* Re: [PATCH net-next] net: dsa: add port fdb dump
From: Florian Fainelli @ 2017-09-21 5:09 UTC (permalink / raw)
To: Vivien Didelot, netdev; +Cc: linux-kernel, kernel, David S. Miller, Andrew Lunn
In-Reply-To: <20170920233214.6936-1-vivien.didelot@savoirfairelinux.com>
On 09/20/2017 04:32 PM, Vivien Didelot wrote:
> Dumping a DSA port's FDB entries is not specific to a DSA slave, so add
> a dsa_port_fdb_dump function, similarly to dsa_port_fdb_add and
> dsa_port_fdb_del.
>
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
--
Florian
^ permalink raw reply
* Re: [PATCH net-next] net: dsa: use dedicated CPU port
From: Florian Fainelli @ 2017-09-21 5:10 UTC (permalink / raw)
To: Vivien Didelot, netdev; +Cc: linux-kernel, kernel, David S. Miller, Andrew Lunn
In-Reply-To: <20170920162805.29503-1-vivien.didelot@savoirfairelinux.com>
On 09/20/2017 09:28 AM, Vivien Didelot wrote:
> Each port in DSA has its own dedicated CPU port currently available in
> its parent switch's ds->ports[port].cpu_dp. Use it instead of getting
> the unique tree CPU port, which will be deprecated soon.
>
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
> ---
> drivers/net/dsa/b53/b53_common.c | 4 ++--
> drivers/net/dsa/bcm_sf2.c | 6 +++---
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
--
Florian
^ permalink raw reply
* Re:Re: [PATCH net-next ] net: Remove useless function skb_header_release
From: Gao Feng @ 2017-09-21 5:07 UTC (permalink / raw)
To: Joe Perches; +Cc: davem, netdev
In-Reply-To: <1505969350.12311.20.camel@perches.com>
>On Thu, 2017-09-21 at 12:39 +0800, gfree.wind@vip.163.com wrote:
>> From: Gao Feng <gfree.wind@vip.163.com>
>>
>> There is no one which would invokes the function skb_header_release.
>> So just remove it now.
>
>This in incomplete.
>There are other references to this function.
>
>$ git grep -w skb_header_release
>drivers/net/usb/asix_common.c: * TCP packets for example are cloned, but skb_header_release()
>include/linux/skbuff.h: * skb_header_release - release reference to header
>include/linux/skbuff.h:static inline void skb_header_release(struct sk_buff *skb)
>include/linux/skbuff.h: * Variant of skb_header_release() assuming skb is private to caller.
>net/batman-adv/soft-interface.c: * data using skb_header_release in our skbs to allow skb_cow_header to
But there are comments, not real codes.
Unless the skb_header_release may be used in the external modules
Best Regards
Feng
^ permalink raw reply
* Re: [PATCH net] bpf: one perf event close won't free bpf program attached by another perf event
From: Yonghong Song @ 2017-09-21 5:17 UTC (permalink / raw)
To: Steven Rostedt; +Cc: peterz, ast, daniel, netdev, kernel-team
In-Reply-To: <20170920214109.11002b7c@vmware.local.home>
On 9/20/17 6:41 PM, Steven Rostedt wrote:
> On Mon, 18 Sep 2017 16:38:36 -0700
> Yonghong Song <yhs@fb.com> wrote:
>
>> This patch fixes a bug exhibited by the following scenario:
>> 1. fd1 = perf_event_open with attr.config = ID1
>> 2. attach bpf program prog1 to fd1
>> 3. fd2 = perf_event_open with attr.config = ID1
>> <this will be successful>
>> 4. user program closes fd2 and prog1 is detached from the tracepoint.
>> 5. user program with fd1 does not work properly as tracepoint
>> no output any more.
>>
>> The issue happens at step 4. Multiple perf_event_open can be called
>> successfully, but only one bpf prog pointer in the tp_event. In the
>> current logic, any fd release for the same tp_event will free
>> the tp_event->prog.
>>
>> The fix is to free tp_event->prog only when the closing fd
>> corresponds to the one which registered the program.
>>
>> Signed-off-by: Yonghong Song <yhs@fb.com>
>> ---
>> Additional context: discussed with Alexei internally but did not find
>> a solution which can avoid introducing the additional field in
>> trace_event_call structure.
>>
>> Peter, could you take a look as well and maybe you could have better
>> alternative? Thanks!
>>
>> include/linux/trace_events.h | 1 +
>> kernel/events/core.c | 3 ++-
>> 2 files changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/include/linux/trace_events.h b/include/linux/trace_events.h
>> index 7f11050..2e0f222 100644
>> --- a/include/linux/trace_events.h
>> +++ b/include/linux/trace_events.h
>> @@ -272,6 +272,7 @@ struct trace_event_call {
>> int perf_refcount;
>> struct hlist_head __percpu *perf_events;
>> struct bpf_prog *prog;
>> + struct perf_event *bpf_prog_owner;
>
> Does this have to be in the trace_event_call structure? Hmm, I'm
> wondering if the prog needs to be there (I should look to see if we can
> move it from it). The trace_event_call is created for *every* event,
> and there's thousands of them now. Every byte to this structure adds
> 1000s of bytes to the kernel. Would it be possible to attach the prog
> and the owner to the perf_event?
Regarding whether we could move the prog and the owner to the
perf_event. It is possible. There is already a "prog" field in the
perf_event structure for overflow handler. We could reuse the "prog"
field and we do not need bpf_prog_owner any more. We can iterate
through trace_event_call->perf_events to find the event which has the
prog and executes it. This will support multiple prog's attaching to
the same trace_event_call as well.
This approach may need careful evaluation though.
(1). It adds runtime overhead although the overhead should be small
since perf_event attaching to the same trace_event_call should be small.
(2). trace_event_call->perf_events are per cpu data structure, that
means, some filtering logic is needed to avoid the same perf_event prog
is executing twice.
(3). since the list is traversed, the locking (rcu?) may be required to
pretect the list. Not sure whether additional locking is needed or not.
Alternative to using trace_event_call->perf_events, we may replace
"struct bpf_prog *prog" to a list (or some other list like data
structure) to just record perf events which have bpf progs attached.
But this will add memory overhead to trace_event_call data structure.
>
> -- Steve
>
>
>>
>> int (*perf_perm)(struct trace_event_call *,
>> struct perf_event *);
>> diff --git a/kernel/events/core.c b/kernel/events/core.c
>> index 3e691b7..6bc21e2 100644
>> --- a/kernel/events/core.c
>> +++ b/kernel/events/core.c
>> @@ -8171,6 +8171,7 @@ static int perf_event_set_bpf_prog(struct perf_event *event, u32 prog_fd)
>> }
>> }
>> event->tp_event->prog = prog;
>> + event->tp_event->bpf_prog_owner = event;
>>
>> return 0;
>> }
>> @@ -8185,7 +8186,7 @@ static void perf_event_free_bpf_prog(struct perf_event *event)
>> return;
>>
>> prog = event->tp_event->prog;
>> - if (prog) {
>> + if (prog && event->tp_event->bpf_prog_owner == event) {
>> event->tp_event->prog = NULL;
>> bpf_prog_put(prog);
>> }
>
^ permalink raw reply
* Re: [PATCH net] bpf: one perf event close won't free bpf program attached by another perf event
From: Yonghong Song @ 2017-09-21 5:20 UTC (permalink / raw)
To: Steven Rostedt; +Cc: peterz, ast, daniel, netdev, kernel-team
In-Reply-To: <94f5a61e-1a88-f285-224b-66c92dc3da7b@fb.com>
On 9/20/17 10:17 PM, Yonghong Song wrote:
>
>
> On 9/20/17 6:41 PM, Steven Rostedt wrote:
>> On Mon, 18 Sep 2017 16:38:36 -0700
>> Yonghong Song <yhs@fb.com> wrote:
>>
>>> This patch fixes a bug exhibited by the following scenario:
>>> 1. fd1 = perf_event_open with attr.config = ID1
>>> 2. attach bpf program prog1 to fd1
>>> 3. fd2 = perf_event_open with attr.config = ID1
>>> <this will be successful>
>>> 4. user program closes fd2 and prog1 is detached from the tracepoint.
>>> 5. user program with fd1 does not work properly as tracepoint
>>> no output any more.
>>>
>>> The issue happens at step 4. Multiple perf_event_open can be called
>>> successfully, but only one bpf prog pointer in the tp_event. In the
>>> current logic, any fd release for the same tp_event will free
>>> the tp_event->prog.
>>>
>>> The fix is to free tp_event->prog only when the closing fd
>>> corresponds to the one which registered the program.
>>>
>>> Signed-off-by: Yonghong Song <yhs@fb.com>
>>> ---
>>> Additional context: discussed with Alexei internally but did not find
>>> a solution which can avoid introducing the additional field in
>>> trace_event_call structure.
>>>
>>> Peter, could you take a look as well and maybe you could have better
>>> alternative? Thanks!
>>>
>>> include/linux/trace_events.h | 1 +
>>> kernel/events/core.c | 3 ++-
>>> 2 files changed, 3 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/include/linux/trace_events.h b/include/linux/trace_events.h
>>> index 7f11050..2e0f222 100644
>>> --- a/include/linux/trace_events.h
>>> +++ b/include/linux/trace_events.h
>>> @@ -272,6 +272,7 @@ struct trace_event_call {
>>> int perf_refcount;
>>> struct hlist_head __percpu *perf_events;
>>> struct bpf_prog *prog;
>>> + struct perf_event *bpf_prog_owner;
>>
>> Does this have to be in the trace_event_call structure? Hmm, I'm
>> wondering if the prog needs to be there (I should look to see if we can
>> move it from it). The trace_event_call is created for *every* event,
>> and there's thousands of them now. Every byte to this structure adds
>> 1000s of bytes to the kernel. Would it be possible to attach the prog
>> and the owner to the perf_event?
>
> Regarding whether we could move the prog and the owner to the
> perf_event. It is possible. There is already a "prog" field in the
> perf_event structure for overflow handler. We could reuse the "prog"
> field and we do not need bpf_prog_owner any more. We can iterate
> through trace_event_call->perf_events to find the event which has the
> prog and executes it. This will support multiple prog's attaching to
> the same trace_event_call as well.
>
> This approach may need careful evaluation though.
> (1). It adds runtime overhead although the overhead should be small
> since perf_event attaching to the same trace_event_call should be small.
> (2). trace_event_call->perf_events are per cpu data structure, that
> means, some filtering logic is needed to avoid the same perf_event prog
> is executing twice.
What I mean here is that the trace_event_call->perf_events need to be
checked on ALL cpus since bpf prog should be executed regardless of
cpu affiliation. It is possible that the same perf_event in different
per_cpu bucket and hence filtering is needed to avoid the same
perf_event bpf_prog is executed twice.
> (3). since the list is traversed, the locking (rcu?) may be required to
> pretect the list. Not sure whether additional locking is needed or not.
>
> Alternative to using trace_event_call->perf_events, we may replace
> "struct bpf_prog *prog" to a list (or some other list like data
> structure) to just record perf events which have bpf progs attached.
> But this will add memory overhead to trace_event_call data structure.
>
>>
>> -- Steve
>>
>>
>>> int (*perf_perm)(struct trace_event_call *,
>>> struct perf_event *);
>>> diff --git a/kernel/events/core.c b/kernel/events/core.c
>>> index 3e691b7..6bc21e2 100644
>>> --- a/kernel/events/core.c
>>> +++ b/kernel/events/core.c
>>> @@ -8171,6 +8171,7 @@ static int perf_event_set_bpf_prog(struct
>>> perf_event *event, u32 prog_fd)
>>> }
>>> }
>>> event->tp_event->prog = prog;
>>> + event->tp_event->bpf_prog_owner = event;
>>> return 0;
>>> }
>>> @@ -8185,7 +8186,7 @@ static void perf_event_free_bpf_prog(struct
>>> perf_event *event)
>>> return;
>>> prog = event->tp_event->prog;
>>> - if (prog) {
>>> + if (prog && event->tp_event->bpf_prog_owner == event) {
>>> event->tp_event->prog = NULL;
>>> bpf_prog_put(prog);
>>> }
>>
^ permalink raw reply
* [PATCH 2/6] e100: Cocci spatch "pool_zalloc-simple"
From: Thomas Meyer @ 2017-09-21 6:15 UTC (permalink / raw)
To: jeffrey.t.kirsher, intel-wired-lan, netdev, linux-kernel
In-Reply-To: <1505974468535-2047238743-0-diffsplit-thomas@m3y3r.de>
Use *_pool_zalloc rather than *_pool_alloc followed by memset with 0.
Found by coccinelle spatch "api/alloc/pool_zalloc-simple.cocci"
Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
---
diff -u -p a/drivers/net/ethernet/intel/e100.c b/drivers/net/ethernet/intel/e100.c
--- a/drivers/net/ethernet/intel/e100.c
+++ b/drivers/net/ethernet/intel/e100.c
@@ -1910,11 +1910,10 @@ static int e100_alloc_cbs(struct nic *ni
nic->cb_to_use = nic->cb_to_send = nic->cb_to_clean = NULL;
nic->cbs_avail = 0;
- nic->cbs = pci_pool_alloc(nic->cbs_pool, GFP_KERNEL,
- &nic->cbs_dma_addr);
+ nic->cbs = pci_pool_zalloc(nic->cbs_pool, GFP_KERNEL,
+ &nic->cbs_dma_addr);
if (!nic->cbs)
return -ENOMEM;
- memset(nic->cbs, 0, count * sizeof(struct cb));
for (cb = nic->cbs, i = 0; i < count; cb++, i++) {
cb->next = (i + 1 < count) ? cb + 1 : nic->cbs;
^ permalink raw reply
* [PATCH 3/3] net: stmmac: Cocci spatch "of_table"
From: Thomas Meyer @ 2017-09-21 6:24 UTC (permalink / raw)
To: peppe.cavallaro, alexandre.torgue, netdev, linux-kernel
In-Reply-To: <1505974912018-1637125348-0-diffsplit-thomas@m3y3r.de>
Make sure (of/i2c/platform)_device_id tables are NULL terminated.
Found by coccinelle spatch "misc/of_table.cocci"
Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
---
diff -u -p a/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c
@@ -315,6 +315,7 @@ static int stmmac_dt_phy(struct plat_stm
{ .compatible = "allwinner,sun8i-h3-emac" },
{ .compatible = "allwinner,sun8i-v3s-emac" },
{ .compatible = "allwinner,sun50i-a64-emac" },
+ {},
};
/* If phy-handle property is passed from DT, use it as the PHY */
^ permalink raw reply
* [patch net-next 00/12] mlxsw: Add support for offloading IPv4 multicast routes
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
From: Jiri Pirko <jiri@mellanox.com>
Yotam says:
This patchset introduces offloading of the kernel IPv4 multicast router
logic in the Spectrum driver.
Patches 1-4 update ipmr to use the FIB notification chain for both MFC and
VIF notifications, and patches 7-11 update the Spectrum driver to register
to these notifications and offload the routes.
Similarly to IPv4 and IPv6, any failure will trigger the abort mechanism
which is updated in this patch-set to eject multicast route tables too.
At this stage, the following limitations apply:
- A multicast MFC route will be offloaded by the driver if all the output
interfaces are Spectrum router interfaces (RIFs). In any other case
(which includes pimreg device, tunnel devices and management ports) the
route will be trapped to the CPU and the packets will be forwarded by
software.
- ipmr proxy routes are not supported and will trigger the abort
mechanism.
- The MFC TTL values are currently treated as boolean: if the value is
different than 255, the traffic is forwarded to the interface and if the
value is 255 it is not forwarded. Dropping packets based on their TTL isn't
currently supported.
To allow users to have visibility on which of the routes are offloaded and
which are not, patch 5 introduces a per-route offload indication similar to
IPv4 and IPv6 routes which is sent to the user via the RTNetlink interface.
The Spectrum driver multicast router offloading support, which is
introduced in patches 7 and 8, is divided into two parts:
- The hardware logic which abstracts the Spectrum hardware and provides a
simple API for the upper levels.
- The offloading logic which gets the MFC and VIF notifications from the
kernel and updates the hardware using the hardware logic part.
Finally, the last patch makes the Spectrum router logic not ignore the
multicast FIB notifications and call the corresponding functions in the
multicast router offloading logic.
Yotam Gigi (12):
fib: notifier: Add VIF add and delete event types
ipmr: Add reference count to MFC entries
ipmr: Add FIB notification access functions
ipmr: Send FIB notifications on MFC and VIF entries
net: ipmr: Add MFC offload indication
net: mroute: Check if rule is a default rule
mlxsw: spectrum: Add the multicast routing offloading logic
mlxsw: spectrum: Add the multicast routing hardware logic
mlxsw: spectrum: router: Squash the default route table to main
mlxsw: spectrum_router: Add multicast routes notification handling
functionality
mlxsw: spectrum: Notify multicast router on RIF MTU changes
mlxsw: spectrum: router: Don't ignore IPMR notifications
drivers/net/ethernet/mellanox/mlxsw/Makefile | 3 +-
drivers/net/ethernet/mellanox/mlxsw/spectrum.h | 2 +
drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c | 1012 ++++++++++++++++++++
drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h | 133 +++
.../net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c | 828 ++++++++++++++++
.../net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h | 43 +
.../net/ethernet/mellanox/mlxsw/spectrum_router.c | 205 +++-
include/linux/mroute.h | 43 +
include/net/fib_notifier.h | 2 +
include/net/netns/ipv4.h | 3 +
net/ipv4/ipmr.c | 209 +++-
11 files changed, 2472 insertions(+), 11 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c
create mode 100644 drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h
create mode 100644 drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c
create mode 100644 drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h
--
2.9.5
^ permalink raw reply
* [patch net-next 01/12] fib: notifier: Add VIF add and delete event types
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
In order for an interface to forward packets according to the kernel
multicast routing table, it must be configured with a VIF index according
to the mroute user API. The VIF index is then used to refer to that
interface in the mroute user API, for example, to set the iif and oifs of
an MFC entry.
In order to allow drivers to be aware and offload multicast routes, they
have to be aware of the VIF add and delete notifications.
Due to the fact that a specific VIF can be deleted and re-added pointing to
another netdevice, and the MFC routes that point to it will forward the
matching packets to the new netdevice, a driver willing to offload MFC
cache entries must be aware of the VIF add and delete events in addition to
MFC routes notifications.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
include/net/fib_notifier.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/include/net/fib_notifier.h b/include/net/fib_notifier.h
index 669b971..54cd6b8 100644
--- a/include/net/fib_notifier.h
+++ b/include/net/fib_notifier.h
@@ -20,6 +20,8 @@ enum fib_event_type {
FIB_EVENT_RULE_DEL,
FIB_EVENT_NH_ADD,
FIB_EVENT_NH_DEL,
+ FIB_EVENT_VIF_ADD,
+ FIB_EVENT_VIF_DEL,
};
struct fib_notifier_ops {
--
2.9.5
^ permalink raw reply related
* [patch net-next 03/12] ipmr: Add FIB notification access functions
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Make the ipmr module register as a FIB notifier. To do that, implement both
the ipmr_seq_read and ipmr_dump ops.
