Re: [PATCH 1/2] dma-mapping: prepare dma_map_ops to conversion to physical address

[Leon Romanovsky <leon@kernel.org>]

On Thu, Jul 17, 2025 at 02:07:58PM -0300, Jason Gunthorpe wrote:
> On Thu, Jul 17, 2025 at 05:52:13PM +0300, Leon Romanovsky wrote:
> > > I would expect 'else if'
> > 
> > I don't expect both of them to exist. There is WARN_ON_ONCE() to catch it.
> > If it is important, I can add else.
> 
> Compiler generates better code with the else.
> 
> > > > @@ -213,8 +221,16 @@ void dma_unmap_phys(struct device *dev, dma_addr_t addr, size_t size,
> > > >  		dma_direct_unmap_phys(dev, addr, size, dir, attrs);
> > > >  	else if (use_dma_iommu(dev))
> > > >  		iommu_dma_unmap_phys(dev, addr, size, dir, attrs);
> > > > -	else
> > > > -		ops->unmap_page(dev, addr, size, dir, attrs);
> > > > +	else {
> > > > +		if (IS_ENABLED(CONFIG_DMA_API_DEBUG))
> > > > +			/* We shouldn't have both functions */
> > > > +			WARN_ON_ONCE(ops->unmap_page && ops->unmap_phys);
> > > > +
> > > > +		if (ops->unmap_phys)
> > > > +			ops->unmap_phys(dev, addr, size, dir, attrs);
> > > > +		if (ops->unmap_page)
> > > > +			ops->unmap_page(dev, addr, size, dir, attrs);
> > > 
> > > Here too
> > > 
> > > What is the practical difference between unmap_phys and unmap_page?
> > 
> > There is no difference. It is provided to provide symmetrical interface
> > to .map_phys, simple rename.
> 
> We really should not be converting phys to page without knowing there
> is a page available..

I didn't want. See this email and followup discussion.
https://lore.kernel.org/all/20250627170213.GL17401@unreal/

Thanks