On Mon, 26 Oct 2015 21:06:33 +0100
Pablo Neira Ayuso wrote:
> Hi,
>
> On Mon, Oct 26, 2015 at 11:55:39AM -0700, Ani Sinha wrote:
> > netfilter: nf_conntrack: fix RCU race in nf_conntrack_find_get
>
> Please, no need to Cc everyone here. Please, submit your Netfilter
> patches to netfilter-de...
On Tue, Oct 27, 2015 at 10:30:19PM -0700, Andy Lutomirski wrote:
> From: Andy Lutomirski
>
> Once virtio starts using the DMA API, we won't be able to safely DMA
> from the stack. virtio-net does a couple of config DMA requests
> from small stack buffers -- switch to using dynamically-allocated
On Wed, Oct 28, 2015 at 11:13:39AM +0800, Jason Wang wrote:
>
>
> On 10/27/2015 04:38 PM, Michael S. Tsirkin wrote:
> > On Mon, Oct 26, 2015 at 10:52:47AM -0700, Ravi Kerur wrote:
> >> Ported earlier patch from Jason Wang (dated 12/26/2014).
> >>
> >> This patch tries to reduce the number of MSIX
On 15/10/28 (水) 13:58, Stephen Hemminger wrote:
On Mon, 26 Oct 2015 12:40:55 +0900
Toshiaki Makita wrote:
...
Thank you for taking a look at the patch set.
I'm not sure if I fully understand you, so please correct me if I
misread you.
The problem is that you require changing network device
On 10/28/2015 03:21 PM, Michael S. Tsirkin wrote:
> On Wed, Oct 28, 2015 at 11:13:39AM +0800, Jason Wang wrote:
>>
>> On 10/27/2015 04:38 PM, Michael S. Tsirkin wrote:
>>> On Mon, Oct 26, 2015 at 10:52:47AM -0700, Ravi Kerur wrote:
Ported earlier patch from Jason Wang (dated 12/26/2014).
>>>
>
> Well, then we should wait for another voice aimed at the complexity of
> the patchset before amending and resending me the patchset.
>
>
Well, I perceive that after Sutter has taken over the maintenance
responsibilitiy and answered accordingly that the outstanding
resentments as resolved.
Tho
On Wed, Oct 28, 2015 at 11:39:41AM +0900, Stephen Hemminger wrote:
> On Tue, 27 Oct 2015 14:21:03 +0100
> Phil Sutter wrote:
>
> > On Thu, Sep 10, 2015 at 09:35:08PM +0200, Matthias Tafelmeier wrote:
> > > This small patch extends the lib json_writer module for formerly
> > > deactivated function
On Tue, Oct 27, 2015 at 09:01:08AM -0700, Johannes Weiner wrote:
...
> > > But regardless of tcp window control, we need to account socket memory
> > > in the main memory accounting pool where pressure is shared (to the
> > > best of our abilities) between all accounted memory consumers.
> > >
> >
On Wed, Oct 28, 2015 at 09:07:47AM +0100, Matthias Tafelmeier wrote:
>
> >
> > Well, then we should wait for another voice aimed at the complexity of
> > the patchset before amending and resending me the patchset.
> >
> >
>
> Well, I perceive that after Sutter has taken over the maintenance
> res
On Wed, 2015-09-23 at 10:58 -0700, David Miller wrote:
> From: David Woodhouse
> Date: Wed, 23 Sep 2015 09:14:09 +0100
>
> > But sure, for now I'll drop this from the series and I can try to
> > convince you separately.
>
> Yes, let's discuss this independantly to the nice bug fixing
> that's ha
From: Ido Schimmel
In certain use cases it is not always desirable for the switch device to
flood traffic to CPU port. Instead, only certain packet types (e.g.
STP, LACP) should be trapped to it.
Signed-off-by: Ido Schimmel
Signed-off-by: Jiri Pirko
---
Documentation/networking/switchdev.txt
From: Ido Schimmel
Allow devices supporting this feature to control the flooding of unknown
unicast traffic, by making switchdev infrastructure propagate this setting
to the switch driver.