The ipmr_seq_read op returns a sequence counter that is incremented on
every notification related operation done by the ipmr. To implement that,
add a sequence counter in the netns_ipv4 struct and increment it whenever a
new MFC route or VIF are added or deleted. The sequence operations are
protected by the RTNL lock.
The ipmr_dump iterates the list of MFC routes and the list of VIF entries
and sends notifications about them. The entries dump is done under RCU.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
include/linux/mroute.h | 15 ++++++
include/net/netns/ipv4.h | 3 ++
net/ipv4/ipmr.c | 135 ++++++++++++++++++++++++++++++++++++++++++++++-
3 files changed, 151 insertions(+), 2 deletions(-)
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index 2f88e3d..ba5a976 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -5,6 +5,7 @@
#include <linux/pim.h>
#include <linux/rhashtable.h>
#include <net/sock.h>
+#include <net/fib_notifier.h>
#include <uapi/linux/mroute.h>
#ifdef CONFIG_IP_MROUTE
@@ -58,6 +59,14 @@ struct vif_device {
int link; /* Physical interface index */
};
+struct vif_entry_notifier_info {
+ struct fib_notifier_info info;
+ struct net_device *dev;
+ vifi_t vif_index;
+ unsigned short vif_flags;
+ u32 tb_id;
+};
+
#define VIFF_STATIC 0x8000
#define VIF_EXISTS(_mrt, _idx) ((_mrt)->vif_table[_idx].dev != NULL)
@@ -145,6 +154,12 @@ struct mfc_cache {
struct rcu_head rcu;
};
+struct mfc_entry_notifier_info {
+ struct fib_notifier_info info;
+ struct mfc_cache *mfc;
+ u32 tb_id;
+};
+
struct rtmsg;
int ipmr_get_route(struct net *net, struct sk_buff *skb,
__be32 saddr, __be32 daddr,
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index 20d061c..5921496 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -162,6 +162,9 @@ struct netns_ipv4 {
struct fib_notifier_ops *notifier_ops;
unsigned int fib_seq; /* protected by rtnl_mutex */
+ struct fib_notifier_ops *ipmr_notifier_ops;
+ unsigned int ipmr_seq; /* protected by rtnl_mutex */
+
atomic_t rt_genid;
};
#endif
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 86dc5f9..9d331a74 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -264,6 +264,16 @@ static void __net_exit ipmr_rules_exit(struct net *net)
fib_rules_unregister(net->ipv4.mr_rules_ops);
rtnl_unlock();
}
+
+static int ipmr_rules_dump(struct net *net, struct notifier_block *nb)
+{
+ return fib_rules_dump(net, nb, RTNL_FAMILY_IPMR);
+}
+
+static unsigned int ipmr_rules_seq_read(struct net *net)
+{
+ return fib_rules_seq_read(net, RTNL_FAMILY_IPMR);
+}
#else
#define ipmr_for_each_table(mrt, net) \
for (mrt = net->ipv4.mrt; mrt; mrt = NULL)
@@ -298,6 +308,16 @@ static void __net_exit ipmr_rules_exit(struct net *net)
net->ipv4.mrt = NULL;
rtnl_unlock();
}
+
+static int ipmr_rules_dump(struct net *net, struct notifier_block *nb)
+{
+ return 0;
+}
+
+static unsigned int ipmr_rules_seq_read(struct net *net)
+{
+ return 0;
+}
#endif
static inline int ipmr_hash_cmp(struct rhashtable_compare_arg *arg,
@@ -587,6 +607,43 @@ static struct net_device *ipmr_reg_vif(struct net *net, struct mr_table *mrt)
}
#endif
+static int call_ipmr_vif_entry_notifier(struct notifier_block *nb,
+ struct net *net,
+ enum fib_event_type event_type,
+ struct vif_device *vif,
+ vifi_t vif_index, u32 tb_id)
+{
+ struct vif_entry_notifier_info info = {
+ .info = {
+ .family = RTNL_FAMILY_IPMR,
+ .net = net,
+ },
+ .dev = vif->dev,
+ .vif_index = vif_index,
+ .vif_flags = vif->flags,
+ .tb_id = tb_id,
+ };
+
+ return call_fib_notifier(nb, net, event_type, &info.info);
+}
+
+static int call_ipmr_mfc_entry_notifier(struct notifier_block *nb,
+ struct net *net,
+ enum fib_event_type event_type,
+ struct mfc_cache *mfc, u32 tb_id)
+{
+ struct mfc_entry_notifier_info info = {
+ .info = {
+ .family = RTNL_FAMILY_IPMR,
+ .net = net,
+ },
+ .mfc = mfc,
+ .tb_id = tb_id
+ };
+
+ return call_fib_notifier(nb, net, event_type, &info.info);
+}
+
/**
* vif_delete - Delete a VIF entry
* @notify: Set to 1, if the caller is a notifier_call
@@ -3050,14 +3107,85 @@ static const struct net_protocol pim_protocol = {
};
#endif
+static unsigned int ipmr_seq_read(struct net *net)
+{
+ ASSERT_RTNL();
+
+ return net->ipv4.ipmr_seq + ipmr_rules_seq_read(net);
+}
+
+static int ipmr_dump(struct net *net, struct notifier_block *nb)
+{
+ struct mr_table *mrt;
+ int err;
+
+ err = ipmr_rules_dump(net, nb);
+ if (err)
+ return err;
+
+ ipmr_for_each_table(mrt, net) {
+ struct vif_device *v = &mrt->vif_table[0];
+ struct mfc_cache *mfc;
+ int vifi;
+
+ /* Notifiy on table VIF entries */
+ for (vifi = 0; vifi < mrt->maxvif; vifi++, v++) {
+ if (!v->dev)
+ continue;
+
+ call_ipmr_vif_entry_notifier(nb, net, FIB_EVENT_VIF_ADD,
+ v, vifi, mrt->id);
+ }
+
+ /* Notify on table MFC entries */
+ list_for_each_entry_rcu(mfc, &mrt->mfc_cache_list, list)
+ call_ipmr_mfc_entry_notifier(nb, net,
+ FIB_EVENT_ENTRY_ADD, mfc,
+ mrt->id);
+ }
+
+ return 0;
+}
+
+static const struct fib_notifier_ops ipmr_notifier_ops_template = {
+ .family = RTNL_FAMILY_IPMR,
+ .fib_seq_read = ipmr_seq_read,
+ .fib_dump = ipmr_dump,
+ .owner = THIS_MODULE,
+};
+
+int __net_init ipmr_notifier_init(struct net *net)
+{
+ struct fib_notifier_ops *ops;
+
+ net->ipv4.ipmr_seq = 0;
+
+ ops = fib_notifier_ops_register(&ipmr_notifier_ops_template, net);
+ if (IS_ERR(ops))
+ return PTR_ERR(ops);
+ net->ipv4.ipmr_notifier_ops = ops;
+
+ return 0;
+}
+
+static void __net_exit ipmr_notifier_exit(struct net *net)
+{
+ fib_notifier_ops_unregister(net->ipv4.ipmr_notifier_ops);
+ net->ipv4.ipmr_notifier_ops = NULL;
+}
+
/* Setup for IP multicast routing */
static int __net_init ipmr_net_init(struct net *net)
{
int err;
+ err = ipmr_notifier_init(net);
+ if (err)
+ goto ipmr_notifier_fail;
+
err = ipmr_rules_init(net);
if (err < 0)
- goto fail;
+ goto ipmr_rules_fail;
#ifdef CONFIG_PROC_FS
err = -ENOMEM;
@@ -3074,7 +3202,9 @@ static int __net_init ipmr_net_init(struct net *net)
proc_vif_fail:
ipmr_rules_exit(net);
#endif
-fail:
+ipmr_rules_fail:
+ ipmr_notifier_exit(net);
+ipmr_notifier_fail:
return err;
}
@@ -3084,6 +3214,7 @@ static void __net_exit ipmr_net_exit(struct net *net)
remove_proc_entry("ip_mr_cache", net->proc_net);
remove_proc_entry("ip_mr_vif", net->proc_net);
#endif
+ ipmr_notifier_exit(net);
ipmr_rules_exit(net);
}
--
2.9.5
^ permalink raw reply related
* [patch net-next 02/12] ipmr: Add reference count to MFC entries
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Next commits will introduce MFC notifications through the atomic
fib_notification chain, thus allowing modules to be aware of MFC entries.
Due to the fact that modules may need to hold a reference to an MFC entry,
add reference count to MFC entries to prevent them from being freed while
these modules use them.
The reference counting is done only on resolved MFC entries currently.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
include/linux/mroute.h | 20 ++++++++++++++++++++
net/ipv4/ipmr.c | 8 +++++---
2 files changed, 25 insertions(+), 3 deletions(-)
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index d7f6333..2f88e3d 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -138,6 +138,7 @@ struct mfc_cache {
unsigned long wrong_if;
unsigned long lastuse;
unsigned char ttls[MAXVIFS];
+ refcount_t refcount;
} res;
} mfc_un;
struct list_head list;
@@ -148,4 +149,23 @@ struct rtmsg;
int ipmr_get_route(struct net *net, struct sk_buff *skb,
__be32 saddr, __be32 daddr,
struct rtmsg *rtm, u32 portid);
+
+#ifdef CONFIG_IP_MROUTE
+void ipmr_cache_free(struct mfc_cache *mfc_cache);
+#else
+static inline void ipmr_cache_free(struct mfc_cache *mfc_cache)
+{
+}
+#endif
+
+static inline void ipmr_cache_put(struct mfc_cache *c)
+{
+ if (refcount_dec_and_test(&c->mfc_un.res.refcount))
+ ipmr_cache_free(c);
+}
+static inline void ipmr_cache_hold(struct mfc_cache *c)
+{
+ refcount_inc(&c->mfc_un.res.refcount);
+}
+
#endif
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index c9b3e6e..86dc5f9 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -652,10 +652,11 @@ static void ipmr_cache_free_rcu(struct rcu_head *head)
kmem_cache_free(mrt_cachep, c);
}
-static inline void ipmr_cache_free(struct mfc_cache *c)
+void ipmr_cache_free(struct mfc_cache *c)
{
call_rcu(&c->rcu, ipmr_cache_free_rcu);
}
+EXPORT_SYMBOL(ipmr_cache_free);
/* Destroy an unresolved cache entry, killing queued skbs
* and reporting error to netlink readers.
@@ -949,6 +950,7 @@ static struct mfc_cache *ipmr_cache_alloc(void)
if (c) {
c->mfc_un.res.last_assert = jiffies - MFC_ASSERT_THRESH - 1;
c->mfc_un.res.minvif = MAXVIFS;
+ refcount_set(&c->mfc_un.res.refcount, 1);
}
return c;
}
@@ -1162,7 +1164,7 @@ static int ipmr_mfc_delete(struct mr_table *mrt, struct mfcctl *mfc, int parent)
rhltable_remove(&mrt->mfc_hash, &c->mnode, ipmr_rht_params);
list_del_rcu(&c->list);
mroute_netlink_event(mrt, c, RTM_DELROUTE);
- ipmr_cache_free(c);
+ ipmr_cache_put(c);
return 0;
}
@@ -1264,7 +1266,7 @@ static void mroute_clean_tables(struct mr_table *mrt, bool all)
rhltable_remove(&mrt->mfc_hash, &c->mnode, ipmr_rht_params);
list_del_rcu(&c->list);
mroute_netlink_event(mrt, c, RTM_DELROUTE);
- ipmr_cache_free(c);
+ ipmr_cache_put(c);
}
if (atomic_read(&mrt->cache_resolve_queue_len) != 0) {
--
2.9.5
^ permalink raw reply related
* [patch net-next 04/12] ipmr: Send FIB notifications on MFC and VIF entries
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Use the newly introduced notification chain to send events upon VIF and MFC
addition and deletion. The MFC notifications are sent only on resolved MFC
entries, as unresolved cannot be offloaded.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
net/ipv4/ipmr.c | 53 +++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 53 insertions(+)
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 9d331a74..7891d95 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -627,6 +627,27 @@ static int call_ipmr_vif_entry_notifier(struct notifier_block *nb,
return call_fib_notifier(nb, net, event_type, &info.info);
}
+static int call_ipmr_vif_entry_notifiers(struct net *net,
+ enum fib_event_type event_type,
+ struct vif_device *vif,
+ vifi_t vif_index, u32 tb_id)
+{
+ struct vif_entry_notifier_info info = {
+ .info = {
+ .family = RTNL_FAMILY_IPMR,
+ .net = net,
+ },
+ .dev = vif->dev,
+ .vif_index = vif_index,
+ .vif_flags = vif->flags,
+ .tb_id = tb_id,
+ };
+
+ ASSERT_RTNL();
+ net->ipv4.ipmr_seq++;
+ return call_fib_notifiers(net, event_type, &info.info);
+}
+
static int call_ipmr_mfc_entry_notifier(struct notifier_block *nb,
struct net *net,
enum fib_event_type event_type,
@@ -644,6 +665,24 @@ static int call_ipmr_mfc_entry_notifier(struct notifier_block *nb,
return call_fib_notifier(nb, net, event_type, &info.info);
}
+static int call_ipmr_mfc_entry_notifiers(struct net *net,
+ enum fib_event_type event_type,
+ struct mfc_cache *mfc, u32 tb_id)
+{
+ struct mfc_entry_notifier_info info = {
+ .info = {
+ .family = RTNL_FAMILY_IPMR,
+ .net = net,
+ },
+ .mfc = mfc,
+ .tb_id = tb_id
+ };
+
+ ASSERT_RTNL();
+ net->ipv4.ipmr_seq++;
+ return call_fib_notifiers(net, event_type, &info.info);
+}
+
/**
* vif_delete - Delete a VIF entry
* @notify: Set to 1, if the caller is a notifier_call
@@ -651,6 +690,7 @@ static int call_ipmr_mfc_entry_notifier(struct notifier_block *nb,
static int vif_delete(struct mr_table *mrt, int vifi, int notify,
struct list_head *head)
{
+ struct net *net = read_pnet(&mrt->net);
struct vif_device *v;
struct net_device *dev;
struct in_device *in_dev;
@@ -660,6 +700,10 @@ static int vif_delete(struct mr_table *mrt, int vifi, int notify,
v = &mrt->vif_table[vifi];
+ if (VIF_EXISTS(mrt, vifi))
+ call_ipmr_vif_entry_notifiers(net, FIB_EVENT_VIF_DEL, v, vifi,
+ mrt->id);
+
write_lock_bh(&mrt_lock);
dev = v->dev;
v->dev = NULL;
@@ -909,6 +953,7 @@ static int vif_add(struct net *net, struct mr_table *mrt,
if (vifi+1 > mrt->maxvif)
mrt->maxvif = vifi+1;
write_unlock_bh(&mrt_lock);
+ call_ipmr_vif_entry_notifiers(net, FIB_EVENT_VIF_ADD, v, vifi, mrt->id);
return 0;
}
@@ -1209,6 +1254,7 @@ static int ipmr_cache_unresolved(struct mr_table *mrt, vifi_t vifi,
static int ipmr_mfc_delete(struct mr_table *mrt, struct mfcctl *mfc, int parent)
{
+ struct net *net = read_pnet(&mrt->net);
struct mfc_cache *c;
/* The entries are added/deleted only under RTNL */
@@ -1220,6 +1266,7 @@ static int ipmr_mfc_delete(struct mr_table *mrt, struct mfcctl *mfc, int parent)
return -ENOENT;
rhltable_remove(&mrt->mfc_hash, &c->mnode, ipmr_rht_params);
list_del_rcu(&c->list);
+ call_ipmr_mfc_entry_notifiers(net, FIB_EVENT_ENTRY_DEL, c, mrt->id);
mroute_netlink_event(mrt, c, RTM_DELROUTE);
ipmr_cache_put(c);
@@ -1248,6 +1295,8 @@ static int ipmr_mfc_add(struct net *net, struct mr_table *mrt,
if (!mrtsock)
c->mfc_flags |= MFC_STATIC;
write_unlock_bh(&mrt_lock);
+ call_ipmr_mfc_entry_notifiers(net, FIB_EVENT_ENTRY_REPLACE, c,
+ mrt->id);
mroute_netlink_event(mrt, c, RTM_NEWROUTE);
return 0;
}
@@ -1297,6 +1346,7 @@ static int ipmr_mfc_add(struct net *net, struct mr_table *mrt,
ipmr_cache_resolve(net, mrt, uc, c);
ipmr_cache_free(uc);
}
+ call_ipmr_mfc_entry_notifiers(net, FIB_EVENT_ENTRY_ADD, c, mrt->id);
mroute_netlink_event(mrt, c, RTM_NEWROUTE);
return 0;
}
@@ -1304,6 +1354,7 @@ static int ipmr_mfc_add(struct net *net, struct mr_table *mrt,
/* Close the multicast socket, and clear the vif tables etc */
static void mroute_clean_tables(struct mr_table *mrt, bool all)
{
+ struct net *net = read_pnet(&mrt->net);
struct mfc_cache *c, *tmp;
LIST_HEAD(list);
int i;
@@ -1322,6 +1373,8 @@ static void mroute_clean_tables(struct mr_table *mrt, bool all)
continue;
rhltable_remove(&mrt->mfc_hash, &c->mnode, ipmr_rht_params);
list_del_rcu(&c->list);
+ call_ipmr_mfc_entry_notifiers(net, FIB_EVENT_ENTRY_DEL, c,
+ mrt->id);
mroute_netlink_event(mrt, c, RTM_DELROUTE);
ipmr_cache_put(c);
}
--
2.9.5
^ permalink raw reply related
* [patch net-next 05/12] net: ipmr: Add MFC offload indication
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Allow drivers, registered to the fib notification chain indicate whether a
multicast MFC route is offloaded or not, similarly to unicast routes. The
indication of whether a route is offloaded is done using the mfc_flags
field on an mfc_cache struct, and the information is sent to the userspace
via the RTNetlink interface only.
Currently, MFC routes are either offloaded or not, thus there is no need to
add per-VIF offload indication.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
include/linux/mroute.h | 1 +
net/ipv4/ipmr.c | 3 +++
2 files changed, 4 insertions(+)
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index ba5a976..9b64587 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -93,6 +93,7 @@ struct mr_table {
*/
enum {
MFC_STATIC = BIT(0),
+ MFC_OFFLOAD = BIT(1),
};
struct mfc_cache_cmp_arg {
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 7891d95..6804885 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -2268,6 +2268,9 @@ static int __ipmr_fill_mroute(struct mr_table *mrt, struct sk_buff *skb,
nla_put_u32(skb, RTA_IIF, mrt->vif_table[c->mfc_parent].dev->ifindex) < 0)
return -EMSGSIZE;
+ if (c->mfc_flags & MFC_OFFLOAD)
+ rtm->rtm_flags |= RTNH_F_OFFLOAD;
+
if (!(mp_attr = nla_nest_start(skb, RTA_MULTIPATH)))
return -EMSGSIZE;
--
2.9.5
^ permalink raw reply related
* [patch net-next 06/12] net: mroute: Check if rule is a default rule
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
When the ipmr starts, it adds one default FIB rule that matches all packets
and sends them to the DEFAULT (multicast) FIB table. A more complex rule
can be added by user to specify that for a specific interface, a packet
should be look up at either an arbitrary table or according to the l3mdev
of the interface.
