> We currently manage IRQs in pool_bm which is a bit field of MAX_MSIX bits.
> Thus, allocating more than MAX_MSIX interrupts can't be managed in pool_bm.
> Fixing this by capping number of requested MSIXs to MAX_MSIX.
Carol, Matan, could U2 elaborate where/when the bug hits us?
Also, missing Fi
On Fri, Aug 28, 2015 at 7:06 AM, Alexey Kardashevskiy wrote:
> 68230242cdb breaks SRIOV on POWER8 system. I am not really suggesting
> reverting the patch, rather asking for a fix.
thanks for the detailed report, we will look into that.
Just to be sure, when going back in time, what is the lates
On Fri, Aug 28, 2015 at 12:46 PM, Joe Perches wrote:
> On Fri, 2015-08-28 at 14:28 -0500, cls...@linux.vnet.ibm.com wrote:
>> From: Carol L Soto
>>
>> Seeing this message with mlx4_eni with IPv6: hw csum failure
>>
>> Changing IPv6 csum calculation to be based on OFED 2.4 code.
>> When calculate
Docker container creation linearly increased from around 1.6 sec to 7.5 sec
(at 1000 containers) and perf data showed 50% ovehead in snmp_fold_field.
reason: currently __snmp6_fill_stats64 calls snmp_fold_field that walks
through per cpu data of an item (iteratively for around 36 items).
idea: Th
Signed-off-by: Raghavendra K T
---
include/net/ip.h | 10 ++
net/ipv4/af_inet.c | 41 +++--
2 files changed, 37 insertions(+), 14 deletions(-)
diff --git a/include/net/ip.h b/include/net/ip.h
index d5fe9f2..93bf12e 100644
--- a/include/net/ip.h
+++ b
While creating 1000 containers, perf is showing lot of time spent in
snmp_fold_field on a large cpu system.
The current patch tries to improve by reordering the statistics gathering.
Please note that similar overhead was also reported while creating
veth pairs https://lkml.org/lkml/2013/3/19/556
From: Pravin B Shelar
Date: Sat, 29 Aug 2015 17:44:04 -0700
> After converting all vport to netdev implmentations there
> is no need for some of vport functionalify.
Looks good, series applied, thanks.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message
After converting all vport to netdev implmentations there
is no need for some of vport functionalify.
Pravin B Shelar (4):
openvswitch: Remove vport get_name()
openvswitch: Remove egress_tun_info.
openvswitch: Remove vport stats.
openvswitch: Remove vport-net
net/openvswitch/actions.c
Remove unused get_name() function pointer from vport ops.
Signed-off-by: Pravin B Shelar
---
net/openvswitch/vport.h |6 +-
1 files changed, 1 insertions(+), 5 deletions(-)
diff --git a/net/openvswitch/vport.h b/net/openvswitch/vport.h
index b88b3ee..2f35244 100644
--- a/net/openvswitch
tun info is passed using skb-dst pointer. Now we have
converted all vports to netdev based implementation so
Now we can remove redundant pointer to tun-info from OVS_CB.
Signed-off-by: Pravin B Shelar
---
net/openvswitch/actions.c |5 -
net/openvswitch/datapath.c |1 -
net/o
This structure is not used anymore.
Signed-off-by: Pravin B Shelar
---
net/openvswitch/datapath.h |2 --
net/openvswitch/flow.c |3 ++-
net/openvswitch/vport.h|4
3 files changed, 2 insertions(+), 7 deletions(-)
diff --git a/net/openvswitch/datapath.h b/net/openvswitch/
Since all vport types are now backed by netdev, we can directly
use netdev stats. Following patch removes redundant stat
from vport.
Signed-off-by: Pravin B Shelar
---
net/openvswitch/vport-internal_dev.c | 57 --
net/openvswitch/vport-netdev.c | 22 +++
net/openvsw
On 8/29/15 4:40 PM, David Miller wrote:
In the output path we go:
struct flowi4 fl4 = {
/* needed to match OIF rule */
.flowi4_oif = vrf_dev->ifindex,
.flowi4_iif = LOOPBACK_IFINDEX,
.flowi4_tos = RT_TOS(ip4h->tos),
In the output path we go:
struct flowi4 fl4 = {
/* needed to match OIF rule */
.flowi4_oif = vrf_dev->ifindex,
.flowi4_iif = LOOPBACK_IFINDEX,
.flowi4_tos = RT_TOS(ip4h->tos),
.flowi4_flags = FLOWI_FLAG_ANYSRC
Hello.
On 8/28/2015 9:50 PM, Tony Lindgren wrote:
The interrupt handler may not be available when smsc911x probes if the
interrupt handler is a GPIO controller for example. Let's fix that
by adding handling for -EPROBE_DEFER.
