Re: [dpdk-dev] [PATCH] bus/vdev: add warning for duplicated vdev name

On Wed, Dec 19, 2018 at 03:26:36AM +0100, Thomas Monjalon wrote: > 29/11/2018 08:16, Yahui Cao: > > If duplicated vdev name is detected, print out a warning message. > > > > Signed-off-by: Yahui Cao > > --- > > drivers/bus/vdev/vdev.c | 2 ++ > > 1 file changed, 2 insertions(+) > > > > diff --g

Re: [dpdk-dev] [PATCH v3 3/3] net/virtio: improve batching in mergeable path

> -Original Message- > From: dev On Behalf Of Maxime Coquelin > Sent: Friday, December 21, 2018 1:27 AM > To: dev@dpdk.org; jfreim...@redhat.com; tiwei@intel.com; > zhihong.w...@intel.com > Cc: Maxime Coquelin > Subject: [dpdk-dev] [PATCH v3 3/3] net/virtio: improve batching in > m

[dpdk-dev] [PATCH v3] devtools: fix wrong headline lowercase for arm

Change to the new wording according to Arm corp: https://mails.dpdk.org/archives/dev/2018-October/115174.html otherwise, "Wrong headline lowercase" may be falsely reported. Fixes: f146ada21d1e ("scripts: add more git log checks") Cc: sta...@dpdk.org Signed-off-by: Gavin Hu --- devtools/check-gi

[dpdk-dev] [PATCH] kni: fix build on RHEL 7.6-ALT on IBM POWER9

Signed-off-by: David Zeng --- kernel/linux/kni/compat.h | 8 +++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/kernel/linux/kni/compat.h b/kernel/linux/kni/compat.h index 5aadebb..1afa3b8 100644 --- a/kernel/linux/kni/compat.h +++ b/kernel/linux/kni/compat.h @@ -102,8 +102,14

[dpdk-dev] [PATCH] app/testpmd: fixed get RSS conf

I40e do not allow to get rss hena only, need to get rss key meanwhile. Fixes: 16321de09396 ("ethdev: allow to get RSS hash functions and key") Cc: sta...@dpdk.org Signed-off-by: Qiming Yang --- app/test-pmd/config.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/app/te

[dpdk-dev] DSW eventdev and multi-process DPDK

Hi, We are considering using a multi-process mode of the DPDK with the event generators and consumers being spread across multiple processes (on different cores). We are also considering using the DSW eventdev. Is the DSW designed for such a use case? If so, are there some restrictions and somet

Re: [dpdk-dev] [PATCH v4] vhost: batch used descs chains write-back with packed ring

On Thu, Dec 20, 2018 at 05:47:55PM +0100, Maxime Coquelin wrote: > Instead of writing back descriptors chains in order, let's > write the first chain flags last in order to improve batching. > > Also, move the write barrier in logging cache sync, so that it > is done only when logging is enabled.

Re: [dpdk-dev] [PATCH v2] devtools: fix error propagation from check-forbidden-tokens.awk

18/12/2018 16:19, Arnon Warshavsky: > Bugzilla ID: 165 > Fixes: 4d4c612e6a30 ("devtools: check wrong svg include in guides") > Signed-off-by: Arnon Warshavsky > > Explicitly collect the error code of the > multiple awk script calls Applied, thanks Note: the explanation must be moved at the begi

Re: [dpdk-dev] [PATCH] hash: fix out-of-bound write while freeing key slot

20/12/2018 14:59, Bruce Richardson: > On Wed, Nov 21, 2018 at 08:51:56PM -0600, Honnappa Nagarahalli wrote: > > Add a debug check for out-of-bound write while freeing the key slot. > > > > Coverity issue: 325733 > > Fixes: e605a1d36ca7 ("hash: add lock-free r/w concurrency") > > Cc: honnappa.nagar

Re: [dpdk-dev] [PATCH v2 2/2] test/compress: add varied buffer input/outputs

> -Original Message- > From: Kovacevic, Marko > Sent: Thursday, December 20, 2018 7:58 AM > To: dev@dpdk.org > Cc: akhil.go...@nxp.com; Daly, Lee ; Jozwiak, TomaszX > ; O'Hare, Cathal ; Trahe, > Fiona > ; Kovacevic, Marko > Subject: [PATCH v2 2/2] test/compress: add varied buffer input

Re: [dpdk-dev] [PATCH] hash: fix __rte_hash_lookup_bulk return value

10/12/2018 11:29, Bruce Richardson: > On Fri, Dec 07, 2018 at 04:01:26PM -0800, Jeff Shaw wrote: > > The __rte_hash_lookup_bulk() function returns void, and therefore > > should not return with an expression. This commit fixes the following > > compiler warning when attempting to compile with "-ped

Re: [dpdk-dev] [PATCH v2 1/2] test/compress: add out of space test

> -Original Message- > From: Kovacevic, Marko > Sent: Thursday, December 20, 2018 7:58 AM > To: dev@dpdk.org > Cc: akhil.go...@nxp.com; Daly, Lee ; Jozwiak, TomaszX > ; O'Hare, Cathal ; Trahe, > Fiona > ; Kovacevic, Marko > Subject: [PATCH v2 1/2] test/compress: add out of space test >

