Ferruh Yigit wrote:
> Sent: Saturday, October 10, 2020 2:53 AM
> To: oulijun ; wenzhuo...@intel.com;
> beilei.x...@intel.com; adrien.mazarg...@6wind.com; Phil Yang
>
> Cc: dev@dpdk.org; linux...@huawei.com
> Subject: Re: [PATCH v4] RSS key use with testpmd
>
> On 10
> -Original Message-
> From: Honnappa Nagarahalli
> Sent: Friday, October 2, 2020 8:07 AM
> To: dev@dpdk.org; Honnappa Nagarahalli
> ; Phil Yang ;
> tho...@monjalon.net; arybche...@solarflare.com; ferruh.yi...@intel.com
> Cc: abhinandan.guj...@intel.com; nd ;
> br
Honnappa Nagarahalli wrote:
> Subject: [PATCH] maintainers: update for MCS lock
>
> Updating MAINTAINERS file for MCS lock.
>
> Signed-off-by: Honnappa Nagarahalli
Thank you, Honnappa.
Acked-by: Phil Yang
> ---
> Phil Yang is leaving Arm. I worked with Phil on th
tpmd dummy RSS key as the default
> key.
>
> Fixes: ac8d22de2394 ("ethdev: flatten RSS configuration in flow API")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Lijun Ou
It looks good to me. Thanks.
Reviewed-by: Phil Yang
> ---
> V3->V4:
> -fix checkpatch warning
Hi Raslan,
It seems that there are no more comments for this patch.
So shall we proceed further?
Thanks,
Phil Yang
> -Original Message-
> From: Alexander Kozyrev
> Sent: Thursday, September 10, 2020 9:37 AM
> To: Honnappa Nagarahalli ; Phil Yang
> ; akozy...@mel
^
>
> Signed-off-by: Ferruh Yigit
Verified with gcc version 11.0.0 20200920 (experimental).
Reviewed-by: Phil Yang
> ---
> app/test-pmd/config.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/app/test-pmd/co
dev On Behalf Of Phil Yang : writes:
> > Subject: [dpdk-dev] [PATCH] [PATCH V2] testpmd: add eeprom/module
> > eeprom display
> >
> > Add module EEPROM/EEPROM dump command
> >"show port (module_eeprom|eeprom)"
> > Commands will dump the c
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic
builtins for link status update.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
Acked-by: Andrew Rybchenko
---
lib/librte_ethdev/rte_ethdev_driver.h | 19 +--
1 file
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic builtins
for device processing counter.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_bbdev/rte_bbdev.c | 5 +++--
lib/librte_bbdev/rte_bbdev.h | 4 +---
2 files changed, 4
:
remove unnecessary rte_atomic.h headers. (David)
v2:
Fix Clang int-conversion warning.
v1:
Initial version.
Phil Yang (4):
eal: use C11 atomic builtins for already initialized check
bbdev: use C11 atomic builtins for device processing counter
power: use C11 atomic builtins for power in use
Since rte_atomicXX APIs are not allowed to be used, use C11 builtins to
check if EAL is already initialized.
Signed-off-by: Phil Yang
Reviewed-by: Honnappa Nagarahalli
Reviewed-by: Ruifeng Wang
---
lib/librte_eal/freebsd/eal.c | 19 ++-
lib/librte_eal/linux/eal.c | 21
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic
builtins for power in use state update.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_power/power_acpi_cpufreq.c | 45 +
lib/librte_power
David Marchand writes:
> Subject: Re: [PATCH v2 0/4] use C11 atomic builtins for libs
>
> On Wed, Sep 16, 2020 at 10:24 AM Phil Yang wrote:
> >
> > Since rte_atomicXX APIs are not allowed to be used[1][2], use C11 atomic
> > builtins instead in eal, bbdev, power,
David Marchand writes:
> Subject: Re: [PATCH v2 1/4] eal: use C11 atomic builtins for already
> initialized
> check
>
> On Wed, Sep 16, 2020 at 10:24 AM Phil Yang wrote:
> >
> > Since rte_atomicXX APIs are not allowed to be used, use C11 builtins to
> > ch
The 'ninja install' needs absolute paths. Use the absolute path for the
default build_dir to avoid repeatedly calling of readlink.
