Re: [E1000-devel] Symmetric hashing for ixgbe driver?

works best with your flows. Thanks, -Don > -Original Message- > From: Cooper F. Nelson [mailto:cnel...@ucsd.edu] > Sent: Wednesday, November 16, 2016 4:23 PM > To: Skidmore, Donald C <donald.c.skidm...@intel.com>; e1000- > de...@lists.sf.net > Subject: Re: [E1000-

Re: [E1000-devel] Symmetric hashing for ixgbe driver?

Hey Cooper, Well RSS is a HW offload so I'm not sure how much can be done to "fix" this issue. That said with more resent drivers versions you can modify the RSS hash key and maybe try out the special "Random Secret Key" mentioned in the Suricata documentation. Likewise ATR may behave better

Re: [E1000-devel] rx queue

6, 2016 3:11 PM > To: Skidmore, Donald C <donald.c.skidm...@intel.com>; Tal Abudi > <talab...@gmail.com> > Cc: e1000-devel@lists.sourceforge.net > Subject: Re: [E1000-devel] rx queue > > Don, thank You for all Your answers. I need SFP modules capable card >

Re: [E1000-devel] rx queue

sure it contains at least 6bit RSS hash if not larger. Thanks, -Don Skidmore <donald.c.skidm...@intel.com> > -Original Message- > From: Łukasz Chrustek [mailto:luk...@chrustek.net] > Sent: Thursday, October 06, 2016 2:42 PM > To: Skidmore, Donald C <donald.c.skidm...@

Re: [E1000-devel] rx queue

for ATR. Which is why we included this counter in the ethtool stats. Thanks, -Don Skidmore <donald.c.skidm...@intel.com> > -Original Message- > From: Michał Purzyński [mailto:michalpurzyns...@gmail.com] > Sent: Thursday, October 06, 2016 2:04 PM > To:

Re: [E1000-devel] ETHTOOL_GLINKSETTINGS/SLINKSETTINGS

Hey Ryan, As for submitting a patch for get/set_link_ksettings the normal way it works is you would submit the patch upstream to Jeff's queue. We would look it over including some testing here. If it ended up getting in we would later back port it into the sourceforge driver and you would

Re: [E1000-devel] rx queue

Thursday, October 06, 2016 1:22 PM > To: Skidmore, Donald C <donald.c.skidm...@intel.com>; Tal Abudi > <talab...@gmail.com> > Cc: e1000-devel@lists.sourceforge.net > Subject: Re: [E1000-devel] rx queue > > Hello Don, > > So if it defaults RSS, is there any chan

Re: [E1000-devel] rx queue

chrustek.net] > Sent: Thursday, October 06, 2016 12:09 PM > To: Skidmore, Donald C <donald.c.skidm...@intel.com>; Tal Abudi > <talab...@gmail.com> > Cc: e1000-devel@lists.sourceforge.net > Subject: Re: [E1000-devel] rx queue > > Hello, > > I didn't change an

Re: [E1000-devel] rx queue

It looks like your just using RSS to spread most of your traffic between queues who's hash is limited to 4 bits (16 queues). This would happen if you turned off ATR or your flows weren't applicably for ATR. You should be able to verify this by the ethtool statistic. Thanks, -Don Skidmore

Re: [E1000-devel] ETHTOOL_GLINKSETTINGS/SLINKSETTINGS

Hey Ryan, Currently only X550 adapters support 2.5G/5G and even then we only support the speed being forced. The PHY was out befor the 802.3bz spec was finished so it doesn't completely support it. I would like to get around to moving to the new get/set_link_ksettings but I just haven't had

Re: [E1000-devel] Ubuntu ixgbe - Fake Tx hang detected with timeout

of your error just more as an example of issues we have seen in Linux. Which is why I think talking to VMware would be the way to go. Thanks, -Don Skidmore <donald.c.skidm...@intel.com> From: Victor Detoni [mailto:victordet...@gmail.com] Sent: Monday, October 03, 2016 6:29 PM To: Skidmore, Do

Re: [E1000-devel] Ubuntu ixgbe - Fake Tx hang detected with timeout

excessive pause frames. But ESX is really the wild card here, I would like to remove it from environment and see if you can recreate it. Thanks, -Don From: Victor Detoni [mailto:victordet...@gmail.com] Sent: Tuesday, September 27, 2016 10:16 AM To: Skidmore, Donald C <donald.c.skidm...@intel.com&

Re: [E1000-devel] Ubuntu ixgbe - Fake Tx hang detected with timeout

Hey Victor, Just to be clear you're talking about the ESX driver correct? Excuses my ignorance but I work primarily with the linux drivers so I wanted to be clear. But if it is ESX I can route your question to that team as I don't believe they regularly monitor this forum. Thanks, -Don

Re: [E1000-devel] AWS enhanced networking need for clarification

Hey Mike, Driver versions from in-distro drivers are kind of black magic at best. The norm seems to be to start with the version of the driver that was in the initial pull from kernel.org. And although they back port fixes/enhancements to the code base they don't change the driver version.

