2015-07-09 14:15, Stephen Hemminger:
> On Thu, 09 Jul 2015 22:41:23 +0200
> Thomas Monjalon wrote:
>
> > REJECTED
> >
> > 2015-07-09 11:15, Stephen Hemminger:
> > > Changes in this version:
> > > - rebase for PCI uio changes
> > > - fix version map for 2.1
> > > - add BCM57840 support
> >
2015-07-09 18:46, Yong Wang:
> On 7/9/15, 11:24 AM, "dev on behalf of stephen at networkplumber.org"
> wrote:
> >From: Stephen Hemminger
> >
> >These are update to earlier vmxnet3 driver changes. Mostly rebasing
> >to current layout. I skipped the jumbo MTU stuff which works for me
> >but
2015-06-04 17:48, Dumitrescu, Cristian:
> From: Stephen Hemminger [mailto:stephen at networkplumber.org]
> > The rte_sched statistics API should allow reading statistics without
> > clearing. Make auto-clear optional. In this version, this is handled
> > by deprecating the old API and adding a new
2015-04-01 11:36, Stephen Hemminger:
> On Wed, 1 Apr 2015 07:45:02 +
> Thomas Monjalon wrote:
>
> > 2015-03-24 23:54, Ananyev, Konstantin:
> > > From: Stephen Hemminger [mailto:stephen at networkplumber.org]
> > > > On Mon, 23 Mar 2015 16:45:44 +
> > > > "Ananyev, Konstantin" wrote:
> >
2015-07-09 16:01, Stephen Hemminger:
> From: Stephen Hemminger
>
> Customers often screen off info level messages, so raise log
> level of significant events
>
> Signed-off-by: Stephen Hemminger
It was acked by Bruce with a (not addressed) comment:
2015-07-09 16:01, Stephen Hemminger:
> From: Stephen Hemminger
>
> The ixgbe driver was not following DPDK convention and
> was leaving loggin always in even if LOG_LEVEL was configured
> to disable debug logs.
>
> Signed-off-by: Stephen Hemminger
It was acked by Bruce:
2015-07-09 16:26, Stephen Hemminger:
> From: Stephen Hemminger
>
> The build of DPDK may be done on a system where Linux headers
> in /usr/include (and therefore kernel version macro) are much
> older than the target runtime system.
It seems strange wanting to build a feature not present in the
2015-07-09 14:30, John McNamara:
> This patchset adds ethdev API to enable and read IEEE1588/802.1AS PTP
> timestamps from devices that support it. The following functions are added:
>
> rte_eth_timesync_enable()
> rte_eth_timesync_disable()
> rte_eth_timesync_read_rx_timestamp()
>
2015-07-09 10:43, Stephen Hemminger:
> From: Stephen Hemminger
>
> The internal API for some functions used by KNI changes in current
> kernel. KNI really needs to be submitted upstream (by Intel)
> to keep it maintainable.
>
> Signed-off-by: Stephen Hemminger
There was already a similar
2015-07-07 22:42, Rahul Lakkireddy:
> This series of patches do the following:
> 1. RX performance fixes for cxgbe Poll Mode Driver.
> 2. Add more supported Chelsio T5 device ids.
> 3. Update cxgbe documentation to include instructions on how to flash
>firmware and enable/disable flow control.
For x550 supports 2 new flow director modes, MAC VLAN and Cloud. The MAC VLAN
mode means the MAC and VLAN are monitored. The Cloud mode is for VxLAN and
NVGRE, and the tunnel type, TNI/VNI, inner MAC and inner VLAN are monitored. So,
there're a few new lookup fields for these 2 new modes, like
Hi Thomas,
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Thomas Monjalon
> Sent: Friday, July 10, 2015 1:54 AM
> To: Stephen Hemminger
> Cc: dev at dpdk.org; Stephen Hemminger
> Subject: Re: [dpdk-dev] [PATCH] kni: fix build with Linux kernel 4.1 or later
Commit 35b3313e322b ("pci: merge mapping functions for linux and bsd")
introduced a bug that all vfio will be
blocked.
