The series has fixes for problems discussed in [1].
Basically it does not allow unsupported offloads to pass.
If fixes regressions for PMDs which carefully check offloads before,
but these checks are removed now in favor of checks in ethdev.
It may break applications which request some offload w
Do not allow to request unsupported Tx offload since all checks
are removed from PMDs because of consistency check in ethdev.
Otherwise application may rely on offload which is not actually
supported and send traffic with, for example, wrong checksums,
truncated packets or packets with garbage.
If
Do not allow allow unsupported offload to be passed to PMD since
in can result in inconsistent NIC configuration and processing
in the driver.
All PMDs are converted to the new offload API and must report its
capabilties correctly. Both device and queue offloads are listed
in [rt]x_offload_capa so
Return of error was removed to mitigate possible breakage of old
applications which are not converted to the new offload API yet.
Old Rx offload API has no per queue controls and Rx queue offloads
are derived from the device Rx mode bitfields exactly in the same
way as it is done on configure and,
Monday, May 14, 2018 10:36 AM, Andrew Rybchenko:
> Subject: [PATCH v2 0/3] ethdev: fail if requested offload is not supported
>
> The series has fixes for problems discussed in [1].
>
> Basically it does not allow unsupported offloads to pass.
>
> If fixes regressions for PMDs which carefully ch
HI Pablo
>-Original Message-
>From: Pablo de Lara [mailto:pablo.de.lara.gua...@intel.com]
>Sent: 27 April 2018 18:54
>To: dev@dpdk.org
>Cc: fiona.tr...@intel.com; Verma, Shally ;
>ahmed.mans...@nxp.com; Gupta, Ashish
>; Pablo de Lara ;
>Verma, Shally ; Gupta,
>Ashish
>Subject: [PATCH v6
The NIC always strips CRC, so advertise the CRC strip offload in
rx_offload_capa.
Fixes: 93fb21fdbe23 ("net/enic: enable overlay offload for VXLAN and GENEVE")
Signed-off-by: Hyong Youb Kim
---
drivers/net/enic/enic_res.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/net/enic/enic
Hi Shally,
> -Original Message-
> From: Verma, Shally [mailto:shally.ve...@cavium.com]
> Sent: Monday, May 14, 2018 8:54 AM
> To: De Lara Guarch, Pablo ; dev@dpdk.org
> Cc: Trahe, Fiona ; ahmed.mans...@nxp.com; Gupta,
> Ashish ; Gupta, Ashish
> ; Sahu, Sunila
> Subject: RE: [PATCH v6 02/1
On 09-May-18 6:08 PM, Mike Stolarchuk wrote:
Hello Dpdk,
rte_eth_dev_socket_id() describes a -1 return value as:
*Returns*
The NUMA socket id to which the Ethernet device is connected or a default
of zero if the socket could not be determined. -1 is returned is the
port_id value is out of rang
> -Original Message-
> From: Gujjar, Abhinandan S
> Sent: Monday, May 14, 2018 7:50 AM
> To: jerin.ja...@caviumnetworks.com; tho...@monjalon.net; dev@dpdk.org
> Cc: De Lara Guarch, Pablo ; Gujjar,
> Abhinandan S
> Subject: [PATCH] test: fix build with icc
>
> This patch provides fix for
>-Original Message-
>From: De Lara Guarch, Pablo [mailto:pablo.de.lara.gua...@intel.com]
>Sent: 14 May 2018 13:34
>To: Verma, Shally ; dev@dpdk.org
>Cc: Trahe, Fiona ; ahmed.mans...@nxp.com; Gupta, Ashish
>; Gupta, Ashish
>; Sahu, Sunila
>Subject: RE: [PATCH v6 02/14] compressdev: add q
On 11-May-18 11:25 AM, Remy Horton wrote:
On 03/05/2018 18:18, Anatoly Burakov wrote:
[..]
Also, fixup unit tests to account for the new behavior.
Tried running the tests but it fails on a boundary check:
RTE>>memzone_autotest
test basic memzone API
[...]
1GB Huge pages available
test alignm
On 10-May-18 2:57 PM, Remy Horton wrote:
On 03/05/2018 18:18, Anatoly Burakov wrote:
[..]
Signed-off-by: Anatoly Burakov
[..]
