-Original Message-
From: Stephen Hemminger
Sent: Friday, March 10, 2023 10:15 PM
To: Ferruh Yigit
Cc: Dumitrescu, Cristian ; Jangra, Yogesh
; Singh, Aman Deep ; Zhang,
Yuying ; dev@dpdk.org; R, Kamalakannan
; Suresh Narayane, Harshad
Subject: Re: [PATCH] app/testpmd: fix closing so
On Thu, Mar 16, 2023 at 10:14:56PM +0300, Pavel Ivashchenko wrote:
> How to reproduce:
>
> 1. Define RTE_LIBRTE_MBUF_DEBUG
> 2. MALLOC_PERTURB_=178 DPDK_TEST=mbuf_autotest gdb --args
> obj-x86_64-linux-gnu/app/test/dpdk-test --file-prefix=mbuf_autotest
>
>PANIC in rte_mbuf_sanity_check():
>
On Fri, Mar 17, 2023 at 9:03 AM Olivier Matz wrote:
> On Thu, Mar 16, 2023 at 10:14:56PM +0300, Pavel Ivashchenko wrote:
> > How to reproduce:
> >
> > 1. Define RTE_LIBRTE_MBUF_DEBUG
> > 2. MALLOC_PERTURB_=178 DPDK_TEST=mbuf_autotest gdb --args
> > obj-x86_64-linux-gnu/app/test/dpdk-test --file-p
This patch adds detection to avoid repeated calls to dev_reset.
And enables the iavf_dev_watchdog to detect some VF reset without PF event
(VIRTCHNL_EVENT_RESET_IMPENDING), one of which is after an NVM update.
Fixes: e74e1bb6280d ("net/iavf: enable port reset")
Fixes: 5e03e316c753 ("net/iavf: han
> -Original Message-
> From: Ye, MingjinX
> Sent: Friday, March 17, 2023 1:10 PM
> To: dev@dpdk.org
> Cc: Yang, Qiming ; sta...@dpdk.org; Zhou, YidingX
> ; Ye, MingjinX ; Zhang,
> Ke1X ; Zhang, Qi Z
> Subject: [PATCH v4] net/ice: fix ice dcf control thread crash
>
> The control thread
On Thu, Mar 16, 2023 at 1:07 AM Tyler Retzlaff
wrote:
>
> In rte_thread_create setting affinity after pthread_create may fail.
> Such a failure should result in the entire rte_thread_create failing
> but doesn't.
>
> Additionally if there is a failure to set affinity a race exists where
> the crea
Ferruh Yigit, Mar 16, 2023 at 18:05:
> Hi Anthony,
>
> What is the motivation here?
Hi Ferruh,
AFAIK testpmd is the reference tool used for CI and tests whether it is
for functional or performance tests and I think it would be in
everyone's interest to consume less CPU during them. Moreover, all
BEGIN:VCALENDAR
METHOD:REQUEST
PRODID:Microsoft Exchange Server 2010
VERSION:2.0
BEGIN:VTIMEZONE
TZID:China Standard Time
BEGIN:STANDARD
DTSTART:16010101T00
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:16010101T00
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
END:DAYLIGHT
E
BEGIN:VCALENDAR
METHOD:CANCEL
PRODID:Microsoft Exchange Server 2010
VERSION:2.0
BEGIN:VTIMEZONE
TZID:China Standard Time
BEGIN:STANDARD
DTSTART:16010101T00
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:16010101T00
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
END:DAYLIGHT
EN
From: Jerin Jacob
Starting from DPDK 23.07, the Marvell QLogic bnx2x driver
will be removed. This decision has been made to alleviate the burden of
maintaining a discontinued product.
Signed-off-by: Jerin Jacob
---
doc/guides/rel_notes/deprecation.rst | 3 +++
1 file changed, 3 insertions(+)
On Fri, Mar 17, 2023 at 11:45:08AM +0100, David Marchand wrote:
> On Thu, Mar 16, 2023 at 1:07 AM Tyler Retzlaff
> wrote:
> >
> > In rte_thread_create setting affinity after pthread_create may fail.
> > Such a failure should result in the entire rte_thread_create failing
> > but doesn't.
> >
> > A
The bus cleanup callback is used to sunset all devices on bus
gracefully. This patch supports the callback by unplug all
devices on auxiliary bus.
Signed-off-by: Xueming Li
---
drivers/bus/auxiliary/auxiliary_common.c | 22 ++
1 file changed, 22 insertions(+)
diff --git a/dr
The bus cleanup callback is used to sunset all devices on bus
gracefully. This patch supports the callback by unplug all
devices on auxiliary bus.
Signed-off-by: Xueming Li
---
drivers/bus/auxiliary/auxiliary_common.c | 21 +
1 file changed, 21 insertions(+)
diff --git a/dri
Hi Stephen,
Thank you for having a look at this.
On Wed, Mar 15, 2023 at 10:31:31AM -0700, Stephen Hemminger wrote:
> The cmdline library poll function is broken on Windows
> and was never tested, don't use it.