For drivers willing to offload the ipmr logic into a hardware but don't
want to offload all the FIB rules functionality, provide a function that
can indicate whether the FIB rule is the default multicast rule, thus only
one routing table is needed.
This way, a driver can register to the FIB notification chain, get
notifications about FIB rules added and trigger some kind of an internal
abort mechanism when a non default rule is added by the user.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
include/linux/mroute.h | 7 +++++++
net/ipv4/ipmr.c | 10 ++++++++++
2 files changed, 17 insertions(+)
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index 9b64587..585369d 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -5,6 +5,7 @@
#include <linux/pim.h>
#include <linux/rhashtable.h>
#include <net/sock.h>
+#include <net/fib_rules.h>
#include <net/fib_notifier.h>
#include <uapi/linux/mroute.h>
@@ -19,6 +20,7 @@ int ip_mroute_getsockopt(struct sock *, int, char __user *, int __user *);
int ipmr_ioctl(struct sock *sk, int cmd, void __user *arg);
int ipmr_compat_ioctl(struct sock *sk, unsigned int cmd, void __user *arg);
int ip_mr_init(void);
+bool ipmr_rule_default(const struct fib_rule *rule);
#else
static inline int ip_mroute_setsockopt(struct sock *sock, int optname,
char __user *optval, unsigned int optlen)
@@ -46,6 +48,11 @@ static inline int ip_mroute_opt(int opt)
{
return 0;
}
+
+static inline bool ipmr_rule_default(const struct fib_rule *rule)
+{
+ return true;
+}
#endif
struct vif_device {
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 6804885..6e1de35 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -320,6 +320,16 @@ static unsigned int ipmr_rules_seq_read(struct net *net)
}
#endif
+bool ipmr_rule_default(const struct fib_rule *rule)
+{
+#if IS_ENABLED(CONFIG_FIB_RULES)
+ return fib_rule_matchall(rule) && rule->table == RT_TABLE_DEFAULT;
+#else
+ return true;
+#endif
+}
+EXPORT_SYMBOL(ipmr_rule_default);
+
static inline int ipmr_hash_cmp(struct rhashtable_compare_arg *arg,
const void *ptr)
{
--
2.9.5
^ permalink raw reply related
* [patch net-next 07/12] mlxsw: spectrum: Add the multicast routing offloading logic
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Add the multicast router offloading logic, which is in charge of handling
the VIF and MFC notifications and translating it to the hardware logic API.
The offloading logic has to overcome several obstacles in order to safely
comply with the kernel multicast router user API:
- It must keep track of the mapping between VIFs to netdevices. The user
can add an MFC cache entry pointing to a VIF, delete the VIF and add
re-add it with a different netdevice. The offloading logic has to handle
this in order to be compatible with the kernel logic.
- It must keep track of the mapping between netdevices to spectrum RIFs,
as the current hardware implementation assume having a RIF for every
port in a multicast router.
- It must handle routes pointing to pimreg device to be trapped to the
kernel, as the packet should be delivered to userspace.
- It must handle routes pointing tunnel VIFs. The current implementation
does not support multicast forwarding to tunnels, thus routes that point
to a tunnel should be trapped to the kernel.
- It must be aware of proxy multicast routes, which include both (*,*)
routes and duplicate routes. Currently proxy routes are not offloaded
and trigger the abort mechanism: removal of all routes from hardware and
triggering the traffic to go through the kernel.
The multicast routing offloading logic also updates the counters of the
offloaded MFC routes in a periodic work.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
drivers/net/ethernet/mellanox/mlxsw/Makefile | 3 +-
drivers/net/ethernet/mellanox/mlxsw/spectrum.h | 1 +
drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c | 1012 +++++++++++++++++++++
drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h | 133 +++
4 files changed, 1148 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c
create mode 100644 drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h
diff --git a/drivers/net/ethernet/mellanox/mlxsw/Makefile b/drivers/net/ethernet/mellanox/mlxsw/Makefile
index 4b88158..9b29764 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/Makefile
+++ b/drivers/net/ethernet/mellanox/mlxsw/Makefile
@@ -17,7 +17,8 @@ mlxsw_spectrum-objs := spectrum.o spectrum_buffers.o \
spectrum_kvdl.o spectrum_acl_tcam.o \
spectrum_acl.o spectrum_flower.o \
spectrum_cnt.o spectrum_fid.o \
- spectrum_ipip.o spectrum_acl_flex_actions.o
+ spectrum_ipip.o spectrum_acl_flex_actions.o \
+ spectrum_mr.o
mlxsw_spectrum-$(CONFIG_MLXSW_SPECTRUM_DCB) += spectrum_dcb.o
mlxsw_spectrum-$(CONFIG_NET_DEVLINK) += spectrum_dpipe.o
obj-$(CONFIG_MLXSW_MINIMAL) += mlxsw_minimal.o
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
index e907ec4..51d8b9f 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
@@ -153,6 +153,7 @@ struct mlxsw_sp {
struct mlxsw_sp_sb *sb;
struct mlxsw_sp_bridge *bridge;
struct mlxsw_sp_router *router;
+ struct mlxsw_sp_mr *mr;
struct mlxsw_afa *afa;
struct mlxsw_sp_acl *acl;
struct mlxsw_sp_fid_core *fid_core;
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c
new file mode 100644
index 0000000..c77febd
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c
@@ -0,0 +1,1012 @@
+/*
+ * drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.c
+ * Copyright (c) 2017 Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2017 Yotam Gigi <yotamg@mellanox.com>
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ * 3. Neither the names of the copyright holders nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * Alternatively, this software may be distributed under the terms of the
+ * GNU General Public License ("GPL") version 2 as published by the Free
+ * Software Foundation.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/rhashtable.h>
+
+#include "spectrum_mr.h"
+#include "spectrum_router.h"
+
+struct mlxsw_sp_mr {
+ const struct mlxsw_sp_mr_ops *mr_ops;
+ void *catchall_route_priv;
+ struct delayed_work stats_update_dw;
+ struct list_head table_list;
+#define MLXSW_SP_MR_ROUTES_COUNTER_UPDATE_INTERVAL 5000 /* ms */
+ unsigned long priv[0];
+ /* priv has to be always the last item */
+};
+
+struct mlxsw_sp_mr_vif {
+ struct net_device *dev;
+ const struct mlxsw_sp_rif *rif;
+ unsigned long vif_flags;
+
+ /* A list of route_vif_entry structs that point to routes that the VIF
+ * instance is used as one of the egress VIFs
+ */
+ struct list_head route_evif_list;
+
+ /* A list of route_vif_entry structs that point to routes that the VIF
+ * instance is used as an ingress VIF
+ */
+ struct list_head route_ivif_list;
+};
+
+struct mlxsw_sp_mr_route_vif_entry {
+ struct list_head vif_node;
+ struct list_head route_node;
+ struct mlxsw_sp_mr_vif *mr_vif;
+ struct mlxsw_sp_mr_route *mr_route;
+};
+
+struct mlxsw_sp_mr_table {
+ struct list_head node;
+ enum mlxsw_sp_l3proto proto;
+ struct mlxsw_sp *mlxsw_sp;
+ u32 vr_id;
+ struct mlxsw_sp_mr_vif vifs[MAXVIFS];
+ struct list_head route_list;
+ struct rhashtable route_ht;
+ char catchall_route_priv[0];
+ /* catchall_route_priv has to be always the last item */
+};
+
+struct mlxsw_sp_mr_route {
+ struct list_head node;
+ struct rhash_head ht_node;
+ struct mlxsw_sp_mr_route_key key;
+ enum mlxsw_sp_mr_route_action route_action;
+ u16 min_mtu;
+ struct mfc_cache *mfc4;
+ void *route_priv;
+ const struct mlxsw_sp_mr_table *mr_table;
+ /* A list of route_vif_entry structs that point to the egress VIFs */
+ struct list_head evif_list;
+ /* A route_vif_entry struct that point to the ingress VIF */
+ struct mlxsw_sp_mr_route_vif_entry ivif;
+};
+
+static const struct rhashtable_params mlxsw_sp_mr_route_ht_params = {
+ .key_len = sizeof(struct mlxsw_sp_mr_route_key),
+ .key_offset = offsetof(struct mlxsw_sp_mr_route, key),
+ .head_offset = offsetof(struct mlxsw_sp_mr_route, ht_node),
+ .automatic_shrinking = true,
+};
+
+static bool mlxsw_sp_mr_vif_regular(const struct mlxsw_sp_mr_vif *vif)
+{
+ return !(vif->vif_flags & (VIFF_TUNNEL | VIFF_REGISTER));
+}
+
+static bool mlxsw_sp_mr_vif_valid(const struct mlxsw_sp_mr_vif *vif)
+{
+ return mlxsw_sp_mr_vif_regular(vif) && vif->dev && vif->rif;
+}
+
+static bool mlxsw_sp_mr_vif_rif_invalid(const struct mlxsw_sp_mr_vif *vif)
+{
+ return mlxsw_sp_mr_vif_regular(vif) && vif->dev && !vif->rif;
+}
+
+static bool
+mlxsw_sp_mr_route_ivif_in_evifs(const struct mlxsw_sp_mr_route *mr_route)
+{
+ vifi_t ivif;
+
+ switch (mr_route->mr_table->proto) {
+ case MLXSW_SP_L3_PROTO_IPV4:
+ ivif = mr_route->mfc4->mfc_parent;
+ return mr_route->mfc4->mfc_un.res.ttls[ivif] != 255;
+ case MLXSW_SP_L3_PROTO_IPV6:
+ /* fall through */
+ default:
+ WARN_ON_ONCE(1);
+ }
+ return false;
+}
+
+static int
+mlxsw_sp_mr_route_valid_evifs_num(const struct mlxsw_sp_mr_route *mr_route)
+{
+ struct mlxsw_sp_mr_route_vif_entry *rve;
+ int valid_evifs = 0;
+
+ valid_evifs = 0;
+ list_for_each_entry(rve, &mr_route->evif_list, route_node)
+ if (mlxsw_sp_mr_vif_valid(rve->mr_vif))
+ valid_evifs++;
+ return valid_evifs;
+}
+
+static bool mlxsw_sp_mr_route_starg(const struct mlxsw_sp_mr_route *mr_route)
+{
+ switch (mr_route->mr_table->proto) {
+ case MLXSW_SP_L3_PROTO_IPV4:
+ return mr_route->key.source_mask.addr4 == INADDR_ANY;
+ case MLXSW_SP_L3_PROTO_IPV6:
+ /* fall through */
+ default:
+ WARN_ON_ONCE(1);
+ }
+ return false;
+}
+
+static enum mlxsw_sp_mr_route_action
+mlxsw_sp_mr_route_action(const struct mlxsw_sp_mr_route *mr_route)
+{
+ struct mlxsw_sp_mr_route_vif_entry *rve;
+
+ /* If the ingress port is not regular and resolved, trap the route */
+ if (!mlxsw_sp_mr_vif_valid(mr_route->ivif.mr_vif))
+ return MLXSW_SP_MR_ROUTE_ACTION_TRAP;
+
+ /* The kernel does not match a (*,G) route that the ingress interface is
+ * not one of the egress interfaces, so trap these kind of routes.
+ */
+ if (mlxsw_sp_mr_route_starg(mr_route) &&
+ !mlxsw_sp_mr_route_ivif_in_evifs(mr_route))
+ return MLXSW_SP_MR_ROUTE_ACTION_TRAP;
+
+ /* If the route has no valid eVIFs, trap it. */
+ if (!mlxsw_sp_mr_route_valid_evifs_num(mr_route))
+ return MLXSW_SP_MR_ROUTE_ACTION_TRAP;
+
+ /* If either one of the eVIFs is not regular (VIF of type pimreg or
+ * tunnel) or one of the VIFs has no matching RIF, trap the packet.
+ */
+ list_for_each_entry(rve, &mr_route->evif_list, route_node) {
+ if (!mlxsw_sp_mr_vif_regular(rve->mr_vif) ||
+ mlxsw_sp_mr_vif_rif_invalid(rve->mr_vif))
+ return MLXSW_SP_MR_ROUTE_ACTION_TRAP;
+ }
+ return MLXSW_SP_MR_ROUTE_ACTION_FORWARD;
+}
+
+static enum mlxsw_sp_mr_route_prio
+mlxsw_sp_mr_route_prio(const struct mlxsw_sp_mr_route *mr_route)
+{
+ return mlxsw_sp_mr_route_starg(mr_route) ?
+ MLXSW_SP_MR_ROUTE_PRIO_STARG : MLXSW_SP_MR_ROUTE_PRIO_SG;
+}
+
+static void mlxsw_sp_mr_route4_key(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route_key *key,
+ const struct mfc_cache *mfc)
+{
+ bool starg = (mfc->mfc_origin == INADDR_ANY);
+
+ memset(key, 0, sizeof(*key));
+ key->vrid = mr_table->vr_id;
+ key->proto = mr_table->proto;
+ key->group.addr4 = mfc->mfc_mcastgrp;
+ key->group_mask.addr4 = 0xffffffff;
+ key->source.addr4 = mfc->mfc_origin;
+ key->source_mask.addr4 = starg ? 0 : 0xffffffff;
+}
+
+static int mlxsw_sp_mr_route_evif_link(struct mlxsw_sp_mr_route *mr_route,
+ struct mlxsw_sp_mr_vif *mr_vif)
+{
+ struct mlxsw_sp_mr_route_vif_entry *rve;
+
+ rve = kzalloc(sizeof(*rve), GFP_KERNEL);
+ if (!rve)
+ return -ENOMEM;
+ rve->mr_route = mr_route;
+ rve->mr_vif = mr_vif;
+ list_add_tail(&rve->route_node, &mr_route->evif_list);
+ list_add_tail(&rve->vif_node, &mr_vif->route_evif_list);
+ return 0;
+}
+
+static void
+mlxsw_sp_mr_route_evif_unlink(struct mlxsw_sp_mr_route_vif_entry *rve)
+{
+ list_del(&rve->route_node);
+ list_del(&rve->vif_node);
+ kfree(rve);
+}
+
+static void mlxsw_sp_mr_route_ivif_link(struct mlxsw_sp_mr_route *mr_route,
+ struct mlxsw_sp_mr_vif *mr_vif)
+{
+ mr_route->ivif.mr_route = mr_route;
+ mr_route->ivif.mr_vif = mr_vif;
+ list_add_tail(&mr_route->ivif.vif_node, &mr_vif->route_ivif_list);
+}
+
+static void mlxsw_sp_mr_route_ivif_unlink(struct mlxsw_sp_mr_route *mr_route)
+{
+ list_del(&mr_route->ivif.vif_node);
+}
+
+static int
+mlxsw_sp_mr_route_info_create(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route *mr_route,
+ struct mlxsw_sp_mr_route_info *route_info)
+{
+ struct mlxsw_sp_mr_route_vif_entry *rve;
+ u16 *erif_indices;
+ u16 irif_index;
+ u16 erif = 0;
+
+ erif_indices = kmalloc_array(MAXVIFS, sizeof(*erif_indices),
+ GFP_KERNEL);
+ if (!erif_indices)
+ return -ENOMEM;
+
+ list_for_each_entry(rve, &mr_route->evif_list, route_node) {
+ if (mlxsw_sp_mr_vif_valid(rve->mr_vif)) {
+ u16 rifi = mlxsw_sp_rif_index(rve->mr_vif->rif);
+
+ erif_indices[erif++] = rifi;
+ }
+ }
+
+ if (mlxsw_sp_mr_vif_valid(mr_route->ivif.mr_vif))
+ irif_index = mlxsw_sp_rif_index(mr_route->ivif.mr_vif->rif);
+ else
+ irif_index = 0;
+
+ route_info->irif_index = irif_index;
+ route_info->erif_indices = erif_indices;
+ route_info->min_mtu = mr_route->min_mtu;
+ route_info->route_action = mr_route->route_action;
+ route_info->erif_num = erif;
+ return 0;
+}
+
+static void
+mlxsw_sp_mr_route_info_destroy(struct mlxsw_sp_mr_route_info *route_info)
+{
+ kfree(route_info->erif_indices);
+}
+
+static int mlxsw_sp_mr_route_write(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route *mr_route,
+ bool replace)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp;
+ struct mlxsw_sp_mr_route_info route_info;
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+ int err;
+
+ err = mlxsw_sp_mr_route_info_create(mr_table, mr_route, &route_info);
+ if (err)
+ return err;
+
+ if (!replace) {
+ struct mlxsw_sp_mr_route_params route_params;
+
+ mr_route->route_priv = kzalloc(mr->mr_ops->route_priv_size,
+ GFP_KERNEL);
+ if (!mr_route->route_priv) {
+ err = -ENOMEM;
+ goto out;
+ }
+
+ route_params.key = mr_route->key;
+ route_params.value = route_info;
+ route_params.prio = mlxsw_sp_mr_route_prio(mr_route);
+ err = mr->mr_ops->route_create(mlxsw_sp, mr->priv,
+ mr_route->route_priv,
+ &route_params);
+ if (err)
+ kfree(mr_route->route_priv);
+ } else {
+ err = mr->mr_ops->route_update(mlxsw_sp, mr_route->route_priv,
+ &route_info);
+ }
+out:
+ mlxsw_sp_mr_route_info_destroy(&route_info);
+ return err;
+}
+
+static void mlxsw_sp_mr_route_erase(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route *mr_route)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp;
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+
+ mr->mr_ops->route_destroy(mlxsw_sp, mr->priv, mr_route->route_priv);
+ kfree(mr_route->route_priv);
+}
+
+static struct mlxsw_sp_mr_route *
+mlxsw_sp_mr_route4_create(struct mlxsw_sp_mr_table *mr_table,
+ struct mfc_cache *mfc)
+{
+ struct mlxsw_sp_mr_route_vif_entry *rve, *tmp;
+ struct mlxsw_sp_mr_route *mr_route;
+ int err;
+ int i;
+
+ /* Allocate and init a new route and fill it with parameters */
+ mr_route = kzalloc(sizeof(*mr_table), GFP_KERNEL);
+ if (!mr_route)
+ return ERR_PTR(-ENOMEM);
+ INIT_LIST_HEAD(&mr_route->evif_list);
+ mlxsw_sp_mr_route4_key(mr_table, &mr_route->key, mfc);
+
+ /* Find min_mtu and link iVIF and eVIFs */
+ mr_route->min_mtu = ETH_MAX_MTU;
+ ipmr_cache_hold(mfc);
+ mr_route->mfc4 = mfc;
+ mr_route->mr_table = mr_table;
+ for (i = 0; i < MAXVIFS; i++) {
+ if (mfc->mfc_un.res.ttls[i] != 255) {
+ err = mlxsw_sp_mr_route_evif_link(mr_route,
+ &mr_table->vifs[i]);
+ if (err)
+ goto err;
+ if (mr_table->vifs[i].dev &&
+ mr_table->vifs[i].dev->mtu < mr_route->min_mtu)
+ mr_route->min_mtu = mr_table->vifs[i].dev->mtu;
+ }
+ }
+ mlxsw_sp_mr_route_ivif_link(mr_route, &mr_table->vifs[mfc->mfc_parent]);
+ if (err)
+ goto err;
+
+ mr_route->route_action = mlxsw_sp_mr_route_action(mr_route);
+ return mr_route;
+err:
+ ipmr_cache_put(mfc);
+ list_for_each_entry_safe(rve, tmp, &mr_route->evif_list, route_node)
+ mlxsw_sp_mr_route_evif_unlink(rve);
+ kfree(mr_route);
+ return ERR_PTR(err);
+}
+
+static void mlxsw_sp_mr_route4_destroy(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route *mr_route)
+{
+ struct mlxsw_sp_mr_route_vif_entry *rve, *tmp;
+
+ mlxsw_sp_mr_route_ivif_unlink(mr_route);
+ ipmr_cache_put(mr_route->mfc4);
+ list_for_each_entry_safe(rve, tmp, &mr_route->evif_list, route_node)
+ mlxsw_sp_mr_route_evif_unlink(rve);
+ kfree(mr_route);
+}
+
+static void mlxsw_sp_mr_route_destroy(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route *mr_route)
+{
+ switch (mr_table->proto) {
+ case MLXSW_SP_L3_PROTO_IPV4:
+ mlxsw_sp_mr_route4_destroy(mr_table, mr_route);
+ break;
+ case MLXSW_SP_L3_PROTO_IPV6:
+ /* fall through */
+ default:
+ WARN_ON_ONCE(1);
+ }
+}
+
+static void mlxsw_sp_mr_mfc_offload_set(struct mlxsw_sp_mr_route *mr_route,
+ bool offload)
+{
+ switch (mr_route->mr_table->proto) {
+ case MLXSW_SP_L3_PROTO_IPV4:
+ if (offload)
+ mr_route->mfc4->mfc_flags |= MFC_OFFLOAD;
+ else
+ mr_route->mfc4->mfc_flags &= ~MFC_OFFLOAD;
+ break;
+ case MLXSW_SP_L3_PROTO_IPV6:
+ /* fall through */
+ default:
+ WARN_ON_ONCE(1);
+ }
+}
+
+static void mlxsw_sp_mr_mfc_offload_update(struct mlxsw_sp_mr_route *mr_route)
+{
+ bool offload;
+
+ offload = mr_route->route_action != MLXSW_SP_MR_ROUTE_ACTION_TRAP;
+ mlxsw_sp_mr_mfc_offload_set(mr_route, offload);
+}
+
+static void __mlxsw_sp_mr_route_del(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route *mr_route)
+{
+ mlxsw_sp_mr_mfc_offload_set(mr_route, false);
+ mlxsw_sp_mr_route_erase(mr_table, mr_route);
+ rhashtable_remove_fast(&mr_table->route_ht, &mr_route->ht_node,
+ mlxsw_sp_mr_route_ht_params);
+ list_del(&mr_route->node);
+ mlxsw_sp_mr_route_destroy(mr_table, mr_route);
+}
+
+int mlxsw_sp_mr_route4_add(struct mlxsw_sp_mr_table *mr_table,
+ struct mfc_cache *mfc, bool replace)
+{
+ struct mlxsw_sp_mr_route *mr_orig_route = NULL;
+ struct mlxsw_sp_mr_route *mr_route;
+ int err;
+
+ /* If the route is a (*,*) route, abort, as these kind of routes are
+ * used for proxy routes.
