ath isn't Ethernet. :)
___
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to freebsd-net-unsubscr...@freebsd.org
On Mon, Apr 8, 2013 at 1:14 PM, Duvvuru,Venkat Kumar
venkatkumar.duvv...@emulex.com wrote:
Hi,
In the transmit path, if enqueue mechanism is used instead of blocking on
the lock, the throughput is not good in some scenarios (especially single
queue, multiple connections).
For example:
if
On Fri, Apr 26, 2013 at 10:49 AM, Erich Weiler wei...@soe.ucsc.edu wrote:
The pf isn't a process, so you can't see it in top. pf has some helper
threads however, but packet processing isn't performed by any of them.
But the work pf does would show up in 'system' on top right? So if I see
On Wed, May 22, 2013 at 12:42 PM, Luigi Rizzo ri...@iet.unipi.it wrote:
all that work on paravirtualization (as presented at bsdcan)
http://info.iet.unipi.it/~luigi/netmap/talk-bsdcan-2013.html
just to discover that simply enabling DEVICE_POLLING (10-years old
technology) gives the same
On Thu, May 23, 2013 at 4:32 AM, Alex Liptsin al...@mellanox.com wrote:
Hello.
I have FreeBSD 9.1 installed.
There is mellanox adapter inside.
OFED support is already installed.
I try to add pkeys on ib0 port.
Usually in Linux I did:
echo 0x800c /sys/class/net/ib0/create_child
On Thu, Aug 1, 2013 at 2:07 PM, Joe Moog joem...@ebureau.com wrote:
We have an iXsystems 1U server (E5) with an Intel 4-port ethernet NIC
installed, model I350-T4 (manufactured May of 2013). We're trying to bind
the 4 ports on this NIC together into a single lagg port, connected LACP to
a
On Thu, Oct 3, 2013 at 11:42 AM, Tugrul Erdogan
h.tugrul.erdo...@gmail.com wrote:
Hi,
I am using FreeBSD alpha2 amd64 on my router and after packet loaded test
attack I am taking
ping: sendto: No buffer space available error when running ping from
router to victim directoion. The nic is
On Thu, Oct 3, 2013 at 2:51 AM, hiren panchasara
hiren.panchas...@gmail.com wrote:
What/where is the exact logic/code of how card determines what traffic goes
to what queue?
Is it based on DST-MAC always?
RSS is based off of a hash of the 4-tuple (src/dst ip, src/dst port).
Note that this is
You might want to take a look at the projects/sv branch, which
implement kernel core dumps over the network. We had to solve a
similar problem there (in lem, em, igb and ixgbe) and ended up
piggybacking on most of the DEVICE_POLLING code to do it. The work
ended up stalling over objections over
The oldest hardware supported by the ixgbe driver is the 82598, which
supports up to 16 RSS queues (see Table 3-48 in the 82598 datasheet).
I believe that the 82599 and X520 are more capable.
I have no idea what the situation is on igb, as there seems to be a
lot more hardware supported by that
On Mon, Aug 18, 2014 at 8:58 AM, Piotr Kubaj p.ku...@pp.com.pl wrote:
Hi. Please see
http://forums.freebsd.org/viewtopic.php?f=15t=45303#p264204 and
http://forums.freebsd.org/viewtopic.php?f=15t=45303#p264249 .
I know I can use web interface or ssh but WinBox is required. In short,
using
I'm unable to get sshuttle to redirect DNS traffic even on a machine
that doesn't have any other ipfw rules running, so I don't think that
it's a conflict with your divert rules causing the problem.
Unfortunately I don't have a solution to your problem. When I need to
use sshuttle I run it from a
systat -ifstat currently truncates byte counters down to 32-bit
integers. The following fixes the issue, but I'm not very happy with
it. u_long is what the rest of our code uses for network counters,
but that ends up meaning that our counters are 32-bits wide on 32-bit
platforms. I could make
On Wed, Sep 17, 2014 at 4:27 AM, Stefano Garzarella
stefanogarzare...@gmail.com wrote:
Much of the advantage of TSO comes from crossing the network stack only
once per (large) segment instead of once per 1500-byte frame.
