* Re: [PATCH] ipv4: arp: process only if ipv4 address configured
From: Florian Westphal @ 2014-02-12 22:01 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Florian Westphal, netdev
In-Reply-To: <1392229587.1752.11.camel@edumazet-glaptop2.roam.corp.google.com>
Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Wed, 2014-02-12 at 18:27 +0100, Florian Westphal wrote:
> > 8030f54499925d073a88c09f ([IPV4] devinet: Register inetdev earlier.)
> > changed arp behaviour (2.6.22 onwards).
> >
> > Before this, inetdev_init() was called only when the first address was
> > added to the interface, i.e. arp_process always dropped incoming arp
> > packets as __in_dev_get_rcu() returned NULL when no IP address was set
> > on the interface.
> >
> > With >= 2.6.22 we now process arp packets even if no address is assigned.
> > It can cause issues if the machine has several interfaces in the same
> > segment; requests receive answers from multiple macs.
>
> What about arp_filter value/meaning ?
Sure, arp_filter=1 avoids this.
If you mean "we don't care, its been like this for years and if you
don't want it then set arp_filter=1" -- fine with me.
Sorry if this wasn't clear -- its more about the change in behaviour
and if we should care.
^ permalink raw reply
* NETIF_F_HIGHDMA?
From: Sergei Shtylyov @ 2014-02-12 23:00 UTC (permalink / raw)
To: David Miller, netdev
Hello.
I fail to understand why this flag is needed at all. netdev-features.txt says:
* Transmit DMA from high memory
On platforms where this is relevant, NETIF_F_HIGHDMA signals that
ndo_start_xmit can handle skbs with frags in high memory.
TTBOMK, highmem is an area of the kernel *virtual* memory that maps
physical memory that can't be directly mapped unto the linear kernel space,
i.e. in case when there's more memory in the system than the kernel memory can
directly map. But what this has to do with DMA which deals with the physical
memory? High memory doesn't have to be beyond 4 GiB limit -- it can exist even
on 32-bit system (consider 4 GiB memory and 2/2 or 3/1 GiB user/kernel VM
split), contrary to the belief of the many Ethernet PCI drivers which only set
NETIF_F_HIGHDMA when the PCI dual address cycle is supported (i.e. the address
on PCI can be 64-bit)...
So, in my opinion, each and every driver should set this flag (except
maybe those arch-specific drivers where CONFIG_HIGHMEM is not supported but
even it this case the flag wouldn't hurt), and hence it's not needed at all.
What am I missing in my picture?
WBR, Sergei
^ permalink raw reply
* socket bind to local address ::ffff:127.0.0.1 question
From: Vincent Li @ 2014-02-12 20:41 UTC (permalink / raw)
To: netdev@vger.kernel.org
Hi,
we have a traffic path like <BIGIP monitor userland process on Linux>
<------> router<------->pool/real server
there is a weird issue that intermittently, a socket is bound to
loopback address '::ffff:127.0.0.1' which caused issue for us because
remote end host can't SYN+ACK the SYN source from 127.0.0.1, strace
the userland process shows:
6797 12:36:48.815296 connect(11, {sa_family=AF_INET6,
sin6_port=htons(8180), inet_pton(AF_INET6, "::ffff:172.16.3.165",
&sin6_addr), sin6_flowinfo=0, sin6_scope_id=875638834}, 28) = -1
EINPROGRESS (Operation now in progress)
6797 12:36:48.815464 gettimeofday({1391744208, 815499}, NULL) = 0
6797 12:36:48.815552 time(NULL) = 1391744208
6797 12:36:48.815657 stat64("/etc/localtime", {st_mode=S_IFREG|0644,
st_size=331, ...}) = 0
...................
6797 12:36:48.816948 getsockname(11, {sa_family=AF_INET6,
sin6_port=htons(45621), inet_pton(AF_INET6, "::ffff:127.0.0.1",
&sin6_addr), sin6_flowinfo=0, sin6_scope_id=0},
[18433145657494601756]) = 0
the working strace is like:
6797 12:30:03.855031 connect(11, {sa_family=AF_INET6,
sin6_port=htons(8180), inet_pton(AF_INET6, "::ffff:172.16.3.165",
&sin6_addr), sin6_flowinfo=0, sin6_scope_id=875638834}, 28) = -1
EINPROGRESS (Operation now in progress)
6797 12:30:03.855198 gettimeofday({1391743803, 855244}, NULL) = 0
6797 12:30:03.855284 time(NULL) = 1391743803
6797 12:30:03.855377 stat64("/etc/localtime", {st_mode=S_IFREG|0644,
st_size=331, ...}) = 0
........................................
6797 12:30:03.856757 getsockname(11, {sa_family=AF_INET6,
sin6_port=htons(59567), inet_pton(AF_INET6, "::ffff:172.16.17.161",
&sin6_addr), sin6_flowinfo=0, sin6_scope_id=0},
[18433145657494601756]) = 0
I looked kernel code through cscope
Cscope tag: LOOPBACK4_IPV6
# line filename / context / line
1 51 include/net/transp_v6.h <<GLOBAL>>
#define LOOPBACK4_IPV6 cpu_to_be32(0x7f000006)
2 640 net/dccp/ipv6.c <<dccp_v6_request_recv_sock>>
newinet->daddr = newinet->saddr = newinet->rcv_saddr =
LOOPBACK4_IPV6;
3 1023 net/dccp/ipv6.c <<dccp_v6_connect>>
inet->rcv_saddr = LOOPBACK4_IPV6;
4 495 net/ipv4/ip_sockglue.c <<do_ip_setsockopt>>
inet->daddr != LOOPBACK4_IPV6)) {
5 342 net/ipv6/af_inet6.c <<inet6_bind>>
v4addr = LOOPBACK4_IPV6;
6 193 net/ipv6/datagram.c <<ip6_datagram_connect>>
inet->rcv_saddr = LOOPBACK4_IPV6;
7 109 net/ipv6/ipv6_sockglue.c <<ipv6_update_options>>
inet_sk(sk)->daddr != LOOPBACK4_IPV6) {
8 287 net/ipv6/raw.c <<rawv6_bind>>
v4addr = LOOPBACK4_IPV6;
9 280 net/ipv6/tcp_ipv6.c <<tcp_v6_connect>>
inet->rcv_saddr = LOOPBACK4_IPV6;
10 1435 net/ipv6/tcp_ipv6.c <<tcp_v6_syn_recv_sock>>
newinet->daddr = newinet->saddr = newinet->rcv_saddr =
LOOPBACK4_IPV6;
is LOOPBACK4_IPV6 (0x7f000006) equivalent to ::ffff:127.0.0.1 ? I
tried to do ipv6 decimal or hex conversion, the result seems does not
match.
any clue?
Vincent
^ permalink raw reply
* Re: [PATCH] staging: r8188eu: Fix Smatch warnings
From: Dan Carpenter @ 2014-02-12 20:11 UTC (permalink / raw)
To: Larry Finger; +Cc: devel, gregkh, netdev
In-Reply-To: <1392226065-21778-1-git-send-email-Larry.Finger@lwfinger.net>
On Wed, Feb 12, 2014 at 11:27:45AM -0600, Larry Finger wrote:
> These warnings were reported by the 0-DAY kernel build testing backend.
It's actually me who goes through those ones and reports them. But it's
probably not the first time someone has called me a backend today. :P
regards,
dan carpenter
^ permalink raw reply
* Re: [RFC 2/2] xen-netback: disable multicast and use a random hw MAC address
From: Luis R. Rodriguez @ 2014-02-12 19:52 UTC (permalink / raw)
To: Bill Fink
Cc: Ian Campbell, netdev@vger.kernel.org, xen-devel, Paul Durrant,
Wei Liu, kvm, linux-kernel@vger.kernel.org
In-Reply-To: <20140212121739.ecb2f222.billfink@mindspring.com>
On Wed, Feb 12, 2014 at 9:17 AM, Bill Fink <billfink@mindspring.com> wrote:
> On Wed, 12 Feb 2014, Ian Campbell wrote:
>> IOW -- enabling/disabling multicast seems to me to be an odd proxy for
>> disabling SLAAC or DAD and AIUI your patch fixes the opposite case,
>> which is to avoid SLAAC and DAD on interfaces which don't do multicast
>> (which makes sense since those protocols involve multicast).
