dpdk.git
11 years agoexamples/dpdk_qat: minor change
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
examples/dpdk_qat: minor change

Signed-off-by: Intel
11 years agoexamples/dpdk_qat: fix cross compile check
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
examples/dpdk_qat: fix cross compile check

Signed-off-by: Intel
11 years agoexamples: work only on defined lcores
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
examples: work only on defined lcores

Signed-off-by: Intel
11 years agoexamples: time fixes
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
examples: time fixes

Signed-off-by: Intel
11 years agoexamples: use global RTE_MAX_ETHPORTS
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
examples: use global RTE_MAX_ETHPORTS

Signed-off-by: Intel
11 years agoexamples: numa updates
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
examples: numa updates

Signed-off-by: Intel
11 years agoapp/testpmd: update csumonly
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app/testpmd: update csumonly

Signed-off-by: Intel
11 years agoapp: fix build with GCC 4.8
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: fix build with GCC 4.8

GCC 4.8 was producing this error:
argument to ‘sizeof’ in ‘strncmp’ call is the same expression as the
second source; did you mean to provide an explicit length?
[-Werror=sizeof-pointer-memaccess]

Signed-off-by: Intel
11 years agoapp: only use one thread to run autotests
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: only use one thread to run autotests

Signed-off-by: Intel
11 years agoapp: fix build when some libs are disabled
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: fix build when some libs are disabled

Signed-off-by: Intel
11 years agoapp: rework ring tests
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: rework ring tests

Signed-off-by: Intel
11 years agoapp: add a test on mbuf alignment
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: add a test on mbuf alignment

Signed-off-by: Intel
11 years agoapp: add tests on --syslog, --no-shconf and --huge-dir flags
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: add tests on --syslog, --no-shconf and --huge-dir flags

Signed-off-by: Intel
11 years agoapp: add some tests
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: add some tests

Signed-off-by: Intel
11 years agoapp: remove pmac from Makefile
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: remove pmac from Makefile

Signed-off-by: Intel
11 years agoapp: fix Makefile
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app: fix Makefile

Signed-off-by: Intel
11 years agokni: disable ETHTOOL_GRXRINGS
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
kni: disable ETHTOOL_GRXRINGS

Signed-off-by: Intel
11 years agokni: fix build with kernel 3.8
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
kni: fix build with kernel 3.8

- timecompare (used for hardware timestamping) has been removed.
(see Linux commit 65f8f9a1c1db831e5159e3e3e50912d1f214cd0c)
Simply disable HW_TIME_STAMP feature because it is not used by KNI.

- annotations __devinit and __devexit have been removed.
(see Linux commit 54b956b903607f8f8878754dd4352da6a54a1da2)

Signed-off-by: Intel
11 years agokni: fix multi-process support
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
kni: fix multi-process support

Signed-off-by: Intel
11 years agoixgbe: force phy power up/down
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: force phy power up/down

Signed-off-by: Intel
11 years agoigb: force phy power up/down
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
igb: force phy power up/down

Signed-off-by: Intel
11 years agoixgbe: update VMDq/DCB support
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: update VMDq/DCB support

Signed-off-by: Intel
11 years agoigb: update VMDq/DCB support
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
igb: update VMDq/DCB support

Signed-off-by: Intel
11 years agoethdev: update VMDq/DCB support
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ethdev: update VMDq/DCB support

Signed-off-by: Intel
11 years agoethdev: set VMDq pool when adding mac address
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ethdev: set VMDq pool when adding mac address

Signed-off-by: Intel
11 years agoixgbe: add PF support
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: add PF support

Signed-off-by: Intel
11 years agoigb: add PF support
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
igb: add PF support

Signed-off-by: Intel
11 years agoethdev: add PF support
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ethdev: add PF support

Signed-off-by: Intel
11 years agoixgbe: support dual vlan
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: support dual vlan

Signed-off-by: Intel
11 years agoigb: fix max RX packet size and support dual VLAN
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
igb: fix max RX packet size and support dual VLAN

When in 'normal' mode, hardware is designed to receive up to 1522 bytes.
When in 'jumbo' mode, RLPML register must be updated so that hardware accepts
a 802.1q vlan header. If dual vlan is enabled, then a second vlan header is
expected, so update RLPML register according to dual vlan support.

