[PATCH for-next 0/1] net: jme: migrate to dma_map_phys instead of map_page

[PATCH for-next 0/1] net: jme: migrate to dma_map_phys instead of map_page

[Chu Guangqing]

There is no functional change.

After introduction of dma_map_phys(), there is no need to convert
    from physical address to struct page in order to map page. So let's
    use it directly

Chu Guangqing (1):
  net: jme: migrate to dma_map_phys instead of map_page

 drivers/net/ethernet/jme.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

-- 
2.43.7

[PATCH for-next 1/1] net: jme: migrate to dma_map_phys instead of map_page

[Chu Guangqing]

After introduction of dma_map_phys(), there is no need to convert
from physical address to struct page in order to map page. So let's
use it directly.

Signed-off-by: Chu Guangqing <chuguangqing@inspur.com>
---
 drivers/net/ethernet/jme.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/jme.c b/drivers/net/ethernet/jme.c
index d8be0e4dcb07..7ceeb706939d 100644
--- a/drivers/net/ethernet/jme.c
+++ b/drivers/net/ethernet/jme.c
@@ -735,9 +735,10 @@ jme_make_new_rx_buf(struct jme_adapter *jme, int i)
 	if (unlikely(!skb))
 		return -ENOMEM;
 
-	mapping = dma_map_page(&jme->pdev->dev, virt_to_page(skb->data),
-			       offset_in_page(skb->data), skb_tailroom(skb),
-			       DMA_FROM_DEVICE);
+	mapping = dma_map_phys(&jme->pdev->dev, virt_to_phys(skb->data),
+			       skb_tailroom(skb),
+			       DMA_FROM_DEVICE,
+			       0);
 	if (unlikely(dma_mapping_error(&jme->pdev->dev, mapping))) {
 		dev_kfree_skb(skb);
 		return -ENOMEM;
-- 
2.43.7

Re: [PATCH for-next 1/1] net: jme: migrate to dma_map_phys instead of map_page

[Simon Horman]

On Fri, Oct 24, 2025 at 03:07:34PM +0800, Chu Guangqing wrote:
> After introduction of dma_map_phys(), there is no need to convert
> from physical address to struct page in order to map page. So let's
> use it directly.
> 
> Signed-off-by: Chu Guangqing <chuguangqing@inspur.com>

Although there have been a number minor updates since, mainly for in-kernel
API changes, it seems to me that the last change of substance to this
driver - a feature or bug fix - was the following commit in 2016.

commit 81422e672f81 ("jme: Fix device PM wakeup API usage")

So, unless there is a tree-wide effort to move to the pattern you describe,
I am wondering if it would be better to just leave this code as-is.

Quoting documentation:

1.6.6. Clean-up patches¶

Netdev discourages patches which perform simple clean-ups, which are not in the context of other work. For example:

    Addressing checkpatch.pl, and other trivial coding style warnings

    Addressing Local variable ordering issues

    Conversions to device-managed APIs (devm_ helpers)

This is because it is felt that the churn that such changes produce comes at a greater cost than the value of such clean-ups.

Conversely, spelling and grammar fixes are not discouraged.

https://docs.kernel.org/process/maintainer-netdev.html#clean-up-patches

Re: [PATCH for-next 1/1] net: jme: migrate to dma_map_phys instead of map_page

[Leon Romanovsky]

On Fri, Oct 24, 2025 at 03:07:34PM +0800, Chu Guangqing wrote:
> After introduction of dma_map_phys(), there is no need to convert
> from physical address to struct page in order to map page. So let's
> use it directly.
> 
> Signed-off-by: Chu Guangqing <chuguangqing@inspur.com>
> ---
>  drivers/net/ethernet/jme.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/net/ethernet/jme.c b/drivers/net/ethernet/jme.c
> index d8be0e4dcb07..7ceeb706939d 100644
> --- a/drivers/net/ethernet/jme.c
> +++ b/drivers/net/ethernet/jme.c
> @@ -735,9 +735,10 @@ jme_make_new_rx_buf(struct jme_adapter *jme, int i)
>  	if (unlikely(!skb))
>  		return -ENOMEM;
>  
> -	mapping = dma_map_page(&jme->pdev->dev, virt_to_page(skb->data),
> -			       offset_in_page(skb->data), skb_tailroom(skb),
> -			       DMA_FROM_DEVICE);
> +	mapping = dma_map_phys(&jme->pdev->dev, virt_to_phys(skb->data),
> +			       skb_tailroom(skb),
> +			       DMA_FROM_DEVICE,
> +			       0);

Same comment as before, dma_map_phys() should be paired with dma_unmap_phys().

Thanks

>  	if (unlikely(dma_mapping_error(&jme->pdev->dev, mapping))) {
>  		dev_kfree_skb(skb);
>  		return -ENOMEM;
> -- 
> 2.43.7
> 
>