Date:   Fri, 25 Nov 2022 16:34:35 +0100
From:   "Fabio M. De Francesco" <fmdefrancesco@...il.com>
To:     netdev@...r.kernel.org,
        Anirudh Venkataramanan <anirudh.venkataramanan@...el.com>
Cc:     Ira Weiny <ira.weiny@...el.com>,
        Anirudh Venkataramanan <anirudh.venkataramanan@...el.com>,
        Ayush Sawal <ayush.sawal@...lsio.com>
Subject: Re: [PATCH v2 net-next 1/6] ch_ktls: Use memcpy_from_page() instead of k[un]map_atomic()

On mercoledì 23 novembre 2022 21:52:14 CET Anirudh Venkataramanan wrote:
> kmap_atomic() is being deprecated in favor of kmap_local_page(). Replace
> the map-memcpy-unmap usage pattern (done using k[un]map_atomic()) with
> memcpy_from_page(), which internally uses kmap_local_page() and
> kunmap_local(). This renders the variables 'data' and 'vaddr' unnecessary,
> and so remove these too.
> 
> Note that kmap_atomic() disables preemption and page-fault processing, but
> kmap_local_page() doesn't. When converting uses of kmap_atomic(), one has
> to check if the code being executed between the map/unmap implicitly
> depends on page-faults and/or preemption being disabled. If yes, then code
> to disable page-faults and/or preemption should also be added for
> functional correctness. That however doesn't appear to be the case here,
> so just memcpy_from_page() is used.

Just two marginal notes:

It looks like you are explaining your mental process and teaching how  
developers should approach these kinds of conversions. Although I'm OK with 
this exposition, a plain assurance that the code being executed between the 
map / unmap did not depend on page-faults and/or preemption being disabled 
would have sufficed. :-)

Ira were suggesting to not use "it appears" and replace with stronger and 
clearer assertions like "it is" or "I can confirm that" (or whatever else like  
these).

As said, no problems at all with regard to the overall goodness of this and 
the other patches.

> 
> Also note that the page being mapped is not allocated by the driver, and so
> the driver doesn't know if the page is in normal memory. This is the reason
> kmap_local_page() is used (via memcpy_from_page()) as opposed to
> page_address().
> 
> I don't have hardware, so this change has only been compile tested.
> 
> Cc: Ayush Sawal <ayush.sawal@...lsio.com>
> Cc: Ira Weiny <ira.weiny@...el.com>
> Cc: Fabio M. De Francesco <fmdefrancesco@...il.com>
> Suggested-by: Fabio M. De Francesco <fmdefrancesco@...il.com>
> Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@...el.com>
> ---
> v1 -> v2:
>  Use memcpy_from_page() as suggested by Fabio
>  Add a "Suggested-by" tag
>  Rework commit message
>  Some emails cc'd in v1 are defunct. Drop them. The MAINTAINERS file for
>  Chelsio drivers likely needs an update
> ---
>  .../chelsio/inline_crypto/ch_ktls/chcr_ktls.c | 26 +++++++++----------
>  1 file changed, 12 insertions(+), 14 deletions(-)

It looks good to me.

Reviewed-by: Fabio M. De Francesco <fmdefrancesco@...il.com>

Thanks,

Fabio

> diff --git a/drivers/net/ethernet/chelsio/inline_crypto/ch_ktls/chcr_ktls.c
> b/drivers/net/ethernet/chelsio/inline_crypto/ch_ktls/chcr_ktls.c index
> da9973b..1a5fdd7 100644
> --- a/drivers/net/ethernet/chelsio/inline_crypto/ch_ktls/chcr_ktls.c
> +++ b/drivers/net/ethernet/chelsio/inline_crypto/ch_ktls/chcr_ktls.c
> @@ -1839,9 +1839,7 @@ static int chcr_short_record_handler(struct
> chcr_ktls_info *tx_info, */
>  		if (prior_data_len) {
>  			int i = 0;
> -			u8 *data = NULL;
>  			skb_frag_t *f;
> -			u8 *vaddr;
>  			int frag_size = 0, frag_delta = 0;
> 
>  			while (remaining > 0) {
> @@ -1853,24 +1851,24 @@ static int chcr_short_record_handler(struct
> chcr_ktls_info *tx_info, i++;
>  			}
>  			f = &record->frags[i];
> -			vaddr = kmap_atomic(skb_frag_page(f));
> -
> -			data = vaddr + skb_frag_off(f)  + remaining;
>  			frag_delta = skb_frag_size(f) - remaining;
> 
>  			if (frag_delta >= prior_data_len) {
> -				memcpy(prior_data, data, 
prior_data_len);
> -				kunmap_atomic(vaddr);
> +				memcpy_from_page(prior_data, 
skb_frag_page(f),
> +						 skb_frag_off(f) + 
remaining,
> +						 prior_data_len);
>  			} else {
> -				memcpy(prior_data, data, frag_delta);
> -				kunmap_atomic(vaddr);
> +				memcpy_from_page(prior_data, 
skb_frag_page(f),
> +						 skb_frag_off(f) + 
remaining,
> +						 frag_delta);
> +
>  				/* get the next page */
>  				f = &record->frags[i + 1];
> -				vaddr = kmap_atomic(skb_frag_page(f));
> -				data = vaddr + skb_frag_off(f);
> -				memcpy(prior_data + frag_delta,
> -				       data, (prior_data_len - 
frag_delta));
> -				kunmap_atomic(vaddr);
> +
> +				memcpy_from_page(prior_data + 
frag_delta,
> +						 skb_frag_page(f),
> +						 skb_frag_off(f),
> +						 prior_data_len - 
frag_delta);
>  			}
>  			/* reset tcp_seq as per the prior_data_required 
len */
>  			tcp_seq -= prior_data_len;
> --
> 2.37.2

Powered by blists - more mailing lists