+ for (idx = 0; idx < RTE_HEAP_NUM_FREELISTS; idx++) {
+ for (elem = LIST_FIRST(&heap->free_head[idx]);
+ !!elem; elem = LIST_NEXT(elem, fre
On 09-May-18 8:03 PM, Thomas Monjalon wrote:
09/05/2018 18:11, Bruce Richardson:
On Wed, May 09, 2018 at 04:59:39PM +0100, Van Haaren, Harry wrote:
From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Anatoly Burakov
Sent: Monday, April 30, 2018 1:09 PM
To: dev@dpdk.org
Cc: Richardson, Bruce ;
>-Original Message-
>From: Pablo de Lara [mailto:pablo.de.lara.gua...@intel.com]
>Sent: 04 May 2018 15:52
>To: dev@dpdk.org
>Cc: fiona.tr...@intel.com; lee.d...@intel.com; Verma, Shally
>; ahmed.mans...@nxp.com; Gupta,
>Ashish ; Pablo de Lara
>; Gupta, Ashish ;
>Verma, Shally
>Subject:
On 08-May-18 8:10 AM, Xiao Wang wrote:
From: Maxime Coquelin
When memory is hot-added or hot-removed, the virtio-user driver has to
notify the vhost-user backend with sending a VHOST_USER_SET_MEM_TABLE
request with the new memory map as payload.
This patch implements and registers a mem_event
> -Original Message-
> From: Verma, Shally [mailto:shally.ve...@cavium.com]
> Sent: Monday, May 14, 2018 9:17 AM
> To: De Lara Guarch, Pablo ; dev@dpdk.org
> Cc: Trahe, Fiona ; ahmed.mans...@nxp.com; Gupta,
> Ashish ; Gupta, Ashish
> ; Sahu, Sunila
> Subject: RE: [PATCH v6 02/14] compres
Wei Dai,
Do you agree with my comments?
Could we have a wording patch to squash in RC3?
10/05/2018 23:39, Thomas Monjalon:
> Hi,
>
> A first general comment: a lot of spaces are still inside parens.
> You can grep '( )'.
>
> 10/05/2018 13:56, Wei Dai:
> > --- a/doc/guides/prog_guide/poll_mode_d
> -Original Message-
> From: Verma, Shally [mailto:shally.ve...@cavium.com]
> Sent: Monday, May 14, 2018 9:29 AM
> To: De Lara Guarch, Pablo ; dev@dpdk.org
> Cc: Trahe, Fiona ; Daly, Lee ;
> ahmed.mans...@nxp.com; Gupta, Ashish ; Gupta,
> Ashish
> Subject: RE: [PATCH v4 1/5] test/compres
> > This patch provides fix for icc compilation issue with event crypto adapter
> > test
> > application.
> > Currently, number of elements in the session mempool is defined by
> > max_nb_sessions & this is replaced with a macro.
> >
> > Fixes: 3c2c535ecfc0 ("test: add event crypto adapter auto-t
With this change, DPAA and DPAA2 configuration compiled applications
are interoperable without any impact on the functional behavior.
Signed-off-by: Shreyansh Jain
---
Note for beyond rc3 patch:
This has no functional impact on DPAA builds.
config/defconfig_arm64-dpaa-linuxapp-gcc | 2 +-
1 fi
The function pthread_create() expects void *(*func) (void *)
for function pointer, however, lthread_func_t was defined as
void (*func) (void *), so now gcc 8.1 warns that the cast is
incorrect, causing a compilation failure. This patch changes
the declaration of lthread_func_t from returning a void
From: "Zhang, Tianfei"
fix compile error on ia32 icc compiler
Fixes: 56bb54ea1bdf ("raw/ifpga: add Intel FPGA bus rawdev driver")
Cc: sta...@dpdk.org
Signed-off-by: Zhang, Tianfei
---
drivers/raw/ifpga_rawdev/base/ifpga_feature_dev.h | 10 +++---
1 file changed, 7 insertions(+), 3 deletio
In strlcpy function parameters there was no allowance for
null terminator, so ip address was copied without last character.