Signed-off-by: Ido Schimmel
Signed-off-by: Jiri Pirko
---
Documentation/networking/switchdev.txt | 3 +++
From: Or Gerlitz
Fixes those places where checkpatch complains that comparisons
should place the constant on the right side of the test.
Signed-off-by: Or Gerlitz
Signed-off-by: Jiri Pirko
Signed-off-by: Ido Schimmel
---
drivers/net/ethernet/mellanox/mlxsw/core.c | 4 ++--
drivers/net/et
From: Jiri Pirko
Fix copy-paste errors.
Signed-off-by: Jiri Pirko
Signed-off-by: Ido Schimmel
---
drivers/net/ethernet/mellanox/mlxsw/reg.h | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h
b/drivers/net/ethernet/mellanox/mlxs
From: Jiri Pirko
Signed-off-by: Jiri Pirko
Signed-off-by: Ido Schimmel
---
drivers/net/ethernet/mellanox/mlxsw/reg.h | 7 +++
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h
b/drivers/net/ethernet/mellanox/mlxsw/reg.h
index 2bf230c..
From: Jiri Pirko
It is a flag anyway, so move it to existing u8 flag and don't waste mem.
Fix the flags to be in single u8 on the way.
Signed-off-by: Jiri Pirko
Signed-off-by: Ido Schimmel
---
drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 4 ++--
drivers/net/ethernet/mellanox/mlxsw/spectru
From: Jiri Pirko
Signed-off-by: Jiri Pirko
---
drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c
b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c
index
From: Jiri Pirko
The value passed through switchdev attr set is not in jiffies, but in
clock_t, so fix the convert.
Reported-by: Sagi Rotem
Fixes: 56ade8fe3f ("mlxsw: spectrum: Add initial support for Spectrum ASIC")
Signed-off-by: Jiri Pirko
Signed-off-by: Ido Schimmel
---
drivers/net/ether
From: Ido Schimmel
Add or remove a bridged port from the flooding domain of unknown unicast
packets according to user configuration.
Signed-off-by: Ido Schimmel
Signed-off-by: Jiri Pirko
---
drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 1 +
drivers/net/ethernet/mellanox/mlxsw/spectr
From: Ido Schimmel
When enabling a range of VLANs on a bridged port we can configure
flooding for these VLANs by one register access instead of calling the
same register for each VLAN. This is accomplished by using the 'range'
field of the Switch Flooding Table Register (SFTR).
Signed-off-by: Id
Hi Arnd,
Since we want to make counting semaphores go away,
Why do we want to make counting semaphores go away? completely?
or just for binary use cases?
I have a use case in iser target code where a counting semaphore is the
best suited synchronizing mechanism.
I have a single thread handli
From: Jiri Pirko
The original description was for LAG, so fix it.
Signed-off-by: Jiri Pirko
Signed-off-by: Ido Schimmel
---
drivers/net/ethernet/mellanox/mlxsw/reg.h | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h
b/drivers/n
From: Jiri Pirko
This driver update mainly brings support for user to be able to setup
flooding on specified port, via bridge flag. Also, there is a fix in ageing
time conversion. The rest is just cosmetics.
Ido Schimmel (4):
switchdev: Add support for flood control
switchdev: Make flood to
From: Or Gerlitz
Fix a place where checkpatch complains that braces should be used
on all arms of this statement.
Signed-off-by: Or Gerlitz
Signed-off-by: Jiri Pirko
Signed-off-by: Ido Schimmel
---
drivers/net/ethernet/mellanox/mlxsw/pci.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion
Hi David,
Hannes Frederic Sowa writes:
> This series fixes wrong checksums on the wire for IPv4 and IPv6. Large
> send buffers and especially NFS lead to wrong checksums in both IPv4
> and IPv6.
>
> CHECKSUM_PARTIAL skbs should not receive the respective fragmentations
> functions, so we add WAR
On Wed, Oct 28, 2015 at 3:32 PM, David Miller wrote:
>
> This may look a bit scary this late in the release cycle, but as is typically
> the case it's predominantly small driver fixes all over the place.
