On Fri, Nov 18 2016, 12:07 PM, gre...@linuxfoundation.org wrote:
> On Fri, Nov 18, 2016 at 08:48:36AM +, Levy, Amir (Jer) wrote:
> > > BTW, it is quite a shame that the Thunderbolt firmware version can't
> > > be read from Linux.
> > >
> >
> > This is WIP, once this patch will be upstream, we w
From: Chris Lesiak Sent: Friday, November 18, 2016
10:37 PM
>To: Andy Duan
>Cc: netdev@vger.kernel.org; linux-ker...@vger.kernel.org; Jaccon
>Bastiaansen
>Subject: Re: [PATCH] net: fec: Detect and recover receive queue hangs
>
>On 11/18/2016 12:44 AM, Andy Duan wrote:
>> From: Chris Lesi
On Fri, Nov 18, 2016 at 06:44:05PM +0100, Pablo Neira Ayuso wrote:
> On Fri, Nov 18, 2016 at 09:17:18AM -0800, Alexei Starovoitov wrote:
> > On Fri, Nov 18, 2016 at 01:37:32PM +0100, Pablo Neira Ayuso wrote:
> > > On Thu, Nov 17, 2016 at 07:27:08PM +0100, Daniel Mack wrote:
> > > [...]
> > > > @@ -
From: John Fastabend
Date: Sat, 19 Nov 2016 19:14:15 -0800
> On 16-11-19 07:08 PM, David Miller wrote:
>> From: Stefan Hajnoczi
>> Date: Thu, 17 Nov 2016 15:49:23 +
>>
>>> @@ -159,6 +199,10 @@ virtio_transport_send_pkt(struct virtio_vsock_pkt *pkt)
>>> return -ENODEV;
>>> }
From: Jon Maloy
Date: Sat, 19 Nov 2016 14:47:07 -0500
> The comment block in socket.c describing the locking policy is
> obsolete, and does not reflect current reality. We remove it in this
> commit.
>
> Since the current locking policy is much simpler and follows a
> mainstream approach, we see
From: Zhang Shengju
Date: Sat, 19 Nov 2016 23:28:32 +0800
> If the link is filtered out, loop index should also be updated. If not,
> loop index will not be correct.
>
> Signed-off-by: Zhang Shengju
Applied to 'net' and queued up for -stable.
Bug fixes should always be targetted at 'net' not
On 16-11-19 07:08 PM, David Miller wrote:
> From: Stefan Hajnoczi
> Date: Thu, 17 Nov 2016 15:49:23 +
>
>> @@ -159,6 +199,10 @@ virtio_transport_send_pkt(struct virtio_vsock_pkt *pkt)
>> return -ENODEV;
>> }
>>
>> +if (le32_to_cpu(pkt->hdr.dst_cid) == vsock->guest_cid)
From: Alexey Dobriyan
Date: Sat, 19 Nov 2016 03:59:07 +0300
> Length of a netlink attribute may be u16 but lengths of basic attributes
> are much smaller, so small we can save 16 bytes of .rodata and pocket
> change inside .text.
>
> 16-bit is worse on x86-64 than 8-bit because of operand size o
From: Alexey Dobriyan
Date: Sat, 19 Nov 2016 03:54:35 +0300
> ->nla_len is unsigned entity (it's length after all) and u16,
> thus it can't overflow when being aligned into int/unsigned int.
>
> (nlmsg_next has the same code, but I didn't yet convince myself
> it is correct to do so).
>
> There
From: Alexey Dobriyan
Date: Sat, 19 Nov 2016 04:08:08 +0300
> 1) cast to "int" is unnecessary:
>u8 will be promoted to int before decrementing,
>small positive numbers fit into "int", so their values won't be changed
>during promotion.
>
>Once everything is int including loop cou
From: Alexey Dobriyan
Date: Sat, 19 Nov 2016 03:47:56 +0300
> size_t is way too much for an integer not exceeding 64.
