Re: [RFC net-next 19/22] eth: bnxt: use queue op config validate

[Jakub Kicinski <kuba@kernel.org>]

On Wed, 23 Apr 2025 10:00:01 +0000 Dragos Tatulea wrote:
> > +static int
> > +bnxt_queue_cfg_validate(struct net_device *dev, int idx,
> > +			struct netdev_queue_config *qcfg,
> > +			struct netlink_ext_ack *extack)
> > +{
> > +	struct bnxt *bp = netdev_priv(dev);
> > +
> > +	/* Older chips need MSS calc so rx_buf_len is not supported,
> > +	 * but we don't set queue ops for them so we should never get here.
> > +	 */
> > +	if (qcfg->rx_buf_len != bp->rx_page_size &&
> > +	    !(bp->flags & BNXT_FLAG_CHIP_P5_PLUS)) {
> > +		NL_SET_ERR_MSG_MOD(extack, "changing rx-buf-len not supported");
> > +		return -EINVAL;
> > +	}
> > +
> > +	if (!is_power_of_2(qcfg->rx_buf_len)) {
> > +		NL_SET_ERR_MSG_MOD(extack, "rx-buf-len is not power of 2");
> > +		return -ERANGE;
> > +	}
> > +	if (qcfg->rx_buf_len < BNXT_RX_PAGE_SIZE ||
> > +	    qcfg->rx_buf_len > BNXT_MAX_RX_PAGE_SIZE) {
> > +		NL_SET_ERR_MSG_MOD(extack, "rx-buf-len out of range");
> > +		return -ERANGE;
> > +	}
> > +	return 0;
> > +}
> > +  
> HDS off and rx_buf_len > 4K seems to be accepted. Is this inteded?

For bnxt rx_buf_len only applies to the "payload buffers".
I should document that, and retest with XDP. 

I posted a doc recently with a "design guide" for API interfaces, 
it said:

  Visibility
  ==========

  To simplify the implementations configuration parameters of disabled features
  do not have to be hidden, or inaccessible.

Which I intended to mean that configuring something that isn't enabled
is okay. IIRC we also don't reject setting hds threshold if hds is off.

Hope I understood what you're getting at.