* Re: [PATCH v3 07/14] net: smc911x: remove the dmaengine compat need
From: Robert Jarzmik @ 2018-06-20 17:17 UTC (permalink / raw)
To: David S. Miller; +Cc: Daniel Mack, linux-kernel, netdev
In-Reply-To: <20180617170217.24177-8-robert.jarzmik@free.fr>
Hi David,
I have converted all the pxa related drivers to DMA slave maps, the only 2 that
remain in my queue are touching your tree, ie. smc911x and smc91x (see [1]).
Once these 2 are done, I have my last dependency on the dma filter function
removed, and can queue the remaining cleanup patches.
Could you (or somebody from netdev) review it and either ack it (and I'll take
it through the pxa tree), or take it for v4.19 please ?
Cheers.
--
Robert
[1] Submission
Robert Jarzmik <robert.jarzmik@free.fr> writes:
> As the pxa architecture switched towards the dmaengine slave map, the
> old compatibility mechanism to acquire the dma requestor line number and
> priority are not needed anymore.
>
> This patch simplifies the dma resource acquisition, using the more
> generic function dma_request_slave_channel().
>
> Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
> ---
> Since v2: converted to NULL filter function and NULL dma parameter call
> ---
> drivers/net/ethernet/smsc/smc911x.c | 13 ++-----------
> 1 file changed, 2 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/net/ethernet/smsc/smc911x.c b/drivers/net/ethernet/smsc/smc911x.c
> index 05157442a980..b1b53f6c452f 100644
> --- a/drivers/net/ethernet/smsc/smc911x.c
> +++ b/drivers/net/ethernet/smsc/smc911x.c
> @@ -74,7 +74,6 @@ static const char version[] =
> #include <linux/skbuff.h>
>
> #include <linux/dmaengine.h>
> -#include <linux/dma/pxa-dma.h>
>
> #include <asm/io.h>
>
> @@ -1795,7 +1794,6 @@ static int smc911x_probe(struct net_device *dev)
> #ifdef SMC_USE_DMA
> struct dma_slave_config config;
> dma_cap_mask_t mask;
> - struct pxad_param param;
> #endif
>
> DBG(SMC_DEBUG_FUNC, dev, "--> %s\n", __func__);
> @@ -1971,15 +1969,8 @@ static int smc911x_probe(struct net_device *dev)
>
> dma_cap_zero(mask);
> dma_cap_set(DMA_SLAVE, mask);
> - param.prio = PXAD_PRIO_LOWEST;
> - param.drcmr = -1UL;
> -
> - lp->rxdma =
> - dma_request_slave_channel_compat(mask, pxad_filter_fn,
> - ¶m, &dev->dev, "rx");
> - lp->txdma =
> - dma_request_slave_channel_compat(mask, pxad_filter_fn,
> - ¶m, &dev->dev, "tx");
> + lp->rxdma = dma_request_channel(mask, NULL, NULL);
> + lp->txdma = dma_request_channel(mask, NULL, NULL);
> lp->rxdma_active = 0;
> lp->txdma_active = 0;
--
Robert
^ permalink raw reply
* Re: [PATCH v3 08/14] net: smc91x: remove the dmaengine compat need
From: Robert Jarzmik @ 2018-06-20 17:17 UTC (permalink / raw)
To: David S. Miller; +Cc: Daniel Mack, linux-kernel, netdev
In-Reply-To: <20180617170217.24177-9-robert.jarzmik@free.fr>
Hi David,
I have converted all the pxa related drivers to DMA slave maps, the only 2 that
remain in my queue are touching your tree, ie. smc911x and smc91x (see [1]).
Once these 2 are done, I have my last dependency on the dma filter function
removed, and can queue the remaining cleanup patches.
Could you (or somebody from netdev) review it and either ack it (and I'll take
it through the pxa tree), or take it for v4.19 please ?
Cheers.
--
Robert
[1] Submission
Robert Jarzmik <robert.jarzmik@free.fr> writes:
> As the pxa architecture switched towards the dmaengine slave map, the
> old compatibility mechanism to acquire the dma requestor line number and
> priority are not needed anymore.
>
> This patch simplifies the dma resource acquisition, using the more
> generic function dma_request_slave_channel().
>
> Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
> ---
> Since v2: converted to NULL filter function and NULL dma parameter call
> ---
> drivers/net/ethernet/smsc/smc91x.c | 9 +--------
> drivers/net/ethernet/smsc/smc91x.h | 1 -
> 2 files changed, 1 insertion(+), 9 deletions(-)
>
> diff --git a/drivers/net/ethernet/smsc/smc91x.c b/drivers/net/ethernet/smsc/smc91x.c
> index 080428762858..b944828f9ea3 100644
> --- a/drivers/net/ethernet/smsc/smc91x.c
> +++ b/drivers/net/ethernet/smsc/smc91x.c
> @@ -2019,17 +2019,10 @@ static int smc_probe(struct net_device *dev, void __iomem *ioaddr,
> # endif
> if (lp->cfg.flags & SMC91X_USE_DMA) {
> dma_cap_mask_t mask;
> - struct pxad_param param;
>
> dma_cap_zero(mask);
> dma_cap_set(DMA_SLAVE, mask);
> - param.prio = PXAD_PRIO_LOWEST;
> - param.drcmr = -1UL;
> -
> - lp->dma_chan =
> - dma_request_slave_channel_compat(mask, pxad_filter_fn,
> - ¶m, &dev->dev,
> - "data");
> + lp->dma_chan = dma_request_channel(mask, NULL, NULL);
> }
> #endif
>
> diff --git a/drivers/net/ethernet/smsc/smc91x.h b/drivers/net/ethernet/smsc/smc91x.h
> index b337ee97e0c0..a27352229fc2 100644
> --- a/drivers/net/ethernet/smsc/smc91x.h
> +++ b/drivers/net/ethernet/smsc/smc91x.h
> @@ -301,7 +301,6 @@ struct smc_local {
> * as RX which can overrun memory and lose packets.
