necessary, since following dev_configure
is required after dev_reset and it will be updated c
On Tue, Dec 11, 2018 at 04:43:31PM +, Anatoly Burakov wrote:
> When running in no-huge mode, we anonymously allocate our memory.
> While this works for regular NICs and vdev's, it's not suitable
> for memory sharing scenarios such as virtio with vhost_user
> backend.
>
> To fix this, allocate
On Tue, Dec 11, 2018 at 04:43:27PM +, Anatoly Burakov wrote:
> It is already possible to use both DPDK in general and
> virtio specifically, without hugetlbfs mounts, but
> currently virtio cannot be used without hugepage memory
> (i.e. with a --no-huge EAL switch) due to the fact that
> it nee
From: Gavin Hu
The __sync builtin based implementation generates full memory barriers
('dmb ish') on Arm platforms. Using C11 atomic builtins to generate one
way barriers.
Here is the assembly code of __sync_compare_and_swap builtin.
__sync_bool_compare_and_swap(dst, exp, src);
0x0090
v1: reimplement rwlock with __atomic builtins, and add a rwlock perf test
on all available cores to benchmark the improvement.
We tested the patches on three arm64 platforms, ThundeX2 gained 20% performance,
Qualcomm gained 36% and the 4-Cortex-A72 Marvell MACCHIATObin gained 19.6%.
Below is t
Add performance test on all available cores to benchmark
the scaling up performance and fairness of rw_lock.
Fixes: af75078faf ("first public release")
Cc: sta...@dpdk.org
Suggested-by: Gavin Hu
Signed-off-by: Joyce Kong
Reviewed-by: Honnappa Nagarahalli
Reviewed-by: Ola Liljedahl
Reviewed-by
Hi Ferruh,
> >
> > > +# this lib depends upon:
> > > +DEPDIRS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += lib/librte_eal
> > > +lib/librte_ether
> > > +DEPDIRS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += lib/librte_mempool
> > > +lib/librte_mbuf
> > > +DEPDIRS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += lib/librte_net
> > > +DEPDIRS-
Hi Ferruh,
> -Original Message-
> From: Yigit, Ferruh
> Sent: Thursday, December 13, 2018 2:18 AM
> To: Lu, Wenzhuo ; dev@dpdk.org
> Cc: Yang, Qiming ; Li, Xiaoyun
> ; Wu, Jingjing
> Subject: Re: [dpdk-dev] [PATCH v3 16/34] net/ice: support device
> initialization
>
> On 12/12/2018 6:59
Hi Reshma,
snipped
>
> Since this patch displaying mepool and its elements both , edit heading to
> suit
> both?
>
> ( or)
>
> What about having iter mempool support as separate patch after this patch?
> That looks more logical as iter-mempool is separate command option now.
This is good sug
Hi Mattias,
> -Original Message-
> From: Mattias Rönnblom [mailto:mattias.ronnb...@ericsson.com]
> Sent: Thursday, December 13, 2018 4:07 AM
> To: Lu, Wenzhuo ; dev@dpdk.org
> Cc: Yang, Qiming ; Li, Xiaoyun
> ; Wu, Jingjing
> Subject: Re: [dpdk-dev] [PATCH v3 17/34] net/ice: support devi
Hi Mattias,
> -Original Message-
> From: Stillwell Jr, Paul M
> Sent: Thursday, December 13, 2018 5:18 AM
> To: Mattias Rönnblom ; Lu, Wenzhuo
> ; dev@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH v3 11/34] net/ice: Add common functions
>
> -Original Message-
> From: Mattias Rönnblom
Vhost lib has already provided a helper for used ring logging, driver
could use it to reduce code.
Signed-off-by: Xiao Wang
---
drivers/net/ifc/ifcvf_vdpa.c | 27 ---
1 file changed, 8 insertions(+), 19 deletions(-)
diff --git a/drivers/net/ifc/ifcvf_vdpa.c b/drivers/net
In SW assisted live migration mode, driver will stop the device and
setup a mediate virtio ring to relay the communication between the
virtio driver and the VDPA device.
This data path intervention will allow SW to help on guest dirty page
logging for live migration.
This SW fallback is event dri
Signed-off-by: Xiao Wang
---
v2:
* Add release note.
