> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Stephen Hemminger
> Sent: Saturday, May 18, 2019 02:47
> To: dev@dpdk.org
> Subject: [dpdk-dev] Instability of port ids
>
> Several customers have reported similar issues with how the owned/stack
> device model
>
https://bugs.dpdk.org/show_bug.cgi?id=281
Bug ID: 281
Summary: BPF: Linking error in librte_bpf
Product: DPDK
Version: 19.08
Hardware: x86
OS: Linux
Status: CONFIRMED
Severity: normal
Priority: N
On 2019-05-17 21:27, Neil Horman wrote:
diff --git a/lib/librte_eal/common/rte_random.c
b/lib/librte_eal/common/rte_random.c
new file mode 100644
index 0..4d3cf5226
--- /dev/null
+++ b/lib/librte_eal/common/rte_random.c
@@ -0,0 +1,139 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Cop
On Fri, May 17, 2019 at 03:27:25PM -0400, Neil Horman wrote:
> On Tue, May 14, 2019 at 04:53:08PM +0200, Mattias Rönnblom wrote:
> > On 2019-05-14 16:16, Neil Horman wrote:
> > > On Tue, May 14, 2019 at 11:20:41AM +0200, Mattias Rönnblom wrote:
> > > > This commit replaces rte_rand()'s use of lrand
On Tue, May 14, 2019 at 04:53:08PM +0200, Mattias Rönnblom wrote:
> On 2019-05-14 16:16, Neil Horman wrote:
> > On Tue, May 14, 2019 at 11:20:41AM +0200, Mattias Rönnblom wrote:
> > > This commit replaces rte_rand()'s use of lrand48() with a DPDK-native
> > > combined Linear Feedback Shift Register
From: Liron Himi
Signed-off-by: Liron Himi
Reviewed-by: Jerin Jacob Kollanukkaran
Tested-by: Liron Himi
---
config/arm/arm64_armada_linux_gcc | 16
config/arm/meson.build| 8 +++-
2 files changed, 23 insertions(+), 1 deletion(-)
create mode 100644 config/arm
From: Liron Himi
This patch introduces armada target to address difference
in number of cores, no numa support
Signed-off-by: Liron Himi
Reviewed-by: Alan Winkowski
Tested-by: Liron Himi
---
config/defconfig_arm64-armada-linux-gcc| 24
config/defconfig_arm64-arma
From: Liron Himi
Introduces armada target to address difference
in number of cores, no numa support
Liron Himi (2):
mk: add Marvell ARMADA architecture based on armv8-a
meson: add Marvell ARMADA architecture based on armv8-a
config/arm/arm64_armada_linux_gcc | 16
From: Liron Himi
Introduces armada target to address difference
in number of cores, no numa support
Liron Himi (2):
mk: add Marvell ARMADA architecture based on armv8-a
meson: add Marvell ARMADA architecture based on armv8-a
config/arm/arm64_armada_linux_gcc | 16
From: Liron Himi
Change-Id: If210dbfeed447df3beffd53339a40e4a9069b6dc
Signed-off-by: Liron Himi
Reviewed-by: Jerin Jacob Kollanukkaran
---
config/arm/arm64_armada_linux_gcc | 16
config/arm/meson.build| 8 +++-
2 files changed, 23 insertions(+), 1 deletion(-)
From: Liron Himi
This patch introduces armada target to address difference
in number of cores, no numa support
Change-Id: Ieaabe77b165000fe1280f98105329a0e63833b92
Signed-off-by: Liron Himi
Reviewed-by: Alan Winkowski
Tested-by: Liron Himi
---
config/defconfig_arm64-armada-linux-gcc| 24 ++
Several customers have reported similar issues with how the owned/stack device
model
works in DPDK. With failsafe/tap and VF or netvsc and VF there are DPDK ports
which
are marked as owned and therefore not visible.
The problem is the application has to guess and workaround these port values in
Add new telemetry mode support for l3fwd-power.
This is a standalone mode, in this mode l3fwd-power
does simple l3fwding along with calculating
empty polls, full polls, and busy percentage for
each forwarding core. The aggregation of these
values of all cores is reported as application
level teleme
telemetry has support for fetching port based stats
from metrics library.