Signed-off-by: Intel
11 years agoixgbe: retrieve RX available descriptors
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: retrieve RX available descriptors

Signed-off-by: Intel
11 years agoigb: retrieve RX available descriptors
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (02:00 +0200)]
igb: retrieve RX available descriptors

Signed-off-by: Intel
11 years agoe1000: retrieve RX available descriptors
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (02:00 +0200)]
e1000: retrieve RX available descriptors

Signed-off-by: Intel
11 years agoethdev: retrieve RX available descriptors
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ethdev: retrieve RX available descriptors

Signed-off-by: Intel
11 years agoixgbe: RSS RETA configuration
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: RSS RETA configuration

Signed-off-by: Intel
11 years agoigb: RSS RETA configuration
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
igb: RSS RETA configuration

Signed-off-by: Intel
11 years agoethdev: RSS RETA configuration
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ethdev: RSS RETA configuration

Signed-off-by: Intel
11 years agoixgbe: allocate structures on numa_node
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: allocate structures on numa_node

Signed-off-by: Intel
11 years agoethdev: store numa_node per device
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ethdev: store numa_node per device

Signed-off-by: Intel
11 years agoixgbe: allow unsupported SFP
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: allow unsupported SFP

Add RTE_LIBRTE_IXGBE_ALLOW_UNSUPPORTED_SFP build option

Signed-off-by: Intel
11 years agoixgbe: force inlining
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: force inlining

Signed-off-by: Intel
11 years agoixgbe: log tx code path
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ixgbe: log tx code path

Signed-off-by: Intel
11 years agoe1000: move workaround for wthresh on 82576 at wrong place
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
e1000: move workaround for wthresh on 82576 at wrong place

The workaround is moved in e1000em code but 82576 is an igb NIC.
It breaks the fix of the commit 7e9e49feea.

Signed-off-by: Intel <intel.com>
11 years agoigb: workaround errata with wthresh on 82576
Stephen Hemminger [Thu, 30 May 2013 17:12:35 +0000 (10:12 -0700)]
igb: workaround errata with wthresh on 82576

The 82576 has known issues which require the write threshold to be set to 1.
See:
http://download.intel.com/design/network/specupdt/82576_SPECUPDATE.pdf

If not then single packets will hang in transmit ring until more arrive.
Simple tests like ping will fail.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Reviewed-by: Vincent Jardin <vincent.jardin@6wind.com>
11 years agoethdev: more doxygen comments about fdir
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ethdev: more doxygen comments about fdir

Signed-off-by: Intel
11 years agopci: store numa_node per device
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
pci: store numa_node per device

Signed-off-by: Intel
11 years agopci: configure SR-IOV with max_vfs
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
pci: configure SR-IOV with max_vfs

Signed-off-by: Intel
11 years agopci: fix probing blacklisted device with RTE_PCI_DRV_MULTIPLE flag
Julien Courtat [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
pci: fix probing blacklisted device with RTE_PCI_DRV_MULTIPLE flag

When blacklisting network adapters that share a PCI bus address with multiple
ports such as Mellanox ConnectX-3 (librte_pmd_mlx4), RTE applications
enter infinite loops while probing PCI devices.

This commit removes this probe on subsequent instances for blacklisted devices
that have RTE_PCI_DRV_MULTIPLE set.

