Skip to contentSkip to conversation

net: fec: remove unnecessary DMA mapping of TSO header

PATCH In review Dheeraj Reddy Jonnalagadda 3 participants 5 messages Jan 22, 2025 – Jan 24, 2025

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net: fec: remove unnecessary DMA mapping of TSO header

The TSO header buffer is pre-allocated DMA memory, so there's no need to map it again with dma_map_single() in fec_enet_txq_put_hdr_tso(). Remove this redundant mapping operation.

Fixes 79f339125ea3 ("net: fec: Add software TSO support")
---
 drivers/net/ethernet/freescale/fec_main.c | 9 ---------
 1 file changed, 9 deletions(-)
drivers/net/ethernet/freescale/fec_main.c +0 -9
index 68725506a095..039de4c5044e 100644
@@ -805,15 +805,6 @@ fec_enet_txq_put_hdr_tso(struct fec_enet_priv_tx_q *txq,
 
 		if (fep->quirks & FEC_QUIRK_SWAP_FRAME)
 			swap_buffer(bufaddr, hdr_len);
-
-		dmabuf = dma_map_single(&fep->pdev->dev, bufaddr,
-					hdr_len, DMA_TO_DEVICE);
-		if (dma_mapping_error(&fep->pdev->dev, dmabuf)) {
-			dev_kfree_skb_any(skb);
-			if (net_ratelimit())
-				netdev_err(ndev, "Tx DMA memory map failed\n");
-			return NETDEV_TX_OK;
-		}
 	}
 
 	bdp->cbd_bufaddr = cpu_to_fec32(dmabuf);
Signature · 1 line
2.34.1
reviewer
#1 lore raw
Quoted 20 lines · Dheeraj Reddy Jonnalagadda

The TSO header buffer is pre-allocated DMA memory, so there's no need to map it again with dma_map_single() in fec_enet_txq_put_hdr_tso(). Remove this redundant mapping operation.

Fixes 79f339125ea3 ("net: fec: Add software TSO support")
---
 drivers/net/ethernet/freescale/fec_main.c | 9 ---------
 1 file changed, 9 deletions(-)
drivers/net/ethernet/freescale/fec_main.c +0 -2
index 68725506a095..039de4c5044e 100644
@@ -805,15 +805,6 @@ fec_enet_txq_put_hdr_tso(struct fec_enet_priv_tx_q *txq,
 
 		if (fep->quirks & FEC_QUIRK_SWAP_FRAME)
 			swap_buffer(bufaddr, hdr_len);
-
-		dmabuf = dma_map_single(&fep->pdev->dev, bufaddr,

drivers/net/ethernet/freescale/fec_main.c fec_enet_txq_put_hdr_tso(struct fec_enet_priv_tx_q *txq, · line 810

Quoted 10 lines
drivers/net/ethernet/freescale/fec_main.c +0 -7
-					hdr_len, DMA_TO_DEVICE);
-		if (dma_mapping_error(&fep->pdev->dev, dmabuf)) {
-			dev_kfree_skb_any(skb);
-			if (net_ratelimit())
-				netdev_err(ndev, "Tx DMA memory map failed\n");
-			return NETDEV_TX_OK;
-		}
 	}
 
 	bdp->cbd_bufaddr = cpu_to_fec32(dmabuf);

Hi Dheeraj,

It seems to me that at this point of execution, without your patch, dmabuf can be one of two values:

1. txq->tso_hdrs_dma + index * TSO_HEADER_SIZE This is the default case.

2. txq->tx_bounce[index]
This is the special case, which is handed in the conditional block that your patch updates.

In case 1, I agree that tso_hdrs_dma is already mapped. But this case is not affected by your patch as the conditional block that includes dma_map_single() is not executed.

In case 2, I do not see where txq->tx_bounce[index] is already mapped.

Am I missing something?

maintainer
#2 lore raw
In reply to Dheeraj Reddy Jonnalagadda
Quoted 31 lines

The TSO header buffer is pre-allocated DMA memory, so there's no need to map it again with dma_map_single() in fec_enet_txq_put_hdr_tso(). Remove this redundant mapping operation.