+ */
+ if (mfc->mfc_origin == INADDR_ANY && mfc->mfc_mcastgrp == INADDR_ANY) {
+ dev_warn(mr_table->mlxsw_sp->bus_info->dev,
+ "Offloading proxy routes is not supported.\n");
+ return -EINVAL;
+ }
+
+ /* Create a new route */
+ mr_route = mlxsw_sp_mr_route4_create(mr_table, mfc);
+ if (IS_ERR(mr_route))
+ return PTR_ERR(mr_route);
+
+ /* Find any route with a matching key */
+ mr_orig_route = rhashtable_lookup_fast(&mr_table->route_ht,
+ &mr_route->key,
+ mlxsw_sp_mr_route_ht_params);
+ if (replace) {
+ /* On replace case, make the route point to the new route_priv.
+ */
+ if (WARN_ON(!mr_orig_route)) {
+ err = -ENOENT;
+ goto err_no_orig_route;
+ }
+ mr_route->route_priv = mr_orig_route->route_priv;
+ } else if (mr_orig_route) {
+ /* On non replace case, if another route with the same key was
+ * found, abort, as duplicate routes are used for proxy routes.
+ */
+ dev_warn(mr_table->mlxsw_sp->bus_info->dev,
+ "Offloading proxy routes is not supported.\n");
+ err = -EINVAL;
+ goto err_duplicate_route;
+ }
+
+ /* Put it in the table data-structures */
+ list_add_tail(&mr_route->node, &mr_table->route_list);
+ err = rhashtable_insert_fast(&mr_table->route_ht,
+ &mr_route->ht_node,
+ mlxsw_sp_mr_route_ht_params);
+ if (err)
+ goto err_rhashtable_insert;
+
+ /* Write the route to the hardware */
+ err = mlxsw_sp_mr_route_write(mr_table, mr_route, replace);
+ if (err)
+ goto err_mr_route_write;
+
+ /* Destroy the original route */
+ if (replace) {
+ rhashtable_remove_fast(&mr_table->route_ht,
+ &mr_orig_route->ht_node,
+ mlxsw_sp_mr_route_ht_params);
+ list_del(&mr_orig_route->node);
+ mlxsw_sp_mr_route4_destroy(mr_table, mr_orig_route);
+ }
+
+ mlxsw_sp_mr_mfc_offload_update(mr_route);
+ return 0;
+
+err_mr_route_write:
+ rhashtable_remove_fast(&mr_table->route_ht, &mr_route->ht_node,
+ mlxsw_sp_mr_route_ht_params);
+err_rhashtable_insert:
+ list_del(&mr_route->node);
+err_no_orig_route:
+err_duplicate_route:
+ mlxsw_sp_mr_route4_destroy(mr_table, mr_route);
+ return err;
+}
+
+void mlxsw_sp_mr_route4_del(struct mlxsw_sp_mr_table *mr_table,
+ struct mfc_cache *mfc)
+{
+ struct mlxsw_sp_mr_route *mr_route;
+ struct mlxsw_sp_mr_route_key key;
+
+ mlxsw_sp_mr_route4_key(mr_table, &key, mfc);
+ mr_route = rhashtable_lookup_fast(&mr_table->route_ht, &key,
+ mlxsw_sp_mr_route_ht_params);
+ if (mr_route)
+ __mlxsw_sp_mr_route_del(mr_table, mr_route);
+}
+
+/* Should be called after the VIF struct is updated */
+static int
+mlxsw_sp_mr_route_ivif_resolve(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route_vif_entry *rve)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp;
+ enum mlxsw_sp_mr_route_action route_action;
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+ u16 irif_index;
+ int err;
+
+ route_action = mlxsw_sp_mr_route_action(rve->mr_route);
+ if (route_action == MLXSW_SP_MR_ROUTE_ACTION_TRAP)
+ return 0;
+
+ /* rve->mr_vif->rif is guaranteed to be valid at this stage */
+ irif_index = mlxsw_sp_rif_index(rve->mr_vif->rif);
+ err = mr->mr_ops->route_irif_update(mlxsw_sp, rve->mr_route->route_priv,
+ irif_index);
+ if (err)
+ return err;
+
+ err = mr->mr_ops->route_action_update(mlxsw_sp,
+ rve->mr_route->route_priv,
+ route_action);
+ if (err)
+ /* No need to rollback here because the iRIF change only takes
+ * place after the action has been updated.
+ */
+ return err;
+
+ rve->mr_route->route_action = route_action;
+ mlxsw_sp_mr_mfc_offload_update(rve->mr_route);
+ return 0;
+}
+
+static void
+mlxsw_sp_mr_route_ivif_unresolve(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route_vif_entry *rve)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp;
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+
+ mr->mr_ops->route_action_update(mlxsw_sp, rve->mr_route->route_priv,
+ MLXSW_SP_MR_ROUTE_ACTION_TRAP);
+ rve->mr_route->route_action = MLXSW_SP_MR_ROUTE_ACTION_TRAP;
+ mlxsw_sp_mr_mfc_offload_update(rve->mr_route);
+}
+
+/* Should be called after the RIF struct is updated */
+static int
+mlxsw_sp_mr_route_evif_resolve(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route_vif_entry *rve)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp;
+ enum mlxsw_sp_mr_route_action route_action;
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+ u16 erif_index = 0;
+ int err;
+
+ /* Update the route action, as the new eVIF can be a tunnel or a pimreg
+ * device which will require updating the action.
+ */
+ route_action = mlxsw_sp_mr_route_action(rve->mr_route);
+ if (route_action != rve->mr_route->route_action) {
+ err = mr->mr_ops->route_action_update(mlxsw_sp,
+ rve->mr_route->route_priv,
+ route_action);
+ if (err)
+ return err;
+ }
+
+ /* Add the eRIF */
+ if (mlxsw_sp_mr_vif_valid(rve->mr_vif)) {
+ erif_index = mlxsw_sp_rif_index(rve->mr_vif->rif);
+ err = mr->mr_ops->route_erif_add(mlxsw_sp,
+ rve->mr_route->route_priv,
+ erif_index);
+ if (err)
+ goto err_route_erif_add;
+ }
+
+ /* Update the minimum MTU */
+ if (rve->mr_vif->dev->mtu < rve->mr_route->min_mtu) {
+ rve->mr_route->min_mtu = rve->mr_vif->dev->mtu;
+ err = mr->mr_ops->route_min_mtu_update(mlxsw_sp,
+ rve->mr_route->route_priv,
+ rve->mr_route->min_mtu);
+ if (err)
+ goto err_route_min_mtu_update;
+ }
+
+ rve->mr_route->route_action = route_action;
+ mlxsw_sp_mr_mfc_offload_update(rve->mr_route);
+ return 0;
+
+err_route_min_mtu_update:
+ if (mlxsw_sp_mr_vif_valid(rve->mr_vif))
+ mr->mr_ops->route_erif_del(mlxsw_sp, rve->mr_route->route_priv,
+ erif_index);
+err_route_erif_add:
+ if (route_action != rve->mr_route->route_action)
+ mr->mr_ops->route_action_update(mlxsw_sp,
+ rve->mr_route->route_priv,
+ rve->mr_route->route_action);
+ return err;
+}
+
+/* Should be called before the RIF struct is updated */
+static void
+mlxsw_sp_mr_route_evif_unresolve(struct mlxsw_sp_mr_table *mr_table,
+ struct mlxsw_sp_mr_route_vif_entry *rve)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp;
+ enum mlxsw_sp_mr_route_action route_action;
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+ u16 rifi;
+
+ /* If the unresolved RIF was not valid, no need to delete it */
+ if (!mlxsw_sp_mr_vif_valid(rve->mr_vif))
+ return;
+
+ /* Update the route action: if there is only one valid eVIF in the
+ * route, set the action to trap as the VIF deletion will lead to zero
+ * valid eVIFs. On any other case, use the mlxsw_sp_mr_route_action to
+ * determine the route action.
+ */
+ if (mlxsw_sp_mr_route_valid_evifs_num(rve->mr_route) == 1)
+ route_action = MLXSW_SP_MR_ROUTE_ACTION_TRAP;
+ else
+ route_action = mlxsw_sp_mr_route_action(rve->mr_route);
+ if (route_action != rve->mr_route->route_action)
+ mr->mr_ops->route_action_update(mlxsw_sp,
+ rve->mr_route->route_priv,
+ route_action);
+
+ /* Delete the erif from the route */
+ rifi = mlxsw_sp_rif_index(rve->mr_vif->rif);
+ mr->mr_ops->route_erif_del(mlxsw_sp, rve->mr_route->route_priv, rifi);
+ rve->mr_route->route_action = route_action;
+ mlxsw_sp_mr_mfc_offload_update(rve->mr_route);
+}
+
+static int mlxsw_sp_mr_vif_resolve(struct mlxsw_sp_mr_table *mr_table,
+ struct net_device *dev,
+ struct mlxsw_sp_mr_vif *mr_vif,
+ unsigned long vif_flags,
+ const struct mlxsw_sp_rif *rif)
+{
+ struct mlxsw_sp_mr_route_vif_entry *irve, *erve;
+ int err;
+
+ /* Update the VIF */
+ mr_vif->dev = dev;
+ mr_vif->rif = rif;
+ mr_vif->vif_flags = vif_flags;
+
+ /* Update all routes where this VIF is used as an unresolved iRIF */
+ list_for_each_entry(irve, &mr_vif->route_ivif_list, vif_node) {
+ err = mlxsw_sp_mr_route_ivif_resolve(mr_table, irve);
+ if (err)
+ goto err_irif_unresolve;
+ }
+
+ /* Update all routes where this VIF is used as an unresolved eRIF */
+ list_for_each_entry(erve, &mr_vif->route_evif_list, vif_node) {
+ err = mlxsw_sp_mr_route_evif_resolve(mr_table, erve);
+ if (err)
+ goto err_erif_unresolve;
+ }
+ return 0;
+
+err_erif_unresolve:
+ list_for_each_entry_from_reverse(erve, &mr_vif->route_evif_list,
+ vif_node)
+ mlxsw_sp_mr_route_evif_unresolve(mr_table, erve);
+err_irif_unresolve:
+ list_for_each_entry_from_reverse(irve, &mr_vif->route_ivif_list,
+ vif_node)
+ mlxsw_sp_mr_route_ivif_unresolve(mr_table, irve);
+ mr_vif->rif = NULL;
+ return err;
+}
+
+static void mlxsw_sp_mr_vif_unresolve(struct mlxsw_sp_mr_table *mr_table,
+ struct net_device *dev,
+ struct mlxsw_sp_mr_vif *mr_vif)
+{
+ struct mlxsw_sp_mr_route_vif_entry *rve;
+
+ /* Update all routes where this VIF is used as an unresolved eRIF */
+ list_for_each_entry(rve, &mr_vif->route_evif_list, vif_node)
+ mlxsw_sp_mr_route_evif_unresolve(mr_table, rve);
+
+ /* Update all routes where this VIF is used as an unresolved iRIF */
+ list_for_each_entry(rve, &mr_vif->route_ivif_list, vif_node)
+ mlxsw_sp_mr_route_ivif_unresolve(mr_table, rve);
+
+ /* Update the VIF */
+ mr_vif->dev = dev;
+ mr_vif->rif = NULL;
+}
+
+int mlxsw_sp_mr_vif_add(struct mlxsw_sp_mr_table *mr_table,
+ struct net_device *dev, vifi_t vif_index,
+ unsigned long vif_flags, const struct mlxsw_sp_rif *rif)
+{
+ struct mlxsw_sp_mr_vif *mr_vif = &mr_table->vifs[vif_index];
+
+ if (WARN_ON(vif_index >= MAXVIFS))
+ return -EINVAL;
+ if (mr_vif->dev)
+ return -EEXIST;
+ return mlxsw_sp_mr_vif_resolve(mr_table, dev, mr_vif, vif_flags, rif);
+}
+
+void mlxsw_sp_mr_vif_del(struct mlxsw_sp_mr_table *mr_table, vifi_t vif_index)
+{
+ struct mlxsw_sp_mr_vif *mr_vif = &mr_table->vifs[vif_index];
+
+ if (WARN_ON(vif_index >= MAXVIFS))
+ return;
+ if (WARN_ON(!mr_vif->dev))
+ return;
+ mlxsw_sp_mr_vif_unresolve(mr_table, NULL, mr_vif);
+}
+
+struct mlxsw_sp_mr_vif *
+mlxsw_sp_mr_dev_vif_lookup(struct mlxsw_sp_mr_table *mr_table,
+ const struct net_device *dev)
+{
+ vifi_t vif_index;
+
+ for (vif_index = 0; vif_index < MAXVIFS; vif_index++)
+ if (mr_table->vifs[vif_index].dev == dev)
+ return &mr_table->vifs[vif_index];
+ return NULL;
+}
+
+int mlxsw_sp_mr_rif_add(struct mlxsw_sp_mr_table *mr_table,
+ const struct mlxsw_sp_rif *rif)
+{
+ const struct net_device *rif_dev = mlxsw_sp_rif_dev(rif);
+ struct mlxsw_sp_mr_vif *mr_vif;
+
+ if (!rif_dev)
+ return 0;
+
+ mr_vif = mlxsw_sp_mr_dev_vif_lookup(mr_table, rif_dev);
+ if (!mr_vif)
+ return 0;
+ return mlxsw_sp_mr_vif_resolve(mr_table, mr_vif->dev, mr_vif,
+ mr_vif->vif_flags, rif);
+}
+
+void mlxsw_sp_mr_rif_del(struct mlxsw_sp_mr_table *mr_table,
+ const struct mlxsw_sp_rif *rif)
+{
+ const struct net_device *rif_dev = mlxsw_sp_rif_dev(rif);
+ struct mlxsw_sp_mr_vif *mr_vif;
+
+ if (!rif_dev)
+ return;
+
+ mr_vif = mlxsw_sp_mr_dev_vif_lookup(mr_table, rif_dev);
+ if (!mr_vif)
+ return;
+ mlxsw_sp_mr_vif_unresolve(mr_table, mr_vif->dev, mr_vif);
+}
+
+void mlxsw_sp_mr_rif_mtu_update(struct mlxsw_sp_mr_table *mr_table,
+ const struct mlxsw_sp_rif *rif, int mtu)
+{
+ const struct net_device *rif_dev = mlxsw_sp_rif_dev(rif);
+ struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp;
+ struct mlxsw_sp_mr_route_vif_entry *rve;
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+ struct mlxsw_sp_mr_vif *mr_vif;
+
+ if (!rif_dev)
+ return;
+
+ /* Search for a VIF that use that RIF */
+ mr_vif = mlxsw_sp_mr_dev_vif_lookup(mr_table, rif_dev);
+ if (!mr_vif)
+ return;
+
+ /* Update all the routes that uses that VIF as eVIF */
+ list_for_each_entry(rve, &mr_vif->route_evif_list, vif_node) {
+ if (mtu < rve->mr_route->min_mtu) {
+ rve->mr_route->min_mtu = mtu;
+ mr->mr_ops->route_min_mtu_update(mlxsw_sp,
+ rve->mr_route->route_priv,
+ mtu);
+ }
+ }
+}
+
+struct mlxsw_sp_mr_table *mlxsw_sp_mr_table_create(struct mlxsw_sp *mlxsw_sp,
+ u32 vr_id,
+ enum mlxsw_sp_l3proto proto)
+{
+ struct mlxsw_sp_mr_route_params catchall_route_params = {
+ .prio = MLXSW_SP_MR_ROUTE_PRIO_CATCHALL,
+ .key = {
+ .vrid = vr_id,
+ },
+ .value = {
+ .route_action = MLXSW_SP_MR_ROUTE_ACTION_TRAP,
+ }
+ };
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+ struct mlxsw_sp_mr_table *mr_table;
+ int err;
+ int i;
+
+ mr_table = kzalloc(sizeof(*mr_table) + mr->mr_ops->route_priv_size,
+ GFP_KERNEL);
+ if (!mr_table)
+ return ERR_PTR(-ENOMEM);
+
+ mr_table->vr_id = vr_id;
+ mr_table->mlxsw_sp = mlxsw_sp;
+ mr_table->proto = proto;
+ INIT_LIST_HEAD(&mr_table->route_list);
+
+ err = rhashtable_init(&mr_table->route_ht,
+ &mlxsw_sp_mr_route_ht_params);
+ if (err)
+ goto err_route_rhashtable_init;
+
+ for (i = 0; i < MAXVIFS; i++) {
+ INIT_LIST_HEAD(&mr_table->vifs[i].route_evif_list);
+ INIT_LIST_HEAD(&mr_table->vifs[i].route_ivif_list);
+ }
+
+ err = mr->mr_ops->route_create(mlxsw_sp, mr->priv,
+ mr_table->catchall_route_priv,
+ &catchall_route_params);
+ if (err)
+ goto err_ops_route_create;
+ list_add_tail(&mr_table->node, &mr->table_list);
+ return mr_table;
+
+err_ops_route_create:
