On 11/5/2020 3:11 PM, Lance Richardson wrote:
With this change, the bnxt driver fails to initialize under testpmd:
Configuring Port 0 (socket 0)
Port 0 failed to enable Rx offload JUMBO_FRAME
Fail to configure port 0
EAL: Error - exiting with code: 1
It appears that the cause is this bit of cod
On Wed, Nov 4, 2020 at 6:28 PM David Marchand wrote:
> 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: Oliv
On Wed, Nov 4, 2020 at 7:59 PM Dharmik Thakkar wrote:
>
> 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 thr
On Thu, Nov 05, 2020 at 09:33:15AM -0600, Dharmik Thakkar wrote:
> With pkg-config support available within AArch64crypto library,
> meson option 'armv8_crypto_dir' can be removed.
> PKG_CONFIG_PATH environment variable should be set appropriately
> to use the crypto library.
>
> Refer [1] for bui
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Ananyev,
> Konstantin
> Sent: Thursday, November 5, 2020 2:56 PM
>
> > >
> > > >
> > > > On Thu, Nov 05, 2020 at 11:34:18AM +, Ananyev, Konstantin
> wrote:
> > > > >
> > > > >
> > > > > > On Thu, Nov 05, 2020 at 11:26:51AM +0300, Andrew Ry
05/11/2020 14:31, Olivier Matz:
> On Fri, Oct 30, 2020 at 04:01:26AM +0300, Dmitry Kozlyuk wrote:
> > cmdline_numtype member names clash with Windows system identifiers.
> > Add RTE_ prefix to cmdline constants to avoid this and possible
> > future conflicts.
> >
> > Suggested-by: Ranjit Menon
>
05/11/2020 14:36, Olivier Matz:
> On Mon, Oct 26, 2020 at 03:42:01PM +0100, Morten Brørup wrote:
> > The mbuf header files had some commenting style errors that affected the
> > API documentation.
> > Also, the RTE_ prefix was missing on a macro and a definition.
> >
> > Note: This patch does not
On Thu, 5 Nov 2020 09:27:59 +0100
David Marchand wrote:
> On Sat, Oct 24, 2020 at 3:02 AM Stephen Hemminger
> wrote:
> >
> > This is a revised version of the earlier RFC patch set for changing
> > the blacklist/whitelist terms in DPDK.
> >
> > The first patch is a duplicate from the other patch
On Thu, Nov 05, 2020 at 12:13:37PM +0100, Thomas Monjalon wrote:
> 05/11/2020 12:07, Bruce Richardson:
> > It's reasonably common for patches to have issues when built on 32-bits, so
> > to prevent this, we can add a 32-bit build (if supported) to the
> > "test-meson-builds.sh" script. The tricky b
It's reasonably common for patches to have issues when built on 32-bits, so
to prevent this, we can add a 32-bit build (if supported) to the
"test-meson-builds.sh" script. The tricky bit is using a valid
PKG_CONFIG_LIBDIR, so for now we use two common possibilities for where that
should point to in
Hi,
> -Original Message-
> From: dev On Behalf Of Matan Azrad
> Sent: Tuesday, November 3, 2020 8:48 AM
> To: Slava Ovsiienko
> Cc: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH] net/mlx5: fix atomic API update
>
> The Txq refcnt 1 value means that there is not real reference to the
> queue
Hi,
> -Original Message-
> From: dev On Behalf Of Matan Azrad
> Sent: Tuesday, November 3, 2020 8:49 AM
> To: Slava Ovsiienko
> Cc: dev@dpdk.org; sta...@dpdk.org
> Subject: [dpdk-dev] [PATCH] net/mlx5: fix Tx queue stop state
>
> The Tx queue stop API doesn't call the PMD callback when t
> > Hi Juraj,
> >
> > I might be missing something, but I don't
> > believe these changes are sufficient to
> > enable vector mode for ARM32. For one
> > thing, bnxt_receive_function() in bnxt_ethdev.c
> > would need to be changed to enable the
> > selection of the vector receive function.
>
> Hi L
The fix of max_rx_pkt_len for allowing VLAN packets in all cases
was breaking configuration of some drivers. Example with virtio:
Ethdev port_id=0 max_rx_pkt_len 11229 > max valid value 9728
Fail to configure port 0
Trying to fix the logic was revealing other issues in some driver
In 'rte_eth_dev_configure()', if 'DEV_RX_OFFLOAD_JUMBO_FRAME' is not set
the max frame size is limited to 'RTE_ETHER_MAX_LEN' (1518).