Root cause is that VFIO_PRESENT is unaccessable in eal
common level.
This patch is to fix this.
Signed-off-by: Michael Qiu
---
lib/librte_eal/common/eal_common_pci.c |
Hi, Stephen
This patch does not work for fm10k with vfio, see error below:
EAL: PCI device :84:00.0 on NUMA socket 1
EAL: probe driver: 8086:15a4 rte_pmd_fm10k
EAL: PCI memory mapped at 0x7f198000
EAL: Trying to map BAR 2 that contains the MSI-X table. Trying offsets:
:,
2015-07-10 07:16, De Lara Guarch, Pablo:
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Thomas Monjalon
> > 2015-07-09 10:43, Stephen Hemminger:
> > > From: Stephen Hemminger
> > >
> > > The internal API for some functions used by KNI changes in current
> > > kernel. KNI really needs
From: "Chen Jing D(Mark)"
fm10k has 128 RETA entries in 32 registers, but it only initialized
first 32 when doing multiple rx queue configurations. This fix will
initialize all 128 entries instead.
Signed-off-by: Chen Jing D(Mark)
---
drivers/net/fm10k/fm10k_ethdev.c |
On 7/10/2015 4:19 PM, Chen, Jing D wrote:
> From: "Chen Jing D(Mark)"
>
> fm10k has 128 RETA entries in 32 registers, but it only initialized
> first 32 when doing multiple rx queue configurations. This fix will
> initialize all 128 entries instead.
>
> Signed-off-by: Chen Jing D(Mark)
On 2015/07/10 16:47, Michael Qiu wrote:
> Commit 35b3313e322b ("pci: merge mapping functions for linux and bsd")
>
> introduced a bug that all vfio will be
> blocked.
>
> Root cause is that VFIO_PRESENT is unaccessable in eal
> common level.
>
> This patch is to fix this.
>
> Signed-off-by:
On Thu, Jul 09, 2015 at 05:54:30PM +0100, Pablo de Lara wrote:
> Add new unit test for calculating the average table utilization,
> using random keys, based on number of entries that can be added
> until we encounter one that cannot be added (bucket if full).
>
> Also, replace current hash_perf
On Thu, Jul 09, 2015 at 04:37:48PM -0700, Stephen Hemminger wrote:
> From: Stephen Hemminger
>
> For applications that use m->userdata the initialization can
> be a signficant (10%) performance penalty.
>
> Rather than taking the cache penalty of initializing userdata
> in the receive handling,
On Tue, Jul 07, 2015 at 10:50:23 +, Burakov, Anatoly wrote:
> Hi Rahul,
>
> > However, unsigned long seems to be working fine for all builds.
>
> unsigned long it is then, if there aren't any other objections.
>
> Thanks,
> Anatoly
Hi Alejandro,
Are you planning to update the original
> > Due to API changes in netdevice.h in 4.1 kernel release, KNI modules
> > would not build. This patch set adds the properly checks to fix
> > compilation.
> >
> > Changes in v2:
> >
> > - Fixed vHost module build errors.
> >
> > Miguel Bernal Marin (4):
> > kni: fix
Hi Stephen,
> --- /dev/null
> +++ b/lib/librte_eal/linuxapp/eal/compat_vfio.h
> @@ -0,0 +1,181 @@
Wouldn't this need a GPL license header?