Fixes: ae943ebe1ed3 ("examples/ipsec-secgw: replace strncpy with strlcpy")
Signed-off-by: Kirill Rybalchenko
---
examples/ipsec-secgw/parser.c | 9 +
1 file change
Hi Thomas,
On 13/5/2018 10:37 PM, Thomas Monjalon wrote:
10/05/2018 07:31, David Hunt:
The function pthread_create() expects void *(*func) (void *)
for function pointer, however, lthread_func_t was defined as
void (*func) (void *), so now gcc 8.1 warns that the cast is
incorrect, causing a comp
On 5/13/2018 6:30 AM, Shahaf Shuler wrote:
> Friday, May 11, 2018 7:09 PM, Andrew Rybchenko:
>
> On 05/11/2018 05:22 PM, Ferruh Yigit wrote:
>
> I think we have the following options:
>
>
>
> A. Rollback corresponding changes which remove checks from PMDs
>
>
Thanks a lot!
> -Original Message-
> From: Rybalchenko, Kirill
> Sent: Monday, May 14, 2018 7:53 AM
> To: dev@dpdk.org
> Cc: sta...@dpdk.org; Rybalchenko, Kirill ;
> Zhang, Roy Fan
> Subject: [PATCH] crypto/scheduler: fix multicore rings re-use
>
> When scheduler mode changed from multic
Hi,
> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Sunday, May 13, 2018 18:07
> To: De Lara Guarch, Pablo
> Cc: dev@dpdk.org; Xu, Rosen
> Subject: Re: [dpdk-dev] [PATCH] bus/fpga: use strlcpy instead of strncpy
>
> 12/05/2018 09:02, Pablo de Lara:
> >
Function mlx4_conv_rss_types() contains constant arrays variables
which are recreated with every call to the function. By changing the
arrays definitions from "const" to "static const" these recreations
can be saved.
Signed-off-by: Ophir Munk
---
v1:
Initial release
v2:
Update based on reviews (s
Advertise mlx4 supported RSS functions as part of dev_infos_get
callback.
Previous to this commit RSS support was reported as none. Since the
introduction of [1] it is required that all RSS configurations will be
verified.
[1] commit 8863a1fbfc66 ("ethdev: add supported hash function check")
Sign
Monday, May 14, 2018 1:08 PM, Ophir Munk:
> Subject: [PATCH v6 1/2] net/mlx4: avoid constant recreations in function
>
> Function mlx4_conv_rss_types() contains constant arrays variables which are
> recreated with every call to the function. By changing the arrays definitions
> from "const" to "st
Fixes: 4c07e0552f0a ("crypto/scheduler: add multicore scheduling mode")
This patch fixes the possible duplicated ring names in multi-core
scheduler. Originally two or more multi-core schedulers may have
same worker ring names thus will cause initialization error.
Signed-off-by: Fan Zhang
---
dr
Monday, May 14, 2018 1:00 PM, Ferruh Yigit:
> >
> > It depends on which PMD is used. Yes, it was no checks in ethdev before.
> > If PMD does not support multi-segment Tx, some checksum or VLAN
> > insertion offload, but application requests it and rely on it, it will
> > result in invalid packets s
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Xu, Rosen
> Sent: Monday, May 14, 2018 10:59 AM
> To: dev@dpdk.org; tho...@monjalon.net
> Cc: Xu, Rosen ; Zhang, Roy Fan
> ; Doherty, Declan ;
> Richardson, Bruce ; shreyansh.j...@nxp.com;
> Yigit, Ferruh ; Ananye
14/05/2018 04:53, David Hunt:
> The function pthread_create() expects void *(*func) (void *)
> for function pointer, however, lthread_func_t was defined as
> void (*func) (void *), so now gcc 8.1 warns that the cast is
> incorrect, causing a compilation failure. This patch changes
> the declaration
> > From: "Zhang, Tianfei"
> >
> > fix compile error on ia32 icc compiler
>
> This is also applicable for 64-bit icc compilation.
>
> >
> > Fixes: 56bb54ea1bdf ("raw/ifpga: add Intel FPGA bus rawdev driver")
> > Cc: sta...@dpdk.org
>
> No need to cc stable, as this is fixing code from this re
Hi Awal,
> -Original Message-
> From: Awal, Mohammad Abdul
> Sent: Friday, May 11, 2018 6:14 PM
> To: dev@dpdk.org; Iremonger, Bernard ;
> adrien.mazarg...@6wind.com
> Subject: [PATCH v2 2/2] doc/guides: updated testpmd app user guide for
> vxlan/nvgre encap/decap
>
> Updated the testpmd
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Andy Green
> Sent: Monday, May 14, 2018 6:11 AM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v4 17/23] test_table_pipeline: repair munged
> indirection level
>
> Signed-off-by: Andy Green
> ---
> test/test/t
Hi
From: zhiyong.y...@intel.com
> The second parameter "name" in the function rte_eth_dev_detach has been
> already redefined as "char *name __rte_unused", "port_id" is printed instead
> of "name" in testpmd.