+ rhashtable_destroy(&mr_table->route_ht);
+err_route_rhashtable_init:
+ kfree(mr_table);
+ return ERR_PTR(err);
+}
+
+void mlxsw_sp_mr_table_destroy(struct mlxsw_sp_mr_table *mr_table)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_table->mlxsw_sp;
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+
+ WARN_ON(!mlxsw_sp_mr_table_empty(mr_table));
+ list_del(&mr_table->node);
+ mr->mr_ops->route_destroy(mlxsw_sp, mr->priv,
+ &mr_table->catchall_route_priv);
+ rhashtable_destroy(&mr_table->route_ht);
+ kfree(mr_table);
+}
+
+void mlxsw_sp_mr_table_flush(struct mlxsw_sp_mr_table *mr_table)
+{
+ struct mlxsw_sp_mr_route *mr_route, *tmp;
+ int i;
+
+ list_for_each_entry_safe(mr_route, tmp, &mr_table->route_list, node)
+ __mlxsw_sp_mr_route_del(mr_table, mr_route);
+
+ for (i = 0; i < MAXVIFS; i++) {
+ mr_table->vifs[i].dev = NULL;
+ mr_table->vifs[i].rif = NULL;
+ }
+}
+
+bool mlxsw_sp_mr_table_empty(const struct mlxsw_sp_mr_table *mr_table)
+{
+ int i;
+
+ for (i = 0; i < MAXVIFS; i++)
+ if (mr_table->vifs[i].dev)
+ return false;
+ return list_empty(&mr_table->route_list);
+}
+
+static void mlxsw_sp_mr_route_stats_update(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_route *mr_route)
+{
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+ u64 packets, bytes;
+
+ if (mr_route->route_action == MLXSW_SP_MR_ROUTE_ACTION_TRAP)
+ return;
+
+ mr->mr_ops->route_stats(mlxsw_sp, mr_route->route_priv, &packets,
+ &bytes);
+
+ switch (mr_route->mr_table->proto) {
+ case MLXSW_SP_L3_PROTO_IPV4:
+ mr_route->mfc4->mfc_un.res.pkt = packets;
+ mr_route->mfc4->mfc_un.res.bytes = bytes;
+ break;
+ case MLXSW_SP_L3_PROTO_IPV6:
+ /* fall through */
+ default:
+ WARN_ON_ONCE(1);
+ }
+}
+
+static void mlxsw_sp_mr_stats_update(struct work_struct *work)
+{
+ struct mlxsw_sp_mr *mr = container_of(work, struct mlxsw_sp_mr,
+ stats_update_dw.work);
+ struct mlxsw_sp_mr_table *mr_table;
+ struct mlxsw_sp_mr_route *mr_route;
+ unsigned long interval;
+
+ rtnl_lock();
+ list_for_each_entry(mr_table, &mr->table_list, node)
+ list_for_each_entry(mr_route, &mr_table->route_list, node)
+ mlxsw_sp_mr_route_stats_update(mr_table->mlxsw_sp,
+ mr_route);
+ rtnl_unlock();
+
+ interval = msecs_to_jiffies(MLXSW_SP_MR_ROUTES_COUNTER_UPDATE_INTERVAL);
+ mlxsw_core_schedule_dw(&mr->stats_update_dw, interval);
+}
+
+int mlxsw_sp_mr_init(struct mlxsw_sp *mlxsw_sp,
+ const struct mlxsw_sp_mr_ops *mr_ops)
+{
+ struct mlxsw_sp_mr *mr;
+ unsigned long interval;
+ int err;
+
+ mr = kzalloc(sizeof(*mr) + mr_ops->priv_size, GFP_KERNEL);
+ if (!mr)
+ return -ENOMEM;
+ mr->mr_ops = mr_ops;
+ mlxsw_sp->mr = mr;
+ INIT_LIST_HEAD(&mr->table_list);
+
+ err = mr_ops->init(mlxsw_sp, mr->priv);
+ if (err)
+ goto err;
+
+ /* Create the delayed work for counter updates */
+ INIT_DELAYED_WORK(&mr->stats_update_dw, mlxsw_sp_mr_stats_update);
+ interval = msecs_to_jiffies(MLXSW_SP_MR_ROUTES_COUNTER_UPDATE_INTERVAL);
+ mlxsw_core_schedule_dw(&mr->stats_update_dw, interval);
+ return 0;
+err:
+ kfree(mr);
+ return err;
+}
+
+void mlxsw_sp_mr_fini(struct mlxsw_sp *mlxsw_sp)
+{
+ struct mlxsw_sp_mr *mr = mlxsw_sp->mr;
+
+ cancel_delayed_work_sync(&mr->stats_update_dw);
+ mr->mr_ops->fini(mr->priv);
+ kfree(mr);
+}
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h
new file mode 100644
index 0000000..c851b23
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h
@@ -0,0 +1,133 @@
+/*
+ * drivers/net/ethernet/mellanox/mlxsw/spectrum_mr.h
+ * Copyright (c) 2017 Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2017 Yotam Gigi <yotamg@mellanox.com>
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ * 3. Neither the names of the copyright holders nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * Alternatively, this software may be distributed under the terms of the
+ * GNU General Public License ("GPL") version 2 as published by the Free
+ * Software Foundation.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _MLXSW_SPECTRUM_MCROUTER_H
+#define _MLXSW_SPECTRUM_MCROUTER_H
+
+#include <linux/mroute.h>
+#include "spectrum_router.h"
+#include "spectrum.h"
+
+enum mlxsw_sp_mr_route_action {
+ MLXSW_SP_MR_ROUTE_ACTION_FORWARD,
+ MLXSW_SP_MR_ROUTE_ACTION_TRAP,
+};
+
+enum mlxsw_sp_mr_route_prio {
+ MLXSW_SP_MR_ROUTE_PRIO_SG,
+ MLXSW_SP_MR_ROUTE_PRIO_STARG,
+ MLXSW_SP_MR_ROUTE_PRIO_CATCHALL,
+ __MLXSW_SP_MR_ROUTE_PRIO_MAX
+};
+
+#define MLXSW_SP_MR_ROUTE_PRIO_MAX (__MLXSW_SP_MR_ROUTE_PRIO_MAX - 1)
+
+struct mlxsw_sp_mr_route_key {
+ int vrid;
+ enum mlxsw_sp_l3proto proto;
+ union mlxsw_sp_l3addr group;
+ union mlxsw_sp_l3addr group_mask;
+ union mlxsw_sp_l3addr source;
+ union mlxsw_sp_l3addr source_mask;
+};
+
+struct mlxsw_sp_mr_route_info {
+ enum mlxsw_sp_mr_route_action route_action;
+ u16 irif_index;
+ u16 *erif_indices;
+ size_t erif_num;
+ u16 min_mtu;
+};
+
+struct mlxsw_sp_mr_route_params {
+ struct mlxsw_sp_mr_route_key key;
+ struct mlxsw_sp_mr_route_info value;
+ enum mlxsw_sp_mr_route_prio prio;
+};
+
+struct mlxsw_sp_mr_ops {
+ int priv_size;
+ int route_priv_size;
+ int (*init)(struct mlxsw_sp *mlxsw_sp, void *priv);
+ int (*route_create)(struct mlxsw_sp *mlxsw_sp, void *priv,
+ void *route_priv,
+ struct mlxsw_sp_mr_route_params *route_params);
+ int (*route_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv,
+ struct mlxsw_sp_mr_route_info *route_info);
+ int (*route_stats)(struct mlxsw_sp *mlxsw_sp, void *route_priv,
+ u64 *packets, u64 *bytes);
+ int (*route_action_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv,
+ enum mlxsw_sp_mr_route_action route_action);
+ int (*route_min_mtu_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv,
+ u16 min_mtu);
+ int (*route_irif_update)(struct mlxsw_sp *mlxsw_sp, void *route_priv,
+ u16 irif_index);
+ int (*route_erif_add)(struct mlxsw_sp *mlxsw_sp, void *route_priv,
+ u16 erif_index);
+ int (*route_erif_del)(struct mlxsw_sp *mlxsw_sp, void *route_priv,
+ u16 erif_index);
+ void (*route_destroy)(struct mlxsw_sp *mlxsw_sp, void *priv,
+ void *route_priv);
+ void (*fini)(void *priv);
+};
+
+struct mlxsw_sp_mr;
+struct mlxsw_sp_mr_table;
+
+int mlxsw_sp_mr_init(struct mlxsw_sp *mlxsw_sp,
+ const struct mlxsw_sp_mr_ops *mr_ops);
+void mlxsw_sp_mr_fini(struct mlxsw_sp *mlxsw_sp);
+int mlxsw_sp_mr_route4_add(struct mlxsw_sp_mr_table *mr_table,
+ struct mfc_cache *mfc, bool replace);
+void mlxsw_sp_mr_route4_del(struct mlxsw_sp_mr_table *mr_table,
+ struct mfc_cache *mfc);
+int mlxsw_sp_mr_vif_add(struct mlxsw_sp_mr_table *mr_table,
+ struct net_device *dev, vifi_t vif_index,
+ unsigned long vif_flags,
+ const struct mlxsw_sp_rif *rif);
+void mlxsw_sp_mr_vif_del(struct mlxsw_sp_mr_table *mr_table, vifi_t vif_index);
+int mlxsw_sp_mr_rif_add(struct mlxsw_sp_mr_table *mr_table,
+ const struct mlxsw_sp_rif *rif);
+void mlxsw_sp_mr_rif_del(struct mlxsw_sp_mr_table *mr_table,
+ const struct mlxsw_sp_rif *rif);
+void mlxsw_sp_mr_rif_mtu_update(struct mlxsw_sp_mr_table *mr_table,
+ const struct mlxsw_sp_rif *rif, int mtu);
+struct mlxsw_sp_mr_table *mlxsw_sp_mr_table_create(struct mlxsw_sp *mlxsw_sp,
+ u32 tb_id,
+ enum mlxsw_sp_l3proto proto);
+void mlxsw_sp_mr_table_destroy(struct mlxsw_sp_mr_table *mr_table);
+void mlxsw_sp_mr_table_flush(struct mlxsw_sp_mr_table *mr_table);
+bool mlxsw_sp_mr_table_empty(const struct mlxsw_sp_mr_table *mr_table);
+
+#endif
--
2.9.5
^ permalink raw reply related
* [patch net-next 08/12] mlxsw: spectrum: Add the multicast routing hardware logic
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Implement the multicast routing hardware API introduced in previous patch
for the specific spectrum hardware.
The spectrum hardware multicast routes are written using the RMFT2 register
and point to an ACL flexible action set. The actions used for multicast
routes are:
- Counter action, which allows counting bytes and packets on multicast
routes.
- Multicast route action, which provide RPF check and do the actual packet
duplication to a list of RIFs.
- Trap action, in the case the route action specified by the called is
trap.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
drivers/net/ethernet/mellanox/mlxsw/Makefile | 2 +-
drivers/net/ethernet/mellanox/mlxsw/spectrum.h | 1 +
.../net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c | 828 +++++++++++++++++++++
.../net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h | 43 ++
4 files changed, 873 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c
create mode 100644 drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h
diff --git a/drivers/net/ethernet/mellanox/mlxsw/Makefile b/drivers/net/ethernet/mellanox/mlxsw/Makefile
index 9b29764..4816504 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/Makefile
+++ b/drivers/net/ethernet/mellanox/mlxsw/Makefile
@@ -18,7 +18,7 @@ mlxsw_spectrum-objs := spectrum.o spectrum_buffers.o \
spectrum_acl.o spectrum_flower.o \
spectrum_cnt.o spectrum_fid.o \
spectrum_ipip.o spectrum_acl_flex_actions.o \
- spectrum_mr.o
+ spectrum_mr.o spectrum_mr_tcam.o
mlxsw_spectrum-$(CONFIG_MLXSW_SPECTRUM_DCB) += spectrum_dcb.o
mlxsw_spectrum-$(CONFIG_NET_DEVLINK) += spectrum_dpipe.o
obj-$(CONFIG_MLXSW_MINIMAL) += mlxsw_minimal.o
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
index 51d8b9f..d06f7fe 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
@@ -139,6 +139,7 @@ struct mlxsw_sp_port_mall_tc_entry {
struct mlxsw_sp_sb;
struct mlxsw_sp_bridge;
struct mlxsw_sp_router;
+struct mlxsw_sp_mr;
struct mlxsw_sp_acl;
struct mlxsw_sp_counter_pool;
struct mlxsw_sp_fid_core;
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c
new file mode 100644
index 0000000..cda9e9a
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c
@@ -0,0 +1,828 @@
+/*
+ * drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c
+ * Copyright (c) 2017 Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2017 Yotam Gigi <yotamg@mellanox.com>
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ * 3. Neither the names of the copyright holders nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * Alternatively, this software may be distributed under the terms of the
+ * GNU General Public License ("GPL") version 2 as published by the Free
+ * Software Foundation.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/kernel.h>
+#include <linux/list.h>
+#include <linux/netdevice.h>
+#include <linux/parman.h>
+
+#include "reg.h"
+#include "spectrum.h"
+#include "core_acl_flex_actions.h"
+#include "spectrum_mr.h"
+
+struct mlxsw_sp_mr_tcam_region {
+ struct mlxsw_sp *mlxsw_sp;
+ enum mlxsw_reg_rtar_key_type rtar_key_type;
+ struct parman *parman;
+ struct parman_prio *parman_prios;
+};
+
+struct mlxsw_sp_mr_tcam {
+ struct mlxsw_sp_mr_tcam_region ipv4_tcam_region;
+};
+
+/* This struct maps to one RIGR2 register entry */
+struct mlxsw_sp_mr_erif_sublist {
+ struct list_head list;
+ u32 rigr2_kvdl_index;
+ int num_erifs;
+ u16 erif_indices[MLXSW_REG_RIGR2_MAX_ERIFS];
+ bool synced;
+};
+
+struct mlxsw_sp_mr_tcam_erif_list {
+ struct list_head erif_sublists;
+ u32 kvdl_index;
+};
+
+static bool
+mlxsw_sp_mr_erif_sublist_full(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_erif_sublist *erif_sublist)
+{
+ int erif_list_entries = MLXSW_CORE_RES_GET(mlxsw_sp->core,
+ MC_ERIF_LIST_ENTRIES);
+
+ return erif_sublist->num_erifs == erif_list_entries;
+}
+
+static void
+mlxsw_sp_mr_erif_list_init(struct mlxsw_sp_mr_tcam_erif_list *erif_list)
+{
+ INIT_LIST_HEAD(&erif_list->erif_sublists);
+}
+
+#define MLXSW_SP_KVDL_RIGR2_SIZE 1
+
+static struct mlxsw_sp_mr_erif_sublist *
+mlxsw_sp_mr_erif_sublist_create(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_tcam_erif_list *erif_list)
+{
+ struct mlxsw_sp_mr_erif_sublist *erif_sublist;
+ int err;
+
+ erif_sublist = kzalloc(sizeof(*erif_sublist), GFP_KERNEL);
+ if (!erif_sublist)
+ return ERR_PTR(-ENOMEM);
+ err = mlxsw_sp_kvdl_alloc(mlxsw_sp, MLXSW_SP_KVDL_RIGR2_SIZE,
+ &erif_sublist->rigr2_kvdl_index);
+ if (err) {
+ kfree(erif_sublist);
+ return ERR_PTR(err);
+ }
+
+ list_add_tail(&erif_sublist->list, &erif_list->erif_sublists);
+ return erif_sublist;
+}
+
+static void
+mlxsw_sp_mr_erif_sublist_destroy(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_erif_sublist *erif_sublist)
+{
+ list_del(&erif_sublist->list);
+ mlxsw_sp_kvdl_free(mlxsw_sp, erif_sublist->rigr2_kvdl_index);
+ kfree(erif_sublist);
+}
+
+static int
+mlxsw_sp_mr_erif_list_add(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_tcam_erif_list *erif_list,
+ u16 erif_index)
+{
+ struct mlxsw_sp_mr_erif_sublist *sublist;
+
+ /* If either there is no erif_entry or the last one is full, allocate a
+ * new one.
