* Re: [PATCH net-next] net: ethernet: ti: cpsw: use for mcast entries only host port
From: Grygorii Strashko @ 2018-10-13 0:41 UTC (permalink / raw)
To: Ivan Khoronzhuk, davem; +Cc: linux-omap, netdev, linux-kernel
In-Reply-To: <20181012160629.7245-1-ivan.khoronzhuk@linaro.org>
On 10/12/2018 11:06 AM, Ivan Khoronzhuk wrote:
> In dual-emac mode the cpsw driver sends directed packets, that means
> that packets go to the directed port, but an ALE lookup is performed
> to determine untagged egress only. It means that on tx side no need
> to add port bit for ALE mcast entry mask, and basically ALE entry
> for port identification is needed only on rx side.
>
> So, add only host port in dual_emac mode as used directed
> transmission, and no need in one more port. For single port boards
> and switch mode all ports used, as usual, so no changes for them.
> Also it simplifies farther changes.
>
> In other words, mcast entries for dual-emac should behave exactly
> like unicast. It also can help avoid leaking packets between ports
> with same vlan on h/w level if ports could became members of same vid.
>
> So now, for instance, if mcast address 33:33:00:00:00:01 is added then
> entries in ALE table:
>
> vid = 1, addr = 33:33:00:00:00:01, port_mask = 0x1
> vid = 2, addr = 33:33:00:00:00:01, port_mask = 0x1
>
> Instead of:
> vid = 1, addr = 33:33:00:00:00:01, port_mask = 0x3
> vid = 2, addr = 33:33:00:00:00:01, port_mask = 0x5
>
> With the same considerations, set only host port for unregistered
> mcast for dual-emac mode in case of IFF_ALLMULTI is set, exactly like
> it's done in cpsw_ale_set_allmulti().
>
> Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
> ---
>
> Based on net-next/master
Thank you.
Reviewed-by: Grygorii Strashko <grygorii.strashko@ti.com>
--
regards,
-grygorii
^ permalink raw reply
* Re: [PATCH 0/2] net: ethernet: ti: cpsw fix mcast packet lost
From: Grygorii Strashko @ 2018-10-13 0:38 UTC (permalink / raw)
To: Ivan Khoronzhuk, davem; +Cc: linux-omap, netdev, linux-kernel
In-Reply-To: <20181012152815.31320-1-ivan.khoronzhuk@linaro.org>
On 10/12/2018 10:28 AM, Ivan Khoronzhuk wrote:
> The patchset omits redundant refresh of mcast address table and
> prevents mcast packet lost.
>
> Based on net-next/master
> tested on am572x evm
>
> Ivan Khoronzhuk (2):
> net: ethernet: ti: cpsw_ale: use const for API having pointer on mac
> address
> net: ethernet: ti: cpsw: fix lost of mcast packets while rx_mode
> update
>
> drivers/net/ethernet/ti/cpsw.c | 46 ++++++++++++++++++------------
> drivers/net/ethernet/ti/cpsw_ale.c | 12 ++++----
> drivers/net/ethernet/ti/cpsw_ale.h | 8 +++---
> 3 files changed, 38 insertions(+), 28 deletions(-)
>
Thank you.
Reviewed-by: Grygorii Strashko <grygorii.strashko@ti.com>
--
regards,
-grygorii
^ permalink raw reply
* Re: [PATCH] fore200e: fix sbus compile
From: David Miller @ 2018-10-12 16:49 UTC (permalink / raw)
To: hch; +Cc: netdev, 3chas3
In-Reply-To: <20181012081751.29076-1-hch@lst.de>
From: Christoph Hellwig <hch@lst.de>
Date: Fri, 12 Oct 2018 10:17:51 +0200
> Fix a stupid typo introduced in the refactoring.
>
> Fixes: 0efe5523 ("fore200e: simplify fore200e_bus usage")
> Signed-off-by: Christoph Hellwig <hch@lst.de>
Applied.
^ permalink raw reply
* Re: [PATCH net-next] net: Evict neighbor entries on carrier down
From: David Miller @ 2018-10-12 16:49 UTC (permalink / raw)
To: dsahern; +Cc: netdev, roopa, idosch, dsahern
In-Reply-To: <20181012033349.5213-1-dsahern@kernel.org>
From: David Ahern <dsahern@kernel.org>
Date: Thu, 11 Oct 2018 20:33:49 -0700
> From: David Ahern <dsahern@gmail.com>
>
> When a link's carrier goes down it could be a sign of the port changing
> networks. If the new network has overlapping addresses with the old one,
> then the kernel will continue trying to use neighbor entries established
> based on the old network until the entries finally age out - meaning a
> potentially long delay with communications not working.
>
> This patch evicts neighbor entries on carrier down with the exception of
> those marked permanent. Permanent entries are managed by userspace (either
> an admin or a routing daemon such as FRR).
>
> Signed-off-by: David Ahern <dsahern@gmail.com>
Applied.
^ permalink raw reply
* Re: [PATCH v2 net-next] net/ipv6: Add knob to skip DELROUTE message on device down
From: David Miller @ 2018-10-12 16:49 UTC (permalink / raw)
To: dsahern; +Cc: netdev, roopa, dsahern
In-Reply-To: <20181012031721.11648-1-dsahern@kernel.org>
From: David Ahern <dsahern@kernel.org>
Date: Thu, 11 Oct 2018 20:17:21 -0700
> From: David Ahern <dsahern@gmail.com>
>
> Another difference between IPv4 and IPv6 is the generation of RTM_DELROUTE
> notifications when a device is taken down (admin down) or deleted. IPv4
> does not generate a message for routes evicted by the down or delete;
> IPv6 does. A NOS at scale really needs to avoid these messages and have
> IPv4 and IPv6 behave similarly, relying on userspace to handle link
> notifications and evict the routes.
>
> At this point existing user behavior needs to be preserved. Since
> notifications are a global action (not per app) the only way to preserve
> existing behavior and allow the messages to be skipped is to add a new
> sysctl (net/ipv6/route/skip_notify_on_dev_down) which can be set to
> disable the notificatioons.
>
> IPv6 route code already supports the option to skip the message (it is
> used for multipath routes for example). Besides the new sysctl we need
> to pass the skip_notify setting through the generic fib6_clean and
> fib6_walk functions to fib6_clean_node and to set skip_notify on calls
> to __ip_del_rt for the addrconf_ifdown path.
>
> Signed-off-by: David Ahern <dsahern@gmail.com>
> ---
> v2
> - removed the changes to addrconf and anycast. addrconf_ifdown calls
> rt6_disable_ip which calls rt6_sync_down_dev. The last one evicts all
> routes for the device, so the delete route calls done later in addrconf
> and anycast are superfluous
Applied.
^ permalink raw reply
* Re: [PATCH net-next] net: cdc_ncm: remove set but not used variable 'ctx'
From: David Miller @ 2018-10-12 16:48 UTC (permalink / raw)
To: yuehaibing; +Cc: oliver, linux-usb, netdev, kernel-janitors
In-Reply-To: <1539308953-185007-1-git-send-email-yuehaibing@huawei.com>
From: YueHaibing <yuehaibing@huawei.com>
Date: Fri, 12 Oct 2018 01:49:13 +0000
> Fixes gcc '-Wunused-but-set-variable' warning:
>
> drivers/net/usb/cdc_ncm.c: In function 'cdc_ncm_status':
> drivers/net/usb/cdc_ncm.c:1603:22: warning:
> variable 'ctx' set but not used [-Wunused-but-set-variable]
> struct cdc_ncm_ctx *ctx;
>
> It not used any more after
> commit fa83dbeee558 ("net: cdc_ncm: remove redundant "disconnected" flag")
>
> Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Applied.