Thanks,Anatoly
2015-07-08 14:21, Bruce Richardson:
> On Wed, Jul 08, 2015 at 12:27:34PM +0100, Pablo de Lara wrote:
> > rte_hash structure should not be a public structure,
> > and therefore it should be moved to the C file and be declared
> > as internal. rte_hash_hash implementation is also moved
> > to the C
2015-07-10 10:11, Bruce Richardson:
> On Thu, Jul 09, 2015 at 05:54:30PM +0100, Pablo de Lara wrote:
> > Add new unit test for calculating the average table utilization,
> > using random keys, based on number of entries that can be added
> > until we encounter one that cannot be added (bucket if
On Sun, Jun 28, 2015 at 10:23:28PM +, De Lara Guarch, Pablo wrote:
> Hi Stephen,
>
> > -Original Message-
> > From: Stephen Hemminger [mailto:stephen at networkplumber.org]
> > Sent: Friday, June 26, 2015 5:50 PM
> > To: De Lara Guarch, Pablo
> > Cc: dev at dpdk.org
> > Subject: Re:
On 08/07/2015 16:53, John McNamara wrote:
> From: Tero Aho
>
> Extend eth_pcap rx and tx to support jumbo frames.
>
> On the receive side read large packets into multiple mbufs and
> on the transmit side convert them back to a single pcap buffer.
>
> Signed-off-by: Tero Aho
> Signed-off-by: John
On Fri, Jul 10, 2015 at 10:34 AM, Tetsuya Mukawa wrote:
> On 2015/07/10 16:47, Michael Qiu wrote:
> > Commit 35b3313e322b ("pci: merge mapping functions for linux and bsd")
> >
> > introduced a bug that all vfio will be
> > blocked.
> >
> > Root cause is that VFIO_PRESENT is unaccessable in eal
2015-07-10 20:25 GMT+09:00 David Marchand :
> On Fri, Jul 10, 2015 at 10:34 AM, Tetsuya Mukawa wrote:
>>
>> On 2015/07/10 16:47, Michael Qiu wrote:
>> > Commit 35b3313e322b ("pci: merge mapping functions for linux and bsd")
>> >
>> > introduced a bug that all vfio will be
>> > blocked.
>> >
>> >
On Fri, Jul 10, 2015 at 01:25:49PM +0200, David Marchand wrote:
> On Fri, Jul 10, 2015 at 10:34 AM, Tetsuya Mukawa wrote:
>
> > On 2015/07/10 16:47, Michael Qiu wrote:
> > > Commit 35b3313e322b ("pci: merge mapping functions for linux and bsd")
> > >
> > > introduced a bug that all vfio will be
2015-07-10 20:38 GMT+09:00 Bruce Richardson :
> On Fri, Jul 10, 2015 at 01:25:49PM +0200, David Marchand wrote:
>> On Fri, Jul 10, 2015 at 10:34 AM, Tetsuya Mukawa
>> wrote:
>>
>> > On 2015/07/10 16:47, Michael Qiu wrote:
>> > > Commit 35b3313e322b ("pci: merge mapping functions for linux and
2015-07-10 12:38, Bruce Richardson:
> On Fri, Jul 10, 2015 at 01:25:49PM +0200, David Marchand wrote:
> > On Fri, Jul 10, 2015 at 10:34 AM, Tetsuya Mukawa
> > wrote:
> > > On 2015/07/10 16:47, Michael Qiu wrote:
> > > > Commit 35b3313e322b ("pci: merge mapping functions for linux and bsd")
> > >
Alan, Pablo, any comment?
By the way, we need a maintainer for rte_power.
2015-07-07 16:36, Yong Liu:
> From: Marvin Liu
>
> Two socket Haswell cpus on cottonwoodPass have 72 cores totally. And this
> number oversize the default value of maximum core number 64.
>
> Total number of channels
v13 changes:
- Rebase beause patch 1588
v12 changes:
- Update coding style to match latest change over rte_ethdev.c (by shemming at
brocade.com)
v11 changes:
- Fixed a typo and clean coding style.
v10 changes:
- Merged const fix back to igb/ixgbe.
v9 changes:
- Fixed checkpatch errors.
v8
add two new apis: rte_eth_dev_default_mac_addr_set
and rte_eth_ethtool_dev_info to enable
reading device parameters (mac, register, eeprom,
pause, ring) based upon ethtool alike
data parameter sepcification.