>
> Space savings: 10 bytes!
>
> add/remove: 0/0 grow/shrink: 0/3 up/down: 0/-10 (-10)
> function old new delta
> nap
From: Guillaume Nault
Date: Fri, 18 Nov 2016 22:13:00 +0100
> Lock socket before checking the SOCK_ZAPPED flag in l2tp_ip6_bind().
> Without lock, a concurrent call could modify the socket flags between
> the sock_flag(sk, SOCK_ZAPPED) test and the lock_sock() call. This way,
> a socket could be
From: Stefan Hajnoczi
Date: Thu, 17 Nov 2016 15:49:23 +
> @@ -159,6 +199,10 @@ virtio_transport_send_pkt(struct virtio_vsock_pkt *pkt)
> return -ENODEV;
> }
>
> + if (le32_to_cpu(pkt->hdr.dst_cid) == vsock->guest_cid) {
> + return virtio_transport_send_pk
This adds support for the XDP_TX action to virtio_net. When an XDP
program is run and returns the XDP_TX action the virtio_net XDP
implementation will transmit the packet on a TX queue that aligns
with the current CPU that the XDP packet was processed on.
Before sending the packet the header is ze
XDP requires using isolated transmit queues to avoid interference
with normal networking stack (BQL, NETDEV_TX_BUSY, etc). This patch
adds a XDP queue per cpu when a XDP program is loaded and does not
expose the queues to the OS via the normal API call to
netif_set_real_num_tx_queues(). This way th
From: Shrijeet Mukherjee
This adds XDP support to virtio_net. Some requirements must be
met for XDP to be enabled depending on the mode. First it will
only be supported with LRO disabled so that data is not pushed
across multiple buffers. The MTU must be less than a page size
to avoid having to h
This adds a warning for drivers to use when encountering an invalid
buffer for XDP. For normal cases this should not happen but to catch
this in virtual/qemu setups that I may not have expected from the
emulation layer having a standard warning is useful.
Signed-off-by: John Fastabend
---
includ
This implements virtio_net for the mergeable buffers and big_packet
modes. I tested this with vhost_net running on qemu and did not see
any issues.
There are some restrictions for XDP to be enabled (see patch 3) for
more details.
1. LRO must be off
2. MTU must be less than PAGE_SIZE
3. queu
This adds support for dynamically setting the LRO feature flag. The
message to control guest features in the backend uses the
CTRL_GUEST_OFFLOADS msg type.
Signed-off-by: John Fastabend
---
drivers/net/virtio_net.c | 45 -
1 file changed, 44 insertio
From: Tushar Dave
Date: Sat, 19 Nov 2016 13:53:58 -0800
> 'struct i40e_dma_mem' defined with 'packed' directive causing kernel
> unaligned errors on sparc.
I really wish people would get out of the habit of just slapping
the __packed attribute onto structures seemingly by default.
On 11/19/16 10:28 AM, Zhang Shengju wrote:
> If the link is filtered out, loop index should also be updated. If not,
> loop index will not be correct.
>
> Signed-off-by: Zhang Shengju
> ---
> net/core/rtnetlink.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/core/
From: Eric Dumazet
barrier() is a big hammer compared to READ_ONCE(),
and requires comments explaining what is protected.
READ_ONCE() is more precise and compiler should generate
better overall code.
Signed-off-by: Eric Dumazet
---
drivers/net/ethernet/broadcom/bnx2.c | 17 +
'struct i40e_dma_mem' defined with 'packed' directive causing kernel
unaligned errors on sparc.
e.g.
i40e: Intel(R) Ethernet Connection XL710 Network Driver - version
1.6.16-k
i40e: Copyright (c) 2013 - 2014 Intel Corporation.