>
> Fixes: b65ecf199324 ("devargs: rename legacy API")
> Cc: sta...@dpdk.org
Really m
Fixes: 0ebce6129bc6 ("net/dpaa2: support new ethdev offload APIs")
Cc: sunil.k...@nxp.com
Signed-off-by: Shreyansh Jain
---
Note: Though checkpatch produces a warning for 80 char, this is more readable
than splitting it across multiple lines.
drivers/net/dpaa2/dpaa2_rxtx.c | 2 +-
1 file
Read IOMMU attributes from linux kernle 4.10.0 error.
Cc: sta...@dpdk.org
Signed-off-by: Rosen Xu
---
doc/guides/rel_notes/known_issues.rst | 21 +
1 file changed, 21 insertions(+)
diff --git a/doc/guides/rel_notes/known_issues.rst
b/doc/guides/rel_notes/known_issues.rst
i
On Friday 11 May 2018 04:15 PM, Ferruh Yigit wrote:
Some PMDs are still using old offloading API related variables, thanks to Andrew
for reporting it.
Maintainers of related PMDs are cc'ed.
Please check following list and provide a fix ASAP for it, thanks.
# git grep header_split drivers/
driv
14/05/2018 12:05, Shreyansh Jain:
> With this change, DPAA and DPAA2 configuration compiled applications
> are interoperable without any impact on the functional behavior.
>
> Signed-off-by: Shreyansh Jain
> ---
> Note for beyond rc3 patch:
> This has no functional impact on DPAA builds.
Applie
On 05/14/2018 06:35 PM, Singh, Jasvinder wrote:
-Original Message-
From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Andy Green
Sent: Monday, May 14, 2018 6:11 AM
To: dev@dpdk.org
Subject: [dpdk-dev] [PATCH v4 17/23] test_table_pipeline: repair munged
indirection level
Signed-off
On Mon, May 14, 2018 at 12:56:20AM -0700, Hyong Youb Kim wrote:
> The NIC always strips CRC, so advertise the CRC strip offload in
> rx_offload_capa.
Hi Ferruh,
Can you discard this patch? We are finding more offload flags are
missing. Will submit a new patch that sets them all instead of one pat
14/05/2018 12:38, Matan Azrad:
>
> Hi
>
> From: zhiyong.y...@intel.com
> > The second parameter "name" in the function rte_eth_dev_detach has been
> > already redefined as "char *name __rte_unused", "port_id" is printed instead
> > of "name" in testpmd.
> >
> > Fixes: b65ecf199324 ("devargs: re
On 5/14/2018 11:50 AM, Hyong Youb Kim wrote:
> On Mon, May 14, 2018 at 12:56:20AM -0700, Hyong Youb Kim wrote:
>> The NIC always strips CRC, so advertise the CRC strip offload in
>> rx_offload_capa.
>
> Hi Ferruh,
>
> Can you discard this patch? We are finding more offload flags are
> missing. Wi
> -Original Message-
> From: Li Han [mailto:han@zte.com.cn]
> Sent: Wednesday, April 18, 2018 9:52 AM
> To: Ananyev, Konstantin
> Cc: dev@dpdk.org; Li Han
> Subject: [PATCH] lib/librte_ip_frag:fix ip frag process log
>
> in ip_frag_process,some IP_FRAG_LOG content is wrong
>
> Sig
Hi Awal,
> -Original Message-
> From: Awal, Mohammad Abdul
> Sent: Friday, May 11, 2018 6:14 PM
> To: dev@dpdk.org; Iremonger, Bernard ;
> adrien.mazarg...@6wind.com
> Subject: [PATCH v2 1/2] app/testpmd: enabled vxlan and nvgre encap/decap
> support for rte_flow
>
> We have enabled testp
On 5/14/2018 6:04 AM, Andy Green wrote:
> /home/agreen/projects/dpdk/drivers/net/bnx2x/elink.c:
> In function ‘elink_check_kr2_wa’:
> /home/agreen/projects/dpdk/drivers/net/bnx2x/elink.c:12922:28:
> error: bitwise comparison always evaluates to false
> [-Werror=tautological-compare]
> ((nex
Hi, Thomas & Ferruh
Thanks for your feedback.