This is mistake because for the PMDs that has frame size bigger than
"RTE_ETHER_HDR_LEN + RTE_ETHER_CRC_LEN" (18 bytes), the MTU becomes
less than 1500, causing a va
On 11/5/2020 5:44 PM, Thomas Monjalon wrote:
The fix of max_rx_pkt_len for allowing VLAN packets in all cases
was breaking configuration of some drivers. Example with virtio:
Ethdev port_id=0 max_rx_pkt_len 11229 > max valid value 9728
Fail to configure port 0
Trying to fix the
05/11/2020 19:11, Ferruh Yigit:
> On 11/5/2020 5:44 PM, Thomas Monjalon wrote:
> > The fix of max_rx_pkt_len for allowing VLAN packets in all cases
> > was breaking configuration of some drivers. Example with virtio:
> >
> > Ethdev port_id=0 max_rx_pkt_len 11229 > max valid value 9728
> >
05/11/2020 15:31, Ferruh Yigit:
> On 11/5/2020 9:51 AM, Steve Yang wrote:
> > If "dev_info.max_mtu" isn't set by driver, the rte_ethdev will set
> > default value "UINT16_MAX", that will cause the overhead value overflow.
> >
> > Add more strict condition to ensure the overhead valid.
> >
> > Fix
With this change, the bnxt driver fails to initialize under testpmd:
Configuring Port 0 (socket 0)
Port 0 failed to enable Rx offload JUMBO_FRAME
Fail to configure port 0
EAL: Error - exiting with code: 1
It appears that the cause is this bit of code in bnxt_ethdev.c:
if (bp->eth_dev->da
On Wed, Nov 4, 2020 at 12:03 PM Stephen Hemminger
wrote:
>
> 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
A new DPDK release candidate is ready for testing:
https://git.dpdk.org/dpdk/tag/?id=v20.11-rc3
There are only 9 new patches since yesterday -rc2, including
a revert which should unblock a lot of testing cases with testpmd.
Release notes:
http://doc.dpdk.org/guides/rel_notes/relea
> First the code cause problem in the driver looks in another place, following
> in
> 'bnxt_mtu_set_op()':
>
>if (new_mtu > RTE_ETHER_MTU) {
>bp->flags |= BNXT_FLAG_JUMBO;
>bp->eth_dev->data->dev_conf.rxmode.offloads |=
>
Existing field ID validity check does not validate the field
descriptor availability. Make it more rigorous to avoid
reading past the buffer containing field descriptors.
Coverity issue: 363742
Fixes: 370ed675a952 ("common/sfc_efx/base: support setting PPORT in match spec")
Signed-off-by: Ivan Ma
A set of patches to fix build with musl libc. I also did a few cleanups wrt
macros and fixed a few scary compiler warnings while at it.
Please note that those are only compile tested on x86_64 with musl libc.
v2 has some code style fixes reported by checkpatch
v3 is a full reroll of the entire p
Improve portability by avoid use non-standard 'uint'.
Use uint8_t for hash_key_len as rss_key_len is a uint8_t type.
This solves following build error when building with musl libc:
In file included from ../app/test-pmd/cmdline.c:74:
../app/test-pmd/testpmd.h:813:29: error: unknown type name 'uin
Improve portability by replacing non-standard 'uint' with 'unsigned int'
This solves the build error with musl libc:
In file included from ../drivers/net/cxgbe/cxgbe.h:9,
from ../drivers/net/cxgbe/cxgbe_ethdev.c:37:
../drivers/net/cxgbe/base/common.h:201:4: error: unknown type na
Add a fallback for non-GNU libc systems like musl libc for the
non-standard functions outl_p, outw_p and outb_p.
This solves the following buildtime errors when building with musl libc:
pci_uio.c:(.text+0xaa1): undefined reference to `outw_p'
pci_uio.c:(.text+0xac5): undefined reference to `outl_
Fix build with musl libc by using warn(3) instead of using the GNU
specific error(3).
`error(0, errno, ...)` is identical to `warn(...)` so replace:
error(0, ret, ...);
with the identical:
errno = ret;
warn(...);
We could have used the identical `warnx("...: %s", strerror(ret))` but
stre
There is no standard saying that __WORDSIZE should be be defined or in
what include it should be defined. Use RTE_ARCH_64 instead.
This solves a warning when building with musl libc:
warning: "__WORDSIZE" is not defined, evaluates to 0 [-Wundef]
Fixes: 847ee3bd0d1f ("bus/dpaa: support FMAN fram
The swab16/swab32/swab64 are Linux specific and not GNU libc specific.
Keep the check for __GLIBC__ just in case other GNU systems depends on
this (Hurd or GNU/kFreeBSD).