Signed-off-by: Liang-Min Larry Wang
---
lib/librte_ether/Makefile | 1 +
add function to support ethtool ops:
- set_mac_addr
- get_reg_length
- get_regs
- get_eeprom_length
- get_eeprom
- set_eeprom
Signed-off-by: Liang-Min Larry Wang
---
drivers/net/ixgbe/ixgbe_ethdev.c | 178 +-
drivers/net/ixgbe/ixgbe_regs.h | 376
add function to support ethtool ops:
- set_mac_addr
- get_reg_length
- get_regs
- get_eeprom_length
- get_eeprom
- set_eeprom
Signed-off-by: Liang-Min Larry Wang
---
drivers/net/e1000/igb_ethdev.c | 175
drivers/net/e1000/igb_regs.h | 223
The example includes an ethtool library and two applications:
one application is a non- DPDK process (nic-control)
and the other is a DPDK l2fwd applicaiton (l2fwd-app).
The nic-control process sends ethtool alike device management
requests to l2fwd-app through a named pipe IPC. This example
is
Stephen,
> -Original Message-
> From: Stephen Hemminger [mailto:stephen at networkplumber.org]
> Sent: Friday, June 26, 2015 9:37 PM
> To: Wang, Liang-min
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v9 5/5] ixgbe/igb: changed register tables to
> const
>
> On Fri, 26 Jun 2015
pcap: add support for rx and tx byte counters in addition to existing rx and tx
packet counters
Signed-off-by: Klaus Degner
---
drivers/net/pcap/rte_eth_pcap.c | 22 --
1 file changed, 20 insertions(+), 2 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Sanford, Robert
> Sent: Thursday, July 2, 2015 10:05 PM
> To: Richardson, Bruce; Thomas Monjalon; dev at dpdk.org
> Subject: [dpdk-dev] Ethernet API - multiple post-RX-burst callbacks' run-
> order is opposite
2015-07-10 13:14, Mcnamara, John:
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Sanford, Robert
> > When one adds multiple post-RX-burst callbacks to a queue, their execution
> > order is the opposite of the order in which they are added.
>
> Hi,
>
> I think you are right and that
Silly questions: Why use rte_pktmbuf_clone()? Assuming that one is not
going to modify the mbuf at all, why not just increment the reference
count with rte_mbuf_refcnt_update()?
--
Thanks,
Robert
>Keith speaks truth. If I were going to do what you're describing, I would
>do the following:
>
On Fri, Jul 10, 2015 at 01:46:21PM +, Sanford, Robert wrote:
> Silly questions: Why use rte_pktmbuf_clone()? Assuming that one is not
> going to modify the mbuf at all, why not just increment the reference
> count with rte_mbuf_refcnt_update()?
>
Yep, that should work fine too.
/Bruce
> --
>
> -Original Message-
> From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> Sent: Friday, July 10, 2015 2:24 PM
> To: Mcnamara, John
> Cc: Sanford, Robert; Richardson, Bruce; dev at dpdk.org
> Subject: Re: Ethernet API - multiple post-RX-burst callbacks' run-order is
> opposite to
2015-07-10 13:52, Mcnamara, John:
> From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> > > If the patch is accepted I'll add a note to the release notes also.
> >
> > Why not doing the release notes change atomicly in the same patch?
>
> Mainly, because there isn't currently a clear
Thomas:
Could we roll back that commit or apply Changchun's patch?
/Huawei
On 7/1/2015 11:53 PM, Xie, Huawei wrote:
> On 7/1/2015 3:49 PM, Ouyang Changchun wrote:
>> This commit breaks virtio basic packets rx functionality:
>> d78deadae4dca240e85054bf2d604a801676becc
>>
>> The QEMU use 256 as
2015-07-10 14:05, Xie, Huawei:
> Thomas:
> Could we roll back that commit or apply Changchun's patch?
