Re: [ovs-dev] [PATCH v3] netdev-dpdk: Set pmd thread priority

2016-07-25 Thread Daniele Di Proietto
I agree with Mark's comments, other than that this looks good to me.

If you agree with the comments would you mind sending an updates version?

Thanks,

Daniele

2016-07-19 7:04 GMT-07:00 Kavanagh, Mark B :

> >
>
> Hi Bhanu,
>
> Thanks for the patch - some comments inline.
>
> Cheers,
> Mark
>
> >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.
> >
> >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 prority gets applied.
>
> Typo above - 'prority'
>
> >
> >* If pmd-cpu-mask is specified, one ore more pmd threads shall be
>
> Typo above - 'ore'
>
> >  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).
> >
> >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 &
> >
> >With this commit OVS control threads and pmd threads can't have same
> >affinity ('dpdk-lcore-mask','pmd-cpu-mask' should be non-overlapping).
> >Also other processes with same affinity as PMD thread will be
> unresponsive.
> >
> >Signed-off-by: Bhanuprakash Bodireddy 
> >---
> >
> >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 | 15 +++
> > lib/dpif-netdev.c|  9 +
> > lib/ovs-numa.c   | 18 ++
> > lib/ovs-numa.h   |  1 +
> > 4 files changed, 39 insertions(+), 4 deletions(-)
> >
> >diff --git a/INSTALL.DPDK-ADVANCED.md b/INSTALL.DPDK-ADVANCED.md
> >index 9ae536d..d828290 100644
> >--- a/INSTALL.DPDK-ADVANCED.md
> >+++ b/INSTALL.DPDK-ADVANCED.md
> >@@ -205,8 +205,10 @@ 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`
> >
> >@@ -234,8 +236,10 @@ 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
>
> There's some repetition in the last paragraph - I'm reviewing this patch
> in isolation, so the text may make sense/be required in the full document.
>
> >
> >   `ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=6`
> >
> >@@ -246,6 +250,9 @@ 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.
>
> Although it's mentioned in the commit message, it might be worth
> mentioning here the consequences of attempting to pin non-PMD processes to
> a pmd-cpu-mask core (i.e. CPU starvation)
>
> >+
> > ### 4.3 DPDK physical port Rx Queues
> >
> >   `ovs-vsctl set Interface  options:n_rxq=`
> >diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
> >index e0107b7..805d0ae 100644
> >--- a/lib/dpif-netdev.c
> >+++ b/lib/dpif-netdev.c
> >@@ -2851,6 +2851,15 @@ 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, _list);
> >+
> >+/* When cpu affinity mask explicitly set using 

Re: [ovs-dev] [PATCH v3] netdev-dpdk: Set pmd thread priority

2016-07-19 Thread Kavanagh, Mark B
>

Hi Bhanu,

Thanks for the patch - some comments inline.

Cheers,
Mark

>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.
>
>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 prority gets applied.

Typo above - 'prority'

>
>* If pmd-cpu-mask is specified, one ore more pmd threads shall be

Typo above - 'ore'

>  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).
>
>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 &
>
>With this commit OVS control threads and pmd threads can't have same
>affinity ('dpdk-lcore-mask','pmd-cpu-mask' should be non-overlapping).
>Also other processes with same affinity as PMD thread will be unresponsive.
>
>Signed-off-by: Bhanuprakash Bodireddy 
>---
>
>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 | 15 +++
> lib/dpif-netdev.c|  9 +
> lib/ovs-numa.c   | 18 ++
> lib/ovs-numa.h   |  1 +
> 4 files changed, 39 insertions(+), 4 deletions(-)
>
>diff --git a/INSTALL.DPDK-ADVANCED.md b/INSTALL.DPDK-ADVANCED.md
>index 9ae536d..d828290 100644
>--- a/INSTALL.DPDK-ADVANCED.md
>+++ b/INSTALL.DPDK-ADVANCED.md
>@@ -205,8 +205,10 @@ 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`
>
>@@ -234,8 +236,10 @@ 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

There's some repetition in the last paragraph - I'm reviewing this patch in 
isolation, so the text may make sense/be required in the full document.

>
>   `ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=6`
>
>@@ -246,6 +250,9 @@ 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.

Although it's mentioned in the commit message, it might be worth mentioning 
here the consequences of attempting to pin non-PMD processes to a pmd-cpu-mask 
core (i.e. CPU starvation)

>+
> ### 4.3 DPDK physical port Rx Queues
>
>   `ovs-vsctl set Interface  options:n_rxq=`
>diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
>index e0107b7..805d0ae 100644
>--- a/lib/dpif-netdev.c
>+++ b/lib/dpif-netdev.c
>@@ -2851,6 +2851,15 @@ 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, _list);
>+
>+/* When cpu affinity mask explicitly set using pmd-cpu-mask, pmd thread's
>+ * scheduling policy is set to SCHED_RR and the 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 (pmd->dp->pmd_cmask) {
>+ovs_numa_thread_setpriority(SCHED_RR);
>+}
> reload:
> emc_cache_init(>flow_cache);
>
>diff --git a/lib/ovs-numa.c b/lib/ovs-numa.c
>index 7652636..d9f8ea1 100644
>--- 

[ovs-dev] [PATCH v3] netdev-dpdk: Set pmd thread priority

2016-07-18 Thread Bhanuprakash Bodireddy
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.

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 prority gets applied.

* If pmd-cpu-mask is specified, one ore 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).

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 &

With this commit OVS control threads and pmd threads can't have same
affinity ('dpdk-lcore-mask','pmd-cpu-mask' should be non-overlapping).
Also other processes with same affinity as PMD thread will be unresponsive.

Signed-off-by: Bhanuprakash Bodireddy 
---

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 | 15 +++
 lib/dpif-netdev.c|  9 +
 lib/ovs-numa.c   | 18 ++
 lib/ovs-numa.h   |  1 +
 4 files changed, 39 insertions(+), 4 deletions(-)

diff --git a/INSTALL.DPDK-ADVANCED.md b/INSTALL.DPDK-ADVANCED.md
index 9ae536d..d828290 100644
--- a/INSTALL.DPDK-ADVANCED.md
+++ b/INSTALL.DPDK-ADVANCED.md
@@ -205,8 +205,10 @@ 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`
 
@@ -234,8 +236,10 @@ 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`
 
@@ -246,6 +250,9 @@ 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.
+
 ### 4.3 DPDK physical port Rx Queues
 
   `ovs-vsctl set Interface  options:n_rxq=`
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index e0107b7..805d0ae 100644
--- a/lib/dpif-netdev.c
+++ b/lib/dpif-netdev.c
@@ -2851,6 +2851,15 @@ 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, _list);
+
+/* When cpu affinity mask explicitly set using pmd-cpu-mask, pmd thread's
+ * scheduling policy is set to SCHED_RR and the 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 (pmd->dp->pmd_cmask) {
+ovs_numa_thread_setpriority(SCHED_RR);
+}
 reload:
 emc_cache_init(>flow_cache);
 
diff --git a/lib/ovs-numa.c b/lib/ovs-numa.c
index 7652636..d9f8ea1 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(, 0, sizeof(threadparam));
+threadparam.sched_priority = sched_get_priority_max(policy);
+err = pthread_setschedparam(pthread_self(), policy, );
+if (err) {
+VLOG_ERR("Thread priority error %d",err);