---
doc/guides/nics/ifc.rst| 7 +++
doc/guides/rel_notes/release_19_02.rst | 5 +
2 files changed, 12 insertions(+)
diff --git a/doc/guides/nics/ifc.rst b/doc/guides/nics/ifc.rst
index 48f9adf1d..858f35f74 100644
--- a/
Driver probe may fail for different causes, debug message is helpful for
debugging issue.
Signed-off-by: Xiao Wang
---
drivers/net/ifc/ifcvf_vdpa.c | 19 +--
1 file changed, 13 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ifc/ifcvf_vdpa.c b/drivers/net/ifc/ifcvf_vdpa.c
If user wants the VF to be used in VDPA (vhost data path acceleration)
mode, then the user can add a "vdpa=1" parameter for the device.
So if driver doesn't not find this option, it should quit and let the
bus continue the probe.
Signed-off-by: Xiao Wang
---
drivers/net/ifc/ifcvf_vdpa.c | 47 ++
If driver fails to register ifc VF device into vhost lib, then this
device should not be stored.
Fixes: a3f8150eac6d ("net/ifcvf: add ifcvf vDPA driver")
cc: sta...@dpdk.org
Signed-off-by: Xiao Wang
---
drivers/net/ifc/ifcvf_vdpa.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
This patch series enables a new method for live migration, i.e. software
assisted live migration. This patch provides a device argument for user
to choose the methold.
When "swlm=1", driver/device will do live migration with a relay thread
dealing with dirty page logging. Without this parameter, d
This patch provides two helpers for vdpa device driver to perform a
relay between the guest virtio ring and a mediate virtio ring.
The available ring relay will synchronize the available entries, and
helps to do desc validity checking.
The used ring relay will synchronize the used entries from me
VDPA driver can decide if it needs to enable/disable the host notifier
mapping, so exposing a API can allow flexibility. A later patch will
base on this.
Signed-off-by: Xiao Wang
---
v2:
* Reword the vdpa host notifier control API comment.
---
drivers/net/ifc/ifcvf_vdpa.c | 3 +++
lib
In the previous VDPA implementation we have enabled live migration support
by HW accelerator doing all the stuff, including dirty page logging and
device status report/restore. In this mode VDPA sample daemon and device
driver just takes care of the control path and does not involve in data
path, s
Hi Shally,
I'm implementing a crypto asymmetric PMD and have some concerns about the API
which I
will work through over the next few months. Starting with modexp and modinv I
have
the following questions / suggestions:
rte_crypto_asym.h:233
rte_crypto_
On 2018-12-12 07:59, Wenzhuo Lu wrote:
/../
+
+ for (i = 0; i < len * sizeof(union ice_rx_desc); i++)
+ ((volatile char *)rxq->rx_ring)[i] = 0;
More of a general question... but why doesn't DPDK has the
READ/WRITE_ONCE() of the Linux kernel? Would reduce the amount of
op
NULL algo algo does not to set counter flag so it should be zeroed.
Fixes: db0e952a5c01 ("crypto/qat: add NULL capability")
Cc: sta...@dpdk.org
Signed-off-by: Arek Kusztal
---
drivers/crypto/qat/qat_sym_session.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/crypto/qat/qat_sym_ses
This patchset fixes invalid message which is caused by not
entirely correct way of checking for block size.
Arek Kusztal (3):
crypto/qat: handle error msg of block size properly
crypto/qat: fix message for CCM when setting unused counter
crypto/qat: fix message for NULL algo setting unused c
AES-CCM algo does not to set counter flag so it should be zeroed.
Fixes: ab56c4d9ed9a ("crypto/qat: support AES-CCM")
Cc: sta...@dpdk.org
Signed-off-by: Arek Kusztal
---
drivers/crypto/qat/qat_sym_session.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/crypto/qat/qat_sym_session.c
Error code of qat_hash_get_block_size needs to be handle properly.
Fixes: 10b49880e3c5 ("crypto/qat: make the session struct variable in size")
Cc: sta...@dpdk.org
Signed-off-by: Arek Kusztal
---
drivers/crypto/qat/qat_sym_session.c | 14 +-
1 file changed, 9 insertions(+), 5 deleti
On 2018-12-12 07:59, Wenzhuo Lu wrote:
From: Paul M Stillwell Jr
Add code that multiple other features use.