Suggested-by: Juraj Linkeš
Signed-off-by: Phil Yang
---
devtools/test-meson-builds.sh | 8
1 file changed, 4 insertions(+), 4 deletions(-)
di
As the DESTDIR variable use as ninja's destination directory when invoking
ninja install. If the DESTDIR is not an absolute path, ninja complains.
Fixes: 777014e56d07 ("devtools: add ABI checks")
Cc: sta...@dpdk.org
Signed-off-by: Phil Yang
---
v2:
Use readlink to ensure consiste
David Marchand writes:
> Subject: Re: [dpdk-dev] [PATCH v2 0/2] remove RTE CIO barriers
>
> On Thu, Sep 17, 2020 at 4:23 AM Phil Yang wrote:
> > > If there will a v3, can it be possible to have it after next-net pulled
> > > to main repo? Which I am hopping one for
Since the 20.08 release deprecated rte_cio_*mb APIs because these APIs
provide the same functionality as rte_io_*mb APIs on all platforms, so
remove them and use rte_io_*mb instead.
Signed-off-by: Phil Yang
Signed-off-by: Joyce Kong
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
barriers behave the same as
the other arches. (David)
v2:
Update 20.11 release note by adding these API change announcement. (Honnappa)
v1:
Initial version.
Phil Yang (1):
eal: remove deprecated coherent IO memory barriers
doc/guides/rel_notes/deprecation.rst | 6 -
doc/guides
dev On Behalf Of Lijun Ou writes:
> >> Subject: [dpdk-dev] [PATCH v3] app/testpmd: fix the default RSS key
> >> configuration
> >
> > Hi Lijun,
> >
> > Please fix the coding style issues.
> >
> > "Must be a reply to the first patch (--in-reply-to)."
> >
> >
> >>
> >> When a user runs a flow crea
> -Original Message-
> From: dev On Behalf Of Lijun Ou
> Sent: Thursday, September 10, 2020 9:51 AM
> To: wenzhuo...@intel.com; beilei.x...@intel.com;
> adrien.mazarg...@6wind.com; ferruh.yi...@intel.com
> Cc: dev@dpdk.org; linux...@huawei.com
> Subject: [dpdk-dev] [PATCH v3] app/testpmd:
> -Original Message-
> From: dev On Behalf Of David Liu
> Sent: Wednesday, September 16, 2020 1:28 AM
> To: ferruh.yi...@intel.com
> Cc: dev@dpdk.org; lylav...@iol.unh.edu; tho...@monjalon.net; David Liu
>
> Subject: [dpdk-dev] [PATCH] [PATCH V2] testpmd: add eeprom/module
> eeprom displa
Jakub Grajciar -X (jgrajcia - PANTHEON TECH SRO at Cisco) writes:
> > Subject: RE: [PATCH] net/memif: relax barrier for zero copy path
> >
> > Hi Jakub,
> > I am trying to review this patch. I am having difficulty in
> understanding
> > the implementation for the queue/ring, appreciate if yo
; should be an
absolute path, but the default 'abiref' value is a relative path.
So should we keep them align?
>
> > Fixes: 777014e56d07 ("devtools: add ABI checks")
> >
> > Signed-off-by: Phil Yang
> > ---
> > devtools/test-meson-builds.sh | 2 +-
Ferruh Yigit wrote:
> >>
> >> Since rte_cio_*mb APIs have been deprecated[1], remove them and use
> >> rte_io_*mb APIs instead.
> >>
> >> [1]
> http://code.dpdk.org/dpdk/latest/source/doc/guides/rel_notes/deprecatio
> n.rst#L100
> >>
> >> v2:
> >> Update 20.11 release note by adding these API cha
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic
builtins for link status update.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_ethdev/rte_ethdev_driver.h | 19 +--
1 file changed, 5 insertions(+), 14
Since rte_atomicXX APIs are not allowed to be used, use C11 builtins to
check if EAL is already initialized.