^ permalink raw reply
* RE: [PATCH 05/16] igbvf: Replace spin_is_locked() with lockdep
From: Brown, Aaron F @ 2018-10-13 0:11 UTC (permalink / raw)
To: Lance Roy, linux-kernel@vger.kernel.org
Cc: Paul E. McKenney, Kirsher, Jeffrey T, David S. Miller,
intel-wired-lan@lists.osuosl.org, netdev@vger.kernel.org
In-Reply-To: <20181003053902.6910-6-ldr709@gmail.com>
> From: netdev-owner@vger.kernel.org [mailto:netdev-
> owner@vger.kernel.org] On Behalf Of Lance Roy
> Sent: Tuesday, October 2, 2018 10:39 PM
> To: linux-kernel@vger.kernel.org
> Cc: Paul E. McKenney <paulmck@linux.ibm.com>; Lance Roy
> <ldr709@gmail.com>; Kirsher, Jeffrey T <jeffrey.t.kirsher@intel.com>; David
> S. Miller <davem@davemloft.net>; intel-wired-lan@lists.osuosl.org;
> netdev@vger.kernel.org
> Subject: [PATCH 05/16] igbvf: Replace spin_is_locked() with lockdep
>
> lockdep_assert_held() is better suited to checking locking requirements,
> since it won't get confused when someone else holds the lock. This is
> also a step towards possibly removing spin_is_locked().
>
> Signed-off-by: Lance Roy <ldr709@gmail.com>
> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
> Cc: "David S. Miller" <davem@davemloft.net>
> Cc: intel-wired-lan@lists.osuosl.org
> Cc: <netdev@vger.kernel.org>
> ---
> drivers/net/ethernet/intel/igbvf/mbx.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
^ permalink raw reply
* Re: [PATCH net] ip6_tunnel: Don't update PMTU on tunnels with collect_md
From: Stefano Brivio @ 2018-10-12 16:34 UTC (permalink / raw)
To: Nicolas Dichtel; +Cc: David S. Miller, Alexei Starovoitov, netdev
In-Reply-To: <8622979d-ea2d-82e2-600e-c0dd7d7e3e8a@6wind.com>
On Fri, 12 Oct 2018 17:58:55 +0200
Nicolas Dichtel <nicolas.dichtel@6wind.com> wrote:
> Le 12/10/2018 à 14:32, Stefano Brivio a écrit :
> > Commit 8d79266bc48c ("ip6_tunnel: add collect_md mode to IPv6
> > tunnels") introduced a check to avoid updating PMTU when
> > collect_md mode is enabled.
> >
> > Later, commit f15ca723c1eb ("net: don't call update_pmtu
> > unconditionally") dropped this check, I guess inadvertently.
> I removed it because update_pmtu() is not set for md_dst_op, thus I assume this
> check was done for that purpose.
Ha, sounds reasonable, yes.
> Could you explain in your commit log which problem your patch fixes?
Nothing really.
The change in f15ca723c1eb looked accidental and I thought it doesn't
make sense to update the PMTU in that case, but I didn't figure out
it's not actually done anyway.
Maybe it makes things a bit more readable, in that case I'd target it
for net-next. What do you think?
--
Stefano
^ permalink raw reply
* Re: [PATCH] dt-bindings: Add bindings for aliases node
From: Christian Lamparter @ 2018-10-12 16:31 UTC (permalink / raw)
To: Matthias Kaehlcke
Cc: Brian Norris, Geert Uytterhoeven, Rob Herring, Mark Rutland,
open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
Linux Kernel Mailing List, linux-wireless, linux-spi, netdev,
swboyd, Florian Fainelli
In-Reply-To: <20181012000837.GP22824@google.com>
On Friday, October 12, 2018 2:08:37 AM CEST Matthias Kaehlcke wrote:
> Maybe the doc should include a recommendation to use aliases
> sparingly? I'm open to input on that from folks who have a better
> understanding of the potential pitfalls
I had a similar discussion with the OpenWrt devs over the
use of "led-$function" aliases in a DTS. I did a bit of digging and
found this wonderful emails from Mark Rutland regarding the general
use and abuse of aliases in a reply to a patch by Christer Weinigel
"devicetree - document using aliases to set spi bus number."
<https://patchwork.kernel.org/patch/9133903/#19207021>
|"If those ports are physically organised and labelled the same, then
|using aliases could make sense, to describe the well-defined physical
|labels. If you've assigned the numbers artificially, or if the physical
|organisation differs across boards, then aliases are not the right tool
|for the job.
|
|In the latter cases we're altering the hardware description to suit an
|application, rather than providing the necessary abstraction, which is
|the kind of (ab)use of aliases which we want to avoid."
And he followed it up with a summary:
<https://patchwork.kernel.org/comment/19207071/>
|Typically, serial ports are much more user-accessible (physically), and
|much more directly useful to a user in a generic fashion. They're often
|labelled (physically or in a manual) with a number, and we use aliases
|to describe those labels to the kernel. The fact that the kernel may use
|that to drive its own internal numbering is immaterial to the binding.
So the gist of this is that aliases are meant for user-accessible /
physically devices/ports/etc... that are labeled as such. And this of
course works perfectly for power/status LEDs and such because they
usually have little "power" symbols/pictograms/lables near them.
^ permalink raw reply
* [PATCH net-next 6/6] dpaa2-eth: remove unused FD field
From: Ioana Ciornei @ 2018-10-12 16:27 UTC (permalink / raw)
To: netdev@vger.kernel.org, davem@davemloft.net
Cc: Ioana Ciocoi Radulescu, Ioana Ciornei
In-Reply-To: <cover.1539361388.git.ioana.ciornei@nxp.com>
From: Ioana Radulescu <ruxandra.radulescu@nxp.com>
According to the hardware ArchDef, the PTV1 field in FD[CTRL]
is ignored by WRIOP, so setting it for Tx FDs is pointless.