Re: [dpdk-dev] [PATCH] app/test-pmd: expand test-pmd queue threshold size of RED parameters

Thanks Cristian. > -Original Message- > From: Dumitrescu, Cristian > Sent: Friday, December 21, 2018 3:08 > To: Xu, Rosen ; Singh, Jasvinder > ; dev@dpdk.org > Cc: Lu, Wenzhuo ; Yigit, Ferruh > ; Wu, Jingjing > Subject: RE: [PATCH] app/test-pmd: expand test-pmd queue threshold size of > R

Re: [dpdk-dev] [PATCH v4] libs/power: add p-state driver compatibility

20/12/2018 15:43, Liang Ma: > Previously, in order to use the power library, it was necessary > for the user to disable the intel_pstate driver by adding > “intel_pstate=disable” to the kernel command line for the system, > which causes the acpi_cpufreq driver to be loaded in its place. > > This p

Re: [dpdk-dev] [PATCH v3] libs/power: add p-state driver compatibility

21/12/2018 01:30, Thomas Monjalon: > 20/12/2018 15:52, Hunt, David: > > On 19/12/2018 8:31 PM, Thomas Monjalon wrote: > > > 19/12/2018 10:09, Hunt, David: > > >> On 19/12/2018 3:18 AM, Thomas Monjalon wrote: > > >>> 14/12/2018 14:11, Liang Ma: > > Previously, in order to use the power library,

Re: [dpdk-dev] [PATCH v3] libs/power: add p-state driver compatibility

20/12/2018 15:52, Hunt, David: > On 19/12/2018 8:31 PM, Thomas Monjalon wrote: > > 19/12/2018 10:09, Hunt, David: > >> On 19/12/2018 3:18 AM, Thomas Monjalon wrote: > >>> 14/12/2018 14:11, Liang Ma: > Previously, in order to use the power library, it was necessary > for the user to disabl

Re: [dpdk-dev] [PATCH v2] eal: close MP socket during cleanup

20/12/2018 13:51, Qi Zhang: > When secondary process quit, the mp_socket* file still exist, that > cause rte_mp_request_sync fail when try to send message on a floating > socket. > > The patch fix the issue by introduce a function rte_mp_channel_cleanup. > This function will be called by rte_eal_c

[dpdk-dev] [PATCH] crypto/aesni_mb: enable out of place processing

Add out-of-place processing, i.e. different source and destination m_bufs, plus related capability update, tests and documentation. Document limitation to cipher-only as lib doesn't support digest generation in encrypt & generate direction for out-of-place yet. Signed-off-by: Fiona Trahe Signed-o

[dpdk-dev] [PATCH] test/crypto: don't attempt unsupported SGL tests on aesni mb PMD

Remove AESNI_MB flag from SGL test cases which it doesn't support. Signed-off-by: Fiona Trahe --- test/test/test_cryptodev_aes_test_vectors.h | 8 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/test/test/test_cryptodev_aes_test_vectors.h b/test/test/test_cryptodev_aes_te

[dpdk-dev] [PATCH] test/crypto: fix misleading trace message

Test was reporting digest verification failed for all operation errors. Fixed so it only reports this if the PMD actually reports an auth failure. Fixes: 9c0eed2f06ae ("app/test: rework crypto AES unit test") CC: sta...@dpdk.org Signed-off-by: Fiona Trahe --- test/test/test_cryptodev_blockciphe

Re: [dpdk-dev] [PATCH] telemetry: fix error when using ports of different types

19/12/2018 12:59, Bruce Richardson: > Different NIC ports can have different numbers of xstats on them, which > means that we can't just use the xstats list from the first port registered > in the telemetry library. Instead, we need to check the type of each port - > by checking its ops structure p

Re: [dpdk-dev] [RFC 00/14] prefix network structures

On Thu, 20 Dec 2018 21:59:37 + Ferruh Yigit wrote: > On 10/24/2018 9:18 AM, olivier.matz at 6wind.com (Olivier Matz) wrote: > > This RFC targets 19.02. > > > > The rte_net headers conflict with the libc headers, because > > some definitions are duplicated, sometimes with few differences. > >

Re: [dpdk-dev] [PATCH v2 1/4] bitmap: remove deprecated bsf64 function

20/12/2018 13:09, Anatoly Burakov: > The function rte_bsf64 was deprecated in a previous release, so > remove the function, and the deprecation notice associated with > it. > > Signed-off-by: Anatoly Burakov Applied with change suggested in patch 4, thanks

Re: [dpdk-dev] [PATCH / RFC ] ethdev: Allow rte_eth_dev_configure with zero RX/TX queues

On 6/23/2016 4:53 PM, thomas.monjalon at 6wind.com (Thomas Monjalon) wrote: > 2016-05-16 11:33, Simon Kagstrom: >> This allows releasing RX/TX queue memory. >> --- >> We're using DPDK 16.04 and have a test suite which performs a sequence >> of separate tests of the type >> >>allocate mempool >>

