Set the DPDK pmd thread scheduling policy to SCHED_RR and static
priority to highest priority value of the policy. This is to deal with
pmd thread starvation case where another cpu hogging process can get
scheduled/affinitized on to the same core the pmd thread is running
there by significantly impacting the datapath performance. To reproduce
the pmd thread starvation case:
$ ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=6
$ taskset 0x2 cat /dev/zero > /dev/null &
Setting the realtime scheduling policy to the pmd threads is one step
towards Fastpath Service Assurance in OVS DPDK.
The realtime scheduling policy is applied only when CPU mask is passed
to 'pmd-cpu-mask'. For example:
* In the absence of pmd-cpu-mask, one pmd thread shall be created
and default scheduling policy and priority gets applied.
* If pmd-cpu-mask is specified, one or more pmd threads shall be
spawned on the corresponding core(s) in the mask and real time
scheduling policy SCHED_RR and highest priority of the policy is
applied to the pmd thread(s).
With this commit, it is recommended that the OVS control thread and pmd
thread shouldn't be pinned to same core ('dpdk-lcore-mask','pmd-cpu-mask'
should be non-overlapping). If dpdk-lcore-mask is set same as pmd-cpu-mask
the pmd thread is not spawned on lowest core of the dpdk-lcore-mask.
Also other processes with same affinity as PMD thread will be unresponsive.
Signed-off-by: Bhanuprakash Bodireddy <bhanuprakash.bodire...@intel.com>
---
v5->v6:
* Prohibit spawning pmd thread on the lowest core in dpdk-lcore-mask if
lcore-mask and pmd-mask affinity are identical.
* Updated Note section in INSTALL.DPDK-ADVANCED doc.
* Tested below cases to verify system stability with pmd priority patch
dpdk-lcore-mask | pmd-cpu-mask | Comment
1. Not set | Not set | control threads affinity: 0-27
pmd thread: core 0
2. 1 | 1 | pmd thread isn't spawned and warning
logged in logfile.
3. 1 | c |
4. F0 | F0 | control threads pinned to core 4.
3 pmd threads created on core 5,6,7 but 4.
v4->v5:
* Reword Note section in DPDK-ADVANCED.md
v3->v4:
* Document update
* Use ovs_strerror for reporting errors in lib-numa.c
v2->v3:
* Move set_priority() function to lib/ovs-numa.c
* Apply realtime scheduling policy and priority to pmd thread only if
pmd-cpu-mask is passed.
* Update INSTALL.DPDK-ADVANCED.
v1->v2:
* Removed #ifdef and introduced dummy function "pmd_thread_setpriority"
in netdev-dpdk.h
* Rebase
INSTALL.DPDK-ADVANCED.md | 24 ++++++++++++++++++++----
lib/dpif-netdev.c | 12 ++++++++++++
lib/netdev-dpdk.c | 22 ++++++++++++++++++++++
lib/ovs-numa.c | 18 ++++++++++++++++++
lib/ovs-numa.h | 1 +
5 files changed, 73 insertions(+), 4 deletions(-)
diff --git a/INSTALL.DPDK-ADVANCED.md b/INSTALL.DPDK-ADVANCED.md
index dd36ae4..f039723 100755
--- a/INSTALL.DPDK-ADVANCED.md
+++ b/INSTALL.DPDK-ADVANCED.md
@@ -208,8 +208,10 @@ needs to be affinitized accordingly.
pmd thread is CPU bound, and needs to be affinitized to isolated
cores for optimum performance.
- By setting a bit in the mask, a pmd thread is created and pinned
- to the corresponding CPU core. e.g. to run a pmd thread on core 2
+ By setting a bit in the mask, a pmd thread is created, pinned
+ to the corresponding CPU core and the scheduling policy SCHED_RR
+ along with maximum priority of the policy applied to the pmd thread.
+ e.g. to pin a pmd thread on core 2
`ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=4`
@@ -237,8 +239,10 @@ needs to be affinitized accordingly.
responsible for different ports/rxq's. Assignment of ports/rxq's to
pmd threads is done automatically.
- A set bit in the mask means a pmd thread is created and pinned
- to the corresponding CPU core. e.g. to run pmd threads on core 1 and 2
+ A set bit in the mask means a pmd thread is created, pinned to the
+ corresponding CPU core and the scheduling policy SCHED_RR with highest
+ priority of the scheduling policy applied to pmd thread.
+ e.g. to run pmd threads on core 1 and 2
`ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=6`
@@ -249,6 +253,18 @@ needs to be affinitized accordingly.