Re: [E1000-devel] Setting speed of ixgbe to be 1000

Try: Ethtool -s eth3 advertise 0x20 Thanks, -Don > -Original Message- > From: Kevin Wilson [mailto:wkev...@gmail.com] > Sent: Thursday, September 15, 2016 11:18 AM > To: e1000-devel@lists.sourceforge.net > Subject: [E1000-devel] Setting speed of ixgbe to

Re: [E1000-devel] Intel 82599 AXX10GBNIAIOM cards for 10G SFPs UDP performance issue

. It would be at least a good place to start. Thanks, -Don From: Hank Liu [mailto:hank.tz...@gmail.com] Sent: Wednesday, September 07, 2016 6:40 PM To: Skidmore, Donald C <donald.c.skidm...@intel.com> Cc: Rustad, Mark D <mark.d.rus...@intel.com>; e1000-devel@lists.sourceforge.net Subject

Re: [E1000-devel] Intel 82599 AXX10GBNIAIOM cards for 10G SFPs UDP performance issue

sessions what kind of throughput do you see then? Thanks, -Don <donald.c.skidm...@intel.com> From: Hank Liu [mailto:hank.tz...@gmail.com] Sent: Wednesday, September 07, 2016 5:42 PM To: Skidmore, Donald C <donald.c.skidm...@intel.com> Cc: Rustad, Mark D <mark.d.rus...@intel.co

Re: [E1000-devel] Intel 82599 AXX10GBNIAIOM cards for 10G SFPs UDP performance issue

. The later makes sense since all of your traffic is going to just two queue that appear to not be getting drained fast enough. Thanks, -Don <donald.c.skidm...@intel.com> From: Hank Liu [mailto:hank.tz...@gmail.com] Sent: Wednesday, September 07, 2016 4:51 PM To: Skidmore, Do

Re: [E1000-devel] Intel 82599 AXX10GBNIAIOM cards for 10G SFPs UDP performance issue

nk Liu [mailto:hank.tz...@gmail.com] Sent: Wednesday, September 07, 2016 3:40 PM To: Rustad, Mark D <mark.d.rus...@intel.com> Cc: Skidmore, Donald C <donald.c.skidm...@intel.com>; e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] Intel 82599 AXX10GBNIAIOM cards for 10G SFPs UDP

Re: [E1000-devel] Intel 82599 AXX10GBNIAIOM cards for 10G SFPs UDP performance issue

option (lspci -vvv to useful information. Thanks, -Don From: Hank Liu [mailto:hank.tz...@gmail.com] Sent: Wednesday, September 07, 2016 10:20 AM To: Skidmore, Donald C <donald.c.skidm...@intel.com> Cc: e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] Intel 82599 AXX10GBNIAIOM

Re: [E1000-devel] Intel 82599 AXX10GBNIAIOM cards for 10G SFPs UDP performance issue

> -Original Message- > From: Hank Liu [mailto:hank.tz...@gmail.com] > Sent: Wednesday, September 07, 2016 9:37 AM > To: e1000-devel@lists.sourceforge.net > Subject: [E1000-devel] Intel 82599 AXX10GBNIAIOM cards for 10G SFPs UDP > performance issue > > I run into 10G card performance

Re: [E1000-devel] [dpdk-dev] Unsupported SFP+ Module - Hardware Initialization Failure -19

The error value of -19 means your EEPROM indicates you're NIC doesn't have support for your SFP+ module. So you're correct this isn't specifically a DPDK issue, the code that verifies this is shared both between the Linux driver as well as DPDK. There are two possible cases here 1) we

Re: [E1000-devel] How does Perfect Filter work?

Any packet that doesn't hit a perfect filter is then subject to the RSS hash. Since you don't have any perfect filters RSS is hashing everything. In fact this is a good way to turn on RSS only (i.e. turn on ntuple but don't make any filters.) Thanks, -Don Skidmore

Re: [E1000-devel] Initializing Fortville nic in zero init mode

Hey Kevin, All I meant in the post you reference bellow is that ixgbe (not the driver that supports Fortville) does NOT support a zero init mode for unknown PHYs. I mention the reason for this in the post as well. We have never support such a mode and to my knowledge Fortville's driver

Re: [E1000-devel] X552 SFP+ link problems

> -Original Message- > From: Andre Tomt [mailto:an...@tomt.net] > Sent: Monday, August 08, 2016 4:41 PM > To: Skidmore, Donald C <donald.c.skidm...@intel.com>; e1000- > de...@lists.sourceforge.net > Subject: Re: [E1000-devel] X552 SFP+ link problems > > On