It is waiting an agreement with Changchun, symbolized by an Acked-by:
> On 7/1/2015 11:53 PM, Xie, Huawei wrote:
> > On 7/1/2015 3:49 PM, Ouyang Changchun wrote:
> >> This commit breaks virtio
Acked-by: Huawei Xie
On 7/6/2015 10:27 AM, Ouyang, Changchun wrote:
> The patch set fix vhost sample fails to start up in second time:
>
> It should call api to unregister vhost driver when sample exit/quit, then
> the socket file will be removed(by calling unlink), and thus make vhost sample
Now; what if we are to encapsulate the packet within a tunnel; such as GRE,
prior to sending it out to a mirrored-port ... Must the packet be copied into a
new buffer or can we use a DPDK primitive that is more efficient?
Sami.
> -Original Message-
> From: Bruce Richardson
On 6/17/2015 11:24 PM, Thomas Monjalon wrote:
> 2015-05-07 06:54, Xie, Huawei:
>> On 4/16/2015 7:48 PM, Pavel Boldin wrote:
>>> + /* Closing the source_fd */
>>> + ret = sys_close(eventfd_copy.source_fd);
>> Pavel:
>> Here we close the fd and re-install a new file on this fd later.
>>
NACK
2015-06-19 14:29, Cyril Chemparathy:
> This commit adds a forward count argument, which is used to terminate
> the test once a certain number of packets have been forwarded.
It is not a test but an example application.
This patch doesn't demonstrate a DPDK feature.
If you need to improve a
Helin, a comment?
2015-06-11 00:19, Damjan Marion:
> This patch improves performance of vectored rx on i40e devices.
>
> Signed-off-by: Damjan Marion
> ---
> drivers/net/i40e/i40e_rxtx.c | 5 +
> 1 file changed, 5 insertions(+)
>
> diff --git a/drivers/net/i40e/i40e_rxtx.c
On Fri, Jul 10, 2015 at 02:21:08PM +, Assaad, Sami (Sami) wrote:
> Now; what if we are to encapsulate the packet within a tunnel; such as GRE,
> prior to sending it out to a mirrored-port ... Must the packet be copied into
> a new buffer or can we use a DPDK primitive that is more efficient?
2015-06-16 23:09, Thomas Monjalon:
> 2015-05-27 13:45, Simon Kagstrom:
> > /proc/version_signature is the version for the host machine, but in
> > e.g., chroots, this does not need to match that DPDK is built for. Use
> > utsrelease.h from the kernel sources instead and fake the upload
> >
We definitely need a maintainer for e1000.
2015-05-24 16:17, motomu:
> From: Motomu Utsumi
>
> in add flexfiler it always write to same register address,
> so when try to adapt some filters only last one is enabled.
> I substitute value based on based on flex_filter->index for reg_off.
>
> in
Xuelin, have you given up with that patch?
2015-04-14 23:11, Ananyev, Konstantin:
> From: xuelin.shi at freescale.com [mailto:xuelin.shi at freescale.com]
> > txd->read.olinfo_status =
> > - rte_cpu_to_le_32(olinfo_status);
> > +
Xie,
Regarding the patches:
1. The replaced code in fourth patch is checked to be a copy-paste of the
`sys_close` syscall.
2. It is not uncommon for the applications to close FD making it allocated
for a different file. In our particular case the file is closed in the
*source* process and *added*
Hi Marion
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Damjan Marion
> Sent: Wednesday, June 10, 2015 3:19 PM
> To: dev at dpdk.org
> Cc: Damjan Marion
> Subject: [dpdk-dev] [PATCH] i40e: prefetch next mbuf in rx alloc code
>
> This patch improves
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Miguel Bernal Marin
> Sent: Friday, June 26, 2015 3:15 PM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH v2 2/4] kni: fix header_ops to build with 4.1
>
> rebuild member was removed from headers_ops in
On 7/10/2015 10:50 PM, Pavel Boldin wrote:
Xie,
Regarding the patches:
1. The replaced code in fourth patch is checked to be a copy-paste of the
`sys_close` syscall.
sys_close does extra cleanup than the replaced code. My concern is, for
example, sys_close will mark the fd as
Could one of you guys help to review the KNI vhost part?