>
> Forgive me if this doesn't make sense in this context since
> I'm not a kernel developer, but I was just wondering if any of
> the sysctls:
>
> /proc/sys/net/ipv6/conf/<ifc>/disable_ipv6
> /proc/sys/net/ipv6/conf/<ifc>/accept_dad
> /proc/sys/net/ipv6/conf/<ifc>/accept_ra
> /proc/sys/net/ipv6/conf/<ifc>/autoconf
>
> would be apropos for the requirement being discussed.
These are run time configuration options, post initialization. What
we're considering is internal net_device capability fields, to even
avoid creating these in the first place.
Luis
^ permalink raw reply
* Re: RFC: bridge get fdb by bridge device
From: Vlad Yasevich @ 2014-02-12 19:02 UTC (permalink / raw)
To: Jamal Hadi Salim, netdev@vger.kernel.org
Cc: Stephen Hemminger, Scott Feldman, John Fastabend
In-Reply-To: <52FA9167.2040305@mojatatu.com>
On 02/11/2014 04:08 PM, Jamal Hadi Salim wrote:
> On 02/11/14 16:00, Vlad Yasevich wrote:
>> On 02/11/2014 03:15 PM, Jamal Hadi Salim wrote:
>
>>
>> Because it just a multi-function nic that isn't tagged with any
>> kine of bridge flag. As John said, this might be useful, but not
>> done yet.
>>
>
> Ok, fair enough. Someone should send a patch - John perhaps.
>
>>
>> Not directly. However, if you put a layered software device in between
>> (vlan, bond, macvlan), then you can add that device to another bridge.
>> In fact, people do that to get GVRP working with VMs.
>>
>
> Do you recall the reasoning behind it?
Before my time. It's there since before 2.6.12 :)
-vlad
>
>
>>> It certainly has some equivalent semantics (looks at dst MAC then
>>> picks the port). Possible to add Vlans as well?
>>
>> I suppose. You can do things like:
>> # ip link add link eth0 dev vlan100 protocol 8021Q id 100
>> # ip link add link vlan0 dev mac100 type macvlan
>>
>> Now, you have a macvlan (mac100) that will only receive vlan100 traffic.
>> Expressing this in terms of fdb would be a bit difficult since each
>> interface is separate and eth0 doesn't really know about the stack.
>> It would require quite a lot of code.
>>
>
> nice.
>
>>> Why dont we tag such a thing as a bridge then?
>>>
>>
>> Because they are not always a bridge. It could be just a nic capable of
>> mac filtering.
>>
>
> I think in one of the modes it is merely a filter.
> But you turn on this other feature it is a bridge.
>
>>
>> Didn't realize it has different connotation for vxlan. The you probably
>> don't want to include and support in the bridge fdb show command.
>
> Thats what i thought you said earlier ;->
>
> cheers,
> jamal
^ permalink raw reply
* Re: [PATCH net-next 04/10] ieee802154: support 100kbps QPSK/EU in at86rf230
From: Sergei Shtylyov @ 2014-02-12 20:01 UTC (permalink / raw)
To: Phoebe Buckheister, netdev; +Cc: linux-zigbee-devel, davem
In-Reply-To: <1392223034-19955-5-git-send-email-phoebe.buckheister@itwm.fraunhofer.de>
Hello.
On 02/12/2014 07:37 PM, Phoebe Buckheister wrote:
> The standard assigns channel 0 on page 2 to be 100kbps QPSK in the
> 868.3MHz band. Add support to the at86rf230 driver for this channel and
> page, at the moment predicated only for the RF212 chip.
> Per the datasheet, configurations for page 0, channels 0 to 10 and page
> 2, channels 0 to 10 differ only in the BPSK_QPSK bit. Support for
> channels 1 to 10 is untested.
> Signed-off-by: Phoebe Buckheister <phoebe.buckheister@itwm.fraunhofer.de>
> ---
> drivers/net/ieee802154/at86rf230.c | 17 ++++++++++++++---
> 1 file changed, 14 insertions(+), 3 deletions(-)
> diff --git a/drivers/net/ieee802154/at86rf230.c b/drivers/net/ieee802154/at86rf230.c
> index 99bc822..f6338c4 100644
> --- a/drivers/net/ieee802154/at86rf230.c
> +++ b/drivers/net/ieee802154/at86rf230.c
[...]
> @@ -595,6 +596,14 @@ at86rf212_set_channel(struct at86rf230_local *lp, int page, int channel)
> if (rc < 0)
> return rc;
>
> + if (page == 0)
> + rc = at86rf230_write_subreg(lp, SR_BPSK_QPSK, 0);
> + else
> + rc = at86rf230_write_subreg(lp, SR_BPSK_QPSK, 1);
> +
Don't think empty line is needed here.
> + if (rc < 0)
> + return rc;
> +
> return at86rf230_write_subreg(lp, SR_CHANNEL, channel);
> }
>
[...]
> @@ -1067,9 +1077,10 @@ static int at86rf230_probe(struct spi_device *spi)
>
> spi_set_drvdata(spi, lp);
>
> - if (is_rf212(lp))
> + if (is_rf212(lp)) {
> dev->phy->channels_supported[0] = 0x00007FF;
> - else
> + dev->phy->channels_supported[2] = 0x00007FF;
> + } else
> dev->phy->channels_supported[0] = 0x7FFF800;
Have to use {} in both arms of the *if* statement now -- see
Documentation/CodingStyle.
WBR, Sergei
^ permalink raw reply
* [PATCH net-next 13/13] sfc: Add/remove blank lines to taste
From: Shradha Shah @ 2014-02-12 19:00 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
Remove trailing blank lines in several files.
Use only one blank line between functions.
Add a blank line as a separator in a few places.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/ethtool.c | 1 -
drivers/net/ethernet/sfc/falcon.c | 4 ++--
drivers/net/ethernet/sfc/farch.c | 2 --
drivers/net/ethernet/sfc/net_driver.h | 1 -
drivers/net/ethernet/sfc/nic.c | 1 -
5 files changed, 2 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/sfc/ethtool.c b/drivers/net/ethernet/sfc/ethtool.c
index 7d5cb86..89fcaff 100644
--- a/drivers/net/ethernet/sfc/ethtool.c
+++ b/drivers/net/ethernet/sfc/ethtool.c
@@ -699,7 +699,6 @@ static void efx_ethtool_get_pauseparam(struct net_device *net_dev,
pause->autoneg = !!(efx->wanted_fc & EFX_FC_AUTO);
}
-
static void efx_ethtool_get_wol(struct net_device *net_dev,
struct ethtool_wolinfo *wol)
{
diff --git a/drivers/net/ethernet/sfc/falcon.c b/drivers/net/ethernet/sfc/falcon.c
index 18d6f76..72652f3 100644
--- a/drivers/net/ethernet/sfc/falcon.c
+++ b/drivers/net/ethernet/sfc/falcon.c
@@ -422,7 +422,6 @@ static inline void falcon_irq_ack_a1(struct efx_nic *efx)
efx_readd(efx, ®, FR_AA_WORK_AROUND_BROKEN_PCI_READS);
}
-
static irqreturn_t falcon_legacy_interrupt_a1(int irq, void *dev_id)
{
struct efx_nic *efx = dev_id;
@@ -467,6 +466,7 @@ static irqreturn_t falcon_legacy_interrupt_a1(int irq, void *dev_id)
efx_schedule_channel_irq(efx_get_channel(efx, 1));
return IRQ_HANDLED;
}
+
/**************************************************************************
*
* RSS
@@ -1358,6 +1358,7 @@ static void falcon_reconfigure_mac_wrapper(struct efx_nic *efx)
case 100: link_speed = 1; break;
default: link_speed = 0; break;
}
+
/* MAC_LINK_STATUS controls MAC backpressure but doesn't work
* as advertised. Disable to ensure packets are not
* indefinitely held and TX queue can be flushed at any point
@@ -2868,4 +2869,3 @@ const struct efx_nic_type falcon_b0_nic_type = {
.mcdi_max_ver = -1,
.max_rx_ip_filters = FR_BZ_RX_FILTER_TBL0_ROWS,
};
-
diff --git a/drivers/net/ethernet/sfc/farch.c b/drivers/net/ethernet/sfc/farch.c
index f72489a..aa1b169 100644
--- a/drivers/net/ethernet/sfc/farch.c
+++ b/drivers/net/ethernet/sfc/farch.c
@@ -311,7 +311,6 @@ static inline void efx_farch_push_tx_desc(struct efx_tx_queue *tx_queue,
*/
void efx_farch_tx_write(struct efx_tx_queue *tx_queue)
{
-
struct efx_tx_buffer *buffer;
efx_qword_t *txd;
unsigned write_ptr;
@@ -1609,7 +1608,6 @@ irqreturn_t efx_farch_msi_interrupt(int irq, void *dev_id)
return IRQ_HANDLED;
}
-
/* Setup RSS indirection table.
