.setup() can run again within one registration: removing another switch
from the tree tears the whole tree down, and re-probing sets every
member switch up again. The hardware reset at the start of .setup()
brings the MIB counters back to zero, but a MIB block kept across that
cycle still holds the old software buffer, so the next yt921x_mib_read()
treats the small value as a 32-bit wrap and adds 2^32 to every counter
it extends.
.port_teardown() keeps the MIB block on purpose and .port_setup() only
zeroes a freshly allocated one, so clear the buffer in .setup() together
with the hardware counters it extends. The published stats need no
handling; every entry is republished from the buffer.
Fixes: 186623f4aa72 ("net: dsa: yt921x: Add support for Motorcomm YT921x")
Signed-off-by: David Yang <mmyangfl@gmail.com>
---
drivers/net/dsa/motorcomm/chip.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/drivers/net/dsa/motorcomm/chip.c b/drivers/net/dsa/motorcomm/chip.c
index 33e5fc127244..ff17233848a8 100644
--- a/drivers/net/dsa/motorcomm/chip.c
+++ b/drivers/net/dsa/motorcomm/chip.c
@@ -2458,6 +2458,16 @@ static int yt921x_dsa_setup(struct dsa_switch *ds)
return -ENODEV;
}
+ for (int port = 0; port < YT921X_PORT_NUM; port++) {
+ struct yt921x_port *pp = &priv->ports[port];
+ struct yt921x_mib *pm = pp->mib;
+
+ if (!pm)
+ continue;
+
+ memset(&pm->data, 0, sizeof(pm->data));
+ }
+
mask = priv->info->serdes_mask;
for_each_set_bit(port, &mask, YT921X_PORT_NUM) {
struct yt921x_port *pp = &priv->ports[port];--
2.53.0