Remove all references to it from the code.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
index 85628e6..88f7acc 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
@@ -437,7 +437,7 @@ static int build_sg_fd(struct dpaa2_eth_priv *priv,
dpaa2_fd_set_format(fd, dpaa2_fd_sg);
dpaa2_fd_set_addr(fd, addr);
dpaa2_fd_set_len(fd, skb->len);
- dpaa2_fd_set_ctrl(fd, FD_CTRL_PTA | FD_CTRL_PTV1);
+ dpaa2_fd_set_ctrl(fd, FD_CTRL_PTA);
if (priv->tx_tstamp && skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)
enable_tx_tstamp(fd, sgt_buf);
@@ -490,7 +490,7 @@ static int build_single_fd(struct dpaa2_eth_priv *priv,
dpaa2_fd_set_offset(fd, (u16)(skb->data - buffer_start));
dpaa2_fd_set_len(fd, skb->len);
dpaa2_fd_set_format(fd, dpaa2_fd_single);
- dpaa2_fd_set_ctrl(fd, FD_CTRL_PTA | FD_CTRL_PTV1);
+ dpaa2_fd_set_ctrl(fd, FD_CTRL_PTA);
if (priv->tx_tstamp && skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)
enable_tx_tstamp(fd, buffer_start);
--
1.9.1
^ permalink raw reply related
* [PATCH net-next 5/6] dpaa2-eth: mark unused parameter in dpaa2_eth_tx_conf
From: Ioana Ciornei @ 2018-10-12 16:27 UTC (permalink / raw)
To: netdev@vger.kernel.org, davem@davemloft.net
Cc: Ioana Ciocoi Radulescu, Ioana Ciornei
In-Reply-To: <cover.1539361388.git.ioana.ciornei@nxp.com>
The ch parameter is never used in the dpaa2_eth_tx_conf function but
since its prototype must match the type defined in the consume field of
struct dpaa2_eth_fq, just mark it as __always_unused.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
index 119551e..85628e6 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
@@ -659,7 +659,7 @@ static netdev_tx_t dpaa2_eth_tx(struct sk_buff *skb, struct net_device *net_dev)
/* Tx confirmation frame processing routine */
static void dpaa2_eth_tx_conf(struct dpaa2_eth_priv *priv,
- struct dpaa2_eth_channel *ch,
+ struct dpaa2_eth_channel *ch __always_unused,
const struct dpaa2_fd *fd,
struct napi_struct *napi __always_unused,
u16 queue_id __always_unused)
--
1.9.1
^ permalink raw reply related
* [PATCH net-next 4/6] dpaa2-eth: remove unused priv parameter
From: Ioana Ciornei @ 2018-10-12 16:27 UTC (permalink / raw)
To: netdev@vger.kernel.org, davem@davemloft.net
Cc: Ioana Ciocoi Radulescu, Ioana Ciornei
In-Reply-To: <cover.1539361388.git.ioana.ciornei@nxp.com>
The priv parameter is never used in the build_linear_skb and
drain_channel function. Remove it from the function definitions.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 10 ++++------
1 file changed, 4 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
index dedbd67..119551e 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
@@ -98,8 +98,7 @@ static void free_rx_fd(struct dpaa2_eth_priv *priv,
}
/* Build a linear skb based on a single-buffer frame descriptor */
-static struct sk_buff *build_linear_skb(struct dpaa2_eth_priv *priv,
- struct dpaa2_eth_channel *ch,
+static struct sk_buff *build_linear_skb(struct dpaa2_eth_channel *ch,
const struct dpaa2_fd *fd,
void *fd_vaddr)
{
@@ -233,7 +232,7 @@ static void dpaa2_eth_rx(struct dpaa2_eth_priv *priv,
percpu_extras = this_cpu_ptr(priv->percpu_extras);
if (fd_format == dpaa2_fd_single) {
- skb = build_linear_skb(priv, ch, fd, vaddr);
+ skb = build_linear_skb(ch, fd, vaddr);
} else if (fd_format == dpaa2_fd_sg) {
skb = build_frag_skb(priv, ch, buf_data);
skb_free_frag(vaddr);
@@ -1085,8 +1084,7 @@ static int dpaa2_eth_open(struct net_device *net_dev)
/* The DPIO store must be empty when we call this,
* at the end of every NAPI cycle.
*/
-static u32 drain_channel(struct dpaa2_eth_priv *priv,
- struct dpaa2_eth_channel *ch)
+static u32 drain_channel(struct dpaa2_eth_channel *ch)
{
u32 drained = 0, total = 0;
@@ -1107,7 +1105,7 @@ static u32 drain_ingress_frames(struct dpaa2_eth_priv *priv)
for (i = 0; i < priv->num_channels; i++) {
ch = priv->channel[i];
- drained += drain_channel(priv, ch);
+ drained += drain_channel(ch);
}
return drained;
--
1.9.1
^ permalink raw reply related
* [PATCH net-next 3/6] dpaa2-eth: fix uninitialized variable warnings
From: Ioana Ciornei @ 2018-10-12 16:27 UTC (permalink / raw)
To: netdev@vger.kernel.org, davem@davemloft.net
Cc: Ioana Ciocoi Radulescu, Ioana Ciornei
In-Reply-To: <cover.1539361388.git.ioana.ciornei@nxp.com>
All 3 cases of possible uninitialized variables are false
positives since they are used only as output parameters.
Nonetheless, fix the warnings.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
index cf6de06..dedbd67 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
@@ -935,7 +935,7 @@ static int dpaa2_eth_poll(struct napi_struct *napi, int budget)
struct dpaa2_eth_channel *ch;
struct dpaa2_eth_priv *priv;
int rx_cleaned = 0, txconf_cleaned = 0;
- enum dpaa2_eth_fq_type type;
+ enum dpaa2_eth_fq_type type = 0;
int store_cleaned;
int err;
@@ -1002,7 +1002,7 @@ static void disable_ch_napi(struct dpaa2_eth_priv *priv)
static int link_state_update(struct dpaa2_eth_priv *priv)
{
- struct dpni_link_state state;
+ struct dpni_link_state state = {0};
int err;
err = dpni_get_link_state(priv->mc_io, 0, priv->mc_token, &state);
@@ -1116,7 +1116,7 @@ static u32 drain_ingress_frames(struct dpaa2_eth_priv *priv)
static int dpaa2_eth_stop(struct net_device *net_dev)
{
struct dpaa2_eth_priv *priv = netdev_priv(net_dev);
- int dpni_enabled;
+ int dpni_enabled = 0;
int retries = 10;
u32 drained;
--
1.9.1
^ permalink raw reply related
* [PATCH net-next 2/6] dpaa2-eth: make dpaa2_eth_set_dist_key static
From: Ioana Ciornei @ 2018-10-12 16:27 UTC (permalink / raw)
To: netdev@vger.kernel.org, davem@davemloft.net
Cc: Ioana Ciocoi Radulescu, Ioana Ciornei
In-Reply-To: <cover.1539361388.git.ioana.ciornei@nxp.com>
The dpaa2_eth_set_dist_key function is only used in a single file.
Make it static.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
index 156080d..cf6de06 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
@@ -2172,8 +2172,8 @@ int dpaa2_eth_cls_fld_off(int prot, int field)
/* Set Rx distribution (hash or flow classification) key
* flags is a combination of RXH_ bits
*/
-int dpaa2_eth_set_dist_key(struct net_device *net_dev,
- enum dpaa2_eth_rx_dist type, u64 flags)
+static int dpaa2_eth_set_dist_key(struct net_device *net_dev,
+ enum dpaa2_eth_rx_dist type, u64 flags)
{
struct device *dev = net_dev->dev.parent;
struct dpaa2_eth_priv *priv = netdev_priv(net_dev);
--
1.9.1
^ permalink raw reply related
* [PATCH net-next 1/6] dpaa2-eth: Fix Kconfig dependencies
From: Ioana Ciornei @ 2018-10-12 16:27 UTC (permalink / raw)
To: netdev@vger.kernel.org, davem@davemloft.net
Cc: Ioana Ciocoi Radulescu, Ioana Ciornei
In-Reply-To: <cover.1539361388.git.ioana.ciornei@nxp.com>
From: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Both ARCH_LAYERSCAPE and COMPILE_TEST dependencies are already implied
through the FSL_MC_BUS dep, so there's no need to state it explicitly.
Also, the fsl-mc bus depends on COMPILE_TEST only for some
architectures (arm, arm64, ppc, x86), so it's not correct to
claim build support unconditionally.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
drivers/net/ethernet/freescale/dpaa2/Kconfig | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/net/ethernet/freescale/dpaa2/Kconfig b/drivers/net/ethernet/freescale/dpaa2/Kconfig
index a7f365d..809a155 100644
--- a/drivers/net/ethernet/freescale/dpaa2/Kconfig
+++ b/drivers/net/ethernet/freescale/dpaa2/Kconfig
@@ -1,7 +1,6 @@
config FSL_DPAA2_ETH
tristate "Freescale DPAA2 Ethernet"
depends on FSL_MC_BUS && FSL_MC_DPIO
- depends on ARCH_LAYERSCAPE || COMPILE_TEST
help
This is the DPAA2 Ethernet driver supporting Freescale SoCs
with DPAA2 (DataPath Acceleration Architecture v2).
--
1.9.1
^ permalink raw reply related
* [PATCH net-next 0/6] dpaa2-eth: code cleanup
From: Ioana Ciornei @ 2018-10-12 16:27 UTC (permalink / raw)
To: netdev@vger.kernel.org, davem@davemloft.net
Cc: Ioana Ciocoi Radulescu, Ioana Ciornei
There are no functional changes in this patch set, only some cleanup
changes such as: unused parameters, uninitialized variables and
unnecessary Kconfig dependencies.