* This maps from the hash value of the packet to RXQ
*/
diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h
index af2b8c5..8a400a0 100644
--- a/drivers/net/ethernet/sfc/net_driver.h
+++ b/drivers/net/ethernet/sfc/net_driver.h
@@ -1323,7 +1323,6 @@ static inline struct efx_rx_buffer *efx_rx_buffer(struct efx_rx_queue *rx_queue,
return &rx_queue->buffer[index];
}
-
/**
* EFX_MAX_FRAME_LEN - calculate maximum frame length
*
diff --git a/drivers/net/ethernet/sfc/nic.c b/drivers/net/ethernet/sfc/nic.c
index 79226b1..32d969e 100644
--- a/drivers/net/ethernet/sfc/nic.c
+++ b/drivers/net/ethernet/sfc/nic.c
@@ -530,4 +530,3 @@ void efx_nic_fix_nodesc_drop_stat(struct efx_nic *efx, u64 *rx_nodesc_drops)
efx->rx_nodesc_drops_prev_state = !!(efx->net_dev->flags & IFF_UP);
*rx_nodesc_drops -= efx->rx_nodesc_drops_while_down;
}
-
^ permalink raw reply related
* [PATCH net-next 12/13] sfc: Fail self-test with -EBUSY, not -EIO, if the device is busy
From: Shradha Shah @ 2014-02-12 19:00 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/ethtool.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/sfc/ethtool.c b/drivers/net/ethernet/sfc/ethtool.c
index 3444dd6..7d5cb86 100644
--- a/drivers/net/ethernet/sfc/ethtool.c
+++ b/drivers/net/ethernet/sfc/ethtool.c
@@ -461,7 +461,7 @@ static void efx_ethtool_self_test(struct net_device *net_dev,
goto fail;
if (efx->state != STATE_READY) {
- rc = -EIO;
+ rc = -EBUSY;
goto out;
}
^ permalink raw reply related
* [PATCH net-next 11/13] sfc: Cosmetic changes to self-test from the out-of-tree driver
From: Shradha Shah @ 2014-02-12 19:00 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/ethtool.c | 18 +++++++++++++-----
1 files changed, 13 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/sfc/ethtool.c b/drivers/net/ethernet/sfc/ethtool.c
index 2294289..3444dd6 100644
--- a/drivers/net/ethernet/sfc/ethtool.c
+++ b/drivers/net/ethernet/sfc/ethtool.c
@@ -251,6 +251,9 @@ static void efx_fill_test(unsigned int test_index, u8 *strings, u64 *data,
* @test_index: Starting index of the test
* @strings: Ethtool strings, or %NULL
* @data: Ethtool test results, or %NULL
+ *
+ * Fill in a block of loopback self-test entries. Return new test
+ * index.
*/
static int efx_fill_loopback_test(struct efx_nic *efx,
struct efx_loopback_self_tests *lb_tests,
@@ -290,6 +293,12 @@ static int efx_fill_loopback_test(struct efx_nic *efx,
* @tests: Efx self-test results structure, or %NULL
* @strings: Ethtool strings, or %NULL
* @data: Ethtool test results, or %NULL
+ *
+ * Get self-test number of strings, strings, and/or test results.
+ * Return number of strings (== number of test results).
+ *
+ * The reason for merging these three functions is to make sure that
+ * they can never be inconsistent.
*/
static int efx_ethtool_fill_self_tests(struct efx_nic *efx,
struct efx_self_tests *tests,
@@ -444,7 +453,7 @@ static void efx_ethtool_self_test(struct net_device *net_dev,
{
struct efx_nic *efx = netdev_priv(net_dev);
struct efx_self_tests *efx_tests;
- int already_up;
+ bool already_up;
int rc = -ENOMEM;
efx_tests = kzalloc(sizeof(*efx_tests), GFP_KERNEL);
@@ -453,7 +462,7 @@ static void efx_ethtool_self_test(struct net_device *net_dev,
if (efx->state != STATE_READY) {
rc = -EIO;
- goto fail1;
+ goto out;
}
netif_info(efx, drv, efx->net_dev, "starting %sline testing\n",
@@ -466,7 +475,7 @@ static void efx_ethtool_self_test(struct net_device *net_dev,
if (rc) {
netif_err(efx, drv, efx->net_dev,
"failed opening device.\n");
- goto fail1;
+ goto out;
}
}
@@ -479,8 +488,7 @@ static void efx_ethtool_self_test(struct net_device *net_dev,
rc == 0 ? "passed" : "failed",
(test->flags & ETH_TEST_FL_OFFLINE) ? "off" : "on");
-fail1:
- /* Fill ethtool results structures */
+out:
efx_ethtool_fill_self_tests(efx, efx_tests, NULL, data);
kfree(efx_tests);
fail:
^ permalink raw reply related
* [PATCH net-next 10/13] sfc: Update product naming
From: Shradha Shah @ 2014-02-12 19:00 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
We don't use 'Solarstorm' or 'Solarflare Communications' in full
any more.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/efx.c | 2 +-
drivers/net/ethernet/sfc/efx.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
index 91acdc2..62d1a78 100644
--- a/drivers/net/ethernet/sfc/efx.c
+++ b/drivers/net/ethernet/sfc/efx.c
@@ -3271,6 +3271,6 @@ module_exit(efx_exit_module);
MODULE_AUTHOR("Solarflare Communications and "
"Michael Brown <mbrown@fensystems.co.uk>");
-MODULE_DESCRIPTION("Solarflare Communications network driver");
+MODULE_DESCRIPTION("Solarflare network driver");
MODULE_LICENSE("GPL");
MODULE_DEVICE_TABLE(pci, efx_pci_table);
diff --git a/drivers/net/ethernet/sfc/efx.h b/drivers/net/ethernet/sfc/efx.h
index dbd7b78..9903258 100644
--- a/drivers/net/ethernet/sfc/efx.h
+++ b/drivers/net/ethernet/sfc/efx.h
@@ -14,7 +14,7 @@
#include "net_driver.h"
#include "filter.h"
-/* Solarstorm controllers use BAR 0 for I/O space and BAR 2(&3) for memory */
+/* All controllers use BAR 0 for I/O space and BAR 2(&3) for memory */
#define EFX_MEM_BAR 2
/* TX */
^ permalink raw reply related
* [PATCH net-next 09/13] sfc: Use canonical pointer type for MAC address in efx_set_mac_address()
From: Shradha Shah @ 2014-02-12 18:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
Functions such as is_valid_ether_addr() expect u8 *, so use that
instead of char *.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/efx.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
index 84a1e11..91acdc2 100644
--- a/drivers/net/ethernet/sfc/efx.c
+++ b/drivers/net/ethernet/sfc/efx.c
@@ -2113,7 +2113,7 @@ static int efx_set_mac_address(struct net_device *net_dev, void *data)
{
struct efx_nic *efx = netdev_priv(net_dev);
struct sockaddr *addr = data;
- char *new_addr = addr->sa_data;
+ u8 *new_addr = addr->sa_data;
if (!is_valid_ether_addr(new_addr)) {
netif_err(efx, drv, efx->net_dev,
^ permalink raw reply related
* [PATCH net-next 05/13] sfc: Rename 'use_options' variable in tso_start() to clearer 'use_opt_desc'
From: Shradha Shah @ 2014-02-12 18:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/tx.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/sfc/tx.c b/drivers/net/ethernet/sfc/tx.c
index 3aa22cd..fa94753 100644
--- a/drivers/net/ethernet/sfc/tx.c
+++ b/drivers/net/ethernet/sfc/tx.c
@@ -1002,7 +1002,7 @@ static void efx_enqueue_unwind(struct efx_tx_queue *tx_queue)
static int tso_start(struct tso_state *st, struct efx_nic *efx,
const struct sk_buff *skb)
{
- bool use_options = efx_nic_rev(efx) >= EFX_REV_HUNT_A0;
+ bool use_opt_desc = efx_nic_rev(efx) >= EFX_REV_HUNT_A0;
struct device *dma_dev = &efx->pci_dev->dev;
unsigned int header_len, in_len;
dma_addr_t dma_addr;
@@ -1028,7 +1028,7 @@ static int tso_start(struct tso_state *st, struct efx_nic *efx,
st->out_len = skb->len - header_len;
- if (!use_options) {
+ if (!use_opt_desc) {
st->header_unmap_len = 0;
if (likely(in_len == 0)) {
^ permalink raw reply related
* [PATCH net-next 08/13] sfc: Preserve rx_frm_trunc counters when resizing DMA rings
From: Shradha Shah @ 2014-02-12 18:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
We allocate efx_channel structures with kzalloc() so we don't need to
zero-initialise individual fields in efx_probe_channel(). Further,
this function will be called again during DMA ring resizing and we
should not reset any statistics then.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/efx.c | 2 --
1 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
index 83d4643..84a1e11 100644
--- a/drivers/net/ethernet/sfc/efx.c
+++ b/drivers/net/ethernet/sfc/efx.c
@@ -503,8 +503,6 @@ static int efx_probe_channel(struct efx_channel *channel)
goto fail;
}
- channel->n_rx_frm_trunc = 0;
-
return 0;
fail:
^ permalink raw reply related
* [PATCH net-next 07/13] sfc: Correct comment about number of TX queues used on EF10
From: Shradha Shah @ 2014-02-12 18:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
EF10 implements option descriptors to switch TX checksum offload
on and off between packets. We could therefore use a single
hardware TX queue per kernel TX queue, although we don't yet.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/ef10.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/sfc/ef10.c b/drivers/net/ethernet/sfc/ef10.c
index 174a92f..3b39798 100644
--- a/drivers/net/ethernet/sfc/ef10.c
+++ b/drivers/net/ethernet/sfc/ef10.c
@@ -172,8 +172,8 @@ static int efx_ef10_probe(struct efx_nic *efx)
struct efx_ef10_nic_data *nic_data;
int i, rc;
- /* We can have one VI for each 8K region. However we need
- * multiple TX queues per channel.