> */
> #include <linux/dma-mapping.h>
> -#include <linux/dma/pxa-dma.h>
>
> #ifdef SMC_insl
> #undef SMC_insl
--
Robert
^ permalink raw reply
* Re: [PATCH iproute2-next v3] ip-xfrm: Add support for OUTPUT_MARK
From: David Ahern @ 2018-06-20 17:18 UTC (permalink / raw)
To: Subash Abhinov Kasiviswanathan, lorenzo, netdev, stephen,
steffen.klassert
In-Reply-To: <1529116375-7578-1-git-send-email-subashab@codeaurora.org>
On 6/15/18 8:32 PM, Subash Abhinov Kasiviswanathan wrote:
> This patch adds support for OUTPUT_MARK in xfrm state to exercise the
> functionality added by kernel commit 077fbac405bf
> ("net: xfrm: support setting an output mark.").
>
...
> v1->v2: Moved the XFRMA_OUTPUT_MARK print after XFRMA_MARK in
> xfrm_xfrma_print() as mentioned by Lorenzo
>
> v2->v3: Fix one help formatting error as mentioned by Lorenzo.
> Keep mark and output-mark on the same line and add man page info as
> mentioned by David.
>
> Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
> ---
> ip/ipxfrm.c | 17 ++++++++++++++++-
> ip/xfrm_state.c | 9 +++++++++
> man/man8/ip-xfrm.8 | 11 +++++++++++
> 3 files changed, 36 insertions(+), 1 deletion(-)
>
applied to iproute2-next. Thanks
^ permalink raw reply
* Re: [PATCH iproute2-next 1/1] tc: jsonify nat action
From: David Ahern @ 2018-06-20 17:24 UTC (permalink / raw)
To: Keara Leibovitz, dsahern; +Cc: stephen, netdev, kernel
In-Reply-To: <1529348269-30039-1-git-send-email-kleib@mojatatu.com>
On 6/18/18 12:57 PM, Keara Leibovitz wrote:
> Add json output support for nat action
>
...
>
> Signed-off-by: Keara Leibovitz <kleib@mojatatu.com>
> ---
> tc/m_nat.c | 32 +++++++++++++++++++-------------
> 1 file changed, 19 insertions(+), 13 deletions(-)
>
applied to iproute2-next. Thanks
^ permalink raw reply
* Re: [PATCH net][RESEND] strparser: Don't schedule in workqueue in paused state
From: Dave Watson @ 2018-06-20 17:32 UTC (permalink / raw)
To: Vakul Garg
Cc: davem, doronrk, tom, john.fastabend, netdev, ebiggers,
linux-kernel
In-Reply-To: <20180620215949.32334-1-vakul.garg@nxp.com>
On 06/21/18 03:29 AM, Vakul Garg wrote:
> In function strp_data_ready(), it is useless to call queue_work if
> the state of strparser is already paused. The state checking should
> be done before calling queue_work. The change reduces the context
> switches and improves the ktls-rx throughput by approx 20% (measured
> on cortex-a53 based platform).
Nice find, works for me. Although current code isn't broken, just
slower, net-next would be fine.
Acked-by: Dave Watson <davejwatson@fb.com>
^ permalink raw reply
* [PATCH net] cls_flower: fix use after free in flower S/W path
From: Paolo Abeni @ 2018-06-20 17:34 UTC (permalink / raw)
To: netdev
Cc: Jamal Hadi Salim, Cong Wang, Jiri Pirko, Marcelo Ricardo Leitner,
Paul Blakey
If flower filter is created without the skip_sw flag, fl_mask_put()
can race with fl_classify() and we can destroy the mask rhashtable
while a lookup operation is accessing it.
BUG: unable to handle kernel paging request at 00000000000911d1
PGD 0 P4D 0
SMP PTI
CPU: 3 PID: 5582 Comm: vhost-5541 Not tainted 4.18.0-rc1.vanilla+ #1950
Hardware name: Dell Inc. PowerEdge R730/072T6D, BIOS 2.1.7 06/16/2016
RIP: 0010:rht_bucket_nested+0x20/0x60
Code: 31 c8 c1 c1 18 29 c8 c3 66 90 8b 4f 04 ba 01 00 00 00 8b 07 48 8b bf 80 00 00 0
RSP: 0018:ffffafc5cfbb7a48 EFLAGS: 00010206
RAX: 0000000000001978 RBX: ffff9f12dff88a00 RCX: 00000000ffff9f12
RDX: 00000000000911d1 RSI: 0000000000000148 RDI: 0000000000000001
RBP: ffff9f12dff88a00 R08: 000000005f1cc119 R09: 00000000a715fae2
R10: ffffafc5cfbb7aa8 R11: ffff9f1cb4be804e R12: ffff9f1265e13000
R13: 0000000000000000 R14: ffffafc5cfbb7b48 R15: ffff9f12dff88b68
FS: 0000000000000000(0000) GS:ffff9f1d3f0c0000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000000911d1 CR3: 0000001575a94006 CR4: 00000000001626e0
Call Trace:
fl_lookup+0x134/0x140 [cls_flower]
fl_classify+0xf3/0x180 [cls_flower]
tcf_classify+0x78/0x150
__netif_receive_skb_core+0x69e/0xa50
netif_receive_skb_internal+0x42/0xf0
tun_get_user+0xdd5/0xfd0 [tun]
tun_sendmsg+0x52/0x70 [tun]
handle_tx+0x2b3/0x5f0 [vhost_net]
vhost_worker+0xab/0x100 [vhost]
kthread+0xf8/0x130
ret_from_fork+0x35/0x40
Modules linked in: act_mirred act_gact cls_flower vhost_net vhost tap sch_ingress
CR2: 00000000000911d1
Fix the above waiting for a RCU grace period before destroying the
rhashtable.