GSO does the same both for segmentation (TCP) and fragmentation (UDP)
by
The patch below converts the ixl(4) and ixlv(4) drivers to use the new
ifcounter interface. I've hidden the interface behind some macros to
ensure that the driver continues to compile for FreeBSD 10 and
earlier. The result of the macros is that the ifcounter
implementation is somewhat clunkier
What netmasks are you using on the FreeBSD machine? I have seen very
strange behaviour (like packets not being able to be routed) in the
FreeBSD network stack if I assign two addresses on the same subnet to
different interfaces.
On Thu, Sep 25, 2014 at 2:53 PM, Tony Moseby tsmos...@yahoo.no
2012/2/2 Alexander V. Chernikov melif...@freebsd.org:
P.S. it is also reasonable to set net.isr.bindthreads to 1
I really don't recommend setting this in any release. There is
currently a bug with binding kernel threads that causes unrelated
threads to be unnecessarily bound to CPUs. In the
On Thu, Feb 2, 2012 at 2:09 PM, Alexander V. Chernikov
melif...@freebsd.org wrote:
As far as I understand, the only effect of setting bindthreads to 1 causes
intr_event_bind() to bind soft netisr to appropriate CPU. Can you point me
to ML discussion or some other info clarifying this bug?
On Thu, Feb 2, 2012 at 2:51 PM, Ryan Stone ryst...@gmail.com wrote:
of the last netisr through to be bound to a CPU.
*sigh*, that should read netisr thread of course.
___
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo
On Fri, Mar 9, 2012 at 4:30 AM, Gleb Smirnoff gleb...@freebsd.org wrote:
The following reply was made to PR kern/165863; it has been noted by GNATS.
From: Gleb Smirnoff gleb...@freebsd.org
To: Eric van Gyzen eric_van_gy...@dell.com,
Eric van Gyzen e...@vangyzen.net, ema...@freebsd.org
The following reply was made to PR kern/165863; it has been noted by GNATS.
From: Ryan Stone ryst...@gmail.com
To: Eric van Gyzen e...@vangyzen.net
Cc: Gleb Smirnoff gleb...@freebsd.org, Eric van Gyzen
eric_van_gy...@dell.com, ema...@freebsd.org,
bug-follo...@freebsd.org
Subject: Re
Currently, if you remove an IPv4 address from an interface, all ARP
cache entries on its subnet are invalidated. However, the same thing
is not done for NDP cache entries when an IPv6 address is removed*.
Is this correct behaviour? It seems weird to have IPv4 and IPv6
behave differently.
* In
Ok, I think that I have an approach that will work. This is heavily
based off of glebius' proposal. The big difference is that instead of
initializing the arptimer callout with the ll_entry's lock, I
initialize it with the if_afdata_lock. This eliminates the LOR
problem in arptimer. It also
On Fri, Mar 30, 2012 at 12:28 AM, Li, Qing qing...@bluecoat.com wrote:
* In a way this is a good thing as in6_lltable_prefix_free() is
guaranteed to crash your kernel in two different ways, and that's not
counting the race conditions that it's subject to.
Could you please elaborate
On Mon, Apr 9, 2012 at 5:46 PM, Li, Qing qing...@bluecoat.com wrote:
Could you please share with me a call path that you believe is
problematic, or a simple test case that uncovers the issue ?
I cannot because it is literally impossible for
in6_lltable_prefix_free() to ever be called.
On Mon, Apr 9, 2012 at 9:08 PM, Li, Qing qing...@bluecoat.com wrote:
I have not followed this thread closely from the beginning, so let me confirm.
Are you reporting multiple bugs: 1) ndp entries remain after the removal of
an address (according to the thread title), and 2) you believe there
2012/4/10 Gleb Smirnoff gleb...@freebsd.org:
Thanks, Ryan!
(snip)
Looks okay from my viewpoint. Have you stress tested it? My snap patch
definitely had problems, AFAIR.