Re: [E1000-devel] Support for Symmetric RSS Hashing?

to hear it worked for you. :) -Don Skidmore <donald.c.skidm...@intel.com> From: Derek Ditch [mailto:de...@criticalstack.com] Sent: Wednesday, June 22, 2016 12:23 PM To: Fujinaka, Todd <todd.fujin...@intel.com> Cc: Skidmore, Donald C <donald.c.skidm...@intel.com>; e1000-devel@lis

Re: [E1000-devel] [PATCH 1/1] ixgbe: add fiber tranceiver plug/unplug notifier

From: zhuyj [mailto:zyjzyj2...@gmail.com] Sent: Thursday, June 16, 2016 3:43 AM To: e1000-devel@lists.sourceforge.net; netdev <net...@vger.kernel.org>; Skidmore, Donald C <donald.c.skidm...@intel.com> Cc: Zhu Yanjun <zyjzyj2...@gmail.com> Subject: Re: [PATCH 1/1] ixgbe: ad

Re: [E1000-devel] [PATCH 1/1] ixgbe: add fiber tranceiver plug/unplug notifier

> -Original Message- > From: zhuyj [mailto:zyjzyj2...@gmail.com] > Sent: Tuesday, June 14, 2016 6:55 AM > To: e1000-devel@lists.sourceforge.net; netdev > Subject: Re: [E1000-devel] [PATCH 1/1] ixgbe: add fiber tranceiver > plug/unplug notifier > > Hi, all > >

Re: [E1000-devel] ixgbe 4.3.13 x Linux 4.5.0 = Compilation Faillure

So this is an Ubuntu kernel not a kernel.org one, right? It doesn't mean it shouldn't work just I don't believe our validation actively checks new Ubuntu kernels. When issues do come up we will however write kcompat for the next release. Likewise the Source Forge driver can only be tested on

Re: [E1000-devel] allowing promiscuous mode between VF interfaces for ixgbe

Hey Tudor, I think you're correct in that the existence of the "trusted" state make something like this possible. However any final patch would have to would have to take at least the following factors in to account: - You would have to create a new version of the mailbox protocol (i.e. 1.3)

Re: [E1000-devel] error while compiling ixgbe

Have you tired the latest ixgbe driver 4.3.13? Thanks, -Don > -Original Message- > From: Srinivasreddy R [mailto:srinivasreddy4...@gmail.com] > Sent: Thursday, February 25, 2016 9:51 PM > To: e1000-devel@lists.sourceforge.net > Subject: [E1000-devel] error

Re: [E1000-devel] Issue report - ixgbe-4.4.0.19.tar.gz

Aizawa, Hiroshi [mailto:hiroshi.aiz...@avnet.com] > Sent: Tuesday, February 02, 2016 4:09 PM > To: Keller, Jacob E; Skidmore, Donald C; e1000-de...@lists.sf.net > Cc: Tsutsui, Akira > Subject: RE: Issue report - ixgbe-4.4.0.19.tar.gz > > Hi Don, > > Let me explain the back

Re: [E1000-devel] Issue report - ixgbe-4.4.0.19.tar.gz

Hey Hiroshi, Couple of things confuse me about your issue below: - You are using an unreleased driver (4.4.0.19). I'm wondering where you even got this driver. - You are using a bypass adapter. The standard linux driver does not support that device because it uses the SDP's for communication

Re: [E1000-devel] multicast promisc on VF not available in version 4.3.13

It doesn't work that way with ixgbe. If you are a trusted VF and you request to join more than 30 MC groups the PF will put you into MC Promisc mode. So you will also need a new version of iproute2 that has the trusted option and an OS that supports it. Thanks, -Don Skidmore

Re: [E1000-devel] multicast promisc on VF not available in version 4.3.13

Gill (ajgill) [mailto:ajg...@cisco.com] > Sent: Friday, January 22, 2016 3:20 PM > To: Skidmore, Donald C; e1000-devel@lists.sourceforge.net > Subject: Re: [E1000-devel] multicast promisc on VF not available in version > 4.3.13 > > Thanks Don. > That¹s makes it clear for me. > >

Re: [E1000-devel] ixgbe: link flapping on two directly connected servers with 82599ES adapter

BTW – I believe this is a link to the supported PHY types. Maybe one of these would work for you? http://www.intel.com/content/www/us/en/support/network-and-i-o/ethernet-products/05528.html?wapkw=compatible+sfp Thanks, -Don <donald.c.skidm...@intel.com> From: Skidmore, Donald