Metrics library also has global stats which are
not fetched by telemetry, so extend telemetry to
fetch the global metrics.
Signed-off-by: Reshma Pattan
---
doc/guides/howto/telemetry.rst| 9 +-
doc/guide
Add support for RFC 4301(5.1.2) to update of
Type of service field and Traffic class field
bits inside ipv4/ipv6 packets for outbound cases
and inbound cases which deals with the update of
the DSCP/ENC bits inside each of the fields.
Signed-off-by: Marko Kovacevic
---
examples/ipsec-secgw/sa.c
On 5/5/19 11:02 AM, Li Lin wrote:
From: Li Lin
This patch introduces two new messages VHOST_USER_GET_INFLIGHT_FD
and VHOST_USER_SET_INFLIGHT_FD to support transferring a shared
buffer between qemu and backend.
Firstly, qemu uses VHOST_USER_GET_INFLIGHT_FD to get the
shared buffer from backe
On 5/9/19 10:03 AM, David Marchand wrote:
This .rx_queue_setup devop is called after ethdev already dereferenced
the mempool pointer.
No need to check and we can remove this rte_exit.
Fixes: 48cec290a3d2 ("net/virtio: move queue configure code to proper place")
Cc: sta...@dpdk.org
Signed-off
Some helpers in the header file are forced inlined other are
only inlined, this patch forces inline for all.
It will avoid it to be embedded as functions when called multiple
times in the same object file. For example, when we added packed
ring support in vhost-user library, rte_memcpy_generic got
Handling of fragmented virtio-net header and indirect descriptors
tables was implemented to fix CVE-2018-1059. It should never
happen with healthy guests and so are already considered as
unlikely code path.
This patch moves these bits into non-inline dedicated functions
to reduce the I-cache press
Now that we have a single function to map the descriptors
buffers, let's prefetch them there as it is the earliest
place we can do it.
Signed-off-by: Maxime Coquelin
---
lib/librte_vhost/virtio_net.c | 32 ++--
1 file changed, 2 insertions(+), 30 deletions(-)
diff --
At runtime either packed Tx/Rx functions will always be called,
or split Tx/Rx functions will always be called.
This patch removes the forced inlining in order to reduce
the I-cache pressure.
Signed-off-by: Maxime Coquelin
---
lib/librte_vhost/virtio_net.c | 8
1 file changed, 4 insert
In order to reduce the I-cache pressure, this patch removes
the inlining of the dirty pages logging functions, that we
can consider as cold path.
Indeed, these functions are only called while doing live
migration, so not called most of the time.
Signed-off-by: Maxime Coquelin
---
lib/librte_vho
Some OVS-DPDK PVP benchmarks show a performance drop
when switching from DPDK v17.11 to v18.11.
With the addition of packed ring layout support,
rte_vhost_enqueue_burst and rte_vhost_dequeue_burst
became very large, and only a part of the instructions
are executed (either packed or split ring used
On 5/17/19 3:00 PM, David Marchand wrote:
On Fri, May 17, 2019 at 2:23 PM Maxime Coquelin
mailto:maxime.coque...@redhat.com>> wrote:
At runtime either packed Tx/Rx functions will always be called,
or split Tx/Rx functions will always be called.
This patch removes the forced in
On 5/17/19 3:04 PM, David Marchand wrote:
On Fri, May 17, 2019 at 2:23 PM Maxime Coquelin
mailto:maxime.coque...@redhat.com>> wrote:
Some OVS-DPDK PVP benchmarks show a performance drop
when switching from DPDK v17.11 to v18.11.
With the addition of packed ring layout support
clang 6.0 and onwards, for the external function call generates
BPF_PSEUDO_CALL instruction:
call pseudo +-off -> call another bpf function.
More details about that change: https://lwn.net/Articles/741773/
DPDK BPF implementation right now doesn't support multiple BPF
functions per module.
To overc
Hello Anatoly,
Not sure what the issue is at the moment.
I too have an error on the eal_flags_autotest but what is worrying me is
that it works fine in 18.11.1 and 19.02.
Is there anything that rings a bell?
Thanks.
