Skip to contentSkip to conversation

net: skb: Remove skb_data_area_size()

PATCH Applied Ricardo Martinez 4 participants 6 messages May 13, 2022 – May 16, 2022

Subsystems (4)
lore
Share & tools
lore

Ricardo Martinez

author reviewer
#0 lore raw

cover net: skb: Remove skb_data_area_size()

This patch series removes the skb_data_area_size() helper, replacing it in t7xx driver with the size used during skb allocation.

https://lore.kernel.org/netdev/CAHNKnsTmH-rGgWi3jtyC=ktM1DW2W1VJkYoTMJV2Z_Bt498bsg@mail.gmail.com/

Ricardo Martinez (2):
net: wwan: t7xx: Avoid calls to skb_data_area_size()
net: skb: Remove skb_data_area_size()

 drivers/net/wwan/t7xx/t7xx_hif_cldma.c     | 7 +++----
 drivers/net/wwan/t7xx/t7xx_hif_dpmaif_rx.c | 6 ++----
 include/linux/skbuff.h                     | 5 -----
 3 files changed, 5 insertions(+), 13 deletions(-)
Signature · 1 line
2.25.1
author reviewer
#1 lore raw

PATCH 1/2 net: wwan: t7xx: Avoid calls to skb_data_area_size()

skb_data_area_size() helper was used to calculate the size of the DMA mapped buffer passed to the HW. Instead of doing this, use the size passed to allocate the skbs.

---
 drivers/net/wwan/t7xx/t7xx_hif_cldma.c     | 7 +++----
 drivers/net/wwan/t7xx/t7xx_hif_dpmaif_rx.c | 6 ++----
 2 files changed, 5 insertions(+), 8 deletions(-)
drivers/net/wwan/t7xx/t7xx_hif_cldma.c +3 -4
index 46066dcd2607..0c52801ed0de 100644
@@ -97,8 +97,7 @@ static int t7xx_cldma_alloc_and_map_skb(struct cldma_ctrl *md_ctrl, struct cldma
 	if (!req->skb)
 		return -ENOMEM;
 
-	req->mapped_buff = dma_map_single(md_ctrl->dev, req->skb->data,
-					  skb_data_area_size(req->skb), DMA_FROM_DEVICE);
+	req->mapped_buff = dma_map_single(md_ctrl->dev, req->skb->data, size, DMA_FROM_DEVICE);
 	if (dma_mapping_error(md_ctrl->dev, req->mapped_buff)) {
 		dev_kfree_skb_any(req->skb);
 		req->skb = NULL;
@@ -154,7 +153,7 @@ static int t7xx_cldma_gpd_rx_from_q(struct cldma_queue *queue, int budget, bool
 
 		if (req->mapped_buff) {
 			dma_unmap_single(md_ctrl->dev, req->mapped_buff,
-					 skb_data_area_size(skb), DMA_FROM_DEVICE);
+					 queue->tr_ring->pkt_size, DMA_FROM_DEVICE);
 			req->mapped_buff = 0;
 		}
 
@@ -376,7 +375,7 @@ static void t7xx_cldma_ring_free(struct cldma_ctrl *md_ctrl,
 	list_for_each_entry_safe(req_cur, req_next, &ring->gpd_ring, entry) {
 		if (req_cur->mapped_buff && req_cur->skb) {
 			dma_unmap_single(md_ctrl->dev, req_cur->mapped_buff,
-					 skb_data_area_size(req_cur->skb), tx_rx);
+					 ring->pkt_size, tx_rx);
 			req_cur->mapped_buff = 0;
 		}
 
drivers/net/wwan/t7xx/t7xx_hif_dpmaif_rx.c +2 -4
index 35a8a0d7c1ee..91a0eb19e0d8 100644
@@ -151,14 +151,12 @@ static bool t7xx_alloc_and_map_skb_info(const struct dpmaif_ctrl *dpmaif_ctrl,
 {
 	dma_addr_t data_bus_addr;
 	struct sk_buff *skb;
-	size_t data_len;
 
 	skb = __dev_alloc_skb(size, GFP_KERNEL);
 	if (!skb)
 		return false;
 