Ioana Ciornei (4):
dpaa2-eth: make dpaa2_eth_set_dist_key static
dpaa2-eth: fix uninitialized variable warnings
dpaa2-eth: remove unused priv parameter
dpaa2-eth: mark unused parameter in dpaa2_eth_tx_conf
Ioana Radulescu (2):
dpaa2-eth: Fix Kconfig dependencies
dpaa2-eth: remove unused FD field
drivers/net/ethernet/freescale/dpaa2/Kconfig | 1 -
drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 26 +++++++++++-------------
2 files changed, 12 insertions(+), 15 deletions(-)
--
1.9.1
^ permalink raw reply
* [PATCH v3 lora-next 3/5] net: lora: sx1301: convert to using regmap fields for bit ops
From: Ben Whitten @ 2018-10-12 16:26 UTC (permalink / raw)
To: afaerber
Cc: starnight, hasnain.virk, netdev, liuxuenetmail, shess,
Ben Whitten, David S. Miller, linux-kernel
In-Reply-To: <1539361567-3602-1-git-send-email-ben.whitten@lairdtech.com>
From: Ben Whitten <ben.whitten@gmail.com>
We convert to using regmap fields to allow bit access to the registers
where regmap handles the read update write.
Signed-off-by: Ben Whitten <ben.whitten@gmail.com>
---
drivers/net/lora/sx1301.c | 234 +++++++++++++---------------------------------
drivers/net/lora/sx1301.h | 46 +++++++++
2 files changed, 110 insertions(+), 170 deletions(-)
diff --git a/drivers/net/lora/sx1301.c b/drivers/net/lora/sx1301.c
index 9c85fe7..339f8d9 100644
--- a/drivers/net/lora/sx1301.c
+++ b/drivers/net/lora/sx1301.c
@@ -24,27 +24,6 @@
#include "sx1301.h"
-#define REG_PAGE_RESET_SOFT_RESET BIT(7)
-
-#define REG_16_GLOBAL_EN BIT(3)
-
-#define REG_17_CLK32M_EN BIT(0)
-
-#define REG_0_105_FORCE_HOST_RADIO_CTRL BIT(1)
-#define REG_0_105_FORCE_HOST_FE_CTRL BIT(2)
-#define REG_0_105_FORCE_DEC_FILTER_GAIN BIT(3)
-
-#define REG_0_MCU_RST_0 BIT(0)
-#define REG_0_MCU_RST_1 BIT(1)
-#define REG_0_MCU_SELECT_MUX_0 BIT(2)
-#define REG_0_MCU_SELECT_MUX_1 BIT(3)
-
-#define REG_2_43_RADIO_A_EN BIT(0)
-#define REG_2_43_RADIO_B_EN BIT(1)
-#define REG_2_43_RADIO_RST BIT(2)
-
-#define REG_EMERGENCY_FORCE_HOST_CTRL BIT(0)
-
static const struct regmap_range_cfg sx1301_regmap_ranges[] = {
{
.name = "Pages",
@@ -74,9 +53,10 @@ static struct regmap_config sx1301_regmap_config = {
.max_register = SX1301_MAX_REGISTER,
};
-static int sx1301_soft_reset(struct sx1301_priv *priv)
+static int sx1301_field_write(struct sx1301_priv *priv,
+ enum sx1301_fields field_id, u8 val)
{
- return regmap_write(priv->regmap, SX1301_PAGE, REG_PAGE_RESET_SOFT_RESET);
+ return regmap_field_write(priv->regmap_fields[field_id], val);
}
static int sx1301_agc_ram_read(struct sx1301_priv *priv, u8 addr, unsigned int *val)
@@ -120,7 +100,7 @@ static int sx1301_arb_ram_read(struct sx1301_priv *priv, u8 addr, unsigned int *
static int sx1301_load_firmware(struct sx1301_priv *priv, int mcu, const struct firmware *fw)
{
u8 *buf;
- u8 rst, select_mux;
+ enum sx1301_fields rst, select_mux;
unsigned int val;
int ret;
@@ -131,29 +111,26 @@ static int sx1301_load_firmware(struct sx1301_priv *priv, int mcu, const struct
switch (mcu) {
case 0:
- rst = REG_0_MCU_RST_0;
- select_mux = REG_0_MCU_SELECT_MUX_0;
+ rst = F_MCU_RST_0;
+ select_mux = F_MCU_SELECT_MUX_0;
break;
case 1:
- rst = REG_0_MCU_RST_1;
- select_mux = REG_0_MCU_SELECT_MUX_1;
+ rst = F_MCU_RST_1;
+ select_mux = F_MCU_SELECT_MUX_1;
break;
default:
return -EINVAL;
}
- ret = regmap_read(priv->regmap, SX1301_MCU_CTRL, &val);
+ ret = sx1301_field_write(priv, rst, 1);
if (ret) {
- dev_err(priv->dev, "MCU read failed\n");
+ dev_err(priv->dev, "MCU reset failed\n");
return ret;
}
- val |= rst;
- val &= ~select_mux;
-
- ret = regmap_write(priv->regmap, SX1301_MCU_CTRL, val);
+ ret = sx1301_field_write(priv, select_mux, 0);
if (ret) {
- dev_err(priv->dev, "MCU reset / select mux write failed\n");
+ dev_err(priv->dev, "MCU RAM select mux failed\n");
return ret;
}
@@ -194,17 +171,9 @@ static int sx1301_load_firmware(struct sx1301_priv *priv, int mcu, const struct
kfree(buf);
- ret = regmap_read(priv->regmap, SX1301_MCU_CTRL, &val);
- if (ret) {
- dev_err(priv->dev, "MCU read (1) failed\n");
- return ret;
- }
-
- val |= select_mux;
-
- ret = regmap_write(priv->regmap, SX1301_MCU_CTRL, val);
+ ret = sx1301_field_write(priv, select_mux, 1);
if (ret) {
- dev_err(priv->dev, "MCU reset / select mux write (1) failed\n");
+ dev_err(priv->dev, "MCU RAM release mux failed\n");
return ret;
}
@@ -230,17 +199,9 @@ static int sx1301_agc_calibrate(struct sx1301_priv *priv)
return ret;
}
- ret = regmap_read(priv->regmap, SX1301_FORCE_CTRL, &val);
- if (ret) {
- dev_err(priv->dev, "0|105 read failed\n");
- return ret;
- }
-
- val &= ~REG_0_105_FORCE_HOST_RADIO_CTRL;
-
- ret = regmap_write(priv->regmap, SX1301_FORCE_CTRL, val);
+ ret = sx1301_field_write(priv, F_FORCE_HOST_RADIO_CTRL, 0);
if (ret) {
- dev_err(priv->dev, "0|105 write failed\n");
+ dev_err(priv->dev, "force host control failed\n");
return ret;
}
@@ -254,17 +215,9 @@ static int sx1301_agc_calibrate(struct sx1301_priv *priv)
return ret;
}
- ret = regmap_read(priv->regmap, SX1301_MCU_CTRL, &val);
- if (ret) {
- dev_err(priv->dev, "MCU read (0) failed\n");
- return ret;
- }
-
- val &= ~REG_0_MCU_RST_1;
-
- ret = regmap_write(priv->regmap, SX1301_MCU_CTRL, val);
+ ret = sx1301_field_write(priv, F_MCU_RST_1, 0);
if (ret) {
- dev_err(priv->dev, "MCU write (0) failed\n");
+ dev_err(priv->dev, "MCU 1 reset failed\n");
return ret;
}
@@ -282,34 +235,18 @@ static int sx1301_agc_calibrate(struct sx1301_priv *priv)
return -ENXIO;
}
- ret = regmap_read(priv->regmap, SX1301_EMERGENCY_FORCE_HOST_CTRL, &val);
- if (ret) {
- dev_err(priv->dev, "emergency force read failed\n");
- return ret;
- }
-
- val &= ~REG_EMERGENCY_FORCE_HOST_CTRL;
-
- ret = regmap_write(priv->regmap, SX1301_EMERGENCY_FORCE_HOST_CTRL, val);
