From patchwork Thu Nov 21 06:18:02 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Frank Sae X-Patchwork-Id: 13881684 X-Patchwork-Delegate: kuba@kernel.org Received: from out198-8.us.a.mail.aliyun.com (out198-8.us.a.mail.aliyun.com [47.90.198.8]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id D91F2230986; Thu, 21 Nov 2024 06:18:20 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=47.90.198.8 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1732169907; cv=none; b=lIBS3xujg6GW41guAc7VNftmG8kwjl7VY+2yiZ93c9V8Pz+EpRGXt5LNgVCD68oqjtHPuQ6yXBwFSE5YiP48MlNU6Cqfpl8XGox0/Dl1d6ykq2awJAPaKtEms9VxeSXlPPd8XFDF9X6Xls3JrYRYJOt5H/6lH5nqEwkHLu+OpLs= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1732169907; c=relaxed/simple; bh=CXCg3h4nW21IyFct/IBlnrAs8YRvDNnbS6yWqqJ/5AE=; h=From:To:Cc:Subject:Date:Message-Id:In-Reply-To:References: MIME-Version:Content-Type; b=S+sxXjParyrrrarSliWiJB2eiqosDQRkSAXV6G80QRFt/gckEwo7f0m1/JukevZIyBP7YzJv2cykfwxO45M7OpR+hl6fm1pM5B4cfB/bHBCSecK7bvozkWcg3e3hsxwVujtz3BXrWoQidiTVMJ/xs5yKSIMCrlDxga4kvoOpFHQ= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=motor-comm.com; spf=pass smtp.mailfrom=motor-comm.com; arc=none smtp.client-ip=47.90.198.8 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=motor-comm.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=motor-comm.com Received: from sun-VirtualBox..(mailfrom:Frank.Sae@motor-comm.com fp:SMTPD_---.aGmppae_1732100203 cluster:ay29) by smtp.aliyun-inc.com; Wed, 20 Nov 2024 18:56:44 +0800 From: Frank Sae To: davem@davemloft.net, edumazet@google.com, kuba@kernel.org, pabeni@redhat.com Cc: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, xiaogang.fan@motor-comm.com, fei.zhang@motor-comm.com, hua.sun@motor-comm.com, Frank.Sae@motor-comm.com Subject: [PATCH net-next v2 10/21] motorcomm:yt6801: Implement .ndo_start_xmit function Date: Thu, 21 Nov 2024 14:18:02 +0800 Message-Id: <20241120105625.22508-11-Frank.Sae@motor-comm.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20241120105625.22508-1-Frank.Sae@motor-comm.com> References: <20241120105625.22508-1-Frank.Sae@motor-comm.com> Precedence: bulk X-Mailing-List: netdev@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Patchwork-Delegate: kuba@kernel.org Implement .ndo_start_xmit function to prepare preliminary packet info for TX， Prepare tso and vlan， then map tx skb, at last it call dev_xmit function to send data. Signed-off-by: Frank Sae --- .../ethernet/motorcomm/yt6801/yt6801_desc.c | 133 ++++++ .../net/ethernet/motorcomm/yt6801/yt6801_hw.c | 404 ++++++++++++++++++ .../ethernet/motorcomm/yt6801/yt6801_net.c | 269 ++++++++++++ 3 files changed, 806 insertions(+) diff --git a/drivers/net/ethernet/motorcomm/yt6801/yt6801_desc.c b/drivers/net/ethernet/motorcomm/yt6801/yt6801_desc.c index 4010e9412..e5e95ee5c 100644 --- a/drivers/net/ethernet/motorcomm/yt6801/yt6801_desc.c +++ b/drivers/net/ethernet/motorcomm/yt6801/yt6801_desc.c @@ -455,6 +455,139 @@ void fxgmac_desc_rx_reset(struct fxgmac_desc_data *desc_data) dma_wmb(); } +int fxgmac_tx_skb_map(struct fxgmac_channel *channel, struct sk_buff *skb) +{ + struct fxgmac_pdata *pdata = channel->pdata; + struct fxgmac_ring *ring = channel->tx_ring; + unsigned int start_index, cur_index; + struct fxgmac_desc_data *desc_data; + unsigned int offset, datalen, len; + struct fxgmac_pkt_info *pkt_info; + unsigned int tso, vlan; + dma_addr_t skb_dma; + skb_frag_t *frag; + + offset = 0; + start_index = ring->cur; + cur_index = ring->cur; + pkt_info = &ring->pkt_info; + pkt_info->desc_count = 0; + pkt_info->length = 0; + + tso = FXGMAC_GET_BITS(pkt_info->attributes, TX_PKT_ATTR_TSO_ENABLE_POS, + TX_PKT_ATTR_TSO_ENABLE_LEN); + vlan = FXGMAC_GET_BITS(pkt_info->attributes, TX_PKT_ATTR_VLAN_CTAG_POS, + TX_PKT_ATTR_VLAN_CTAG_LEN); + + /* Save space for a context descriptor if needed */ + if ((tso && pkt_info->mss != ring->tx.cur_mss) || + (vlan && pkt_info->vlan_ctag != ring->tx.cur_vlan_ctag)) + cur_index = FXGMAC_GET_ENTRY(cur_index, ring->dma_desc_count); + + desc_data = FXGMAC_GET_DESC_DATA(ring, cur_index); + + if (tso) { + /* Map the TSO header */ + skb_dma = dma_map_single(pdata->dev, skb->data, + pkt_info->header_len, DMA_TO_DEVICE); + if (dma_mapping_error(pdata->dev, skb_dma)) { + yt_err(pdata, "dma_map_single err\n"); + goto err_out; + } + desc_data->skb_dma = skb_dma; + desc_data->skb_dma_len = pkt_info->header_len; + yt_dbg(pdata, "skb header: index=%u, dma=%pad, len=%u\n", + cur_index, &skb_dma, pkt_info->header_len); + + offset = pkt_info->header_len; + pkt_info->length += pkt_info->header_len; + + cur_index = FXGMAC_GET_ENTRY(cur_index, ring->dma_desc_count); + desc_data = FXGMAC_GET_DESC_DATA(ring, cur_index); + } + + /* Map the (remainder of the) packet */ + for (datalen = skb_headlen(skb) - offset; datalen;) { + len = min_t(unsigned int, datalen, FXGMAC_TX_MAX_BUF_SIZE); + skb_dma = dma_map_single(pdata->dev, skb->data + offset, len, + DMA_TO_DEVICE); + if (dma_mapping_error(pdata->dev, skb_dma)) { + yt_err(pdata, "dma_map_single err\n"); + goto err_out; + } + desc_data->skb_dma = skb_dma; + desc_data->skb_dma_len = len; + yt_dbg(pdata, "skb data: index=%u, dma=%pad, len=%u\n", + cur_index, &skb_dma, len); + + datalen -= len; + offset += len; + pkt_info->length += len; + + cur_index = FXGMAC_GET_ENTRY(cur_index, ring->dma_desc_count); + desc_data = FXGMAC_GET_DESC_DATA(ring, cur_index); + } + + for (u32 i = 0; i < skb_shinfo(skb)->nr_frags; i++) { + yt_dbg(pdata, "mapping frag %u\n", i); + + frag = &skb_shinfo(skb)->frags[i]; + offset = 0; + + for (datalen = skb_frag_size(frag); datalen;) { + len = min_t(unsigned int, datalen, + FXGMAC_TX_MAX_BUF_SIZE); + skb_dma = skb_frag_dma_map(pdata->dev, frag, offset, + len, DMA_TO_DEVICE); + if (dma_mapping_error(pdata->dev, skb_dma)) { + yt_err(pdata, "skb_frag_dma_map err\n"); + goto err_out; + } + desc_data->skb_dma = skb_dma; + desc_data->skb_dma_len = len; + desc_data->mapped_as_page = 1; + + yt_dbg(pdata, "skb frag: index=%u, dma=%pad, len=%u\n", + cur_index, &skb_dma, len); + + datalen -= len; + offset += len; + pkt_info->length += len; + + cur_index = FXGMAC_GET_ENTRY(cur_index, + ring->dma_desc_count); + desc_data = FXGMAC_GET_DESC_DATA(ring, cur_index); + } + } + + /* Save the skb address in the last entry. We always have some data + * that has been mapped so desc_data is always advanced past the last + * piece of mapped data - use the entry pointed to by cur_index - 1. + */ + desc_data = FXGMAC_GET_DESC_DATA(ring, (cur_index - 1) & + (ring->dma_desc_count - 1)); + desc_data->skb = skb; + + /* Save the number of descriptor entries used */ + if (start_index <= cur_index) + pkt_info->desc_count = cur_index - start_index; + else + pkt_info->desc_count = + ring->dma_desc_count - start_index + cur_index; + + return pkt_info->desc_count; + +err_out: + while (start_index < cur_index) { + desc_data = FXGMAC_GET_DESC_DATA(ring, start_index); + start_index = + FXGMAC_GET_ENTRY(start_index, ring->dma_desc_count); + fxgmac_desc_data_unmap(pdata, desc_data); + } + + return 0; +} + void fxgmac_dump_rx_desc(struct fxgmac_pdata *pdata, struct fxgmac_ring *ring, unsigned int idx) { diff --git a/drivers/net/ethernet/motorcomm/yt6801/yt6801_hw.c b/drivers/net/ethernet/motorcomm/yt6801/yt6801_hw.c index 1af26b0b4..791dd69b7 100644 --- a/drivers/net/ethernet/motorcomm/yt6801/yt6801_hw.c +++ b/drivers/net/ethernet/motorcomm/yt6801/yt6801_hw.c @@ -2156,6 +2156,408 @@ static void fxgmac_clear_misc_int_status(struct fxgmac_pdata *pdata) wr32_mac(pdata, val, DMA_ECC_INT_SR); } +static void fxgmac_dev_xmit(struct fxgmac_channel *channel) +{ + struct fxgmac_pdata *pdata = channel->pdata; + struct fxgmac_ring *ring = channel->tx_ring; + unsigned int tso_context, vlan_context; + unsigned int csum, tso, vlan, attr; + struct fxgmac_desc_data *desc_data; + struct fxgmac_dma_desc *dma_desc; + struct fxgmac_pkt_info *pkt_info; + int start_index = ring->cur; + int cur_index = ring->cur; + int i; + + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, "%s, desc cur=%d\n", __func__, cur_index); + + pkt_info = &ring->pkt_info; + attr = pkt_info->attributes; + csum = FXGMAC_GET_BITS(attr, TX_PKT_ATTR_CSUM_ENABLE_POS, + TX_PKT_ATTR_CSUM_ENABLE_LEN); + tso = FXGMAC_GET_BITS(attr, TX_PKT_ATTR_TSO_ENABLE_POS, + TX_PKT_ATTR_TSO_ENABLE_LEN); + vlan = FXGMAC_GET_BITS(attr, TX_PKT_ATTR_VLAN_CTAG_POS, + TX_PKT_ATTR_VLAN_CTAG_LEN); + + if (tso && pkt_info->mss != ring->tx.cur_mss) + tso_context = 1; + else + tso_context = 0; + + if ((tso_context) && (netif_msg_tx_done(pdata))) { + yt_dbg(pdata, "%s, tso_%s tso=0x%x,pkt_mss=%d,cur_mss=%d\n", + __func__, (pkt_info->mss) ? "start" : "stop", tso, + pkt_info->mss, ring->tx.cur_mss); + } + + if (vlan && pkt_info->vlan_ctag != ring->tx.cur_vlan_ctag) + vlan_context = 1; + else + vlan_context = 0; + + if (vlan && (netif_msg_tx_done(pdata))) + yt_dbg(pdata, + "%s, pkt vlan=%d, ring vlan=%d, vlan_context=%d\n", + __func__, pkt_info->vlan_ctag, ring->tx.cur_vlan_ctag, + vlan_context); + + desc_data = FXGMAC_GET_DESC_DATA(ring, cur_index); + dma_desc = desc_data->dma_desc; + + /* Create a context descriptor if this is a TSO pkt_info */ + if (tso_context) { + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, "tso context descriptor,mss=%u\n", + pkt_info->mss); + + /* Set the MSS size */ + fxgmac_set_bits_le(&dma_desc->desc2, TX_CONTEXT_DESC2_MSS_POS, + TX_CONTEXT_DESC2_MSS_LEN, pkt_info->mss); + + /* Mark it as a CONTEXT descriptor */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_CONTEXT_DESC3_CTXT_POS, + TX_CONTEXT_DESC3_CTXT_LEN, 1); + + /* Indicate this descriptor contains the MSS */ + fxgmac_set_bits_le(&dma_desc->desc3, + TX_CONTEXT_DESC3_TCMSSV_POS, + TX_CONTEXT_DESC3_TCMSSV_LEN, 1); + + ring->tx.cur_mss = pkt_info->mss; + } + + if (vlan_context) { + yt_dbg(pdata, "VLAN context descriptor, ctag=%u\n", + pkt_info->vlan_ctag); + + /* Mark it as a CONTEXT descriptor */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_CONTEXT_DESC3_CTXT_POS, + TX_CONTEXT_DESC3_CTXT_LEN, 1); + + /* Set the VLAN tag */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_CONTEXT_DESC3_VT_POS, + TX_CONTEXT_DESC3_VT_LEN, + pkt_info->vlan_ctag); + + /* Indicate this descriptor contains the VLAN tag */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_CONTEXT_DESC3_VLTV_POS, + TX_CONTEXT_DESC3_VLTV_LEN, 1); + + ring->tx.cur_vlan_ctag = pkt_info->vlan_ctag; + } + if (tso_context || vlan_context) { + cur_index = FXGMAC_GET_ENTRY(cur_index, ring->dma_desc_count); + desc_data = FXGMAC_GET_DESC_DATA(ring, cur_index); + dma_desc = desc_data->dma_desc; + } + + /* Update buffer address (for TSO this is the header) */ + dma_desc->desc0 = cpu_to_le32(lower_32_bits(desc_data->skb_dma)); + dma_desc->desc1 = cpu_to_le32(upper_32_bits(desc_data->skb_dma)); + + /* Update the buffer length */ + fxgmac_set_bits_le(&dma_desc->desc2, TX_NORMAL_DESC2_HL_B1L_POS, + TX_NORMAL_DESC2_HL_B1L_LEN, desc_data->skb_dma_len); + + /* VLAN tag insertion check */ + if (vlan) { + fxgmac_set_bits_le(&dma_desc->desc2, TX_NORMAL_DESC2_VTIR_POS, + TX_NORMAL_DESC2_VTIR_LEN, + TX_NORMAL_DESC2_VLAN_INSERT); + pdata->stats.tx_vlan_packets++; + } + + /* Timestamp enablement check */ + if (FXGMAC_GET_BITS(attr, TX_PKT_ATTR_PTP_POS, TX_PKT_ATTR_PTP_LEN)) + fxgmac_set_bits_le(&dma_desc->desc2, TX_NORMAL_DESC2_TTSE_POS, + TX_NORMAL_DESC2_TTSE_LEN, 1); + + /* Mark it as First Descriptor */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_FD_POS, + TX_NORMAL_DESC3_FD_LEN, 1); + + /* Mark it as a NORMAL descriptor */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_CTXT_POS, + TX_NORMAL_DESC3_CTXT_LEN, 0); + + /* Set OWN bit if not the first descriptor */ + if (cur_index != start_index) + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_OWN_POS, + TX_NORMAL_DESC3_OWN_LEN, 1); + + if (tso) { + /* Enable TSO */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_TSE_POS, + TX_NORMAL_DESC3_TSE_LEN, 1); + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_TCPPL_POS, + TX_NORMAL_DESC3_TCPPL_LEN, + pkt_info->tcp_payload_len); + fxgmac_set_bits_le(&dma_desc->desc3, + TX_NORMAL_DESC3_TCPHDRLEN_POS, + TX_NORMAL_DESC3_TCPHDRLEN_LEN, + pkt_info->tcp_header_len / 4); + + pdata->stats.tx_tso_packets++; + } else { + /* Enable CRC and Pad Insertion */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_CPC_POS, + TX_NORMAL_DESC3_CPC_LEN, 0); + + /* Enable HW CSUM */ + if (csum) + fxgmac_set_bits_le(&dma_desc->desc3, + TX_NORMAL_DESC3_CIC_POS, + TX_NORMAL_DESC3_CIC_LEN, 0x3); + + /* Set the total length to be transmitted */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_FL_POS, + TX_NORMAL_DESC3_FL_LEN, pkt_info->length); + } + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, + "%s, before more descs, desc cur=%d, start=%d, desc=%#x,%#x,%#x,%#x\n", + __func__, cur_index, start_index, dma_desc->desc0, + dma_desc->desc1, dma_desc->desc2, dma_desc->desc3); + + if (start_index <= cur_index) + i = cur_index - start_index + 1; + else + i = ring->dma_desc_count - start_index + cur_index; + + for (; i < pkt_info->desc_count; i++) { + cur_index = FXGMAC_GET_ENTRY(cur_index, ring->dma_desc_count); + desc_data = FXGMAC_GET_DESC_DATA(ring, cur_index); + dma_desc = desc_data->dma_desc; + + /* Update buffer address */ + dma_desc->desc0 = + cpu_to_le32(lower_32_bits(desc_data->skb_dma)); + dma_desc->desc1 = + cpu_to_le32(upper_32_bits(desc_data->skb_dma)); + + /* Update the buffer length */ + fxgmac_set_bits_le(&dma_desc->desc2, TX_NORMAL_DESC2_HL_B1L_POS, + TX_NORMAL_DESC2_HL_B1L_LEN, + desc_data->skb_dma_len); + + /* Set OWN bit */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_OWN_POS, + TX_NORMAL_DESC3_OWN_LEN, 1); + + /* Mark it as NORMAL descriptor */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_CTXT_POS, + TX_NORMAL_DESC3_CTXT_LEN, 0); + + /* Enable HW CSUM */ + if (csum) + fxgmac_set_bits_le(&dma_desc->desc3, + TX_NORMAL_DESC3_CIC_POS, + TX_NORMAL_DESC3_CIC_LEN, 0x3); + } + + /* Set LAST bit for the last descriptor */ + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_LD_POS, + TX_NORMAL_DESC3_LD_LEN, 1); + + fxgmac_set_bits_le(&dma_desc->desc2, TX_NORMAL_DESC2_IC_POS, + TX_NORMAL_DESC2_IC_LEN, 1); + + /* Save the Tx info to report back during cleanup */ + desc_data->tx.packets = pkt_info->tx_packets; + desc_data->tx.bytes = pkt_info->tx_bytes; + + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, + "%s, last descs, desc cur=%d, desc=%#x,%#x,%#x,%#x\n", + __func__, cur_index, dma_desc->desc0, dma_desc->desc1, + dma_desc->desc2, dma_desc->desc3); + + /* In case the Tx DMA engine is running, make sure everything + * is written to the descriptor(s) before setting the OWN bit + * for the first descriptor + */ + dma_wmb(); + + /* Set OWN bit for the first descriptor */ + desc_data = FXGMAC_GET_DESC_DATA(ring, start_index); + dma_desc = desc_data->dma_desc; + fxgmac_set_bits_le(&dma_desc->desc3, TX_NORMAL_DESC3_OWN_POS, + TX_NORMAL_DESC3_OWN_LEN, 1); + + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, + "%s, first descs, start=%d, desc=%#x,%#x,%#x,%#x\n", + __func__, start_index, dma_desc->desc0, dma_desc->desc1, + dma_desc->desc2, dma_desc->desc3); + + if (netif_msg_tx_queued(pdata)) + fxgmac_dump_tx_desc(pdata, ring, start_index, + pkt_info->desc_count, 1); + + /* Make sure ownership is written to the descriptor */ + smp_wmb(); + + ring->cur = FXGMAC_GET_ENTRY(cur_index, ring->dma_desc_count); + fxgmac_tx_start_xmit(channel, ring); + + if (netif_msg_tx_done(pdata)) { + yt_dbg(pdata, "%s, %s: descriptors %u to %u written\n", + __func__, channel->name, + start_index & (ring->dma_desc_count - 1), + (ring->cur - 1) & (ring->dma_desc_count - 1)); + } +} + +static void fxgmac_get_rx_tstamp(struct fxgmac_pkt_info *pkt_info, + struct fxgmac_dma_desc *dma_desc) +{ + u64 nsec; + + nsec = le32_to_cpu(dma_desc->desc1); + nsec <<= 32; + nsec |= le32_to_cpu(dma_desc->desc0); + if (nsec != 0xffffffffffffffffULL) { + pkt_info->rx_tstamp = nsec; + fxgmac_set_bits(&pkt_info->attributes, + RX_PKT_ATTR_RX_TSTAMP_POS, + RX_PKT_ATTR_RX_TSTAMP_LEN, 1); + } +} + +static int fxgmac_dev_read(struct fxgmac_channel *channel) +{ + struct fxgmac_pdata *pdata = channel->pdata; + struct fxgmac_ring *ring = channel->rx_ring; + struct net_device *netdev = pdata->netdev; + static unsigned int cnt_incomplete; + struct fxgmac_desc_data *desc_data; + struct fxgmac_dma_desc *dma_desc; + struct fxgmac_pkt_info *pkt_info; + u32 ipce, iphe, rxparser; + unsigned int err, etlt; + unsigned int *attr; + + desc_data = FXGMAC_GET_DESC_DATA(ring, ring->cur); + dma_desc = desc_data->dma_desc; + pkt_info = &ring->pkt_info; + attr = &pkt_info->attributes; + + /* Check for data availability */ + if (FXGMAC_GET_BITS_LE(dma_desc->desc3, RX_NORMAL_DESC3_OWN_POS, + RX_NORMAL_DESC3_OWN_LEN)) + return 1; + + /* Make sure descriptor fields are read after reading the OWN bit */ + dma_rmb(); + + if (netif_msg_rx_status(pdata)) + fxgmac_dump_rx_desc(pdata, ring, ring->cur); + + if (FXGMAC_GET_BITS_LE(dma_desc->desc3, RX_NORMAL_DESC3_CTXT_POS, + RX_NORMAL_DESC3_CTXT_LEN)) { + /* Timestamp Context Descriptor */ + fxgmac_get_rx_tstamp(pkt_info, dma_desc); + + fxgmac_set_bits(attr, RX_PKT_ATTR_CONTEXT_POS, + RX_PKT_ATTR_CONTEXT_LEN, 1); + fxgmac_set_bits(attr, RX_PKT_ATTR_CONTEXT_NEXT_POS, + RX_PKT_ATTR_CONTEXT_NEXT_LEN, 0); + if (netif_msg_rx_status(pdata)) + yt_dbg(pdata, "%s, context desc ch=%s\n", __func__, + channel->name); + return 0; + } + + /* Normal Descriptor, be sure Context Descriptor bit is off */ + fxgmac_set_bits(attr, RX_PKT_ATTR_CONTEXT_POS, RX_PKT_ATTR_CONTEXT_LEN, + 0); + + /* Indicate if a Context Descriptor is next */ + /* Get the header length */ + if (FXGMAC_GET_BITS_LE(dma_desc->desc3, RX_NORMAL_DESC3_FD_POS, + RX_NORMAL_DESC3_FD_LEN)) { + desc_data->rx.hdr_len = + FXGMAC_GET_BITS_LE(dma_desc->desc2, + RX_NORMAL_DESC2_HL_POS, + RX_NORMAL_DESC2_HL_LEN); + if (desc_data->rx.hdr_len) + pdata->stats.rx_split_header_packets++; + } + + /* Get the pkt_info length */ + desc_data->rx.len = FXGMAC_GET_BITS_LE(dma_desc->desc3, + RX_NORMAL_DESC3_PL_POS, + RX_NORMAL_DESC3_PL_LEN); + + if (!FXGMAC_GET_BITS_LE(dma_desc->desc3, RX_NORMAL_DESC3_LD_POS, + RX_NORMAL_DESC3_LD_LEN)) { + /* Not all the data has been transferred for this pkt_info */ + fxgmac_set_bits(attr, RX_PKT_ATTR_INCOMPLETE_POS, + RX_PKT_ATTR_INCOMPLETE_LEN, 1); + cnt_incomplete++; + if (cnt_incomplete < 2 && netif_msg_rx_status(pdata)) + yt_dbg(pdata, + "%s, not last