Signed-off-by: Phil Yang
Reviewed-by: Honnappa Nagarahalli
Reviewed-by: Ruifeng Wang
---
lib/librte_eal/freebsd/eal.c | 18 ++
lib/librte_eal/linux/eal.c | 20
Clang int-conversion warning.
v1:
Initial version.
Phil Yang (4):
eal: use C11 atomic builtins for already initialized check
bbdev: use C11 atomic builtins for device processing counter
power: use C11 atomic builtins for power in use state update
ethdev: use C11 atomic builtins for link
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic builtins
for device processing counter.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_bbdev/rte_bbdev.c | 5 +++--
lib/librte_bbdev/rte_bbdev.h | 4 +---
2 files changed, 4
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic
builtins for power in use state update.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_power/power_acpi_cpufreq.c | 45 +
lib/librte_power
David Marchand writes:
> Subject: Re: [dpdk-dev] [PATCH 0/4] use C11 atomic builtins for libs
>
> On Fri, Sep 11, 2020 at 5:29 AM Phil Yang wrote:
> >
> > Since rte_atomicXX APIs are not allowed to be used[1][2], use C11 atomic
> > builtins instead in eal, bb
Thomas Monjalon writes:
> >
> > Remove the deprecated refcnt_atomic union fields in
> > rte_mbuf and rte_mbuf_ext_shared_info structures.
> >
> > Signed-off-by: Phil Yang
> > Reviewed-by: Ruifeng wang
> > Signed-off-by: Thomas Monjalon
> > ---
Andrew Rybchenko writes:
> On 9/11/20 10:35 AM, Phil Yang wrote:
> > Remove the deprecated[1] refcnt_atomic member in structures rte_mbuf
> and
> > rte_mbuf_ext_shared_info.
> >
> > [1]
> http://code.dpdk.org/dpdk/latest/source/doc/guides/rel_notes/deprecat
Remove the deprecated[1] refcnt_atomic member in structures rte_mbuf and
rte_mbuf_ext_shared_info.
[1]
http://code.dpdk.org/dpdk/latest/source/doc/guides/rel_notes/deprecation.rst#L161
Phil Yang (2):
mbuf: remove generic mbuf atomically accessed refcnt
mbuf: remove external buffer
Remove the deprecated refcnt_atomic union fields in rte_mbuf
structure.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
---
doc/guides/rel_notes/release_20_11.rst | 2 ++
lib/librte_mbuf/rte_mbuf_core.h| 7 +--
2 files changed, 3 insertions(+), 6 deletions(-)
diff --git a/doc
Remove the deprecated refcnt_atomic union fields in
rte_mbuf_ext_shared_info structure.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng wang
---
doc/guides/rel_notes/release_20_11.rst | 4 ++--
lib/librte_mbuf/rte_mbuf_core.h| 6 +-
2 files changed, 3 insertions(+), 7 deletions
ormance on throughput.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
---
drivers/net/memif/rte_eth_memif.c | 35 +--
1 file changed, 25 insertions(+), 10 deletions(-)
diff --git a/drivers/net/memif/rte_eth_memif.c
b/drivers/net/memif/rte_eth_memif.c
ind
Since rte_cio_*mb APIs have been deprecated[1], remove them and use
rte_io_*mb APIs instead.
[1]
http://code.dpdk.org/dpdk/latest/source/doc/guides/rel_notes/deprecation.rst#L100
v2:
Update 20.11 release note by adding these API change announcement. (Honnappa)
v1:
Initial version.
Phil Yang
Replace the deprecated rte_cio_[rw]mb barriers with rte_io_[rw]mb
barriers.