+ ret = sx1301_field_write(priv, F_EMERGENCY_FORCE_HOST_CTRL, 0);
if (ret) {
- dev_err(priv->dev, "emergency force write failed\n");
+ dev_err(priv->dev, "emergency force failed\n");
return ret;
}
dev_err(priv->dev, "starting calibration...\n");
msleep(2300);
- ret = regmap_read(priv->regmap, SX1301_EMERGENCY_FORCE_HOST_CTRL, &val);
- if (ret) {
- dev_err(priv->dev, "emergency force read (1) failed\n");
- return ret;
- }
-
- val |= REG_EMERGENCY_FORCE_HOST_CTRL;
-
- ret = regmap_write(priv->regmap, SX1301_EMERGENCY_FORCE_HOST_CTRL, val);
+ ret = sx1301_field_write(priv, F_EMERGENCY_FORCE_HOST_CTRL, 1);
if (ret) {
- dev_err(priv->dev, "emergency force write (1) failed\n");
+ dev_err(priv->dev, "emergency force release failed\n");
return ret;
}
@@ -356,19 +293,15 @@ static int sx1301_load_all_firmware(struct sx1301_priv *priv)
if (ret)
return ret;
- ret = regmap_read(priv->regmap, SX1301_FORCE_CTRL, &val);
- if (ret) {
- dev_err(priv->dev, "0|105 read failed\n");
+ ret = sx1301_field_write(priv, F_FORCE_HOST_RADIO_CTRL, 0);
+ if (ret)
return ret;
- }
-
- val &= ~(REG_0_105_FORCE_HOST_RADIO_CTRL | REG_0_105_FORCE_HOST_FE_CTRL | REG_0_105_FORCE_DEC_FILTER_GAIN);
-
- ret = regmap_write(priv->regmap, SX1301_FORCE_CTRL, val);
- if (ret) {
- dev_err(priv->dev, "0|105 write failed\n");
+ ret = sx1301_field_write(priv, F_FORCE_HOST_FE_CTRL, 0);
+ if (ret)
+ return ret;
+ ret = sx1301_field_write(priv, F_FORCE_DEC_FILTER_GAIN, 0);
+ if (ret)
return ret;
- }
ret = regmap_write(priv->regmap, SX1301_CHRS, 0);
if (ret) {
@@ -376,17 +309,15 @@ static int sx1301_load_all_firmware(struct sx1301_priv *priv)
return ret;
}
- ret = regmap_read(priv->regmap, SX1301_MCU_CTRL, &val);
+ ret = sx1301_field_write(priv, F_MCU_RST_0, 0);
if (ret) {
- dev_err(priv->dev, "MCU read (0) failed\n");
+ dev_err(priv->dev, "MCU 0 release failed\n");
return ret;
}
- val &= ~(REG_0_MCU_RST_1 | REG_0_MCU_RST_0);
-
- ret = regmap_write(priv->regmap, SX1301_MCU_CTRL, val);
+ ret = sx1301_field_write(priv, F_MCU_RST_1, 0);
if (ret) {
- dev_err(priv->dev, "MCU write (0) failed\n");
+ dev_err(priv->dev, "MCU 1 release failed\n");
return ret;
}
@@ -438,7 +369,6 @@ static netdev_tx_t sx130x_loradev_start_xmit(struct sk_buff *skb, struct net_dev
static int sx130x_loradev_open(struct net_device *netdev)
{
struct sx1301_priv *priv = netdev_priv(netdev);
- unsigned int val;
int ret;
netdev_dbg(netdev, "%s", __func__);
@@ -448,31 +378,15 @@ static int sx130x_loradev_open(struct net_device *netdev)
return -ENXIO;
}
- ret = regmap_read(priv->regmap, SX1301_GEN, &val);
- if (ret) {
- netdev_err(netdev, "16 read (1) failed\n");
- return ret;
- }
-
- val |= REG_16_GLOBAL_EN;
-
- ret = regmap_write(priv->regmap, SX1301_GEN, val);
- if (ret) {
- netdev_err(netdev, "16 write (1) failed\n");
- return ret;
- }
-
- ret = regmap_read(priv->regmap, SX1301_CKEN, &val);
+ ret = sx1301_field_write(priv, F_GLOBAL_EN, 1);
if (ret) {
- netdev_err(netdev, "17 read (1) failed\n");
+ dev_err(priv->dev, "enable global clocks failed\n");
return ret;
}
- val |= REG_17_CLK32M_EN;
-
- ret = regmap_write(priv->regmap, SX1301_CKEN, val);
+ ret = sx1301_field_write(priv, F_CLK32M_EN, 1);
if (ret) {
- netdev_err(netdev, "17 write (1) failed\n");
+ dev_err(priv->dev, "enable 32M clock failed\n");
return ret;
}
@@ -519,6 +433,7 @@ static int sx1301_probe(struct spi_device *spi)
struct sx1301_priv *priv;
struct gpio_desc *rst;
int ret;
+ int i;
unsigned int ver;
unsigned int val;
@@ -557,6 +472,19 @@ static int sx1301_probe(struct spi_device *spi)
return ret;
}
+ for (i = 0; i < ARRAY_SIZE(sx1301_regmap_fields); i++) {
+ const struct reg_field *reg_fields = sx1301_regmap_fields;
+
+ priv->regmap_fields[i] = devm_regmap_field_alloc(&spi->dev,
+ priv->regmap,
+ reg_fields[i]);
+ if (IS_ERR(priv->regmap_fields[i])) {
+ ret = PTR_ERR(priv->regmap_fields[i]);
+ dev_err(&spi->dev, "Cannot allocate regmap field: %d\n", ret);
+ return ret;
+ }
+ }
+
ret = regmap_read(priv->regmap, SX1301_VER, &ver);
if (ret) {
dev_err(&spi->dev, "version read failed\n");
@@ -574,83 +502,49 @@ static int sx1301_probe(struct spi_device *spi)
return ret;
}
- ret = sx1301_soft_reset(priv);
+ ret = sx1301_field_write(priv, F_SOFT_RESET, 1);
if (ret) {
dev_err(&spi->dev, "soft reset failed\n");
return ret;
}
- ret = regmap_read(priv->regmap, SX1301_GEN, &val);
+ ret = sx1301_field_write(priv, F_GLOBAL_EN, 0);
if (ret) {
- dev_err(&spi->dev, "16 read failed\n");
+ dev_err(&spi->dev, "gate global clocks failed\n");
return ret;
}
- val &= ~REG_16_GLOBAL_EN;
-
- ret = regmap_write(priv->regmap, SX1301_GEN, val);
+ ret = sx1301_field_write(priv, F_CLK32M_EN, 0);
if (ret) {
- dev_err(&spi->dev, "16 write failed\n");
+ dev_err(&spi->dev, "gate 32M clock failed\n");
return ret;
}
- ret = regmap_read(priv->regmap, SX1301_CKEN, &val);
+ ret = sx1301_field_write(priv, F_RADIO_A_EN, 1);
if (ret) {
- dev_err(&spi->dev, "17 read failed\n");
+ dev_err(&spi->dev, "radio a enable failed\n");
return ret;
}
- val &= ~REG_17_CLK32M_EN;
-
- ret = regmap_write(priv->regmap, SX1301_CKEN, val);
+ ret = sx1301_field_write(priv, F_RADIO_B_EN, 1);
if (ret) {
- dev_err(&spi->dev, "17 write failed\n");
- return ret;
- }
-
- ret = regmap_read(priv->regmap, SX1301_RADIO_CFG, &val);
- if (ret) {
- dev_err(&spi->dev, "2|43 read failed\n");
- return ret;
- }
-
- val |= REG_2_43_RADIO_B_EN | REG_2_43_RADIO_A_EN;
-
- ret = regmap_write(priv->regmap, SX1301_RADIO_CFG, val);
- if (ret) {
- dev_err(&spi->dev, "2|43 write failed\n");
+ dev_err(&spi->dev, "radio b enable failed\n");
return ret;
}
msleep(500);
- ret = regmap_read(priv->regmap, SX1301_RADIO_CFG, &val);
+ ret = sx1301_field_write(priv, F_RADIO_RST, 1);
if (ret) {
- dev_err(&spi->dev, "2|43 read failed\n");
