[PATCH 5.18 1084/1095] net: dsa: felix: fix min gate len calculation for tc when its first gate is closed

[Greg Kroah-Hartman]

From: Vladimir Oltean <vladimir.oltean@xxxxxxx>

commit 7e4babffa6f340a74c820d44d44d16511e666424 upstream.

min_gate_len[tc] is supposed to track the shortest interval of
continuously open gates for a traffic class. For example, in the
following case:

TC 76543210

t0 00000001b 200000 ns
t1 00000010b 200000 ns

min_gate_len[0] and min_gate_len[1] should be 200000, while
min_gate_len[2-7] should be 0.

However what happens is that min_gate_len[0] is 200000, but
min_gate_len[1] ends up being 0 (despite gate_len[1] being 200000 at the
point where the logic detects the gate close event for TC 1).

The problem is that the code considers a "gate close" event whenever it
sees that there is a 0 for that TC (essentially it's level rather than
edge triggered). By doing that, any time a gate is seen as closed
without having been open prior, gate_len, which is 0, will be written
into min_gate_len. Once min_gate_len becomes 0, it's impossible for it
to track anything higher than that (the length of actually open
intervals).

To fix this, we make the writing to min_gate_len[tc] be edge-triggered,
which avoids writes for gates that are closed in consecutive intervals.
However what this does is it makes us need to special-case the
permanently closed gates at the end.

Fixes: 55a515b1f5a9 ("net: dsa: felix: drop oversized frames with tc-taprio instead of hanging the port")
Signed-off-by: Vladimir Oltean <vladimir.oltean@xxxxxxx>
Link: https://lore.kernel.org/r/20220804202817.1677572-1-vladimir.oltean@xxxxxxx
Signed-off-by: Jakub Kicinski <kuba@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
 drivers/net/dsa/ocelot/felix_vsc9959.c |   15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

--- a/drivers/net/dsa/ocelot/felix_vsc9959.c
+++ b/drivers/net/dsa/ocelot/felix_vsc9959.c
@@ -1133,6 +1133,7 @@ static void vsc9959_tas_min_gate_lengths
 {
 	struct tc_taprio_sched_entry *entry;
 	u64 gate_len[OCELOT_NUM_TC];
+	u8 gates_ever_opened = 0;
 	int tc, i, n;
 
 	/* Initialize arrays */
@@ -1160,16 +1161,28 @@ static void vsc9959_tas_min_gate_lengths
 		for (tc = 0; tc < OCELOT_NUM_TC; tc++) {
 			if (entry->gate_mask & BIT(tc)) {
 				gate_len[tc] += entry->interval;
+				gates_ever_opened |= BIT(tc);
 			} else {
 				/* Gate closes now, record a potential new
 				 * minimum and reinitialize length
 				 */
-				if (min_gate_len[tc] > gate_len[tc])
+				if (min_gate_len[tc] > gate_len[tc] &&
+				    gate_len[tc])
 					min_gate_len[tc] = gate_len[tc];
 				gate_len[tc] = 0;
 			}
 		}
 	}
+
+	/* min_gate_len[tc] actually tracks minimum *open* gate time, so for
+	 * permanently closed gates, min_gate_len[tc] will still be U64_MAX.
+	 * Therefore they are currently indistinguishable from permanently
+	 * open gates. Overwrite the gate len with 0 when we know they're
+	 * actually permanently closed, i.e. after the loop above.
+	 */
+	for (tc = 0; tc < OCELOT_NUM_TC; tc++)
+		if (!(gates_ever_opened & BIT(tc)))
+			min_gate_len[tc] = 0;
 }
 
 /* Update QSYS_PORT_MAX_SDU to make sure the static guard bands added by the