If this patch fixes panics observed by kern/165863 and passes stress
testing, then it should be committed ASAP, and
On Thu, Apr 12, 2012 at 2:38 PM, Konstantin Belousov
kostik...@gmail.com wrote:
And, on receive, the atom board does receive interrupts, em0:rx 0 counter
in vmstat -i increases. Even more fun, the sysctl dev.em.0.debug
shows increasing hw rdh (as I understand, this is hardware 'last
received'
On Wed, Apr 25, 2012 at 3:59 AM, Li, Qing qing...@bluecoat.com wrote:
The patch is located at
http://people.freebsd.org/~qingli/nd6_prefix.diff
Please give it a try. I did only basic testing as of now and
will do more tomorrow.
--Qing
I tested this last night. Unfortunately this seems
On Mon, Jul 9, 2012 at 4:12 AM, Gleb Smirnoff gleb...@freebsd.org wrote:
This looks very much related to a known race in ARP code.
See this email and related thread:
http://lists.freebsd.org/pipermail/freebsd-net/2012-March/031865.html
Ryan didn't check in any patches since, and I failed to
On Sat, Jul 28, 2012 at 2:15 AM, Jason Hellenthal
jhellent...@dataix.net wrote:
You should be able to dtrace that out and uncover whats going on with a
little intellect.
The D language itself might take a little learning but it would be worth
your while to pick it up now instead of later.
I know that there have been a lot of discussions about fixing how
packets are handed off to ifnets due to the current methods being
extremely race-prone. Has there been any consensus on how the problem
is going to be solved?
In my particular case, I've seen an if_bridge interface whose if_snd
On Wed, Sep 26, 2012 at 9:55 AM, John Baldwin j...@freebsd.org wrote:
You only have to drop the RX lock around if_input() if you use the same lock
for both TX and RX (as if_transmit() / if_start() can be invoked while locks
in the network stack are held).
Last time I checked(FreeBSD 8.2), this
On Wed, Sep 26, 2012 at 3:38 PM, John Baldwin j...@freebsd.org wrote:
ix:rx - udp - udpinp - in_multi_mtx - ix:core - ix:rx
Hmm, I'm not sure where the 'in_multi_mtx - ix:core' bit comes from.
I think that is the broken part of this. The SIOCADDMULTI and SIOCDELMULTI
ioctls are invoked
On Mon, Oct 15, 2012 at 12:29 PM, Gleb Smirnoff gleb...@freebsd.org wrote:
To me this unlock/lock looks like a legacy from times, when the driver
had a single mutex for both TX and RX parts.
And removing this re-locking in foo_rxeof() was one of the aims for separate
TX/RX locking.
Really,
On Tue, Oct 16, 2012 at 8:47 AM, Gleb Smirnoff gleb...@freebsd.org wrote:
Can you please provide hints how can SIOCADDMULTI lead to obtaining RX
lock in the stock driver?
It doesn't. But it does acquire the core lock, and the core lock is
acquired before the RX lock (in ixgbe_init, for
On Sun, Oct 21, 2012 at 2:21 PM, Alexander V. Chernikov
melif...@freebsd.org wrote:
ix:rx - udp is also fairly obvious (here's one backtrace):
The same question, where udp - ix:rx can happen ?
It can't happen directly as far as I can tell. But to trigger a
deadlock, all that has to happen is
On Fri, Dec 21, 2012 at 7:13 AM, Tsaregorodtsev Denis tele...@extrim.itwrote:
Hello,
I maintenance ISP's DNS server which works under FreeBSD 7.3 and BIND
9.9.1-P4. The network adapter is Intel(R) PRO/1000 Network Connection,
driver - em. The server is connected to a Cisco 6500 switch.
I don't believe that this is fixed in later versions of the driver. The
problem is that when the interface loses link the transmit queue can fill
up. Once that happens the driver never gets any more calls from the network
stack to make it send packets. Pinging the interface fixes it because the
On Wed, Jan 23, 2013 at 2:13 AM, Neel Natu neeln...@gmail.com wrote:
Hi,
I am running into a problem in head with the e1000 link state
detection logic attached to a 82571EB serdes controller.