- return ret;
- }
-
- val |= REG_2_43_RADIO_RST;
-
- ret = regmap_write(priv->regmap, SX1301_RADIO_CFG, val);
- if (ret) {
- dev_err(&spi->dev, "2|43 write failed\n");
+ dev_err(&spi->dev, "radio asert reset failed\n");
return ret;
}
msleep(5);
- ret = regmap_read(priv->regmap, SX1301_RADIO_CFG, &val);
- if (ret) {
- dev_err(&spi->dev, "2|43 read failed\n");
- return ret;
- }
-
- val &= ~REG_2_43_RADIO_RST;
-
- ret = regmap_write(priv->regmap, SX1301_RADIO_CFG, val);
+ ret = sx1301_field_write(priv, F_RADIO_RST, 0);
if (ret) {
- dev_err(&spi->dev, "2|43 write failed\n");
+ dev_err(&spi->dev, "radio deasert reset failed\n");
return ret;
}
diff --git a/drivers/net/lora/sx1301.h b/drivers/net/lora/sx1301.h
index e6400f8..0bbd948 100644
--- a/drivers/net/lora/sx1301.h
+++ b/drivers/net/lora/sx1301.h
@@ -60,11 +60,57 @@
#define SX1301_MAX_REGISTER (SX1301_PAGE_BASE(3) + 0x7F)
+enum sx1301_fields {
+ F_SOFT_RESET,
+ F_GLOBAL_EN,
+ F_CLK32M_EN,
+ F_RADIO_A_EN,
+ F_RADIO_B_EN,
+ F_RADIO_RST,
+
+ F_MCU_RST_0,
+ F_MCU_RST_1,
+ F_MCU_SELECT_MUX_0,
+ F_MCU_SELECT_MUX_1,
+
+ F_FORCE_HOST_RADIO_CTRL,
+ F_FORCE_HOST_FE_CTRL,
+ F_FORCE_DEC_FILTER_GAIN,
+
+ F_EMERGENCY_FORCE_HOST_CTRL,
+};
+
+static const struct reg_field sx1301_regmap_fields[] = {
+ /* PAGE */
+ [F_SOFT_RESET] = REG_FIELD(SX1301_PAGE, 7, 7),
+ /* GEN */
+ [F_GLOBAL_EN] = REG_FIELD(SX1301_GEN, 3, 3),
+ /* CKEN */
+ [F_CLK32M_EN] = REG_FIELD(SX1301_CKEN, 0, 0),
+ /* RADIO_CFG */
+ [F_RADIO_A_EN] = REG_FIELD(SX1301_RADIO_CFG, 0, 0),
+ [F_RADIO_B_EN] = REG_FIELD(SX1301_RADIO_CFG, 1, 1),
+ [F_RADIO_RST] = REG_FIELD(SX1301_RADIO_CFG, 2, 2),
+ /* MCU_CTRL */
+ [F_MCU_RST_0] = REG_FIELD(SX1301_MCU_CTRL, 0, 0),
+ [F_MCU_RST_1] = REG_FIELD(SX1301_MCU_CTRL, 1, 1),
+ [F_MCU_SELECT_MUX_0] = REG_FIELD(SX1301_MCU_CTRL, 2, 2),
+ [F_MCU_SELECT_MUX_1] = REG_FIELD(SX1301_MCU_CTRL, 3, 3),
+ /* FORCE_CTRL */
+ [F_FORCE_HOST_RADIO_CTRL] = REG_FIELD(SX1301_FORCE_CTRL, 1, 1),
+ [F_FORCE_HOST_FE_CTRL] = REG_FIELD(SX1301_FORCE_CTRL, 2, 2),
+ [F_FORCE_DEC_FILTER_GAIN] = REG_FIELD(SX1301_FORCE_CTRL, 3, 3),
+ /* EMERGENCY_FORCE_HOST_CTRL */
+ [F_EMERGENCY_FORCE_HOST_CTRL] =
+ REG_FIELD(SX1301_EMERGENCY_FORCE_HOST_CTRL, 0, 0),
+};
+
struct sx1301_priv {
struct lora_dev_priv lora;
struct device *dev;
struct gpio_desc *rst_gpio;
struct regmap *regmap;
+ struct regmap_field *regmap_fields[ARRAY_SIZE(sx1301_regmap_fields)];
};
int __init sx130x_radio_init(void);
--
2.7.4
^ permalink raw reply related
* [PATCH v3 lora-next 0/5] based on 4.19-rc7 migration to regmap and clk framewor
From: Ben Whitten @ 2018-10-12 16:26 UTC (permalink / raw)
To: afaerber; +Cc: starnight, hasnain.virk, netdev, liuxuenetmail, shess,
Ben Whitten
This series is designed to be applied to a lora-next which is based on
v4.19-rc7, not what is currently upstream, there were no additional changes
in this rebase.
In this series we add the write variant to regmap_noinc then use it within
the sx1301 driver to load firmware.
We complete the sx1301 migration to regmap by replacing the SPI burst read
and write with regmap_noinc_read and write.
The various fields of the sx1301 are converted to regmap to pass the read
modify write down to the regmap layer and reduce our code.
The downside is that in the cases of the hand optimised multiple field writes
at once within one register, however if it is desirable these can be replaced
with regmap_update_bits.
Apply the same regmap treatment for the radio device and allow it to register as
a clock provider, this clock is enabled when the lora link is brought up as it
is required prior to calibration of the concentrator.
modprobe lora-sx125x
lora-dev: init
sx1301 spi0.0: SX1301 module probed
sx125x_con spi0.0-a: SX125x version: 21
sx125x_con spi0.0-a: SX125x module probed
sx125x_con spi0.0-b: SX125x version: 21
sx125x_con spi0.0-b: registering clkout
sx125x_con spi0.0-b: SX125x module probed
ip link set lora0 up
sx125x_con spi0.0-b: enabling clkout
Ben Whitten (5):
regmap: Add regmap_noinc_write API
net: lora: sx1301: replace burst spi functions with regmap_noinc
net: lora: sx1301: convert to using regmap fields for bit ops
net: lora: sx125x: convert to regmap fields
net: lora: sx125x sx1301: allow radio to register as a clk provider
drivers/base/regmap/internal.h | 3 +
drivers/base/regmap/regmap.c | 77 ++++++++++++
drivers/net/lora/sx125x.c | 159 +++++++++++++++++++++++--
drivers/net/lora/sx1301.c | 257 ++++++++++++-----------------------------
drivers/net/lora/sx1301.h | 50 +++++++-
include/linux/regmap.h | 19 +++
6 files changed, 369 insertions(+), 196 deletions(-)
--
2.7.4
^ permalink raw reply
* Re: [PATCH net-next] net: bridge: add support for per-port vlan stats
From: Nikolay Aleksandrov @ 2018-10-12 16:04 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev, davem, bridge, Roopa Prabhu
In-Reply-To: <20181012083932.1753c93e@xeon-e3>
On October 12, 2018 6:39:32 PM GMT+03:00, Stephen Hemminger <stephen@networkplumber.org> wrote:
>On Fri, 12 Oct 2018 13:41:16 +0300
>Nikolay Aleksandrov <nikolay@cumulusnetworks.com> wrote:
>
>> This patch adds an option to have per-port vlan stats instead of the
>> default global stats. The option can be set only when there are no
>port
>> vlans in the bridge since we need to allocate the stats if it is set
>> when vlans are being added to ports (and respectively free them
>> when being deleted). Also bump RTNL_MAX_TYPE as the bridge is the
>> largest user of options. The current stats design allows us to add
>> these without any changes to the fast-path, it all comes down to
>> the per-vlan stats pointer which, if this option is enabled, will
>> be allocated for each port vlan instead of using the global
>bridge-wide
>> one.
