Re: [Intel-wired-lan] [PATCH net-next 07/11] net: page_pool: add DMA-sync-for-CPU inline helpers

[Jakub Kicinski]

On Thu, 18 May 2023 15:45:33 +0200 Alexander Lobakin wrote:
> >> diff --git a/include/net/page_pool.h b/include/net/page_pool.h
> >> index 8435013de06e..f740c50b661f 100644
> >> --- a/include/net/page_pool.h
> >> +++ b/include/net/page_pool.h
> >> @@ -32,7 +32,7 @@
> >>  
> >>  #include <linux/mm.h> /* Needed by ptr_ring */
> >>  #include <linux/ptr_ring.h>
> >> -#include <linux/dma-direction.h>
> >> +#include <linux/dma-mapping.h>  
> > 
> > highly nit picky - but isn't dma-mapping.h pretty heavy?
> > And we include page_pool.h in skbuff.h. Not that it matters
> > today, but maybe one day we'll succeed putting skbuff.h
> > on a diet -- so perhaps it's better to put "inline helpers
> > with non-trivial dependencies" into a new header?  
> 
> Maybe we could rather stop including page_pool.h into skbuff.h? It's
> used there only for  1 external, which could be declared directly in
> skbuff.h. When Matteo was developing PP recycling, he was storing
> mem_info in skb as well, but then it was optimized and we don't do that
> anymore.
> It annoys sometimes to see the whole kernel rebuilt each time I edit
> pag_pool.h :D In fact, only PP-enabled drivers and core code need it.

Or maybe we can do both? I think that separating types, defines and
simple wrappers from helpers should be considered good code hygiene.

> >>  #define PP_FLAG_DMA_MAP		BIT(0) /* Should page_pool do the DMA
> >>  					* map/unmap  
> >   
> >> +/**
> >> + * page_pool_dma_sync_for_cpu - sync Rx page for CPU after it's written by HW
> >> + * @pool: page_pool which this page belongs to
> >> + * @page: page to sync
> >> + * @dma_sync_size: size of the data written to the page
> >> + *
> >> + * Can be used as a shorthand to sync Rx pages before accessing them in the
> >> + * driver. Caller must ensure the pool was created with %PP_FLAG_DMA_MAP.
> >> + */
> >> +static inline void page_pool_dma_sync_for_cpu(const struct page_pool *pool,
> >> +					      const struct page *page,
> >> +					      u32 dma_sync_size)
> >> +{
> >> +	dma_sync_single_range_for_cpu(pool->p.dev,
> >> +				      page_pool_get_dma_addr(page),
> >> +				      pool->p.offset, dma_sync_size,
> >> +				      page_pool_get_dma_dir(pool));  
> > 
> > Likely a dumb question but why does this exist?
> > Is there a case where the "maybe" version is not safe?  
> 
> If the driver doesn't set DMA_SYNC_DEV flag, then the "maybe" version
> will never do anything. But we may want to use these helpers in such
> drivers too?

Oh, I see, the polarity of the flag is awkward. Hm.
Maybe just rename things, drop the "maybe_" and prefix the non-checking
version with __ ? We expect drivers to call the version which check the
flag mostly (AFAIU), so it should have the most obvious name.
Plus perhaps a sentence in the kdoc explaining why __ exists would be
good, if it wasn't obvious to me it may not be obvious to others..