Signed-off-by: Paul M Stillwell Jr
---
drivers/net/ice/base/ice_common.c | 3521 +
drivers/net/ice/base/ice_common.h | 186 ++
2 files changed,
On 2018-12-12 07:47, Gavin Hu wrote:
Within rte_hash_reset, calling a while loop to dequeue one by
one from the ring, while not using them at all, is wasting cycles,
The patch just flush the ring by resetting the indices can save cpu
cycles.
Fixes: b26473ff8f4a ("hash: add reset function")
Fixes
On Wed, Dec 12, 2018 at 05:34:31PM +0100, Maxime Coquelin wrote:
> Hi Ilya,
>
> On 12/12/18 4:23 PM, Ilya Maximets wrote:
> > On 12.12.2018 11:24, Maxime Coquelin wrote:
> > > Instead of writing back descriptors chains in order, let's
> > > write the first chain flags last in order to improve batc
Hi Reshma,
More comments on this patch.
> diff --git a/lib/librte_pipeline/rte_table_action.c
> b/lib/librte_pipeline/rte_table_action.c
> index 7c7c8dd82..f9768b9cc 100644
> --- a/lib/librte_pipeline/rte_table_action.c
> +++ b/lib/librte_pipeline/rte_table_action.c
> @@ -107,14 +107,6 @@ mtr_c
On 12/12/2018 6:59 AM, Wenzhuo Lu wrote:
> Signed-off-by: Wenzhuo Lu
> Signed-off-by: Qiming Yang
> Signed-off-by: Xiaoyun Li
> Signed-off-by: Jingjing Wu
<...>
> @@ -297,6 +297,15 @@ CONFIG_RTE_LIBRTE_FM10K_RX_OLFLAGS_ENABLE=y
> CONFIG_RTE_LIBRTE_FM10K_INC_VECTOR=y
>
> #
> +# Compile bur
Only some minor some cosmetic questions
On 06/12/2018 3:38 PM, Konstantin Ananyev wrote:
Provide implementation for rte_ipsec_pkt_crypto_prepare() and
...
+/*
+ * Move preceding (L3) headers up to free space for ESP header and IV.
+ */
+static inline void
+insert_esph(char *np, char *op, uint
Meeting notes for the DPDK technical board meeting
held on 2018-11-21
Attendees:
- Bruce Richardson
- Ferruh Yigit
- Hemant Agrawal
- Konstantin Ananyev
- Maxime Coquelin
- Olivier Matz
- Thomas Monjalon
0) PPC support status
- Power 8 is
On 12/12/2018 4:54 PM, Stillwell Jr, Paul M wrote:
> The "ice_hw_autogen.h" file is an auto generated file from the HW engineers.
> It is not created dynamically when the driver is built. I'm not sure why they
> didn't call it ice_registers.h, but that was their decision.
my bad, it is in first
On 12/12/2018 3:19 PM, Ferruh Yigit wrote:
> On 12/12/2018 6:59 AM, Wenzhuo Lu wrote:
>> From: Paul M Stillwell Jr
>>
>> Add the structures required by the NIC.
>>
>> Signed-off-by: Paul M Stillwell Jr
>
> <...>
>
>> @@ -0,0 +1,869 @@
>> +/* SPDX-License-Identifier: BSD-3-Clause
>> + * Copyrigh
Added deprecation notice to replace rte_meter_color
with rte_color.
Signed-off-by: Reshma Pattan
Acked-by: Cristian Dumitrescu
---
doc/guides/rel_notes/deprecation.rst | 5 +
1 file changed, 5 insertions(+)
diff --git a/doc/guides/rel_notes/deprecation.rst
b/doc/guides/rel_notes/deprecati
Hi Ilya,
On 12/12/18 4:23 PM, Ilya Maximets wrote:
On 12.12.2018 11:24, Maxime Coquelin wrote:
Instead of writing back descriptors chains in order, let's
write the first chain flags last in order to improve batching.
With Kernel's pktgen benchmark, ~3% performance gain is measured.
Signed-off
> -Original Message-
> From: Varghese, Vipin
> Sent: Monday, December 3, 2018 5:50 AM
Since this patch displaying mepool and its elements both , edit heading to suit
both?
( or)
What about having iter mempool support as separate patch after this patch?