Re: [dpdk-dev] [PATCH v1 1/2] rte_memcmp functions using Intel AVX and SSE intrinsics

On 5/26/2016 9:57 AM, zhihong.wang at intel.com (Wang, Zhihong) wrote: >> -Original Message- >> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Ravi Kerur >> Sent: Tuesday, March 8, 2016 7:01 AM >> To: dev at dpdk.org >> Subject: [dpdk-dev] [PATCH v1 1/2] rte_memcmp functions using

Re: [dpdk-dev] [PATCH v2 4/4] common: add 64-bit log2 function

20/12/2018 13:09, Anatoly Burakov: > --- a/app/test-pmd/testpmd.c > +++ b/app/test-pmd/testpmd.c> > -static inline uint32_t > -bsf64(uint64_t v) > -{ > - return (uint32_t)__builtin_ctzll(v); > -} > - [...] > --- a/lib/librte_eal/linuxapp/eal/eal_memalloc.c > +++ b/lib/librte_eal/linuxapp/eal/ea

Re: [dpdk-dev] [PATCH] eal: fix runtime dir cleanup in noshconf mode

20/12/2018 12:11, Anatoly Burakov: > When using --no-shconf or --in-memory modes, there is no runtime > directory to be created, so there is no point in attempting to > clean it. > > Fixes: 0a529578f162 ("eal: clean up unused files on initialization") > > Signed-off-by: Anatoly Burakov Applied,

Re: [dpdk-dev] [dpdk-stable] [PATCH] test: add missing fbarray autotest to meson

> > Currently, fbarray autotest is only built by make, but is missing from > > meson build files. > > > > Fixes: 7985860c18af ("test/fbarray: add autotests") > > Cc: sta...@dpdk.org > > > > Signed-off-by: Anatoly Burakov > > Acked-by: Bruce Richardson Applied, thanks

Re: [dpdk-dev] [dpdk-stable] [PATCH] test: add missing external mem autotest to meson

05/12/2018 17:25, Anatoly Burakov: > The 'external_mem_autotest' was defined in the meson build, but > the actual source file was not being compiled by meson. > > Fixes: b270daa43b3d ("test: support external memory") > Cc: sta...@dpdk.org > > Signed-off-by: Anatoly Burakov Applied, thanks

Re: [dpdk-dev] [PATCH] doc: remove note on limitations for memory mode in secondary

06/12/2018 10:30, Anatoly Burakov: > Memory mode flags are now shared between primary and secondary > processes, so the in documentation about limitations is no longer > necessary. > > Fixes: 64cdfc35aaad ("mem: store memory mode flags in shared config") > Cc: sta...@dpdk.org > > Signed-off-by: A

Re: [dpdk-dev] [RFC 03/14] net: add rte prefix to ether structures

On 10/24/2018 9:18 AM, olivier.matz at 6wind.com (Olivier Matz) wrote: > diff --git a/lib/librte_cmdline/cmdline_parse_etheraddr.c > b/lib/librte_cmdline/cmdline_parse_etheraddr.c > index 24e04755c..c9cf33f19 100644 > --- a/lib/librte_cmdline/cmdline_parse_etheraddr.c > +++ b/lib/librte_cmdline/cm

Re: [dpdk-dev] [PATCH v3 0/5] Allow using virtio-user without hugepages

> Anatoly Burakov (5): > mem: fix error code for segment fd API for external segs > memalloc: check for memfd support in segment fd API > memalloc: allow setting up segment list fd's > mem: use memfd for no-huge mode > test: add segment fd API test Applied, thanks

Re: [dpdk-dev] [RFC 00/14] prefix network structures

On 10/24/2018 9:18 AM, olivier.matz at 6wind.com (Olivier Matz) wrote: > This RFC targets 19.02. > > The rte_net headers conflict with the libc headers, because > some definitions are duplicated, sometimes with few differences. > This was discussed in [1], and more recently at the techboard. > >

Re: [dpdk-dev] [PATCH v1] test/ring: ring perf test case enhancement

> > +Cc Olivier, maintainer of the ring library. > > 20/12/2018 12:33, Gavin Hu: > > From: Joyce Kong > > > > Run ring perf test on all available cores to really verify MPMC operations. > > The old way of running on a pair of cores is not enough for MPMC > > rings. We used this test case for rin

Re: [dpdk-dev] [PATCH v2] devtools: fix wrong headline lowercase for arm

> > Change to the new wording according to Arm corp: > https://mails.dpdk.org/archives/dev/2018-October/115174.html > otherwise, "Wrong headline lowercase" may be falsely reported. > > Fixes: f146ada21d1e ("scripts: add more git log checks") > Cc: sta...@dpdk.org > > Signed-off-by: Gavin Hu > -

Re: [dpdk-dev] [PATCH v2 0/4] hash: deprecate lock ellision and read/write concurreny flags