Signed-off-by: Phil Yang
Signed-off-by: Joyce Kong
Reviewed-by: Ruifeng Wang
---
drivers/common/mlx5/mlx5_common.h | 2 +-
drivers/crypto/octeontx2/otx2_cryptodev_ops.c | 2 +-
drivers/crypto/octeontx2
Remove the deprecated rte_cio_[rw]mb APIs.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
---
doc/guides/rel_notes/release_20_11.rst | 3 +++
lib/librte_eal/arm/include/rte_atomic_32.h | 4 ---
lib/librte_eal/arm/include/rte_atomic_64.h | 4 ---
lib/librte_eal/include/generic
Since rte_atomicXX APIs are not allowed to be used[1][2], use C11 atomic
builtins instead in eal, bbdev, power, and ethdev libs.
[1]
http://code.dpdk.org/dpdk/latest/source/doc/guides/rel_notes/deprecation.rst#L87
[2] http://code.dpdk.org/dpdk/latest/source/devtools/checkpatches.sh#L80
Phil
Since rte_atomicXX APIs are not allowed to be used, use C11 builtins to
check if EAL is already initialized.
Signed-off-by: Phil Yang
Reviewed-by: Honnappa Nagarahalli
Reviewed-by: Ruifeng Wang
---
lib/librte_eal/freebsd/eal.c | 18 ++
lib/librte_eal/linux/eal.c | 20
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic builtins
for device processing counter.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_bbdev/rte_bbdev.c | 5 +++--
lib/librte_bbdev/rte_bbdev.h | 4 +---
2 files changed, 4
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic
builtins for power in use state update.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_power/power_acpi_cpufreq.c | 45 +
lib/librte_power
Since rte_atomicXX APIs are not allowed to be used, use C11 atomic
builtins for link status update.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Nagarahalli
---
lib/librte_ethdev/rte_ethdev_driver.h | 19 +--
1 file changed, 5 insertions(+), 14
> >
> > > >
> > > > > > > > > > @@ -1790,9 +1792,9 @@ mlx5_rx_burst_mprq(void
> > *dpdk_rxq,
> > > > > > struct
> > > > > > > > > > rte_mbuf **pkts, uint16_t pkts_n) void *buf_addr;
> > > > > > > > > >
> > > > > > > > > > /* Increment the refcnt of the whole chunk. */
> > > > > > > > > > -rte_at
Use c11 atomics with RELAXED ordering instead of the rte_atomic ops
which enforce unnecessary barriers on aarch64.
Signed-off-by: Phil Yang
---
v4:
Remove the unnecessary ACQUIRE barrier in rx burst path. (Honnappa)
v3:
Split from the patchset:
http://patchwork.dpdk.org/cover/68159/
drivers
operation
is atomic.
Thanks,
Phil Yang
> -Original Message-
> From: Diogo Behrens
> Sent: Thursday, August 27, 2020 4:57 PM
> To: Phil Yang
> Cc: dev@dpdk.org; nd ; Honnappa Nagarahalli
>
> Subject: RE: [PATCH] librte_eal: fix mcslock hang on weak memory
>
Diogo Behrens writes:
> Subject: [PATCH] librte_eal: fix mcslock hang on weak memory
>
> The initialization me->locked=1 in lock() must happen before
> next->locked=0 in unlock(), otherwise a thread may hang forever,
> waiting me->locked become 0. On weak memory systems (such as ARMv
Aaron Conole writes:
> Subject: Re: [PATCH v8 5/5] doc: add aarch32 build guidance
>
> Juraj Linkeš writes:
>
> > From: Phil Yang
> >
> > Add cross-compiling guidance for 32-bit aarch32 DPDK on aarch64 host.
> >
> > Signed-off
Hi Honnappa,
> Hi Phil,
> Can you please add a statement in release notes indicating that these
> APIs are removed.
Sure. I will add it to the 2/2 patch in the new version.
Thanks,
Phil
>
> Thank you,
> Honnappa
>
> > -Original Message-
> >
Remove the deprecated rte_cio_[rw]mb APIs.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
---
lib/librte_eal/arm/include/rte_atomic_32.h | 4 ---
lib/librte_eal/arm/include/rte_atomic_64.h | 4 ---
lib/librte_eal/include/generic/rte_atomic.h | 39 -
lib
Replace the deprecated rte_cio_[rw]mb barriers with rte_io_[rw]mb
barriers.
