> > The DPDK is mostly consistent about using the POSIX standard
> > memset instead of bzero; but there seem to be some leftover BSD
> > hold outs.
> >
> > Stephen Hemminger (3):
> > test-pmd: use memset not bzero
> > xen-virt: use memset not bzero
> > qat: use memset instead of bzero
>
> S
2015-12-04 15:14, Bernard Iremonger:
> call rte_eth_copy_pci_info() after the RTE_PCI_DRV_INTR_LSC
> has been initialised.
>
> Fixes: eeefe73f0af1("drivers: copy PCI device info to ethdev data")
>
> Reported-by: Stephen Hemminger
> Signed-off-by: Bernard Iremonger
Applied, thanks
>
>The original was always setting unicast. While here, clean up some
>other references that also point into the Ethernet header.
>
>Signed-off-by: Chas Williams <3chas3 at gmail.com>
>---
> drivers/net/bnx2x/bnx2x.c | 23 +++
> drivers/net/bnx2x/ecore_hsi.h | 5 +++--
> 2 f
2015-12-06 23:20, Thomas Monjalon:
> 2015-11-18 15:58, Chas Williams:
> > The original was always setting unicast. While here, clean up some
> > other references that also point into the Ethernet header.
> >
> > Signed-off-by: Chas Williams <3chas3 at gmail.com>
>
> Applied, thanks
No sorry, no
2015-11-18 15:58, Chas Williams:
> The original was always setting unicast. While here, clean up some
> other references that also point into the Ethernet header.
>
> Signed-off-by: Chas Williams <3chas3 at gmail.com>
Applied, thanks
> >> Signed-off-by: Chas Williams <3chas3 at gmail.com>
>
> Acked-by: Harish Patil
Applied, thanks
2015-12-03 12:36, Rahul Lakkireddy:
> On Wednesday, December 12/02/15, 2015 at 14:16:18 -0800, Stephen Hemminger
> wrote:
> > The upcoming Hyper-V driver converts the pci_drv element
> > in struct eth_driver to a union. When vmbus is added the
> > pci_drv needs to be explicit. Easier to fix the i
2015-11-30 11:54, Jingjing Wu:
> When dev_stop is called in i40evf pmd driver, queues are switched off
> to stop receiving and transmitting. But the mac address of this VF
> still exists in VEB switch.
> To stop the traffic from VSI level, the mac address need to be removed
> too. Then the bandwidt
2015-11-30 08:33, Stephen Hemminger:
> On Mon, 30 Nov 2015 11:54:09 +0800
> Jingjing Wu wrote:
>
> > + (void)rte_memcpy(mac_addr.addr_bytes, hw->mac.addr,
> > + sizeof(mac_addr.addr_bytes))
>
> Please don't add useless (void) cast.
> I haven't seen that since lint on
2015-12-02 15:52, Andrey Chilikin:
> Fix wrong copy-paste which led to one bit missing from
> I40E_INSET_FLEX_PAYLOAD mask
>
> Signed-off-by: Andrey Chilikin
Applied, thanks
> > This patch fixes tx byte statistics when transmitting packets
> > with link down.
> >
> > Previously, the counter would decrement 4 bytes for each packet that
> > was transmitted with link down, causing the uint64 to wrap around.
> >
> > Fixes: c03fcee9abbd ("ixgbe: remove CRC size from byte
2015-12-03 16:35, Pablo de Lara:
> rte_hash_create function was accidentally duplicated in
> DPDK_2.1 in rte_hash_version.map.
>
> Fixes: 473d1beb ("hash: allow to store data in hash table")
>
> Reported-by: Ferruh Yigit
> Signed-off-by: Pablo de Lara
Applied, thanks
> > We can free memzones now, so remove incorrect doxygen documentation
> > stating otherwise.
