Patch "net: dsa: sja1105: fix -ENOSPC when replacing the same tc-cbs too many times" has been added to the 6.1-stable tree

[Sasha Levin <sashal@xxxxxxxxxx>]

This is a note to let you know that I've just added the patch titled

    net: dsa: sja1105: fix -ENOSPC when replacing the same tc-cbs too many times

to the 6.1-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     net-dsa-sja1105-fix-enospc-when-replacing-the-same-t.patch
and it can be found in the queue-6.1 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.



commit 072bd8e8f4f1aa0897b58ced9bc86bdde4c1ee30
Author: Vladimir Oltean <vladimir.oltean@xxxxxxx>
Date:   Wed Sep 6 00:53:37 2023 +0300

    net: dsa: sja1105: fix -ENOSPC when replacing the same tc-cbs too many times
    
    [ Upstream commit 894cafc5c62ccced758077bd4e970dc714c42637 ]
    
    After running command [2] too many times in a row:
    
    [1] $ tc qdisc add dev sw2p0 root handle 1: mqprio num_tc 8 \
            map 0 1 2 3 4 5 6 7 queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 hw 0
    [2] $ tc qdisc replace dev sw2p0 parent 1:1 cbs offload 1 \
            idleslope 120000 sendslope -880000 locredit -1320 hicredit 180
    
    (aka more than priv->info->num_cbs_shapers times)
    
    we start seeing the following error message:
    
    Error: Specified device failed to setup cbs hardware offload.
    
    This comes from the fact that ndo_setup_tc(TC_SETUP_QDISC_CBS) presents
    the same API for the qdisc create and replace cases, and the sja1105
    driver fails to distinguish between the 2. Thus, it always thinks that
    it must allocate the same shaper for a {port, queue} pair, when it may
    instead have to replace an existing one.
    
    Fixes: 4d7525085a9b ("net: dsa: sja1105: offload the Credit-Based Shaper qdisc")
    Signed-off-by: Vladimir Oltean <vladimir.oltean@xxxxxxx>
    Signed-off-by: David S. Miller <davem@xxxxxxxxxxxxx>
    Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>

diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c
index 377f177502003..9dd5cdcda2843 100644
--- a/drivers/net/dsa/sja1105/sja1105_main.c
+++ b/drivers/net/dsa/sja1105/sja1105_main.c
@@ -2123,6 +2123,18 @@ static void sja1105_bridge_leave(struct dsa_switch *ds, int port,
 
 #define BYTES_PER_KBIT (1000LL / 8)
 
+static int sja1105_find_cbs_shaper(struct sja1105_private *priv,
+				   int port, int prio)
+{
+	int i;
+
+	for (i = 0; i < priv->info->num_cbs_shapers; i++)
+		if (priv->cbs[i].port == port && priv->cbs[i].prio == prio)
+			return i;
+
+	return -1;
+}
+
 static int sja1105_find_unused_cbs_shaper(struct sja1105_private *priv)
 {
 	int i;
@@ -2163,9 +2175,14 @@ static int sja1105_setup_tc_cbs(struct dsa_switch *ds, int port,
 	if (!offload->enable)
 		return sja1105_delete_cbs_shaper(priv, port, offload->queue);
 
-	index = sja1105_find_unused_cbs_shaper(priv);
-	if (index < 0)
-		return -ENOSPC;
+	/* The user may be replacing an existing shaper */
+	index = sja1105_find_cbs_shaper(priv, port, offload->queue);
+	if (index < 0) {
+		/* That isn't the case - see if we can allocate a new one */
+		index = sja1105_find_unused_cbs_shaper(priv);
+		if (index < 0)
+			return -ENOSPC;
+	}
 
 	cbs = &priv->cbs[index];
 	cbs->port = port;