Signed-off-by: Phil Yang
Signed-off-by: Joyce Kong
---
drivers/common/mlx5/mlx5_common.h | 2 +-
drivers/crypto/octeontx2/otx2_cryptodev_ops.c | 2 +-
drivers/crypto/octeontx2/otx2_cryptodev_sec.c | 4
> -Original Message-
> From: oulijun
> Sent: Tuesday, August 4, 2020 9:29 AM
> To: Phil Yang ; wenzhuo...@intel.com;
> beilei.x...@intel.com; bernard.iremon...@intel.com;
> adrien.mazarg...@6wind.com; ferruh.yi...@intel.com
> Cc: dev@dpdk.org; linux...@huawei.com; nd
> -Original Message-
> From: Viacheslav Ovsiienko
> Sent: Wednesday, July 29, 2020 8:30 PM
> To: dev@dpdk.org
> Cc: ma...@mellanox.com; rasl...@mellanox.com; tho...@monjalon.net;
> ferruh.yi...@intel.com; Phil Yang
> Subject: [PATCH v2] app/testpmd: fix
Hi Slava,
Thanks for your analysis. Some comments inline.
> -Original Message-
> From: Slava Ovsiienko
> Sent: Wednesday, July 29, 2020 4:08 PM
> To: Phil Yang
> Cc: Matan Azrad ; Raslan Darawsheh
> ; tho...@monjalon.net; ferruh.yi...@intel.com;
> Honnappa Nag
> -Original Message-
> From: dev On Behalf Of Viacheslav Ovsiienko
> Sent: Monday, July 27, 2020 11:27 PM
> To: dev@dpdk.org
> Cc: ma...@mellanox.com; rasl...@mellanox.com; tho...@monjalon.net;
> ferruh.yi...@intel.com
> Subject: [dpdk-dev] [PATCH] app/testpmd: fix txonly mode timestamp
>
Reduce the number of loops to 10K to avoid the meson test timeout
warning.
For example:
$ sudo meson test -C build --suite DPDK:fast-tests / atomic_autotest -t 50
[...]
1/1 DPDK:fast-tests / atomic_autotest OK 187.35s
Signed-off-by: Phil Yang
---
v2:
Split from the original patchset
Move the MCS lock performance test into perf tests.
Signed-off-by: Phil Yang
Reviewed-by: Gavin Hu
---
v2:
1. Rebase code.
2. Split from the original patchset.
MAINTAINERS | 1 +
app/test/Makefile| 1 +
app/test/autotest_data.py| 6 +++
app/test
Alexander Kozyrev writes:
> > > > > > @@ -1790,9 +1792,9 @@ mlx5_rx_burst_mprq(void *dpdk_rxq,
> > struct
> > > > > > rte_mbuf **pkts, uint16_t pkts_n) void *buf_addr;
> > > > > >
> > > > > > /* Increment the refcnt of the whole chunk. */
> > > > > > -rte_atomic16_add_return(&buf->refcnt, 1);
>
> > > Subject: Re: [dpdk-dev] [PATCH v3] net/mlx5: relaxed ordering for
> > > multi-packet RQ buffer refcnt
> > >
> > > Hi,
> > >
> > > We are also doing C11 atomics converting for other components.
> > > Your insi
platforms. Move the performance test
> into perf tests.
>
> Signed-off-by: Phil Yang
> Reviewed-by: Gavin Hu
> ---
> MAINTAINERS | 1 +
> app/test/Makefile| 1 +
> app/test/autotest_data.py| 6 +++
> app/test/meson.build
Alexander Kozyrev writes:
> >
> > > > > Subject: RE: [dpdk-dev] [PATCH v3] net/mlx5: relaxed ordering for
> > > > > multi- packet RQ buffer refcnt
> > > > >
> > > > > Hi Phil Yang, we noticed that this patch gives us 10% of
&
> Reported-by: David Marchand
> Signed-off-by: Harry van Haaren
Minor nit, otherwise it looks good to me.