This fixes a build error with musl libc.
Fixes: 04711d41a872 ("crypto/dpaa2_sec: add run-time assembler for descriptor")
Cc: s
Solve the build error with musl libc:
../lib/librte_eal/unix/eal_file.c: In function 'eal_file_open':
../lib/librte_eal/unix/eal_file.c:22:15: error: 'O_RDONLY' undeclared
(first use in this function)
22 | sys_flags = O_RDONLY;
| ^~~~
Fixes: 176bb37ca6f3 ("eal: introd
Simplify pr_debug, pr_err and pr_warn macros by add an intermediate
pr_msg macro. This way we only need test for SUPPRESS_PRINTS once.
Signed-off-by: Natanael Copa
Acked-by: Hemant Agrawal
---
drivers/common/dpaax/caamflib/compat.h | 28 +++---
1 file changed, 12 insertions(
Fix memory corruption bug caused by passing address of a 16b value,
instead of address of a 32b value.
Fixes: 1998008 ("event/dlb: add eventdev probe")
Coverity issue: 363715
Cc: sta...@dpdk.org
Signed-off-by: Timothy McDaniel
---
drivers/event/dlb/pf/dlb_main.c | 4 ++--
1 file changed, 2 inse
Remove nonproductive assignment of qmport->cq_depth.
Fixes: ee57517 ("event/dlb: add port setup")
Coverity issue: 363717
Cc: sta...@dpdk.org
Signed-off-by: Timothy McDaniel
---
drivers/event/dlb/dlb.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/event/dlb/dlb.c b/drivers/event/dlb
Add queue depth threshold and class of service sections
to DLB2 rst file.
Fixes:5433956 ("event/dlb2: add eventdev probe")
Cc: sta...@dpdk.org
Signed-off-by: Timothy McDaniel
---
doc/guides/eventdevs/dlb2.rst | 47 +++
1 file changed, 47 insertions(+)
di
Add NULL check before using t->mbuf_pool.
Fixes: 6f1b828 ("event/dlb2: add self-tests")
Coverity issue: 363719
Cc: sta...@dpdk.org
Signed-off-by: Timothy McDaniel
---
drivers/event/dlb2/dlb2_selftest.c | 7 ++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/event/dlb2/
When an age action becomes aged-out the rte_flow_get_aged_flows should
return the action context supplied by the configuration structure.
In case the age action created by the shared action API, the shared
action context of the Testpmd application was not set.
In addition, the application handler
On Thu, Nov 5, 2020 at 12:51 PM Thomas Monjalon wrote:
>
> The fix of max_rx_pkt_len for allowing VLAN packets in all cases
> was breaking configuration of some drivers. Example with virtio:
>
> Ethdev port_id=0 max_rx_pkt_len 11229 > max valid value 9728
> Fail to configure port 0
This patch series changes the use of the terms blacklist and whitelist
and instead uses block and allow. The change is most visible
in the options passed to DPDK programs for marking which devices
to use.
The options are also shortened because they can be used for
not just PCI devices. When DPDK
This patch renames the enum values in the EAL include files.
As a backward compatible temporary migration tool, define
a replacement mapping for old values.
The old names relating to blacklist and whitelist are replaced
by block list and allow list, but applications may be using the
older compatib
Use the new terminology blocked to describe when devices
are excluded from being used.
Signed-off-by: Stephen Hemminger
Acked-by: Luca Boccassi
Acked-by: Hemant Agrawal
---
drivers/bus/dpaa/dpaa_bus.c| 7 +++
drivers/bus/fslmc/fslmc_bus.c | 9 -
drivers/bus/fslmc/fsl
Replace -w / --pci-whitelist with -a / --allow options
and --pci-blacklist with --block.
The -b short option remains unchanged.
Allow the old options for now, but print a nag
warning since old options are deprecated.
Signed-off-by: Stephen Hemminger
Acked-by: Luca Boccassi
---
lib/librte_eal/c
Remove the deprecation notice and add description to the release notes.
Signed-off-by: Stephen Hemminger
---
doc/guides/rel_notes/deprecation.rst | 23 ---
doc/guides/rel_notes/release_20_11.rst | 11 +++
2 files changed, 11 insertions(+), 23 deletions(-)
diff --gi
Test the renamed blocklist and allowlist arguments.
Use new terms in test variable names as well.
Signed-off-by: Stephen Hemminger
Acked-by: Luca Boccassi
---
app/test/autotest.py| 16 +--
app/test/autotest_runner.py | 18 ++---
app/test/test.c | 2 +-
app/t
There are two areas where documentation needed update.