Fixes: 05cd271fd61a ("cls_flower: Support multiple masks per priority")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
net/sched/cls_flower.c | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/net/sched/cls_flower.c b/net/sched/cls_flower.c
index 2b5be42a9f1c..5e7333c6472d 100644
--- a/net/sched/cls_flower.c
+++ b/net/sched/cls_flower.c
@@ -203,6 +203,17 @@ static int fl_init(struct tcf_proto *tp)
return rhashtable_init(&head->ht, &mask_ht_params);
}
+static void fl_mask_free(struct fl_flow_mask *mask)
+{
+ rhashtable_destroy(&mask->ht);
+ kfree(mask);
+}
+
+static void fl_mask_free_rcu(struct rcu_head *entry)
+{
+ fl_mask_free(container_of(entry, struct fl_flow_mask, rcu));
+}
+
static bool fl_mask_put(struct cls_fl_head *head, struct fl_flow_mask *mask,
bool async)
{
@@ -210,12 +221,11 @@ static bool fl_mask_put(struct cls_fl_head *head, struct fl_flow_mask *mask,
return false;
rhashtable_remove_fast(&head->ht, &mask->ht_node, mask_ht_params);
- rhashtable_destroy(&mask->ht);
list_del_rcu(&mask->list);
if (async)
- kfree_rcu(mask, rcu);
+ call_rcu(&mask->rcu, fl_mask_free_rcu);
else
- kfree(mask);
+ fl_mask_free(mask);
return true;
}
--
2.17.1
^ permalink raw reply related
* Re: [PATCH net] ipvlan: call dev_change_flags when reset ipvlan mode
From: Cong Wang @ 2018-06-20 17:45 UTC (permalink / raw)
To: David Miller
Cc: Hangbin Liu, Linux Kernel Network Developers, Stefano Brivio,
Paolo Abeni, Mahesh Bandewar
In-Reply-To: <20180620.143147.2291173423483856091.davem@davemloft.net>
On Tue, Jun 19, 2018 at 10:31 PM, David Miller <davem@davemloft.net> wrote:
> From: Hangbin Liu <liuhangbin@gmail.com>
> Date: Wed, 20 Jun 2018 11:22:54 +0800
>
>> The only case dev_change_flags() return an err is when we change IFF_UP flag.
>> Since we only set/reset IFF_NOARP, do you think we still need to check the
>> return value?
>
> It is bad to try and take shortcuts on error handling using assumptions
> like that.
>
> If dev_change_flags() is adjusted to return error codes in more
> situations, nobody is going to remember to undo your "optimziation"
> here.
>
> Please check for errors, thank you.
Yeah. Also since the notifier is triggered in this case:
if (dev->flags & IFF_UP &&
(changes & ~(IFF_UP | IFF_PROMISC | IFF_ALLMULTI | IFF_VOLATILE))) {
struct netdev_notifier_change_info change_info = {
.info = {
.dev = dev,
},
.flags_changed = changes,
};
call_netdevice_notifiers_info(NETDEV_CHANGE, &change_info.info);
}
the return value of call_netdevice_notifiers_info() isn't captured
either, but it should be.
^ permalink raw reply
* Re: [RFC v2, net-next, PATCH 4/4] net/cpsw_switchdev: add switchdev mode of operation on cpsw driver
From: Ilias Apalodimas @ 2018-06-20 17:51 UTC (permalink / raw)
To: Ivan Vecera
Cc: Andrew Lunn, netdev, grygorii.strashko, ivan.khoronzhuk, nsekhar,
jiri, f.fainelli, francois.ozog, yogeshs, spatton, Jose.Abreu
In-Reply-To: <671c83ea-6227-cc09-e604-14cfa6804726@redhat.com>
Hello Ivan,
On Wed, Jun 20, 2018 at 02:56:48PM +0200, Ivan Vecera wrote:
> On 18.6.2018 22:19, Ilias Apalodimas wrote:
> > Jiri proposed using devlink, which makes sense, but i am not sure it's
> > applicable on this patchset. This will change the driver completely and will
> > totally break backwards compatibility.
>
> Another good reason for a new driver.
>
> I.
This is actually conflicting at least to my understanding. Jiri proposed using
devlink was used as an alternative method to enable a new mode instead of
adding it on a .config option. A new driver wouldn't have a need for that right?
Thanks
Ilias
^ permalink raw reply
* Re: [RFC v2, net-next, PATCH 4/4] net/cpsw_switchdev: add switchdev mode of operation on cpsw driver
From: Andrew Lunn @ 2018-06-20 17:57 UTC (permalink / raw)
To: Ilias Apalodimas
Cc: Ivan Vecera, netdev, grygorii.strashko, ivan.khoronzhuk, nsekhar,
jiri, f.fainelli, francois.ozog, yogeshs, spatton, Jose.Abreu
In-Reply-To: <20180620175128.GA27235@apalos>
> > Another good reason for a new driver.
> >
> > I.
> This is actually conflicting at least to my understanding. Jiri proposed using
> devlink was used as an alternative method to enable a new mode instead of
> adding it on a .config option. A new driver wouldn't have a need for that right?
A new driver would only do switchdev. So correct, there would not be
any configuration need. It would also give you a chance to have a new
device tree binding, if that helps at all.
Andrew
^ permalink raw reply
* Re: [RFC v2, net-next, PATCH 4/4] net/cpsw_switchdev: add switchdev mode of operation on cpsw driver
From: Florian Fainelli @ 2018-06-20 17:58 UTC (permalink / raw)
To: Ilias Apalodimas, Ivan Vecera
Cc: Andrew Lunn, netdev, grygorii.strashko, ivan.khoronzhuk, nsekhar,
jiri, francois.ozog, yogeshs, spatton, Jose.Abreu
In-Reply-To: <20180620175128.GA27235@apalos>
On 06/20/2018 10:51 AM, Ilias Apalodimas wrote:
> Hello Ivan,
> On Wed, Jun 20, 2018 at 02:56:48PM +0200, Ivan Vecera wrote:
>> On 18.6.2018 22:19, Ilias Apalodimas wrote:
>>> Jiri proposed using devlink, which makes sense, but i am not sure it's
>>> applicable on this patchset. This will change the driver completely and will
>>> totally break backwards compatibility.