Christ people. This is just sh*t.
The conflict I get is due to stupid new gcc header file cr
Hi all,
I was unable to figure out who's maintaining
drivers/net/ethernet/freescale/fec_main.c, so I hope someone can help out on
this list...
We have a board with a RMII phy connected to an i.MX6S. The hardware seems to
be ok, since I can receive and transmit ethernet frames without drops or
er
This patch allows perf record setting event's attr.inherit bit by
config terms like:
# perf record -e cycles/no-inherit/ ...
# perf record -e cycles/inherit/ ...
So user can control inherit bit for each event separately.
In following example, a.out fork()s in main then do some complex
CPU inte
Alexei provided a patchset to allow BPF programs output data to ring
buffer using helper bpf_perf_event_output() [1]. and have been merged
into net-next as commit a43eec304259a6c637f4014a6d4767159b6a3aa3 (bpf:
introduce bpf_perf_event_output() helper).
This patchset introduces perf side code to u
bpf_perf_event_output() outputs data through sample->raw_data. This
patch adds support to convert those data into CTF. A python script
then can be used to process output data from BPF programs.
Example:
BPF program #
struct bpf_map_def SEC("maps") map_channel = {
Add hexdamical u32 to base data type, which is useful for raw output
because raw data are u32 aligned.
Signed-off-by: Wang Nan
Cc: Alexei Starovoitov
Cc: Arnaldo Carvalho de Melo
Cc: Brendan Gregg
Cc: Jiri Olsa
---
tools/perf/util/data-convert-bt.c | 2 ++
1 file changed, 2 insertions(+)
di
Commit a43eec304259a6c637f4014a6d4767159b6a3aa3 (bpf: introduce
bpf_perf_event_output() helper) add a helper to enable BPF program
output data to perf ring buffer through a new type of perf event
PERF_COUNT_SW_BPF_OUTPUT. This patch enable perf to create perf
event of that type. Now perf user can u
Hi Linus,
On Wed, Oct 28, 2015, at 10:39, Linus Torvalds wrote:
> On Wed, Oct 28, 2015 at 3:32 PM, David Miller
> wrote:
> >
> > This may look a bit scary this late in the release cycle, but as is
> > typically
> > the case it's predominantly small driver fixes all over the place.
>
> Christ pe
On 2015/10/28 18:55, Wang Nan wrote:
Alexei provided a patchset to allow BPF programs output data to ring
buffer using helper bpf_perf_event_output() [1]. and have been merged
into net-next as commit a43eec304259a6c637f4014a6d4767159b6a3aa3 (bpf:
introduce bpf_perf_event_output() helper).
Thi
Hi Scott,
> I wonder why it was 64-bit specific in the first place.
I think it was part of a series where I added my 64bit assembly checksum
routines, and I didn't step back and think that the wrapper code would
be useful on 32 bit.
Anton
--
To unsubscribe from this list: send the line "unsubscr
On Wed, Oct 28, 2015 at 10:48:54AM +0100, David Jander wrote:
>
> Hi all,
>
> I was unable to figure out who's maintaining
> drivers/net/ethernet/freescale/fec_main.c, so I hope someone can help out on
> this list...
>
> We have a board with a RMII phy connected to an i.MX6S. The hardware seems
On Wed, 28 Oct 2015 11:14:14 +
Russell King - ARM Linux wrote:
> On Wed, Oct 28, 2015 at 10:48:54AM +0100, David Jander wrote:
> >
> > Hi all,
> >
> > I was unable to figure out who's maintaining
> > drivers/net/ethernet/freescale/fec_main.c, so I hope someone can help out
> > on this list.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA384
> I did not take over maintenance responsibility (whatever that means
> to you precisely). I merely reviewed the patches, focussing on the
> technical aspects of both implementation and patch management.