> >
> > Fixes: ff909fe21f0a ("mem: introduce memzone freeing")
> >
> > Signed-off-by: Sergio Gonzalez Monroy
>
> Acked-by: John McNamara
Applied, thanks
2015-12-06 19:46, Thomas Monjalon:
> From: Panu Matilainen
>
> Similar to commit 113c8e13c4201eee207723571f83aaf285277d75, but
> for bnx2x, pcap, sze2data and xenvirt PMDs.
>
> Requiring applications to know about library internal details like
> dependencies to external helper libraries is a lim
existing rte_bitmap library implementation optimally configured to run on
64-bytes cache line, extending to 128-bytes cache line targets.
Signed-off-by: Jerin Jacob
---
lib/librte_sched/rte_bitmap.h | 12 +---
1 file changed, 9 insertions(+), 3 deletions(-)
diff --git a/lib/librte_sched
No need to split mbuf structure to two cache lines for 128-byte cache line
size targets as it can fit on a single 128-byte cache line.
Signed-off-by: Jerin Jacob
---
app/test/test_mbuf.c | 4
lib/librte_eal/linuxapp/eal/include/exec-env/rte_kni_commo
This patchset fixes performance/cache resource issues with 128-byte cache line
targets
found in mbuf and bitmap DPDK libraries
Currently, we have two DPDK targets(ThunderX and ppc_64) which are based on
128-bytes cache line size target.
This patchset doesn't introduce any performance degradation
use rte_prefetch_non_temporal() abstraction instead of _mm_prefetch(x, 0)
to in-order to build distributor application for non x86 platforms
Signed-off-by: Jerin Jacob
---
examples/distributor/main.c | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/examples/distributo
non-temporal/transient/stream version of rte_prefetch0()
The non-temporal prefetch is intended as a prefetch hint that processor
will use the prefetched data only once or short period,
unlike the rte_prefetch0() function which imply that
prefetched data to use repeatedly.
Signed-off-by: Jerin Jac
Introduced rte_prefetch_non_temporal() to remove IA specific _mm_prefect(addr,
0)
gcc intrinsic and build examples/distributor for non 86 platform
Not sure the rte_prefetch_non_temporal mapping correct for
all the platforms. Architecture maintainers please check the mapping for
rte_prefetch_non_t
From: Panu Matilainen
Similar to commit 113c8e13c4201eee207723571f83aaf285277d75, but
for bnx2x, pcap, sze2data and xenvirt PMDs.
Requiring applications to know about library internal details like
dependencies to external helper libraries is a limitation of
static linkage, shared libraries shoul
Some DPDK libraries have a dependency.
The Mellanox drivers embed this declaration in shared library case.
So the application do not need to know the dependency when linking.
But it cannot work with static libraries or the combined one.
Note that Mellanox drivers are currently not supported in a sh
> > Lookup burst size was changed for exact match from 4 to 8, for both ipv4 and
> > ipv6, but actually only
> > 4 keys were being looked up for ipv6, instead of 8, causing random
> > segmentation faults.
> >
> > Fixes: 80fcb4d4 ("examples/l3fwd: increase lookup burst size to 8")
> >
> > Signed-o
2015-12-02 14:29, Bruce Richardson:
> On Wed, Dec 02, 2015 at 01:06:20PM +, Sergio Gonzalez Monroy wrote:
> > Claim responsability for:
> > - Secondary Process as maintainer.
> > - FreeBSD EAL, FreeBSD contigmem and FreeBSD UIO as co-maintainer.
> >
> > Signed-off-by: Sergio Gonzalez Monroy
>
2015-11-27 14:31, Bruce Richardson:
> On Fri, Nov 27, 2015 at 02:14:04PM +, Pablo de Lara wrote:
> > Co-maintain helloworld, l2fwd and dpdk-qat sample apps.
> >
> > Signed-off-by: Pablo de Lara
> > ---
> > MAINTAINERS | 3 +++
> > 1 file changed, 3 insertions(+)
> >
> Thanks Pablo.