On 11/2/2018 5:38 PM, Honnappa Nagarahalli wrote: >> >> On Thu, Nov 01, 2018 at 06:25:18PM -0500, Honnappa Nagarahalli wrote: >>> Various configuration flags in rte_hash library result in increase of >>> number of test cases. Configuration flags for enabling transactional >>> memory use and read/wr

Re: [dpdk-dev] [PATCH 1/3] rte_ethdev: Add API function to read dev clock

On 12/19/2018 1:49 PM, Tom Barbette wrote: > +/** > + * @warning > + * @b EXPERIMENTAL: this API may change without prior notice. > + * > + * Read the current clock counter of an Ethernet device > + * > + * This returns the current raw clock value of an Ethernet device. > + * The value returned her

Re: [dpdk-dev] [PATCH 2/2] net/tap: add buffer overflow checks before checksum

> On Dec 17, 2018, at 9:50 AM, Richardson, Bruce > wrote: > > The checksum calculation APIs take only the packet headers pointers as > parameters, so they assume that the lengths reported in those headers are > correct. However, a malicious packet could claim to be far larger than it > is, so

Re: [dpdk-dev] [PATCH] net/cxgbe: implement reset hit counters for offloaded flows

On 12/14/2018 7:01 PM, Rahul Lakkireddy wrote: > Implement logic to reset hit counters for offloaded flows. > > Signed-off-by: Rahul Lakkireddy Applied to dpdk-next-net/master, thanks.

Re: [dpdk-dev] [PATCH] net/failsafe: remove unnecessary debug messages

On 12/20/2018 7:14 PM, Ferruh Yigit wrote: > On 12/14/2018 7:35 PM, Stephen Hemminger wrote: >> The failsafe driver spews messages every time it is polling >> making debugging other things more difficult. Just remove it. >> >> Example: >> net_failsafe: Hot-plug mutex was locked by thread 1401820664

Re: [dpdk-dev] [PATCH] net/failsafe: remove unnecessary debug messages

On 12/14/2018 7:35 PM, Stephen Hemminger wrote: > The failsafe driver spews messages every time it is polling > making debugging other things more difficult. Just remove it. > > Example: > net_failsafe: Hot-plug mutex was locked by thread 140182066431744 by the > hot-plug alarm > net_failsafe: Ho

Re: [dpdk-dev] [PATCH v7 2/2] mbuf: implement generic format for sched field

> -Original Message- > From: Olivier Matz [mailto:olivier.m...@6wind.com] > Sent: Thursday, December 20, 2018 12:43 PM > To: Pattan, Reshma > Cc: dev@dpdk.org; jerin.ja...@caviumnetworks.com; Rao, Nikhil > ; tho...@monjalon.net; Singh, Jasvinder > ; Dumitrescu, Cristian > ; Ananyev, Kon

Re: [dpdk-dev] [PATCH 1/2] sched: refine get base helper function

> -Original Message- > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Tonghao Zhang > Sent: Wednesday, November 28, 2018 1:56 PM > To: dev@dpdk.org > Cc: Tonghao Zhang > Subject: [dpdk-dev] [PATCH 1/2] sched: refine get base helper function > > use switch instead of if, and it is

Re: [dpdk-dev] [PATCH 0/2] prevent out of bounds read with checksum

On 12/17/2018 3:50 PM, Bruce Richardson wrote: > The functions for checksumming the packet payload don't perform bounds > checks, and are used by the TAP driver which does not do any bounds checks > on the incoming packet either. This means a packet received with an > incorrect IP header can read b

Re: [dpdk-dev] [PATCH 2/2] sched: fix possible mem leak on initialize

> -Original Message- > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Tonghao Zhang > Sent: Tuesday, December 11, 2018 2:54 AM > To: dev@dpdk.org > Subject: Re: [dpdk-dev] [PATCH 2/2] sched: fix possible mem leak on initialize > > ping > On Wed, Nov 28, 2018 at 9:56 PM Tonghao Zhan

Re: [dpdk-dev] [PATCH 2/2] net/tap: add buffer overflow checks before checksum

On 12/17/2018 3:50 PM, Bruce Richardson wrote: > The checksum calculation APIs take only the packet headers pointers as > parameters, so they assume that the lengths reported in those headers are > correct. However, a malicious packet could claim to be far larger than it > is, so we need to check t

Re: [dpdk-dev] [PATCH] app/test-pmd: expand test-pmd queue threshold size of RED parameters

> > > -Original Message- > > > From: Xu, Rosen > > > Sent: Monday, November 26, 2018 8:36 AM > > > To: dev@dpdk.org > > > Cc: Dumitrescu, Cristian ; Lu, Wenzhuo > > > ; Singh, Jasvinder ; > > > Xu, Rosen ; Yigit, Ferruh > > > ; Wu, Jingjing > > > Subject: [PATCH] app/test-pmd: expand tes

Re: [dpdk-dev] [dpdk-next-net] [PATCH] net/af_packet: fix docs guide

On 12/18/2018 7:10 AM, Rami Rosen wrote: > The framecnt option represents a counter (integer) that does not > reflect size. It is somewhat parallel in this sense to the framecount > member of pkt_rx_queue/pkt_tx_queue (or tp_frame_nr in tpacket_req). > This patch removes the size indicator of byte