desc,pkt incomplete yet,%u\n", + __func__, cnt_incomplete); + + return 0; + } + if ((cnt_incomplete) && netif_msg_rx_status(pdata)) + yt_dbg(pdata, "%s, rx back to normal and incomplete cnt=%u\n", + __func__, cnt_incomplete); + cnt_incomplete = 0; + + /* This is the last of the data for this pkt_info */ + fxgmac_set_bits(attr, RX_PKT_ATTR_INCOMPLETE_POS, + RX_PKT_ATTR_INCOMPLETE_LEN, 0); + + /* Set checksum done indicator as appropriate */ + if (netdev->features & NETIF_F_RXCSUM) { + ipce = FXGMAC_GET_BITS_LE(dma_desc->desc1, + RX_NORMAL_DESC1_WB_IPCE_POS, + RX_NORMAL_DESC1_WB_IPCE_LEN); + iphe = FXGMAC_GET_BITS_LE(dma_desc->desc1, + RX_NORMAL_DESC1_WB_IPHE_POS, + RX_NORMAL_DESC1_WB_IPHE_LEN); + if (!ipce && !iphe) + fxgmac_set_bits(attr, RX_PKT_ATTR_CSUM_DONE_POS, + RX_PKT_ATTR_CSUM_DONE_LEN, 1); + else + return 0; + } + + /* Check for errors (only valid in last descriptor) */ + err = FXGMAC_GET_BITS_LE(dma_desc->desc3, RX_NORMAL_DESC3_ES_POS, + RX_NORMAL_DESC3_ES_LEN); + + /* b111: Incomplete parsing due to ECC error */ + rxparser = FXGMAC_GET_BITS_LE(dma_desc->desc2, + RX_NORMAL_DESC2_WB_RAPARSER_POS, + RX_NORMAL_DESC2_WB_RAPARSER_LEN); + if (err || rxparser == 0x7) { + fxgmac_set_bits(&pkt_info->errors, RX_PACKET_ERRORS_FRAME_POS, + RX_PACKET_ERRORS_FRAME_LEN, 1); + return 0; + } + + etlt = FXGMAC_GET_BITS_LE(dma_desc->desc3, RX_NORMAL_DESC3_ETLT_POS, + RX_NORMAL_DESC3_ETLT_LEN); + + if (etlt == 0x4 && (netdev->features & NETIF_F_HW_VLAN_CTAG_RX)) { + fxgmac_set_bits(attr, RX_PKT_ATTR_VLAN_CTAG_POS, + RX_PKT_ATTR_VLAN_CTAG_LEN, 1); + pkt_info->vlan_ctag = + FXGMAC_GET_BITS_LE(dma_desc->desc0, + RX_NORMAL_DESC0_OVT_POS, + RX_NORMAL_DESC0_OVT_LEN); + yt_dbg(pdata, "vlan-ctag=%#06x\n", pkt_info->vlan_ctag); + } + + return 0; +} + void fxgmac_hw_ops_init(struct fxgmac_hw_ops *hw_ops) { hw_ops->pcie_init = fxgmac_pcie_init; @@ -2168,6 +2570,8 @@ void fxgmac_hw_ops_init(struct fxgmac_hw_ops *hw_ops) hw_ops->disable_tx = fxgmac_disable_tx; hw_ops->enable_rx = fxgmac_enable_rx; hw_ops->disable_rx = fxgmac_disable_rx; + hw_ops->dev_read = fxgmac_dev_read; + hw_ops->dev_xmit = fxgmac_dev_xmit; hw_ops->enable_channel_irq = fxgmac_enable_channel_irq; hw_ops->disable_channel_irq = fxgmac_disable_channel_irq; diff --git a/drivers/net/ethernet/motorcomm/yt6801/yt6801_net.c b/drivers/net/ethernet/motorcomm/yt6801/yt6801_net.c index c3baa06b0..fa1587e69 100644 --- a/drivers/net/ethernet/motorcomm/yt6801/yt6801_net.c +++ b/drivers/net/ethernet/motorcomm/yt6801/yt6801_net.c @@ -38,6 +38,206 @@ static unsigned int fxgmac_desc_rx_dirty(struct fxgmac_ring *ring) return dirty; } +void fxgmac_tx_start_xmit(struct fxgmac_channel *channel, + struct fxgmac_ring *ring) +{ + struct fxgmac_pdata *pdata = channel->pdata; + struct fxgmac_desc_data *desc_data; + + /* Make sure everything is written before the register write */ + wmb(); + + /* Issue a poll command to Tx DMA by writing address + * of next immediate free descriptor + */ + desc_data = FXGMAC_GET_DESC_DATA(ring, ring->cur); + wr32_mac(pdata, lower_32_bits(desc_data->dma_desc_addr), + FXGMAC_DMA_REG(channel, DMA_CH_TDTR_LO)); + + if (netif_msg_tx_done(pdata)) { + yt_dbg(pdata, + "tx_start_xmit: dump before wr reg, dma base=0x%016llx,reg=0x%08x,", + desc_data->dma_desc_addr, + rd32_mac(pdata, FXGMAC_DMA_REG(channel, DMA_CH_TDTR_LO))); + + yt_dbg(pdata, "tx timer usecs=%u,tx_timer_active=%u\n", + pdata->tx_usecs, channel->tx_timer_active); + } + + ring->tx.xmit_more = 0; +} + +static netdev_tx_t fxgmac_maybe_stop_tx_queue(struct fxgmac_channel *channel, + struct fxgmac_ring *ring, + unsigned int count) +{ + struct fxgmac_pdata *pdata = channel->pdata; + + if (count > fxgmac_desc_tx_avail(ring)) { + /* Avoid