Re: [E1000-devel] ixgbe: link flapping on two directly connected servers with 82599ES adapter

Hey Peter, Sorry our hardware doesn't support LX4 modules, which is what I'm assuming you mean by WDM SFP+. The unsupported module parameter is only used for modules that we recognize but haven't validated. This way we at least know the module won't hurt the hardware even if we can't vouch

Re: [E1000-devel] 404 response for some downloads

I'm having the same issue with anything I attempt to download from Source Forge. We don't directly control the hosting site, it might be worthwhile to report this to Source Forge itself. Thanks, -Don Skidmore > -Original Message- > From: Steven Noonan

Re: [E1000-devel] ixgbe and using iptables/SYNPROXY causes random system resets

Hey Steffen, Sorry to hear about your system reset problem. Since you're not seeing anything in the system logs is you thinking that the NIC's are involved due to the frequency of the failures seems to correspond to amount of system traffic? I'm asking as in my 7 years of maintaining the

Re: [E1000-devel] ixgbe and using iptables/SYNPROXY causes random system resets

randeburg, Jesse > Sent: Tuesday, December 08, 2015 12:36 PM > To: Steffen Weber; Skidmore, Donald C; e1000-devel@lists.sourceforge.net > Subject: RE: [E1000-devel] ixgbe and using iptables/SYNPROXY causes > random system resets > > Also because the problem is maybe related to SY

Re: [E1000-devel] ixgbe and using iptables/SYNPROXY causes random system resets

ay, December 08, 2015 12:12 PM > To: Skidmore, Donald C; e1000-devel@lists.sourceforge.net > Subject: Re: [E1000-devel] ixgbe and using iptables/SYNPROXY causes > random system resets > > Hi Don, > > thank you for the reply. > > We've been using the X520 NICs since

Re: [E1000-devel] Random crashes with the IXGBE driver.

Hey Nikolay, Sorry to hear about you random crashes, it is strange they are only occurring on certain HW. I didn't notice the driver (ixgbe) in any of the stack dumps. Are you just inferring that the driver is a likely candidate as the skb head looks to be messed up? A couple of things come

Re: [E1000-devel] Fake TX hangs

Hey Tal Abudi, The Fake Tx hang message means that the stack is trying to reset the driver since it "thinks" we are hung however the driver doesn't believe that there is anything it can transmit. This is most often cased but excessive flow control or a faulty switch. What does your ethtool

Re: [E1000-devel] new ixgbe driver for kernel 4.1 and newer?

These version numbers don't mean as much as you would expect them too. We try very hard to keep the source forge driver has close as possible to the current in kernel driver but there are differences for various reasons. (i.e. kcompat issues, differing module parameters, etc). That said we

Re: [E1000-devel] [linux-nics] bad pages when up/down network cable

Hey Xishi, You are using a really old driver, our current is 4.1.1 and 3.9 dates to over 3 years ago. We have makes changes to support surprise removal that very well may address this issue. Have you attempted to recreate this failure with the latest out of tree driver? Thanks, -Don

Re: [E1000-devel] 82599, SR-IOV and ntuple filters

-Original Message- From: Eugene L. Vorokov [mailto:v...@pidarasy.org] Sent: Tuesday, June 30, 2015 3:54 PM To: Skidmore, Donald C Cc: e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] 82599, SR-IOV and ntuple filters Hi Donald, Thank you very much for your reply

Re: [E1000-devel] VF inside VM with bypassed PCI-E ixgbe

Hey Pavel, In order to support the bypass drive you need a special driver. I'm surprised you weren't supplied with one when you bought the NIC. Without it you can't communicate with the uC that handles all the bypass functionality. Likewise the SDP aren't connected to what the upstream

Re: [E1000-devel] [PATCH v2 2/3] if_link: Add VF multicast promiscuous control

-Original Message- From: Edward Cree [mailto:ec...@solarflare.com] Sent: Monday, February 23, 2015 5:53 AM To: Skidmore, Donald C Cc: Hiroshi Shimamoto; vyase...@redhat.com; Kirsher, Jeffrey T; Alexander Duyck; Bjørn Mork; e1000-devel@lists.sourceforge.net; net...@vger.kernel.org

Re: [E1000-devel] [PATCH v2 2/3] if_link: Add VF multicast promiscuous control

-Original Message- From: Edward Cree [mailto:ec...@solarflare.com] Sent: Friday, February 20, 2015 5:52 AM To: Hiroshi Shimamoto Cc: Skidmore, Donald C; vyase...@redhat.com; Kirsher, Jeffrey T; Alexander Duyck; Bjørn Mork; e1000-devel@lists.sourceforge.net; net...@vger.kernel.org

Re: [E1000-devel] [PATCH 1/3] ixgbe, ixgbevf: Add new mbox API to enable MC promiscuous mode

