Received-SPF: pass (google.com: domain of linux-kernel+bounces-4871-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
From: Faizal Rahim <faizal.abdul.rahim@linux.intel.com>
To: Vladimir Oltean <vladimir.oltean@nxp.com>,
	Vinicius Costa Gomes <vinicius.gomes@intel.com>,
	Jamal Hadi Salim <jhs@mojatatu.com>,
	Cong Wang <xiyou.wangcong@gmail.com>,
	Jiri Pirko <jiri@resnulli.us>,
	"David S . Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>,
	Paolo Abeni <pabeni@redhat.com>
Cc: netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: [PATCH v3 net 1/4] net/sched: taprio: fix too early schedules switching
Date: Tue, 19 Dec 2023 03:14:50 -0500
Message-Id: <20231219081453.718489-2-faizal.abdul.rahim@linux.intel.com>
In-Reply-To: <20231219081453.718489-1-faizal.abdul.rahim@linux.intel.com>
References: <20231219081453.718489-1-faizal.abdul.rahim@linux.intel.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

Switching the schedule prematurely leads to a situation where the last
entry from oper schedule is still running, during this period, calls
to taprio_skb_exceeds_queue_max_sdu() in the enqueue path, such as
taprio_enqueue_segmented(), will inspect q->oper_sched. At this point,
q->oper_sched refers to the new admin schedule instead of the ongoing
oper schedule.

Fixes: a878fd46fe43 ("net/sched: keep the max_frm_len information inside struct sched_gate_list")
Signed-off-by: Faizal Rahim <faizal.abdul.rahim@linux.intel.com>
---
 net/sched/sch_taprio.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
index 31a8252bd09c..bbcaf05d40ba 100644
--- a/net/sched/sch_taprio.c
+++ b/net/sched/sch_taprio.c
@@ -41,6 +41,7 @@ static struct static_key_false taprio_have_working_mqprio;
 #define TXTIME_ASSIST_IS_ENABLED(flags) ((flags) & TCA_TAPRIO_ATTR_FLAG_TXTIME_ASSIST)
 #define FULL_OFFLOAD_IS_ENABLED(flags) ((flags) & TCA_TAPRIO_ATTR_FLAG_FULL_OFFLOAD)
 #define TAPRIO_FLAGS_INVALID U32_MAX
+#define CYCLE_TIME_CORRECTION_UNSPEC S64_MIN
 
 struct sched_entry {
 	/* Durations between this GCL entry and the GCL entry where the
@@ -75,6 +76,7 @@ struct sched_gate_list {
 	ktime_t cycle_end_time;
 	s64 cycle_time;
 	s64 cycle_time_extension;
+	s64 cycle_time_correction;
 	s64 base_time;
 };
 
@@ -213,6 +215,11 @@ static void switch_schedules(struct taprio_sched *q,
 	*admin = NULL;
 }
 
+static bool sched_switch_pending(const struct sched_gate_list *oper)
+{
+	return oper->cycle_time_correction != CYCLE_TIME_CORRECTION_UNSPEC;
+}
+
 /* Get how much time has been already elapsed in the current cycle. */
 static s32 get_cycle_time_elapsed(struct sched_gate_list *sched, ktime_t time)
 {
@@ -940,7 +947,7 @@ static enum hrtimer_restart advance_sched(struct hrtimer *timer)
 	admin = rcu_dereference_protected(q->admin_sched,
 					  lockdep_is_held(&q->current_entry_lock));
 
-	if (!oper)
+	if (!oper || sched_switch_pending(oper))
 		switch_schedules(q, &admin, &oper);
 
 	/* This can happen in two cases: 1. this is the very first run
@@ -981,7 +988,7 @@ static enum hrtimer_restart advance_sched(struct hrtimer *timer)
 		 * schedule runs.
 		 */
 		end_time = sched_base_time(admin);
-		switch_schedules(q, &admin, &oper);
+		oper->cycle_time_correction = 0;
 	}
 
 	next->end_time = end_time;
@@ -1174,6 +1181,7 @@ static int parse_taprio_schedule(struct taprio_sched *q, struct nlattr **tb,
 	}
 
 	taprio_calculate_gate_durations(q, new);
+	new->cycle_time_correction = CYCLE_TIME_CORRECTION_UNSPEC;
 
 	return 0;
 }
-- 
2.25.1