wrongly optimistic queue wake-up: tx poll thread must + * not miss a ring update when it notices a stopped queue. + */ + smp_wmb(); + netif_stop_subqueue(pdata->netdev, channel->queue_index); + ring->tx.queue_stopped = 1; + + /* Sync with tx poll: + * - publish queue status and cur ring index (write barrier) + * - refresh dirty ring index (read barrier). + * May the current thread have a pessimistic view of the ring + * status and forget to wake up queue, a racing tx poll thread + * can't. + */ + smp_mb(); + if (count <= fxgmac_desc_tx_avail(ring)) { + ring->tx.queue_stopped = 0; + netif_start_subqueue(pdata->netdev, + channel->queue_index); + fxgmac_tx_start_xmit(channel, ring); + } else { + /* If we haven't notified the hardware because of + * xmit_more support, tell it now + */ + if (ring->tx.xmit_more) + fxgmac_tx_start_xmit(channel, ring); + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, "about stop tx q, ret BUSY\n"); + return NETDEV_TX_BUSY; + } + } + + return NETDEV_TX_OK; +} + +static void fxgmac_prep_vlan(struct sk_buff *skb, + struct fxgmac_pkt_info *pkt_info) +{ + if (skb_vlan_tag_present(skb)) + pkt_info->vlan_ctag = skb_vlan_tag_get(skb); +} + +static int fxgmac_prep_tso(struct fxgmac_pdata *pdata, struct sk_buff *skb, + struct fxgmac_pkt_info *pkt_info) +{ + int ret; + + if (!FXGMAC_GET_BITS(pkt_info->attributes, TX_PKT_ATTR_TSO_ENABLE_POS, + TX_PKT_ATTR_TSO_ENABLE_LEN)) + return 0; + + ret = skb_cow_head(skb, 0); + if (ret) + return ret; + + pkt_info->header_len = skb_transport_offset(skb) + tcp_hdrlen(skb); + pkt_info->tcp_header_len = tcp_hdrlen(skb); + pkt_info->tcp_payload_len = skb->len - pkt_info->header_len; + pkt_info->mss = skb_shinfo(skb)->gso_size; + + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, + "header_len=%u, tcp_header_len=%u, tcp_payload_len=%u, mss=%u\n", + pkt_info->header_len, pkt_info->tcp_header_len, + pkt_info->tcp_payload_len, pkt_info->mss); + + /* Update the number of packets that will ultimately be transmitted + * along with the extra bytes for each extra packet + */ + pkt_info->tx_packets = skb_shinfo(skb)->gso_segs; + pkt_info->tx_bytes += (pkt_info->tx_packets - 1) * pkt_info->header_len; + + return 0; +} + +static int fxgmac_is_tso(struct sk_buff *skb) +{ + if (skb->ip_summed != CHECKSUM_PARTIAL) + return 0; + + if (!skb_is_gso(skb)) + return 0; + + return 1; +} + +static void fxgmac_prep_tx_pkt(struct fxgmac_pdata *pdata, + struct fxgmac_ring *ring, struct sk_buff *skb, + struct fxgmac_pkt_info *pkt_info) +{ + u32 *attr = &pkt_info->attributes; + u32 len, context_desc = 0; + + pkt_info->skb = skb; + pkt_info->desc_count = 0; + pkt_info->tx_packets = 1; + pkt_info->tx_bytes = skb->len; + + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, "%s, pkt desc cnt=%d,skb len=%d, skbheadlen=%d\n", + __func__, pkt_info->desc_count, skb->len, + skb_headlen(skb)); + + if (fxgmac_is_tso(skb)) { + /* TSO requires an extra descriptor if mss is different */ + if (skb_shinfo(skb)->gso_size != ring->tx.cur_mss) { + context_desc = 1; + pkt_info->desc_count++; + } + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, + "fxgmac_is_tso=%d, ip_summed=%d,skb gso=%d\n", + ((skb->ip_summed == CHECKSUM_PARTIAL) && + (skb_is_gso(skb))) ? 1 : 0, + skb->ip_summed, skb_is_gso(skb) ? 1 : 0); + + /* TSO requires an extra descriptor for TSO header */ + pkt_info->desc_count++; + fxgmac_set_bits(attr, TX_PKT_ATTR_TSO_ENABLE_POS, + TX_PKT_ATTR_TSO_ENABLE_LEN, 1); + fxgmac_set_bits(attr, TX_PKT_ATTR_CSUM_ENABLE_POS, + TX_PKT_ATTR_CSUM_ENABLE_LEN, 1); + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, "%s,tso, pkt desc cnt=%d\n", __func__, + pkt_info->desc_count); + } else if (skb->ip_summed == CHECKSUM_PARTIAL) + fxgmac_set_bits(attr, TX_PKT_ATTR_CSUM_ENABLE_POS, + TX_PKT_ATTR_CSUM_ENABLE_LEN, 1); + + if (skb_vlan_tag_present(skb)) { + /* VLAN requires an extra descriptor if tag is different */ + if (skb_vlan_tag_get(skb) != ring->tx.cur_vlan_ctag) + /* We can share with the TSO context descriptor */ + if (!context_desc) { + context_desc = 1; + pkt_info->desc_count++; + } + + fxgmac_set_bits(attr, TX_PKT_ATTR_VLAN_CTAG_POS, + TX_PKT_ATTR_VLAN_CTAG_LEN, 1); + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, "%s,VLAN, pkt desc cnt=%d,vlan=0x%04x\n", + __func__, pkt_info->desc_count, + skb_vlan_tag_get(skb)); + } + + for (len = skb_headlen(skb); len;) { + pkt_info->desc_count++; + len -= min_t(unsigned int, len, FXGMAC_TX_MAX_BUF_SIZE); + } + + for (u32 i = 0; i < skb_shinfo(skb)->nr_frags; i++) + for (len = skb_frag_size(&skb_shinfo(skb)->frags[i]); len;) { + pkt_info->desc_count++; + len -= min_t(unsigned int, len, FXGMAC_TX_MAX_BUF_SIZE); + } + + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, + "%s,pkt desc cnt%d,skb len%d, skbheadlen=%d,frags=%d\n", + __func__, pkt_info->desc_count, skb->len, + skb_headlen(skb), skb_shinfo(skb)->nr_frags); +} + static int fxgmac_calc_rx_buf_size(struct fxgmac_pdata *pdata, unsigned int mtu) { u32 rx_buf_size, max_mtu; @@ -1713,8 +1913,77 @@ void fxgmac_dbg_pkt(struct fxgmac_pdata *pdata, struct sk_buff *skb, bool tx_rx) yt_dbg(pdata, "\n************** SKB dump ****************\n"); } +static netdev_tx_t fxgmac_xmit(struct sk_buff *skb, struct net_device *netdev) +{ + struct fxgmac_pdata *pdata = netdev_priv(netdev); + struct fxgmac_pkt_info *tx_pkt_info; + struct fxgmac_channel *channel; + struct fxgmac_hw_ops *hw_ops; + struct netdev_queue *txq; + struct fxgmac_ring *ring; + int ret; + + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, "%s, skb->len=%d,q=%d\n", __func__, skb->len, + skb->queue_mapping); + + channel = pdata->channel_head + skb->queue_mapping; + txq = netdev_get_tx_queue(netdev, channel->queue_index); + ring = channel->tx_ring; + tx_pkt_info = &ring->pkt_info; + + hw_ops = &pdata->hw_ops; + + if (skb->len == 0) { + yt_err(pdata, "empty skb received from stack\n"); + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; + } + + /* Prepare preliminary packet info for TX */ + memset(tx_pkt_info, 0, sizeof(*tx_pkt_info)); + fxgmac_prep_tx_pkt(pdata, ring, skb, tx_pkt_info); + + /* Check that there are enough descriptors available */ + ret = fxgmac_maybe_stop_tx_queue(channel, ring, + tx_pkt_info->desc_count); + if (ret == NETDEV_TX_BUSY) + return ret; + + ret = fxgmac_prep_tso(pdata, skb, tx_pkt_info); + if (ret < 0) { + yt_err(pdata, "error processing TSO packet\n"); + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; + } + fxgmac_prep_vlan(skb, tx_pkt_info); + + if (!fxgmac_tx_skb_map(channel, skb)) { + dev_kfree_skb_any(skb); + yt_err(pdata, "xmit, map tx skb err\n"); + return NETDEV_TX_OK; + } + + /* Report on the actual number of bytes (to be) sent */ + netdev_tx_sent_queue(txq, tx_pkt_info->tx_bytes); + if (netif_msg_tx_done(pdata)) + yt_dbg(pdata, "xmit,before hw_xmit, byte len=%d\n", + tx_pkt_info->tx_bytes); + + /* Configure required descriptor fields for transmission */ + hw_ops->dev_xmit(channel); + + if (netif_msg_pktdata(pdata)) + fxgmac_dbg_pkt(pdata, skb, true); + + /* Stop the queue in advance if there may not be enough descriptors */ + fxgmac_maybe_stop_tx_queue(channel, ring, FXGMAC_TX_MAX_DESC_NR); + + return NETDEV_TX_OK; +} static const struct net_device_ops fxgmac_netdev_ops = { .ndo_open = fxgmac_open, + .ndo_start_xmit = fxgmac_xmit, }; const struct net_device_ops *fxgmac_get_netdev_ops(void)