Thanks,
Helin
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Miguel Bernal Marin
> Sent: Friday, June 26, 2015 3:15 PM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH v2 3/4] kni: fix function
Could one of you guys help to review the vhost part?
Thanks,
Helin
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Miguel Bernal Marin
> Sent: Friday, June 26, 2015 3:15 PM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH v2 4/4] kni: fix missing
> -Original Message-
> From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> Sent: Thursday, July 2, 2015 5:21 PM
> To: Mcnamara, John
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 2/3] doc: added guidelines on dpdk
> documentation
>
> Good idea to have guidelines for doc.
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Miguel Bernal Marin
> Sent: Friday, June 26, 2015 3:15 PM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH v2 1/4] kni: fix igb_ndo_bridge_getlink to build
> with
> 4.1
>
> ndo_bridge_getlink has changed
Don't know why previous mail get messed.
On 7/10/2015 10:50 PM, Pavel Boldin wrote:
Xie,
Regarding the patches:
1. The replaced code in fourth patch is checked to be a copy-paste of the
`sys_close` syscall.
sys_close does extra cleanup than the replaced coe. My concern is, for example,
2015-07-10 15:34, Zhang, Helin:
> Could one of you guys help to review the KNI vhost part?
This series is already applied.
Thanks
On Fri, 10 Jul 2015 10:32:17 +0100
Bruce Richardson wrote:
> On Thu, Jul 09, 2015 at 04:37:48PM -0700, Stephen Hemminger wrote:
> > From: Stephen Hemminger
> >
> > For applications that use m->userdata the initialization can
> > be a signficant (10%) performance penalty.
> >
> > Rather than
On Fri, 10 Jul 2015 10:13:41 +
"Burakov, Anatoly" wrote:
> Hi Stephen,
>
> > --- /dev/null
> > +++ b/lib/librte_eal/linuxapp/eal/compat_vfio.h
> > @@ -0,0 +1,181 @@
>
> Wouldn't this need a GPL license header?
>
> Thanks,Anatoly
No. It is perfectly valid to create new headers with
This patchset:
* Adds guidelines on the purpose and structure of the DPDK documentation,
how to build it and guidelines for creating it and adding to it.
* Moves and refactors the Doxygen guidelines from the Coding Style doc to
the Documentation Guidelines doc.
As with the other guidelines
Set the default literal block format for Sphinx docs
to 'none' instead of the implicit default of 'python'.
This means that default :: blocks won't have any
highlighting applied. Explicit highlighting via
the code-block:: directive is already available and
in use in the docs.
Signed-off-by: John
Added guidelines on the purpose and structure of the DPDK
documentation, how to build it and guidelines for creating it.
Signed-off-by: John McNamara
---
doc/guides/guidelines/documentation.rst | 579
doc/guides/guidelines/index.rst | 2 +
2 files
Moved and refactored the Doxygen guidelines from the Coding Style doc
to the Documentation Guidelines doc. Replaced the existing Doxygen
references with links.
Signed-off-by: John McNamara
---
doc/guides/guidelines/coding_style.rst | 134 ++
2015-07-10 15:42, Xie, Huawei:
> Don't know why previous mail get messed.
Same problem with this one.
It's maybe because you send HTML mail which is wrongly translated in plain text.
> On 7/10/2015 10:50 PM, Pavel Boldin wrote:
> Xie,
>
> Regarding the patches:
> 1. The replaced code in fourth
> > When initialize the hardware, the stat should be reset.