[dpdk-dev] DPDK Release Status Meeting 20/12/2018

Minutes 20 December 2018 Agenda: * Release Dates * RC1 deadline * Subtrees * OvS * Opens Participants: * Debian * Intel * Mellanox * RedHat Release Dates - * v19.02 dates: * First release candidate: December 21, 2018 * Last release candidate: January 2

Re: [dpdk-dev] [PATCH] net/vhost: fix double free of MAC addrs

On 12/19/18 5:37 AM, Hideyuki Yamashita wrote: The common data freeing has been moved to rte_eth_dev_release_port(), so freeing mac_addrs like this in eth_dev_close() is unnecessary and will cause double free. Fixes: e16adf08e54d ("ethdev: free all common data when releasing port") Cc: sta...

Re: [dpdk-dev] [PATCH v2 0/4] vhost: add missing barriers, move prefetching

On 12/19/18 9:21 AM, Maxime Coquelin wrote: This series adds missing read barriers after reading avail index for split ring and desc flags for packed ring. Also, it turns out that some descriptors prefetching are either badly placed, or useless, last part of the series fixes that. With the s

Re: [dpdk-dev] [PATCH] net/virtio: avoid double accounting of bytes

On 12/20/18 11:56 AM, Jens Freimann wrote: Accounting of bytes was moved to a common function, so at the moment we do it twice. This patches fixes it for sending packets with packed virtqueues. Signed-off-by: Jens Freimann --- drivers/net/virtio/virtio_rxtx.c | 1 - 1 file changed, 1 dele

Re: [dpdk-dev] [PATCH v4] vhost: batch used descs chains write-back with packed ring

On Thu, Dec 20, 2018 at 05:47:55PM +0100, Maxime Coquelin wrote: > Instead of writing back descriptors chains in order, let's > write the first chain flags last in order to improve batching. > > Also, move the write barrier in logging cache sync, so that it > is done only when logging is enabled.

Re: [dpdk-dev] [PATCH v4 04/10] lib: introduce ipsec library

> > > + > > > +static int > > > +fill_crypto_xform(struct crypto_xform *xform, > > > + const struct rte_ipsec_sa_prm *prm) > > > +{ > > > + struct rte_crypto_sym_xform *xf; > > > + > > > + memset(xform, 0, sizeof(*xform)); > > > + > > > + for (xf = prm->crypto_xform; xf != NULL; xf = xf->next) { >

Re: [dpdk-dev] [PATCH] net/virtio: avoid double accounting of bytes

On 12/20/18 11:56 AM, Jens Freimann wrote: Accounting of bytes was moved to a common function, so at the moment we do it twice. This patches fixes it for sending packets with packed virtqueues. Signed-off-by: Jens Freimann --- drivers/net/virtio/virtio_rxtx.c | 1 - 1 file changed, 1 dele

Re: [dpdk-dev] [PATCH v4] crypto/aesni_mb: support plain SHA

Hi Fan, > -Original Message- > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Fan Zhang > Sent: Thursday, December 20, 2018 5:22 AM > To: dev@dpdk.org > Cc: akhil.go...@nxp.com; De Lara Guarch, Pablo > ; Nowak, DamianX > ; Krakowiak, LukaszX > Subject: [dpdk-dev] [PATCH v4] crypto/

[dpdk-dev] [PATCH v2 4/5] spinlock: reimplement with atomic one-way barrier builtins

The __sync builtin based implementation generates full memory barriers ('dmb ish') on Arm platforms. Using C11 atomic builtins to generate one way barriers. Here is the assembly code of __sync_compare_and_swap builtin. __sync_bool_compare_and_swap(dst, exp, src); 0x0090f1b0 <+16>:e0

[dpdk-dev] [PATCH v2 5/5] eal: fix clang compilation error on x86

When CONFIG_RTE_FORCE_INTRINSICS is enabled for x86, the clang compilation error was: include/generic/rte_atomic.h:215:9: error: implicit declaration of function '__atomic_exchange_2' is invalid in C99 include/generic/rte_atomic.h:494:9: error:

[dpdk-dev] [PATCH v2 3/5] test/spinlock: amortize the cost of getting time

Instead of getting timestamps per iteration, amortize its overhead can help getting more precise benchmarking results. Signed-off-by: Gavin Hu Reviewed-by: Joyce Kong --- test/test/test_spinlock.c | 28 +++- 1 file changed, 15 insertions(+), 13 deletions(-) diff --git a

[dpdk-dev] [PATCH v2 2/5] test/spinlock: get timestamp more precisely

To precisely benchmark the spinlock performance, uses the precise version of getting timestamps, which enforces the timestamps are obtained at the expected places. Signed-off-by: Gavin Hu Reviewed-by: Phil Yang --- test/test/test_spinlock.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions

[dpdk-dev] [PATCH v2 0/5] spinlock optimization and test case enhancements

V2: 1. FORCE_INTRINCIS is still an option for ppc/x86, although not is use by default, so don't remove it from generic file. 2. Fix the clang compiler error on x86 when the above FORCE_INTRINSICS is enabled. V1: 1. Remove the 1us delay outside of the locked region to really benchmark the

[dpdk-dev] [PATCH v2 1/5] test/spinlock: remove 1us delay for correct benchmarking

The test is to benchmark the performance of spinlock by counting the number of spinlock acquire and release operations within the specified time. A typical pair of lock and unlock operations costs tens or hundreds of nano seconds, in comparison to this, delaying 1 us outside of the locked region is

[dpdk-dev] [PATCH v3 3/3] net/virtio: improve batching in mergeable path

This patch improves both descriptors dequeue and refill, by using the same batching strategy as done in in-order path. Signed-off-by: Maxime Coquelin Tested-by: Jens Freimann Reviewed-by: Jens Freimann --- drivers/net/virtio/virtio_rxtx.c | 239 +-- 1 file changed,

[dpdk-dev] [PATCH v3 1/3] net/virtio: inline refill and offload helpers

Signed-off-by: Maxime Coquelin Reviewed-by: Jens Freimann --- drivers/net/virtio/virtio_rxtx.c | 8 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/net/virtio/virtio_rxtx.c b/drivers/net/virtio/virtio_rxtx.c index 8564f18a7..95d74d000 100644 --- a/drivers/net/virti

[dpdk-dev] [PATCH v3 0/3] net/virtio: Rx paths improvements

First version of this series did merge out-of-order mergeable and non-mergeable receive paths, but Intel STV team highlighted some performance regression when using multiqueue with two cores enqueueing descs on host, while a single core dequeues the two queues. I didn't manage to close the perform

[dpdk-dev] [PATCH v3 2/3] net/virtio: add non-mergeable support to in-order path

This patch adds support for in-order path when meargeable buffers feature hasn't been negotiated. Signed-off-by: Maxime Coquelin Reviewed-by: Tiwei Bie --- drivers/net/virtio/virtio_ethdev.c | 14 -- drivers/net/virtio/virtio_ethdev.h | 2 +- drivers/net/virtio/virtio_rxtx.c | 10

Re: [dpdk-dev] [PATCH v3 0/4] Allow using external memory without malloc

20/12/2018 17:16, Stephen Hemminger: > On Thu, 20 Dec 2018 15:32:37 + > Anatoly Burakov wrote: > > > Currently, the only way to use externally allocated memory > > is through rte_malloc API's. While this is fine for a lot > > of use cases, it may not be suitable for certain other use > > case

Re: [dpdk-dev] [PATCH v3 0/4] Allow using external memory without malloc

> Anatoly Burakov (4): > malloc: separate creating memseg list and malloc heap > malloc: separate destroying memseg list and heap data > mem: allow registering external memory areas > mem: allow usage of non-heap external memory in multiprocess Applied, thanks

[dpdk-dev] [PATCH v1 8/8] option: return an error when register fail

Make rte_option_register return a negative value when an error occur. Signed-off-by: Gaetan Rivet --- lib/librte_eal/common/include/rte_option.h | 6 +- lib/librte_eal/common/rte_option.c | 8 +--- 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/lib/librte_eal/com

[dpdk-dev] [PATCH v1 5/8] option: check against common option on register

Not only check against other registered options, but also common EAL options. This will mitigate user confusion. Signed-off-by: Gaetan Rivet --- lib/librte_eal/common/rte_option.c | 14 ++ 1 file changed, 14 insertions(+) diff --git a/lib/librte_eal/common/rte_option.c b/lib/librte

[dpdk-dev] [PATCH v1 6/8] option: fix log level of error in register

INFO is not correct when logging an error. Signed-off-by: Gaetan Rivet --- lib/librte_eal/common/rte_option.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/librte_eal/common/rte_option.c b/lib/librte_eal/common/rte_option.c index ae8a0e2ca..b2c93b220 100644 --- a/lib/l

[dpdk-dev] [PATCH v1 7/8] option: improve library documentation

Use doxygen to describe the main structure and describe a little more why it exists. Signed-off-by: Gaetan Rivet --- lib/librte_eal/common/include/rte_option.h | 8 ++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/lib/librte_eal/common/include/rte_option.h b/lib/librte_eal

[dpdk-dev] [PATCH v1 4/8] option: rename name field

option->opt_* is redundant. The field should also be constant. Signed-off-by: Gaetan Rivet --- lib/librte_eal/common/include/rte_option.h | 2 +- lib/librte_eal/common/rte_option.c | 8 lib/librte_telemetry/rte_telemetry.c | 2 +- 3 files changed, 6 insertions(+), 6 deleti

[dpdk-dev] [PATCH v1 3/8] option: add usage string

Add a usage string field in rte_option, allowing to display help to the user and describe which options are currently available. Signed-off-by: Gaetan Rivet --- lib/librte_eal/common/eal_common_options.c | 1 + lib/librte_eal/common/eal_private.h| 7 +++ lib/librte_eal/common/inclu

