| Message ID | ZgSzhZBJSUyme1Lk@tpad (mailing list archive) |
|---|---|
| State | Changes Requested |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [net-next,-v6] net: enable timestamp static key if CPU isolation is configured | expand |
Marcelo Tosatti wrote: > > For systems that use CPU isolation (via nohz_full), creating or destroying > a socket with SO_TIMESTAMP, SO_TIMESTAMPNS or SO_TIMESTAMPING with flag > SOF_TIMESTAMPING_RX_SOFTWARE will cause a static key to be enabled/disabled. > This in turn causes undesired IPIs to isolated CPUs. > > So enable the static key unconditionally, if CPU isolation is enabled, > thus avoiding the IPIs. > > Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com> Reviewed-by: Willem de Bruijn <willemb@google.com>
On Wed, 27 Mar 2024 21:02:13 -0300 Marcelo Tosatti wrote: > #include <net/page_pool/types.h> > #include <net/page_pool/helpers.h> > #include <net/rps.h> > +#include <linux/sched/isolation.h> Sorry for a late nit, could you put this include next to the other two linux/sched* includes earlier in the file?
diff --git a/net/core/dev.c b/net/core/dev.c index 5d36a634f468..48c725caa130 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -157,6 +157,7 @@ #include <net/page_pool/types.h> #include <net/page_pool/helpers.h> #include <net/rps.h> +#include <linux/sched/isolation.h> #include "dev.h" #include "net-sysfs.h" @@ -11890,6 +11891,10 @@ static int __init net_dev_init(void) NULL, dev_cpu_dead); WARN_ON(rc < 0); rc = 0; + + /* avoid static key IPIs to isolated CPUs */ + if (housekeeping_enabled(HK_TYPE_MISC)) + net_enable_timestamp(); out: if (rc < 0) { for_each_possible_cpu(i) {
For systems that use CPU isolation (via nohz_full), creating or destroying a socket with SO_TIMESTAMP, SO_TIMESTAMPNS or SO_TIMESTAMPING with flag SOF_TIMESTAMPING_RX_SOFTWARE will cause a static key to be enabled/disabled. This in turn causes undesired IPIs to isolated CPUs. So enable the static key unconditionally, if CPU isolation is enabled, thus avoiding the IPIs. Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com> --- v2: mention SOF_TIMESTAMPING_OPT_TX_SWHW in the commit log (Willem de Bruijn / Paolo Abeni) v3: SOF_TIMESTAMPING_OPT_TX_SWHW is irrelevant (Willem de Bruijn) v4: additional changelog improvements (Willem de Bruijn) v5: late initcall not necessary, can use subsys initcall (Willem de Bruijn) v6: rebase against net-next, change subject (Jakub Kicinski)