Signed-off-by: Julien Courtat <julien.courtat@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
11 years agopci: don't unbind resources on exit
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
pci: don't unbind resources on exit

unbind operations must be done before starting eal application

Signed-off-by: Intel
11 years agopci: rework interrupt handling
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
pci: rework interrupt handling

Signed-off-by: Intel <intel.com>
11 years agopci: support multiple PCI regions per device
Stephen Hemminger [Thu, 30 May 2013 17:12:39 +0000 (19:12 +0200)]
pci: support multiple PCI regions per device

Need to change PCI code to support multiple I/O regions on a single device.
Some devices like VMXNET3 have multiple PCI memory regions, and some
have none.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Intel
11 years agombuf: force inlining
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
mbuf: force inlining

Signed-off-by: Intel
11 years agomem: force inlining
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
mem: force inlining

Signed-off-by: Intel
11 years agomem: remove RLIMIT_NOFILE update
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
mem: remove RLIMIT_NOFILE update

Memory init won't try to increase its RLIMIT_NOFILE limit, this must be done
prior to starting it.

Signed-off-by: Intel
11 years agomem: find most little element from heap
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
mem: find most little element from heap

Allocations now return the most little element from heap matching size
constraint instead of the first one available.

Signed-off-by: Intel
11 years agoring: optimisations
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
ring: optimisations

Signed-off-by: Intel
11 years agotimer: add a rte_pause in a tight loop
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
timer: add a rte_pause in a tight loop

Signed-off-by: Intel
11 years agotimer: prefer TSC to HPET
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
timer: prefer TSC to HPET

Signed-off-by: Intel
11 years agotimer: optimize for empty case
Stephen Hemminger [Thu, 30 May 2013 17:12:36 +0000 (17:12 +0000)]
timer: optimize for empty case

In many application there are no timers queued, and the call to
rte_timer_managecan be optimized in that case avoid reading HPET and
lock overhead.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Reviewed-by: Vincent Jardin <vincent.jardin@6wind.com>
11 years agotimer: option --vmware-tsc-map for VMware guest
Damien Millescamps [Fri, 21 Sep 2012 11:36:40 +0000 (11:36 +0000)]
timer: option --vmware-tsc-map for VMware guest

The VMWare TSC mapping uses a hook to RDPMC to read the physical TSC
in the case of VMware ESXi.

Signed-off-by: Damien Millescamps <damien.millescamps@6wind.com>
Acked-by: Jean-Mickael Guerin <jmg@6wind.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Introduce new option --vmware-tsc-map, ignored if
CONFIG_RTE_LIBRTE_EAL_VMWARE_TSC_MAP_SUPPORT is not set.

Default is CONFIG_RTE_LIBRTE_EAL_VMWARE_TSC_MAP_SUPPORT=y.

if CONFIG_RTE_LIBRTE_EAL_VMWARE_TSC_MAP_SUPPORT is set:
    rte_rdtsc() selects at runtime between Vmware mapping of
    TSC or native TSC
else
    rte_rdtsc() always uses native rdtsc.

When running DPDK on VMware guest, enable --vmware-tsc-map to
read the physical TSC.
Caution: ESXi should pass monitor_control.pseudo_perfctr = TRUE
othewise it results in general protection fault.

Signed-off-by: Jean-Mickael Guerin <jean-mickael.guerin@6wind.com>
Acked-by: Vincent Jardin <vincent.jardin@6wind.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
11 years agohash: jhash optimisation
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
hash: jhash optimisation

Signed-off-by: Intel
11 years agohash: fix for multi-process apps
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
hash: fix for multi-process apps

Signed-off-by: Intel
11 years agosched: initial import
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
sched: initial import

Signed-off-by: Intel
11 years agometer: initial import
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
meter: initial import

Signed-off-by: Intel
11 years agopower: initial import
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
power: initial import

Signed-off-by: Intel
11 years agolpm: add IPv6 support
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
lpm: add IPv6 support

Signed-off-by: Intel
11 years agolog: optimize log/panic with attribute cold
Stephen Hemminger [Thu, 30 May 2013 17:12:37 +0000 (17:12 +0000)]
log: optimize log/panic with attribute cold

Both logging and calls to panic are never in the critical path.
Use the GCC attribute cold to mark these functions as cold,
which generates more optimised code.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Reviewed-by: Vincent Jardin <vincent.jardin@6wind.com>
11 years agolog: get full path as syslog id
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
log: get full path as syslog id

It partially reverts commit 04210699eee.