-Original Message- From: Hiroshi Shimamoto [mailto:h-shimam...@ct.jp.nec.com] Sent: Sunday, February 15, 2015 8:54 PM To: vyase...@redhat.com; Skidmore, Donald C; Kirsher, Jeffrey T Cc: Alexander Duyck; Bjørn Mork; e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy

Re: [E1000-devel] [PATCH 1/3] ixgbe, ixgbevf: Add new mbox API to enable MC promiscuous mode

-Original Message- From: Hiroshi Shimamoto [mailto:h-shimam...@ct.jp.nec.com] Sent: Thursday, February 12, 2015 8:45 PM To: Skidmore, Donald C; Kirsher, Jeffrey T Cc: Alexander Duyck; Bjørn Mork; e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy Jong; linux-ker

Re: [E1000-devel] [PATCH 1/3] ixgbe, ixgbevf: Add new mbox API to enable MC promiscuous mode

-Original Message- From: Hiroshi Shimamoto [mailto:h-shimam...@ct.jp.nec.com] Sent: Monday, February 09, 2015 6:29 PM To: Kirsher, Jeffrey T Cc: Alexander Duyck; Skidmore, Donald C; Bjørn Mork; e1000- de...@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy Jong; linux- ker

Re: [E1000-devel] [PATCH 1/3] ixgbe, ixgbevf: Add new mbox API to enable MC promiscuous mode

-Original Message- From: Hiroshi Shimamoto [mailto:h-shimam...@ct.jp.nec.com] Sent: Friday, January 30, 2015 3:37 AM To: Alexander Duyck; Skidmore, Donald C; Bjørn Mork Cc: e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy Jong; linux-ker...@vger.kernel.org; David

Re: [E1000-devel] [PATCH 1/2] if_link: Add VF multicast promiscuous mode control

-Original Message- From: Alexander Duyck [mailto:alexander.du...@gmail.com] Sent: Thursday, January 22, 2015 11:18 PM To: Hiroshi Shimamoto; Skidmore, Donald C; Bjørn Mork Cc: e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy Jong; linux-ker...@vger.kernel.org

Re: [E1000-devel] UDP RSS for fragmented packet

-Original Message- From: Fan Du [mailto:fengyuleidian0...@gmail.com] Sent: Wednesday, January 21, 2015 9:20 PM To: Skidmore, Donald C Cc: e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] UDP RSS for fragmented packet 于 2015年01月22日 01:34, Skidmore, Donald C 写道: Hey

Re: [E1000-devel] [PATCH 1/2] if_link: Add VF multicast promiscuous mode control

-Original Message- From: David Laight [mailto:david.lai...@aculab.com] Sent: Thursday, January 22, 2015 1:50 AM To: Skidmore, Donald C; Hiroshi Shimamoto; Bjørn Mork Cc: e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy Jong; linux-ker...@vger.kernel.org; Hayato

Re: [E1000-devel] Issue with linuxptp and ixgbe 3.23.2

Hey Scott, First thanks for the heads up, we appreciate it. It looks like we are able to recreate this in internally so we are adding it to our bug list. Hopefully we can get a fix out soon. Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: Scott Silverman

Re: [E1000-devel] UDP RSS for fragmented packet

Hey Fan, The issue is with our HW RSS doesn't work with frag packets. Sorry. Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: Fan Du [mailto:fengyuleidian0...@gmail.com] Sent: Wednesday, January 21, 2015 12:02 AM To: e1000-devel@lists.sourceforge.net

Re: [E1000-devel] [PATCH 1/2] if_link: Add VF multicast promiscuous mode control

-Original Message- From: Hiroshi Shimamoto [mailto:h-shimam...@ct.jp.nec.com] Sent: Wednesday, January 21, 2015 4:18 AM To: David Laight; Skidmore, Donald C; Bjørn Mork Cc: e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy Jong; linux-ker...@vger.kernel.org

Re: [E1000-devel] [PATCH 1/2] if_link: Add VF multicast promiscuous mode control

-Original Message- From: Hiroshi Shimamoto [mailto:h-shimam...@ct.jp.nec.com] Sent: Tuesday, January 20, 2015 3:40 PM To: Bjørn Mork Cc: e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy Jong; linux-ker...@vger.kernel.org; Hayato Momma Subject: Re: [E1000-devel]

Re: [E1000-devel] [PATCH 1/2] if_link: Add VF multicast promiscuous mode control

-Original Message- From: Hiroshi Shimamoto [mailto:h-shimam...@ct.jp.nec.com] Sent: Tuesday, January 20, 2015 5:07 PM To: Skidmore, Donald C; Bjørn Mork Cc: e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Choi, Sy Jong; linux-ker...@vger.kernel.org; Hayato Momma Subject

Re: [E1000-devel] Question - Adding per CPU counters for Multi Queue

Hey Tal, The easiest way to see all the packets per CPU would be to bind the queues to a given CPU they the queue counters would contain the information you're looking for. You could do this with the set_irq_affinity script we provide. Currently the driver doesn't have any structure that is

Re: [E1000-devel] Driver 4 Intel NetEffect 10Ge ?