Cc: Steve Glendinning
Signed-off-by: Tony Lindgren
---
drivers
On Sat, Aug 29, 2015 at 1:46 PM, David Miller wrote:
> From: Peter Nørlund
> Date: Sat, 29 Aug 2015 22:31:15 +0200
>
>> On Sat, 29 Aug 2015 13:14:29 -0700 (PDT)
>> David Miller wrote:
>>
>>> From: p...@ordbogen.com
>>> Date: Fri, 28 Aug 2015 22:00:47 +0200
>>>
>>> > When the routing cache was re
On Sat, Aug 29, 2015 at 1:46 PM, David Miller wrote:
> From: Peter Nørlund
> Date: Sat, 29 Aug 2015 22:31:15 +0200
>
>> On Sat, 29 Aug 2015 13:14:29 -0700 (PDT)
>> David Miller wrote:
>>
>>> From: p...@ordbogen.com
>>> Date: Fri, 28 Aug 2015 22:00:47 +0200
>>>
>>> > When the routing cache was re
On 08/28/2015 11:18 PM, David Miller wrote:
...
I don't know about this.
You're adding a new user visible feature bit, but...
The user can set it and silently the kernel will accept it, but this
bit is ignored.
Regardless of it's internal value, we never publish it to the user.
This is just a
From: Peter Nørlund
Date: Sat, 29 Aug 2015 22:31:15 +0200
> On Sat, 29 Aug 2015 13:14:29 -0700 (PDT)
> David Miller wrote:
>
>> From: p...@ordbogen.com
>> Date: Fri, 28 Aug 2015 22:00:47 +0200
>>
>> > When the routing cache was removed in 3.6, the IPv4 multipath
>> > algorithm changed from mor
From: Jesse Gross
Date: Fri, 28 Aug 2015 16:54:40 -0700
> Geneve can benefit from GRO at the device level in a manner similar
> to other tunnels, especially as hardware offloads are still emerging.
>
> After this patch, aggregated frames are seen on the tunnel interface.
> Single stream throughp
From: Simon Horman
Date: Sat, 29 Aug 2015 09:02:21 +0900
> When an error occurs skipping IPv6 extension headers retain the already
> parsed IP protocol and IPv6 addresses in the flow. Also assume that the
> packet is not a fragment in the absence of information to the contrary;
> that is always u
On Sat, 29 Aug 2015 13:14:29 -0700 (PDT)
David Miller wrote:
> From: p...@ordbogen.com
> Date: Fri, 28 Aug 2015 22:00:47 +0200
>
> > When the routing cache was removed in 3.6, the IPv4 multipath
> > algorithm changed from more or less being destination-based into
> > being quasi-random per-packe
From: Johan Hedberg
Date: Fri, 28 Aug 2015 23:13:51 +0300
> One more bunch of Bluetooth patches for 4.3:
>
> - Crash fix for hci_bcm driver
> - Enhancements to hci_intel driver (e.g. baudrate configuration)
> - Fix for SCO link type after multiple connect attempts
> - Cleanups & minor fixes
From: p...@ordbogen.com
Date: Fri, 28 Aug 2015 22:00:47 +0200
> When the routing cache was removed in 3.6, the IPv4 multipath algorithm
> changed
> from more or less being destination-based into being quasi-random per-packet
> scheduling. This increases the risk of out-of-order packets and makes
From: Tony Lindgren
Date: Fri, 28 Aug 2015 11:50:15 -0700
> The interrupt handler may not be available when smsc911x probes if the
> interrupt handler is a GPIO controller for example. Let's fix that
> by adding handling for -EPROBE_DEFER.
>
> Cc: Steve Glendinning
> Signed-off-by: Tony Lindgre
From: Jiri Benc
Date: Fri, 28 Aug 2015 20:48:18 +0200
> With tunneling, it is currently possible to get an IPv6 header and interpret
> it as an IPv4 header, or to interpret an IPv6 address as an IPv4 address
> (and vice versa). This leads to things like sending packets to incorrect
> address, IPv
From: David Ahern
Date: Fri, 28 Aug 2015 08:42:09 -0700
> A few useful tracepoints developing VRF driver.
>
> Signed-off-by: David Ahern
Applied, thanks David.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majord
Hello.
On 8/27/2015 10:21 PM, Phil Sutter wrote:
This way users can attach noqueue just like any other qdisc using tc
without having to mess with tx_queue_len first.