>>
>> CC: bridge@lists.linux-foundation.org
>> CC: Roopa Prabhu <roopa@cumulusnetworks.com>
>> Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
>
>Yes, per-vlan stats could be quite useful.
>Most cases of statistics in the kernel are always on, and some API's
>get them (and skip others). Other than the additional memory overhead,
>why
>not make the statistics as always on.
>
The additional overhead can be quite a lot and it also affects performance, besides that currently users expect the same accumulated stats.
>Also, is there any chance of creating too much data in a netlink
>message if there are 4K-1 VLAN's?
No, that's okay because we expose the stats via the xstats API which
continues the dump on filling the current buffer.
We can export as much info as needed.
Cheers,
Nik
^ permalink raw reply
* Re: [PATCH net] ipv6: rate-limit probes for neighbourless routes
From: Sabrina Dubroca @ 2018-10-12 16:03 UTC (permalink / raw)
To: Eric Dumazet; +Cc: netdev, Stefano Brivio
In-Reply-To: <ba734a42-a509-e1b7-141f-d5f4ea252039@gmail.com>
2018-10-12, 08:17:28 -0700, Eric Dumazet wrote:
>
>
> On 10/12/2018 07:22 AM, Sabrina Dubroca wrote:
> > When commit 270972554c91 ("[IPV6]: ROUTE: Add Router Reachability
> > Probing (RFC4191).") introduced router probing, the rt6_probe() function
> > required that a neighbour entry existed. This neighbour entry is used to
> > record the timestamp of the last probe via the ->updated field.
> >
> > Later, commit 2152caea7196 ("ipv6: Do not depend on rt->n in rt6_probe().")
> > removed the requirement for a neighbour entry. Neighbourless routes skip
> > the interval check and are not rate-limited.
> >
>
> Interesting. Can you describe a little more what happens here ?
The machine gets two default routes, one valid, one via a gateway that
doesn't exist. Every time a packet is sent via the default route, we
probe the dead gateway, which doesn't have a neighbour entry. Since
there's no rate-limiting, we send too many neighbour solicitations,
disobeying the RFC.
> Could this explain gazillions of syzbot/syzkaller reports that have all in common :
That seems completely unrelated. That particular trace is about router
solicitations, not neighbour solicitations. This seems more like an
RCU issue.
> rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> Call Trace:
> rcu: (detected by 0, t=10712 jiffies, g=90369, q=135)
> <IRQ>
> __dump_stack lib/dump_stack.c:77 [inline]
> dump_stack+0x1c4/0x2b4 lib/dump_stack.c:113
> rcu: All QSes seen, last rcu_preempt kthread activity 10548 (4295003843-4294993295), jiffies_till_next_fqs=1, root ->qsmask 0x0
> syz-executor0 R
> running task
> warn_alloc.cold.119+0xb7/0x1bd mm/page_alloc.c:3426
> 22896 7592 5826 0x8010000c
> Call Trace:
> <IRQ>
> sched_show_task.cold.83+0x2b6/0x30a kernel/sched/core.c:5296
> __alloc_pages_slowpath+0x2667/0x2d80 mm/page_alloc.c:4297
> print_other_cpu_stall.cold.79+0xa83/0xba5 kernel/rcu/tree.c:1430
> check_cpu_stall kernel/rcu/tree.c:1557 [inline]
> __rcu_pending kernel/rcu/tree.c:3276 [inline]
> rcu_pending kernel/rcu/tree.c:3319 [inline]
> rcu_check_callbacks+0xafc/0x1990 kernel/rcu/tree.c:2665
> __alloc_pages_nodemask+0xa80/0xde0 mm/page_alloc.c:4390
> __alloc_pages include/linux/gfp.h:473 [inline]
> __alloc_pages_node include/linux/gfp.h:486 [inline]
> kmem_getpages mm/slab.c:1409 [inline]
> cache_grow_begin+0x91/0x8c0 mm/slab.c:2677
> fallback_alloc+0x203/0x2e0 mm/slab.c:3219
> ____cache_alloc_node+0x1c7/0x1e0 mm/slab.c:3287
> slab_alloc_node mm/slab.c:3327 [inline]
> kmem_cache_alloc_node+0xe3/0x730 mm/slab.c:3642
> __alloc_skb+0x119/0x770 net/core/skbuff.c:193
> alloc_skb include/linux/skbuff.h:997 [inline]
> ndisc_alloc_skb+0x144/0x340 net/ipv6/ndisc.c:403
> ndisc_send_rs+0x331/0x6e0 net/ipv6/ndisc.c:669
> update_process_times+0x2d/0x70 kernel/time/timer.c:1636
> addrconf_rs_timer+0x314/0x690 net/ipv6/addrconf.c:3836
> tick_sched_handle+0x9f/0x180 kernel/time/tick-sched.c:164
> tick_sched_timer+0x45/0x130 kernel/time/tick-sched.c:1274
> __run_hrtimer kernel/time/hrtimer.c:1398 [inline]
> __hrtimer_run_queues+0x41c/0x10d0 kernel/time/hrtimer.c:1460
> call_timer_fn+0x272/0x920 kernel/time/timer.c:1326
> hrtimer_interrupt+0x313/0x780 kernel/time/hrtimer.c:1518
--
Sabrina
^ permalink raw reply
* Re: [PATCH iproute2 net-next] bridge: add support for backup port
From: Nikolay Aleksandrov @ 2018-10-12 16:01 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev, roopa, dsahern
In-Reply-To: <20181012084028.5a64747f@xeon-e3>
On October 12, 2018 6:40:28 PM GMT+03:00, Stephen Hemminger <stephen@networkplumber.org> wrote:
>On Fri, 12 Oct 2018 14:42:55 +0300
>Nikolay Aleksandrov <nikolay@cumulusnetworks.com> wrote:
>
>> This patch adds support for the new backup port option that can be
>set
>> on a bridge port. If the port's carrier goes down all of the traffic
>> gets redirected to the configured backup port. We add the following
>new
>> arguments:
>> $ ip link set dev brport type bridge_slave backup_port brport2
>> $ ip link set dev brport type bridge_slave nobackup_port
>>
>> $ bridge link set dev brport backup_port brport2
>> $ bridge link set dev brport nobackup_port
>>
>> The man pages are updated respectively.
>> Also 2 minor style adjustments:
>> - add missing space to bridge man page's state argument
>> - use lower starting case for vlan_tunnel in ip-link man page (to be
>> consistent with the rest)
>>
>> Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
>
>This seems a bit like feature creep in the bridge.
>Why not use team or bonding?