The ixgbe driver doesn't support any of the neteffect devices, they have their own driver. Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: John Donnelly [mailto:b24warb...@gmail.com] Sent: Monday, January 05, 2015 3:45 PM To:

Re: [E1000-devel] ixgbe use only one queue for outgoing vlan when AtrSampleRate=0

donald.c.skidm...@intel.com -Original Message- From: Viktor Khomyuk [mailto:v.khom...@office.ngs.ru] Sent: Friday, December 19, 2014 6:20 AM To: Skidmore, Donald C; Fujinaka, Todd; Tantilov, Emil S; e1000- de...@lists.sourceforge.net Subject: Re: [E1000-devel] ixgbe use only one queue

Re: [E1000-devel] RSS related questions

Hey Liu, I’ll attempt to answer your questions inline below: Thanks, -Don Skidmore donald.c.skidm...@intel.com From: lzh [mailto:lhqlzh...@163.com] Sent: Thursday, December 18, 2014 10:12 PM To: Skidmore, Donald C; Tom Barbette Cc: e1000-devel@lists.sourceforge.net Subject: Re:RE: [E1000-devel

Re: [E1000-devel] RSS related questions

Hey Liu, Couple of things you're probably seeing. First off just because you have 20 queues doesn't mean your flows will hash to use all of those queues. In fact if you are testing this by sending traffic from just one system I would expect you to be using only one queue as you would only

Re: [E1000-devel] ixgbe use only one queue for outgoing vlan when AtrSampleRate=0

, December 17, 2014 11:23 PM To: Fujinaka, Todd; Skidmore, Donald C; Tantilov, Emil S; e1000- de...@lists.sourceforge.net Subject: Re: [E1000-devel] ixgbe use only one queue for outgoing vlan when AtrSampleRate=0 For simplicity's sake we can talk only about traffic generator. When AtrSampleRate=1

Re: [E1000-devel] filtering on mac dst address

and ixgbevf drivers but I don't believe anything would be different for 82576. Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: alex nln [mailto:alex...@yahoo.com] Sent: Monday, December 15, 2014 10:11 PM To: Skidmore, Donald C; e1000-devel@lists.sourceforge.net

Re: [E1000-devel] filtering on mac dst address

work for you? Sorry, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: alex nln [mailto:alex...@yahoo.com] Sent: Tuesday, December 16, 2014 4:13 PM To: Skidmore, Donald C; e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] filtering on mac dst address

Re: [E1000-devel] filtering on mac dst address

I'm a little confused on what you're asking for here. You want the VF driver to filter traffic it is transmitting to a certain queue or drop the packet all together? Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: alex nln [mailto:alex...@yahoo.com]

Re: [E1000-devel] ixgbe use only one queue for outgoing vlan when AtrSampleRate=0

donald.c.skidm...@intel.com -Original Message- From: Viktor Khomyuk [mailto:v.khom...@office.ngs.ru] Sent: Sunday, December 14, 2014 8:14 PM To: Tantilov, Emil S; Skidmore, Donald C; e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] ixgbe use only one queue for outgoing vlan when

Re: [E1000-devel] ixgbe use only one queue for outgoing vlan when AtrSampleRate=0

Khomyuk [mailto:v.khom...@office.ngs.ru] Sent: Wednesday, December 10, 2014 7:54 PM To: Skidmore, Donald C; e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] ixgbe use only one queue for outgoing vlan when AtrSampleRate=0 Thank You for answer. I have application with 8 threads

Re: [E1000-devel] rx_no_dma_resources woes

Skidmore donald.c.skidm...@intel.com -Original Message- From: Stefan Puiu [mailto:stefan.p...@gmail.com] Sent: Tuesday, December 09, 2014 2:28 AM To: Skidmore, Donald C Cc: e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] rx_no_dma_resources woes Hi Don, Thanks for your

Re: [E1000-devel] rx_no_dma_resources woes

Hey Stefan, I'm not sure why you are seeing more rx_no_dma_resource on one version of Ubuntu than another. We sadly don't validate against Ubuntu. What I can tell you is that this count goes up when ether: - The Rx queue is disabled (i.e. during a reset) - No free descriptors in the Rx queue