I agree with your comments and I am working on the latest commit of the repo
dpdk-next-net.
I will submit a new patch to adopt your suggestion.
Please wait for a while ...
> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.n
Currently, reserving zero-length memzones is done by looking at
malloc statistics, and reserving biggest sized element found in those
statistics. This has two issues.
First, there is a race condition. The heap is unlocked between the
time we check stats, and the time we reserve malloc element for
User might be interested to find out what is the biggest chunk of
IOVA-contiguous free space that can be allocated from malloc. Add
relevant malloc-internal functions and expose this through malloc
stats calculation call.
Signed-off-by: Anatoly Burakov
Acked-by: Remy Horton
---
Notes:
v4:
Add an internal-only function to allocate biggest element from
the heap. Nominally, it supports SOCKET_ID_ANY as its socket
argument, but it's essentially useless because other sockets
will only be allocated from if the entire heap on current or
specified socket is busy.
Still, asking to reserve a
This patchset does two things. First, it enables reserving
memzones of zero-length that are IOVA-contiguous. Second,
it fixes a long-standing race condition in reserving
zero-length memzones, where malloc heap is not locked between
stats collection and reservation, and will instead allocate
biggest
> -Original Message-
> From: Rybalchenko, Kirill
> Sent: Monday, May 14, 2018 11:01 AM
> To: dev@dpdk.org
> Cc: sta...@dpdk.org; Rybalchenko, Kirill ;
> Nicolau, Radu ; akhil.go...@nxp.com; Zhang, Roy Fan
> ; Pattan, Reshma ; De
> Lara Guarch, Pablo
> Subject: [PATCH] examples/ipsec-secg
On 14-May-18 9:21 AM, Burakov, Anatoly wrote:
On 11-May-18 11:25 AM, Remy Horton wrote:
On 03/05/2018 18:18, Anatoly Burakov wrote:
[..]
Also, fixup unit tests to account for the new behavior.
Tried running the tests but it fails on a boundary check:
RTE>>memzone_autotest
test basic memzone
On 24/04/2018 12:29 PM, Matan Azrad wrote:
The bonding PMD decides to activate\deactivate its slaves according to
the slaves link statuses.
Thus, it registers to the LSC events of the slaves ports and
activates\deactivates them from its LSC callbacks called asynchronously
by the host thread when
On 5/12/2018 2:35 AM, Yongseok Koh wrote:
> Offset in a MPRQ buffer must be calculated before updating the stride
> index.
>
> Fixes: f0d61f8f8953 ("net/mlx5: add Multi-Packet Rx support")
>
> Signed-off-by: Yongseok Koh
Squashed into relevant commit in next-net, thanks.
On 5/12/2018 2:35 AM, Yongseok Koh wrote:
> Fixes: f0d61f8f8953 ("net/mlx5: add Multi-Packet Rx support")
>
> Signed-off-by: Yongseok Koh
Squashed into relevant commit in next-net, thanks.
> > In strlcpy function parameters there was no allowance for null terminator,
> > so ip
> > address was copied without last character.
> >
> > Fixes: ae943ebe1ed3 ("examples/ipsec-secgw: replace strncpy with strlcpy")
> >
> > Signed-off-by: Kirill Rybalchenko
>
> Acked-by: Pablo de Lara
>
>
On 5/12/2018 12:50 PM, Rahul Lakkireddy wrote:
> Move freeing up resources from dev_close() to dev_uninit(). This fixes
> NULL pointer de-reference when accessing adapter context needed by
> other ports under same PF, but had been freed up by the first port.
> This can happen if only the first port
This patch check if a input requested offloading is valid or not.
Any reuqested offloading must be supported in the device capabilities.
