From patchwork Fri Oct 18 08:40:02 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Kurt Kanzenbach X-Patchwork-Id: 1998982 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=osuosl.org header.i=@osuosl.org header.a=rsa-sha256 header.s=default header.b=LIDUvPbu; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=osuosl.org (client-ip=2605:bc80:3010::136; helo=smtp3.osuosl.org; envelope-from=intel-wired-lan-bounces@osuosl.org; receiver=patchwork.ozlabs.org) Received: from smtp3.osuosl.org (smtp3.osuosl.org [IPv6:2605:bc80:3010::136]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (secp384r1) server-digest SHA384) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4XVJ793zNQz1xwD for ; Fri, 18 Oct 2024 19:40:21 +1100 (AEDT) Received: from localhost (localhost [127.0.0.1]) by smtp3.osuosl.org (Postfix) with ESMTP id E54376062F; Fri, 18 Oct 2024 08:40:18 +0000 (UTC) X-Virus-Scanned: amavis at osuosl.org Received: from smtp3.osuosl.org ([127.0.0.1]) by localhost (smtp3.osuosl.org [127.0.0.1]) (amavis, port 10024) with ESMTP id SSrQ3EyCe_1F; Fri, 18 Oct 2024 08:40:18 +0000 (UTC) X-Comment: SPF check N/A for local connections - client-ip=140.211.166.142; helo=lists1.osuosl.org; envelope-from=intel-wired-lan-bounces@osuosl.org; receiver= DKIM-Filter: OpenDKIM Filter v2.11.0 smtp3.osuosl.org EA80F60AAD DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=osuosl.org; s=default; t=1729240818; bh=4vj4+7zQnr0qpPNcFDIF3nADIgFtRl4ND+JkIeZZpyo=; h=From:Date:References:In-Reply-To:To:Cc:Subject:List-Id: List-Unsubscribe:List-Archive:List-Post:List-Help:List-Subscribe: From; b=LIDUvPbuWVLEroyVdT4VijGzagRZ6jtUQ1+Z2AgkZmXn3hm56UhTsDdZzHuawoN57 XYgaj2/iTcWTmJimFlWFnZj0YFY895cfMBBQeliDxV1fOvtY2JPg1/9AKeKOXbAMHA TPF8b8kEe2fYAx1ED29c+PZva6NSbTyu3qvCIq6rNT0u7gpG7dNu9d9RMlCE0oV1QX L6J39y/v+c9TFlmRLs7yypnvb695M1WcYrBnccwLJR+wDgewz4oXdWIaVJbs+gyS/k ngQsejW2oVBsLVJnVMHn1fdyOHFuzUeNgj6MFClNtC73lntQUhMIsmb7OMtsMJC+I1 vpUyCp549mRFg== Received: from lists1.osuosl.org (lists1.osuosl.org [140.211.166.142]) by smtp3.osuosl.org (Postfix) with ESMTP id EA80F60AAD; Fri, 18 Oct 2024 08:40:17 +0000 (UTC) X-Original-To: intel-wired-lan@lists.osuosl.org Delivered-To: intel-wired-lan@lists.osuosl.org Received: from smtp3.osuosl.org (smtp3.osuosl.org [140.211.166.136]) by lists1.osuosl.org (Postfix) with ESMTP id D20A52804 for ; Fri, 18 Oct 2024 08:40:15 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by smtp3.osuosl.org (Postfix) with ESMTP id C0CB160628 for ; Fri, 18 Oct 2024 08:40:15 +0000 (UTC) X-Virus-Scanned: amavis at osuosl.org Received: from smtp3.osuosl.org ([127.0.0.1]) by localhost (smtp3.osuosl.org [127.0.0.1]) (amavis, port 10024) with ESMTP id tvPOHWuq7zI3 for ; Fri, 18 Oct 2024 08:40:14 +0000 (UTC) Received-SPF: Pass (mailfrom) identity=mailfrom; client-ip=2a0a:51c0:0:12e:550::1; helo=galois.linutronix.de; envelope-from=kurt@linutronix.de; receiver= DMARC-Filter: OpenDMARC Filter v1.4.2 smtp3.osuosl.org 583B56062F DKIM-Filter: OpenDKIM Filter v2.11.0 smtp3.osuosl.org 583B56062F Received: from galois.linutronix.de (Galois.linutronix.de [IPv6:2a0a:51c0:0:12e:550::1]) by smtp3.osuosl.org (Postfix) with ESMTPS id 583B56062F for ; Fri, 18 Oct 2024 08:40:14 +0000 (UTC) From: Kurt Kanzenbach Date: Fri, 18 Oct 2024 10:40:02 +0200 MIME-Version: 1.0 Message-Id: <20241018-b4-igb_zero_copy-v9-6-da139d78d796@linutronix.de> References: <20241018-b4-igb_zero_copy-v9-0-da139d78d796@linutronix.de> In-Reply-To: <20241018-b4-igb_zero_copy-v9-0-da139d78d796@linutronix.de> To: Tony Nguyen , Przemek Kitszel Cc: "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Alexei Starovoitov , Daniel Borkmann , Jesper Dangaard Brouer , John Fastabend , Richard Cochran , Sriram Yagnaraman , Benjamin Steinke , Sebastian Andrzej Siewior , Maciej Fijalkowski , intel-wired-lan@lists.osuosl.org, netdev@vger.kernel.org, bpf@vger.kernel.org, Sriram Yagnaraman , Kurt Kanzenbach X-Developer-Signature: v=1; a=openpgp-sha256; l=9844; i=kurt@linutronix.de; h=from:subject:message-id; bh=2VOrkHfABIRZuuqUvnmAIdU7/YEtW6QOicEf0xDcLEM=; b=owEBbQKS/ZANAwAKAcGT0fKqRnOCAcsmYgBnEh7jOsfYkYvomE1ad11pClK8jqYdCVCug6N/R NiNMfZApq6JAjMEAAEKAB0WIQS8ub+yyMN909/bWZLBk9HyqkZzggUCZxIe4wAKCRDBk9HyqkZz gohND/9Xc1d0XdlrZemsZXV9O8chTnoNiNqAk7KqRHZ/DBLvYMeIchUGPcyymeaSL+9GGrobPea OhutwWMYRyYrSAvyy2tQTsy4tZ8juJNnuekn/pr54ny0om2+H1j+URynHsfhc73daE6bn7+xYHx D7jkhPasECdWS/oBmAKsCPIvnDYoWq2FnDBmZg8PempmwbHY9mUGVEZFg0M4BY46R6xFbflg4ts J+zp0Bd93tVTBhMiPnzjzrqPGHuo2eLCIkHSJDIlmWvaqR+aKDjIgUs2ibJIc/Yi/swg4k+VlHm HUtb90wxSe74Tj02yfZV29mL4h1K+cfQL1jJgkDFBv4JtDKECFnrGwi63R1+bxzo3NqR/Nv87u2 fsBZrVx1oWvkM6B7D57n64E9bmKYprCvJg1C3hLVVaW7lS5lo6p47xL4vAAOUsm1GG6HNossACW /8eNABdJzxoVOFw/i28X+MTKEwADIF8aZ6rqyt3UH2e3aZ3/pcgYt1/0MotJhCklXOmtYUaFpdh rtMHRJ8CILKj1Y/HaFVzLLJx730mf0ogd8moMnduFvTz+MFAXb7l8ajbFQCSagiYAbj3ypfSc2w iENg7+M935DxwhMaopHcNaXBBqLbTuBjvSuY9NY70hnpXHh09ssBP2hpmn4xDm43oApuvBY/VqD e/zjQ9q28pgL1nQ== X-Developer-Key: i=kurt@linutronix.de; a=openpgp; fpr=BCB9BFB2C8C37DD3DFDB5992C193D1F2AA467382 X-Mailman-Original-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linutronix.de; s=2020; t=1729240812; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=4vj4+7zQnr0qpPNcFDIF3nADIgFtRl4ND+JkIeZZpyo=; b=YWJtcxLUOSqhufC69cBU/U5CkNvriPglLxz16cGNT25GWVNy7/VXMtXCLCSNU/2q7SgM6/ RPW/CkvjtNmLWsW3k6PuqOsXGL0lM4Vz/b2CR2GeEaqhgjhJFfHXDsky9aF1FPE6lHX/g5 7kkj33mqAFFz0K5tYL6KBkop9DNpyJnwWmJec1xi11tyXtPOSgzLVVaBvw3FnxCiOVu0FB uc1X6iWW+DMTMkXxwX2oaoLdrYAJM+2Hui4fPeUWgug6Oa/k4UL9cdmohhbMYEQtUaJnDL diCS8ZrlVlyhR5Z2xklLEY0c1FTsk0Roud0lE14AgPk7D9o/AkhkJyo8brvaUA== X-Mailman-Original-DKIM-Signature: v=1; a=ed25519-sha256; c=relaxed/relaxed; d=linutronix.de; s=2020e; t=1729240812; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=4vj4+7zQnr0qpPNcFDIF3nADIgFtRl4ND+JkIeZZpyo=; b=kG8/CkPD59fJiTR4X/zl6qq1ZQfYFXClDK29pZ64OrfFscoUAT/Q1t3RCI6rcbBIq8158K YI+QhoxumAk2m2BA== X-Mailman-Original-Authentication-Results: smtp3.osuosl.org; dmarc=pass (p=none dis=none) header.from=linutronix.de X-Mailman-Original-Authentication-Results: smtp3.osuosl.org; dkim=pass (2048-bit key) header.d=linutronix.de header.i=@linutronix.de header.a=rsa-sha256 header.s=2020 header.b=YWJtcxLU; dkim=pass header.d=linutronix.de header.i=@linutronix.de header.a=ed25519-sha256 header.s=2020e