Re: [PATCH v7 20/50] powerpc/powernv: Improve DMA32 segment calculation

[Daniel Axtens <dja@axtens.net>]

[-- Attachment #1: Type: text/plain, Size: 1234 bytes --]


> +	total_weight = pnv_pci_ioda_total_dma_weight(phb);
> +	weight = pnv_pci_ioda_pe_dma_weight(pe);
> +	if (!total_weight || !weight)
> +		return;
> +
> +	segs = (weight * phb->ioda.dma32_count) / total_weight;
> +	if (!segs)
> +		segs = 1;

I'm a little bit concerned about rounding here. Having said that I've
also lost track of dma32_count: if it's big then rounding won't
matter. What's a typical dma32_count?

> +
> +	/*
> +	 * Allocate continuous DMA32 segments. We begin with the expected
Very much a nit pick, but I think you mean s/continuous/contiguous/.

> +	 * number of segments. With one more attempt, the number of DMA32
> +	 * segments to be allocated is decreased by one until one segment
> +	 * is allocated successfully.
> +	 */
> +	while (segs) {
> +		found = false;
> +		for (base = 0; base <= phb->ioda.dma32_count - segs; base++) {
> +			for (i = base; i < base + segs; i++) {
> +				if (phb->ioda.dma32_segmap[i] !=
> +				    IODA_INVALID_PE)
> +					break;
> +			}
> +
> +			if (i >= base + segs) {
How would `i' ever be greater than base + segs? Should the test just
be 'if (i == base + segs) {' 
> +				found = true;
> +				break;
> +			}
> +		}

Regards,
Daniel

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 859 bytes --]