Any offloading is disabled by default if it is not set in the parameter
dev_conf->[rt]xmode.offloads to rte_eth_dev_configure() and
[rt]x_conf->offloads to rte_et
On 14-May-18 12:29 PM, Burakov, Anatoly wrote:
On 14-May-18 9:21 AM, Burakov, Anatoly wrote:
On 11-May-18 11:25 AM, Remy Horton wrote:
On 03/05/2018 18:18, Anatoly Burakov wrote:
[..]
Also, fixup unit tests to account for the new behavior.
Tried running the tests but it fails on a boundary
On 5/14/2018 11:58 AM, Shreyansh Jain wrote:
> Fixes: 0ebce6129bc6 ("net/dpaa2: support new ethdev offload APIs")
> Cc: sunil.k...@nxp.com
>
> Signed-off-by: Shreyansh Jain
Applied to dpdk-next-net/master, thanks.
> ---
> Note: Though checkpatch produces a warning for 80 char, this is more rea
> On May 7, 2018, at 3:36 AM, Ophir Munk wrote:
>
> Running testpmd command "flow isolae 0" (i.e. disabling flow
> isolation) followed by command "flow isolate 1" (i.e. enabling
> flow isolation) may result in a TAP error:
> PMD: Kernel refused TC filter rule creation (17): File exists
>
> R
On 5/11/2018 4:09 PM, Jingjing Wu wrote:
> Vector used for rx mapping is different if WB_ON_ITR
> is supported. The mapping table need to be updated.
>
> Fixes: d6bde6b5eae9 ("net/avf: enable Rx interrupt")
>
> Signed-off-by: Jingjing Wu
> Cc: sta...@dpdk.org
Applied to dpdk-next-net/master, th
On 5/14/2018 12:45 PM, Doherty, Declan wrote:
> On 24/04/2018 12:29 PM, Matan Azrad wrote:
>> The bonding PMD decides to activate\deactivate its slaves according to
>> the slaves link statuses.
>> Thus, it registers to the LSC events of the slaves ports and
>> activates\deactivates them from its LS
On 5/11/2018 10:16 AM, Andrew Rybchenko wrote:
> On 03/28/2018 03:15 AM, Rasesh Mody wrote:
>> From: Harish Patil
>>
>> Ethdev RX/TX offloads API has changed since:
>> commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API")
>> commit cba7f53b717d ("ethdev: introduce Tx queue offloads API")
14/05/2018 14:00, Wei Dai:
> --- a/doc/guides/prog_guide/poll_mode_drv.rst
> +++ b/doc/guides/prog_guide/poll_mode_drv.rst
> @@ -303,12 +303,12 @@ In the DPDK offload API, offloads are divided into
> per-port and per-queue offloa
> * A pure per-port offloading can't be enabled on a queue and disa
- add EXPERIMENTAL tag for the section in MAINTAINERS.
- add EXPERIMENTAL tag to BPF public API files.
- add attribute __rte_experimental to BPF public API declarations.
Fixes: 94972f35a02e ("bpf: add BPF loading and execution framework")
Fixes: 5dba93ae5f2d ("bpf: add ability to load eBPF program
Hi, Thomas
Thanks for your quick feedback.
> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Monday, May 14, 2018 8:54 PM
> To: Dai, Wei
> Cc: dev@dpdk.org; Yigit, Ferruh ; Zhang, Qi Z
>
> Subject: Re: [dpdk-dev] [PATCH v13] ethdev: new Rx/Tx offloads API
Fixes the wrong argument in table action miss function.
Fixes: 4c387fcdf777 ("pipeline: add new functions for action handlers")
Signed-off-by: Jasvinder Singh
---
test/test/test_table_pipeline.c | 10 --
1 file changed, 4 insertions(+), 6 deletions(-)
diff --git a/test/test/test_table_p
This patch check if a input requested offloading is valid or not.
Any reuqested offloading must be supported in the device capabilities.
Any offloading is disabled by default if it is not set in the parameter
dev_conf->[rt]xmode.offloads to rte_eth_dev_configure() and
[rt]x_conf->offloads to rte_et
Add support for MPLS over GRE and MPLS over UDP tunnel types as
described in the next RFCs:
1. https://tools.ietf.org/html/rfc4023
2. https://tools.ietf.org/html/rfc7510
3. https://tools.ietf.org/html/rfc4385
Signed-off-by: Matan Azrad
---
doc/guides/nics/mlx5.rst | 4 +-
drivers/net/mlx5/
Add an internal-only function to allocate biggest element from
the heap. Nominally, it supports SOCKET_ID_ANY as its socket
argument, but it's essentially useless because other sockets
will only be allocated from if the entire heap on current or
specified socket is busy.
Still, asking to reserve a
User might be interested to find out what is the biggest chunk of
IOVA-contiguous free space that can be allocated from malloc. Add
relevant malloc-internal functions and expose this through malloc
stats calculation call.
Signed-off-by: Anatoly Burakov
Acked-by: Remy Horton
---
Notes:
v4:
This patchset does two things. First, it enables reserving
memzones of zero-length that are IOVA-contiguous. Second,
it fixes a long-standing race condition in reserving
zero-length memzones, where malloc heap is not locked between
stats collection and reservation, and will instead allocate
biggest
Currently, reserving zero-length memzones is done by looking at
malloc statistics, and reserving biggest sized element found in those
statistics. This has two issues.
First, there is a race condition. The heap is unlocked between the
time we check stats, and the time we reserve malloc element for
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Anatoly Burakov
> Sent: Monday, May 14, 2018 7:17 PM
> To: dev@dpdk.org
> Cc: remy.hor...@intel.com
> Subject: [dpdk-dev] [PATCH v5 1/3] malloc: add biggest free IOVA-
> contiguous element to stats
>
> User might b
On 5/14/2018 11:19 AM, Shahaf Shuler wrote:
> Monday, May 14, 2018 1:00 PM, Ferruh Yigit:
>>>
>>> It depends on which PMD is used. Yes, it was no checks in ethdev before.
>>> If PMD does not support multi-segment Tx, some checksum or VLAN
>>> insertion offload, but application requests it and rely
14/05/2018 15:20, Wei Dai:
> Signed-off-by: Wei Dai
> Signed-off-by: Ferruh Yigit
> Signed-off-by: Qi Zhang
[...]
> * hasn't been set in the input argument eth_conf->rxmode.offloads
> * to rte_eth_dev_configure(), it is a new added offloading, it must be
> * per-queue type and it is
Add the following missing flags to the advertised offloads.
- DEV_RX_OFFLOAD_CRC_STRIP
CRC is always stripped.
- DEV_RX_OFFLOAD_JUMBO_FRAME
Jumbo support is always enabled on the NIC.
- DEV_RX_OFFLOAD_SCATTER
Scatter Rx is currently supported.
- DEV_TX_OFFLOAD_MULTI_SEGS
Multiple-segment tr
14/05/2018 16:10, Ferruh Yigit:
> On 5/14/2018 11:19 AM, Shahaf Shuler wrote:
> > Monday, May 14, 2018 1:00 PM, Ferruh Yigit:
> >>>
> >>> It depends on which PMD is used. Yes, it was no checks in ethdev before.
> >>> If PMD does not support multi-segment Tx, some checksum or VLAN
> >>> insertion of
Hi,
We have just discovered an issue in master branch when starting testpmd.
The core number is wrongly detected after this commit:
http://dpdk.org/commit/e505d84c64
Command used for the test:
/x86_64-native-linuxapp-gcc/build/app/test-pmd/testpmd -n 4 -w
00:07.0,l3_vxlan_en=1,rx_vec_en=
On 5/14/2018 2:15 PM, Konstantin Ananyev wrote:
> - add EXPERIMENTAL tag for the section in MAINTAINERS.
> - add EXPERIMENTAL tag to BPF public API files.
> - add attribute __rte_experimental to BPF public API declarations.
>
> Fixes: 94972f35a02e ("bpf: add BPF loading and execution framework")
>
Hi.
This patch adds two features to the librte_ip_frag library:
1) it keeps track of number of mbufs holded in the fragmentation table.
rte_frag_table_mbuf_count()
There might be situations (kind of attack when a lot of fragmented packets are
sent
to a dpdk application in order to flood the fr
14/05/2018 16:20, Ferruh Yigit:
> On 5/14/2018 2:15 PM, Konstantin Ananyev wrote:
> > - add EXPERIMENTAL tag for the section in MAINTAINERS.
> > - add EXPERIMENTAL tag to BPF public API files.
> > - add attribute __rte_experimental to BPF public API declarations.