Ah, I meant the maintenance of iproute2 as a
From: Fabio Estevam
We are in a context where we can sleep, and the FEC PHY reset gpio
may be on an I2C expander. Use the cansleep() variant when
setting the GPIO value.
Based on a patch from Russell King for pci-mvebu.c.
Signed-off-by: Fabio Estevam
---
drivers/net/ethernet/freescale/fec_mai
Raw sockets with hdrincl enabled can insert ipv6 extension headers
right into the data stream. In case we need to fragment those packets,
we reparse the options header to find the place where we can insert
the fragment header. If the extension headers exceed the link's MTU we
actually cannot make p
Linus dislikes these changes. To not hold up the net-merge let's revert
it for now and fix the bug like Linus suggested.
This reverts commit ec3661b42257d9a06cf0d318175623ac7a660113, reversing
changes made to c80dbe04612986fd6104b4a1be21681b113b5ac9.
Cc: Linus Torvalds
Signed-off-by: Hannes Fred
On Wed, Oct 28, 2015 at 9:19 AM, David Jander wrote:
> Sorry, I somehow assumed it was obvious I'd report against latest mainline...
> I'm on 4.3-rc7.
Are you able to find out a previous kernel version that does not
exhibit this failure?
--
To unsubscribe from this list: send the line "unsubscri
[Linus and Dave added, Solaris and NetBSD folks dropped from Cc]
On Tue, Oct 27, 2015 at 05:13:56PM -0700, Eric Dumazet wrote:
> On Tue, 2015-10-27 at 23:17 +, Al Viro wrote:
>
> > * [Linux-specific aside] our __alloc_fd() can degrade quite badly
> > with some use patterns. The cacheline
On Wed, 28 Oct 2015 10:31:17 -0200
Fabio Estevam wrote:
> On Wed, Oct 28, 2015 at 9:19 AM, David Jander wrote:
>
> > Sorry, I somehow assumed it was obvious I'd report against latest
> > mainline... I'm on 4.3-rc7.
>
> Are you able to find out a previous kernel version that does not
> exhibit
From: Igal Liberman
The Storage Profiles contain parameters that are used
by the FMan for frame reception and transmission.
Signed-off-by: Igal Liberman
---
drivers/net/ethernet/freescale/fman/Makefile |2 +-
drivers/net/ethernet/freescale/fman/fman_sp.c | 167 +
From: Igal Liberman
Add the Data Path Acceleration Architecture Frame Manger Driver.
The FMan embeds a series of hardware blocks that implement a group
of Ethernet interfaces. This patch adds The FMan configuration,
initialization and runtime control routines.
The FMan driver supports several ha
From: Igal Liberman
Add Frame Manager Multi-User RAM support.
This internal FMan memory block is used by the
FMan hardware modules, the management being made
through the generic allocator.
The FMan Internal memory, for example, is used for
allocating transmit and receive FIFOs.
Signed-off-by: I
From: Igal Liberman
Add the Data Path Acceleration Architecture Frame Manger MAC support.
This patch adds The FMan MAC configuration, initialization and
runtime control routines.
This patch contains support for these types of MACs:
- dTSEC: Three speed Ethernet controller (10/100/1000 Mbp
From: Igal Liberman
This patch adds the Ethernet MAC driver supporting the three
different types of MACs: dTSEC, tGEC and mEMAC.
Signed-off-by: Igal Liberman
---
drivers/net/ethernet/freescale/fman/Makefile |3 +-
drivers/net/ethernet/freescale/fman/mac.c| 980
From: Igal Liberman
The Freescale Data Path Acceleration Architecture (DPAA) is a set
of hardware components on specific QorIQ multicore processors.
This architecture provides the infrastructure to support
simplified sharing of networking interfaces and accelerators
by multiple CPU cores and the
From: Igal Liberman
Add the Data Path Acceleration Architecture Frame Manger Port Driver.
The FMan driver uses a module called "Port" to represent the physical
TX and RX ports.
Each FMan version has different number of physical ports.