+ /* We can have one VI for each 8K region. However, until we
+ * use TX option descriptors we need two TX queues per channel.
*/
efx->max_channels =
min_t(unsigned int,
^ permalink raw reply related
* [PATCH net-next 06/13] sfc: Remove unused definitions of EF10 user-mode DMA descriptors
From: Shradha Shah @ 2014-02-12 18:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
These DMA descriptor types will only be used by the userland
networking stack.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/ef10_regs.h | 61 ----------------------------------
1 files changed, 0 insertions(+), 61 deletions(-)
diff --git a/drivers/net/ethernet/sfc/ef10_regs.h b/drivers/net/ethernet/sfc/ef10_regs.h
index 207ac9a..62a55dd 100644
--- a/drivers/net/ethernet/sfc/ef10_regs.h
+++ b/drivers/net/ethernet/sfc/ef10_regs.h
@@ -227,36 +227,6 @@
#define ESF_DZ_RX_KER_BUF_ADDR_LBN 0
#define ESF_DZ_RX_KER_BUF_ADDR_WIDTH 48
-/* RX_USER_DESC */
-#define ESF_DZ_RX_USR_RESERVED_LBN 62
-#define ESF_DZ_RX_USR_RESERVED_WIDTH 2
-#define ESF_DZ_RX_USR_BYTE_CNT_LBN 48
-#define ESF_DZ_RX_USR_BYTE_CNT_WIDTH 14
-#define ESF_DZ_RX_USR_BUF_PAGE_SIZE_LBN 44
-#define ESF_DZ_RX_USR_BUF_PAGE_SIZE_WIDTH 4
-#define ESE_DZ_USR_BUF_PAGE_SZ_4MB 10
-#define ESE_DZ_USR_BUF_PAGE_SZ_1MB 8
-#define ESE_DZ_USR_BUF_PAGE_SZ_64KB 4
-#define ESE_DZ_USR_BUF_PAGE_SZ_4KB 0
-#define ESF_DZ_RX_USR_BUF_ID_OFFSET_LBN 0
-#define ESF_DZ_RX_USR_BUF_ID_OFFSET_WIDTH 44
-#define ESF_DZ_RX_USR_4KBPS_BUF_ID_LBN 12
-#define ESF_DZ_RX_USR_4KBPS_BUF_ID_WIDTH 32
-#define ESF_DZ_RX_USR_64KBPS_BUF_ID_LBN 16
-#define ESF_DZ_RX_USR_64KBPS_BUF_ID_WIDTH 28
-#define ESF_DZ_RX_USR_1MBPS_BUF_ID_LBN 20
-#define ESF_DZ_RX_USR_1MBPS_BUF_ID_WIDTH 24
-#define ESF_DZ_RX_USR_4MBPS_BUF_ID_LBN 22
-#define ESF_DZ_RX_USR_4MBPS_BUF_ID_WIDTH 22
-#define ESF_DZ_RX_USR_4MBPS_BYTE_OFFSET_LBN 0
-#define ESF_DZ_RX_USR_4MBPS_BYTE_OFFSET_WIDTH 22
-#define ESF_DZ_RX_USR_1MBPS_BYTE_OFFSET_LBN 0
-#define ESF_DZ_RX_USR_1MBPS_BYTE_OFFSET_WIDTH 20
-#define ESF_DZ_RX_USR_64KBPS_BYTE_OFFSET_LBN 0
-#define ESF_DZ_RX_USR_64KBPS_BYTE_OFFSET_WIDTH 16
-#define ESF_DZ_RX_USR_4KBPS_BYTE_OFFSET_LBN 0
-#define ESF_DZ_RX_USR_4KBPS_BYTE_OFFSET_WIDTH 12
-
/* TX_CSUM_TSTAMP_DESC */
#define ESF_DZ_TX_DESC_IS_OPT_LBN 63
#define ESF_DZ_TX_DESC_IS_OPT_WIDTH 1
@@ -338,37 +308,6 @@
#define ESF_DZ_TX_TSO_TCP_SEQNO_LBN 0
#define ESF_DZ_TX_TSO_TCP_SEQNO_WIDTH 32
-/* TX_USER_DESC */
-#define ESF_DZ_TX_USR_TYPE_LBN 63
-#define ESF_DZ_TX_USR_TYPE_WIDTH 1
-#define ESF_DZ_TX_USR_CONT_LBN 62
-#define ESF_DZ_TX_USR_CONT_WIDTH 1
-#define ESF_DZ_TX_USR_BYTE_CNT_LBN 48
-#define ESF_DZ_TX_USR_BYTE_CNT_WIDTH 14
-#define ESF_DZ_TX_USR_BUF_PAGE_SIZE_LBN 44
-#define ESF_DZ_TX_USR_BUF_PAGE_SIZE_WIDTH 4
-#define ESE_DZ_USR_BUF_PAGE_SZ_4MB 10
-#define ESE_DZ_USR_BUF_PAGE_SZ_1MB 8
-#define ESE_DZ_USR_BUF_PAGE_SZ_64KB 4
-#define ESE_DZ_USR_BUF_PAGE_SZ_4KB 0
-#define ESF_DZ_TX_USR_BUF_ID_OFFSET_LBN 0
-#define ESF_DZ_TX_USR_BUF_ID_OFFSET_WIDTH 44
-#define ESF_DZ_TX_USR_4KBPS_BUF_ID_LBN 12
-#define ESF_DZ_TX_USR_4KBPS_BUF_ID_WIDTH 32
-#define ESF_DZ_TX_USR_64KBPS_BUF_ID_LBN 16
-#define ESF_DZ_TX_USR_64KBPS_BUF_ID_WIDTH 28
-#define ESF_DZ_TX_USR_1MBPS_BUF_ID_LBN 20
-#define ESF_DZ_TX_USR_1MBPS_BUF_ID_WIDTH 24
-#define ESF_DZ_TX_USR_4MBPS_BUF_ID_LBN 22
-#define ESF_DZ_TX_USR_4MBPS_BUF_ID_WIDTH 22
-#define ESF_DZ_TX_USR_4MBPS_BYTE_OFFSET_LBN 0
-#define ESF_DZ_TX_USR_4MBPS_BYTE_OFFSET_WIDTH 22
-#define ESF_DZ_TX_USR_1MBPS_BYTE_OFFSET_LBN 0
-#define ESF_DZ_TX_USR_1MBPS_BYTE_OFFSET_WIDTH 20
-#define ESF_DZ_TX_USR_64KBPS_BYTE_OFFSET_LBN 0
-#define ESF_DZ_TX_USR_64KBPS_BYTE_OFFSET_WIDTH 16
-#define ESF_DZ_TX_USR_4KBPS_BYTE_OFFSET_LBN 0
-#define ESF_DZ_TX_USR_4KBPS_BYTE_OFFSET_WIDTH 12
/*************************************************************************/
/* TX_DESC_UPD_REG: Transmit descriptor update register.
