Re: [PATCH 8/8] ath10k: allow explicit MSI/MSI-X disabling

[Kalle Valo <kvalo@qca.qualcomm.com>]

Michal Kazior <michal.kazior@tieto.com> writes:

> This can be useful for testing and debugging.
>
> Two new ath10k_pci module options are now available:
>
>  disable_msix - disable just MSI-X
>  disable_msi - disable MSI (along with MSI-X)
>
> Signed-off-by: Michal Kazior <michal.kazior@tieto.com>

This is useful but I think it's a bit too much to have two more options,
it would be simpler to have just one. Few quick ideas how to do this a
bit differently:

irq_mode: 0 = automatic, 2 = MSI, 1 = legacy

Here the problem is of course that we can't force use of MSI if the host
only supports legacy interrupts, but IMHO that's a minor nuisance and in
that case we should just enable legacy interrupts.

disable_msi: 0x1 = disable MSI-X, 0x2 = disable MSI

Basically this is just your booleans converted to a bitfield.

I vote for irq_mode because I think that's easier for the user to
understand. But I'm sure there is a better way to do this.

> +	if (msix_supported && !ath10k_disable_msix && !ath10k_disable_msi) {
> +		ar_pci->num_msi_intrs = MSI_NUM_REQUEST;
> +		ret = pci_enable_msi_block(ar_pci->pdev, ar_pci->num_msi_intrs);
> +		if (ret == 0)
> +			return 0;
> +		if (ret > 0)
> +			pci_disable_msi(ar_pci->pdev);
> +
> +		/* fall-through */
> +	}

If we force some other irq mode than the default it would be good to
have an info print for that.

-- 
Kalle Valo

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k