+ */
+ if (list_empty(&erif_list->erif_sublists)) {
+ sublist = mlxsw_sp_mr_erif_sublist_create(mlxsw_sp, erif_list);
+ if (IS_ERR(sublist))
+ return PTR_ERR(sublist);
+ erif_list->kvdl_index = sublist->rigr2_kvdl_index;
+ } else {
+ sublist = list_last_entry(&erif_list->erif_sublists,
+ struct mlxsw_sp_mr_erif_sublist,
+ list);
+ sublist->synced = false;
+ if (mlxsw_sp_mr_erif_sublist_full(mlxsw_sp, sublist)) {
+ sublist = mlxsw_sp_mr_erif_sublist_create(mlxsw_sp,
+ erif_list);
+ if (IS_ERR(sublist))
+ return PTR_ERR(sublist);
+ }
+ }
+
+ /* Add the eRIF to the last entry's last index */
+ sublist->erif_indices[sublist->num_erifs++] = erif_index;
+ return 0;
+}
+
+static void
+mlxsw_sp_mr_erif_list_flush(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_tcam_erif_list *erif_list)
+{
+ struct mlxsw_sp_mr_erif_sublist *erif_sublist, *tmp;
+
+ list_for_each_entry_safe(erif_sublist, tmp, &erif_list->erif_sublists,
+ list)
+ mlxsw_sp_mr_erif_sublist_destroy(mlxsw_sp, erif_sublist);
+}
+
+static int
+mlxsw_sp_mr_erif_list_commit(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_tcam_erif_list *erif_list)
+{
+ struct mlxsw_sp_mr_erif_sublist *curr_sublist;
+ char rigr2_pl[MLXSW_REG_RIGR2_LEN];
+ int err;
+ int i;
+
+ list_for_each_entry(curr_sublist, &erif_list->erif_sublists, list) {
+ if (curr_sublist->synced)
+ continue;
+
+ /* If the sublist is not the last one, pack the next index */
+ if (list_is_last(&curr_sublist->list,
+ &erif_list->erif_sublists)) {
+ mlxsw_reg_rigr2_pack(rigr2_pl,
+ curr_sublist->rigr2_kvdl_index,
+ false, 0);
+ } else {
+ struct mlxsw_sp_mr_erif_sublist *next_sublist;
+
+ next_sublist = list_next_entry(curr_sublist, list);
+ mlxsw_reg_rigr2_pack(rigr2_pl,
+ curr_sublist->rigr2_kvdl_index,
+ true,
+ next_sublist->rigr2_kvdl_index);
+ }
+
+ /* Pack all the erifs */
+ for (i = 0; i < curr_sublist->num_erifs; i++) {
+ u16 erif_index = curr_sublist->erif_indices[i];
+
+ mlxsw_reg_rigr2_erif_entry_pack(rigr2_pl, i, true,
+ erif_index);
+ }
+
+ /* Write the entry */
+ err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rigr2),
+ rigr2_pl);
+ if (err)
+ /* No need of a rollback here because this
+ * hardware entry should not be pointed yet.
+ */
+ return err;
+ curr_sublist->synced = true;
+ }
+ return 0;
+}
+
+static void mlxsw_sp_mr_erif_list_move(struct mlxsw_sp_mr_tcam_erif_list *to,
+ struct mlxsw_sp_mr_tcam_erif_list *from)
+{
+ list_splice(&from->erif_sublists, &to->erif_sublists);
+ to->kvdl_index = from->kvdl_index;
+}
+
+struct mlxsw_sp_mr_tcam_route {
+ struct mlxsw_sp_mr_tcam_erif_list erif_list;
+ struct mlxsw_afa_block *afa_block;
+ u32 counter_index;
+ struct parman_item parman_item;
+ struct parman_prio *parman_prio;
+ enum mlxsw_sp_mr_route_action action;
+ struct mlxsw_sp_mr_route_key key;
+ u16 irif_index;
+ u16 min_mtu;
+};
+
+static struct mlxsw_afa_block *
+mlxsw_sp_mr_tcam_afa_block_create(struct mlxsw_sp *mlxsw_sp,
+ enum mlxsw_sp_mr_route_action route_action,
+ u16 irif_index, u32 counter_index,
+ u16 min_mtu,
+ struct mlxsw_sp_mr_tcam_erif_list *erif_list)
+{
+ struct mlxsw_afa_block *afa_block;
+ int err;
+
+ afa_block = mlxsw_afa_block_create(mlxsw_sp->afa);
+ if (IS_ERR(afa_block))
+ return afa_block;
+
+ err = mlxsw_afa_block_append_counter(afa_block, counter_index);
+ if (err)
+ goto err;
+
+ switch (route_action) {
+ case MLXSW_SP_MR_ROUTE_ACTION_TRAP:
+ err = mlxsw_afa_block_append_trap(afa_block,
+ MLXSW_TRAP_ID_ACL1);
+ if (err)
+ goto err;
+ break;
+ case MLXSW_SP_MR_ROUTE_ACTION_FORWARD:
+ /* If we are about to append a multicast router action, commit
+ * the erif_list.
+ */
+ err = mlxsw_sp_mr_erif_list_commit(mlxsw_sp, erif_list);
+ if (err)
+ goto err;
+
+ err = mlxsw_afa_block_append_mcrouter(afa_block, irif_index,
+ min_mtu, false,
+ erif_list->kvdl_index);
+ if (err)
+ goto err;
+ break;
+ default:
+ err = -EINVAL;
+ goto err;
+ }
+
+ err = mlxsw_afa_block_commit(afa_block);
+ if (err)
+ goto err;
+ return afa_block;
+err:
+ mlxsw_afa_block_destroy(afa_block);
+ return ERR_PTR(err);
+}
+
+static void
+mlxsw_sp_mr_tcam_afa_block_destroy(struct mlxsw_afa_block *afa_block)
+{
+ mlxsw_afa_block_destroy(afa_block);
+}
+
+static int mlxsw_sp_mr_tcam_route_replace(struct mlxsw_sp *mlxsw_sp,
+ struct parman_item *parman_item,
+ struct mlxsw_sp_mr_route_key *key,
+ struct mlxsw_afa_block *afa_block)
+{
+ char rmft2_pl[MLXSW_REG_RMFT2_LEN];
+
+ switch (key->proto) {
+ case MLXSW_SP_L3_PROTO_IPV4:
+ mlxsw_reg_rmft2_ipv4_pack(rmft2_pl, true, parman_item->index,
+ key->vrid,
+ MLXSW_REG_RMFT2_IRIF_MASK_IGNORE, 0,
+ ntohl(key->group.addr4),
+ ntohl(key->group_mask.addr4),
+ ntohl(key->source.addr4),
+ ntohl(key->source_mask.addr4),
+ mlxsw_afa_block_first_set(afa_block));
+ break;
+ case MLXSW_SP_L3_PROTO_IPV6:
+ default:
+ WARN_ON_ONCE(1);
+ }
+
+ return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rmft2), rmft2_pl);
+}
+
+static int mlxsw_sp_mr_tcam_route_remove(struct mlxsw_sp *mlxsw_sp, int vrid,
+ struct parman_item *parman_item)
+{
+ char rmft2_pl[MLXSW_REG_RMFT2_LEN];
+
+ mlxsw_reg_rmft2_ipv4_pack(rmft2_pl, false, parman_item->index, vrid,
+ 0, 0, 0, 0, 0, 0, NULL);
+
+ return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rmft2), rmft2_pl);
+}
+
+static int
+mlxsw_sp_mr_tcam_erif_populate(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_tcam_erif_list *erif_list,
+ struct mlxsw_sp_mr_route_info *route_info)
+{
+ int err;
+ int i;
+
+ for (i = 0; i < route_info->erif_num; i++) {
+ u16 erif_index = route_info->erif_indices[i];
+
+ err = mlxsw_sp_mr_erif_list_add(mlxsw_sp, erif_list,
+ erif_index);
+ if (err)
+ return err;
+ }
+ return 0;
+}
+
+static int
+mlxsw_sp_mr_tcam_route_parman_item_add(struct mlxsw_sp_mr_tcam *mr_tcam,
+ struct mlxsw_sp_mr_tcam_route *route,
+ enum mlxsw_sp_mr_route_prio prio)
+{
+ struct parman_prio *parman_prio = NULL;
+ int err;
+
+ switch (route->key.proto) {
+ case MLXSW_SP_L3_PROTO_IPV4:
+ parman_prio = &mr_tcam->ipv4_tcam_region.parman_prios[prio];
+ err = parman_item_add(mr_tcam->ipv4_tcam_region.parman,
+ parman_prio, &route->parman_item);
+ if (err)
+ return err;
+ break;
+ case MLXSW_SP_L3_PROTO_IPV6:
+ default:
+ WARN_ON_ONCE(1);
+ }
+ route->parman_prio = parman_prio;
+ return 0;
+}
+
+static void
+mlxsw_sp_mr_tcam_route_parman_item_remove(struct mlxsw_sp_mr_tcam *mr_tcam,
+ struct mlxsw_sp_mr_tcam_route *route)
+{
+ switch (route->key.proto) {
+ case MLXSW_SP_L3_PROTO_IPV4:
+ parman_item_remove(mr_tcam->ipv4_tcam_region.parman,
+ route->parman_prio, &route->parman_item);
+ break;
+ case MLXSW_SP_L3_PROTO_IPV6:
+ default:
+ WARN_ON_ONCE(1);
+ }
+}
+
+static int
+mlxsw_sp_mr_tcam_route_create(struct mlxsw_sp *mlxsw_sp, void *priv,
+ void *route_priv,
+ struct mlxsw_sp_mr_route_params *route_params)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+ struct mlxsw_sp_mr_tcam *mr_tcam = priv;
+ int err;
+
+ route->key = route_params->key;
+ route->irif_index = route_params->value.irif_index;
+ route->min_mtu = route_params->value.min_mtu;
+ route->action = route_params->value.route_action;
+
+ /* Create the egress RIFs list */
+ mlxsw_sp_mr_erif_list_init(&route->erif_list);
+ err = mlxsw_sp_mr_tcam_erif_populate(mlxsw_sp, &route->erif_list,
+ &route_params->value);
+ if (err)
+ goto err_erif_populate;
+
+ /* Create the flow counter */
+ err = mlxsw_sp_flow_counter_alloc(mlxsw_sp, &route->counter_index);
+ if (err)
+ goto err_counter_alloc;
+
+ /* Create the flexible action block */
+ route->afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp,
+ route->action,
+ route->irif_index,
+ route->counter_index,
+ route->min_mtu,
+ &route->erif_list);
+ if (IS_ERR(route->afa_block)) {
+ err = PTR_ERR(route->afa_block);
+ goto err_afa_block_create;
+ }
+
+ /* Allocate place in the TCAM */
+ err = mlxsw_sp_mr_tcam_route_parman_item_add(mr_tcam, route,
+ route_params->prio);
+ if (err)
+ goto err_parman_item_add;
+
+ /* Write the route to the TCAM */
+ err = mlxsw_sp_mr_tcam_route_replace(mlxsw_sp, &route->parman_item,
+ &route->key, route->afa_block);
+ if (err)
+ goto err_route_replace;
+ return 0;
+
+err_route_replace:
+ mlxsw_sp_mr_tcam_route_parman_item_remove(mr_tcam, route);
+err_parman_item_add:
+ mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block);
+err_afa_block_create:
+ mlxsw_sp_flow_counter_free(mlxsw_sp, route->counter_index);
+err_erif_populate:
+err_counter_alloc:
+ mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &route->erif_list);
+ return err;
+}
+
+static void mlxsw_sp_mr_tcam_route_destroy(struct mlxsw_sp *mlxsw_sp,
+ void *priv, void *route_priv)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+ struct mlxsw_sp_mr_tcam *mr_tcam = priv;
+
+ mlxsw_sp_mr_tcam_route_remove(mlxsw_sp, route->key.vrid,
+ &route->parman_item);
+ mlxsw_sp_mr_tcam_route_parman_item_remove(mr_tcam, route);
+ mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block);
+ mlxsw_sp_flow_counter_free(mlxsw_sp, route->counter_index);
+ mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &route->erif_list);
+}
+
+static int mlxsw_sp_mr_tcam_route_stats(struct mlxsw_sp *mlxsw_sp,
+ void *route_priv, u64 *packets,
+ u64 *bytes)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+
+ return mlxsw_sp_flow_counter_get(mlxsw_sp, route->counter_index,
+ packets, bytes);
+}
+
+static int
+mlxsw_sp_mr_tcam_route_action_update(struct mlxsw_sp *mlxsw_sp,
+ void *route_priv,
+ enum mlxsw_sp_mr_route_action route_action)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+ struct mlxsw_afa_block *afa_block;
+ int err;
+
+ /* Create a new flexible action block */
+ afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp, route_action,
+ route->irif_index,
+ route->counter_index,
+ route->min_mtu,
+ &route->erif_list);
+ if (IS_ERR(afa_block))
+ return PTR_ERR(afa_block);
+
+ /* Update the TCAM route entry */
+ err = mlxsw_sp_mr_tcam_route_replace(mlxsw_sp, &route->parman_item,
+ &route->key, afa_block);
+ if (err)
+ goto err;
+
+ /* Delete the old one */
+ mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block);
+ route->afa_block = afa_block;
+ route->action = route_action;
+ return 0;
+err:
+ mlxsw_sp_mr_tcam_afa_block_destroy(afa_block);
+ return err;
+}
+
+static int mlxsw_sp_mr_tcam_route_min_mtu_update(struct mlxsw_sp *mlxsw_sp,
+ void *route_priv, u16 min_mtu)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+ struct mlxsw_afa_block *afa_block;
+ int err;
+
+ /* Create a new flexible action block */
+ afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp,
+ route->action,
+ route->irif_index,
+ route->counter_index,
+ min_mtu,
+ &route->erif_list);
+ if (IS_ERR(afa_block))
+ return PTR_ERR(afa_block);
+
+ /* Update the TCAM route entry */
+ err = mlxsw_sp_mr_tcam_route_replace(mlxsw_sp, &route->parman_item,
+ &route->key, afa_block);
+ if (err)
+ goto err;
+
+ /* Delete the old one */
+ mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block);
+ route->afa_block = afa_block;
+ route->min_mtu = min_mtu;
+ return 0;
+err:
+ mlxsw_sp_mr_tcam_afa_block_destroy(afa_block);
+ return err;
+}
+
+static int mlxsw_sp_mr_tcam_route_irif_update(struct mlxsw_sp *mlxsw_sp,
+ void *route_priv, u16 irif_index)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+
+ if (route->action != MLXSW_SP_MR_ROUTE_ACTION_TRAP)
+ return -EINVAL;
+ route->irif_index = irif_index;
+ return 0;
+}
+
+static int mlxsw_sp_mr_tcam_route_erif_add(struct mlxsw_sp *mlxsw_sp,
+ void *route_priv, u16 erif_index)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+ int err;
+
+ err = mlxsw_sp_mr_erif_list_add(mlxsw_sp, &route->erif_list,
+ erif_index);
+ if (err)
+ return err;
+
+ /* Commit the action only if the route action is not TRAP */
+ if (route->action != MLXSW_SP_MR_ROUTE_ACTION_TRAP)
+ return mlxsw_sp_mr_erif_list_commit(mlxsw_sp,
+ &route->erif_list);
+ return 0;
+}
+
+static int mlxsw_sp_mr_tcam_route_erif_del(struct mlxsw_sp *mlxsw_sp,
+ void *route_priv, u16 erif_index)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+ struct mlxsw_sp_mr_erif_sublist *erif_sublist;
+ struct mlxsw_sp_mr_tcam_erif_list erif_list;
+ struct mlxsw_afa_block *afa_block;
+ int err;
+ int i;
+
+ /* Create a copy of the original erif_list without the deleted entry */
+ mlxsw_sp_mr_erif_list_init(&erif_list);
+ list_for_each_entry(erif_sublist, &route->erif_list.erif_sublists, list) {
+ for (i = 0; i < erif_sublist->num_erifs; i++) {
+ u16 curr_erif = erif_sublist->erif_indices[i];
+
+ if (curr_erif == erif_index)
+ continue;
+ err = mlxsw_sp_mr_erif_list_add(mlxsw_sp, &erif_list,
+ curr_erif);
+ if (err)
+ goto err_erif_list_add;
+ }
+ }
+
+ /* Create the flexible action block pointing to the new erif_list */
+ afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp, route->action,
+ route->irif_index,
+ route->counter_index,
+ route->min_mtu,
+ &erif_list);
+ if (IS_ERR(afa_block)) {
+ err = PTR_ERR(afa_block);
+ goto err_afa_block_create;
+ }
+
+ /* Update the TCAM route entry */
+ err = mlxsw_sp_mr_tcam_route_replace(mlxsw_sp, &route->parman_item,
+ &route->key, afa_block);
+ if (err)
+ goto err_route_write;
+
+ mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block);
+ mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &route->erif_list);
+ route->afa_block = afa_block;
+ mlxsw_sp_mr_erif_list_move(&route->erif_list, &erif_list);
+ return 0;
+
+err_route_write:
+ mlxsw_sp_mr_tcam_afa_block_destroy(afa_block);
+err_afa_block_create:
+err_erif_list_add:
+ mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &erif_list);
+ return err;
+}
+
+static int
+mlxsw_sp_mr_tcam_route_update(struct mlxsw_sp *mlxsw_sp, void *route_priv,
+ struct mlxsw_sp_mr_route_info *route_info)