>>
>> Another good reason for a new driver.
>>
>> I.
> This is actually conflicting at least to my understanding. Jiri proposed using
> devlink was used as an alternative method to enable a new mode instead of
> adding it on a .config option. A new driver wouldn't have a need for that right?
Correct, with a new driver would likely behave correctly upon being
probed such that you could have your switch ports act as normal network
devices from which you could run IP-config and do NFS boot.
--
Florian
^ permalink raw reply
* Re: [RFC v2, net-next, PATCH 4/4] net/cpsw_switchdev: add switchdev mode of operation on cpsw driver
From: Ilias Apalodimas @ 2018-06-20 18:03 UTC (permalink / raw)
To: Florian Fainelli
Cc: Ivan Vecera, Andrew Lunn, netdev, grygorii.strashko,
ivan.khoronzhuk, nsekhar, jiri, francois.ozog, yogeshs, spatton,
Jose.Abreu
In-Reply-To: <edad81cb-fb30-fede-1629-b32295979e95@gmail.com>
Hi Florian,
On Wed, Jun 20, 2018 at 10:58:26AM -0700, Florian Fainelli wrote:
> On 06/20/2018 10:51 AM, Ilias Apalodimas wrote:
> > Hello Ivan,
> > On Wed, Jun 20, 2018 at 02:56:48PM +0200, Ivan Vecera wrote:
> >> On 18.6.2018 22:19, Ilias Apalodimas wrote:
> >>> Jiri proposed using devlink, which makes sense, but i am not sure it's
> >>> applicable on this patchset. This will change the driver completely and will
> >>> totally break backwards compatibility.
> >>
> >> Another good reason for a new driver.
> >>
> >> I.
> > This is actually conflicting at least to my understanding. Jiri proposed using
> > devlink was used as an alternative method to enable a new mode instead of
> > adding it on a .config option. A new driver wouldn't have a need for that right?
>
> Correct, with a new driver would likely behave correctly upon being
> probed such that you could have your switch ports act as normal network
> devices from which you could run IP-config and do NFS boot.
The current driver also does NFS properly and the 2 ethernet ports act as normal
network interfaces. The NFS section in the cover letter
is to cover the cases were users running on NFS need to change the running
switch configuration(starting from adding the 2 interfaces on a bridge).
Since iproute2 is located on the NFS filesystem the moment
network connectivity is lost, you loose the ability to perform further
configuration and in certian configuration scenarios render the device
unusable.
> --
> Florian
Thanks
Ilias
^ permalink raw reply
* Re: [PATCH net] cls_flower: fix use after free in flower S/W path
From: Cong Wang @ 2018-06-20 18:06 UTC (permalink / raw)
To: Paolo Abeni
Cc: Linux Kernel Network Developers, Jamal Hadi Salim, Jiri Pirko,
Marcelo Ricardo Leitner, Paul Blakey
In-Reply-To: <e3a9263c79766595ee701b3d66c885ea3d2febe6.1529515725.git.pabeni@redhat.com>
On Wed, Jun 20, 2018 at 10:34 AM, Paolo Abeni <pabeni@redhat.com> wrote:
>
> +static void fl_mask_free(struct fl_flow_mask *mask)
> +{
> + rhashtable_destroy(&mask->ht);
I don't believe you can call rhashtable_destroy() in BH
context, it acquires a mutex...
^ permalink raw reply
* Re: [PATCH v4 net-next] net:sched: add action inheritdsfield to skbedit
From: Marcelo Ricardo Leitner @ 2018-06-20 18:40 UTC (permalink / raw)
To: Davide Caratti
Cc: Michel Machado, Fu, Qiaobin, davem@davemloft.net,
netdev@vger.kernel.org, jhs@mojatatu.com,
xiyou.wangcong@gmail.com
In-Reply-To: <a1716a3accbbf9a72bbedb011862ac57eaddf481.camel@redhat.com>
On Wed, Jun 20, 2018 at 07:02:41PM +0200, Davide Caratti wrote:
...
> > I agree that we should update the drop counter, but given that
> > you're already converting the stats to be per-cpu counters, whatever we
> > add now will be just symbolic since you're going to change it anyway.
It wouldn't be symbolic. One thing is to convert a given increment
into something else, another is to start increasing it for some (new)
reason.
>
> that's ok for me also, as I can use the current v4 code for the rebase
> (and not wait for another respin) _ but let's hear what reviewers think.
>
> > If
> > reviewers think that Qiaobin's patch must add the update line, could you
> > provide the exact line and location so we avoid going to v6 of this patch?
>
> In case, I was thinking of something like:
>
> https://elixir.bootlin.com/linux/v4.18-rc1/source/net/sched/act_ipt.c#L249
>
> so, between 'err:' and 'spin_unlock(&d->tcf_lock)', insert a line like:
>
> d->tcf_qstats.drop++;
I prefer the more complete version. Then it will have a more complete
(git) history and help people when troubleshooting.
Thanks,
Marcelo
^ permalink raw reply
* [PATCH bpf 0/2] tools: bpftool: small fixes for error handling in prog load
From: Jakub Kicinski @ 2018-06-20 18:42 UTC (permalink / raw)
To: alexei.starovoitov, daniel; +Cc: netdev, oss-drivers, Jakub Kicinski
Hi!
Two small fixes for error handling in bpftool prog load, first patch
removes a duplicated message, second one frees resources correctly.