Re: [E1000-devel] ixgbe, hw filtering, five-tuple issue

Hi Jaroslaw, What driver are you using it wouldn't happen to be DPDK would it? If so I don't believe and of the developers of that driver monitor this list. I ask since the functions you mentioned below don't exist in the ixgbe (linux driver). Thanks, -Don -Original Message- From:

Re: [E1000-devel] what's the latest stable version of ixgbevf

The latest ixgbevf driver is 2.14.2. While you don't have to pair up ixgbe and ixgbevf, the new ixgbevf has a lot of performance and bug fixes in it. So if your having bandwidth issues I would try upgrading first. Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message-

Re: [E1000-devel] X520SR2BPL linux driver

The X520SR2BPL card is a bypass adapter based on X520. It would mostly work with the standard driver but you would not be able to use any of it's bypass functionality. You should get the driver from you ever you purchased the NIC from. They will have access to the driver from Intel. Thanks,

Re: [E1000-devel] What's the difference between mac_addr perm_addr?

-Original Message- From: opendaylight [mailto:opendayli...@126.com] Sent: Monday, May 26, 2014 8:26 PM To: e1000-devel@lists.sourceforge.net Subject: [E1000-devel] What's the difference between mac_addr perm_addr? HI list In the ixgbevf / vf.c struct

Re: [E1000-devel] Error when reset the ixgbevf NIC

What version of the driver are you looking at? The code below doesn't look like the latest net-next or out of tree driver. Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: nu li [mailto:ixgb...@gmail.com] Sent: Tuesday, May 27, 2014 3:55 AM To:

Re: [E1000-devel] ixgbe: how to balance PPPoE traffic via RSS to multiple queues

it can filter L2 fields. Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: Fujinaka, Todd Sent: Thursday, May 15, 2014 9:26 AM To: Jack Spinov; Skidmore, Donald C Cc: e1000-devel@lists.sourceforge.net Subject: RE: [E1000-devel] ixgbe: how to balance PPPoE

Re: [E1000-devel] ixgbe: how to balance PPPoE traffic via RSS to multiple queues

Skidmore donald.c.skidm...@intel.com -Original Message- From: Jack Spinov [mailto:spi...@timegroup.ae] Sent: Thursday, May 15, 2014 5:46 AM To: Skidmore, Donald C Cc: e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] ixgbe: how to balance PPPoE traffic via RSS to multiple

Re: [E1000-devel] ixgbe: how to balance PPPoE traffic via RSS to multiple queues

Maybe try VMDq which could sort by L2 (MAC address and VLAN). Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: Jack Spinov [mailto:spi...@timegroup.ae] Sent: Tuesday, May 13, 2014 7:39 AM To: e1000-devel@lists.sourceforge.net Subject: [E1000-devel] ixgbe:

Re: [E1000-devel] Unable to receive multicast packet on VF

-Original Message- From: Madoka Komatsubara [mailto:m-komatsub...@ab.jp.nec.com] Sent: Friday, March 28, 2014 4:49 AM To: linux-ker...@vger.kernel.org; e1000-devel@lists.sourceforge.net; net...@vger.kernel.org Cc: Hiroshi Shimamoto; Hiroshi Baba Subject: [E1000-devel] Unable to

Re: [E1000-devel] [PATCH net-next] ixgbevf: fix skb-pkt_type checks

-Original Message- From: David Miller [mailto:da...@redhat.com] Sent: Thursday, February 27, 2014 1:08 PM To: f.faine...@gmail.com Cc: e1000-devel@lists.sourceforge.net; net...@vger.kernel.org; Brandeburg, Jesse Subject: Re: [E1000-devel] [PATCH net-next] ixgbevf: fix skb-pkt_type

Re: [E1000-devel] Unable to use 10GB NIC cards

Hey Bill, The reason the your 10G adapters aren't getting recognized is because you have undefined SFP+ modules in the cage. :07:00.0: setting latency timer to 64 ixgbe :07:00.0: failed to load because an unsupported SFP+ module type was detected. ixgbe :07:00.0: Reload the driver

Re: [E1000-devel] questions on ixgbe and 10G performance expectations

-Original Message- From: John-Paul Robinson [mailto:j...@uab.edu] Sent: Tuesday, February 18, 2014 2:59 PM To: Brandeburg, Jesse Cc: e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] questions on ixgbe and 10G performance expectations On 02/17/2014 08:19 PM,

Re: [E1000-devel] Access to cards serial numbers

I don't understand what is meant by serial numbers, maybe device ID's. If so you could find them with lspci. -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: Fujinaka, Todd [mailto:todd.fujin...@intel.com] Sent: Wednesday, February 12, 2014 7:46 AM To: Nomad Esst;