--
David Marchand
On Fri, 2019-05-17 at 14:45 +0100, Bruce Richardson wrote:
> Small cleanups to shorten the meson.build file in the kernel/linux
> directory, removing the unneeded version check and taking advantage
> of the
> subdir_done() function introduced in meson 0.46.
>
> V2:
> resubmit without the 19.
Since meson 0.46, meson has supported the subdir_done() function, which
allows us to abort processing of a file early. Using this we can reduce the
indentation in our files by eliminating unnecessary else blocks.
Signed-off-by: Bruce Richardson
Reviewed-by: David Marchand
---
kernel/linux/meson
Small cleanups to shorten the meson.build file in the kernel/linux
directory, removing the unneeded version check and taking advantage of the
subdir_done() function introduced in meson 0.46.
V2:
resubmit without the 19.08 patch prefix and with the typo
removed in patch 1.
Bruce Ri
The check for meson version 0.44 is redundant since the minimum
supported version for the project as a whole is 0.47.
Signed-off-by: Bruce Richardson
Reviewed-by: David Marchand
---
kernel/linux/meson.build | 7 +--
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/kernel/linux/m
The local variables for the error message aren't needed, since the messages
aren't used more than once, and the indent levels are now such that the
lines printing the message are not much longer than the lines defining the
variables to hold the messages themselves. Therefore the use of the
variable
On Fri, 2019-05-17 at 12:47 +0100, Bruce Richardson wrote:
> This patchset adds support for testing a number of examples using the
> DPDK pkgconfig file, allowing us to check that the file works. It
> also
> includes some more general cleanups for the examples, including
> adding
> support for buil
On Fri, 2019-05-17 at 13:02 +0100, Bruce Richardson wrote:
> A small set of fixes for building examples, that was previously part
> of
> the larger set, but pulled out separately for easier apply.
>
> V3: changed patch 1, to properly quit processing examples/meson.build
> if the examples list
On Fri, May 17, 2019 at 2:23 PM Maxime Coquelin
wrote:
> Some OVS-DPDK PVP benchmarks show a performance drop
> when switching from DPDK v17.11 to v18.11.
>
> With the addition of packed ring layout support,
> rte_vhost_enqueue_burst and rte_vhost_dequeue_burst
> became very large, and only a par
On Fri, May 17, 2019 at 2:23 PM Maxime Coquelin
wrote:
> At runtime either packed Tx/Rx functions will always be called,
> or split Tx/Rx functions will always be called.
>
> This patch removes the forced inlining in order to reduce
> the I-cache pressure.
>
I just wonder if the compiler can't d
On 5/17/19 2:22 PM, Maxime Coquelin wrote:
Handling of fragmented virtio-net header and indirect descriptors
tables was implemented to fix CVE-2018-1059. It should not never
happen with healthy guests and so are already considered as
unlikely code path.
This patch moves these bits into non-in
From: root
Some helpers in the header file are forced inlined other are
only inlined, this patch forces inline for all.
It will avoid it to be embedded as functions when called multiple
times in the same object file. For example, when we added packed
ring support in vhost-user library, rte_memcp
Now that we have a single function to map the descriptors
buffers, let's prefetch them there as it is the earliest
place we can do it.
Signed-off-by: Maxime Coquelin
---
lib/librte_vhost/virtio_net.c | 32 ++--
1 file changed, 2 insertions(+), 30 deletions(-)
diff --
At runtime either packed Tx/Rx functions will always be called,
or split Tx/Rx functions will always be called.
This patch removes the forced inlining in order to reduce
the I-cache pressure.
Signed-off-by: Maxime Coquelin
---
lib/librte_vhost/virtio_net.c | 8
1 file changed, 4 insert
Some OVS-DPDK PVP benchmarks show a performance drop
when switching from DPDK v17.11 to v18.11.
With the addition of packed ring layout support,
rte_vhost_enqueue_burst and rte_vhost_dequeue_burst
became very large, and only a part of the instructions
are executed (either packed or split ring used
In order to reduce the I-cache pressure, this patch removes
the inlining of the dirty pages logging functions, that we
can consider as cold path.
Indeed, these functions are only called while doing live
migration, so not called most of the time.