>
> Instead, use sigaction() to cancel read character on Unix OS's
> and a new helper
On Wed, Mar 15, 2023 at 10:31:32AM -0700, Stephen Hemminger wrote:
> The setting in terminal handling for both Unix style and
> Windows was not ensuring that Ctrl-C character would
> cause interrupt.
>
> This is a first release bug. Testpmd interactive mode has
> always disabled control-c handling
On Fri, 17 Mar 2023 12:09:04 +0100
"Anthony Harivel" wrote:
> Ferruh Yigit, Mar 16, 2023 at 18:05:
> > Hi Anthony,
> >
> > What is the motivation here?
>
> Hi Ferruh,
>
> AFAIK testpmd is the reference tool used for CI and tests whether it is
> for functional or performance tests and I think
On Fri, 17 Mar 2023 17:20:59 +0100
Olivier Matz wrote:
> On Wed, Mar 15, 2023 at 10:31:32AM -0700, Stephen Hemminger wrote:
> > The setting in terminal handling for both Unix style and
> > Windows was not ensuring that Ctrl-C character would
> > cause interrupt.
> >
> > This is a first release b
On Fri, 17 Mar 2023 17:20:48 +0100
Olivier Matz wrote:
> >
> > +#ifdef RTE_EXEC_ENV_WINDOWS
> > signal(SIGINT, signal_handler);
> > signal(SIGTERM, signal_handler);
> > +#else
> > + /* Want read() not to be restarted on signal */
> > + struct sigaction action = {
> > + .sa
The cmdline_poll() function is broken and was
not fully tested, don't use it.
Instead, use sigaction() to cancel read character on Unix OS's
and a new helper to cancel I/O on Windows.
Fixes: 0fd1386c30c3 ("app/testpmd: cleanup cleanly from signal")
Bugzilla ID: 1180
Signed-off-by: Stephen Hemming
> -Original Message-
> From: jer...@marvell.com
> Sent: 17 March 2023 18:00
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; david.march...@redhat.com; ferruh.yi...@amd.com;
> andrew.rybche...@oktetlabs.ru; Alok Prasad
> ; Devendra Singh Rawat ; Jerin
> Jacob Kollanukkaran
> Subject: [dpdk
On Fri, Mar 17, 2023 at 3:50 PM Tyler Retzlaff
wrote:
> > > -struct thread_routine_ctx {
> > > +struct thread_start_context {
> > > rte_thread_func thread_func;
> > > - void *routine_args;
> > > + void *thread_args;
> > > + const rte_thread_attr_t *thread_attr;
> > > +
From: Tyler Retzlaff
In rte_thread_create setting affinity after pthread_create may fail.
Such a failure should result in the entire rte_thread_create failing
but doesn't.
Additionally if there is a failure to set affinity a race exists where
the creating thread will free ctx and depending on sc
Anthony Harivel, Mar 16, 2023 at 16:14:
> Sleep for an incremental amount of time if the fwd engine has processed
> less than at least half a burst of packets (i.e 16pkts with default
> setting) on a polling iteration of testpmd.
>
> Upon detecting the threshold of >= 16 pkts on an Rxq, reset the s
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
This series covers the libraries and drivers that are built on Windows.
The code has be converted to use the __atomic builtins but there are
additional during conversion i notice that there
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
---
lib/ring/rte_ring_core.h| 1 -
lib/ring/rte_ring_generic_pvt.h | 10 ++
2 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/lib
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
---
drivers/dma/idxd/idxd_internal.h | 3 +--
drivers/dma/idxd/idxd_pci.c | 6 +++---
2 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/drive
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
---
lib/stack/rte_stack_lf_generic.h | 11 +--
1 file changed, 5 insertions(+), 6 deletions(-)
diff --git a/lib/stack/rte_stack_lf_generic.h b/lib/sta
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
---
drivers/net/ice/ice_dcf.c| 1 -
drivers/net/ice/ice_dcf_ethdev.c | 1 -
drivers/net/ice/ice_ethdev.c | 10 ++
3 files changed, 6 inse
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
---
drivers/net/null/rte_eth_null.c | 20 ++--
1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/drivers/net/null/rte_eth_null.c
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
---
drivers/net/ring/rte_eth_ring.c | 20 ++--
1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ring/rte_eth_ring.c
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
---
drivers/net/ixgbe/ixgbe_bypass.c | 1 -
drivers/net/ixgbe/ixgbe_ethdev.c | 12 ++--
drivers/net/ixgbe/ixgbe_ethdev.h | 3 ++-
drivers/net/ixgbe/i
On Fri, Mar 17, 2023 at 01:19:42PM -0700, Tyler Retzlaff wrote:
> Replace the use of rte_atomic.h types and functions, instead use GCC
> supplied C++11 memory model builtins.
>
> Signed-off-by: Tyler Retzlaff
> ---
> lib/ring/rte_ring_core.h| 1 -
> lib/ring/rte_ring_generic_pvt.h | 10
On Fri, Mar 17, 2023 at 01:19:45PM -0700, Tyler Retzlaff wrote:
> Replace the use of rte_atomic.h types and functions, instead use GCC
> supplied C++11 memory model builtins.
