| Message ID | 1560966016-28254-3-git-send-email-vedang.patel@intel.com |
|---|---|
| State | Superseded |
| Headers | show |
| Series | net/sched: Add txtime-assist support for taprio. | expand |
On 19.06.2019 20:40, Vedang Patel wrote: > Currently, etf expects a socket with SO_TXTIME option set for each packet > it encounters. So, it will drop all other packets. But, in the future > commits we are planning to add functionality which where tstamp value will One of "which" and "where", not both. :-) > be set by another qdisc. Also, some packets which are generated from within > the kernel (e.g. ICMP packets) do not have any socket associated with them. > > So, this commit adds support for skip_sock_check. When this option is set, > etf will skip checking for a socket and other associated options for all > skbs. > > Signed-off-by: Vedang Patel <vedang.patel@intel.com> [...] MBR, Sergei
> On Jun 20, 2019, at 1:16 AM, Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> wrote: > > On 19.06.2019 20:40, Vedang Patel wrote: > >> Currently, etf expects a socket with SO_TXTIME option set for each packet >> it encounters. So, it will drop all other packets. But, in the future >> commits we are planning to add functionality which where tstamp value will > > One of "which" and "where", not both. :-) > Yeah. It’s a typo. Will fix it in next version. Thanks, Vedang >> be set by another qdisc. Also, some packets which are generated from within >> the kernel (e.g. ICMP packets) do not have any socket associated with them. >> So, this commit adds support for skip_sock_check. When this option is set, >> etf will skip checking for a socket and other associated options for all >> skbs. >> Signed-off-by: Vedang Patel <vedang.patel@intel.com> > [...] > > MBR, Sergei
diff --git a/include/uapi/linux/pkt_sched.h b/include/uapi/linux/pkt_sched.h index 8b2f993cbb77..409d1616472d 100644 --- a/include/uapi/linux/pkt_sched.h +++ b/include/uapi/linux/pkt_sched.h @@ -990,6 +990,7 @@ struct tc_etf_qopt { __u32 flags; #define TC_ETF_DEADLINE_MODE_ON BIT(0) #define TC_ETF_OFFLOAD_ON BIT(1) +#define TC_ETF_SKIP_SOCK_CHECK BIT(2) }; enum { diff --git a/net/sched/sch_etf.c b/net/sched/sch_etf.c index db0c2ba1d156..cebfb65d8556 100644 --- a/net/sched/sch_etf.c +++ b/net/sched/sch_etf.c @@ -22,10 +22,12 @@ #define DEADLINE_MODE_IS_ON(x) ((x)->flags & TC_ETF_DEADLINE_MODE_ON) #define OFFLOAD_IS_ON(x) ((x)->flags & TC_ETF_OFFLOAD_ON) +#define SKIP_SOCK_CHECK_IS_SET(x) ((x)->flags & TC_ETF_SKIP_SOCK_CHECK) struct etf_sched_data { bool offload; bool deadline_mode; + bool skip_sock_check; int clockid; int queue; s32 delta; /* in ns */ @@ -77,6 +79,9 @@ static bool is_packet_valid(struct Qdisc *sch, struct sk_buff *nskb) struct sock *sk = nskb->sk; ktime_t now; + if (q->skip_sock_check) + goto skip; + if (!sk) return false; @@ -92,6 +97,7 @@ static bool is_packet_valid(struct Qdisc *sch, struct sk_buff *nskb) if (sk->sk_txtime_deadline_mode != q->deadline_mode) return false; +skip: now = q->get_time(); if (ktime_before(txtime, now) || ktime_before(txtime, q->last)) return false; @@ -385,6 +391,7 @@ static int etf_init(struct Qdisc *sch, struct nlattr *opt, q->clockid = qopt->clockid; q->offload = OFFLOAD_IS_ON(qopt); q->deadline_mode = DEADLINE_MODE_IS_ON(qopt); + q->skip_sock_check = SKIP_SOCK_CHECK_IS_SET(qopt); switch (q->clockid) { case CLOCK_REALTIME: @@ -473,6 +480,9 @@ static int etf_dump(struct Qdisc *sch, struct sk_buff *skb) if (q->deadline_mode) opt.flags |= TC_ETF_DEADLINE_MODE_ON; + if (q->skip_sock_check) + opt.flags |= TC_ETF_SKIP_SOCK_CHECK; + if (nla_put(skb, TCA_ETF_PARMS, sizeof(opt), &opt)) goto nla_put_failure;
Currently, etf expects a socket with SO_TXTIME option set for each packet it encounters. So, it will drop all other packets. But, in the future commits we are planning to add functionality which where tstamp value will be set by another qdisc. Also, some packets which are generated from within the kernel (e.g. ICMP packets) do not have any socket associated with them. So, this commit adds support for skip_sock_check. When this option is set, etf will skip checking for a socket and other associated options for all skbs. Signed-off-by: Vedang Patel <vedang.patel@intel.com> --- include/uapi/linux/pkt_sched.h | 1 + net/sched/sch_etf.c | 10 ++++++++++ 2 files changed, 11 insertions(+)