> > Otherwise when detach then attach port, the stat will not be re-init to
> > zero.
> >
> > Signed-off-by: Michael Qiu
>
> Acked-by: Bernard Iremonger
Applied, thanks
On 7/10/2015 11:50 PM, Thomas Monjalon wrote:
> 2015-07-10 15:42, Xie, Huawei:
>> Don't know why previous mail get messed.
> Same problem with this one.
> It's maybe because you send HTML mail which is wrongly translated in plain
> text.
Have something to to with Pavel? :). I remember the format
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Neil Horman
> Sent: Tuesday, July 7, 2015 2:44 PM
> To: Thomas Monjalon
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] mk: enable next abi in static libs
>
> On Tue, Jul 07, 2015 at 05:46:08AM -0700,
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Damjan Marion
> Sent: Wednesday, June 10, 2015 3:19 PM
> To: dev at dpdk.org
> Cc: Damjan Marion
> Subject: [dpdk-dev] [PATCH] i40e: prefetch next mbuf in rx alloc code
>
> This patch improves performance of
> > There's a parameter "autoneg on|off" in testpmd CLI "set flow_ctrl ...".
> > This
> > parameter is used to enable/disable auto negotiation for flow control. But
> > it's
> > not supported yet.
> > The auto negotiation is enabled by default, we have no way to disable it.
> > This
> > patch
> > If a descriptor the device drive is handling is the context descriptor, its
> > type
> > value will be 0x1.
> > When using the not operator ! to do the conditional check, if the expression
> > value is zero, the device driver will consider the transaction for this
> > descriptor
> > has been
This patchset is to replace the existing hash library with
a more efficient and functional approach, using the Cuckoo hash
method to deal with collisions. This method is based on using
two different hash functions to have two possible locations
in the hash table where an entry can be.
So, if a
This patch replaces the existing hash library with another approach,
using the Cuckoo Hash method to resolve collisions (open addressing),
which pushes items from a full bucket when a new entry tries
to be added in it, storing the evicted entry in an alternative location,
using a secondary hash
Added reset function to be able to empty the table,
without having to destroy and create it again.
Signed-off-by: Pablo de Lara
---
app/test/test_hash.c | 4 +---
app/test/test_hash_perf.c| 12 +++-
lib/librte_hash/rte_cuckoo_hash.c| 21
Since now rte_hash structure is private, a new function
has been added to let the user iterate through the hash table,
returning next key and data associated on each iteration,
plus the position where they were stored.
Signed-off-by: Pablo de Lara
---
app/test/test_hash.c | 66
Signed-off-by: Pablo de Lara
---
MAINTAINERS | 1 +
1 file changed, 1 insertion(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 5476a73..75a27e9 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -345,6 +345,7 @@ F: doc/guides/sample_app_ug/l3_forward_access_ctrl.rst
Hashes
M: Bruce Richardson
Two of the macros in rte_hash.h are now deprecated, so this patch
adds notice that they will be removed in 2.2.
Signed-off-by: Pablo de Lara
---
doc/guides/rel_notes/abi.rst | 1 +
1 file changed, 1 insertion(+)
diff --git a/doc/guides/rel_notes/abi.rst b/doc/guides/rel_notes/abi.rst
index
Updates hash library documentation, reflecting
the new implementation changes.
Signed-off-by: Pablo de Lara
---
doc/guides/prog_guide/hash_lib.rst | 138 +
doc/guides/prog_guide/index.rst| 4 ++
2 files changed, 129 insertions(+), 13 deletions(-)
diff
On Fri, 10 Jul 2015 02:53:51 +0200
Thomas Monjalon wrote:
> 2015-07-09 10:43, Stephen Hemminger:
> > From: Stephen Hemminger
> >
> > The internal API for some functions used by KNI changes in current
> > kernel. KNI really needs to be submitted upstream (by Intel)
> > to keep it maintainable.