This patch adds The FMan Port configuration, initialization an
From: Linus Torvalds
Date: Wed, 28 Oct 2015 18:39:56 +0900
> Get rid of it. And I don't *ever* want to see that shit again.
No problem, I'll revert it all.
I asked Hannes to repost his patches to linux-kernel hoping someone
would review and say it stunk or not, give him some feedback, or
whatev
Hello.
On 10/28/2015 1:55 PM, Wang Nan wrote:
Commit a43eec304259a6c637f4014a6d4767159b6a3aa3 (bpf: introduce
bpf_perf_event_output() helper) add a helper to enable BPF program
You haven't run the patch thru scripts/checkpath.pl, I guess? It now
enforces the certain style of citing a comm
On Wed, Oct 28, 2015 at 10:55:02AM +, Wang Nan wrote:
SNIP
> diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
> index f820906..397fb4e 100644
> --- a/tools/perf/util/evsel.c
> +++ b/tools/perf/util/evsel.c
> @@ -653,6 +653,15 @@ static void apply_config_terms(struct perf_evsel *
On Wed, 2015-10-28 at 12:35 +, Al Viro wrote:
> [Linus and Dave added, Solaris and NetBSD folks dropped from Cc]
>
> On Tue, Oct 27, 2015 at 05:13:56PM -0700, Eric Dumazet wrote:
> > On Tue, 2015-10-27 at 23:17 +, Al Viro wrote:
> >
> > > * [Linux-specific aside] our __alloc_fd() can de
On Tue, Oct 27, 2015 at 12:15 PM, wrote:
> From: Dan Streetman
>
> The ipv4 and ipv6 xfrms each create a template dst_ops object, and
> perform dst_entries_init() on the template objects. Then each net
> namespace has its net.xfrm.xfrm[46]_dst_ops field set to the template
> values. The proble
Hello.
On 10/27/2015 10:02 PM, Punit Vara wrote:
This patch is to the 3945-mac.c file that fixes up following warning
by coccicheck:
drivers/net/wireless/iwlegacy/3945-mac.c:247:5-8: Unneeded variable:
"ret". Return "- EOPNOTSUPP" on line 249
Return -EOPNOTSUPP directly instead of return usin
On 10/28/2015 4:35 PM, Sergei Shtylyov wrote:
This patch is to the 3945-mac.c file that fixes up following warning
by coccicheck:
drivers/net/wireless/iwlegacy/3945-mac.c:247:5-8: Unneeded variable:
"ret". Return "- EOPNOTSUPP" on line 249
Return -EOPNOTSUPP directly instead of return using re
Em Wed, Oct 28, 2015 at 02:21:26PM +0100, Jiri Olsa escreveu:
> On Wed, Oct 28, 2015 at 10:55:02AM +, Wang Nan wrote:
> > @@ -55,6 +56,7 @@ struct perf_evsel_config_term {
> > booltime;
> > char*callgraph;
> > u64 stack_user;
> > + u64
Hello,
On Wed, Oct 28, 2015, at 14:32, Dan Streetman wrote:
> On Tue, Oct 27, 2015 at 12:15 PM, wrote:
> > From: Dan Streetman
> >
> > The ipv4 and ipv6 xfrms each create a template dst_ops object, and
> > perform dst_entries_init() on the template objects. Then each net
> > namespace has its
Em Wed, Oct 28, 2015 at 10:42:13AM -0300, Arnaldo Carvalho de Melo escreveu:
> Em Wed, Oct 28, 2015 at 02:21:26PM +0100, Jiri Olsa escreveu:
> > On Wed, Oct 28, 2015 at 10:55:02AM +, Wang Nan wrote:
> > > @@ -55,6 +56,7 @@ struct perf_evsel_config_term {
> > > booltime;
> > >
On 10/27/2015 09:50 PM, David Miller wrote:
From: Tom Lendacky
Date: Mon, 26 Oct 2015 17:13:54 -0500
During Tx cleanup it's still possible for the descriptor data to be
read ahead of the descriptor index. A memory barrier is required between
the read of the descriptor index and the start of th
From: Dan Streetman
Date: Wed, 28 Oct 2015 09:32:47 -0400
> Well I'm not sure why my test kernel booted, while the test robot
> found the bug of GFP_KERNEL percpu counter alloc during atomic
> context. Thanks test robot!