Kernel unaligned access at TPC[44894c] dma_4v_alloc_coherent+0x1ac/0x30
On 16-11-18 07:23 PM, Jakub Kicinski wrote:
> On Fri, 18 Nov 2016 19:20:58 -0800, Eric Dumazet wrote:
>> On Fri, 2016-11-18 at 18:57 -0800, Jakub Kicinski wrote:
>>> On Fri, 18 Nov 2016 18:43:55 -0800, John Fastabend wrote:
On 16-11-18 06:10 PM, Jakub Kicinski wrote:
>> [...]
The comment block in socket.c describing the locking policy is
obsolete, and does not reflect current reality. We remove it in this
commit.
Since the current locking policy is much simpler and follows a
mainstream approach, we see no need to add a new description.
Signed-off-by: Jon Maloy
---
n
On Fri, Nov 18, 2016 at 02:42:10PM -0800, Vishwanathapura, Niranjana wrote:
> +HFI-VNIC DRIVER
> +M: Dennis Dalessandro
> +M: Niranjana Vishwanathapura
> +L: linux-r...@vger.kernel.org
> +S: Supported
> +F: drivers/infiniband/sw/intel/vnic
This is either a net driver or a ULP, no idea
From: Simon Wunderlich
Date: Sat, 19 Nov 2016 12:06:12 +0100
> here are some bugfix patches which we would like to have integrated
> into net.
>
> Please pull or let me know of any problem!
Pulled.
From: Simon Wunderlich
Date: Sat, 19 Nov 2016 14:35:20 +0100
> this should be our last feature pull request for batman-adv in this round.
>
> Please pull or let me know of any problem!
Also pulled, thanks.
This patch reduce the size of message sent to kernel space. Before this
patch, for command: 'ip link show', we will sent 1056 bytes. With this
patch, we only need to send 40 bytes.
Signed-off-by: Zhang Shengju
---
lib/libnetlink.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
From: Jarno Rajahalme
Date: Fri, 18 Nov 2016 15:40:41 -0800
> Use the common virtio_net_hdr_to_skb() instead of open coding it.
> Other call sites were changed by commit fd2a0437dc, but this one was
> missed, maybe because it is split in two parts of the source code.
>
> Interim comparisons of '
From: Jarno Rajahalme
Date: Fri, 18 Nov 2016 15:40:40 -0800
> virtio_net_hdr_from_skb() clears the memory for the header, so there
> is no point for the callers to do the same.
>
> Signed-off-by: Jarno Rajahalme
Applied.
From: Jarno Rajahalme
Date: Fri, 18 Nov 2016 15:40:42 -0800
> Remove static function __packet_rcv_vnet(), which only called
> virtio_net_hdr_from_skb() and BUG()ged out if an error code was
> returned. Instead, call virtio_net_hdr_from_skb() from the former
> call sites of __packet_rcv_vnet() an
From: Jarno Rajahalme
Date: Fri, 18 Nov 2016 15:40:39 -0800
> Fix incorrent comment after the final #endif.
>
> Signed-off-by: Jarno Rajahalme
Applied.
From: Jarno Rajahalme
Date: Fri, 18 Nov 2016 15:40:38 -0800
> No point storing the return value of virtio_net_hdr_to_skb() or
> virtio_net_hdr_from_skb() to a variable when the value is used only
> once as a boolean in an immediately following if statement.
>
> Signed-off-by: Jarno Rajahalme
A
From: Alexey Khoroshilov
Date: Sat, 19 Nov 2016 01:40:10 +0300
> at91ether_start_xmit() does not check for dma mapping errors.
>
> Found by Linux Driver Verification project (linuxtesting.org).
>
> Signed-off-by: Alexey Khoroshilov
Applied, thanks.
If the link is filtered out, loop index should also be updated. If not,
loop index will not be correct.