That looks more logical as i
On 12.12.2018 11:24, Maxime Coquelin wrote:
> Instead of writing back descriptors chains in order, let's
> write the first chain flags last in order to improve batching.
>
> With Kernel's pktgen benchmark, ~3% performance gain is measured.
>
> Signed-off-by: Maxime Coquelin
> ---
> lib/librte_v
On 12/12/2018 6:59 AM, Wenzhuo Lu wrote:
> From: Paul M Stillwell Jr
>
> Add the structures required by the NIC.
>
> Signed-off-by: Paul M Stillwell Jr
For consistency, base driver updates subsystem can be used as "net/ice/base: "
and it should start with lowercase, so something like:
"net/ic
On 12/12/2018 6:59 AM, Wenzhuo Lu wrote:
> From: Paul M Stillwell Jr
>
> Add the structures required by the NIC.
>
> Signed-off-by: Paul M Stillwell Jr
<...>
> @@ -0,0 +1,869 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2001-2018
> + */
> +
> +#ifndef _ICE_TYPE_H_
> +#def
> > @@ -83,7 +87,13 @@ proc_info_usage(const char *prgname)
> > " --stats-reset: to reset port statistics\n"
> > " --xstats-reset: to reset port extended statistics\n"
> > " --collectd-format: to print statistics to STDOUT in expected
> by
> > collectd format\
> -Original Message-
> From: Varghese, Vipin
> Sent: Monday, December 3, 2018 5:50 AM
> Subject: [PATCH v6 1/9] app/procinfo: add usage for new debug
>
> @@ -83,7 +87,13 @@ proc_info_usage(const char *prgname)
> " --stats-reset: to reset port statistics\n"
>
On 12-Dec-18 12:55 PM, Yongseok Koh wrote:
On Mon, Dec 03, 2018 at 10:23:03AM +, Burakov, Anatoly wrote:
On 02-Dec-18 11:28 PM, Yongseok Koh wrote:
On Dec 1, 2018, at 9:48 PM, Shahaf Shuler wrote:
Hi Anatoly,
Thursday, November 29, 2018 3:49 PM, Anatoly Burakov:
Subject: [PATCH 0/4] A
On Mon, Dec 03, 2018 at 10:23:03AM +, Burakov, Anatoly wrote:
> On 02-Dec-18 11:28 PM, Yongseok Koh wrote:
> >
> > > On Dec 1, 2018, at 9:48 PM, Shahaf Shuler wrote:
> > >
> > > Hi Anatoly,
> > >
> > > Thursday, November 29, 2018 3:49 PM, Anatoly Burakov:
> > > > Subject: [PATCH 0/4] Allow
On 12/11/2018 1:00 PM, Didier Pallard wrote:
> Since below commit, several tx_prep functions are broken, they fail to
> pass supported Tx offload features check:
> PKT_TX_IPVx must be set when any PKT_TX_L4 checksum is requested,
> but these values are not present in the mask of supported Tx offloa
On 12/11/2018 12:01 PM, Alejandro Lucero wrote:
> This patch introduces changes for supporting multiprocess support.
> This is trivial for VFs but comes with some limitations for the PF.
>
> Due to restrictions when using NFP CPP interface, just one secondary
> process is supported by now for the
Code refactoring to separate validation from benchmarking part.
Added op's status checking after rte_compressdev_dequeue_burst
function.
Signed-off-by: Tomasz Jozwiak
---
app/test-compress-perf/Makefile | 2 +
app/test-compress-perf/comp_perf_options.h| 12 +
app/tes
Added:
- initial version of compression performance test
description file.
- release note in release_18_11.rst
Updated index.rst file
Signed-off-by: De Lara Guarch, Pablo
Signed-off-by: Tomasz Jozwiak
---
MAINTAINERS| 5 +++
doc/guides/rel_notes/release_19_
Added dynamic compression feature into compression perf. test.
Signed-off-by: Tomasz Jozwiak
---
app/test-compress-perf/comp_perf_options_parse.c | 10 +++---
doc/guides/rel_notes/release_19_02.rst | 3 +--
doc/guides/tools/comp_perf.rst | 4 ++--
3 files change
This patchset adds initial version of compression performance
test.
v6 changes:
- Updated the 19.02 release notes
v5 changes:
- Fixed documentation
- Added op's status checking after rte_compressdev_dequeue_burst
- code cleanup
v4 changes:
- fixed checkpatch issues
- code cleanup
v3
Added performance measurement part into compression perf. test.