Hi Rahul,
Go ahead. That's fine for me.
Thanks
On Fri, Jul 10, 2015 at 10:54 AM, Rahul Lakkireddy <
rahul.lakkireddy at chelsio.com> wrote:
> On Tue, Jul 07, 2015 at 10:50:23 +, Burakov, Anatoly wrote:
> > Hi Rahul,
> >
> > > However, unsigned long seems to be working fine for all builds.
On Fri, 10 Jul 2015 01:43:17 +0200
Thomas Monjalon wrote:
> 2015-07-09 16:26, Stephen Hemminger:
> > From: Stephen Hemminger
> >
> > The build of DPDK may be done on a system where Linux headers
> > in /usr/include (and therefore kernel version macro) are much
> > older than the target runtime
2015-07-02 12:18, Qiu, Michael:
> On 6/26/2015 10:37 AM, He, Shaopeng wrote:
> > The default MAC address is directly copied to Device Ethernet
> > Link address array in the device initialize phase, which
> > bypasses fm10k MAC address number check mechanism, and will
> > cause an error message
2015-07-10 08:20, Qiu, Michael:
> On 7/10/2015 4:19 PM, Chen, Jing D wrote:
> > From: "Chen Jing D(Mark)"
> >
> > fm10k has 128 RETA entries in 32 registers, but it only initialized
> > first 32 when doing multiple rx queue configurations. This fix will
> > initialize all 128 entries instead.
> >
On Fri, 10 Jul 2015 22:19:58 +0200
Thomas Monjalon wrote:
> 2015-07-10 10:29, Stephen Hemminger:
> > On Fri, 10 Jul 2015 01:43:17 +0200
> > Thomas Monjalon wrote:
> >
> > > 2015-07-09 16:26, Stephen Hemminger:
> > > > From: Stephen Hemminger
> > > >
> > > > The build of DPDK may be done on a
> -Original Message-
> From: Iremonger, Bernard
> Sent: Friday, July 3, 2015 7:04 AM
> To: dev at dpdk.org
> Cc: Zhang, Helin; Qiu, Michael; Iremonger, Bernard
> Subject: [PATCH v6 0/7] i40e: PCI Port Hotplug Changes
>
> Changes in V6:
> Rebased to latest code.
> Removed release of rx
On Fri, Jul 10, 2015 at 06:24:17PM +0100, Pablo de Lara wrote:
> This patchset is to replace the existing hash library with
> a more efficient and functional approach, using the Cuckoo hash
> method to deal with collisions. This method is based on using
> two different hash functions to have two
Hello:
Couple of questions for the 540-Twinville & "testpmd" app:
1. How do I specify multiple receive queues/port -
Used the command line option -rxq=2 and get this error..
***
Fail: nb_rxq(2) is greater than max_rx_queues(1)
EAL: Error - exiting with code: 1
This patchset is to replace the existing hash library with
a more efficient and functional approach, using the Cuckoo hash
method to deal with collisions. This method is based on using
two different hash functions to have two possible locations
in the hash table where an entry can be.
So, if a
This patch replaces the existing hash library with another approach,
using the Cuckoo Hash method to resolve collisions (open addressing),
which pushes items from a full bucket when a new entry tries
to be added in it, storing the evicted entry in an alternative location,
using a secondary hash
Added reset function to be able to empty the table,
without having to destroy and create it again.
Signed-off-by: Pablo de Lara
---
app/test/test_hash.c | 4 +---
app/test/test_hash_perf.c| 12 +++-
lib/librte_hash/rte_cuckoo_hash.c| 21
Since now rte_hash structure is private, a new function
has been added to let the user iterate through the hash table,
returning next key and data associated on each iteration,
plus the position where they were stored.
Signed-off-by: Pablo de Lara
---
app/test/test_hash.c | 66
1 - 100 of 123 matches
Mail list logo