The first was use of whitelist when describing address
filtering.
The other is the legacy -w whitelist option for PCI
which is used in many examples
Signed-off-by: Stephen Hemminger
Acked-by: Luca Boccassi
---
doc/guides/cryptodevs/dpaa2_s
> > > > > > > > >> 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:
> > > > > > > > >>>
> > > > > > > > >>
Numbers of descriptors configured is returned to a user
via rxq_info_get API. This number is incorrect for MPRQ.
For SPRQ this number matches the number of mbufs allocated.
For MPRQ we have fewer external MPRQ buffers that can hold
multiple packets in strides os this big buffer. Take that
into acco
There is a performance penalty for the replenish scheme
used in vectorized Rx burst for both MPRQ and SPRQ.
Mbuf elements are being filled at the end of the mbufs
array and being replenished at the beginning. That leads
to an increase in cache misses and the performance drop.
The more Rx descriptor
> -Original Message-
> From: Dharmik Thakkar
> Sent: Thursday, November 5, 2020 11:33 PM
> To: Ruifeng Wang
> Cc: dev@dpdk.org; nd ; Dharmik Thakkar
>
> Subject: [PATCH 2/2] doc: update build guide for armv8 crypto PMD
>
> Remove build steps about using meson option armv8_crypto_dir.
> -Original Message-
> From: Dharmik Thakkar
> Sent: Thursday, November 5, 2020 11:33 PM
> To: Ruifeng Wang ; Bruce Richardson
>
> Cc: dev@dpdk.org; nd ; Dharmik Thakkar
>
> Subject: [PATCH 1/2] crypto/armv8: update meson build
>
> With pkg-config support available within AArch64crypt
Remove rte_panic(), and use rte_atomic_thread_fence()
instead of rte_smp_[r/w]mb.
Signed-off-by: Jiawen Wu
---
drivers/net/txgbe/base/txgbe_osdep.h | 2 +-
drivers/net/txgbe/txgbe_rxtx.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/txgbe/base/txgbe_osd
On Thu, Nov 5, 2020 at 7:18 AM Lance Richardson <
lance.richard...@broadcom.com> wrote:
> On Wed, Nov 4, 2020 at 12:03 PM Stephen Hemminger
> wrote:
> >
> > Prefetching only helps performance if it is done several 100
> > instructions before the actual use. The purpose of the prefetch
> > is to r
Update release note for large VF, supporting up to 256 queue pairs per VF.
Signed-off-by: Ting Xu
---
doc/guides/rel_notes/release_20_11.rst | 1 +
1 file changed, 1 insertion(+)
diff --git a/doc/guides/rel_notes/release_20_11.rst
b/doc/guides/rel_notes/release_20_11.rst
index 57e4b13ec0..e882
Hi Maxime,
> -Original Message-
> From: Maxime Coquelin
> Sent: Thursday, November 5, 2020 7:46 PM
> To: dev@dpdk.org; Ding, Xuan ;
> step...@networkplumber.org; Yigit, Ferruh ;
> tho...@monjalon.net; Xia, Chenbo
> Cc: sta...@dpdk.org; Maxime Coquelin
> Subject: [PATCH 1/3] vhost: fix e
Hi Maxime,
> -Original Message-
> From: Maxime Coquelin
> Sent: Thursday, November 5, 2020 7:46 PM
> To: dev@dpdk.org; Ding, Xuan ;
> step...@networkplumber.org; Yigit, Ferruh ;
> tho...@monjalon.net; Xia, Chenbo
> Cc: sta...@dpdk.org; Maxime Coquelin
> Subject: [PATCH 2/3] vhost: fix f
> -Original Message-
> From: Maxime Coquelin
> Sent: Thursday, November 5, 2020 7:46 PM
> To: dev@dpdk.org; Ding, Xuan ;
> step...@networkplumber.org; Yigit, Ferruh ;
> tho...@monjalon.net; Xia, Chenbo
> Cc: sta...@dpdk.org; Maxime Coquelin
> Subject: [PATCH 3/3] vhost: fix fd leak in ki
> -Original Message-
> From: Steve Yang
> Sent: Thursday, November 5, 2020 4:32 PM
> To: dev@dpdk.org
> Cc: Yang, Qiming ; Xing, Beilei
> ; Wu, Jingjing ; Yang, SteveX
>
> Subject: [PATCH v1] net/iavf: fix reset VF doesn't notify PF to reset
>
> When VF closed, the VF should notify PF
The entry variable assert in the mlx5_hlist_register() function is not
correct. Remove the invalid entry variable.