Signed-off-by: Maxime Coquelin
---
lib/librte_vho
Handling of fragmented virtio-net header and indirect descriptors
tables was implemented to fix CVE-2018-1059. It should not never
happen with healthy guests and so are already considered as
unlikely code path.
This patch moves these bits into non-inline dedicated functions
to reduce the I-cache p
The "make clean" command had a number of issues:
- the "--ignore-fail-on-non-empty" flag is not present on BSD
- the call to remove the build folder would fail if there was no build
folder present.
These are fixed by only removing the build folder if it exists, and by
using -p flag to rmdir in p
The examples/meson.build file scanned the filesystem to find all examples
to build (for examples=all option) and install. However, using run_command
and scanning the filesystem prevented ninja from properly detecting the
addition or removal of any examples - one had to recreate the build
directory
A small set of fixes for building examples, that was previously part of
the larger set, but pulled out separately for easier apply.
V3: changed patch 1, to properly quit processing examples/meson.build
if the examples list was empty.
Bruce Richardson (3):
examples: fix install of sample app
While the examples were being installed into the appropriate install path
when processing the examples/meson.build file, that file was only processed
if the "examples" meson parameter was non-empty. Since we now do more than
just build the examples, we need to unconditionally process the file.
Fix
On Fri, May 17, 2019 at 12:46:11PM +0100, Bruce Richardson wrote:
> While the examples were being installed into the appropriate install path
> when processing the examples/meson.build file, that file was only processed
> if the "examples" meson parameter was non-empty. Since we now do more than
>
The vdpa example app did not check for a libdpdk pkg-config file and
attempt to build using that. Add support for that method of compile to
align the app with the other examples.
Signed-off-by: Bruce Richardson
---
examples/vdpa/Makefile | 52 +++---
1 file ch
For testing of DPDK, we want to override the prefix given by the
pkg-config file, so that we can get correct paths for DPDK installed
in an unusual location.
Signed-off-by: Bruce Richardson
---
examples/bbdev_app/Makefile | 10 ++
examples/bond/Makefile | 10
Allow the script to run with a reduced set of builds if clang, or
other compilers, are missing.
Signed-off-by: Bruce Richardson
---
devtools/test-meson-builds.sh | 38 ++-
1 file changed, 20 insertions(+), 18 deletions(-)
diff --git a/devtools/test-meson-builds.s
The vhost_crypto example app did not check for a libdpdk pkg-config file
and attempt to build using that. Add support for that method of compile to
align the app with the other examples.
Signed-off-by: Bruce Richardson
---
examples/vhost_crypto/Makefile | 52 --
1
The pkg-config file generated as part of the build of DPDK should allow
applications to be built with an installed DPDK. We can test this as
part of the build by doing an install of DPDK to a temporary directory
within the build folder, and by then compiling up a few sample apps
using make working
While the examples were being installed into the appropriate install path
when processing the examples/meson.build file, that file was only processed
if the "examples" meson parameter was non-empty. Since we now do more than
just build the examples, we need to unconditionally process the file.
Fix
The examples/meson.build file scanned the filesystem to find all examples
to build (for examples=all option) and install. However, using run_command
and scanning the filesystem prevented ninja from properly detecting the
addition or removal of any examples - one had to recreate the build
directory
The "make clean" command had a number of issues:
- the "--ignore-fail-on-non-empty" flag is not present on BSD
- the call to remove the build folder would fail if there was no build
folder present.
These are fixed by only removing the build folder if it exists, and by
using -p flag to rmdir in p
This patchset adds support for testing a number of examples using the
DPDK pkgconfig file, allowing us to check that the file works. It also
includes some more general cleanups for the examples, including adding
support for building from pkg-config to a couple of examples which were
missing that.
A small set of fixes for building examples, that was previously part of
the larger set, but pulled out separately for easier apply.
Bruce Richardson (3):
examples: fix install of sample apps if setting not provided
examples: remove auto-generation of examples list
examples: fix make clean wh
On Fri, May 17, 2019 at 10:15:08AM +0200, David Marchand wrote:
We currently have no check on the mempool pointer passed to
rte_eth_rx_queue_setup.
Let's avoid a plain crash when dereferencing it.