-	data_len = skb_data_area_size(skb);
-	data_bus_addr = dma_map_single(dpmaif_ctrl->dev, skb->data, data_len, DMA_FROM_DEVICE);
+	data_bus_addr = dma_map_single(dpmaif_ctrl->dev, skb->data, size, DMA_FROM_DEVICE);
 	if (dma_mapping_error(dpmaif_ctrl->dev, data_bus_addr)) {
 		dev_err_ratelimited(dpmaif_ctrl->dev, "DMA mapping error\n");
 		dev_kfree_skb_any(skb);
@@ -167,7 +165,7 @@ static bool t7xx_alloc_and_map_skb_info(const struct dpmaif_ctrl *dpmaif_ctrl,
 
 	cur_skb->skb = skb;
 	cur_skb->data_bus_addr = data_bus_addr;
-	cur_skb->data_len = data_len;
+	cur_skb->data_len = size;
 
 	return true;
 }
Signature · 1 line
2.25.1
author reviewer
#2 lore raw

PATCH 2/2 net: skb: Remove skb_data_area_size()

In reply to Ricardo Martinez

skb_data_area_size() is not needed. As Jakub pointed out [1]:
For Rx, drivers can use the size passed during skb allocation or use skb_tailroom().
For Tx, drivers should use skb_headlen().

[1] https://lore.kernel.org/netdev/CAHNKnsTmH-rGgWi3jtyC=ktM1DW2W1VJkYoTMJV2Z_Bt498bsg@mail.gmail.com/

---
 include/linux/skbuff.h | 5 -----
 1 file changed, 5 deletions(-)
include/linux/skbuff.h +0 -5
index 9d82a8b6c8f1..2810a3abe81a 100644
@@ -1764,11 +1764,6 @@ static inline void skb_set_end_offset(struct sk_buff *skb, unsigned int offset)
 }
 #endif
 
-static inline unsigned int skb_data_area_size(struct sk_buff *skb)
-{
-	return skb_end_pointer(skb) - skb->data;
-}
-
 struct ubuf_info *msg_zerocopy_realloc(struct sock *sk, size_t size,
 				       struct ubuf_info *uarg);
 
Signature · 1 line
2.25.1
maintainer Reviewed-by on the v1 series
#4 lore raw
In reply to Ricardo Martinez

Hello Ricardo,

Quoted 5 lines · Ricardo Martinez

This patch series removes the skb_data_area_size() helper, replacing it in t7xx driver with the size used during skb allocation.

https://lore.kernel.org/netdev/CAHNKnsTmH-rGgWi3jtyC=ktM1DW2W1VJkYoTMJV2Z_Bt498bsg@mail.gmail.com/

Quoted 8 lines

Ricardo Martinez (2):
net: wwan: t7xx: Avoid calls to skb_data_area_size()
net: skb: Remove skb_data_area_size()

 drivers/net/wwan/t7xx/t7xx_hif_cldma.c     | 7 +++----
 drivers/net/wwan/t7xx/t7xx_hif_dpmaif_rx.c | 6 ++----
 include/linux/skbuff.h                     | 5 -----
 3 files changed, 5 insertions(+), 13 deletions(-)

Thank you for taking care of this!

#5 lore raw
In reply to Ricardo Martinez
This series was applied to netdev/net-next.git (master) by Jakub Kicinski <[email protected]>:

Hello:

This series was applied to netdev/net-next.git (master) by Jakub Kicinski <[email protected]>:

Quoted 10 lines · On Fri, 13 May 2022 10:33:58 -0700 you wrote:

This patch series removes the skb_data_area_size() helper, replacing it in t7xx driver with the size used during skb allocation.

https://lore.kernel.org/netdev/CAHNKnsTmH-rGgWi3jtyC=ktM1DW2W1VJkYoTMJV2Z_Bt498bsg@mail.gmail.com/

Ricardo Martinez (2):
net: wwan: t7xx: Avoid calls to skb_data_area_size()
net: skb: Remove skb_data_area_size()

[...]

Here is the summary with links:
- [net-next,1/2] net: wwan: t7xx: Avoid calls to skb_data_area_size()

    https://git.kernel.org/netdev/net-next/c/262d98b1193f
  - [net-next,2/2] net: skb: Remove skb_data_area_size()
    https://git.kernel.org/netdev/net-next/c/89af2ce2d95c

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html