^ permalink raw reply related
* [PATCH net-next 04/13] sfc: Replace TSOH_OFFSET with the equivalent NET_IP_ALIGN
From: Shradha Shah @ 2014-02-12 18:58 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
If CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS is defined then NET_IP_ALIGN
will be defined as 0, so this macro is redundant.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/tx.c | 17 ++++-------------
1 files changed, 4 insertions(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/sfc/tx.c b/drivers/net/ethernet/sfc/tx.c
index 75d11fa..3aa22cd 100644
--- a/drivers/net/ethernet/sfc/tx.c
+++ b/drivers/net/ethernet/sfc/tx.c
@@ -787,15 +787,6 @@ void efx_remove_tx_queue(struct efx_tx_queue *tx_queue)
* Requires TX checksum offload support.
*/
-/* Number of bytes inserted at the start of a TSO header buffer,
- * similar to NET_IP_ALIGN.
- */
-#ifdef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS
-#define TSOH_OFFSET 0
-#else
-#define TSOH_OFFSET NET_IP_ALIGN
-#endif
-
#define PTR_DIFF(p1, p2) ((u8 *)(p1) - (u8 *)(p2))
/**
@@ -882,13 +873,13 @@ static u8 *efx_tsoh_get_buffer(struct efx_tx_queue *tx_queue,
EFX_BUG_ON_PARANOID(buffer->flags);
EFX_BUG_ON_PARANOID(buffer->unmap_len);
- if (likely(len <= TSOH_STD_SIZE - TSOH_OFFSET)) {
+ if (likely(len <= TSOH_STD_SIZE - NET_IP_ALIGN)) {
unsigned index =
(tx_queue->insert_count & tx_queue->ptr_mask) / 2;
struct efx_buffer *page_buf =
&tx_queue->tsoh_page[index / TSOH_PER_PAGE];
unsigned offset =
- TSOH_STD_SIZE * (index % TSOH_PER_PAGE) + TSOH_OFFSET;
+ TSOH_STD_SIZE * (index % TSOH_PER_PAGE) + NET_IP_ALIGN;
if (unlikely(!page_buf->addr) &&
efx_nic_alloc_buffer(tx_queue->efx, page_buf, PAGE_SIZE,
@@ -901,10 +892,10 @@ static u8 *efx_tsoh_get_buffer(struct efx_tx_queue *tx_queue,
} else {
tx_queue->tso_long_headers++;
- buffer->heap_buf = kmalloc(TSOH_OFFSET + len, GFP_ATOMIC);
+ buffer->heap_buf = kmalloc(NET_IP_ALIGN + len, GFP_ATOMIC);
if (unlikely(!buffer->heap_buf))
return NULL;
- result = (u8 *)buffer->heap_buf + TSOH_OFFSET;
+ result = (u8 *)buffer->heap_buf + NET_IP_ALIGN;
buffer->flags = EFX_TX_BUF_CONT | EFX_TX_BUF_HEAP;
}
^ permalink raw reply related
* [PATCH net-next 03/13] sfc: Rewrite adjustment of PPS event in a clearer way
From: Shradha Shah @ 2014-02-12 18:58 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
There is substantial latency in generation and handling of PPS events
from the NIC, which we have to correct for before passing a host
timestamp to the PPS subsystem. We compare clocks with the MC,
giving us two offsets to subtract from the timestamp generated by
pps_get_ts():
(a) Time from the last good sync (where we got host and NIC timestamps
for nearly the same instant) to the time we called pps_get_ts()
(b) Time from NIC top of second to the last good sync
We currently calculate (a) + (b) in a quite confusing way.
Instead, calculate (a) completely, then add (b) to it.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/ptp.c | 47 +++++++++++++++++++--------------------
1 files changed, 23 insertions(+), 24 deletions(-)
diff --git a/drivers/net/ethernet/sfc/ptp.c b/drivers/net/ethernet/sfc/ptp.c
index 7d0de50..28275e3 100644
--- a/drivers/net/ethernet/sfc/ptp.c
+++ b/drivers/net/ethernet/sfc/ptp.c
@@ -766,37 +766,36 @@ efx_ptp_process_times(struct efx_nic *efx, MCDI_DECLARE_STRUCT_PTR(synch_buf),
return -EAGAIN;
}
- /* Convert the NIC time into kernel time. No correction is required-
- * this time is the output of a firmware process.
- */
- mc_time = ptp->nic_to_kernel_time(ptp->timeset[last_good].major,
- ptp->timeset[last_good].minor, 0);
-
- /* Calculate delay from actual PPS to last_time */
- delta = ktime_to_timespec(mc_time);
- delta.tv_nsec +=
- last_time->ts_real.tv_nsec -
- (ptp->timeset[last_good].host_start & MC_NANOSECOND_MASK);
-
- /* It is possible that the seconds rolled over between taking
+ /* Calculate delay from last good sync (host time) to last_time.
+ * It is possible that the seconds rolled over between taking
* the start reading and the last value written by the host. The
* timescales are such that a gap of more than one second is never
- * expected.
+ * expected. delta is *not* normalised.
*/
start_sec = ptp->timeset[last_good].host_start >> MC_NANOSECOND_BITS;
last_sec = last_time->ts_real.tv_sec & MC_SECOND_MASK;
- if (start_sec != last_sec) {
- if (((start_sec + 1) & MC_SECOND_MASK) != last_sec) {
- netif_warn(efx, hw, efx->net_dev,
- "PTP bad synchronisation seconds\n");
- return -EAGAIN;
- } else {
- delta.tv_sec = 1;
- }
- } else {
- delta.tv_sec = 0;
+ if (start_sec != last_sec &&
+ ((start_sec + 1) & MC_SECOND_MASK) != last_sec) {
+ netif_warn(efx, hw, efx->net_dev,
+ "PTP bad synchronisation seconds\n");
+ return -EAGAIN;
}
+ delta.tv_sec = (last_sec - start_sec) & 1;
+ delta.tv_nsec =
+ last_time->ts_real.tv_nsec -
+ (ptp->timeset[last_good].host_start & MC_NANOSECOND_MASK);
+
+ /* Convert the NIC time at last good sync into kernel time.
+ * No correction is required - this time is the output of a
+ * firmware process.
+ */
+ mc_time = ptp->nic_to_kernel_time(ptp->timeset[last_good].major,
+ ptp->timeset[last_good].minor, 0);
+
+ /* Calculate delay from NIC top of second to last_time */
+ delta.tv_nsec += ktime_to_timespec(mc_time).tv_nsec;
+ /* Set PPS timestamp to match NIC top of second */
ptp->host_time_pps = *last_time;
pps_sub_ts(&ptp->host_time_pps, delta);
^ permalink raw reply related
* [PATCH net-next 02/13] sfc: Cache skb->data in local variable in efx_ptp_rx()
From: Shradha Shah @ 2014-02-12 18:58 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/ptp.c | 23 +++++++++++++----------
1 files changed, 13 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/sfc/ptp.c b/drivers/net/ethernet/sfc/ptp.c
index 52be63d..7d0de50 100644
--- a/drivers/net/ethernet/sfc/ptp.c
+++ b/drivers/net/ethernet/sfc/ptp.c
@@ -1366,6 +1366,7 @@ static bool efx_ptp_rx(struct efx_channel *channel, struct sk_buff *skb)
struct efx_ptp_match *match = (struct efx_ptp_match *)skb->cb;
u8 *match_data_012, *match_data_345;
unsigned int version;
+ u8 *data;
match->expiry = jiffies + msecs_to_jiffies(PKT_EVENT_LIFETIME_MS);
@@ -1374,7 +1375,8 @@ static bool efx_ptp_rx(struct efx_channel *channel, struct sk_buff *skb)
if (!pskb_may_pull(skb, PTP_V1_MIN_LENGTH)) {
return false;
}
- version = ntohs(*(__be16 *)&skb->data[PTP_V1_VERSION_OFFSET]);
+ data = skb->data;
+ version = ntohs(*(__be16 *)&data[PTP_V1_VERSION_OFFSET]);
if (version != PTP_VERSION_V1) {
return false;
}
@@ -1382,13 +1384,14 @@ static bool efx_ptp_rx(struct efx_channel *channel, struct sk_buff *skb)
/* PTP V1 uses all six bytes of the UUID to match the packet
* to the timestamp
*/
- match_data_012 = skb->data + PTP_V1_UUID_OFFSET;
- match_data_345 = skb->data + PTP_V1_UUID_OFFSET + 3;
+ match_data_012 = data + PTP_V1_UUID_OFFSET;
+ match_data_345 = data + PTP_V1_UUID_OFFSET + 3;
} else {
if (!pskb_may_pull(skb, PTP_V2_MIN_LENGTH)) {
return false;
}
- version = skb->data[PTP_V2_VERSION_OFFSET];
+ data = skb->data;
+ version = data[PTP_V2_VERSION_OFFSET];
if ((version & PTP_VERSION_V2_MASK) != PTP_VERSION_V2) {
return false;
}
@@ -1400,17 +1403,17 @@ static bool efx_ptp_rx(struct efx_channel *channel, struct sk_buff *skb)
* enhanced mode fixes this issue and uses bytes 0-2
* and byte 5-7 of the UUID.