Signed-off-by: De Lara Guarch, Pablo
Signed-off-by: Tomasz Jozwiak
---
app/test-compress-perf/comp_perf_options_parse.c | 2 +-
app/test-compress-perf/main.c| 888 ++-
2 files changed, 884
Added parser part into compression perf. test.
Signed-off-by: De Lara Guarch, Pablo
Signed-off-by: Tomasz Jozwiak
---
app/Makefile | 4 +
app/meson.build | 1 +
app/test-compress-perf/Makefile | 16 +
app
On 12/10/2018 6:28 PM, Hyong Youb Kim wrote:
> These are patches for 19.02.
>
> Hyong Youb Kim (4):
> net/enic: release port upon close
> net/enic: add handler to return firmware version string
> net/enic: support multicast filtering
> doc: update release notes for enic
Series applied to
This patchset adds initial version of compression performance
test.
v6 changes:
- Updated the 19.02 release notes
v5 changes:
- Fixed documentation
- Added op's status checking after rte_compressdev_dequeue_burst
- code cleanup
v4 changes:
- fixed checkpatch issues
- code cleanup
v3
Add a programmer's guide section for meson ut enhancements
Signed-off-by: Hari Kumar Vemula
---
doc/guides/prog_guide/index.rst | 1 +
.../prog_guide/meson_ut_enhancements.rst | 158 ++
2 files changed, 159 insertions(+)
create mode 100644 doc/guides/prog_gu
On 12-Dec-18 11:25 AM, tone.zhang wrote:
With a larger PAGE_SIZE it is possible for the MSI table to very
close to the end of the BAR s.t. when we align the start and end
of the MSI table to the PAGE_SIZE, the end offset of the MSI
table is out of the PCI BAR boundary.
This patch addresses the i
On 12-Dec-18 11:10 AM, Yongseok Koh wrote:
malloc_elem_find_max_iova_contig() could return invalid size due to a
missing sanity check. The following gdb output shows how 'cur_size' can be
invalid in find_biggest_element().
(gdb) p/x cur_size
$4 = 0xffe42900
(gdb)
With a larger PAGE_SIZE it is possible for the MSI table to very
close to the end of the BAR s.t. when we align the start and end
of the MSI table to the PAGE_SIZE, the end offset of the MSI
table is out of the PCI BAR boundary.
This patch addresses the issue by comparing both the start and the
en
> -Original Message-
> From: Burakov, Anatoly
> Sent: Tuesday, December 11, 2018 5:57 PM
> To: dev@dpdk.org
> Cc: Lu, Wenzhuo ; Wu, Jingjing
> ; Iremonger, Bernard ;
> tho...@monjalon.net; Singh, Jasvinder ;
> Dumitrescu, Cristian ;
> jerin.ja...@caviumnetworks.com
> Subject: [PATCH 5/6]
Within rte_hash_reset, calling a while loop to dequeue one by
one from the ring, while not using them at all, is wasting cycles,
The patch just flush the ring by resetting the indices can save cpu
cycles.
Fixes: b26473ff8f4a ("hash: add reset function")
Fixes: 75706568a7eb ("hash: add extendable b
Currently, the flush is done by dequeuing the ring in a while loop. It is
much simpler to flush the queue by resetting the head and tail indices.
Signed-off-by: Gavin Hu
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_ring/rte_ring.h | 20 ++
V3: Allow experimental API for meson build
V2: Fix the coding style issue(commit message line too long)
V1: To flush a ring not in use, dequeue one by one is wasting cpu cycles.
The patch is to just resetting the head and tail indices to save cpu
cycle.
Gavin Hu (2):
ring: add reset api to flu
malloc_elem_find_max_iova_contig() could return invalid size due to a
missing sanity check. The following gdb output shows how 'cur_size' can be
invalid in find_biggest_element().
(gdb) p/x cur_size
$4 = 0xffe42900
(gdb) p elem
$1 = (struct malloc_elem *) 0x
When MPRQ mempool is freed, the pointer stored in priv structure must be
reset to null. Otherwise, the mempool can be freed again if the port is
restarted.
