[PATCH net-next] net: dsa: mt7530: drop paranoid checks in .get_tag_protocol()

It is desirable to reduce the surface of DSA_TAG_PROTO_NONE as much as
we can, because we now have options for switches without hardware
support for DSA tagging, and the occurrence in the mt7530 driver is in
fact quite gratuitout and easy to remove. Since ds->ops->get_tag_protocol()
is only called for CPU ports, the checks for a CPU port in
mtk_get_tag_protocol() are redundant and can be removed.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
---
 drivers/net/dsa/mt7530.c | 10 +---------
 1 file changed, 1 insertion(+), 9 deletions(-)

diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c
index 69f21b71614c..b6e0b347947e 100644
--- a/drivers/net/dsa/mt7530.c
+++ b/drivers/net/dsa/mt7530.c

@@ -1717,15 +1717,7 @@ static enum dsa_tag_protocol
 mtk_get_tag_protocol(struct dsa_switch *ds, int port,
 		     enum dsa_tag_protocol mp)
 {
-	struct mt7530_priv *priv = ds->priv;
-
-	if (port != MT7530_CPU_PORT) {
-		dev_warn(priv->dev,
-			 "port not matched with tagging CPU port\n");
-		return DSA_TAG_PROTO_NONE;
-	} else {
-		return DSA_TAG_PROTO_MTK;
-	}
+	return DSA_TAG_PROTO_MTK;
 }
 
 #ifdef CONFIG_GPIOLIB

-- 
2.25.1

On Sat, Jul 31, 2021 at 01:57:14AM +0300, Vladimir Oltean wrote:

Acked-by: DENG Qingfang <dqfext@gmail.com>

On 7/30/2021 3:57 PM, Vladimir Oltean wrote:

It is desirable to reduce the surface of DSA_TAG_PROTO_NONE as much as
we can, because we now have options for switches without hardware
support for DSA tagging, and the occurrence in the mt7530 driver is in
fact quite gratuitout and easy to remove. Since ds->ops->get_tag_protocol()
is only called for CPU ports, the checks for a CPU port in
mtk_get_tag_protocol() are redundant and can be removed.

The point of the check was in case the designated CPU port from device 
tree/platform data would not match what the Mediatek driver supports, 
similar to what b53 does in the same vein. I am fine with removing that 
check for mt7530 as it does not look like there is an use case where the 
CPU port is not actually MT7530_CPU_PORT.

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

On Fri, Jul 30, 2021 at 08:08:43PM -0700, Florian Fainelli wrote:

On 7/30/2021 3:57 PM, Vladimir Oltean wrote:

quoted

It is desirable to reduce the surface of DSA_TAG_PROTO_NONE as much as
we can, because we now have options for switches without hardware
support for DSA tagging, and the occurrence in the mt7530 driver is in
fact quite gratuitout and easy to remove. Since ds->ops->get_tag_protocol()
is only called for CPU ports, the checks for a CPU port in
mtk_get_tag_protocol() are redundant and can be removed.

The point of the check was in case the designated CPU port from device
tree/platform data would not match what the Mediatek driver supports,
similar to what b53 does in the same vein. I am fine with removing that
check for mt7530 as it does not look like there is an use case where the CPU
port is not actually MT7530_CPU_PORT.

So if mt7530 only supports port 6 as a CPU port, how is Qingfang testing
multiple CPU ports on it?

Hello:

This patch was applied to netdev/net-next.git (refs/heads/master):

On Sat, 31 Jul 2021 01:57:14 +0300 you wrote:

It is desirable to reduce the surface of DSA_TAG_PROTO_NONE as much as
we can, because we now have options for switches without hardware
support for DSA tagging, and the occurrence in the mt7530 driver is in
fact quite gratuitout and easy to remove. Since ds->ops->get_tag_protocol()
is only called for CPU ports, the checks for a CPU port in
mtk_get_tag_protocol() are redundant and can be removed.

[...]

Here is the summary with links:
  - [net-next] net: dsa: mt7530: drop paranoid checks in .get_tag_protocol()
    https://git.kernel.org/netdev/net-next/c/244f8a802911

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html