Suggested-by: Jens Freimann
Signed-off-by: David Marchand
---
lib/librte_ethdev/rte_ethdev.c | 5
On Thu, May 16, 2019 at 02:28:03PM -0400, David Harton wrote:
> Use of weak symbols can hide makefile errors especially when
> custom makefiles are used. Removing the use of weak symbols
> to avoid a stub function being linked in production code.
>
> Signed-off-by: David Harton
> ---
>
> v3 - a
On Fri, 2019-05-17 at 10:43 +0200, Thomas Monjalon wrote:
> 16/05/2019 23:43, Luca Boccassi:
> > On Thu, 2019-05-16 at 18:02 +0100, Ferruh Yigit wrote:
> > > * Can be good to build contact information of the involved
> > > testers
> > > from
> > > various companies, for faster/better communic
17/05/2019 10:55, Nithin Dabilpuram:
> On Wed, May 15, 2019 at 09:27:22AM +0200, Thomas Monjalon wrote:
> > 15/05/2019 08:52, Nithin Dabilpuram:
> > > Hi Thomas,
> > > On Tue, May 14, 2019 at 05:39:30PM +0200, Thomas Monjalon wrote:
> > > > Hi,
> > > >
> > > > 13/05/2019 13:21, Nithin Dabilpuram:
On Wed, May 15, 2019 at 09:27:22AM +0200, Thomas Monjalon wrote:
> 15/05/2019 08:52, Nithin Dabilpuram:
> > Hi Thomas,
> > On Tue, May 14, 2019 at 05:39:30PM +0200, Thomas Monjalon wrote:
> > > Hi,
> > >
> > > 13/05/2019 13:21, Nithin Dabilpuram:
> > > > With the latest published interface of
> >
16/05/2019 23:43, Luca Boccassi:
> On Thu, 2019-05-16 at 18:02 +0100, Ferruh Yigit wrote:
> > * Can be good to build contact information of the involved testers
> > from
> > various companies, for faster/better communication, if you are
> > doing LTS
> > testing please reach out to Kevin
On Thu, May 9, 2019 at 3:29 PM Jens Freimann wrote:
> On Thu, May 09, 2019 at 10:03:48AM +0200, David Marchand wrote:
> >This .rx_queue_setup devop is called after ethdev already dereferenced
> >the mempool pointer.
> >No need to check and we can remove this rte_exit.
> >
> >Fixes: 48cec290a3d2 (
We currently have no check on the mempool pointer passed to
rte_eth_rx_queue_setup.
Let's avoid a plain crash when dereferencing it.
Suggested-by: Jens Freimann
Signed-off-by: David Marchand
---
lib/librte_ethdev/rte_ethdev.c | 5 +
lib/librte_ethdev/rte_ethdev.h | 6 +++---
2 files changed
On 17/05/2019 00:23, Stephen Hemminger wrote:
> On Thu, 16 May 2019 23:12:45 +0200
> Thomas Monjalon wrote:
>
>> 16/05/2019 20:57, Kevin Traynor:
>>> Plan is to have a 18.11.2 RC1 ready early next week for validation
>>> teams. There are a small few outstanding backports that need authors
>>> hel
On 16/05/2019 22:12, Thomas Monjalon wrote:
> 16/05/2019 20:57, Kevin Traynor:
>> Plan is to have a 18.11.2 RC1 ready early next week for validation
>> teams. There are a small few outstanding backports that need authors
>> help - please send by next Monday so I can put into RC1.
>
> Please could
On Thu, May 16, 2019 at 8:05 PM Stephen Hemminger <
step...@networkplumber.org> wrote:
> Make a function that coresponds with eth_aton_r which can
>
corresponds
be used to convert string to ether_addr.
>
> This also allows rte_ethdev to no longer depend on the
> cmdline library.
>
> Signed-off-b
On Thu, May 16, 2019 at 8:05 PM Stephen Hemminger <
step...@networkplumber.org> wrote:
> Formatting ethernet address and getting a random value are
> not in critical path so they should not be inlined.
>
> Signed-off-by: Stephen Hemminger
>
Same comments as on the rfc patch.
Please have a look.
70 matches
Mail list logo