Multiple error messages break JSON:
# bpftool -jp prog load tracex1_kern.o /sys/fs/bpf/a
{
"error": "can't pin the object (/sys/fs/bpf/a): File exists"
},{
"error": "failed to pin program"
}
Jakub Kicinski (2):
tools: bpftool: remove duplicated error message on prog load
tools: bpftool: remember to close the libbpf object on prog load error
tools/bpf/bpftool/prog.c | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
--
2.17.1
^ permalink raw reply
* [PATCH bpf 1/2] tools: bpftool: remove duplicated error message on prog load
From: Jakub Kicinski @ 2018-06-20 18:42 UTC (permalink / raw)
To: alexei.starovoitov, daniel; +Cc: netdev, oss-drivers, Jakub Kicinski
In-Reply-To: <20180620184246.18672-1-jakub.kicinski@netronome.com>
do_pin_fd() will already print out an error message if something
goes wrong. Printing another error is unnecessary and will break
JSON output, since error messages are full objects:
$ bpftool -jp prog load tracex1_kern.o /sys/fs/bpf/a
{
"error": "can't pin the object (/sys/fs/bpf/a): File exists"
},{
"error": "failed to pin program"
}
Fixes: 49a086c201a9 ("bpftool: implement prog load command")
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
---
tools/bpf/bpftool/prog.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c
index 05f42a46d6ed..12b694fe0404 100644
--- a/tools/bpf/bpftool/prog.c
+++ b/tools/bpf/bpftool/prog.c
@@ -694,10 +694,8 @@ static int do_load(int argc, char **argv)
return -1;
}
- if (do_pin_fd(prog_fd, argv[1])) {
- p_err("failed to pin program");
+ if (do_pin_fd(prog_fd, argv[1]))
return -1;
- }
if (json_output)
jsonw_null(json_wtr);
--
2.17.1
^ permalink raw reply related
* [PATCH bpf 2/2] tools: bpftool: remember to close the libbpf object after prog load
From: Jakub Kicinski @ 2018-06-20 18:42 UTC (permalink / raw)
To: alexei.starovoitov, daniel; +Cc: netdev, oss-drivers, Jakub Kicinski
In-Reply-To: <20180620184246.18672-1-jakub.kicinski@netronome.com>
Remembering to close all descriptors and free memory may not seem
important in a user space tool like bpftool, but if we were to run
in batch mode the consumed resources start to add up quickly. Make
sure program load closes the libbpf object (which unloads and frees
it).
Fixes: 49a086c201a9 ("bpftool: implement prog load command")
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
---
tools/bpf/bpftool/prog.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c
index 12b694fe0404..959aa53ab678 100644
--- a/tools/bpf/bpftool/prog.c
+++ b/tools/bpf/bpftool/prog.c
@@ -695,12 +695,18 @@ static int do_load(int argc, char **argv)
}
if (do_pin_fd(prog_fd, argv[1]))
- return -1;
+ goto err_close_obj;
if (json_output)
jsonw_null(json_wtr);
+ bpf_object__close(obj);
+
return 0;
+
+err_close_obj:
+ bpf_object__close(obj);
+ return -1;
}
static int do_help(int argc, char **argv)
--
2.17.1
^ permalink raw reply related
* Re: [PATCH] r8169: Fix netpoll oops
From: Heiner Kallweit @ 2018-06-20 18:51 UTC (permalink / raw)
To: Ville Syrjala, netdev; +Cc: Realtek linux nic maintainers, David S . Miller
In-Reply-To: <20180620120153.11676-1-ville.syrjala@linux.intel.com>
On 20.06.2018 14:01, Ville Syrjala wrote:
> From: Ville Syrjälä <ville.syrjala@linux.intel.com>
>
> Pass the correct thing to rtl8169_interrupt() from netpoll.
>
Indeed, this place I missed to change.
Only small remark: The patch should be annotated as "net" so
that it can go to 4.18.
> Cc: Realtek linux nic maintainers <nic_swsd@realtek.com>
> Cc: netdev@vger.kernel.org
> Cc: Heiner Kallweit <hkallweit1@gmail.com>
> Cc: David S. Miller <davem@davemloft.net>
> Fixes: ebcd5daa7ffd ("r8169: change interrupt handler argument type")
> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> ---
> drivers/net/ethernet/realtek/r8169.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/realtek/r8169.c b/drivers/net/ethernet/realtek/r8169.c
> index 75dfac0248f4..f4cae2be0fda 100644
> --- a/drivers/net/ethernet/realtek/r8169.c
> +++ b/drivers/net/ethernet/realtek/r8169.c
> @@ -7148,7 +7148,7 @@ static void rtl8169_netpoll(struct net_device *dev)
> {
> struct rtl8169_private *tp = netdev_priv(dev);
>
> - rtl8169_interrupt(pci_irq_vector(tp->pci_dev, 0), dev);
> + rtl8169_interrupt(pci_irq_vector(tp->pci_dev, 0), tp);
> }
> #endif
>
>
^ permalink raw reply
* Re: Route fallback issue
From: Julian Anastasov @ 2018-06-20 19:00 UTC (permalink / raw)
To: Akshat Kakkar; +Cc: netdev, cronolog+lartc, lartc, Erik Auerswald
In-Reply-To: <CAA5aLPiPBTEp4eMs9T6xA3dkOPry3H=oD2svo3Ybon3dEgh3gA@mail.gmail.com>
Hello,
On Wed, 20 Jun 2018, Akshat Kakkar wrote:
> Hi netdev community,
>
> I have 2 interfaces
> eno1 : 192.168.1.10/24
> eno2 : 192.168.2.10/24
>
> I added routes as
> 172.16.0.0/12 via 192.168.1.254 metric 1
> 172.16.0.0/12 via 192.168.2.254 metric 2
>
> My intention : All traffic to 172.16.0.0/12 should go thru eno1. If
> 192.168.1.254 is not reachable (no arp entry or link down), then it
> should fall back to eno2.