The symptom is that the link state keeps flapping between up and down.
After I enabled the debug
On Tue, Mar 12, 2013 at 5:31 PM, Boris Kochergin sp...@acm.poly.edu wrote:
Additionally, can someone clarify the meaning of total vs. max for
these values?
-Boris
It measures the amount of memory being used as that type of data
structure. This is the total of the number of allocated items
What's the benefit in having a both an interrupt thread and task that
performs the same function? It seems to me that having two threads that do
the same job is what is making this so complicated.
___
freebsd-net@freebsd.org mailing list
On Fri, Oct 8, 2010 at 9:15 PM, Robert Watson rwat...@freebsd.org wrote:
+ /*
+* get and fill a header mbuf, then chain data as an
extended
+* mbuf.
+*/
+ MGETHDR(m, M_DONTWAIT, MT_DATA);
The idea of calling into
It looks to me that you're getting a ton of input drops. That's
presumably the cause of your issue. You can get the em driver to
print debug information to the console by running:
# sysctl dev.em.3.stats=1
# sysctl.dev.em.3.debug=1
The output should be available in dmesg and /var/log/messages
This one bit me hard several months ago. Your analysis is correct.
It's a hardware bug. The solution is to track in the driver whether
the VME bit is set for the given queue, and if it isn't, ignore the VP
bit.
I meant to report this one to Jack but forgot, evidently.
Ryan Stone
will see that the VP bit is set and set the
M_VLANTAG on the mbuf, and put 0 in m_pkthdr.ether_vtag. This means
that the stack will treat the packet as if it was double vlan tagged,
with outer tag being 0.
Ryan Stone
___
freebsd-net@freebsd.org mailing list
On Fri, Nov 26, 2010 at 3:46 PM, Jack Vogel jfvo...@gmail.com wrote:
Ahh, OK, so how would this solution work:
In rxeof, in addition to requiring VP to be set, also check that
adapter-num_vlans is non-zero, in order to store the tag.
Jack
Yes, I believe that will work.
Ryan Stone
2010/11/26 Jack Vogel jfvo...@gmail.com:
Just for the record, I was not aware of a hardware bug, and
for right now no one is around :) I will ask around next week
to see if something was known that I missed.
I doubt that anybody at Intel will know about this one. To my
knowledge, it's not in
As near as I can tell, the SR version of the X520 is going to be the
same card, only with an SR SFP+ installed.
___
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to
If it has an SFP+, won't the it be LR or SR depending on the type of
SFP+ installed?
___
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to freebsd-net-unsubscr...@freebsd.org
A couple of things that I forgot:
- Note that DEVICE_POLLING is not included in GENERIC. You'll have to
add the following to your kernel config to get polling support:
options DEVICE_POLLING
- you can enable polling on a interface with ifconfig(provided the
driver supports polling)
ifconfig
I've put together a patch against HEAD that replaces the
device_polling algorithm(it should apply cleanly to stable/8 as well
-- nothing has changed with polling in some time). The patch can be
found here:
http://people.freebsd.org/~rstone/kern_poll.diff
The new algorithm makes use of the
Ah, you've anticipated me. This is just the kind of thing that I had
in mind. I have some comments:
- Why allocate the poll_if_t in ether_poll_register_? If you let the
driver allocate it you don't have to worry about failure. And the
driver can embed it in its rx_ring so it doesn't have to
On Wed, Apr 20, 2011 at 9:37 AM, Steven Hartland
kill...@multiplay.co.uk wrote:
Seems there's an issue with the intel ix driver which causes
it to bin connections when you manipulate ip aliases.
This causes issues on boot when jails start as it bins other
active sessions.
Is this a know
On Sat, Jun 11, 2011 at 2:55 PM, Luigi Rizzo ri...@iet.unipi.it wrote:
I am also looking at other options, eg. a mixed mode where
certain packets (e.g. depending on the MAC address) retain
the usual path, whereas others stay in the netmap ring.