>
> Signed-off-by: Tyler Retzlaff
> ---
> drivers/net/ice/ice_dcf.c| 1 -
> drivers/net/ice/ice_dcf_ethdev.c |
On Fri, Mar 17, 2023 at 01:19:47PM -0700, Tyler Retzlaff wrote:
> Replace the use of rte_atomic.h types and functions, instead use GCC
> supplied C++11 memory model builtins.
>
> Signed-off-by: Tyler Retzlaff
> ---
> drivers/net/null/rte_eth_null.c | 20 ++--
> 1 file changed, 10
On Fri, Mar 17, 2023 at 09:59:41AM -0700, Stephen Hemminger wrote:
> The cmdline_poll() function is broken and was
> not fully tested, don't use it.
>
> Instead, use sigaction() to cancel read character on Unix OS's
> and a new helper to cancel I/O on Windows.
>
> Fixes: 0fd1386c30c3 ("app/testpm
On Fri, Mar 17, 2023 at 07:51:25PM +0100, David Marchand wrote:
> On Fri, Mar 17, 2023 at 3:50 PM Tyler Retzlaff
> wrote:
> > > > -struct thread_routine_ctx {
> > > > +struct thread_start_context {
> > > > rte_thread_func thread_func;
> > > > - void *routine_args;
> > > > + voi
On Fri, Mar 17, 2023 at 07:52:29PM +0100, David Marchand wrote:
> From: Tyler Retzlaff
>
> In rte_thread_create setting affinity after pthread_create may fail.
> Such a failure should result in the entire rte_thread_create failing
> but doesn't.
>
> Additionally if there is a failure to set affi
On Fri, 17 Mar 2023 13:19:41 -0700
Tyler Retzlaff wrote:
> Replace the use of rte_atomic.h types and functions, instead use GCC
> supplied C++11 memory model builtins.
>
> This series covers the libraries and drivers that are built on Windows.
>
> The code has be converted to use the __atomic b
On Fri, Mar 17, 2023 at 02:42:26PM -0700, Stephen Hemminger wrote:
> On Fri, 17 Mar 2023 13:19:41 -0700
> Tyler Retzlaff wrote:
>
> > Replace the use of rte_atomic.h types and functions, instead use GCC
> > supplied C++11 memory model builtins.
> >
> > This series covers the libraries and driver
Update driver to use rte thread API where available instead of pthread
as a prerequisite to removing pthread stubs on Windows.
Signed-off-by: Tyler Retzlaff
---
drivers/net/ixgbe/ixgbe_ethdev.c | 10 +-
drivers/net/ixgbe/ixgbe_ethdev.h | 2 +-
2 files changed, 6 insertions(+), 6 deletio
Adopt rte thread APIs in code built for Windows to decouple it from the
pthread shim.
Remove most of the pthread_xxx lifetime shim functions, only
pthread_create remains while we wait for rte_ctrl_thread_create removal.
Tyler Retzlaff (6):
dma/skeleton: use rte thread API
net/ixgbe: use rte t
Update driver to use rte thread API where available instead of pthread
as a prerequisite to removing pthread stubs on Windows.
Signed-off-by: Tyler Retzlaff
---
drivers/net/ice/ice_dcf_parent.c | 11 +--
1 file changed, 5 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ice/ice_dc
Update driver to use rte thread API where available instead of pthread
as a prerequisite to removing pthread stubs on Windows.
Signed-off-by: Tyler Retzlaff
---
drivers/dma/skeleton/skeleton_dmadev.c | 15 ---
drivers/dma/skeleton/skeleton_dmadev.h | 4 ++--
2 files changed, 10 inse
Update driver to use rte thread API where available instead of pthread
as a prerequisite to removing pthread stubs on Windows.
Signed-off-by: Tyler Retzlaff
---
drivers/net/iavf/iavf_vchnl.c | 12 ++--
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/net/iavf/iavf_vc
Update driver to use rte thread API where available instead of pthread
as a prerequisite to removing pthread stubs on Windows.
There is a single pthread_create still in use until
rte_ctrl_thread_create is removed.
Signed-off-by: Tyler Retzlaff
---
lib/eal/common/eal_common_thread.c | 4 ++--
l
Remove most of the pthread_xxx lifetime shim functions, only
pthread_create remains while we wait for rte_ctrl_thread_create removal.
Signed-off-by: Tyler Retzlaff
---
lib/eal/windows/include/pthread.h | 99 ---
1 file changed, 99 deletions(-)
diff --git a/li
From: Long Li
Unnecessary assignments involve memset and waste CPU cycles.
Removing them to reduce CPU usage.
Fixes: 517ed6e2d590 ("net/mana: add basic driver with build environment")
Cc: sta...@dpdk.org
Signed-off-by: Long Li
---
drivers/net/mana/gdma.c | 11 ++-
drivers/net/mana/mana
From: Long Li
We can poll completion queues in a batch to speed up completion processing.
Also, the completion data doesn't need to be copied out of the hardware
queue and they can be passed as pointers to be consumed by the RX/TX code.
Fixes: 517ed6e2d590 ("net/mana: add basic driver with build
48 matches
Mail list logo