You can also try alternative routes. But as the
kernel supports only default alternative routes, you can
put them in their own table:
# Alternative routes use same metric!!!
ip route append default via 192.168.1.254 dev eno1 table 100
ip route append default via 192.168.2.254 dev eno2 table 100
ip rule add prio 100 to 172.16.0.0/12 table 100
Of course, you will get better results if an user space
tool puts only alive routes in service after doing health
checks of all near gateways.
Regards
^ permalink raw reply
* Re: [PATCH net-next 0/2] fixes for ipsec selftests
From: Anders Roxell @ 2018-06-20 19:09 UTC (permalink / raw)
To: shannon.nelson; +Cc: Networking, David Miller
In-Reply-To: <1529473363-4036-1-git-send-email-shannon.nelson@oracle.com>
On Wed, 20 Jun 2018 at 07:42, Shannon Nelson <shannon.nelson@oracle.com> wrote:
>
> A couple of bad behaviors in the ipsec selftest were pointed out
> by Anders Roxell <anders.roxell@linaro.org> and are addressed here.
>
> Shannon Nelson (2):
> selftests: rtnetlink: hide complaint from terminated monitor
> selftests: rtnetlink: use a local IP address for IPsec tests
>
> tools/testing/selftests/net/rtnetlink.sh | 11 +++++++----
> 1 file changed, 7 insertions(+), 4 deletions(-)
>
> --
> 2.7.4
>
Hi Shannon,
With this patches applied and my config patch.
I still get this error when I run the ipsec test:
FAIL: can't add fou port 7777, skipping test
RTNETLINK answers: Operation not supported
FAIL: can't add macsec interface, skipping test
RTNETLINK answers: Protocol not supported
RTNETLINK answers: No such process
RTNETLINK answers: No such process
FAIL: ipsec
Can you please cc the kselftest list when sending patches to
tools/testing/selftests/ ?
Cheers,
Anders
^ permalink raw reply
* net/usb: Use irqsave in USB's complete callback
From: Sebastian Andrzej Siewior @ 2018-06-20 19:31 UTC (permalink / raw)
To: netdev; +Cc: linux-usb, tglx, David S. Miller
This is about using _irqsave() primitives in the completion callback in
order to get rid of local_irq_save() in __usb_hcd_giveback_urb().
Sebastian
^ permalink raw reply
* [PATCH 2/5] net: usb: hso: use irqsave() in USB's complete callback
From: Sebastian Andrzej Siewior @ 2018-06-20 19:31 UTC (permalink / raw)
To: netdev; +Cc: linux-usb, tglx, David S. Miller, Sebastian Andrzej Siewior
In-Reply-To: <20180620193121.24967-1-bigeasy@linutronix.de>
The USB completion callback does not disable interrupts while acquiring
the lock. We want to remove the local_irq_disable() invocation from
__usb_hcd_giveback_urb() and therefore it is required for the callback
handler to disable the interrupts while acquiring the lock.
The callback may be invoked either in IRQ or BH context depending on the
USB host controller.
Use the _irqsave() variant of the locking primitives.
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
drivers/net/usb/hso.c | 35 +++++++++++++++++++++--------------
1 file changed, 21 insertions(+), 14 deletions(-)
diff --git a/drivers/net/usb/hso.c b/drivers/net/usb/hso.c
index e53883ad6107..de305ead32e6 100644
--- a/drivers/net/usb/hso.c
+++ b/drivers/net/usb/hso.c
@@ -999,6 +999,7 @@ static void read_bulk_callback(struct urb *urb)
struct hso_net *odev = urb->context;
struct net_device *net;
int result;
+ unsigned long flags;
int status = urb->status;
/* is al ok? (Filip: Who's Al ?) */
@@ -1028,11 +1029,11 @@ static void read_bulk_callback(struct urb *urb)
if (urb->actual_length) {
/* Handle the IP stream, add header and push it onto network
* stack if the packet is complete. */
- spin_lock(&odev->net_lock);
+ spin_lock_irqsave(&odev->net_lock, flags);
packetizeRx(odev, urb->transfer_buffer, urb->actual_length,
(urb->transfer_buffer_length >
urb->actual_length) ? 1 : 0);
- spin_unlock(&odev->net_lock);
+ spin_unlock_irqrestore(&odev->net_lock, flags);
}
/* We are done with this URB, resubmit it. Prep the USB to wait for
@@ -1193,6 +1194,7 @@ static void hso_std_serial_read_bulk_callback(struct urb *urb)
{
struct hso_serial *serial = urb->context;
int status = urb->status;
+ unsigned long flags;
hso_dbg(0x8, "--- Got serial_read_bulk callback %02x ---\n", status);
@@ -1216,10 +1218,10 @@ static void hso_std_serial_read_bulk_callback(struct urb *urb)
if (serial->parent->port_spec & HSO_INFO_CRC_BUG)
fix_crc_bug(urb, serial->in_endp->wMaxPacketSize);
/* Valid data, handle RX data */
- spin_lock(&serial->serial_lock);
+ spin_lock_irqsave(&serial->serial_lock, flags);
serial->rx_urb_filled[hso_urb_to_index(serial, urb)] = 1;
put_rxbuf_data_and_resubmit_bulk_urb(serial);
- spin_unlock(&serial->serial_lock);
+ spin_unlock_irqrestore(&serial->serial_lock, flags);
}
/*
@@ -1502,12 +1504,13 @@ static void tiocmget_intr_callback(struct urb *urb)
DUMP(serial_state_notification,
sizeof(struct hso_serial_state_notification));
} else {
+ unsigned long flags;
UART_state_bitmap = le16_to_cpu(serial_state_notification->
UART_state_bitmap);
prev_UART_state_bitmap = tiocmget->prev_UART_state_bitmap;
icount = &tiocmget->icount;
- spin_lock(&serial->serial_lock);
+ spin_lock_irqsave(&serial->serial_lock, flags);
if ((UART_state_bitmap & B_OVERRUN) !=
(prev_UART_state_bitmap & B_OVERRUN))