Fixes: 7d6bf6b866b8 ("net/mlx5: add Multi-Packet Rx support")
Cc: sta...@dpdk.org
Signed-off-by: Yongseok Koh
---
drivers/net/mlx5/mlx5_rxq
Hi Stephen,
Thanks for the comments.
> -Original Message-
> From: Stephen Hemminger
> Sent: Monday, December 10, 2018 11:56 PM
> To: Burakov, Anatoly
> Cc: Tone Zhang (Arm Technology China) ;
> dev@dpdk.org; Gavin Hu (Arm Technology China) ;
> Honnappa Nagarahalli ; Steve Capper
> ; nd
Hi Anatoly,
Thanks for the comments.
> -Original Message-
> From: Burakov, Anatoly
> Sent: Monday, December 10, 2018 7:46 PM
> To: Tone Zhang (Arm Technology China) ;
> dev@dpdk.org
> Cc: Gavin Hu (Arm Technology China) ; Honnappa
> Nagarahalli ; Steve Capper
> ; nd
> Subject: Re: [PATC
On 12/6/2018 10:13 AM, Rahul Lakkireddy wrote:
> When multiple Chelsio adapters are present, the current naming
> convention of using underlying physical port number of the adapter
> creates conflicts when allocating control queue mbuf pool for
> multiple adapters and hence results in allocation fa
On Wed, Dec 12, 2018 at 10:19:02AM +, Burakov, Anatoly wrote:
> On 10-Dec-18 10:50 AM, Richardson, Bruce wrote:
> >
> >
> > > -Original Message-
> > > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Burakov, Anatoly
> > > Sent: Monday, December 10, 2018 10:43 AM
> > > To: Jakub G
Hi Alejandro,
Yes, this mode datapath is through the relay thread when LM happens, it’s not
the direct interaction between VM and vdpa device.
BRs,
Xiao
From: Alejandro Lucero [mailto:alejandro.luc...@netronome.com]
Sent: Wednesday, December 12, 2018 2:15 AM
To: Wang, Xiao W
Cc: Bie, Tiwei ; M
On 10-Dec-18 10:50 AM, Richardson, Bruce wrote:
-Original Message-
From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Burakov, Anatoly
Sent: Monday, December 10, 2018 10:43 AM
To: Jakub Grajciar ; dev@dpdk.org
Subject: Re: [dpdk-dev] [RFC v2] /net: memory interface (memif)
On 10-Dec
On Wed, Dec 12, 2018 at 02:47:33PM +0800, Gavin Hu wrote:
> Within rte_hash_reset, calling a while loop to dequeue one by
> one from the ring, while not using them at all, is wasting cycles,
> The patch just flush the ring by resetting the indices can save cpu
> cycles.
>
> Fixes: b26473ff8f4a ("h
On Wed, Nov 28, 2018 at 9:56 AM Xiao Wang wrote:
> This patch series enables a new method for live migration, i.e. software
> assisted live migration. This patch provides a device argument for user
> to choose the methold.
>
> When "swlm=1", driver/device will do live migration with a relay threa
On Tue, Dec 11, 2018 at 06:02:01AM +, Honnappa Nagarahalli wrote:
> + Bruce
>
> > -Original Message-
> > From: Gavin Hu
> > Sent: Monday, December 10, 2018 9:56 PM
> > To: dev@dpdk.org
> > Cc: tho...@monjalon.net; jerin.ja...@caviumnetworks.com;
> > hemant.agra...@nxp.com; Honnappa Na
Sorry, just notice I missed to add v2 to the commit message prefix.
On 12/12/18 9:24 AM, Maxime Coquelin wrote:
Instead of writing back descriptors chains in order, let's
write the first chain flags last in order to improve batching.
With Kernel's pktgen benchmark, ~3% performance gain is measu
> > >
> > > > > +
> > > > > +/* Add a reader thread, running on an lcore, to the list of
> > > > > +threads
> > > > > + * reporting their quiescent state on a TQS variable.
> > > > > + */
> > > > > +int __rte_experimental
> > > > > +rte_tqs_register_lcore(struct rte_tqs *v, unsigned int lcore_id)
Instead of writing back descriptors chains in order, let's
write the first chain flags last in order to improve batching.
With Kernel's pktgen benchmark, ~3% performance gain is measured.
Signed-off-by: Maxime Coquelin
---
lib/librte_vhost/virtio_net.c | 39 +--
99 matches
Mail list logo
