Re: [PATCH net-next v2 3/6] net: dsa: lantiq: allow arbitrary MII registers

On Thu, Jan 15, 2026 at 01:40:50AM +0000, Daniel Golle wrote:

On Thu, Jan 15, 2026 at 12:57:07AM +0000, Daniel Golle wrote:

quoted

The Lantiq GSWIP and MaxLinear GSW1xx drivers are currently relying on a
hard-coded mapping of MII ports to their respective MII_CFG and MII_PCDU
registers and only allow applying an offset to the port index.

While this is sufficient for the currently supported hardware, the very
similar Intel GSW150 (aka. Lantiq PEB7084) cannot be described using
this arrangement.

Introduce two arrays to specify the MII_CFG and MII_PCDU registers for
each port, replacing the current bitmap used to safeguard MII ports as
well as the port index offset.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
---
v2:
 * introduce GSWIP_MAX_PORTS macro

 drivers/net/dsa/lantiq/lantiq_gswip.c        | 30 ++++++++++++++++----
 drivers/net/dsa/lantiq/lantiq_gswip.h        |  6 ++--
 drivers/net/dsa/lantiq/lantiq_gswip_common.c | 27 +++---------------
 drivers/net/dsa/lantiq/mxl-gsw1xx.c          | 30 ++++++++++++++++----
 4 files changed, 56 insertions(+), 37 deletions(-)

diff --git a/drivers/net/dsa/lantiq/lantiq_gswip.c b/drivers/net/dsa/lantiq/lantiq_gswip.c
index b094001a7c805..4a1be6a1df6fe 100644
--- a/drivers/net/dsa/lantiq/lantiq_gswip.c
+++ b/drivers/net/dsa/lantiq/lantiq_gswip.c

@@ -463,10 +463,20 @@ static void gswip_shutdown(struct platform_device *pdev)
 }
 
 static const struct gswip_hw_info gswip_xrx200 = {
-	.max_ports = 7,
+	.max_ports = GSWIP_MAX_PORTS,
 	.allowed_cpu_ports = BIT(6),
-	.mii_ports = BIT(0) | BIT(1) | BIT(5),
-	.mii_port_reg_offset = 0,
+	.mii_cfg = {
+		[0 ... GSWIP_MAX_PORTS - 1] = -1,
+		[0] = GSWIP_MII_CFGp(0),
+		[1] = GSWIP_MII_CFGp(1),
+		[5] = GSWIP_MII_CFGp(5),
+	},

Kernel CI trips with
warning: initialized field overwritten [-Woverride-init]

Would it be ok to enclose the gswip_hw_info initializers with
__diag_push();
__diag_ignore_all("-Woverride-init",
		  "logic to initialize all and then override some is OK");

like it is done in drivers/net/ethernet/renesas/sh_eth.c?

Or should I rather keep the .mii_ports bitmap in addition to the array
to indicate the indexes with valid values?

This is new syntax for me. I don't have an issue either way. It seems
netdev maintainers did already accept it for another Ethernet driver, so
I suppose you could go with __diag_ignore_all().

FWIW, when I had to handle similar things, I opted for the simpler
"spell them out one by one" approach, see sja1110_regs in
drivers/net/dsa/sja1105/sja1105_spi.c. But that shouldn't necessarily be
the only way.