Reviewed-by: Phil Yang
>
> ---
>
> Thanks for discussion on v1, this v2 fixup for the CI
> including previous feedback on ML.
> ---
> app/test/test_service_cor
d from the service_runner_func() function.
>
> Suggested-by: Lukasz Wojciechowski
> Signed-off-by: Harry van Haaren
>
It looks good to me.
Reviewed-by: Phil Yang
Thanks,
Phil
<...>
Alexander Kozyrev writes:
<...>
> Subject: RE: [dpdk-dev] [PATCH v3] net/mlx5: relaxed ordering for multi-
> packet RQ buffer refcnt
>
> Hi Phil Yang, we noticed that this patch gives us 10% of performance
> degradation on ARM.
> x86 seems to be unaffected though. Do
> -Original Message-
> From: David Marchand
> Sent: Friday, July 17, 2020 9:59 PM
> To: Phil Yang
> Cc: tho...@monjalon.net; dev ; Mcnamara, John
> ; Honnappa Nagarahalli
> ; David Christensen
> ; jer...@marvell.com; Ananyev, Konstantin
> ; Ola Liljedahl ;
&
> -Original Message-
> From: Aaron Conole
> Sent: Saturday, July 18, 2020 6:39 AM
> To: Lukasz Wojciechowski
> Cc: David Marchand ; Van Haaren Harry
> ; Igor Romanov
> ; Honnappa Nagarahalli
> ; Phil Yang ; dev
> ; Ferruh Yigit
> Subject: Re
: Phil Yang
Reviewed-by: Ruifeng Wang
---
devtools/checkpatches.sh | 35 +++
1 file changed, 35 insertions(+)
diff --git a/devtools/checkpatches.sh b/devtools/checkpatches.sh
index 58021aa..990a4d0 100755
--- a/devtools/checkpatches.sh
+++ b/devtools/checkpatches.sh
Provide a wrapper for __atomic_thread_fence builtins to support
optimized code for __ATOMIC_SEQ_CST memory order for x86 platforms.
Suggested-by: Honnappa Nagarahalli
Signed-off-by: Phil Yang
Reviewed-by: Ola Liljedahl
Acked-by: Konstantin Ananyev
---
lib/librte_eal/arm/include/rte_atomic_32
Add information about possible optimizations using C11 atomic builtins.
Signed-off-by: Phil Yang
Signed-off-by: Honnappa Nagarahalli
Reviewed-by: Honnappa Nagarahalli
---
doc/guides/prog_guide/writing_efficient_code.rst | 59 +++-
1 file changed, 58 insertions(+), 1
changes of libraries which make use of c11 atomic APIs out of this
patchset.
v3:
add libatomic dependency for 32-bit clang
v2:
1. fix Clang '-Wincompatible-pointer-types' WARNING.
2. fix typos.
Phil Yang (3):
doc: add optimizations using C11 atomic builtins
eal/atomic: add wrapper for
Thomas Monjalon wrties:
> > In order to deprecate the rte_atomic and rte_smp barrier APIs, prevent
> > the patches from using these APIs and __sync builtins in new code.
> >
> > On x86 the __atomic_thread_fence(__ATOMIC_SEQ_CST) is quite
> expensive
> > for SMP case. Flag the new code which use
Thomas Monjalon writes:
> > Provide a wrapper for __atomic_thread_fence builtins to support
> > optimized code for __ATOMIC_SEQ_CST memory order for x86 platforms.
>
> Logically, it should be the patch 2.
> So you can recommend using this wrapper in the explanation
> of the actual patch 2.
Inde
> >> Subject: [dpdk-dev] [PATCH] app/testpmd: fix the default RSS key
> >> configuration
> >>
> >> When an user runs a flow create cmd to configure an RSS rule
> >> without specifying the empty rss actions in testpmd, this mean
> >> that the flow gets the default RSS functions. However, the
> >>
Add information about possible optimizations using C11 atomic builtins.