Signed-off-by: Zhang Shengju
---
net/core/rtnetlink.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index db313ec..f4b9350 100
On Fri, Nov 18, 2016 at 02:20:27PM +, Joao Pinto wrote:
> For now we are interesting in improving the synopsys QoS driver under
> /nect/ethernet/synopsys. For now the driver structure consists of a single
> file
> called dwc_eth_qos.c, containing synopsys ethernet qos common ops and platform
>
From: Sven Eckelmann
batman-adv is requiring the type of wifi device in different contexts. Some
of them can take the rtnl semaphore and some of them already have the
semaphore taken. But even others don't allow that the semaphore will be
taken.
The data has to be retrieved when the hardif is ad
From: Marek Lindner
In a few situations batman-adv tries to determine whether a given interface
is a WiFi interface to enable specific WiFi optimizations. If the interface
batman-adv has been configured with is a virtual interface (e.g. VLAN) it
would not be properly detected as WiFi interface an
From: Sven Eckelmann
The batman-adv module has to be loaded to fulfill genl request by the
userspace. When it is not loaded then requests will fail. It is therefore
useful to get the module automatically loaded when such a request is made.
Signed-off-by: Sven Eckelmann
Signed-off-by: Simon Wund
From: Marek Lindner
Signed-off-by: Marek Lindner
[sven.eckelm...@open-mesh.com: re-add batadv_get_real_netdev to take rtnl
semaphore for batadv_get_real_netdevice]
Signed-off-by: Sven Eckelmann
Signed-off-by: Simon Wunderlich
---
net/batman-adv/bat_v_elp.c | 10 --
net/batman-ad
From: Sven Eckelmann
Things like VLANs don't have their link set when they are created. Thus
the wifi flags have to be evaluated later to fix their contents for the
link interface.
Signed-off-by: Sven Eckelmann
Signed-off-by: Simon Wunderlich
---
net/batman-adv/hard-interface.c | 5 +
1 f
From: Sven Eckelmann
The returned net_namespace of batadv_getlink_net may be used with functions
that potentially modify the struct. Thus it must return the pointer as
non-const like rtnl_link_ops::get_link_net does.
Signed-off-by: Sven Eckelmann
Signed-off-by: Simon Wunderlich
---
net/batman
(2016-11-08
19:02:36 +0100)
are available in the git repository at:
git://git.open-mesh.org/linux-merge.git tags/batadv-next-for-davem-20161119
for you to fetch changes up to 9b4aec647a92a2464337db10507348aecf0f0fd7:
batman-adv: fix rare race conditions on interface removal (2016-11-08
19
From: Linus Lüssing
In rare cases during shutdown the following general protection fault can
happen:
general protection fault: [#1] SMP
Modules linked in: batman_adv(O-) [...]
CPU: 3 PID: 1714 Comm: rmmod Tainted: G O4.6.0-rc6+ #1
[...]
Call Trace:
[] batadv_hardi
From: Marek Lindner
The ELP protocol requires cfg80211 to auto-detect the WiFi througput
to a given neighbor. Use batadv_is_cfg80211_netdev() to determine
whether or not an interface is eligible.
Signed-off-by: Marek Lindner
Signed-off-by: Sven Eckelmann
Signed-off-by: Simon Wunderlich
---
n
From: Sven Eckelmann
The commit 9799c50372b2 ("batman-adv: fix splat on disabling an interface")
fixed a warning but at the same time broke the rtnl function add_slave for
devices which were temporarily removed.
batadv_softif_slave_add requires soft_iface of and hard_iface to be NULL
before it i
14:47:02 +0200)
are available in the git repository at:
git://git.open-mesh.org/linux-merge.git tags/batadv-net-for-davem-20161119
for you to fetch changes up to e13258f38e927b61cdb5f4ad25309450d3b127d1:
batman-adv: Detect missing primaryif during tp_send as error (2016-11-04
12:27:39 +0100
From: Sven Eckelmann
The throughput meter detects different situations as problems for the
current test. It stops the test after these and reports it to userspace.
This also has to be done when the primary interface disappeared during the
test.
Fixes: 33a3bb4a3345 ("batman-adv: throughput meter
49 matches
Mail list logo