*/
- match_data_345 = skb->data + PTP_V2_UUID_OFFSET + 5;
+ match_data_345 = data + PTP_V2_UUID_OFFSET + 5;
if (ptp->mode == MC_CMD_PTP_MODE_V2) {
- match_data_012 = skb->data + PTP_V2_UUID_OFFSET + 2;
+ match_data_012 = data + PTP_V2_UUID_OFFSET + 2;
} else {
- match_data_012 = skb->data + PTP_V2_UUID_OFFSET + 0;
+ match_data_012 = data + PTP_V2_UUID_OFFSET + 0;
BUG_ON(ptp->mode != MC_CMD_PTP_MODE_V2_ENHANCED);
}
}
/* Does this packet require timestamping? */
- if (ntohs(*(__be16 *)&skb->data[PTP_DPORT_OFFSET]) == PTP_EVENT_PORT) {
+ if (ntohs(*(__be16 *)&data[PTP_DPORT_OFFSET]) == PTP_EVENT_PORT) {
match->state = PTP_PACKET_STATE_UNMATCHED;
/* We expect the sequence number to be in the same position in
@@ -1426,8 +1429,8 @@ static bool efx_ptp_rx(struct efx_channel *channel, struct sk_buff *skb)
(match_data_345[0] << 24));
match->words[1] = (match_data_345[1] |
(match_data_345[2] << 8) |
- (skb->data[PTP_V1_SEQUENCE_OFFSET +
- PTP_V1_SEQUENCE_LENGTH - 1] <<
+ (data[PTP_V1_SEQUENCE_OFFSET +
+ PTP_V1_SEQUENCE_LENGTH - 1] <<
16));
} else {
match->state = PTP_PACKET_STATE_MATCH_UNWANTED;
^ permalink raw reply related
* [PATCH net-next 01/13] sfc: Removed adhoc scheme to rate limit PTP event queue overflow message
From: Shradha Shah @ 2014-02-12 18:58 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <52FBC3C2.7030103@solarflare.com>
From: Laurence Evans <levans@solarflare.com>
Use conventional net_ratelimit() instead.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
---
drivers/net/ethernet/sfc/ptp.c | 22 ++--------------------
1 files changed, 2 insertions(+), 20 deletions(-)
diff --git a/drivers/net/ethernet/sfc/ptp.c b/drivers/net/ethernet/sfc/ptp.c
index eb75fbd..52be63d 100644
--- a/drivers/net/ethernet/sfc/ptp.c
+++ b/drivers/net/ethernet/sfc/ptp.c
@@ -223,7 +223,6 @@ struct efx_ptp_timeset {
* @evt_list: List of MC receive events awaiting packets
* @evt_free_list: List of free events
* @evt_lock: Lock for manipulating evt_list and evt_free_list
- * @evt_overflow: Boolean indicating that event list has overflowed
* @rx_evts: Instantiated events (on evt_list and evt_free_list)
* @workwq: Work queue for processing pending PTP operations
* @work: Work task
@@ -275,7 +274,6 @@ struct efx_ptp_data {
struct list_head evt_list;
struct list_head evt_free_list;
spinlock_t evt_lock;
- bool evt_overflow;
struct efx_ptp_event_rx rx_evts[MAX_RECEIVE_EVENTS];
struct workqueue_struct *workwq;
struct work_struct work;
@@ -941,11 +939,6 @@ static void efx_ptp_drop_time_expired_events(struct efx_nic *efx)
}
}
}
- /* If the event overflow flag is set and the event list is now empty
- * clear the flag to re-enable the overflow warning message.
- */
- if (ptp->evt_overflow && list_empty(&ptp->evt_list))
- ptp->evt_overflow = false;
spin_unlock_bh(&ptp->evt_lock);
}
@@ -989,11 +982,6 @@ static enum ptp_packet_state efx_ptp_match_rx(struct efx_nic *efx,
break;
}
}
- /* If the event overflow flag is set and the event list is now empty
- * clear the flag to re-enable the overflow warning message.
- */
- if (ptp->evt_overflow && list_empty(&ptp->evt_list))
- ptp->evt_overflow = false;
spin_unlock_bh(&ptp->evt_lock);
return rc;
@@ -1147,7 +1135,6 @@ static int efx_ptp_stop(struct efx_nic *efx)
list_for_each_safe(cursor, next, &efx->ptp_data->evt_list) {
list_move(cursor, &efx->ptp_data->evt_free_list);
}
- ptp->evt_overflow = false;
spin_unlock_bh(&efx->ptp_data->evt_lock);
return rc;
@@ -1253,7 +1240,6 @@ int efx_ptp_probe(struct efx_nic *efx, struct efx_channel *channel)
spin_lock_init(&ptp->evt_lock);
for (pos = 0; pos < MAX_RECEIVE_EVENTS; pos++)
list_add(&ptp->rx_evts[pos].link, &ptp->evt_free_list);
- ptp->evt_overflow = false;
/* Get the NIC PTP attributes and set up time conversions */
rc = efx_ptp_get_attributes(efx);
@@ -1635,13 +1621,9 @@ static void ptp_event_rx(struct efx_nic *efx, struct efx_ptp_data *ptp)
list_add_tail(&evt->link, &ptp->evt_list);
queue_work(ptp->workwq, &ptp->work);
- } else if (!ptp->evt_overflow) {
- /* Log a warning message and set the event overflow flag.
- * The message won't be logged again until the event queue
- * becomes empty.
- */
+ } else if (net_ratelimit()) {
+ /* Log a rate-limited warning message. */
netif_err(efx, rx_err, efx->net_dev, "PTP event queue overflow\n");
- ptp->evt_overflow = true;
}
spin_unlock_bh(&ptp->evt_lock);
}
^ permalink raw reply related
* [PATCH net-next 00/13] Cleanup patches for the SFC driver
From: Shradha Shah @ 2014-02-12 18:56 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-net-drivers
This patch set consists of some cleanup and housekeeping
patches for the sfc driver.
These patches help to reduce the differences between the in-
tree and out-of-tree driver.
Ben Hutchings (12):
sfc: Cache skb->data in local variable in efx_ptp_rx()
sfc: Rewrite adjustment of PPS event in a clearer way
sfc: Replace TSOH_OFFSET with the equivalent NET_IP_ALIGN
sfc: Rename 'use_options' variable in tso_start() to clearer
'use_opt_desc'
sfc: Remove unused definitions of EF10 user-mode DMA descriptors
sfc: Correct comment about number of TX queues used on EF10
sfc: Preserve rx_frm_trunc counters when resizing DMA rings
sfc: Use canonical pointer type for MAC address in
efx_set_mac_address()
sfc: Update product naming
sfc: Cosmetic changes to self-test from the out-of-tree driver
sfc: Fail self-test with -EBUSY, not -EIO, if the device is busy
sfc: Add/remove blank lines to taste
Laurence Evans (1):
sfc: Removed adhoc scheme to rate limit PTP event queue overflow
message
drivers/net/ethernet/sfc/ef10.c | 4 +-
drivers/net/ethernet/sfc/ef10_regs.h | 61 ----------------------
drivers/net/ethernet/sfc/efx.c | 6 +--
drivers/net/ethernet/sfc/efx.h | 2 +-
drivers/net/ethernet/sfc/ethtool.c | 21 +++++---
drivers/net/ethernet/sfc/falcon.c | 4 +-
drivers/net/ethernet/sfc/farch.c | 2 -
drivers/net/ethernet/sfc/net_driver.h | 1 -
drivers/net/ethernet/sfc/nic.c | 1 -
drivers/net/ethernet/sfc/ptp.c | 92 ++++++++++++++-------------------
drivers/net/ethernet/sfc/tx.c | 21 ++-----
11 files changed, 65 insertions(+), 150 deletions(-)
^ permalink raw reply
* Re: RFC: bridge get fdb by bridge device
From: John Fastabend @ 2014-02-12 18:50 UTC (permalink / raw)
To: Jamal Hadi Salim
Cc: vyasevic, netdev@vger.kernel.org, Stephen Hemminger,
Scott Feldman
In-Reply-To: <52FA9074.2060900@mojatatu.com>
On 2/11/2014 1:04 PM, Jamal Hadi Salim wrote:
> On 02/11/14 15:30, John Fastabend wrote:
>> On 2/11/2014 12:15 PM, Jamal Hadi Salim wrote:
>
> Thanks for the example on the other email.