Signed-off-by: Phil Sutter
---
include/net/sch_generic.h | 1 +
net/sched/sch_api.c | 1 +
net/sched/sch_generic.c
* "vti" and "vti6" tunnel modes added to ip-tunnel.8 manual page
* Added "hoplimit" terminology for IPv6
* Corrected usage line
* Minor language fix
---
man/man8/ip-tunnel.8 | 24
1 file changed, 12 insertions(+), 12 deletions(-)
diff --git a/man/man8/ip-tunnel.8 b/man/ma
> Cumulus Networks is invested in the VRF solution, and we will be here for
> the long haul. We want a feature complete, performant and stable solution
> for open networking. My preference is for a built-in solution rather than a
> bolted on one and I am trying to do that by engaging the community
On 08/29/2015 08:51 PM, Joe Perches wrote:
On Sat, 2015-08-29 at 07:32 -0700, Eric Dumazet wrote:
On Sat, 2015-08-29 at 14:37 +0530, Raghavendra K T wrote:
static inline void __snmp6_fill_stats64(u64 *stats, void __percpu *mib,
- int items, int bytes, siz
On Sat, 2015-08-29 at 07:32 -0700, Eric Dumazet wrote:
> On Sat, 2015-08-29 at 14:37 +0530, Raghavendra K T wrote:
>
> >
> > static inline void __snmp6_fill_stats64(u64 *stats, void __percpu *mib,
> > - int items, int bytes, size_t syncpoff)
> > +
On 8/28/15 10:14 PM, David Miller wrote:
From: Eric Dumazet
Date: Fri, 28 Aug 2015 18:31:07 -0700
On Fri, 2015-08-28 at 17:34 -0700, David Ahern wrote:
Currently the VRF driver registers an Rx handler for enslaved devices.
The handler switches the skb->dev to the VRF device and sends it back
On 8/28/15 10:14 PM, David Miller wrote:
From: David Ahern
Date: Fri, 28 Aug 2015 17:34:20 -0700
Currently the VRF driver registers an Rx handler for enslaved devices.
The handler switches the skb->dev to the VRF device and sends it back for
another pass. While this works fine a side effect is
On Sat, 2015-08-29 at 14:37 +0530, Raghavendra K T wrote:
>
> static inline void __snmp6_fill_stats64(u64 *stats, void __percpu *mib,
> - int items, int bytes, size_t syncpoff)
> + int items, int bytes, size_t syncpoff)
> {
Corcodel Marian :
> Deny functions member of array rtl_work to run when link is
> ok.After rtl8169_init_phy exist 10 sec timeout.On many cases link is ok and
> no need to destroy all work.
I won't figure what happens outside of the "many cases" statement.
rtl_reset_work is a hard reset. It is
Deny functions member of array rtl_work to run when link is
ok.After rtl8169_init_phy exist 10 sec timeout.On many cases link is ok and
no need to destroy all work.
Signed-off-by: Corcodel Marian
diff --git a/drivers/net/ethernet/realtek/r8169.c
b/drivers/net/ethernet/realtek/r8169.c
index
On Sat, Aug 29, 2015 at 12:43:03AM +0200, Thomas Graf wrote:
> On 08/28/15 at 03:34pm, Phil Sutter wrote:
> > Quite ugly, IMHO: rhashtable_insert_fast() may return -ENOMEM as
> > non-permanent error, if allocation in GFP_ATOMIC failed. In this case,
> > allocation in GFP_KERNEL is retried by rht_de
Signed-off-by: Raghavendra K T
---
include/net/ip.h | 10 ++
net/ipv4/af_inet.c | 41 +++--
2 files changed, 37 insertions(+), 14 deletions(-)
diff --git a/include/net/ip.h b/include/net/ip.h
index d5fe9f2..93bf12e 100644
--- a/include/net/ip.h
+++ b
Docker container creation linearly increased from around 1.6 sec to 7.5 sec
(at 1000 containers) and perf data showed 50% ovehead in snmp_fold_field.
reason: currently __snmp6_fill_stats64 calls snmp_fold_field that walks
through per cpu data of an item (iteratively for around 36 items).
idea: Th
While creating 1000 containers, perf is showing lot of time spent in
snmp_fold_field on a large cpu system.
The current patch tries to improve by reordering the statistics gathering.
Please note that similar overhead was also reported while creating
veth pairs https://lkml.org/lkml/2013/3/19/556
On Sat, 29 Aug 2015 00:39:29 +0200, Thomas Graf wrote:
> Can we fix VXLAN instead? It should be possible to receive and send both
> v4 and v6 frames and represent it with metadata on a single socket.
That's actually what I intend to do next. But I need the flag and this
patch for that. If IPv6 is
On Sat, 29 Aug 2015 00:24:11 +0200, Thomas Graf wrote:
> On 08/28/15 at 08:48pm, Jiri Benc wrote:
> > diff --git a/net/openvswitch/vport.c b/net/openvswitch/vport.c
> > index e2dc9dac59e6..40164037928e 100644
> > --- a/net/openvswitch/vport.c
> > +++ b/net/openvswitch/vport.c
> > @@ -587,6 +587,8 @
On 08/29/2015 10:41 AM, David Miller wrote:
From: Raghavendra K T
Date: Sat, 29 Aug 2015 08:27:15 +0530
resending the patch with memset. Please let me know if you want to
resend all the patches.
Do not post patches as replies to existing discussion threads.
Instead, make a new, fresh, patch
On 08/29/2015 08:56 AM, Eric Dumazet wrote:
On Sat, 2015-08-29 at 08:27 +0530, Raghavendra K T wrote:
/* Use put_unaligned() because stats may not be aligned for u64. */
put_unaligned(items, &stats[0]);
for (i = 1; i < items; i++)
- put_unaligned(snmp_f
45 matches
Mail list logo