Fixes 79f339125ea3 ("net: fec: Add software TSO support")
---
 drivers/net/ethernet/freescale/fec_main.c | 9 ---------
 1 file changed, 9 deletions(-)
drivers/net/ethernet/freescale/fec_main.c +0 -9
index 68725506a095..039de4c5044e 100644
@@ -805,15 +805,6 @@ fec_enet_txq_put_hdr_tso(struct fec_enet_priv_tx_q
 *txq,
 
 		if (fep->quirks & FEC_QUIRK_SWAP_FRAME)
 			swap_buffer(bufaddr, hdr_len);
-
-		dmabuf = dma_map_single(&fep->pdev->dev, bufaddr,
-					hdr_len, DMA_TO_DEVICE);
-		if (dma_mapping_error(&fep->pdev->dev, dmabuf)) {
-			dev_kfree_skb_any(skb);
-			if (net_ratelimit())
-				netdev_err(ndev, "Tx DMA memory map failed\n");
-			return NETDEV_TX_OK;
-		}
 	}
 
Quoted 3 lines

bdp->cbd_bufaddr = cpu_to_fec32(dmabuf);
--
2.34.1

Hi Dheeraj,

I must admit that I misread it too. There is another case in the TSO header where txq->tx_bounce may be used in some cases. I think the most correct fix is to make txq->tso_hdrs aligned to 32/64 bytes when allocating tso_hdrs, then we do not need to use txq->tx_bounce in fec_enet_txq_put_hdr_tso(), because (bufaddr) & fep->tx_align) will not be true. This way we can safely remove dma_map_single() from fec_enet_txq_put_hdr_tso().

#3 lore raw
Quoted 10 lines · Wei Fang

Hi Dheeraj,

I must admit that I misread it too. There is another case in the TSO header where txq->tx_bounce may be used in some cases. I think the most correct fix is to make txq->tso_hdrs aligned to 32/64 bytes when allocating tso_hdrs, then we do not need to use txq->tx_bounce in fec_enet_txq_put_hdr_tso(), because (bufaddr) & fep->tx_align) will not be true. This way we can safely remove dma_map_single() from fec_enet_txq_put_hdr_tso().

Hi Fang, Simon,

Thank you for the feedback. I have a clarification question regarding the alignment of txq->tso_hdrs.

In the current code, txq->tso_hdrs is allocated using fec_dma_alloc(), which internally calls dma_alloc_coherent(). As I understand it, dma_alloc_coherent() guarantees that the allocated buffer is properly aligned.

Given this, should we remove the alignment check
((unsigned long)bufaddr) & fep->tx_align and the associated dma_map_single() logic entirely from fec_enet_txq_put_hdr_tso() as you have suggested?

-Dheeraj

maintainer
#4 lore raw
Quoted 15 lines
Quoted 9 lines

Hi Dheeraj,

I must admit that I misread it too. There is another case in the TSO header where txq->tx_bounce may be used in some cases. I think the most correct fix is to make txq->tso_hdrs aligned to 32/64 bytes when allocating tso_hdrs, then we do not need to use txq->tx_bounce in fec_enet_txq_put_hdr_tso(), because (bufaddr) & fep->tx_align) will not be true. This way we can safely remove dma_map_single() from fec_enet_txq_put_hdr_tso().

Hi Fang, Simon,

Thank you for the feedback. I have a clarification question regarding the alignment of txq->tso_hdrs.

Quoted 8 lines

In the current code, txq->tso_hdrs is allocated using fec_dma_alloc(), which internally calls dma_alloc_coherent(). As I understand it, dma_alloc_coherent() guarantees that the allocated buffer is properly aligned.

Given this, should we remove the alignment check
((unsigned long)bufaddr) & fep->tx_align and the associated dma_map_single() logic entirely from fec_enet_txq_put_hdr_tso() as you have suggested?

The maximum of fep->tx_align is 0xf, so we need to ensure the buffer allocated by dma_alloc_coherent() is at least 16-byte aligned, but I'm not sure whether dma_alloc_coherent() guarantees this alignment. A safe approach is to apply for a memory block longer than expected, for example, if we want the alignment is 32-byte, we can apply for a memory block that is 32 bytes longer than the expected length. Then use ALGIN and PTR_ALIGN to get the aligned addr.

txq->tso_hdrs_base = fec_dma_alloc(&fep->pdev->dev,

					     	  txq->bd.ring_size * TSO_HEADER_SIZE + 32,
					     	  &txq->tso_hdrs_dma_base, GFP_KERNEL);

txq->tso_hdrs = PTR_ALIGN(txq->tso_hdrs_base, 32); txq->tso_hdrs_dma = ALIGN(txq->tso_hdrs_dma_base, 32);

This is just my personal experience, if you have a better way that would be great.