Re: [E1000-devel] A fully polling mode ixgbevf driver

I not sure exactly what you are trying to show for this class, but if you want to the driver in polling mode you should try a busy poll socket which will pretty much give you what you're asking for. Thanks, -Don -Original Message- From: William Tu [mailto:u9012...@gmail.com] Sent:

Re: [E1000-devel] RSS queues number limit on 82599 controller

While the hardware is capable of supporting the larger number of queues, the RSS hash is only 4bit long (so only 16 queues) can be supported. ATR and perfect filters could support the larger queue sizes. Not sure that will work for you though. :( Thanks, -Don Skidmore

Re: [E1000-devel] ixgbe/82599 + Flow Director + Low Latency Polling patch

Hey David, I'm sure what you mean by the Low Latency Polling patch, but I'm guessing it refers to the busy poll sockets that we recently pushed upstream. Please correct me if I'm wrong. Given that I believe you could set up perfect filter rules (via. ethtool --config-ntuple) to direct

Re: [E1000-devel] ixgbe master disable timeout

Hey Martin, The error you are seeing here is the driver disabling PCIe Master during the reset path and the transaction pending bit not clearing while we waited for it. We corrected a similar problem a few releases ago changing from a static timeout value to more dynamicly configured. So yes

Re: [E1000-devel] ixgbe master disable timeout

:0e:00.0 Do you think I've not installed it properly? The box isn't a SPARC it's a Intel box a Dell PowerEdge 1950. Best Regards Martin On 2014-01-07 18:58, Skidmore, Donald C wrote: Hey Martin, The error you are seeing here is the driver disabling PCIe Master during the reset

Re: [E1000-devel] Possible bug in ixgbe-3.19.1 ixgbe_enumerate_functions()

We may be able to detect when we get number of physical functions that can't be correct (i.e. not 1 or 2) and at least log a different message rather than the current warning about insufficient bandwidth. Thanks, -Don -Original Message- From: Rustad, Mark D

Re: [E1000-devel] RSS Configuration in ixgbe (Queue limit?)

-Original Message- From: Scott Silverman [mailto:ssilver...@simplexinvestments.com] Sent: Tuesday, December 24, 2013 6:28 AM To: e1000-devel@lists.sourceforge.net Subject: [E1000-devel] RSS Configuration in ixgbe (Queue limit?) In attempting to troubleshoot the issue with ring

Re: [E1000-devel] RSS Configuration in ixgbe (Queue limit?)

From: Scott Silverman [mailto:ssilver...@simplexinvestments.com] Sent: Tuesday, December 24, 2013 10:03 AM To: Skidmore, Donald C Cc: e1000-devel@lists.sourceforge.net Subject: Re: [E1000-devel] RSS Configuration in ixgbe (Queue limit?) Don, I'm not sure I understand your answer. If the driver

Re: [E1000-devel] ETDEV WATCHDOG: eth0 (ixgbe): transmit queue 0 timed out

Those last fake Tx Hang messages from the driver show that the upper stack is timing out while the driver doesn't think it has anything to do. The stack dump previous to this is most likely the culprit. Since we don't see any ixgbe call's in the trace I wonder what is going on with the rest

Re: [E1000-devel] [PATCH] drivers/base/core.c: always output device renaming messages.

-Original Message- From: Greg KH [mailto:gre...@linuxfoundation.org] Sent: Friday, October 11, 2013 9:12 AM To: Bjorn Helgaas Cc: ethan.zhao; linux-ker...@vger.kernel.org; Skidmore, Donald C; e1000- de...@lists.sourceforge.net Subject: Re: [PATCH] drivers/base/core.c: always output

Re: [E1000-devel] Traffic forwarding rate of 10G interfaces is lower than that of 1G interfaces

Hey Abhinay, I would be interested in seeing your statics on the ports before/after a test (ethtool -s ethX). Likewise what your interrupt distribution looks like before/after (cat /proc/interrupts | grep ethX). lscpi -vvv could be useful as well. Thanks, -Don Skidmore

Re: [E1000-devel] 3.11-rc4 ixgbevf: endless Last Request of type 00 to PF Nacked messages

-Original Message- From: Bjorn Helgaas [mailto:bhelg...@google.com] Sent: Thursday, September 12, 2013 3:27 PM To: Skidmore, Donald C Cc: e1000-devel@lists.sourceforge.net; linux-...@vger.kernel.org; linux- ker...@vger.kernel.org; Don Dutile Subject: Re: [E1000-devel] 3.11-rc4

Re: [E1000-devel] Fwd: libvirt-qemu/kvm passthru w/Intel 82598EB

Hey Larry, It almost sounds like you are running out of memory, although I have never seen that libvirt error myself. Could you verify your VM's has enough memory? Thanks, -Don Skidmore donald.c.skidm...@intel.com -Original Message- From: laurence.schuler

  1   2   >