>
Just to wrap things up in one email. Changing between VEB and VEPA
modes already triggers an event. So management applications can listen
for this.
And I can send out a patch to add a flag to hardware bridge devices
I'll likely get to it next week sometime unless someone beats me
to it.
>
>> What do you mean by "bridge device" are you specifically talking about
>> IFF_BRIDGE flag? This flag is used only for ./net/bridge devices.
>
> Right - the simple definition is this thing has an fdb.
> Yes, I know weve added vlan filtering and multicast snooping
> but thats all lipstick. If it has an (ethernet) fdb it is a bridge.
>
Sure, IEEE802.1Q would call these edge relays.
>> For
>> example macvlan uses its own flag. I think there is a good case to be
>> made for netdevices which are acting as the management interface for a
>> hardware bridge to set an identifying flag. Perhaps IFF_HWBRIDGE.
>>
>
> If you introduce IFF_HWBRIDGE - I think that would satisfy the
> distinction. The question then is why not just tag it IFF_BRIDGE?
>
Because it is not the same type of object as the software bridge.
Most notably it doesn't do learning. If anything its more like a
macvlan device and we could just as easily tag it IFF_MACVLAN. So
because it doesn't really match 1:1 with either of those object I
would just presume give its own flag. Userspace can always create
a small macro call it is_bridge_like() and check for any of the
handful of bridge like objects.
>>
>> # ip link set dev bridge0 master bridge1
>> RTNETLINK answers: Too many levels of symbolic links
>>
>
> pourquoi? If the original rationale was to limit the
> broadcast domain scope it sounds strange that a bridge in
> the form a macvlan is allowed.
>
Agreed. But there it is.
>> in the bridge case this doesn't work. But you can stack a macvlan
>> on top of the bridge port,
>>
>> # ip link add link bridge0 type macvlan mode vepa
>>
>> 11: macvlan0@bridge0: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop
>> state DOWN mode DEFAULT group default
>>
>> And macvlans on macvlans is OK as well.
>>
>> # ip link add link macvlan0 type macvlan mode vepa
>>
>> [...]
>>
>
> Ok, I need to let that sink in. Cool actually.
>
Also note you can mix VEB and VEPA modes and if you do it correctly
can create blocks of virtual ports that can do east-west traffic and
isolate others.
>
>>
>> If its useful then we should. You can track them down in userspace
>> via /sys/class/net/ or looking for offloaded netdevices that point
>> to the interface but a flag is definitely more direct.
>>
>
> I prefer a flag. Then i can deal with it via netlink.
>
OK. I'll add one here shortly.
> cheers,
> jamal
>
^ permalink raw reply
* Re: [PATCH net 1/3] kref: add kref_sub_return
From: Anatol Pomozov @ 2014-02-12 18:39 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: Greg Kroah-Hartman, LKML, Jason Wang, virtio-dev, virtualization,
netdev, davem, qinchuanyu, Joern Engel, Nicholas Bellinger
In-Reply-To: <20140212173524.GA26860@redhat.com>
Hi
On Wed, Feb 12, 2014 at 9:35 AM, Michael S. Tsirkin <mst@redhat.com> wrote:
> On Wed, Feb 12, 2014 at 08:56:30AM -0800, Greg Kroah-Hartman wrote:
>> On Wed, Feb 12, 2014 at 06:38:21PM +0200, Michael S. Tsirkin wrote:
>> > It is sometimes useful to get the value of the reference count after
>> > decrement.
>> > For example, vhost wants to execute some periodic cleanup operations
>> > once number of references drops below a specific value, before it
>> > reaches zero (for efficiency).
>>
>> You should never care about what the value of the kref is, if you are
>> using it correctly :)
>>
>> So I really don't want to add this function, as I'm sure people will use
>> it incorrectly. You should only care if the reference drops to 0, if
>> not, then your usage doesn't really fit into the "kref" model, and so,
>> just use an atomic variable.
>
> This happens when you have code that keeps
> reference itself implicitly or explicitly.
>
> foo(struct kref *k, int bar) {
>
> sub = kref_sub(k)
>
> if (sub == 1)
At this moment you cannot be sure that refcount is 1. It might be
changed between kref_sub call and this point. The refcount might be 0,
1, ... or any other value.
The only value that you can be sure is zero. Once refcount becomes
zero there is no way to increase it to positive value as there are no
alive pointers to the object.
> FOO(k, bar) /* Here I am the only one
> with a reference */
>
> }
>
> kref_get(k)
> foo(k, bar);
> ....
> kref_put(k)
>
> Why not do FOO in destructor you ask?
> Absolutely but this will be called much later.
>
> Maybe you will reconsider if I document this
> as the only legal use?
>
>>
>> I really want to know why it matters for "efficiency" that you know this
>> number. How does that help anything, as the number could then go up
>> later on, and the work you did at a "lower" number is obsolete, right?
>>
>> thanks,
>>
>> greg k-h
>
> The issue is that if number dropped to 1, this means
> we must do the cleanup work since there are
> no outstanding buffers, (last user is ourselves)
> if we do not cleanup,
> guest will hang waiting for us.
>
> But it never drops to 0 since we have our own reference
> in the device.
> If it goes up again this means we didn't have
> to do cleanup, but an alternative is doing
> it all the time and that is slow.
>
> Yes I can rework vhost to open-code this kref use, it's
> no big deal.
> Alternatively since most of the use does match kref
> model, maybe __kref_sub_return with disclaimers
> that you must know what you are doing?
> Please let me know.
>
> Thanks!
>
> --
> MST
^ permalink raw reply
* Re: [PATCH net 1/3] kref: add kref_sub_return
From: Greg Kroah-Hartman @ 2014-02-12 18:37 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: virtio-dev, Joern Engel, Anatol Pomozov, netdev, linux-kernel,
virtualization, qinchuanyu, davem
In-Reply-To: <20140212173524.GA26860@redhat.com>
On Wed, Feb 12, 2014 at 07:35:24PM +0200, Michael S. Tsirkin wrote:
> On Wed, Feb 12, 2014 at 08:56:30AM -0800, Greg Kroah-Hartman wrote:
> > On Wed, Feb 12, 2014 at 06:38:21PM +0200, Michael S. Tsirkin wrote:
> > > It is sometimes useful to get the value of the reference count after
> > > decrement.
> > > For example, vhost wants to execute some periodic cleanup operations
> > > once number of references drops below a specific value, before it
> > > reaches zero (for efficiency).
> >
> > You should never care about what the value of the kref is, if you are
> > using it correctly :)
> >
> > So I really don't want to add this function, as I'm sure people will use
> > it incorrectly. You should only care if the reference drops to 0, if
> > not, then your usage doesn't really fit into the "kref" model, and so,
> > just use an atomic variable.
>
> This happens when you have code that keeps
> reference itself implicitly or explicitly.
>
> foo(struct kref *k, int bar) {
>
> sub = kref_sub(k)
>
> if (sub == 1)
> FOO(k, bar) /* Here I am the only one
> with a reference */
Why do you care if you are the only one with a reference?
If you do, then just don't grab that reference and do the work in the
cleanup callback :)
> }
>
> kref_get(k)
> foo(k, bar);
> ....
> kref_put(k)
>
> Why not do FOO in destructor you ask?
> Absolutely but this will be called much later.
>
> Maybe you will reconsider if I document this
> as the only legal use?
No one reads documentation :(
> > I really want to know why it matters for "efficiency" that you know this
> > number. How does that help anything, as the number could then go up
> > later on, and the work you did at a "lower" number is obsolete, right?
> >
> > thanks,
> >
> > greg k-h
>
> The issue is that if number dropped to 1, this means
> we must do the cleanup work since there are
> no outstanding buffers, (last user is ourselves)
> if we do not cleanup,
> guest will hang waiting for us.
This doesn't make sense, nor does it sound like a use for a kref (or you
are using it wrong.)
> But it never drops to 0 since we have our own reference
> in the device.
Then don't do that.
> If it goes up again this means we didn't have
> to do cleanup, but an alternative is doing
> it all the time and that is slow.