It's because of the kernel config options you (don't) have
enabled.
--
To
On 10/27/2015 05:59 PM, Vivien Didelot wrote:
> On Oct. Tuesday 27 (44) 04:43 PM, Neil Armstrong wrote:
>>
>> Yes, I didn't know how to handle this since it was part of a larger patch.
>>
>> I forgot to add this into the cover-letter but I wanted to send an RFC serie
>> with
>> your bcm remove pat
> > void phy_disconnect(struct phy_device *phydev)
> > {
> > if (phydev->irq > 0)
> > phy_stop_interrupts(phydev);
> >
> > phy_stop_machine(phydev);
> >
> > phydev->adjust_link = NULL;
> >
> > phy_detach(phydev)
Make usage of previously introduced mv88e6xxx common remove
function in all mv88e6xxx drivers.
Signed-off-by: Neil Armstrong
---
drivers/net/dsa/mv88e6123_61_65.c | 1 +
drivers/net/dsa/mv88e6131.c | 8
drivers/net/dsa/mv88e6171.c | 1 +
drivers/net/dsa/mv88e6352.c | 1
Implement a remove callback allowing the switch driver to cleanup
resources it used: interrupts and remapped register ranges.
Signed-off-by: Florian Fainelli
Signed-off-by: Neil Armstrong
---
drivers/net/dsa/bcm_sf2.c | 20
1 file changed, 20 insertions(+)
diff --git a/dri
With the previously introduced remove callback, add a
mv88e6xxx common remove function to cleanup all resources.
Signed-off-by: Neil Armstrong
---
drivers/net/dsa/mv88e6xxx.c | 18 ++
drivers/net/dsa/mv88e6xxx.h | 2 ++
2 files changed, 20 insertions(+)
diff --git a/drivers/net
Some switch drivers might request interrupts, remap register ranges,
allow such drivers to implement a "remove" callback doing just that.
Signed-off-by: Florian Fainelli
Signed-off-by: Neil Armstrong
---
include/net/dsa.h | 1 +
net/dsa/dsa.c | 4
2 files changed, 5 insertions(+)
diff
On Wed, Oct 28, 2015 at 10:42:13AM -0300, Arnaldo Carvalho de Melo wrote:
> Em Wed, Oct 28, 2015 at 02:21:26PM +0100, Jiri Olsa escreveu:
> > On Wed, Oct 28, 2015 at 10:55:02AM +, Wang Nan wrote:
> > > @@ -55,6 +56,7 @@ struct perf_evsel_config_term {
> > > booltime;
> > >
Introduce a new remove callback to allow DSA drivers to cleanup their
ressources.
Then add a remove implementation for bcm_sf2 and mv88e6xxx.
This patch was not tested due of a lack of hardware.
v2: add remove callback patch to the serie
Neil Armstrong (4):
net: dsa: allow switch drivers to cl
On Wed, Oct 28, 2015 at 03:12:57PM +0100, Neil Armstrong wrote:
> Some switch drivers might request interrupts, remap register ranges,
> allow such drivers to implement a "remove" callback doing just that.
>
> Signed-off-by: Florian Fainelli
> Signed-off-by: Neil Armstrong
> ---
> include/net/d
On Wed, Oct 28 2015, Hannes Frederic Sowa wrote:
> Hi Linus,
>
> On Wed, Oct 28, 2015, at 10:39, Linus Torvalds wrote:
>> Get rid of it. And I don't *ever* want to see that shit again.
>
> I don't want to give up on that this easily:
>
> In future I would like to see an interface like this. It is
On Wed, Oct 28, 2015 at 03:13:16PM +0100, Neil Armstrong wrote:
> With the previously introduced remove callback, add a
> mv88e6xxx common remove function to cleanup all resources.