Signed-off-by: Phil Yang
Signed-off-by: Honnappa Nagarahalli
Reviewed-by: Honnappa Nagarahalli
---
doc/guides/prog_guide/writing_efficient_code.rst | 59 +++-
1 file changed, 58 insertions(+), 1
In order to deprecate the rte_atomic and rte_smp barrier APIs, prevent
the patches from using these APIs and __sync builtins in new code.
On x86 the __atomic_thread_fence(__ATOMIC_SEQ_CST) is quite expensive
for SMP case. Flag the new code which use __atomic_thread_fence API.
Signed-off-by: Phil
Provide a wrapper for __atomic_thread_fence builtins to support
optimized code for __ATOMIC_SEQ_CST memory order for x86 platforms.
Suggested-by: Honnappa Nagarahalli
Signed-off-by: Phil Yang
Reviewed-by: Ola Liljedahl
Acked-by: Konstantin Ananyev
---
lib/librte_eal/arm/include/rte_atomic_32
libatomic dependency for 32-bit clang
v2:
1. fix Clang '-Wincompatible-pointer-types' WARNING.
2. fix typos.
Phil Yang (3):
doc: add optimizations using C11 atomic builtins
devtools: prevent use of rte atomic APIs in future patches
eal/atomic: add wrapper for C11 atomic thread fence
Honnappa Nagarahalli writes:
>
> > >
> > > > Subject: Re: [dpdk-dev] [PATCH v8 2/3] devtools: prevent use of rte
> > > > atomic APIs in future patches
> > > >
> > > > 16/07/2020 12:48, David Marchand
Honnappa Nagarahalli writes:
>
> > Subject: [PATCH v8 1/3] doc: add optimizations using C11 atomic built-ins
> >
> > Add information about possible optimizations using C11 atomic built-ins.
> >
> > Signed-off-by: Phil Yang
> > Signed-off-by: Honnappa N
> On Thu, Jul 16, 2020 at 3:21 PM Dodji Seketeli wrote:
> > Just for the sake of precision, I'd like to say that in the coming 1.8
> > version of libabigail, this change won't be reported by the tooling as a
> > problem anymore. This is thanks to David filing the feature request
> > https://sou
refcnt_atomic member in structures rte_mbuf and rte_mbuf_ext_shared_info
will be deprecated in 20.11 release.
Suggested-by: Honnappa Nagarahalli
Signed-off-by: Phil Yang
Acked-by: Konstantin Ananyev
---
doc/guides/rel_notes/deprecation.rst | 6 ++
1 file changed, 6 insertions(+)
diff
Use C11 atomic builtins with explicit ordering instead of rte_atomic
ops which enforce unnecessary barriers on aarch64.
Suggested-by: Olivier Matz
Suggested-by: Dodji Seketeli
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
Acked-by: Olivier Matz
---
v5:
1. Change built-ins to builtins.
2
Hello,
Thomas Monjalon writes:
> Subject: Re: [dpdk-dev] [PATCH v8 2/3] devtools: prevent use of rte atomic
> APIs in future patches
>
> 16/07/2020 12:48, David Marchand:
> > On Thu, Jul 16, 2020 at 6:58 AM Phil Yang wrote:
> > > check_forbidden_additions
> -Original Message-
> From: David Marchand
> Sent: Thursday, July 16, 2020 6:49 PM
> To: Phil Yang
> Cc: tho...@monjalon.net; Mcnamara, John ;
> Honnappa Nagarahalli ; David Christensen
> ; dev ; jer...@marvell.com;
> Ananyev, Konstantin ; Ola Liljedahl
> ; B
Provide a wrapper for __atomic_thread_fence built-in to support
optimized code for __ATOMIC_SEQ_CST memory order for x86 platforms.
Suggested-by: Honnappa Nagarahalli
Signed-off-by: Phil Yang
Reviewed-by: Ola Liljedahl
Acked-by: Konstantin Ananyev
---
lib/librte_eal/arm/include/rte_atomic_32
libraries which make use of c11 atomic APIs out of this
patchset.
v3:
add libatomic dependency for 32-bit clang
v2:
1. fix Clang '-Wincompatible-pointer-types' WARNING.
2. fix typos.