> While t
2015-12-06 21:29, Jerin Jacob:
> This patchset fixes performance/cache resource issues with 128-byte cache
> line targets
> found in mbuf and bitmap DPDK libraries
>
> Currently, we have two DPDK targets(ThunderX and ppc_64) which are based on
> 128-bytes cache line size target.
When introducing
2015-12-06 15:13, Yaacov Hazan:
> RTE_PMD_DEBUG_TRACE was changed also to support pedantic flag,
> but not completely.
>
> the macro changed only under the RTE_LIBRTE_ETHDEV_DEBUG define,
> but when RTE_LIBRTE_ETHDEV_DEBUG is not defined the old format
> was left.
>
> fix the macro to support ped
2015-12-03 14:45, Thomas Monjalon:
> Following the recent discussions, this is a proposal to have a standard
> installation process while keeping compatibility with most of the old
> behaviours.
[...]
> Thomas Monjalon (13):
> mk: remove testall
> mk: remove multi-target install
> mk: move in
2015-12-03 13:51, Ferruh Yigit:
> Fixes following error (observed when versioning macros used):
> LD libdpdk.so
> /usr/bin/ld: /root/dpdk/build/lib/libdpdk.so: version node not found
> for symbol @DPDK_x.y
>
> Also resulting combined library contains symbol version information:
> $ readelf -
RTE_PMD_DEBUG_TRACE was changed also to support pedantic flag,
but not completely.
the macro changed only under the RTE_LIBRTE_ETHDEV_DEBUG define,
but when RTE_LIBRTE_ETHDEV_DEBUG is not defined the old format
was left.
fix the macro to support pedantic flag in any case.
Signed-off-by: Yaacov H
-Original Message-
From: Thomas Monjalon [mailto:thomas.monja...@6wind.com]
Sent: Saturday, December 5, 2015 9:21 PM
To: Stephen Hemminger; Glynn, Michael J; Betts, Ian
Cc: dev at dpdk.org; O'Driscoll, Tim; Richardson, Bruce
Subject: Re: [dpdk-dev] [PATCH v8 0/4] examples: add performance-
?? ??!
???/viber/whatsapp +79133913837
Skype: prodawez389
ICQ: 6288862
Email: gorskova32 at gmail.com
??? ?!
2015-12-03 05:25, Thomas Monjalon:
> There is a new function in the EAL API for internal use.
> It has neither a proper prefix nor a .map export:
> libethdev.so: undefined reference to `is_xen_dom0_supported'
>
> Fixes: 719dbebceb81 ("xen: allow determining DOM0 at runtime")
>
> Signed-off-by: Th
2015-10-30 16:37, Ferruh Yigit:
> On Fri, Oct 30, 2015 at 02:17:58PM +, Harry van Haaren wrote:
> > This patch adds -s as an alias to --status in dpdk_nic_bind.py,
> > providing a convienient shorthand.
> >
> > v2:
> > - Corrected --summary to --status
> > - Fixed bug regarding -u usage (Thank
2015-12-04 18:07, Thomas Monjalon:
> The function rte_mempool_obj_iter used in mlx drivers
> was not exported. So the driver loading was failing:
>
> EAL: open shared lib librte_pmd_mlx4.so
> EAL: x86_64-native-linuxapp-gcc/lib/librte_pmd_mlx4.so:
> undefined symbol: rte_mempool_obj_iter
>
2015-12-05 17:27, Mike Sowka:
> Thanks for the pointers Thomas. Here is a signed-off patch
> re-submission with some explanation, to the best of my experience.
Thanks, it is really well detailed.
It would have been perfect with -v2 --in-reply-to :)
> Irrelevant of the target, the preprocessor #if
2015-12-01 07:20, Stephen Hemminger:
> On Tue, 1 Dec 2015 16:13:23 +0100
> Christian Ehrhardt wrote:
>
> > As it causes issues when building with RTE_MACHINE=default due to SSE4.x
> > requirements and in other discussions was so far rated "lightly tested and
> > doesn't provide really significan
37 matches
Mail list logo