>
> Signed-off-by: Neil Armstrong
> ---
> drivers/net/dsa/mv88e6xxx.c | 18 ++
> drivers/net/dsa/mv
On Wed, Oct 28, 2015 at 03:37:02PM +0100, Neil Armstrong wrote:
> Hi Andrew,
>
> On 10/28/2015 03:35 PM, Andrew Lunn wrote:
> > On Wed, Oct 28, 2015 at 03:13:16PM +0100, Neil Armstrong wrote:
> >> diff --git a/drivers/net/dsa/mv88e6xxx.c b/drivers/net/dsa/mv88e6xxx.c
> >> index b1b14f5..6287096 10
Hi Andrew,
On 10/28/2015 03:35 PM, Andrew Lunn wrote:
> On Wed, Oct 28, 2015 at 03:13:16PM +0100, Neil Armstrong wrote:
>> diff --git a/drivers/net/dsa/mv88e6xxx.c b/drivers/net/dsa/mv88e6xxx.c
>> index b1b14f5..6287096 100644
>> --- a/drivers/net/dsa/mv88e6xxx.c
>> +++ b/drivers/net/dsa/mv88e6xxx
On Wed, 2015-10-28 at 06:24 -0700, Eric Dumazet wrote:
> Before I take a deep look at your suggestion, are you sure plain use of
> include/linux/percpu-refcount.h infra is not possible for struct cred ?
BTW, I am not convinced we need to spend so much energy and per-cpu
memory for struct cred ref
This patch series adds support for
- allocating rx pages local to the interrupt
- setting affinity hint to influence IRQs to be allocated on the
same NUMA node as the one where the card resides.
Alexandra Kossovsky (1):
sfc: use __GFP_NOWARN when allocating RX pages from atomic context.
Bert K
From: Bert Kenward
With CONFIG_HAVE_MEMORYLESS_NODES cpu_to_node() may return
nodes without memory, which is not a good choice when later
using that to allocate memory. cpu_to_mem() instead provides
the most appropriate NUMA node to allocate from.
Signed-off-by: Shradha Shah
---
drivers/net/et
From: Daniel Pieczko
When the interrupt servicing a channel is on a NUMA node that is
not local to the device, performance is improved by allocating
rx pages on the node local to the interrupt (remote to the device)
The performance-optimal case, where interrupts and applications
are pinned to CP
From: Bert Kenward
Use cpumask_local_spread to provide interrupt affinity hints
for each queue. This will spread interrupts across NUMA local
CPUs first, extending to remote nodes if needed.
Signed-off-by: Shradha Shah
---
drivers/net/ethernet/sfc/efx.c | 35 +++
On Wed, Oct 28, 2015 at 12:57:28PM +0100, Matthias Tafelmeier wrote:
> >> Those resentments were related to the patchsets complexity and
> >> size.
> >
> > I didn't see any problem with that in the first place. It is indeed
> > a big change, achieving something like that without a big patch set
>
From: Alexandra Kossovsky
If we fail to allocate a page when in atomic context this is
handled by scheduling a fill in non-atomic context.
As such, a warning is not needed.
Signed-off-by: Shradha Shah
---
drivers/net/ethernet/sfc/rx.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(
On Wed, Oct 28, 2015 at 9:42 AM, Hannes Frederic Sowa
wrote:
> Hello,
>
> On Wed, Oct 28, 2015, at 14:32, Dan Streetman wrote:
>> On Tue, Oct 27, 2015 at 12:15 PM, wrote:
>> > From: Dan Streetman
>> >
>> > The ipv4 and ipv6 xfrms each create a template dst_ops object, and
>> > perform dst_entri
> Yeah, sorry for not having looked into this earlier. Also, I neither
> have nor claim any power of veto.
No big issue. Maybe Stephen can clarifiy things. I mean acknowledge no
further objections.