icount->parity++;
@@ -1530,7 +1533,7 @@ static void tiocmget_intr_callback(struct urb *urb)
(prev_UART_state_bitmap & B_RX_CARRIER))
icount->dcd++;
tiocmget->prev_UART_state_bitmap = UART_state_bitmap;
- spin_unlock(&serial->serial_lock);
+ spin_unlock_irqrestore(&serial->serial_lock, flags);
tiocmget->intr_completed = 1;
wake_up_interruptible(&tiocmget->waitq);
}
@@ -1852,6 +1855,7 @@ static void intr_callback(struct urb *urb)
struct hso_serial *serial;
unsigned char *port_req;
int status = urb->status;
+ unsigned long flags;
int i;
usb_mark_last_busy(urb->dev);
@@ -1879,7 +1883,7 @@ static void intr_callback(struct urb *urb)
if (serial != NULL) {
hso_dbg(0x1, "Pending read interrupt on port %d\n",
i);
- spin_lock(&serial->serial_lock);
+ spin_lock_irqsave(&serial->serial_lock, flags);
if (serial->rx_state == RX_IDLE &&
serial->port.count > 0) {
/* Setup and send a ctrl req read on
@@ -1893,7 +1897,8 @@ static void intr_callback(struct urb *urb)
hso_dbg(0x1, "Already a read pending on port %d or port not open\n",
i);
}
- spin_unlock(&serial->serial_lock);
+ spin_unlock_irqrestore(&serial->serial_lock,
+ flags);
}
}
}
@@ -1920,6 +1925,7 @@ static void hso_std_serial_write_bulk_callback(struct urb *urb)
{
struct hso_serial *serial = urb->context;
int status = urb->status;
+ unsigned long flags;
/* sanity check */
if (!serial) {
@@ -1927,9 +1933,9 @@ static void hso_std_serial_write_bulk_callback(struct urb *urb)
return;
}
- spin_lock(&serial->serial_lock);
+ spin_lock_irqsave(&serial->serial_lock, flags);
serial->tx_urb_used = 0;
- spin_unlock(&serial->serial_lock);
+ spin_unlock_irqrestore(&serial->serial_lock, flags);
if (status) {
handle_usb_error(status, __func__, serial->parent);
return;
@@ -1971,14 +1977,15 @@ static void ctrl_callback(struct urb *urb)
struct hso_serial *serial = urb->context;
struct usb_ctrlrequest *req;
int status = urb->status;
+ unsigned long flags;
/* sanity check */
if (!serial)
return;
- spin_lock(&serial->serial_lock);
+ spin_lock_irqsave(&serial->serial_lock, flags);
serial->tx_urb_used = 0;
- spin_unlock(&serial->serial_lock);
+ spin_unlock_irqrestore(&serial->serial_lock, flags);
if (status) {
handle_usb_error(status, __func__, serial->parent);
return;
@@ -1994,9 +2001,9 @@ static void ctrl_callback(struct urb *urb)
(USB_DIR_IN | USB_TYPE_OPTION_VENDOR | USB_RECIP_INTERFACE)) {
/* response to a read command */
serial->rx_urb_filled[0] = 1;
- spin_lock(&serial->serial_lock);
+ spin_lock_irqsave(&serial->serial_lock, flags);
put_rxbuf_data_and_resubmit_ctrl_urb(serial);
- spin_unlock(&serial->serial_lock);
+ spin_unlock_irqrestore(&serial->serial_lock, flags);
} else {
hso_put_activity(serial->parent);
tty_port_tty_wakeup(&serial->port);
--
2.17.1
^ permalink raw reply related
* [PATCH 3/5] net: usb: kaweth: use irqsave() in USB's complete callback
From: Sebastian Andrzej Siewior @ 2018-06-20 19:31 UTC (permalink / raw)
To: netdev; +Cc: linux-usb, tglx, David S. Miller, Sebastian Andrzej Siewior
In-Reply-To: <20180620193121.24967-1-bigeasy@linutronix.de>
The USB completion callback does not disable interrupts while acquiring
the lock. We want to remove the local_irq_disable() invocation from
__usb_hcd_giveback_urb() and therefore it is required for the callback
handler to disable the interrupts while acquiring the lock.
The callback may be invoked either in IRQ or BH context depending on the
USB host controller.
Use the _irqsave() variant of the locking primitives.
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
drivers/net/usb/kaweth.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/usb/kaweth.c b/drivers/net/usb/kaweth.c
index f1605833c5cf..913e50bab0a2 100644
--- a/drivers/net/usb/kaweth.c
+++ b/drivers/net/usb/kaweth.c
@@ -587,7 +587,7 @@ static void kaweth_usb_receive(struct urb *urb)
struct kaweth_device *kaweth = urb->context;
struct net_device *net = kaweth->net;
int status = urb->status;
-
+ unsigned long flags;
int count = urb->actual_length;
int count2 = urb->transfer_buffer_length;
@@ -619,12 +619,12 @@ static void kaweth_usb_receive(struct urb *urb)
net->stats.rx_errors++;
dev_dbg(dev, "Status was -EOVERFLOW.\n");
}
- spin_lock(&kaweth->device_lock);
+ spin_lock_irqsave(&kaweth->device_lock, flags);
if (IS_BLOCKED(kaweth->status)) {
- spin_unlock(&kaweth->device_lock);
+ spin_unlock_irqrestore(&kaweth->device_lock, flags);
return;
}
- spin_unlock(&kaweth->device_lock);
+ spin_unlock_irqrestore(&kaweth->device_lock, flags);
if(status && status != -EREMOTEIO && count != 1) {
dev_err(&kaweth->intf->dev,
--
2.17.1
^ permalink raw reply related
* [PATCH 4/5] net: usb: r8152: use irqsave() in USB's complete callback
From: Sebastian Andrzej Siewior @ 2018-06-20 19:31 UTC (permalink / raw)
To: netdev; +Cc: linux-usb, tglx, David S. Miller, Sebastian Andrzej Siewior
In-Reply-To: <20180620193121.24967-1-bigeasy@linutronix.de>
The USB completion callback does not disable interrupts while acquiring
the lock. We want to remove the local_irq_disable() invocation from
__usb_hcd_giveback_urb() and therefore it is required for the callback
handler to disable the interrupts while acquiring the lock.