header.b=kG8/CkPD Subject: [Intel-wired-lan] [PATCH iwl-next v9 6/6] igb: Add AF_XDP zero-copy Tx support X-BeenThere: intel-wired-lan@osuosl.org X-Mailman-Version: 2.1.30 Precedence: list List-Id: Intel Wired Ethernet Linux Kernel Driver Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-wired-lan-bounces@osuosl.org Sender: "Intel-wired-lan" From: Sriram Yagnaraman Add support for AF_XDP zero-copy transmit path. A new TX buffer type IGB_TYPE_XSK is introduced to indicate that the Tx frame was allocated from the xsk buff pool, so igb_clean_tx_ring() and igb_clean_tx_irq() can clean the buffers correctly based on type. igb_xmit_zc() performs the actual packet transmit when AF_XDP zero-copy is enabled. We share the TX ring between slow path, XDP and AF_XDP zero-copy, so we use the netdev queue lock to ensure mutual exclusion. Signed-off-by: Sriram Yagnaraman [Kurt: Set olinfo_status in igb_xmit_zc() so that frames are transmitted, Use READ_ONCE() for xsk_pool and check Tx disabled and carrier in igb_xmit_zc(), Add FIXME for RS bit] Signed-off-by: Kurt Kanzenbach Reviewed-by: Maciej Fijalkowski --- drivers/net/ethernet/intel/igb/igb.h | 2 + drivers/net/ethernet/intel/igb/igb_main.c | 61 +++++++++++++++++++++++++----- drivers/net/ethernet/intel/igb/igb_xsk.c | 63 +++++++++++++++++++++++++++++++ 3 files changed, 116 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/intel/igb/igb.h b/drivers/net/ethernet/intel/igb/igb.h index e4a85867aa18..02f340280d20 100644 --- a/drivers/net/ethernet/intel/igb/igb.h +++ b/drivers/net/ethernet/intel/igb/igb.h @@ -258,6 +258,7 @@ enum igb_tx_flags { enum igb_tx_buf_type { IGB_TYPE_SKB = 0, IGB_TYPE_XDP, + IGB_TYPE_XSK }; /* wrapper around a pointer to a socket buffer, @@ -859,6 +860,7 @@ bool igb_alloc_rx_buffers_zc(struct igb_ring *rx_ring, void igb_clean_rx_ring_zc(struct igb_ring *rx_ring); int igb_clean_rx_irq_zc(struct igb_q_vector *q_vector, struct xsk_buff_pool *xsk_pool, const int budget); +bool igb_xmit_zc(struct igb_ring *tx_ring, struct xsk_buff_pool *xsk_pool); int igb_xsk_wakeup(struct net_device *dev, u32 qid, u32 flags); #endif /* _IGB_H_ */ diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index 711b60cab594..4587877d1761 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -2979,6 +2979,9 @@ static int igb_xdp_xmit(struct net_device *dev, int n, if (unlikely(!tx_ring)) return -ENXIO; + if (unlikely(test_bit(IGB_RING_FLAG_TX_DISABLED, &tx_ring->flags))) + return -ENXIO; + nq = txring_txq(tx_ring); __netif_tx_lock(nq, cpu); @@ -3326,7 +3329,8 @@ static int igb_probe(struct pci_dev *pdev, const struct pci_device_id *ent) netdev->priv_flags |= IFF_SUPP_NOFCS; netdev->priv_flags |= IFF_UNICAST_FLT; - netdev->xdp_features = NETDEV_XDP_ACT_BASIC | NETDEV_XDP_ACT_REDIRECT; + netdev->xdp_features = NETDEV_XDP_ACT_BASIC | NETDEV_XDP_ACT_REDIRECT | + NETDEV_XDP_ACT_XSK_ZEROCOPY; /* MTU range: 68 - 9216 */ netdev->min_mtu = ETH_MIN_MTU; @@ -4900,15 +4904,20 @@ void igb_clean_tx_ring(struct igb_ring *tx_ring) { u16 i = tx_ring->next_to_clean; struct igb_tx_buffer *tx_buffer = &tx_ring->tx_buffer_info[i]; + u32 xsk_frames = 0; while (i != tx_ring->next_to_use) { union e1000_adv_tx_desc *eop_desc, *tx_desc; /* Free all the Tx ring sk_buffs or xdp frames */ - if (tx_buffer->type == IGB_TYPE_SKB) + if (tx_buffer->type == IGB_TYPE_SKB) { dev_kfree_skb_any(tx_buffer->skb); - else + } else if (tx_buffer->type == IGB_TYPE_XDP) { xdp_return_frame(tx_buffer->xdpf); + } else if (tx_buffer->type == IGB_TYPE_XSK) { + xsk_frames++; + goto skip_for_xsk; + } /* unmap skb header data */ dma_unmap_single(tx_ring->dev, @@ -4939,6 +4948,7 @@ void igb_clean_tx_ring(struct igb_ring *tx_ring) DMA_TO_DEVICE); } +skip_for_xsk: tx_buffer->next_to_watch = NULL; /* move us one more past the eop_desc for start of next pkt */ @@ -4953,6 +4963,9 @@ void igb_clean_tx_ring(struct igb_ring *tx_ring) /* reset BQL for queue */ netdev_tx_reset_queue(txring_txq(tx_ring)); + if (tx_ring->xsk_pool && xsk_frames) + xsk_tx_completed(tx_ring->xsk_pool, xsk_frames); + /* reset next_to_use and next_to_clean */ tx_ring->next_to_use = 0; tx_ring->next_to_clean = 0; @@ -6486,6 +6499,9 @@ netdev_tx_t igb_xmit_frame_ring(struct sk_buff *skb, return NETDEV_TX_BUSY; } + if (unlikely(test_bit(IGB_RING_FLAG_TX_DISABLED, &tx_ring->flags))) + return NETDEV_TX_BUSY; + /* record the location of the first descriptor for this packet */ first = &tx_ring->tx_buffer_info[tx_ring->next_to_use]; first->type = IGB_TYPE_SKB; @@ -8260,13 +8276,18 @@ static int igb_poll(struct napi_struct *napi, int budget) **/ static bool igb_clean_tx_irq(struct igb_q_vector *q_vector, int napi_budget) { - struct igb_adapter *adapter = q_vector->adapter; - struct igb_ring *tx_ring = q_vector->tx.ring; - struct igb_tx_buffer *tx_buffer; - union e1000_adv_tx_desc *tx_desc; unsigned int total_bytes = 0, total_packets = 0; + struct igb_adapter *adapter = q_vector->adapter; unsigned int budget = q_vector->tx.work_limit; + struct igb_ring *tx_ring = q_vector->tx.ring; unsigned int i = tx_ring->next_to_clean; + union e1000_adv_tx_desc *tx_desc; + struct igb_tx_buffer *tx_buffer; + struct xsk_buff_pool *xsk_pool; + int cpu = smp_processor_id(); + bool xsk_xmit_done = true; + struct netdev_queue *nq; + u32 xsk_frames = 0; if (test_bit(__IGB_DOWN, &adapter->state)) return true; @@ -8297,10 +8318,14 @@ static bool igb_clean_tx_irq(struct igb_q_vector *q_vector, int napi_budget) total_packets += tx_buffer->gso_segs; /* free the skb */ - if (tx_buffer->type == IGB_TYPE_SKB) + if (tx_buffer->type == IGB_TYPE_SKB) { napi_consume_skb(tx_buffer->skb, napi_budget); - else + } else if (tx_buffer->type == IGB_TYPE_XDP) { xdp_return_frame(tx_buffer->xdpf); + } else if (tx_buffer->type == IGB_TYPE_XSK) { + xsk_frames++; + goto skip_for_xsk; + } /* unmap skb header data */ dma_unmap_single(tx_ring->dev, @@ -8332,6 +8357,7 @@ static bool igb_clean_tx_irq(struct igb_q_vector *q_vector, int napi_budget) } } +skip_for_xsk: /* move us one more past the eop_desc for start of next pkt */ tx_buffer++; tx_desc++; @@ -8360,6 +8386,21 @@ static bool igb_clean_tx_irq(struct igb_q_vector *q_vector, int napi_budget) q_vector->tx.total_bytes += total_bytes; q_vector->tx.total_packets += total_packets; + xsk_pool = READ_ONCE(tx_ring->xsk_pool); + if (xsk_pool) { + if (xsk_frames) + xsk_tx_completed(xsk_pool, xsk_frames); + if (xsk_uses_need_wakeup(xsk_pool)) + xsk_set_tx_need_wakeup(xsk_pool); + + nq = txring_txq(tx_ring); + __netif_tx_lock(nq, cpu); + /* Avoid transmit queue timeout since we share it with the slow path */ + txq_trans_cond_update(nq); + xsk_xmit_done = igb_xmit_zc(tx_ring, xsk_pool); + __netif_tx_unlock(nq); + } + if (test_bit(IGB_RING_FLAG_TX_DETECT_HANG, &tx_ring->flags)) { struct e1000_hw *hw = &adapter->hw; @@ -8422,7 +8463,7 @@ static bool igb_clean_tx_irq(struct igb_q_vector *q_vector, int napi_budget) } } - return !!budget; + return !!budget && xsk_xmit_done; } /** diff --git a/drivers/net/ethernet/intel/igb/igb_xsk.c b/drivers/net/ethernet/intel/igb/igb_xsk.c index 3d64a9f6360c..157d43787fa0 100644 --- a/drivers/net/ethernet/intel/igb/igb_xsk.c +++ b/drivers/net/ethernet/intel/igb/igb_xsk.c @@ -461,6 +461,69 @@ int igb_clean_rx_irq_zc(struct igb_q_vector *q_vector, return failure ? budget : (int)total_packets; } +bool igb_xmit_zc(struct igb_ring *tx_ring, struct xsk_buff_pool *xsk_pool) +{ + unsigned int budget = igb_desc_unused(tx_ring); + u32 cmd_type, olinfo_status, nb_pkts, i = 0; + struct xdp_desc *descs = xsk_pool->tx_descs; + union e1000_adv_tx_desc *tx_desc = NULL; + struct igb_tx_buffer *tx_buffer_info; + unsigned int total_bytes = 0; + dma_addr_t dma; + + if (!netif_carrier_ok(tx_ring->netdev)) + return true; + + if (test_bit(IGB_RING_FLAG_TX_DISABLED, &tx_ring->flags)) + return true; + + nb_pkts = xsk_tx_peek_release_desc_batch(xsk_pool, budget); + if (!nb_pkts) + return true; + + while (nb_pkts-- > 0) { + dma = xsk_buff_raw_get_dma(xsk_pool, descs[i].addr); + xsk_buff_raw_dma_sync_for_device(xsk_pool, dma, descs[i].len); + + tx_buffer_info = &tx_ring->tx_buffer_info[tx_ring->next_to_use]; + tx_buffer_info->bytecount = descs[i].len; + tx_buffer_info->type = IGB_TYPE_XSK; + tx_buffer_info->xdpf = NULL; + tx_buffer_info->gso_segs = 1; + tx_buffer_info->time_stamp = jiffies; + + tx_desc = IGB_TX_DESC(tx_ring, tx_ring->next_to_use); + tx_desc->read.buffer_addr = cpu_to_le64(dma); + + /* put descriptor type bits */ + cmd_type = E1000_ADVTXD_DTYP_DATA | E1000_ADVTXD_DCMD_DEXT | + E1000_ADVTXD_DCMD_IFCS; + olinfo_status = descs[i].len << E1000_ADVTXD_PAYLEN_SHIFT; + + /* FIXME: This sets the Report Status (RS) bit for every + * descriptor. One nice to have optimization would be to set it + * only for the last descriptor in the whole batch. See Intel + * ice driver for an example on how to do it. + */ + cmd_type |= descs[i].len | IGB_TXD_DCMD; + tx_desc->read.cmd_type_len = cpu_to_le32(cmd_type); + tx_desc->read.olinfo_status = cpu_to_le32(olinfo_status); + + total_bytes += descs[i].len; + + i++; + tx_ring->next_to_use++; + tx_buffer_info->next_to_watch = tx_desc; + if (tx_ring->next_to_use == tx_ring->count) + tx_ring->next_to_use = 0; + } + + netdev_tx_sent_queue(txring_txq(tx_ring), total_bytes); + igb_xdp_ring_update_tail(tx_ring); + + return nb_pkts < budget; +} + int igb_xsk_wakeup(struct net_device *dev, u32 qid, u32 flags) { struct igb_adapter *adapter = netdev_priv(dev);