[dpdk-dev] [PATCH v1 1/8] option: use bare option string as name

Current options name can be passed with arbitrary format. Force the use of "--" prefix and thus POSIX long options format. This restricts the ability to introduce surprising options and will help future additional checks. Signed-off-by: Gaetan Rivet --- lib/librte_eal/common/rte_option.c | 6

[dpdk-dev] [PATCH v1 2/8] option: do not use static iterator

This is rather weird. Someone should have caught that during review. Signed-off-by: Gaetan Rivet --- lib/librte_eal/common/rte_option.c | 8 ++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/lib/librte_eal/common/rte_option.c b/lib/librte_eal/common/rte_option.c index 3fbc1

[dpdk-dev] [PATCH v1 0/8] Clean up rte_option

rte_option lib might benefit from a few improvements. Overview: - list registered options on /rte/app --help - check duplicates within base EAL options - rename some fields to be closer to standard struct option and just cleaner overall. - fix some strange code (static iterator?) -

[dpdk-dev] [PATCH v4] vhost: batch used descs chains write-back with packed ring

Instead of writing back descriptors chains in order, let's write the first chain flags last in order to improve batching. Also, move the write barrier in logging cache sync, so that it is done only when logging is enabled. It means there is now one more barrier for split ring when logging is enabl

Re: [dpdk-dev] [PATCH v3 0/4] Allow using external memory without malloc

On Thu, 20 Dec 2018 15:32:37 + Anatoly Burakov wrote: > Currently, the only way to use externally allocated memory > is through rte_malloc API's. While this is fine for a lot > of use cases, it may not be suitable for certain other use > cases like manual memory management, etc. > > This pat

[dpdk-dev] [PATCH v2] devtools: fix wrong headline lowercase for arm

Change to the new wording according to Arm corp: https://mails.dpdk.org/archives/dev/2018-October/115174.html otherwise, "Wrong headline lowercase" may be falsely reported. Fixes: f146ada21d1e ("scripts: add more git log checks") Cc: sta...@dpdk.org Signed-off-by: Gavin Hu --- devtools/check-gi

Re: [dpdk-dev] [PATCH v1] devtools: fix wrong headline lowercase for arm

> -Original Message- > From: Thomas Monjalon > Sent: Thursday, December 20, 2018 11:57 PM > To: Gavin Hu (Arm Technology China) > Cc: dev@dpdk.org; bruce.richard...@intel.com; ferruh.yi...@intel.com; > Honnappa Nagarahalli ; nd > ; sta...@dpdk.org > Subject: Re: [PATCH v1] devtools: fi

Re: [dpdk-dev] [PATCH v1 4/5] spinlock: move the implementation to arm specific file

On Thu, Dec 20, 2018 at 4:58 PM Gavin Hu (Arm Technology China) < gavin...@arm.com> wrote: > > > This change breaks the use of intrinsics in x86 case at least. > > This is true, if intrinsics is still an option for x86 even not is use by > default. > I will fix this in v2. > Please, also check pp

Re: [dpdk-dev] [PATCH v1 4/5] spinlock: move the implementation to arm specific file

> This change breaks the use of intrinsics in x86 case at least. This is true, if intrinsics is still an option for x86 even not is use by default. I will fix this in v2. Don't know why mail mail client has this issue, I worked with other mail threads correctly.

Re: [dpdk-dev] [PATCH v1] devtools: fix wrong headline lowercase for arm

20/12/2018 16:52, Gavin Hu: > Change to the new wording according to Arm corp: > https://mails.dpdk.org/archives/dev/2018-October/115174.html > otherwise, "Wrong headline lowercase" may be falsely reported. > > Fixes: f146ada21d1e ("scripts: add more git log checks") > Cc: sta...@dpdk.org > > Sig

[dpdk-dev] [PATCH v1] devtools: fix wrong headline lowercase for arm

Change to the new wording according to Arm corp: https://mails.dpdk.org/archives/dev/2018-October/115174.html otherwise, "Wrong headline lowercase" may be falsely reported. Fixes: f146ada21d1e ("scripts: add more git log checks") Cc: sta...@dpdk.org Signed-off-by: Gavin Hu --- devtools/check-gi

[dpdk-dev] [PATCH v3 2/4] malloc: separate destroying memseg list and heap data

Currently, destroying external heap chunk and its memseg list is part of one process. When we will gain the ability to unregister external memory from DPDK that doesn't have any heap structures associated with it, we need to be able to find and destroy memseg lists as well as heap data separately.