The callback may be invoked either in IRQ or BH context depending on the
USB host controller.
Use the _irqsave() variant of the locking primitives.
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
drivers/net/usb/r8152.c | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c
index 86f7196f9d91..c08c0d633407 100644
--- a/drivers/net/usb/r8152.c
+++ b/drivers/net/usb/r8152.c
@@ -1252,6 +1252,7 @@ static void read_bulk_callback(struct urb *urb)
int status = urb->status;
struct rx_agg *agg;
struct r8152 *tp;
+ unsigned long flags;
agg = urb->context;
if (!agg)
@@ -1281,9 +1282,9 @@ static void read_bulk_callback(struct urb *urb)
if (urb->actual_length < ETH_ZLEN)
break;
- spin_lock(&tp->rx_lock);
+ spin_lock_irqsave(&tp->rx_lock, flags);
list_add_tail(&agg->list, &tp->rx_done);
- spin_unlock(&tp->rx_lock);
+ spin_unlock_irqrestore(&tp->rx_lock, flags);
napi_schedule(&tp->napi);
return;
case -ESHUTDOWN:
@@ -1311,6 +1312,7 @@ static void write_bulk_callback(struct urb *urb)
struct net_device *netdev;
struct tx_agg *agg;
struct r8152 *tp;
+ unsigned long flags;
int status = urb->status;
agg = urb->context;
@@ -1332,9 +1334,9 @@ static void write_bulk_callback(struct urb *urb)
stats->tx_bytes += agg->skb_len;
}
- spin_lock(&tp->tx_lock);
+ spin_lock_irqsave(&tp->tx_lock, flags);
list_add_tail(&agg->list, &tp->tx_free);
- spin_unlock(&tp->tx_lock);
+ spin_unlock_irqrestore(&tp->tx_lock, flags);
usb_autopm_put_interface_async(tp->intf);
--
2.17.1
^ permalink raw reply related
* [PATCH 1/5] net: usb: cdc-phonet: use irqsave() in USB's complete callback
From: Sebastian Andrzej Siewior @ 2018-06-20 19:31 UTC (permalink / raw)
To: netdev; +Cc: linux-usb, tglx, David S. Miller, Sebastian Andrzej Siewior
In-Reply-To: <20180620193121.24967-1-bigeasy@linutronix.de>
The USB completion callback does not disable interrupts while acquiring
the lock. We want to remove the local_irq_disable() invocation from
__usb_hcd_giveback_urb() and therefore it is required for the callback
handler to disable the interrupts while acquiring the lock.
The callback may be invoked either in IRQ or BH context depending on the
USB host controller.
Use the _irqsave() variant of the locking primitives.
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
drivers/net/usb/cdc-phonet.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/net/usb/cdc-phonet.c b/drivers/net/usb/cdc-phonet.c
index 288ecd999171..3c40312fa453 100644
--- a/drivers/net/usb/cdc-phonet.c
+++ b/drivers/net/usb/cdc-phonet.c
@@ -99,6 +99,7 @@ static void tx_complete(struct urb *req)
struct net_device *dev = skb->dev;
struct usbpn_dev *pnd = netdev_priv(dev);
int status = req->status;
+ unsigned long flags;
switch (status) {
case 0:
@@ -115,10 +116,10 @@ static void tx_complete(struct urb *req)
}
dev->stats.tx_packets++;
- spin_lock(&pnd->tx_lock);
+ spin_lock_irqsave(&pnd->tx_lock, flags);
pnd->tx_queue--;
netif_wake_queue(dev);
- spin_unlock(&pnd->tx_lock);
+ spin_unlock_irqrestore(&pnd->tx_lock, flags);
dev_kfree_skb_any(skb);
usb_free_urb(req);
--
2.17.1
^ permalink raw reply related
* [PATCH 5/5] net: usb: rtl8150: use irqsave() in USB's complete callback
From: Sebastian Andrzej Siewior @ 2018-06-20 19:31 UTC (permalink / raw)
To: netdev
Cc: linux-usb, tglx, David S. Miller, Sebastian Andrzej Siewior,
Petko Manolov
In-Reply-To: <20180620193121.24967-1-bigeasy@linutronix.de>
The USB completion callback does not disable interrupts while acquiring
the lock. We want to remove the local_irq_disable() invocation from
__usb_hcd_giveback_urb() and therefore it is required for the callback
handler to disable the interrupts while acquiring the lock.
The callback may be invoked either in IRQ or BH context depending on the
USB host controller.
Use the _irqsave() variant of the locking primitives.
Cc: Petko Manolov <petkan@nucleusys.com>
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
drivers/net/usb/rtl8150.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/net/usb/rtl8150.c b/drivers/net/usb/rtl8150.c
index 5f565bd574da..0e81d4c441d9 100644
--- a/drivers/net/usb/rtl8150.c
+++ b/drivers/net/usb/rtl8150.c
@@ -391,6 +391,7 @@ static void read_bulk_callback(struct urb *urb)
u16 rx_stat;
int status = urb->status;
int result;
+ unsigned long flags;
dev = urb->context;
if (!dev)
@@ -432,9 +433,9 @@ static void read_bulk_callback(struct urb *urb)
netdev->stats.rx_packets++;
netdev->stats.rx_bytes += pkt_len;
- spin_lock(&dev->rx_pool_lock);
+ spin_lock_irqsave(&dev->rx_pool_lock, flags);
skb = pull_skb(dev);
- spin_unlock(&dev->rx_pool_lock);
+ spin_unlock_irqrestore(&dev->rx_pool_lock, flags);
if (!skb)
goto resched;
--
2.17.1
^ 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