Signed-off-by: Intel <intel.com>
11 years agolog: add ability to override syslog parameters
Stephen Hemminger [Thu, 30 May 2013 17:12:41 +0000 (10:12 -0700)]
log: add ability to override syslog parameters

By default, DPDK based applications would only allow logging
to syslog as "rte", DAEMON; but for any production application more
control is desired to allow using actual application name and
overriding the facility.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Reviewed-by: Vincent Jardin <vincent.jardin@6wind.com>
11 years agoeal: cosmetic changes
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
eal: cosmetic changes

Signed-off-by: Intel
11 years agoeal: add application usage hook
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
eal: add application usage hook

Signed-off-by: Intel
11 years agoeal: memory barriers use intrinsic functions
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
eal: memory barriers use intrinsic functions

Signed-off-by: Intel
11 years agoeal: use intrinsic functions from compiler
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
eal: use intrinsic functions from compiler

RTE_FORCE_INTRINSICS makes it possible to force use of intrinsic
functions (defaults to n).

Signed-off-by: Intel
11 years agolib: add rte_hexdump and remove duplicated code
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
lib: add rte_hexdump and remove duplicated code

Signed-off-by: Intel
11 years agolib: fix various compilation warnings
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
lib: fix various compilation warnings

Signed-off-by: Intel
11 years agolib: 32/64-bit cleanups
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
lib: 32/64-bit cleanups

Signed-off-by: Intel
11 years agoconfig: change max lcore to 64
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
config: change max lcore to 64

Signed-off-by: Intel
11 years agoapp/testpmd: cosmetic changes
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
app/testpmd: cosmetic changes

Signed-off-by: Intel
11 years agodoc: fix file format (dos to unix)
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
doc: fix file format (dos to unix)

Signed-off-by: Intel
11 years agofile mode changes
Intel [Mon, 3 Jun 2013 00:00:00 +0000 (00:00 +0000)]
file mode changes

Signed-off-by: Intel
11 years agoset version to 1.3.1
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
set version to 1.3.1

Signed-off-by: Intel
11 years agoupdate copyright date to 2013
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
update copyright date to 2013

Signed-off-by: Intel
11 years agoexamples: various changes
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
examples: various changes

Signed-off-by: Intel
11 years agoapp: various changes
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
app: various changes

Signed-off-by: Intel
11 years agoapp: fix unused values
Zijie Pan [Wed, 5 Dec 2012 06:13:34 +0000 (14:13 +0800)]
app: fix unused values

Fix warnings of type "Value stored to 'xxx' is never read".

Signed-off-by: Zijie Pan <zijie.pan@6wind.com>
Acked-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
11 years agoapp: fix volatile read for GCC >= 4.6
Ivan Boule [Thu, 22 Sep 2011 16:01:35 +0000 (18:01 +0200)]
app: fix volatile read for GCC >= 4.6

When a memory address must be read, for instance a [mapped] PCI register,
the read value is assigned to a local variable that is not used after,
as for instance:
    x = *((uint8_t *) mem_addr);

Such instructions do not compile with gcc 4.6.
The fix consists in adding the "volatile" attribute to the accessed data type
and to not assign the read value:
    *((volatile uint8_t *) mem_addr);

Signed-off-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
11 years agoapp/testpmd: fix compliance with __rte_mbuf_sanity_check()
Ivan Boule [Tue, 26 Apr 2011 14:45:43 +0000 (16:45 +0200)]
app/testpmd: fix compliance with __rte_mbuf_sanity_check()

Build mbuf and lists of mbufs in a way compliant with the checks performed
by the function __rte_mbuf_sanity_check() when CONFIG_RTE_LIBRTE_MBUF_DEBUG=y

Signed-off-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
11 years agoapp/testpmd: fix config crash
Damien Millescamps [Tue, 15 May 2012 14:27:04 +0000 (16:27 +0200)]
app/testpmd: fix config crash

A chained topology must always have an increment of 1.
Here, it was 2 if ports number is even.