NIC port0 <-> OVS <-> VM <-> OVS <-> NIC port 1
+ Note: 'dpdk-lcore-mask' and 'pmd-cpu-mask' cpu mask settings should be
+ non-overlapping. If you set dpdk-lcore-mask same as pmd-cpu-mask, the pmd
+ thread is not spawned on lowest core of the dpdk-lcore-mask and in case of
+ cpu mask setting where both the control and the pmd thread pinned to same
+ core, pmd thread is not spawned at all if it is the only cpu in the affinity
+ mask.
+
+ For example, if dpdk-lcore-mask=F0 and pmd-cpu-mask=F0, pmd threads are
+ created on cores 5, 6, and 7 but not on 4. Setting both lcore,pmd mask to
+ core 4 shall pin control threads to core 4 but will prevent pmd thread from
+ spawning.
+
### 4.3 DPDK physical port Rx Queues
`ovs-vsctl set Interface <DPDK interface> options:n_rxq=<integer>`
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index 96504f5..abe1fde 100644
--- a/lib/dpif-netdev.c
+++ b/lib/dpif-netdev.c
@@ -3100,6 +3100,18 @@ pmd_thread_main(void *f_)
ovs_numa_thread_setaffinity_core(pmd->core_id);
dpdk_set_lcore_id(pmd->core_id);
poll_cnt = pmd_load_queues_and_ports(pmd, &poll_list);
+
+ /* When cpu affinity mask explicitly set using pmd-cpu-mask, pmd thread's
+ * scheduling policy is set to SCHED_RR and priority to highest priority
+ * of SCHED_RR policy. In the absence of pmd-cpu-mask, default scheduling
+ * policy and priority shall apply to pmd thread.
+ *
+ * If dpdk-lcore-mask is set same as pmd-cpu-mask, the pmd thread is not
+ * spawned on lowest core of the dpdk-lcore-mask.
+ */
+ if (pmd->dp->pmd_cmask) {
+ ovs_numa_thread_setpriority(SCHED_RR);
+ }
reload:
emc_cache_init(&pmd->flow_cache);
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index c767fd4..5d606a2 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
@@ -3432,6 +3432,28 @@ dpdk_init__(const struct smap *ovs_other_config)
}
}
+ /* If dpdk-lcore-mask is explicitly specified, auto_determine is 'false'
+ * and control threads are pinned to lowest core of the affinity mask.
+ * Mark the corresponding control thread 'core->pinned' status to 'true'
+ * so that pmd thread can't be spawned on the same core.
+ */
+ if (!auto_determine) {
+ const char *lcore_mask;
+ unsigned int core_id;
+ int dl_mask;
+
+ lcore_mask = smap_get(ovs_other_config, "dpdk-lcore-mask");
+ dl_mask = strtol(lcore_mask, NULL, 16);
+
+ /* Get the lowest core_id of dpdk-lcore-mask */
+ core_id = rightmost_1bit_idx(dl_mask);
+
+ /* Set the control thread core 'pinned' status to true */
+ if (!ovs_numa_try_pin_core_specific(core_id)) {
+ VLOG_ERR("Unable to pin core %d", core_id);
+ }
+ }
+
dpdk_argv = argv;
dpdk_argc = argc;
diff --git a/lib/ovs-numa.c b/lib/ovs-numa.c
index c8173e0..428f274 100644
--- a/lib/ovs-numa.c
+++ b/lib/ovs-numa.c
@@ -613,3 +613,21 @@ int ovs_numa_thread_setaffinity_core(unsigned core_id
OVS_UNUSED)
return EOPNOTSUPP;
#endif /* __linux__ */
}
+
+void
+ovs_numa_thread_setpriority(int policy)
+{
+ if (dummy_numa) {
+ return;
+ }
+
+ struct sched_param threadparam;
+ int err;
+
+ memset(&threadparam, 0, sizeof(threadparam));
+ threadparam.sched_priority = sched_get_priority_max(policy);
+ err = pthread_setschedparam(pthread_self(), policy, &threadparam);
+ if (err) {
+ VLOG_ERR("Thread priority error %s",ovs_strerror(err));
+ }
+}
diff --git a/lib/ovs-numa.h b/lib/ovs-numa.h
index be836b2..94f0884 100644
--- a/lib/ovs-numa.h
+++ b/lib/ovs-numa.h
@@ -56,6 +56,7 @@ void ovs_numa_unpin_core(unsigned core_id);
struct ovs_numa_dump *ovs_numa_dump_cores_on_numa(int numa_id);
void ovs_numa_dump_destroy(struct ovs_numa_dump *);
int ovs_numa_thread_setaffinity_core(unsigned core_id);
+void ovs_numa_thread_setpriority(int policy);
#define FOR_EACH_CORE_ON_NUMA(ITER, DUMP) \
LIST_FOR_EACH((ITER), list_node, &(DUMP)->dump)
--
2.4.11
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