+{
+ struct mlxsw_sp_mr_tcam_route *route = route_priv;
+ struct mlxsw_sp_mr_tcam_erif_list erif_list;
+ struct mlxsw_afa_block *afa_block;
+ int err;
+
+ /* Create a new erif_list */
+ mlxsw_sp_mr_erif_list_init(&erif_list);
+ err = mlxsw_sp_mr_tcam_erif_populate(mlxsw_sp, &erif_list, route_info);
+ if (err)
+ goto err_erif_populate;
+
+ /* Create the flexible action block pointing to the new erif_list */
+ afa_block = mlxsw_sp_mr_tcam_afa_block_create(mlxsw_sp,
+ route_info->route_action,
+ route_info->irif_index,
+ route->counter_index,
+ route_info->min_mtu,
+ &erif_list);
+ if (IS_ERR(afa_block)) {
+ err = PTR_ERR(afa_block);
+ goto err_afa_block_create;
+ }
+
+ /* Update the TCAM route entry */
+ err = mlxsw_sp_mr_tcam_route_replace(mlxsw_sp, &route->parman_item,
+ &route->key, afa_block);
+ if (err)
+ goto err_route_write;
+
+ mlxsw_sp_mr_tcam_afa_block_destroy(route->afa_block);
+ mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &route->erif_list);
+ route->afa_block = afa_block;
+ mlxsw_sp_mr_erif_list_move(&route->erif_list, &erif_list);
+ route->action = route_info->route_action;
+ route->irif_index = route_info->irif_index;
+ route->min_mtu = route_info->min_mtu;
+ return 0;
+
+err_route_write:
+ mlxsw_sp_mr_tcam_afa_block_destroy(afa_block);
+err_afa_block_create:
+err_erif_populate:
+ mlxsw_sp_mr_erif_list_flush(mlxsw_sp, &erif_list);
+ return err;
+}
+
+#define MLXSW_SP_MR_TCAM_REGION_BASE_COUNT 16
+#define MLXSW_SP_MR_TCAM_REGION_RESIZE_STEP 16
+
+static int
+mlxsw_sp_mr_tcam_region_alloc(struct mlxsw_sp_mr_tcam_region *mr_tcam_region)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_tcam_region->mlxsw_sp;
+ char rtar_pl[MLXSW_REG_RTAR_LEN];
+
+ mlxsw_reg_rtar_pack(rtar_pl, MLXSW_REG_RTAR_OP_ALLOCATE,
+ mr_tcam_region->rtar_key_type,
+ MLXSW_SP_MR_TCAM_REGION_BASE_COUNT);
+ return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtar), rtar_pl);
+}
+
+static void
+mlxsw_sp_mr_tcam_region_free(struct mlxsw_sp_mr_tcam_region *mr_tcam_region)
+{
+ struct mlxsw_sp *mlxsw_sp = mr_tcam_region->mlxsw_sp;
+ char rtar_pl[MLXSW_REG_RTAR_LEN];
+
+ mlxsw_reg_rtar_pack(rtar_pl, MLXSW_REG_RTAR_OP_DEALLOCATE,
+ mr_tcam_region->rtar_key_type, 0);
+ mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtar), rtar_pl);
+}
+
+static int mlxsw_sp_mr_tcam_region_parman_resize(void *priv,
+ unsigned long new_count)
+{
+ struct mlxsw_sp_mr_tcam_region *mr_tcam_region = priv;
+ struct mlxsw_sp *mlxsw_sp = mr_tcam_region->mlxsw_sp;
+ char rtar_pl[MLXSW_REG_RTAR_LEN];
+ u64 max_tcam_rules;
+
+ max_tcam_rules = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_TCAM_RULES);
+ if (new_count > max_tcam_rules)
+ return -EINVAL;
+ mlxsw_reg_rtar_pack(rtar_pl, MLXSW_REG_RTAR_OP_RESIZE,
+ mr_tcam_region->rtar_key_type, new_count);
+ return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtar), rtar_pl);
+}
+
+static void mlxsw_sp_mr_tcam_region_parman_move(void *priv,
+ unsigned long from_index,
+ unsigned long to_index,
+ unsigned long count)
+{
+ struct mlxsw_sp_mr_tcam_region *mr_tcam_region = priv;
+ struct mlxsw_sp *mlxsw_sp = mr_tcam_region->mlxsw_sp;
+ char rrcr_pl[MLXSW_REG_RRCR_LEN];
+
+ mlxsw_reg_rrcr_pack(rrcr_pl, MLXSW_REG_RRCR_OP_MOVE,
+ from_index, count,
+ mr_tcam_region->rtar_key_type, to_index);
+ mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rrcr), rrcr_pl);
+}
+
+static const struct parman_ops mlxsw_sp_mr_tcam_region_parman_ops = {
+ .base_count = MLXSW_SP_MR_TCAM_REGION_BASE_COUNT,
+ .resize_step = MLXSW_SP_MR_TCAM_REGION_RESIZE_STEP,
+ .resize = mlxsw_sp_mr_tcam_region_parman_resize,
+ .move = mlxsw_sp_mr_tcam_region_parman_move,
+ .algo = PARMAN_ALGO_TYPE_LSORT,
+};
+
+static int
+mlxsw_sp_mr_tcam_region_init(struct mlxsw_sp *mlxsw_sp,
+ struct mlxsw_sp_mr_tcam_region *mr_tcam_region,
+ enum mlxsw_reg_rtar_key_type rtar_key_type)
+{
+ struct parman_prio *parman_prios;
+ struct parman *parman;
+ int err;
+ int i;
+
+ mr_tcam_region->rtar_key_type = rtar_key_type;
+ mr_tcam_region->mlxsw_sp = mlxsw_sp;
+
+ err = mlxsw_sp_mr_tcam_region_alloc(mr_tcam_region);
+ if (err)
+ return err;
+
+ parman = parman_create(&mlxsw_sp_mr_tcam_region_parman_ops,
+ mr_tcam_region);
+ if (!parman) {
+ err = -ENOMEM;
+ goto err_parman_create;
+ }
+ mr_tcam_region->parman = parman;
+
+ parman_prios = kmalloc_array(MLXSW_SP_MR_ROUTE_PRIO_MAX + 1,
+ sizeof(*parman_prios), GFP_KERNEL);
+ if (!parman_prios)
+ goto err_parman_prios_alloc;
+ mr_tcam_region->parman_prios = parman_prios;
+
+ for (i = 0; i < MLXSW_SP_MR_ROUTE_PRIO_MAX + 1; i++)
+ parman_prio_init(mr_tcam_region->parman,
+ &mr_tcam_region->parman_prios[i], i);
+ return 0;
+
+err_parman_prios_alloc:
+ parman_destroy(parman);
+err_parman_create:
+ mlxsw_sp_mr_tcam_region_free(mr_tcam_region);
+ return err;
+}
+
+static void
+mlxsw_sp_mr_tcam_region_fini(struct mlxsw_sp_mr_tcam_region *mr_tcam_region)
+{
+ int i;
+
+ for (i = 0; i < MLXSW_SP_MR_ROUTE_PRIO_MAX + 1; i++)
+ parman_prio_fini(&mr_tcam_region->parman_prios[i]);
+ kfree(mr_tcam_region->parman_prios);
+ parman_destroy(mr_tcam_region->parman);
+ mlxsw_sp_mr_tcam_region_free(mr_tcam_region);
+}
+
+static int mlxsw_sp_mr_tcam_init(struct mlxsw_sp *mlxsw_sp, void *priv)
+{
+ struct mlxsw_sp_mr_tcam *mr_tcam = priv;
+
+ if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MC_ERIF_LIST_ENTRIES) ||
+ !MLXSW_CORE_RES_VALID(mlxsw_sp->core, ACL_MAX_TCAM_RULES))
+ return -EIO;
+
+ return mlxsw_sp_mr_tcam_region_init(mlxsw_sp,
+ &mr_tcam->ipv4_tcam_region,
+ MLXSW_REG_RTAR_KEY_TYPE_IPV4_MULTICAST);
+}
+
+static void mlxsw_sp_mr_tcam_fini(void *priv)
+{
+ struct mlxsw_sp_mr_tcam *mr_tcam = priv;
+
+ mlxsw_sp_mr_tcam_region_fini(&mr_tcam->ipv4_tcam_region);
+}
+
+const struct mlxsw_sp_mr_ops mlxsw_sp_mr_tcam_ops = {
+ .priv_size = sizeof(struct mlxsw_sp_mr_tcam),
+ .route_priv_size = sizeof(struct mlxsw_sp_mr_tcam_route),
+ .init = mlxsw_sp_mr_tcam_init,
+ .route_create = mlxsw_sp_mr_tcam_route_create,
+ .route_update = mlxsw_sp_mr_tcam_route_update,
+ .route_stats = mlxsw_sp_mr_tcam_route_stats,
+ .route_action_update = mlxsw_sp_mr_tcam_route_action_update,
+ .route_min_mtu_update = mlxsw_sp_mr_tcam_route_min_mtu_update,
+ .route_irif_update = mlxsw_sp_mr_tcam_route_irif_update,
+ .route_erif_add = mlxsw_sp_mr_tcam_route_erif_add,
+ .route_erif_del = mlxsw_sp_mr_tcam_route_erif_del,
+ .route_destroy = mlxsw_sp_mr_tcam_route_destroy,
+ .fini = mlxsw_sp_mr_tcam_fini,
+};
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h
new file mode 100644
index 0000000..f9b59ee
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h
@@ -0,0 +1,43 @@
+/*
+ * drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.h
+ * Copyright (c) 2017 Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2017 Yotam Gigi <yotamg@mellanox.com>
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ * 3. Neither the names of the copyright holders nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * Alternatively, this software may be distributed under the terms of the
+ * GNU General Public License ("GPL") version 2 as published by the Free
+ * Software Foundation.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _MLXSW_SPECTRUM_MCROUTER_TCAM_H
+#define _MLXSW_SPECTRUM_MCROUTER_TCAM_H
+
+#include "spectrum.h"
+#include "spectrum_mr.h"
+
+extern const struct mlxsw_sp_mr_ops mlxsw_sp_mr_tcam_ops;
+
+#endif
--
2.9.5
^ permalink raw reply related
* [patch net-next 09/12] mlxsw: spectrum: router: Squash the default route table to main
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Currently, the mlxsw Spectrum driver offloads only either the RT_TABLE_MAIN
FIB table or the VRF tables, so the RT_TABLE_LOCAL table is squashed to the
RT_TABLE_MAIN table to allow local routes to be offloaded too.
By default, multicast MFC routes which are not assigned to any user
requested table are put in the RT_TABLE_DEFAULT table.
Due to the fact that offloading multicast MFC routes support in Spectrum
router logic is going to be introduced soon, squash the default table to
MAIN too.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 0bd93dc..1e6122f 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -692,8 +692,8 @@ static int mlxsw_sp_vr_lpm_tree_unbind(struct mlxsw_sp *mlxsw_sp,
static u32 mlxsw_sp_fix_tb_id(u32 tb_id)
{
- /* For our purpose, squash main and local table into one */
- if (tb_id == RT_TABLE_LOCAL)
+ /* For our purpose, squash main, default and local tables into one */
+ if (tb_id == RT_TABLE_LOCAL || tb_id == RT_TABLE_DEFAULT)
tb_id = RT_TABLE_MAIN;
return tb_id;
}
--
2.9.5
^ permalink raw reply related
* [patch net-next 10/12] mlxsw: spectrum_router: Add multicast routes notification handling functionality
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Add functionality for calling the multicast routing offloading logic upon
MFC and VIF add and delete notifications. In addition, call the multicast
routing upon RIF addition and deletion events.
As the multicast routing offload logic may sleep, the actual calls are done
in a deferred work. To ensure the MFC object is not freed in that interval,
a reference is held to it. In case of a failure, the abort mechanism is
used, which ejects all the routes from the hardware and triggers the
traffic to flow through the kernel.
Note: At that stage, the FIB notifications are still ignored, and will be
enabled in a further patch.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
.../net/ethernet/mellanox/mlxsw/spectrum_router.c | 187 ++++++++++++++++++++-
1 file changed, 185 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 1e6122f..b36ec63 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -65,6 +65,8 @@
#include "spectrum_cnt.h"
#include "spectrum_dpipe.h"
#include "spectrum_ipip.h"
+#include "spectrum_mr.h"
+#include "spectrum_mr_tcam.h"
#include "spectrum_router.h"
struct mlxsw_sp_vr;
@@ -458,6 +460,7 @@ struct mlxsw_sp_vr {
unsigned int rif_count;
struct mlxsw_sp_fib *fib4;
struct mlxsw_sp_fib *fib6;
+ struct mlxsw_sp_mr_table *mr4_table;
};
static const struct rhashtable_params mlxsw_sp_fib_ht_params;
@@ -652,7 +655,7 @@ static void mlxsw_sp_lpm_fini(struct mlxsw_sp *mlxsw_sp)
static bool mlxsw_sp_vr_is_used(const struct mlxsw_sp_vr *vr)
{
- return !!vr->fib4 || !!vr->fib6;
+ return !!vr->fib4 || !!vr->fib6 || !!vr->mr4_table;
}
static struct mlxsw_sp_vr *mlxsw_sp_vr_find_unused(struct mlxsw_sp *mlxsw_sp)
@@ -743,9 +746,18 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_create(struct mlxsw_sp *mlxsw_sp,
err = PTR_ERR(vr->fib6);
goto err_fib6_create;
}
+ vr->mr4_table = mlxsw_sp_mr_table_create(mlxsw_sp, vr->id,
+ MLXSW_SP_L3_PROTO_IPV4);
+ if (IS_ERR(vr->mr4_table)) {
+ err = PTR_ERR(vr->mr4_table);
+ goto err_mr_table_create;
+ }
vr->tb_id = tb_id;
return vr;
+err_mr_table_create:
+ mlxsw_sp_fib_destroy(vr->fib6);
+ vr->fib6 = NULL;
err_fib6_create:
mlxsw_sp_fib_destroy(vr->fib4);
vr->fib4 = NULL;
@@ -754,6 +766,8 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_create(struct mlxsw_sp *mlxsw_sp,
static void mlxsw_sp_vr_destroy(struct mlxsw_sp_vr *vr)
{
+ mlxsw_sp_mr_table_destroy(vr->mr4_table);
+ vr->mr4_table = NULL;
mlxsw_sp_fib_destroy(vr->fib6);
vr->fib6 = NULL;
mlxsw_sp_fib_destroy(vr->fib4);
@@ -774,7 +788,8 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_get(struct mlxsw_sp *mlxsw_sp, u32 tb_id)
static void mlxsw_sp_vr_put(struct mlxsw_sp_vr *vr)
{
if (!vr->rif_count && list_empty(&vr->fib4->node_list) &&
- list_empty(&vr->fib6->node_list))
+ list_empty(&vr->fib6->node_list) &&
+ mlxsw_sp_mr_table_empty(vr->mr4_table))
mlxsw_sp_vr_destroy(vr);
}
@@ -4606,6 +4621,75 @@ static int __mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp,
return 0;
}
+static int mlxsw_sp_router_fibmr_add(struct mlxsw_sp *mlxsw_sp,
+ struct mfc_entry_notifier_info *men_info,
+ bool replace)
+{
+ struct mlxsw_sp_vr *vr;
+
+ if (mlxsw_sp->router->aborted)
+ return 0;
+
+ vr = mlxsw_sp_vr_get(mlxsw_sp, men_info->tb_id);
+ if (IS_ERR(vr))
+ return PTR_ERR(vr);
+
+ return mlxsw_sp_mr_route4_add(vr->mr4_table, men_info->mfc, replace);
+}
+
+static void mlxsw_sp_router_fibmr_del(struct mlxsw_sp *mlxsw_sp,
+ struct mfc_entry_notifier_info *men_info)
+{
+ struct mlxsw_sp_vr *vr;
+
+ if (mlxsw_sp->router->aborted)
+ return;
+
+ vr = mlxsw_sp_vr_find(mlxsw_sp, men_info->tb_id);
+ if (WARN_ON(!vr))
+ return;
+
+ mlxsw_sp_mr_route4_del(vr->mr4_table, men_info->mfc);
+ mlxsw_sp_vr_put(vr);
+}
+
+static int
+mlxsw_sp_router_fibmr_vif_add(struct mlxsw_sp *mlxsw_sp,
+ struct vif_entry_notifier_info *ven_info)
+{
+ struct mlxsw_sp_rif *rif;
+ struct mlxsw_sp_vr *vr;
+
+ if (mlxsw_sp->router->aborted)
+ return 0;
+
+ vr = mlxsw_sp_vr_get(mlxsw_sp, ven_info->tb_id);
+ if (IS_ERR(vr))
+ return PTR_ERR(vr);
+
+ rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, ven_info->dev);
+ return mlxsw_sp_mr_vif_add(vr->mr4_table, ven_info->dev,
+ ven_info->vif_index,
+ ven_info->vif_flags, rif);
+}
+
+static void
+mlxsw_sp_router_fibmr_vif_del(struct mlxsw_sp *mlxsw_sp,
+ struct vif_entry_notifier_info *ven_info)
+{
+ struct mlxsw_sp_vr *vr;
+
+ if (mlxsw_sp->router->aborted)
+ return;
+
+ vr = mlxsw_sp_vr_find(mlxsw_sp, ven_info->tb_id);
+ if (WARN_ON(!vr))
+ return;
+
+ mlxsw_sp_mr_vif_del(vr->mr4_table, ven_info->vif_index);
+ mlxsw_sp_vr_put(vr);
+}
+
static int mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp)
{
enum mlxsw_reg_ralxx_protocol proto = MLXSW_REG_RALXX_PROTOCOL_IPV4;
@@ -4616,6 +4700,10 @@ static int mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp)
if (err)
return err;
+ /* The multicast router code does not need an abort trap as by default,
+ * packets that don't match any routes are trapped to the CPU.