Phil Yang (3):
doc: add optimizations using C11 atomic built-ins
devtools: prevent use of rte atom
__atomic_thread_fence(__ATOMIC_SEQ_CST) is quite expensive
for SMP case. Flag the new code which use __atomic_thread_fence API.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
---
devtools/checkpatches.sh | 40
1 file changed, 40 insertions(+)
diff --git a/devtools
Add information about possible optimizations using C11 atomic built-ins.
Signed-off-by: Phil Yang
Signed-off-by: Honnappa Nagarahalli
---
doc/guides/prog_guide/writing_efficient_code.rst | 59 +++-
1 file changed, 58 insertions(+), 1 deletion(-)
diff --git a/doc/guides
David Marchand writes:
> Subject: Re: [PATCH v4 1/2] mbuf: use C11 atomic built-ins for refcnt
> operations
>
> On Thu, Jul 9, 2020 at 5:59 PM Phil Yang wrote:
> >
> > Use C11 atomic built-ins with explicit ordering instead of rte_atomic
> > ops which enforce u
> -Original Message-
> From: David Marchand
> Sent: Wednesday, July 15, 2020 8:29 PM
> To: Phil Yang
> Cc: Olivier Matz ; dev ; Stephen
> Hemminger ; David Christensen
> ; Honnappa Nagarahalli
> ; Ruifeng Wang
> ; nd ; Dodji Seketeli
> ; Aaron Conole
>
> > -Original Message-
> > From: dev On Behalf Of Lijun Ou
> > Sent: Wednesday, July 1, 2020 3:43 PM
> > To: wenzhuo...@intel.com; beilei.x...@intel.com;
> > bernard.iremon...@intel.com; ferruh.yi...@intel.com
> > Cc: dev@dpdk.org; linux...@huawei.com
> > Subject: [dpdk-dev] [PATCH v2] a
> -Original Message-
> From: dev On Behalf Of Lijun Ou
> Sent: Thursday, July 9, 2020 9:48 PM
> To: wenzhuo...@intel.com; beilei.x...@intel.com;
> bernard.iremon...@intel.com; adrien.mazarg...@6wind.com;
> ferruh.yi...@intel.com
> Cc: dev@dpdk.org; linux...@huawei.com
> Subject: [dpdk-dev]
ls in 5tswap
>
> Enable empty polls in burst stats within 5tswap.c
>
> Signed-off-by: Dharmik Thakkar
> Reviewed-by: Honnappa Nagarahalli
Reviewed-by: Phil Yang
; I think same here, something like following would help:
> "Add information about possible optimizations using C11 atomic built-ins"
>
> >
> > Signed-off-by: Phil Yang
> > Signed-off-by: Honnappa Nagarahalli
> > ---
> > doc/guides/prog_guide/
tput format in
> flow query
>
> This patch fix the error line break in the output format of flow query
>
> Fixes: bdb1d61690f7 ("app/testpmd: support RSS config in flow query")
>
> Signed-off-by: Chenxu Di
Tested-by: Phil Yang
> ---
> app/test-pmd/
20.08 release and remove it in 20.11
> release.
Yes. Just like Oliver suggested, we will deprecate it in 20.08 release and
remove it in 20.11.
http://patchwork.dpdk.org/patch/73632/
Let's see if we have any other comments.
+Jerin and Konstantin.
Thanks,
Phil
>
> >
> > Signed
Honnappa Nagarahalli writes:
>
> >
> > 07/07/2020 11:50, Phil Yang:
> > > Add the maintainership of C11 atomics code. Different maintainers
> > > focus on different platforms C11 atomics.
> > > --- a/MAINTAINERS
> > > +++ b/MAINTAINERS
&g
__atomic_thread_fence(__ATOMIC_SEQ_CST) is quite expensive
for SMP case. Flag the new code which use __atomic_thread_fence API.
Signed-off-by: Phil Yang
Reviewed-by: Ruifeng Wang
---
devtools/checkpatches.sh | 40
1 file changed, 40 insertions(+)
diff --git a/devtools
1 - 100 of 465 matches
Mail list logo