Then just cleanup when it hits 0, like the rest of the world does.
> Yes I can rework vhost to open-code this kref use, it's
> no big deal.
> Alternatively since most of the use does match kref
> model, maybe __kref_sub_return with disclaimers
> that you must know what you are doing?
No, no one reads documentation, sorry. Either fix your use of kref
(i.e. don't care about the count), or do something else, as you don't
want a kref, but rather, an atomic count of what is going on and "1"
means something "special" to you.
sorry,
greg k-h
^ permalink raw reply
* [PATCH net-next 14/14] tipc: add node_lock protection to link lookup function
From: Jon Maloy @ 2014-02-12 18:31 UTC (permalink / raw)
To: davem; +Cc: Jon Maloy, netdev, tipc-discussion
In-Reply-To: <1392229874-29675-1-git-send-email-jon.maloy@ericsson.com>
In an earlier commit, ("tipc: remove links list from bearer struct")
we described three issues that need to be pre-emptively resolved before
we can remove tipc_net_lock. Here we resolve issue a) described in that
commit:
"a) In access method #2, we access the link before taking the
protecting node_lock. This will not work once net_lock is gone,
so we will have to change the access order. We will deal with
this in a later commit in this series."
Here, we change that access order, by ensuring that the function
link_find_link() returns only a safe reference for finding
the link, i.e., a node pointer and an index into its 'links' array,
not the link pointer itself. We also change all callers of this
function to first take the node lock before they can check if there
still is a valid link pointer at the returned index. Since the
function now returns a node pointer rather than a link pointer,
we rename it to the more appropriate 'tipc_link_find_owner().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
---
net/tipc/link.c | 110 ++++++++++++++++++++++++++++++++-----------------------
1 file changed, 65 insertions(+), 45 deletions(-)
diff --git a/net/tipc/link.c b/net/tipc/link.c
index a13b90d..11f2222 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -2389,35 +2389,40 @@ void tipc_link_set_queue_limits(struct tipc_link *l_ptr, u32 window)
l_ptr->queue_limit[MSG_FRAGMENTER] = 4000;
}
-/**
- * link_find_link - locate link by name
- * @name: ptr to link name string
- * @node: ptr to area to be filled with ptr to associated node
- *
+/* tipc_link_find_owner - locate owner node of link by link's name
+ * @name: pointer to link name string
+ * @bearer_id: pointer to index in 'node->links' array where the link was found.
* Caller must hold 'tipc_net_lock' to ensure node and bearer are not deleted;
* this also prevents link deletion.
*
- * Returns pointer to link (or 0 if invalid link name).
+ * Returns pointer to node owning the link, or 0 if no matching link is found.
*/
-static struct tipc_link *link_find_link(const char *name,
- struct tipc_node **node)
+static struct tipc_node *tipc_link_find_owner(const char *link_name,
+ unsigned int *bearer_id)
{
struct tipc_link *l_ptr;
struct tipc_node *n_ptr;
+ struct tipc_node *tmp_n_ptr;
+ struct tipc_node *found_node = 0;
+
int i;
- list_for_each_entry(n_ptr, &tipc_node_list, list) {
+ *bearer_id = 0;
+ list_for_each_entry_safe(n_ptr, tmp_n_ptr, &tipc_node_list, list) {
+ spin_lock(&n_ptr->lock);
for (i = 0; i < MAX_BEARERS; i++) {
l_ptr = n_ptr->links[i];
- if (l_ptr && !strcmp(l_ptr->name, name))
- goto found;
+ if (l_ptr && !strcmp(l_ptr->name, link_name)) {
+ *bearer_id = i;
+ found_node = n_ptr;
+ break;
+ }
}
+ spin_unlock(&n_ptr->lock);
+ if (found_node)
+ break;
}
- l_ptr = NULL;
- n_ptr = NULL;
-found:
- *node = n_ptr;
- return l_ptr;
+ return found_node;
}
/**
@@ -2459,32 +2464,33 @@ static int link_cmd_set_value(const char *name, u32 new_value, u16 cmd)
struct tipc_link *l_ptr;
struct tipc_bearer *b_ptr;
struct tipc_media *m_ptr;
+ int bearer_id;
int res = 0;
- l_ptr = link_find_link(name, &node);
- if (l_ptr) {
- /*
- * acquire node lock for tipc_link_send_proto_msg().
- * see "TIPC locking policy" in net.c.
- */
+ node = tipc_link_find_owner(name, &bearer_id);
+ if (node) {
tipc_node_lock(node);
- switch (cmd) {
- case TIPC_CMD_SET_LINK_TOL:
- link_set_supervision_props(l_ptr, new_value);
- tipc_link_send_proto_msg(l_ptr,
- STATE_MSG, 0, 0, new_value, 0, 0);
- break;
- case TIPC_CMD_SET_LINK_PRI:
- l_ptr->priority = new_value;
- tipc_link_send_proto_msg(l_ptr,
- STATE_MSG, 0, 0, 0, new_value, 0);
- break;
- case TIPC_CMD_SET_LINK_WINDOW:
- tipc_link_set_queue_limits(l_ptr, new_value);
- break;
- default:
- res = -EINVAL;
- break;
+ l_ptr = node->links[bearer_id];
+
+ if (l_ptr) {
+ switch (cmd) {
+ case TIPC_CMD_SET_LINK_TOL:
+ link_set_supervision_props(l_ptr, new_value);
+ tipc_link_send_proto_msg(l_ptr, STATE_MSG, 0,
+ 0, new_value, 0, 0);
+ break;
+ case TIPC_CMD_SET_LINK_PRI:
+ l_ptr->priority = new_value;
+ tipc_link_send_proto_msg(l_ptr, STATE_MSG, 0,
+ 0, 0, new_value, 0);
+ break;
+ case TIPC_CMD_SET_LINK_WINDOW:
+ tipc_link_set_queue_limits(l_ptr, new_value);
+ break;
+ default:
+ res = -EINVAL;
+ break;
+ }
}
tipc_node_unlock(node);
return res;
@@ -2579,6 +2585,7 @@ struct sk_buff *tipc_link_cmd_reset_stats(const void *req_tlv_area, int req_tlv_
char *link_name;
struct tipc_link *l_ptr;
struct tipc_node *node;
+ unsigned int bearer_id;
if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_NAME))
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
@@ -2589,15 +2596,19 @@ struct sk_buff *tipc_link_cmd_reset_stats(const void *req_tlv_area, int req_tlv_
return tipc_cfg_reply_error_string("link not found");
return tipc_cfg_reply_none();
}
-
read_lock_bh(&tipc_net_lock);
- l_ptr = link_find_link(link_name, &node);
+ node = tipc_link_find_owner(link_name, &bearer_id);
+ if (!node) {
+ read_unlock_bh(&tipc_net_lock);
+ return tipc_cfg_reply_error_string("link not found");
+ }
+ spin_lock(&node->lock);
+ l_ptr = node->links[bearer_id];
if (!l_ptr) {
+ tipc_node_unlock(node);
read_unlock_bh(&tipc_net_lock);
return tipc_cfg_reply_error_string("link not found");
}
-
- tipc_node_lock(node);
link_reset_statistics(l_ptr);
tipc_node_unlock(node);
read_unlock_bh(&tipc_net_lock);
@@ -2627,18 +2638,27 @@ static int tipc_link_stats(const char *name, char *buf, const u32 buf_size)
struct tipc_node *node;
char *status;
u32 profile_total = 0;
+ unsigned int bearer_id;
int ret;
if (!strcmp(name, tipc_bclink_name))
return tipc_bclink_stats(buf, buf_size);
read_lock_bh(&tipc_net_lock);
- l = link_find_link(name, &node);
- if (!l) {
+ node = tipc_link_find_owner(name, &bearer_id);
+ if (!node) {
read_unlock_bh(&tipc_net_lock);
return 0;
}
tipc_node_lock(node);
+
+ l = node->links[bearer_id];
+ if (!l) {
+ tipc_node_unlock(node);
+ read_unlock_bh(&tipc_net_lock);
+ return 0;
+ }
+
s = &l->stats;
if (tipc_link_is_active(l))
--
1.7.9.5
------------------------------------------------------------------------------
Android apps run on BlackBerry 10
Introducing the new BlackBerry 10.2.1 Runtime for Android apps.
Now with support for Jelly Bean, Bluetooth, Mapview and more.
Get your Android app in front of a whole new audience. Start now.
http://pubads.g.doubleclick.net/gampad/clk?id=124407151&iu=/4140/ostg.clktrk
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox