This is a note to let you know that I've just added the patch titled net:enetc: allocate CBD ring data memory using DMA coherent methods to the 5.17-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary The filename of the patch is: net-enetc-allocate-cbd-ring-data-memory-using-dma-co.patch and it can be found in the queue-5.17 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable@xxxxxxxxxxxxxxx> know about it. commit c1d9717412330d9d150530372d702ba00d41afd1 Author: Po Liu <po.liu@xxxxxxx> Date: Wed Feb 9 20:33:01 2022 +0800 net:enetc: allocate CBD ring data memory using DMA coherent methods [ Upstream commit b3a723dbc94a6e38f67669d03b521edd766ad895 ] To replace the dma_map_single() stream DMA mapping with DMA coherent method dma_alloc_coherent() which is more simple. dma_map_single() found by Tim Gardner not proper. Suggested by Claudiu Manoil and Jakub Kicinski to use dma_alloc_coherent(). Discussion at: https://lore.kernel.org/netdev/AM9PR04MB8397F300DECD3C44D2EBD07796BD9@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/t/ Fixes: 888ae5a3952ba ("net: enetc: add tc flower psfp offload driver") cc: Claudiu Manoil <claudiu.manoil@xxxxxxx> Reported-by: Tim Gardner <tim.gardner@xxxxxxxxxxxxx> Signed-off-by: Po Liu <po.liu@xxxxxxx> Signed-off-by: David S. Miller <davem@xxxxxxxxxxxxx> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx> diff --git a/drivers/net/ethernet/freescale/enetc/enetc_qos.c b/drivers/net/ethernet/freescale/enetc/enetc_qos.c index 3555c12edb45..d3d7172e0fcc 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_qos.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_qos.c @@ -45,6 +45,7 @@ void enetc_sched_speed_set(struct enetc_ndev_priv *priv, int speed) | pspeed); } +#define ENETC_QOS_ALIGN 64 static int enetc_setup_taprio(struct net_device *ndev, struct tc_taprio_qopt_offload *admin_conf) { @@ -52,10 +53,11 @@ static int enetc_setup_taprio(struct net_device *ndev, struct enetc_cbd cbd = {.cmd = 0}; struct tgs_gcl_conf *gcl_config; struct tgs_gcl_data *gcl_data; + dma_addr_t dma, dma_align; struct gce *gce; - dma_addr_t dma; u16 data_size; u16 gcl_len; + void *tmp; u32 tge; int err; int i; @@ -82,9 +84,16 @@ static int enetc_setup_taprio(struct net_device *ndev, gcl_config = &cbd.gcl_conf; data_size = struct_size(gcl_data, entry, gcl_len); - gcl_data = kzalloc(data_size, __GFP_DMA | GFP_KERNEL); - if (!gcl_data) + tmp = dma_alloc_coherent(&priv->si->pdev->dev, + data_size + ENETC_QOS_ALIGN, + &dma, GFP_KERNEL); + if (!tmp) { + dev_err(&priv->si->pdev->dev, + "DMA mapping of taprio gate list failed!\n"); return -ENOMEM; + } + dma_align = ALIGN(dma, ENETC_QOS_ALIGN); + gcl_data = (struct tgs_gcl_data *)PTR_ALIGN(tmp, ENETC_QOS_ALIGN); gce = (struct gce *)(gcl_data + 1); @@ -110,16 +119,8 @@ static int enetc_setup_taprio(struct net_device *ndev, cbd.length = cpu_to_le16(data_size); cbd.status_flags = 0; - dma = dma_map_single(&priv->si->pdev->dev, gcl_data, - data_size, DMA_TO_DEVICE); - if (dma_mapping_error(&priv->si->pdev->dev, dma)) { - netdev_err(priv->si->ndev, "DMA mapping failed!\n"); - kfree(gcl_data); - return -ENOMEM; - } - - cbd.addr[0] = cpu_to_le32(lower_32_bits(dma)); - cbd.addr[1] = cpu_to_le32(upper_32_bits(dma)); + cbd.addr[0] = cpu_to_le32(lower_32_bits(dma_align)); + cbd.addr[1] = cpu_to_le32(upper_32_bits(dma_align)); cbd.cls = BDCR_CMD_PORT_GCL; cbd.status_flags = 0; @@ -132,8 +133,8 @@ static int enetc_setup_taprio(struct net_device *ndev, ENETC_QBV_PTGCR_OFFSET, tge & (~ENETC_QBV_TGE)); - dma_unmap_single(&priv->si->pdev->dev, dma, data_size, DMA_TO_DEVICE); - kfree(gcl_data); + dma_free_coherent(&priv->si->pdev->dev, data_size + ENETC_QOS_ALIGN, + tmp, dma); return err; } @@ -463,8 +464,9 @@ static int enetc_streamid_hw_set(struct enetc_ndev_priv *priv, struct enetc_cbd cbd = {.cmd = 0}; struct streamid_data *si_data; struct streamid_conf *si_conf; + dma_addr_t dma, dma_align; u16 data_size; - dma_addr_t dma; + void *tmp; int port; int err; @@ -485,21 +487,20 @@ static int enetc_streamid_hw_set(struct enetc_ndev_priv *priv, cbd.status_flags = 0; data_size = sizeof(struct streamid_data); - si_data = kzalloc(data_size, __GFP_DMA | GFP_KERNEL); - if (!si_data) + tmp = dma_alloc_coherent(&priv->si->pdev->dev, + data_size + ENETC_QOS_ALIGN, + &dma, GFP_KERNEL); + if (!tmp) { + dev_err(&priv->si->pdev->dev, + "DMA mapping of stream identify failed!\n"); return -ENOMEM; - cbd.length = cpu_to_le16(data_size); - - dma = dma_map_single(&priv->si->pdev->dev, si_data, - data_size, DMA_FROM_DEVICE); - if (dma_mapping_error(&priv->si->pdev->dev, dma)) { - netdev_err(priv->si->ndev, "DMA mapping failed!\n"); - err = -ENOMEM; - goto out; } + dma_align = ALIGN(dma, ENETC_QOS_ALIGN); + si_data = (struct streamid_data *)PTR_ALIGN(tmp, ENETC_QOS_ALIGN); - cbd.addr[0] = cpu_to_le32(lower_32_bits(dma)); - cbd.addr[1] = cpu_to_le32(upper_32_bits(dma)); + cbd.length = cpu_to_le16(data_size); + cbd.addr[0] = cpu_to_le32(lower_32_bits(dma_align)); + cbd.addr[1] = cpu_to_le32(upper_32_bits(dma_align)); eth_broadcast_addr(si_data->dmac); si_data->vid_vidm_tg = (ENETC_CBDR_SID_VID_MASK + ((0x3 << 14) | ENETC_CBDR_SID_VIDM)); @@ -539,8 +540,8 @@ static int enetc_streamid_hw_set(struct enetc_ndev_priv *priv, cbd.length = cpu_to_le16(data_size); - cbd.addr[0] = cpu_to_le32(lower_32_bits(dma)); - cbd.addr[1] = cpu_to_le32(upper_32_bits(dma)); + cbd.addr[0] = cpu_to_le32(lower_32_bits(dma_align)); + cbd.addr[1] = cpu_to_le32(upper_32_bits(dma_align)); /* VIDM default to be 1. * VID Match. If set (b1) then the VID must match, otherwise @@ -561,10 +562,8 @@ static int enetc_streamid_hw_set(struct enetc_ndev_priv *priv, err = enetc_send_cmd(priv->si, &cbd); out: - if (!dma_mapping_error(&priv->si->pdev->dev, dma)) - dma_unmap_single(&priv->si->pdev->dev, dma, data_size, DMA_FROM_DEVICE); - - kfree(si_data); + dma_free_coherent(&priv->si->pdev->dev, data_size + ENETC_QOS_ALIGN, + tmp, dma); return err; } @@ -633,8 +632,9 @@ static int enetc_streamcounter_hw_get(struct enetc_ndev_priv *priv, { struct enetc_cbd cbd = { .cmd = 2 }; struct sfi_counter_data *data_buf; - dma_addr_t dma; + dma_addr_t dma, dma_align; u16 data_size; + void *tmp; int err; cbd.index = cpu_to_le16((u16)index); @@ -643,19 +643,19 @@ static int enetc_streamcounter_hw_get(struct enetc_ndev_priv *priv, cbd.status_flags = 0; data_size = sizeof(struct sfi_counter_data); - data_buf = kzalloc(data_size, __GFP_DMA | GFP_KERNEL); - if (!data_buf) + tmp = dma_alloc_coherent(&priv->si->pdev->dev, + data_size + ENETC_QOS_ALIGN, + &dma, GFP_KERNEL); + if (!tmp) { + dev_err(&priv->si->pdev->dev, + "DMA mapping of stream counter failed!\n"); return -ENOMEM; - - dma = dma_map_single(&priv->si->pdev->dev, data_buf, - data_size, DMA_FROM_DEVICE); - if (dma_mapping_error(&priv->si->pdev->dev, dma)) { - netdev_err(priv->si->ndev, "DMA mapping failed!\n"); - err = -ENOMEM; - goto exit; } - cbd.addr[0] = cpu_to_le32(lower_32_bits(dma)); - cbd.addr[1] = cpu_to_le32(upper_32_bits(dma)); + dma_align = ALIGN(dma, ENETC_QOS_ALIGN); + data_buf = (struct sfi_counter_data *)PTR_ALIGN(tmp, ENETC_QOS_ALIGN); + + cbd.addr[0] = cpu_to_le32(lower_32_bits(dma_align)); + cbd.addr[1] = cpu_to_le32(upper_32_bits(dma_align)); cbd.length = cpu_to_le16(data_size); @@ -684,7 +684,9 @@ static int enetc_streamcounter_hw_get(struct enetc_ndev_priv *priv, data_buf->flow_meter_dropl; exit: - kfree(data_buf); + dma_free_coherent(&priv->si->pdev->dev, data_size + ENETC_QOS_ALIGN, + tmp, dma); + return err; } @@ -723,9 +725,10 @@ static int enetc_streamgate_hw_set(struct enetc_ndev_priv *priv, struct sgcl_conf *sgcl_config; struct sgcl_data *sgcl_data; struct sgce *sgce; - dma_addr_t dma; + dma_addr_t dma, dma_align; u16 data_size; int err, i; + void *tmp; u64 now; cbd.index = cpu_to_le16(sgi->index); @@ -772,24 +775,20 @@ static int enetc_streamgate_hw_set(struct enetc_ndev_priv *priv, sgcl_config->acl_len = (sgi->num_entries - 1) & 0x3; data_size = struct_size(sgcl_data, sgcl, sgi->num_entries); - - sgcl_data = kzalloc(data_size, __GFP_DMA | GFP_KERNEL); - if (!sgcl_data) - return -ENOMEM; - - cbd.length = cpu_to_le16(data_size); - - dma = dma_map_single(&priv->si->pdev->dev, - sgcl_data, data_size, - DMA_FROM_DEVICE); - if (dma_mapping_error(&priv->si->pdev->dev, dma)) { - netdev_err(priv->si->ndev, "DMA mapping failed!\n"); - kfree(sgcl_data); + tmp = dma_alloc_coherent(&priv->si->pdev->dev, + data_size + ENETC_QOS_ALIGN, + &dma, GFP_KERNEL); + if (!tmp) { + dev_err(&priv->si->pdev->dev, + "DMA mapping of stream counter failed!\n"); return -ENOMEM; } + dma_align = ALIGN(dma, ENETC_QOS_ALIGN); + sgcl_data = (struct sgcl_data *)PTR_ALIGN(tmp, ENETC_QOS_ALIGN); - cbd.addr[0] = cpu_to_le32(lower_32_bits(dma)); - cbd.addr[1] = cpu_to_le32(upper_32_bits(dma)); + cbd.length = cpu_to_le16(data_size); + cbd.addr[0] = cpu_to_le32(lower_32_bits(dma_align)); + cbd.addr[1] = cpu_to_le32(upper_32_bits(dma_align)); sgce = &sgcl_data->sgcl[0]; @@ -844,7 +843,8 @@ static int enetc_streamgate_hw_set(struct enetc_ndev_priv *priv, err = enetc_send_cmd(priv->si, &cbd); exit: - kfree(sgcl_data); + dma_free_coherent(&priv->si->pdev->dev, data_size + ENETC_QOS_ALIGN, + tmp, dma); return err; }