Fixes: e69a59227db0 ("net/mlx5: support concurrent access for hash list")
Signed-off-by: Suanming Mou
---
drivers/net/mlx5/mlx5_utils.c | 2 +-
1 file changed, 1 insertion(+), 1 de
As shared RSS action is global standalone action, the shared action
should be maintained only by the shared action management functions.
Currently, the shared RSS action hrxq uses the same fate action type
with existed general queues, and implicitly uses the deprecate refcnt
increment to indicate
Add args length check before copying to fix the coverity issue.
Fixes: 3a04ecb21420 ("examples/vhost: add async vhost args parsing")
Coverity issue: 363741
Signed-off-by: Cheng Jiang
---
examples/vhost/main.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/examples/vhost/main.c b/exam
Add checking return value of string split function to fix the
coverity issue.
Fixes: 3a04ecb21420 ("examples/vhost: add async vhost args parsing")
Coverity issue: 363739
Signed-off-by: Cheng Jiang
---
examples/vhost/ioat.c | 11 ---
1 file changed, 8 insertions(+), 3 deletions(-)
diff
Here series include four coverity tool warnings fixes
and a bug fixes.
V1->V2:
1. fix some warnings by checkpatch tool.
Chengchang Tang (1):
net/hns3: fix queue enabling status not store after FLR
Hongbo Zheng (3):
net/hns3: use correct logging format symbol
net/hns3: use unsigned types fo
From: Hongbo Zheng
According to bit operator reliability style, variables in
the right expression participating int bit operation
cannot be of unsigned type.
Signed-off-by: Hongbo Zheng
Signed-off-by: Lijun Ou
---
drivers/net/hns3/hns3_ethdev_vf.c | 2 +-
drivers/net/hns3/hns3_rxtx_vec_n
From: Hongbo Zheng
In current driver print log function, some print format
symbols does not match with the actual variable types.
Signed-off-by: Hongbo Zheng
Signed-off-by: Lijun Ou
---
drivers/net/hns3/hns3_dcb.c | 10 +-
drivers/net/hns3/hns3_ethdev.c| 32 -
From: Chengchang Tang
The FLR will resets the queue enabling status. In the
current code, the queue enabling status is not restored
after the reset. Therefore, if upper layer users have
called queue start/stop function before the reset, the
behavior after the reset is not as expected.
This patch
From: Hongbo Zheng
Here adjusts some code style for making the lines more
compact and removes some static check tool warnings.
Signed-off-by: Hongbo Zheng
Signed-off-by: Lijun Ou
---
V1->V2:
- fix checkpatch warning
---
drivers/net/hns3/hns3_cmd.c| 1 -
drivers/net/hns3/hns3_ethdev.c |
Here adds a check for the return value when calling
rte_pci_write_config.
Coverity issue: 363714
Fixes: cea37e513329 ("net/hns3: fix FLR reset")
Cc: sta...@dpdk.org
Signed-off-by: Lijun Ou
---
V1->V2:
- rte_pci_wirte_config -> rte_pci_write_config
---
drivers/net/hns3/hns3_ethdev_vf.c | 10
On Thu, Oct 29, 2020 at 02:47:43PM +0800, yang_y...@163.com wrote:
> From: Yi Yang
>
> Many NICs can't offload VXLAN UFO, so it is very important
> to do VXLAN UDP GSO by software to improve VM-to-VM UDP
> performance, especially for the case that VM MTU is just
> 1500 but not 9000.
>
> With thi
>
> +Cc Konstantin and Honnappa for guidance
>
> 05/11/2020 09:55, Jiawen Wu:
> > On Thursday, November 5, 2020 9:55 AM, Jiawen Wu wrote:
> > > On Thursday, November 5, 2020 1:24 AM, Ferruh Yigit wrote:
> > > > On 11/3/2020 11:08 PM, Thomas Monjalon wrote:
> > > > > When pulling in the main bra
The configuration of FDIR flex mask and flex pit should not be set
during flow validate. It should be set when flow create.
Fixes: 6ced3dd72f5f ("net/i40e: support flexible payload parsing for FDIR")
Cc: sta...@dpdk.org
Signed-off-by: Chenxu Di
---
v2:
-Merge two patches into one patch.
---
dri
> From: Ananyev, Konstantin [mailto:konstantin.anan...@intel.com]
> Sent: Friday, November 6, 2020 12:55 AM
>
> > > > > > > > > >> Hi Olivier,
> > > > > > > > > >>
> > > > > > > > > >>> m->nb_seg must be reset on mbuf free whatever the
> value
> > > of m->next,
> > > > > > > > > >>> because it can
101 - 171 of 171 matches
Mail list logo