Of course when you need to replicate packets
On Sun, Jul 17, 2011 at 7:59 AM, Vadim Goncharov vadim_nucli...@mail.ru wrote:
Ways to improve are to be found from this starting point. However,
are that +2 atomic ops per packet really so expensive? How many of
atomic ops are already on that path? Any measures?
On high-performance multiqueue
I notice that the interface not up in your test. If you ifconfig it
up does it detect that it has lost link? The driver isn't really
running anymore when you bring the link down -- interrupts are
disabled, for example, so it will never get a link down interrupt.
I ran headlong into this issue today when trying to test some network
stack changes. It's depressingly easy to crash HEAD by periodically
destroying vlan interfaces while you are sending traffic over those
interfaces -- I get a panic within minutes.
as to what could have changed that triggers this BIOS bug. I've
tried building with -DOLD_NFSv2 but that didn't resolve the problem.
Ryan Stone
___
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe
On Wed, Jul 27, 2011 at 12:45 PM, Ryan Stone ryst...@gmail.com wrote:
I'm going to try a binary search to narrow down what set of changes
introduced the problem,
On that note, is there an faster way to cleanly build the boot code
than a full buildworld? I can just make in sys/boot
On Wed, Sep 7, 2011 at 10:52 AM, Alexander V. Chernikov
melif...@yandex-team.ru wrote:
Particularly I'm trying to figure out how can I use all this variety of
filters to get MPLS traffic split to different RX queues.
I've got bad news for you on this. From Section 7.1.2.7 of the 82599
On Tue, Sep 13, 2011 at 2:36 PM, Arnaud Lacombe lacom...@gmail.com wrote:
It did not crash, yet. The only downside is that after 3h30 and ~4h,
igb(4) queues' handler started spinning infinitely, breaking network
connectivity.
I saw a similar issue on HEAD last week. The attached patch fix the
On Wed, Sep 14, 2011 at 7:40 AM, John Baldwin j...@freebsd.org wrote:
You should probably commit that.
I've been meaning to post it to -current to see if somebody could come
up with a better approach to solving the problem, but if you're happy
with it, I'll submit it to re@.
I wonder if it
On Wed, Sep 14, 2011 at 7:40 AM, John Baldwin j...@freebsd.org wrote:
You should probably commit that.
I've been meaning to post it to -current to see if somebody could come
up with a better approach to solving the problem, but if you're happy
with it, I'll submit it to re@.
I wonder if it
On Sat, Sep 24, 2011 at 4:56 PM, Juli Mallett jmall...@freebsd.org wrote:
It's actually not as trivial to grep for 10G NICs as
implied, since some drivers have silly bugs, like ixgbe which sets
if_baudrate to '10' (i.e. 1 000 000 000).
I'm pretty sure that the bug is FreeBSD's.
Currently when FreeBSD responds to a ICMP Echo Request, it takes the
original mbuf, rewrites a couple of fields (like the src/dst IP and
the ICMP type), and then sends that mbuf back. As things are
currently implemented, the Don't Fragment bit is kept in the ICMP
replay. This can cause problems
ether_demux currently assumes that all vlan-tagged packets that it
sees have had the vlan stripped out and the M_VLAN tag is set, so it
never checks the ether type for a vlan. However ng_ether_rcv_upper
currently does not guarantee that this is the case(and there may be
other code paths where
2011/10/18 rozhuk...@gmail.com:
In what cases vlan-tagged packet can be received by ng_ether_rcv_upper ?
If another node attaches to an ng_ether node's upper hook and sends a
vlan tagged packet to the hook.