[dpdk-dev] [PATCH v3 4/4] mem: allow usage of non-heap external memory in multiprocess

Add multiprocess support for externally allocated memory areas that are not added to DPDK heap (and add relevant doc sections). Signed-off-by: Anatoly Burakov Acked-by: Yongseok Koh --- .../prog_guide/env_abstraction_layer.rst | 3 + lib/librte_eal/common/eal_common_memory.c | 42

[dpdk-dev] [PATCH v3 3/4] mem: allow registering external memory areas

The general use-case of using external memory is well covered by existing external memory API's. However, certain use cases require manual management of externally allocated memory areas, so this memory should not be added to the heap. It should, however, be added to DPDK's internal structures, so

[dpdk-dev] [PATCH v3 0/4] Allow using external memory without malloc

Currently, the only way to use externally allocated memory is through rte_malloc API's. While this is fine for a lot of use cases, it may not be suitable for certain other use cases like manual memory management, etc. This patchset adds another API to register memory segments with DPDK (so that AP

[dpdk-dev] [PATCH v3 1/4] malloc: separate creating memseg list and malloc heap

Currently, creating external malloc heap involves also creating a memseg list backing that malloc heap. We need to have them as separate functions, to allow creating memseg lists without creating a malloc heap. Signed-off-by: Anatoly Burakov Acked-by: Yongseok Koh --- lib/librte_eal/common/mall

Re: [dpdk-dev] [PATCH v3] vhost: batch used descs chains write-back with packed ring

On 12/20/18 3:30 PM, Michael S. Tsirkin wrote: On Thu, Dec 20, 2018 at 11:00:22AM +0100, Maxime Coquelin wrote: Instead of writing back descriptors chains in order, let's write the first chain flags last in order to improve batching. With Kernel's pktgen benchmark, ~3% performance gain is me

Re: [dpdk-dev] [PATCH v1 4/5] spinlock: move the implementation to arm specific file

On Thu, Dec 20, 2018 at 3:36 PM Gavin Hu (Arm Technology China) < gavin...@arm.com> wrote: > >> On Thu, Dec 20, 2018 at 11:44 AM Gavin Hu > wrote: > >> remove the hard code #ifdef RTE_FORCE_INTRINSICS, move the > implementation > >> to the arm specific file, x86 and POWER

[dpdk-dev] [PATCH v2 2/2] test/compress: add varied buffer input/outputs

Added unit test to check if a SGL buffer was added as an input and a Linear Buffer as output and vice versa so we can test if the application would process the different buffers properly. Signed-off-by: Marko Kovacevic Acked-by: Lee Daly --- test/test/test_compressdev.c | 173 ++

[dpdk-dev] [PATCH v2 1/2] test/compress: add out of space test

From: "Kovacevic, Marko" This patch adds new out of space testcase to check that the destination mbuf is smaller than required for the output of compression to ensure the driver doesn't crash and returns the valid error case. Signed-off-by: Marko Kovacevic Acked-by: Lee Daly --- V2: Added c

Re: [dpdk-dev] [PATCH v7 2/2] mbuf: implement generic format for sched field

> -Original Message- > From: Pattan, Reshma > Sent: Thursday, December 20, 2018 5:46 PM > To: dev@dpdk.org; jerin.ja...@caviumnetworks.com; Rao, Nikhil > ; olivier.m...@6wind.com; tho...@monjalon.net; > Singh, Jasvinder ; Dumitrescu, Cristian > ; Ananyev, Konstantin > > Cc: Pattan, Reshma

Re: [dpdk-dev] [PATCH v3] libs/power: add p-state driver compatibility

On 19/12/2018 8:31 PM, Thomas Monjalon wrote: 19/12/2018 10:09, Hunt, David: On 19/12/2018 3:18 AM, Thomas Monjalon wrote: 14/12/2018 14:11, Liang Ma: Previously, in order to use the power library, it was necessary for the user to disable the intel_pstate driver by adding “intel_pstate=disab

[dpdk-dev] [PATCH v4] libs/power: add p-state driver compatibility

Previously, in order to use the power library, it was necessary for the user to disable the intel_pstate driver by adding “intel_pstate=disable” to the kernel command line for the system, which causes the acpi_cpufreq driver to be loaded in its place. This patch adds the ability for the power libr

Re: [dpdk-dev] [PATCH v1 4/5] spinlock: move the implementation to arm specific file

>> > On Thu, Dec 20, 2018 at 1:47 PM David Marchand >> > wrote: >> > On Thu, Dec 20, 2018 at 11:44 AM Gavin Hu wrote: >>> remove the hard code #ifdef RTE_FORCE_INTRINSICS, move the implementation >>> to the arm specific file, x86 and P

Re: [dpdk-dev] [PATCH v1 4/5] spinlock: move the implementation to arm specific file

>> On Thu, Dec 20, 2018 at 11:44 AM Gavin Hu wrote: >> remove the hard code #ifdef RTE_FORCE_INTRINSICS, move the implementation >> to the arm specific file, x86 and POWER have their own implementations. > No, x86 and ppc define their own implementation when _not_ having

Re: [dpdk-dev] [dpdk-stable] [PATCH] malloc: make alignment requirements more stringent

14/12/2018 19:51, Yongseok Koh: > > > On Dec 14, 2018, at 3:54 AM, Anatoly Burakov > > wrote: > > > > The external heaps API already implicitly expects start address > > of the external memory area to be page-aligned, but it is not > > enforced or documented. Fix this by implementing additional

  1   2   >