Re: [PATCH v5 2/2] net: dsa: qca: ar9331: export stats64

[Jakub Kicinski]

On Fri, 11 Dec 2020 11:53:22 +0100 Oleksij Rempel wrote:
> Add stats support for the ar9331 switch.
> 
> Signed-off-by: Oleksij Rempel <o.rempel@xxxxxxxxxxxxxx>
> ---
>  drivers/net/dsa/qca/ar9331.c | 256 ++++++++++++++++++++++++++++++++++-
>  1 file changed, 255 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/dsa/qca/ar9331.c b/drivers/net/dsa/qca/ar9331.c
> index 4d49c5f2b790..5baef0ec6410 100644
> --- a/drivers/net/dsa/qca/ar9331.c
> +++ b/drivers/net/dsa/qca/ar9331.c
> @@ -101,6 +101,9 @@
>  	 AR9331_SW_PORT_STATUS_RX_FLOW_EN | AR9331_SW_PORT_STATUS_TX_FLOW_EN | \
>  	 AR9331_SW_PORT_STATUS_SPEED_M)
>  
> +/* MIB registers */
> +#define AR9331_MIB_COUNTER(x)			(0x20000 + ((x) * 0x100))
> +
>  /* Phy bypass mode
>   * ------------------------------------------------------------------------
>   * Bit:   | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |10 |11 |12 |13 |14 |15 |
> @@ -154,6 +157,111 @@
>  #define AR9331_SW_MDIO_POLL_SLEEP_US		1
>  #define AR9331_SW_MDIO_POLL_TIMEOUT_US		20
>  
> +/* The interval should be small enough to avoid overflow of 32bit MIBs */
> +/*
> + * FIXME: as long as we can't read MIBs from stats64 call directly, we should
> + * poll stats more frequently then it is actually needed. In normal case
> + * 100 sec interval should be OK.

This comment is a little confusing, if you don't mind.

Should it says something like:

 FIXME: until we can read MIBs from stats64 call directly (i.e. sleep
 there), we have to poll stats more frequently then it is actually needed.
 For overflow protection, normally, 100 sec interval should have been OK.