+ */
+
proto = MLXSW_REG_RALXX_PROTOCOL_IPV6;
return __mlxsw_sp_router_set_abort_trap(mlxsw_sp, proto,
MLXSW_SP_LPM_TREE_MIN + 1);
@@ -4697,6 +4785,8 @@ static void mlxsw_sp_router_fib_flush(struct mlxsw_sp *mlxsw_sp)
if (!mlxsw_sp_vr_is_used(vr))
continue;
+
+ mlxsw_sp_mr_table_flush(vr->mr4_table);
mlxsw_sp_vr_fib_flush(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV4);
/* If virtual router was only used for IPv4, then it's no
@@ -4729,6 +4819,8 @@ struct mlxsw_sp_fib_event_work {
struct fib_entry_notifier_info fen_info;
struct fib_rule_notifier_info fr_info;
struct fib_nh_notifier_info fnh_info;
+ struct mfc_entry_notifier_info men_info;
+ struct vif_entry_notifier_info ven_info;
};
struct mlxsw_sp *mlxsw_sp;
unsigned long event;
@@ -4815,6 +4907,55 @@ static void mlxsw_sp_router_fib6_event_work(struct work_struct *work)
kfree(fib_work);
}
+static void mlxsw_sp_router_fibmr_event_work(struct work_struct *work)
+{
+ struct mlxsw_sp_fib_event_work *fib_work =
+ container_of(work, struct mlxsw_sp_fib_event_work, work);
+ struct mlxsw_sp *mlxsw_sp = fib_work->mlxsw_sp;
+ struct fib_rule *rule;
+ bool replace;
+ int err;
+
+ rtnl_lock();
+ switch (fib_work->event) {
+ case FIB_EVENT_ENTRY_REPLACE: /* fall through */
+ case FIB_EVENT_ENTRY_ADD:
+ replace = fib_work->event == FIB_EVENT_ENTRY_REPLACE;
+
+ err = mlxsw_sp_router_fibmr_add(mlxsw_sp, &fib_work->men_info,
+ replace);
+ if (err)
+ mlxsw_sp_router_fib_abort(mlxsw_sp);
+ ipmr_cache_put(fib_work->men_info.mfc);
+ break;
+ case FIB_EVENT_ENTRY_DEL:
+ mlxsw_sp_router_fibmr_del(mlxsw_sp, &fib_work->men_info);
+ ipmr_cache_put(fib_work->men_info.mfc);
+ break;
+ case FIB_EVENT_VIF_ADD:
+ err = mlxsw_sp_router_fibmr_vif_add(mlxsw_sp,
+ &fib_work->ven_info);
+ if (err)
+ mlxsw_sp_router_fib_abort(mlxsw_sp);
+ dev_put(fib_work->ven_info.dev);
+ break;
+ case FIB_EVENT_VIF_DEL:
+ mlxsw_sp_router_fibmr_vif_del(mlxsw_sp,
+ &fib_work->ven_info);
+ dev_put(fib_work->ven_info.dev);
+ break;
+ case FIB_EVENT_RULE_ADD: /* fall through */
+ case FIB_EVENT_RULE_DEL:
+ rule = fib_work->fr_info.rule;
+ if (!ipmr_rule_default(rule) && !rule->l3mdev)
+ mlxsw_sp_router_fib_abort(mlxsw_sp);
+ fib_rule_put(rule);
+ break;
+ }
+ rtnl_unlock();
+ kfree(fib_work);
+}
+
static void mlxsw_sp_router_fib4_event(struct mlxsw_sp_fib_event_work *fib_work,
struct fib_notifier_info *info)
{
@@ -4860,6 +5001,30 @@ static void mlxsw_sp_router_fib6_event(struct mlxsw_sp_fib_event_work *fib_work,
}
}
+static void
+mlxsw_sp_router_fibmr_event(struct mlxsw_sp_fib_event_work *fib_work,
+ struct fib_notifier_info *info)
+{
+ switch (fib_work->event) {
+ case FIB_EVENT_ENTRY_REPLACE: /* fall through */
+ case FIB_EVENT_ENTRY_ADD: /* fall through */
+ case FIB_EVENT_ENTRY_DEL:
+ memcpy(&fib_work->men_info, info, sizeof(fib_work->men_info));
+ ipmr_cache_hold(fib_work->men_info.mfc);
+ break;
+ case FIB_EVENT_VIF_ADD: /* fall through */
+ case FIB_EVENT_VIF_DEL:
+ memcpy(&fib_work->ven_info, info, sizeof(fib_work->ven_info));
+ dev_hold(fib_work->ven_info.dev);
+ break;
+ case FIB_EVENT_RULE_ADD: /* fall through */
+ case FIB_EVENT_RULE_DEL:
+ memcpy(&fib_work->fr_info, info, sizeof(fib_work->fr_info));
+ fib_rule_get(fib_work->fr_info.rule);
+ break;
+ }
+}
+
/* Called with rcu_read_lock() */
static int mlxsw_sp_router_fib_event(struct notifier_block *nb,
unsigned long event, void *ptr)
@@ -4889,6 +5054,10 @@ static int mlxsw_sp_router_fib_event(struct notifier_block *nb,
INIT_WORK(&fib_work->work, mlxsw_sp_router_fib6_event_work);
mlxsw_sp_router_fib6_event(fib_work, info);
break;
+ case RTNL_FAMILY_IPMR:
+ INIT_WORK(&fib_work->work, mlxsw_sp_router_fibmr_event_work);
+ mlxsw_sp_router_fibmr_event(fib_work, info);
+ break;
}
mlxsw_core_schedule_work(&fib_work->work);
@@ -5102,12 +5271,18 @@ mlxsw_sp_rif_create(struct mlxsw_sp *mlxsw_sp,
if (err)
goto err_configure;
+ err = mlxsw_sp_mr_rif_add(vr->mr4_table, rif);
+ if (err)
+ goto err_mr_rif_add;
+
mlxsw_sp_rif_counters_alloc(rif);
mlxsw_sp->router->rifs[rif_index] = rif;
vr->rif_count++;
return rif;
+err_mr_rif_add:
+ ops->deconfigure(rif);
err_configure:
if (fid)
mlxsw_sp_fid_put(fid);
@@ -5132,6 +5307,7 @@ void mlxsw_sp_rif_destroy(struct mlxsw_sp_rif *rif)
vr->rif_count--;
mlxsw_sp->router->rifs[rif->rif_index] = NULL;
mlxsw_sp_rif_counters_free(rif);
+ mlxsw_sp_mr_rif_del(vr->mr4_table, rif);
ops->deconfigure(rif);
if (fid)
/* Loopback RIFs are not associated with a FID. */
@@ -5994,6 +6170,10 @@ int mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp)
if (err)
goto err_lpm_init;
+ err = mlxsw_sp_mr_init(mlxsw_sp, &mlxsw_sp_mr_tcam_ops);
+ if (err)
+ goto err_mr_init;
+
err = mlxsw_sp_vrs_init(mlxsw_sp);
if (err)
goto err_vrs_init;
@@ -6015,6 +6195,8 @@ int mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp)
err_neigh_init:
mlxsw_sp_vrs_fini(mlxsw_sp);
err_vrs_init:
+ mlxsw_sp_mr_fini(mlxsw_sp);
+err_mr_init:
mlxsw_sp_lpm_fini(mlxsw_sp);
err_lpm_init:
rhashtable_destroy(&mlxsw_sp->router->nexthop_group_ht);
@@ -6036,6 +6218,7 @@ void mlxsw_sp_router_fini(struct mlxsw_sp *mlxsw_sp)
unregister_fib_notifier(&mlxsw_sp->router->fib_nb);
mlxsw_sp_neigh_fini(mlxsw_sp);
mlxsw_sp_vrs_fini(mlxsw_sp);
+ mlxsw_sp_mr_fini(mlxsw_sp);
mlxsw_sp_lpm_fini(mlxsw_sp);
rhashtable_destroy(&mlxsw_sp->router->nexthop_group_ht);
rhashtable_destroy(&mlxsw_sp->router->nexthop_ht);
--
2.9.5
^ permalink raw reply related
* [patch net-next 11/12] mlxsw: spectrum: Notify multicast router on RIF MTU changes
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Due to the fact that multicast routes hold the minimum MTU of all the
egress RIFs and trap packets that don't meet it, notify the mulitcast
router code on RIF MTU changes.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c | 11 +++++++++++
1 file changed, 11 insertions(+)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index b36ec63..edc6462 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -5648,6 +5648,17 @@ int mlxsw_sp_netdevice_router_port_event(struct net_device *dev)
if (err)
goto err_rif_fdb_op;
+ if (rif->mtu != dev->mtu) {
+ struct mlxsw_sp_vr *vr;
+
+ /* The RIF is relevant only to its mr_table instance, as unlike
+ * unicast routing, in multicast routing a RIF cannot be shared
+ * between several multicast routing tables.
+ */
+ vr = &mlxsw_sp->router->vrs[rif->vr_id];
+ mlxsw_sp_mr_rif_mtu_update(vr->mr4_table, rif, dev->mtu);
+ }
+
ether_addr_copy(rif->addr, dev->dev_addr);
rif->mtu = dev->mtu;
--
2.9.5
^ permalink raw reply related
* [patch net-next 12/12] mlxsw: spectrum: router: Don't ignore IPMR notifications
From: Jiri Pirko @ 2017-09-21 6:43 UTC (permalink / raw)
To: netdev; +Cc: davem, yotamg, idosch, mlxsw
In-Reply-To: <20170921064338.1282-1-jiri@resnulli.us>
From: Yotam Gigi <yotamg@mellanox.com>
Make the Spectrum router logic not ignore the RTNL_FAMILY_IPMR FIB
notifications.
Past commits added the IPMR VIF and MFC add/del notifications via the
fib_notifier chain. In addition, a code for handling these notifications in
the Spectrum router logic was added. Make the Spectrum router logic not
ignore these notifications and forward the requests to the Spectrum
multicast router offloading logic.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index edc6462..16c041b 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -5034,7 +5034,8 @@ static int mlxsw_sp_router_fib_event(struct notifier_block *nb,
struct mlxsw_sp_router *router;
if (!net_eq(info->net, &init_net) ||
- (info->family != AF_INET && info->family != AF_INET6))
+ (info->family != AF_INET && info->family != AF_INET6 &&
+ info->family != RTNL_FAMILY_IPMR))
return NOTIFY_DONE;
fib_work = kzalloc(sizeof(*fib_work), GFP_ATOMIC);
--
2.9.5
^ permalink raw reply related
* [PATCH net 1/9] net/smc: add missing dev_put
From: Ursula Braun @ 2017-09-21 7:16 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, linux-rdma-u79uwXL29TY76Z2rM5mHXA,
linux-s390-u79uwXL29TY76Z2rM5mHXA,
jwi-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
schwidefsky-tA70FqPdS9bQT0dZR+AlfA,
heiko.carstens-tA70FqPdS9bQT0dZR+AlfA,
raspl-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8
In-Reply-To: <20170921071634.16883-1-ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
From: Hans Wippel <hwippel-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
In the infiniband part, SMC currently uses get_netdev which calls
dev_hold on the returned net device. However, the SMC code never calls
dev_put on that net device resulting in a wrong reference count.
This patch adds a dev_put after the usage of the net device to fix the
issue.
Signed-off-by: Hans Wippel <hwippel-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
Signed-off-by: Ursula Braun <ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
---
net/smc/smc_ib.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/net/smc/smc_ib.c b/net/smc/smc_ib.c
index 547e0e113b17..0b5852299158 100644
--- a/net/smc/smc_ib.c
+++ b/net/smc/smc_ib.c
@@ -380,6 +380,7 @@ static int smc_ib_fill_gid_and_mac(struct smc_ib_device *smcibdev, u8 ibport)
ndev = smcibdev->ibdev->get_netdev(smcibdev->ibdev, ibport);
if (ndev) {
memcpy(&smcibdev->mac, ndev->dev_addr, ETH_ALEN);
+ dev_put(ndev);
} else if (!rc) {
memcpy(&smcibdev->mac[ibport - 1][0],
&smcibdev->gid[ibport - 1].raw[8], 3);
--
2.13.5
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH net 5/9] net/smc: adapt send request completion notification
From: Ursula Braun @ 2017-09-21 7:16 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, linux-rdma-u79uwXL29TY76Z2rM5mHXA,
linux-s390-u79uwXL29TY76Z2rM5mHXA,
jwi-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
schwidefsky-tA70FqPdS9bQT0dZR+AlfA,
heiko.carstens-tA70FqPdS9bQT0dZR+AlfA,
raspl-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8
In-Reply-To: <20170921071634.16883-1-ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
The solicited flag is meaningful for the receive completion queue.
Ask for next work completion of any type on the send queue.
Signed-off-by: Ursula Braun <ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
---
net/smc/smc_wr.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/smc/smc_wr.c b/net/smc/smc_wr.c
index ab56bda66783..525d91e0d57e 100644
--- a/net/smc/smc_wr.c
+++ b/net/smc/smc_wr.c
@@ -244,7 +244,7 @@ int smc_wr_tx_send(struct smc_link *link, struct smc_wr_tx_pend_priv *priv)
int rc;
ib_req_notify_cq(link->smcibdev->roce_cq_send,
- IB_CQ_SOLICITED_MASK | IB_CQ_REPORT_MISSED_EVENTS);
+ IB_CQ_NEXT_COMP | IB_CQ_REPORT_MISSED_EVENTS);
pend = container_of(priv, struct smc_wr_tx_pend, priv);
rc = ib_post_send(link->roce_qp, &link->wr_tx_ibs[pend->idx],
&failed_wr);
--
2.13.5
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH net 6/9] net/smc: longer delay for client link group removal
From: Ursula Braun @ 2017-09-21 7:16 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, linux-rdma-u79uwXL29TY76Z2rM5mHXA,
linux-s390-u79uwXL29TY76Z2rM5mHXA,
jwi-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
schwidefsky-tA70FqPdS9bQT0dZR+AlfA,
heiko.carstens-tA70FqPdS9bQT0dZR+AlfA,
raspl-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8
In-Reply-To: <20170921071634.16883-1-ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
Client link group creation always follows the server linkgroup creation.
If peer creates a new server link group, client has to create a new
client link group. If peer reuses a server link group for a new
connection, client has to reuse its client link group as well. This
patch introduces a longer delay for client link group removal to make
sure this link group still exists, once the peer decides to reuse a
server link group. This avoids out-of-sync conditions for link groups.
If already scheduled, modify the delay.
Signed-off-by: Ursula Braun <ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
---
net/smc/smc_core.c | 16 ++++++++++++----
1 file changed, 12 insertions(+), 4 deletions(-)
diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c
index 1a16d51e2330..20b66e79c5d6 100644
--- a/net/smc/smc_core.c
+++ b/net/smc/smc_core.c
@@ -25,8 +25,9 @@
#include "smc_cdc.h"
#include "smc_close.h"
-#define SMC_LGR_NUM_INCR 256
-#define SMC_LGR_FREE_DELAY (600 * HZ)
+#define SMC_LGR_NUM_INCR 256
+#define SMC_LGR_FREE_DELAY_SERV (600 * HZ)
+#define SMC_LGR_FREE_DELAY_CLNT (SMC_LGR_FREE_DELAY_SERV + 10)
static u32 smc_lgr_num; /* unique link group number */
@@ -107,8 +108,15 @@ static void smc_lgr_unregister_conn(struct smc_connection *conn)
__smc_lgr_unregister_conn(conn);
}
write_unlock_bh(&lgr->conns_lock);
- if (reduced && !lgr->conns_num)
- schedule_delayed_work(&lgr->free_work, SMC_LGR_FREE_DELAY);
+ if (!reduced || lgr->conns_num)
+ return;
+ /* client link group creation always follows the server link group
+ * creation. For client use a somewhat higher removal delay time,
+ * otherwise there is a risk of out-of-sync link groups.
+ */
+ mod_delayed_work(system_wq, &lgr->free_work,
+ lgr->role == SMC_CLNT ? SMC_LGR_FREE_DELAY_CLNT :
+ SMC_LGR_FREE_DELAY_SERV);
}
static void smc_lgr_free_work(struct work_struct *work)
--
2.13.5
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH net 7/9] net/smc: terminate link group if out-of-sync is received
From: Ursula Braun @ 2017-09-21 7:16 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, linux-rdma-u79uwXL29TY76Z2rM5mHXA,
linux-s390-u79uwXL29TY76Z2rM5mHXA,
jwi-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
schwidefsky-tA70FqPdS9bQT0dZR+AlfA,
heiko.carstens-tA70FqPdS9bQT0dZR+AlfA,
raspl-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8
In-Reply-To: <20170921071634.16883-1-ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
An out-of-sync condition can just be detected by the client.
If the server receives a CLC DECLINE message indicating an out-of-sync
condition for the link groups, the server must clean up the out-of-sync
link group.
There is no need for an extra third parameter in smc_clc_send_decline().
Signed-off-by: Ursula Braun <ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
---
net/smc/af_smc.c | 6 ++----
net/smc/smc_clc.c | 10 +++++-----
net/smc/smc_clc.h | 3 +--
3 files changed, 8 insertions(+), 11 deletions(-)
diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index 2e8d2dabac0c..745f145d4c4d 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -513,7 +513,7 @@ static int smc_connect_rdma(struct smc_sock *smc)
/* RDMA setup failed, switch back to TCP */
smc->use_fallback = true;
if (reason_code && (reason_code != SMC_CLC_DECL_REPLY)) {
- rc = smc_clc_send_decline(smc, reason_code, 0);
+ rc = smc_clc_send_decline(smc, reason_code);
if (rc < sizeof(struct smc_clc_msg_decline))
goto out_err;
}
@@ -808,8 +808,6 @@ static void smc_listen_work(struct work_struct *work)
rc = local_contact;
if (rc == -ENOMEM)
reason_code = SMC_CLC_DECL_MEM;/* insufficient memory*/
- else if (rc == -ENOLINK)
- reason_code = SMC_CLC_DECL_SYNCERR; /* synchr. error */
goto decline_rdma;
}
link = &new_smc->conn.lgr->lnk[SMC_SINGLE_LINK];
@@ -903,7 +901,7 @@ static void smc_listen_work(struct work_struct *work)
smc_conn_free(&new_smc->conn);
new_smc->use_fallback = true;
if (reason_code && (reason_code != SMC_CLC_DECL_REPLY)) {
- rc = smc_clc_send_decline(new_smc, reason_code, 0);
+ rc = smc_clc_send_decline(new_smc, reason_code);
if (rc < sizeof(struct smc_clc_msg_decline))
goto out_err;
}
diff --git a/net/smc/smc_clc.c b/net/smc/smc_clc.c
index 3934913ab835..b7dd2743fb5c 100644
--- a/net/smc/smc_clc.c
+++ b/net/smc/smc_clc.c
@@ -95,9 +95,10 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
}
if (clcm->type == SMC_CLC_DECLINE) {
reason_code = SMC_CLC_DECL_REPLY;
- if (ntohl(((struct smc_clc_msg_decline *)buf)->peer_diagnosis)
- == SMC_CLC_DECL_SYNCERR)
+ if (((struct smc_clc_msg_decline *)buf)->hdr.flag) {
smc->conn.lgr->sync_err = true;
+ smc_lgr_terminate(smc->conn.lgr);
+ }
}
out:
@@ -105,8 +106,7 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
}
/* send CLC DECLINE message across internal TCP socket */
-int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info,
- u8 out_of_sync)
+int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info)
{
struct smc_clc_msg_decline dclc;
struct msghdr msg;
@@ -118,7 +118,7 @@ int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info,
dclc.hdr.type = SMC_CLC_DECLINE;
dclc.hdr.length = htons(sizeof(struct smc_clc_msg_decline));
dclc.hdr.version = SMC_CLC_V1;
- dclc.hdr.flag = out_of_sync ? 1 : 0;
+ dclc.hdr.flag = (peer_diag_info == SMC_CLC_DECL_SYNCERR) ? 1 : 0;
memcpy(dclc.id_for_peer, local_systemid, sizeof(local_systemid));
dclc.peer_diagnosis = htonl(peer_diag_info);
memcpy(dclc.trl.eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER));
diff --git a/net/smc/smc_clc.h b/net/smc/smc_clc.h
index 13db8ce177c9..1c55414041d4 100644
--- a/net/smc/smc_clc.h
+++ b/net/smc/smc_clc.h
@@ -106,8 +106,7 @@ struct smc_ib_device;
int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
u8 expected_type);
-int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info,
- u8 out_of_sync);
+int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info);
int smc_clc_send_proposal(struct smc_sock *smc, struct smc_ib_device *smcibdev,
u8 ibport);
int smc_clc_send_confirm(struct smc_sock *smc);
--
2.13.5
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH net 9/9] net/smc: no close wait in case of process shut down
From: Ursula Braun @ 2017-09-21 7:16 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, linux-rdma-u79uwXL29TY76Z2rM5mHXA,
linux-s390-u79uwXL29TY76Z2rM5mHXA,
jwi-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
schwidefsky-tA70FqPdS9bQT0dZR+AlfA,
heiko.carstens-tA70FqPdS9bQT0dZR+AlfA,
raspl-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8
In-Reply-To: <20170921071634.16883-1-ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
Usually socket closing is delayed if there is still data available in
the send buffer to be transmitted. If a process is killed, the delay
should be avoided.
Signed-off-by: Ursula Braun <ubraun-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
---
net/smc/smc_close.c | 15 ++++++++-------
1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/net/smc/smc_close.c b/net/smc/smc_close.c
index 5201bc103bd8..f0d16fb825f7 100644
--- a/net/smc/smc_close.c
+++ b/net/smc/smc_close.c
@@ -174,15 +174,15 @@ int smc_close_active(struct smc_sock *smc)
{
struct smc_cdc_conn_state_flags *txflags =
&smc->conn.local_tx_ctrl.conn_state_flags;
- long timeout = SMC_MAX_STREAM_WAIT_TIMEOUT;
struct smc_connection *conn = &smc->conn;
struct sock *sk = &smc->sk;
int old_state;
+ long timeout;
int rc = 0;
- if (sock_flag(sk, SOCK_LINGER) &&
- !(current->flags & PF_EXITING))
- timeout = sk->sk_lingertime;
+ timeout = current->flags & PF_EXITING ?
+ 0 : sock_flag(sk, SOCK_LINGER) ?
+ sk->sk_lingertime : SMC_MAX_STREAM_WAIT_TIMEOUT;
again:
old_state = sk->sk_state;
@@ -413,13 +413,14 @@ void smc_close_sock_put_work(struct work_struct *work)
int smc_close_shutdown_write(struct smc_sock *smc)
{
struct smc_connection *conn = &smc->conn;
- long timeout = SMC_MAX_STREAM_WAIT_TIMEOUT;
struct sock *sk = &smc->sk;
int old_state;
+ long timeout;
int rc = 0;
- if (sock_flag(sk, SOCK_LINGER))
- timeout = sk->sk_lingertime;
+ timeout = current->flags & PF_EXITING ?
+ 0 : sock_flag(sk, SOCK_LINGER) ?
+ sk->sk_lingertime : SMC_MAX_STREAM_WAIT_TIMEOUT;
again:
old_state = sk->sk_state;
--
2.13.5
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ 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