> Apart from that, I'm not against this
> patch series in general, just trying to help raise it's
On Wed, 2015-10-28 at 15:01 +, Shradha Shah wrote:
> From: Daniel Pieczko
>
> When the interrupt servicing a channel is on a NUMA node that is
> not local to the device, performance is improved by allocating
> rx pages on the node local to the interrupt (remote to the device)
>
> The perform
---
net/bridge/netfilter/ebtables.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c
index f46ca41..26922e9 100644
--- a/net/bridge/netfilter/ebtables.c
+++ b/net/bridge/netfilter/ebtables.c
@@ -989,7 +989,7 @
---
net/bridge/netfilter/ebtables.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c
index f46ca41..26922e9 100644
--- a/net/bridge/netfilter/ebtables.c
+++ b/net/bridge/netfilter/ebtables.c
@@ -989,7 +989,7 @
On 27/10/2015 23:17, Al Viro wrote:
Frankly, as far as I'm concerned, the bottom line is
* there are two variants of semantics in that area and there's not
much that could be done about that.
Yes, that seems to be the case.
* POSIX is vague enough for both variants to comply
Rainer Weikusat writes:
> Jason Baron writes:
[...]
>> 2)
>>
>> For the case of epoll() in edge triggered mode we need to ensure that
>> when we return -EAGAIN from unix_dgram_sendmsg() when unix_recvq_full()
>> is true, we need to add a unix_peer_wake_connect() call to guarantee a
>> wakeup. O
Testing of the new UDP bearer has revealed that reception of
NAME_DISTRIBUTOR, LINK_PROTOCOL/RESET and LINK_PROTOCOL/ACTIVATE
message buffers is not prepared for the case that those may be
non-linear.
We now linearize all such buffers before they are delivered up to the
generic reception layer.
I
This patch changes how the multipath hash is computed for locally
generated UDP or TCP flows: now the hash comprises also l4 information
(source and destination port).
This allows better utilization of the available paths when the existing
flows have the same source IP and the same destination IP:
On 10/28/2015 12:46 PM, Rainer Weikusat wrote:
> Rainer Weikusat writes:
>> Jason Baron writes:
>
> [...]
>
>>> 2)
>>>
>>> For the case of epoll() in edge triggered mode we need to ensure that
>>> when we return -EAGAIN from unix_dgram_sendmsg() when unix_recvq_full()
>>> is true, we need to ad
> -Original Message-
> From: Richard Weinberger [mailto:richard.weinber...@gmail.com]
> Sent: Tuesday, October 27, 2015 6:36 PM
> To: David Miller
> Cc: Haiyang Zhang ; o...@aepfle.de; Greg Kroah-
> Hartman ; netdev@vger.kernel.org;
> jasow...@redhat.com; driverdev-de...@linuxdriverproje
Hello.
Your subject is too long, it should have been placed in the changelog
partially. You you didn't sign off on the patch, so it can't applied.
MBR, Sergei
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More maj
Hello.
On 10/28/2015 06:02 PM, Shradha Shah wrote:
From: Bert Kenward
Use cpumask_local_spread to provide interrupt affinity hints
for each queue. This will spread interrupts across NUMA local
CPUs first, extending to remote nodes if needed.
Signed-off-by: Shradha Shah
---
drivers/net/eth
counterstmp is not cleared before it is used in get_counters(). it might be
leaked partially when it is sent to userland later on.
Signed-off-by: Loganaden Velvindron
---
net/bridge/netfilter/ebtables.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/bridge/netfilter
On Wed, Oct 28, 2015 at 09:10:20PM +0300, Sergei Shtylyov wrote:
> Hello.
>
>Your subject is too long, it should have been placed in the changelog
> partially. You you didn't sign off on the patch, so it can't applied.
>
> MBR, Sergei
Thank you. Please reject this patch. I re-sent a proper o
Please reject this patch. I sent a proper one with the sign-on later on.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Oct 28, 2015 at 11:20:03AM +0300, Vladimir Davydov wrote:
> Then you'd better not touch existing tcp limits at all, because they
> just work, and the logic behind them is very close to that of global tcp
> limits. I don't think one can simplify it somehow.
Uhm, no, there is a crapload of b
1 - 100 of 122 matches
Mail list logo