> On Thu, Nov 14, 2019 at 10:43:09PM +0100, Jesper Dangaard Brouer wrote:
> > On Thu, 14 Nov 2019 13:04:26 -0800
> > "Jonathan Lemon" <jonathan.lemon@gmail.com> wrote:
> > 
> > > On 14 Nov 2019, at 12:42, Ilias Apalodimas wrote:
> > > 
> > > > Hi Jonathan,
> > > >
> > > > On Thu, Nov 14, 2019 at 12:27:40PM -0800, Jonathan Lemon wrote:  
> > > >>
> > > >>
> > > >> On 14 Nov 2019, at 10:53, Ilias Apalodimas wrote:
> > > >>  
> > > >>> [...]  
> > > >>>>> index 2cbcdbdec254..defbfd90ab46 100644
> > > >>>>> --- a/include/net/page_pool.h
> > > >>>>> +++ b/include/net/page_pool.h
> > > >>>>> @@ -65,6 +65,9 @@ struct page_pool_params {
> > > >>>>>  	int		nid;  /* Numa node id to allocate from pages from */
> > > >>>>>  	struct device	*dev; /* device, for DMA pre-mapping purposes */
> > > >>>>>  	enum dma_data_direction dma_dir; /* DMA mapping direction */
> > > >>>>> +	unsigned int	max_len; /* max DMA sync memory size */
> > > >>>>> +	unsigned int	offset;  /* DMA addr offset */
> > > >>>>> +	u8 sync;
> > > >>>>>  };  
> > > >>>>
> > > >>>> How about using PP_FLAG_DMA_SYNC instead of another flag word?
> > > >>>> (then it can also be gated on having DMA_MAP enabled)  
> > > >>>
> > > >>> You mean instead of the u8?
> > > >>> As you pointed out on your V2 comment of the mail, some cards don't 
> > > >>> sync back to device.
> > > >>>
> > > >>> As the API tries to be generic a u8 was choosen instead of a flag
> > > >>> to cover these use cases. So in time we'll change the semantics of
> > > >>> this to 'always sync', 'dont sync if it's an skb-only queue' etc.
> > > >>>
> > > >>> The first case Lorenzo covered is sync the required len only instead 
> > > >>> of the full buffer  
> > > >>
> > > >> Yes, I meant instead of:
> > > >> +		.sync = 1,
> > > >>
> > > >> Something like:
> > > >>         .flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC
> > > >>
> > 
> > I actually agree and think we could use a flag. I suggest
> > PP_FLAG_DMA_SYNC_DEV to indicate that this DMA-sync-for-device.
> > 
> > Ilias notice that the change I requested to Lorenzo, that dma_sync_size
> > default value is 0xFFFFFFFF (-1).  That makes dma_sync_size==0 a valid
> > value, which you can use in the cases, where you know that nobody have
> > written into the data-area.  This allow us to selectively choose it for
> > these cases.
> 
> Okay, then i guess the flag is a better fit for this.
> The only difference would be that the sync semantics will be done on 'per
> packet' basis,  instead of 'per pool', but that should be fine for our cases.

Ack, fine for me.
Do you think when checking for PP_FLAG_DMA_SYNC_DEV we should even verify
PP_FLAG_DMA_MAP? Something like:

if ((pool->p.flags & (PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV)) ==
    (PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV))
	page_pool_dma_sync_for_device();

Regards,
Lorenzo

> 
> Cheers
> /Ilias
> > 
> > -- 
> > Best regards,
> >   Jesper Dangaard Brouer
> >   MSc.CS, Principal Kernel Engineer at Red Hat
> >   LinkedIn: http://www.linkedin.com/in/brouer
> > 
>