Signed-off-by: Damien Millescamps <damien.millescamps@6wind.com>
Acked-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
11 years agolib: minor changes
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
lib: minor changes

Signed-off-by: Intel
11 years agolib: fix non-C99 macros definitions in exported headers
Adrien Mazarguil [Wed, 19 Dec 2012 16:05:54 +0000 (17:05 +0100)]
lib: fix non-C99 macros definitions in exported headers

The original definitions prevent external programs/libraries from compiling
without warnings when using these headers and -std=gnu99 (relaxed C99 mode).

Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Acked-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Damien Millescamps <damien.millescamps@6wind.com>
11 years agolib: fix uninitialized value
Zijie Pan [Thu, 6 Dec 2012 07:36:51 +0000 (15:36 +0800)]
lib: fix uninitialized value

Fix warning "The left expression of the compound assignment
is an uninitialized value".

Signed-off-by: Zijie Pan <zijie.pan@6wind.com>
Acked-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
11 years agolib: fix uninitialized variables
Zijie Pan [Fri, 1 Feb 2013 11:43:29 +0000 (12:43 +0100)]
lib: fix uninitialized variables

Fix compilation errors caused by uninitialized variables.

Signed-off-by: Krzysztof Witek <krzysztof.witek@6wind.com>
Signed-off-by: Zijie Pan <zijie.pan@6wind.com>
Acked-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
11 years agolib: fix unused values
Zijie Pan [Tue, 4 Dec 2012 09:17:43 +0000 (17:17 +0800)]
lib: fix unused values

Fix warnings of type "Value stored to 'xxx' is never read".

Signed-off-by: Zijie Pan <zijie.pan@6wind.com>
Acked-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
11 years agokni: various changes
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
kni: various changes

Signed-off-by: Intel
11 years agokni: disable IGB_DCA
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
kni: disable IGB_DCA

Signed-off-by: Intel
11 years agokni: fix disabling
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
kni: fix disabling

Signed-off-by: Intel
11 years agokni: fix build with kernel < 2.6.34
Ivan Boule [Wed, 23 Jan 2013 13:22:14 +0000 (14:22 +0100)]
kni: fix build with kernel < 2.6.34

Signed-off-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
11 years agoigb_uio: fix build with kernel >= 3.8
Nicolas Dichtel [Wed, 20 Feb 2013 08:56:56 +0000 (09:56 +0100)]
igb_uio: fix build with kernel >= 3.8

__devinit has been removed in linux commit 54b956b903
("Remove __dev* markings from init.h", 3.8-rc4).

Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
11 years agoigb: fix RX for Springville (i210)
Intel [Tue, 12 Mar 2013 11:03:00 +0000 (12:03 +0100)]
igb: fix RX for Springville (i210)

Signed-off-by: Intel
11 years agopci: allow drivers to be bound several times to the same PCI device
Adrien Mazarguil [Wed, 19 Dec 2012 16:55:37 +0000 (17:55 +0100)]
pci: allow drivers to be bound several times to the same PCI device

Drivers with the flag RTE_PCI_DRV_MULTIPLE enabled will be bound several
times to the same device until they return an error.

These drivers must internally keep track of each device state in order for
this to work reliably.

This flag currently required for Ethernet adapters with several physical
ports but a single address on the PCI bus.

Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Acked-by: Ivan Boule <ivan.boule@6wind.com>
Acked-by: Damien Millescamps <damien.millescamps@6wind.com>