___
freebsd-net@freebsd.org mailing list
2011/10/19 rozhuk...@gmail.com:
This may be a wrong configuration or QinQ: packet may have M_VLAN tag is
set and still vlan-tagged (ether_type = VLAN)
It is not QinQ in my case. The interface that the upper hook exports
is that you send it a valid ethernet frame and it passes that frame up
On Wed, Oct 19, 2011 at 3:30 PM, Juli Mallett jmall...@freebsd.org wrote:
Why should the requirements for the Netgraph path be any different to
the non-Netgraph path? If drivers must ensure that frames that reach
ether_demux have had their VLAN tags stripped, so should Netgraph
things that
On Sun, Oct 23, 2011 at 5:31 PM, David DeSimone f...@verio.net wrote:
Ethernet cards filter their traffic based on MAC address, not based on
IP address.
Use tcpdump -e to examine the destination MAC of the packets you are
receiving, in order to determine whether you should receive them.
On Mon, Oct 24, 2011 at 3:08 PM, Sergey Saley sergeysa...@gmail.com wrote:
Nothing has changed after increasing mbufs count increasing:
irq256: ix0:que 0 3013004 2431
irq257: ix0:que 1 970295 783
irq258: ix0:que 2 574782 463
On Mon, Oct 24, 2011 at 3:51 PM, Sergey Saley sergeysa...@gmail.com wrote:
MPD5, netgraph, pppoe.Types of traffic - any (customer traffic).
Bying this card I counted on a 3-4G traffic at 3-4K pppoe sessions.
It turned to 600-700Mbit/s, about 50K pps at 700-800 pppoe sessions.
PPPoE is your
On Sun, Oct 30, 2011 at 4:57 AM, Hooman Fazaeli hoomanfaza...@gmail.com wrote:
IMHO, this is not a driver issue and the real fix would be to change
IFQ_HANDOFF to call if_start when the queue is full.
I'm not sure that's the right approach. 99% of the time, calling
if_start when the queue is
Currently you have to disable DEVICE_POLLING in your kernel config in
order to use multiple netisr threads. Back in the spring I was
working on a patch to fix this limitation, but at the time I was
having trouble getting access to equipment to test it out on. I
should be able to get some systems
This was fixed on HEAD in r207337. I believe that it was MFC'ed some time ago.
http://svnweb.freebsd.org/base?view=revisionrevision=207337
___
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe,
...@pci0:2:0:0: class=0x02 card=0x125e8086 chip=0x105e8086 rev=0x00
hdr=0x00
e...@pci0:2:0:1: class=0x02 card=0x125e8086 chip=0x105e8086 rev=0x00
hdr=0x00
Please let me know if there's anything else that you need. I'm
willing to try patches or get more debug output.
Thanks,
Ryan Stone
The em driver version 6.2.9(that's the Intel version number) worked
just fine, but it stopped working on multiple machines as soon as I
upgraded, so I doubt that it's the hardware.
___
freebsd-net@freebsd.org mailing list
So it seems the problem is buggy hardware that incorrectly reports
that is has a copper PHY instead of an internal serdes. For whatever
reason this didn't matter to the old driver, but the new driver is
tripping over the problem. *sigh*
___
I just bought a laptop with a Atheros AR9XXX and as far as I can tell
there is no driver support yet in FreeBSD. Some googling has seemed
to indicate that there is a driver in OpenBSD(athn) that supports this
family of cards. I was wondering if there was any work underway to
port this to
Hm, I see that you're using the ticks variable. To the best of my
knowledge, changing the current time should not cause ticks to change.
A lot of things, like the callout wheel, would be pretty broken if
that were happening. My knowledge of this stuff is several major
versions out of date,
No, defining EM_WATCHDOG as 10 * hz should mean that the watchdog
expires after 10 seconds no matter what your kern.hz is. hz is set to
the number of ticks in a second.
___
freebsd-net@freebsd.org mailing list
You have ifconfig'ed the interfaces up, right?
___
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to freebsd-net-unsubscr...@freebsd.org
If you look at the datasheet you will see that the statistics do not
measure the same thing(why would Intel implement two hardware counters
that measure the same thing?). The Good Packets Received Count
measures the number of valid packets that passed one of the receive
filters. In other words,
The extra packets counted in tpr are packets received by the interface
that were not addressed to that interface. The fact that your machine
with a 82571 has not seen any such packets is just a coincidence.