Not really, the bond/team cannot work in such way. We did discuss it when the kernel patches were sent, for more information please check
https://www.spinics.net/lists/netdev/msg514642.html
Thanks,
Nik
^ permalink raw reply
* Re: [PATCH net] ip6_tunnel: Don't update PMTU on tunnels with collect_md
From: Nicolas Dichtel @ 2018-10-12 15:58 UTC (permalink / raw)
To: Stefano Brivio, David S. Miller; +Cc: Alexei Starovoitov, netdev
In-Reply-To: <5893664877d77075e32b33f72a7784f539555da0.1539266541.git.sbrivio@redhat.com>
Le 12/10/2018 à 14:32, Stefano Brivio a écrit :
> Commit 8d79266bc48c ("ip6_tunnel: add collect_md mode to IPv6
> tunnels") introduced a check to avoid updating PMTU when
> collect_md mode is enabled.
>
> Later, commit f15ca723c1eb ("net: don't call update_pmtu
> unconditionally") dropped this check, I guess inadvertently.
I removed it because update_pmtu() is not set for md_dst_op, thus I assume this
check was done for that purpose.
Could you explain in your commit log which problem your patch fixes?
^ permalink raw reply
* Re: [PATCH iproute2 net-next] bridge: add support for backup port
From: Stephen Hemminger @ 2018-10-12 15:40 UTC (permalink / raw)
To: Nikolay Aleksandrov; +Cc: netdev, roopa, dsahern
In-Reply-To: <20181012114255.17217-1-nikolay@cumulusnetworks.com>
On Fri, 12 Oct 2018 14:42:55 +0300
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> wrote:
> This patch adds support for the new backup port option that can be set
> on a bridge port. If the port's carrier goes down all of the traffic
> gets redirected to the configured backup port. We add the following new
> arguments:
> $ ip link set dev brport type bridge_slave backup_port brport2
> $ ip link set dev brport type bridge_slave nobackup_port
>
> $ bridge link set dev brport backup_port brport2
> $ bridge link set dev brport nobackup_port
>
> The man pages are updated respectively.
> Also 2 minor style adjustments:
> - add missing space to bridge man page's state argument
> - use lower starting case for vlan_tunnel in ip-link man page (to be
> consistent with the rest)
>
> Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
This seems a bit like feature creep in the bridge.
Why not use team or bonding?
^ permalink raw reply
* Re: [PATCH bpf-next 2/2] bpf, libbpf: simplify perf RB walk and do incremental updates
From: Jakub Kicinski @ 2018-10-12 15:39 UTC (permalink / raw)
To: Daniel Borkmann; +Cc: alexei.starovoitov, netdev
In-Reply-To: <531acdb6-aef3-fe19-64aa-4255daab9cc1@iogearbox.net>
On Fri, 12 Oct 2018 15:30:57 +0200, Daniel Borkmann wrote:
> On 10/12/2018 10:39 AM, Daniel Borkmann wrote:
> > On 10/12/2018 05:04 AM, Jakub Kicinski wrote:
> >> On Thu, 11 Oct 2018 16:02:07 +0200, Daniel Borkmann wrote:
> >>> Clean up and improve bpf_perf_event_read_simple() ring walk a bit
> >>> to use similar tail update scheme as in perf and bcc allowing the
> >>> kernel to make forward progress not only after full timely walk.
> >>
> >> The extra memory barriers won't impact performance? If I read the code
> >> correctly we now have:
> >>
> >> while (bla) {
> >> head = HEAD
> >> rmb()
> >>
> >> ...
> >>
> >> mb()
> >> TAIL = tail
> >> }
> >>
> >> Would it make sense to try to piggy back on the mb() for head re-read
> >> at least? Perhaps that's a non-issue, just wondering.
> >
> > From the scheme specified in the comment in prior patch my understanding
> > would be that they don't pair (see B and C) so there would be no guarantee
> > that load of head would happen before load of data. Fwiw, I've been using
> > the exact same semantics as user space perf tool walks the perf mmap'ed
> > ring buffer (tools/perf/util/mmap.{h,c}) here. Given kernel doesn't stop
>
> On that note, I'll also respin, after some clarification with PeterZ on
> why perf is using {rmb,mb}() barriers today as opposed to more lightweight
> smp_{rmb,mb}() ones it turns out there is no real reason other than
> historic one and perf can be changed and made more efficient as well. ;)
Cool, to be clear I meant something like:
head = HEAD
rmb()
while (bla) {
...
head = HEAD
mb()
TAIL = tail
}
> > pushing into ring buffer while user space walks it and indicates how far
> > it has consumed data via tail update, it would allow for making room
> > successively and not only after full run has complete, so we don't make
> > any assumptions in the generic libbpf library helper on how slow/quick
> > the callback would be processing resp. how full ring is, etc, and kernel
> > pushing new data can be processed in the same run if necessary. One thing
> > we could consider is to batch tail updates, say, every 8 elements and a
> > final update once we break out walking the ring; probably okay'ish as a
> > heuristic..
Makes sense, I would be tempted to do some batching, but since I'm
mostly working with I/O devices I may think that memory barriers cost
more than they do for main memory-only accesses :)
> >>> Also few other improvements to use realloc() instead of free() and
> >>> malloc() combination and for the callback use proper perf_event_header
> >>> instead of void pointer, so that real applications can use container_of()
> >>> macro with proper type checking.
> >>
> >> FWIW the free() + malloc() was to avoid the the needless copy of the
> >> previous event realloc() may do. It makes sense to use realloc()
> >> especially if you want to put extra info in front of the buffer, just
> >> sayin' it wasn't a complete braino ;)
> >
> > No strong preference from my side, I'd think that it might be sensible in
> > any case from applications to call the bpf_perf_event_read_simple() with a
> > already preallocated buffer, depending on the expected max element size from
> > BPF could e.g. be a buffer of 1 page or so. Given 512 byte stack space from
> > the BPF prog and MTU 1500 this would more than suffice to avoid new
> > allocations altogether. Anyway, given we only grow the new memory area I
> > did some testing on realloc() with an array of pointers to prior malloc()'ed
> > buffers, running randomly 10M realloc()s to increase size over them and
> > saw <1% where area had to be moved, so we're hitting corner case of a corner
> > case, I'm also ok to leave the combination, though. :)
Thanks for doing the experiment, not strong preference here either, so
realloc() seems good!
^ permalink raw reply
* Re: [PATCH net-next] net: bridge: add support for per-port vlan stats
From: Stephen Hemminger @ 2018-10-12 15:39 UTC (permalink / raw)
To: Nikolay Aleksandrov; +Cc: netdev, davem, bridge, Roopa Prabhu
In-Reply-To: <20181012104116.12751-1-nikolay@cumulusnetworks.com>
On Fri, 12 Oct 2018 13:41:16 +0300
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> wrote:
> This patch adds an option to have per-port vlan stats instead of the
> default global stats. The option can be set only when there are no port
> vlans in the bridge since we need to allocate the stats if it is set
> when vlans are being added to ports (and respectively free them
> when being deleted). Also bump RTNL_MAX_TYPE as the bridge is the
> largest user of options. The current stats design allows us to add
> these without any changes to the fast-path, it all comes down to
> the per-vlan stats pointer which, if this option is enabled, will
> be allocated for each port vlan instead of using the global bridge-wide
> one.
>
> CC: bridge@lists.linux-foundation.org
> CC: Roopa Prabhu <roopa@cumulusnetworks.com>
> Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Yes, per-vlan stats could be quite useful.
Most cases of statistics in the kernel are always on, and some API's
get them (and skip others). Other than the additional memory overhead, why
not make the statistics as always on.
Also, is there any chance of creating too much data in a netlink
message if there are 4K-1 VLAN's?
^ permalink raw reply
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