> >
> > Fixes: 94972f35a02e ("bpf:
14/05/2018 15:15, Konstantin Ananyev:
> - add EXPERIMENTAL tag for the section in MAINTAINERS.
> - add EXPERIMENTAL tag to BPF public API files.
> - add attribute __rte_experimental to BPF public API declarations.
>
> Fixes: 94972f35a02e ("bpf: add BPF loading and execution framework")
> Fixes: 5d
On 5/14/2018 3:11 PM, Hyong Youb Kim wrote:
> Add the following missing flags to the advertised offloads.
> - DEV_RX_OFFLOAD_CRC_STRIP
> CRC is always stripped.
> - DEV_RX_OFFLOAD_JUMBO_FRAME
> Jumbo support is always enabled on the NIC.
> - DEV_RX_OFFLOAD_SCATTER
> Scatter Rx is currently su
On 5/14/2018 3:11 PM, Thomas Monjalon wrote:
> 14/05/2018 15:20, Wei Dai:
>> Signed-off-by: Wei Dai
>> Signed-off-by: Ferruh Yigit
>> Signed-off-by: Qi Zhang
>
> [...]
>> * hasn't been set in the input argument eth_conf->rxmode.offloads
>> * to rte_eth_dev_configure(), it is a new added
For the stable versions of DPDK, the NFP driver's NSP user space
driver was modified to account for non-root usage. When that
happened, commit 515933ad8385 ("nfp: allow for non-root user")
inadvertently removed the unlink() call.
Fixes: 515933ad8385 ("nfp: allow for non-root user")
Cc: Alejandro
On 5/14/2018 8:51 AM, Shahaf Shuler wrote:
> Monday, May 14, 2018 10:36 AM, Andrew Rybchenko:
>> Subject: [PATCH v2 0/3] ethdev: fail if requested offload is not supported
>>
>> The series has fixes for problems discussed in [1].
>>
>> Basically it does not allow unsupported offloads to pass.
>>
>>
> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Monday, May 14, 2018 3:31 PM
> To: Ananyev, Konstantin
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] bpf: add missed EXPERIMENTAL tags
>
> 14/05/2018 15:15, Konstantin Ananyev:
> > - add EXPERIMENTA
14/05/2018 16:55, Ananyev, Konstantin:
>
> > -Original Message-
> > From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > Sent: Monday, May 14, 2018 3:31 PM
> > To: Ananyev, Konstantin
> > Cc: dev@dpdk.org
> > Subject: Re: [dpdk-dev] [PATCH] bpf: add missed EXPERIMENTAL tags
> >
> > 14/
There's possibly an issue here:
/* If the device isn't started don't handle interrupts */
if (!bonded_eth_dev->data->dev_started)
return rc;
/* verify that port_id is a valid slave of bonded port */
for (i = 0; i < internals->slave_count; i++) {
On Wed, May 09, 2018 at 09:57:31AM +, Shahaf Shuler wrote:
> Wednesday, May 9, 2018 11:47 AM, Ophir Munk:
> > Subject: Please revert: [dpdk-dev] [PATCH 2/2] net/mlx4: fix useless default
> > in RSS converter
> >
> > Hi Shahaf, Adrien,
> >
> > This commit (in next-net/master since 8-May-18) is
On 14/05/18 16:46, Aaron Conole wrote:
For the stable versions of DPDK, the NFP driver's NSP user space
driver was modified to account for non-root usage. When that
happened, commit 515933ad8385 ("nfp: allow for non-root user")
inadvertently removed the unlink() call.
Acked-by: Eelco Chaudron
Hi Thomas,
> > 2. Fixed in autotest_test_funcs.py to handle test cases which returns
> > "Skipped" as result.
> > The issue was skipped test cases got timed out, causing delay in
> > sanity script execution.
> > 3. Enhanced support for FreeBSD,
> > as FreeBSD doesn't support socket-mem, file-pre
14/05/2018 17:26, Pattan, Reshma:
> Hi Thomas,
>
> > > 2. Fixed in autotest_test_funcs.py to handle test cases which returns
> > > "Skipped" as result.
> > > The issue was skipped test cases got timed out, causing delay in
> > > sanity script execution.
> > > 3. Enhanced support for FreeBSD,
> >
1 - 100 of 163 matches
Mail list logo