Whatever switch that machine is connected to has happened not to have
flooded unicast
Currently there's interface for giving access to queues to userspace.
One problem that you're going to run into is that the registers for
all of the queues are mapped consecutively. There wouldn't be any way
of preventing the application from messing around with queues that
don't belong to it.
On Tue, Jul 20, 2010 at 5:55 PM, Ryan Stone ryst...@gmail.com wrote:
Currently there's interface for giving access to queues to userspace.
That should of course read Currently there's *no* interface...
___
freebsd-net@freebsd.org mailing list
http
It's not a big thing but it would be nice to replace the m_next and
m_nextpkt fields with queue.h macros.
___
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to
It seems to me that the mistake that you are making is that you are
setting the vlan portion (e.g. vlan 18 vlandev em0) twice. The second
time that the configuration is applied it fails. I believe that you
want instead to put the following in rc.conf:
ipv6_ifconfig_vlan18=xxx:xxx:xxx:460::
From a quick look at the code, whenever an interface is added to a
bridge, if that interface does not support a feature currently enabled
on the bridge then it has to disable that feature on all member
interfaces of that bridge. That would re-init the em interface, which
would cause a link flap.
I've just finished committing support for PCI Single Root I/O
Virtualization in the pci subsystem to head. This should be a no-op
for everyone right now, but there were some minor refactorings in the
pci code that could have a lingering bug. I did make sure to test
that it boots on a variety of
This is great! Thanks to both you and Intel. I'm planning on getting
SR-IOV support into head this week, which would allow you to create
ixlv instances (on the same hardware). Any chance that you'd have the
time to look into supporting SR-IOV for that driver too?
I think that you might be a bit confused about the behaviour. An ix
port will only be missing if
a) You have a non-Intel SFP+ installed
b) hw.ix.unsupported_sfp=1 is not set in loader.conf
ix ports that have no SFP+ installed do show up in ifconfig
On Wed, Jun 17, 2015 at 7:00 AM, Lakshmi Narasimhan Sundararajan
lakshm...@msystechnologies.com wrote:
[lakshmis@mau-bsd-10a ~/fortville/hol/sys/dev/ixl]$ diff -c5pt ixl_txrx.c
ixl_txrx.c.mod
*** ixl_txrx.c Fri Jun 12 06:56:51 2015
--- ixl_txrx.c.mod Fri Jun 12 06:56:33 2015
***
On Jul 28, 2015 11:07 AM, Sergey Kandaurov pluk...@freebsd.org wrote:
On 28 July 2015 at 20:55, Alexander V. Chernikov melif...@freebsd.org
wrote:
28.07.2015, 15:32, Arseny Nasokin eir...@gmail.com:
Separate ifconfig_IF_description will be simpler to use.
I was also insisting on this..
On Wed, Nov 11, 2015 at 11:18 AM, Steven Hartland
wrote:
> Comparing this to the Linux driver which does detect the link down I've
> discovered it actually polls the link status by default in its watchdog.
>
> Disabling this with "ethtool --set-priv-flags eth1 LinkPolling
At $WORK, we have observed that if you attempt to add a route with the
-iface flag, but you specify an incorrect interface name, the route command
will perform a DNS lookup of the interface name. It appears that if the
DNS lookup succeeds that it will treat the parameter to -iface as if it
were a
An easier way to block ICMP redirects would be to set the sysctl:
sysctl net.inet.icmp.drop_redirect=1
___
freebsd-net@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to
The use of mbuf clusters larger than a single page really doesn't work.
The problem is that over time physical memory becomes fragmented and
eventually 9K of contiguous memory can't be allocated anymore. This is why
many drivers now limit themselves to page-sized clusters.
On Mon, Jun 6, 2016 at
From a quick look at the vlan code, I can identify a few cases that might
cause that counter to increment:
1) Error from the underlying ixgbe device. Does "netstat -dI ix0" show
that the driver has been dropping packets?
2) Link down events on the underlying NIC. I believe that link flaps will
1 - 100 of 252 matches
Mail list logo