From nobody Sun Feb 8 02:41:22 2026 Received: from mail-pj1-f74.google.com (mail-pj1-f74.google.com [209.85.216.74]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 8884D204C36 for ; Sat, 8 Mar 2025 21:40:49 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.74 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470052; cv=none; b=jZTG8+hbPB5ZLxjV4b1PpBp0/Q7BDeM72RGGujbmPa4BI28tkRM7d4iIOJmY+P8D61h6YDkDX5LhdUKw8wUdyKyyZseUKhtG6vsl3x5cTthZUxJxQiGXglOJqpWaosTw3a+57zIDWUy0QODcuGJstzYq+l5x5hsArBh4jeocwvE= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470052; c=relaxed/simple; bh=1gKSJ1zhftBSMc9cuS8nUiycP8+YXffYCs0yXilMSBo=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=pP5zbVmWDfYxm9olhyzDSgq+iv2dad1toF2x60iXV5WODgoQWB6HbWAqFTwWKTRu0KhuwDXOgbiaYs28qefGqGf+JVfKORj66H4CCA4fV1IUq3CZlXKj2V7G41QSCtJz2So0HxEHsfAG1NvCxuLqyh0d48Qz2pBFXm8nFU2fkt4= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=xQ2uiICG; arc=none smtp.client-ip=209.85.216.74 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="xQ2uiICG" Received: by mail-pj1-f74.google.com with SMTP id 98e67ed59e1d1-2ff581215f7so4980355a91.3 for ; Sat, 08 Mar 2025 13:40:49 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470049; x=1742074849; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=tX6BmoNjuzIzQpeeDWDPAAnCBo+DQIywGklr81dGywI=; b=xQ2uiICG9X8jqkFh8qGewG1I0erAW8qDUyBKZg3VOhUkkDYrbeUZZlo4jJNofukfBN apkTcMxVMKow5Fw2hOZ5C/XoLl2eq2feHKoEClImlwLXTz6YoRD7Rg4VkTNxbjsOHY98 Fl9BlAkh3xDdUk1E9hipKGd19rIb/O3CORtopeH6ZqJ6g3unGjHtdbj1IYI4BI/rchx4 tQP6/s2UpQuGpuceL66VxVieyRvSSJTbaNnRl74AlE3iQKbKxUvado9Z1nwxnOimYpeH n9ytSLGTHh1snO0xZmSZAzivVqiRe/dW70stXB+wixWJNm47uctgoPkI9fDCLsAmIxCF Bx8g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470049; x=1742074849; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=tX6BmoNjuzIzQpeeDWDPAAnCBo+DQIywGklr81dGywI=; b=uLHv4owxC+srXPoaG+Vx93OacXouZpNVG1Gh0oVw55KiSQam+C4jzKkRFsxMsaKtYH fn7mrEPCucZnWQOKApug6z/1E2TyJ2CEgErzTYrg2z/aCoNVflzji+Vs3hhkUIGnfUcq 9X8kPSJ6uBw927sTSMsmWm64wySYliiFm3tBE/TaqJA1U4ygqHSDTYu546CXz8lyP8oY f6zmdNVf92fvwGm92FeO0MtFxINIjJqN9t2f29WGAp/5XUHreBDf0gLnTCY/bYumBnSG 7lh9I+xUJ8v6XO26Yk6nTFMvp9Kvkn3nPVLKJEiR/ySQ24cyAvf1rsQfKQGEL1M3Y3GP FuHg== X-Forwarded-Encrypted: i=1; AJvYcCUNqZ8bORCCaa45PpKWQxRI/DRQ7W+U634f0+7C/RobwOIETAWBcx0W9MAyBbl6jlavr/LtfWTC2Drz+Oo=@vger.kernel.org X-Gm-Message-State: AOJu0YzpKYYZTAJF0BoMs7Oh09yxukW/XteXBp0HDrplQIGnIk6bLVDX sCv5gcgipSvyqhI/akkjXkRM2kg/OKd/pUlFTFk5qtglAroRIAbrVvmxLcGiYbL14Ku37acCiGj jrvIW+/oL4/c0QxjPI5sJCQ== X-Google-Smtp-Source: AGHT+IGosKIoAKeyr/SG8ND6M+mWt52v5RM08qZWQV+Ah7+Pc16NBy99FOItw+U5mA3+iLTD73K+eFUwb0dExaCk+w== X-Received: from pgbdo7.prod.google.com ([2002:a05:6a02:e87:b0:af3:27c:5603]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6a20:7483:b0:1ee:d317:48e9 with SMTP id adf61e73a8af0-1f544ad8152mr15572734637.6.1741470048718; Sat, 08 Mar 2025 13:40:48 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:37 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-2-almasrymina@google.com> Subject: [PATCH net-next v7 1/9] netmem: add niov->type attribute to distinguish different net_iov types From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Later patches in the series adds TX net_iovs where there is no pp associated, so we can't rely on niov->pp->mp_ops to tell what is the type of the net_iov. Add a type enum to the net_iov which tells us the net_iov type. Signed-off-by: Mina Almasry Tested-by: Lei Yang --- v7: - New patch --- include/net/netmem.h | 15 ++++++++++++++- net/core/devmem.c | 3 ++- 2 files changed, 16 insertions(+), 2 deletions(-) diff --git a/include/net/netmem.h b/include/net/netmem.h index c61d5b21e7b4..16ef53ea713a 100644 --- a/include/net/netmem.h +++ b/include/net/netmem.h @@ -20,8 +20,21 @@ DECLARE_STATIC_KEY_FALSE(page_pool_mem_providers); */ #define NET_IOV 0x01UL =20 +enum net_iov_type { + /* The unspecified type is temporary until the io_uring net_iovs being + * worked on in parallel are migrated to specify their type, then this + * can be deprecated. + */ + NET_IOV_UNSPECIFIED =3D 0, + NET_IOV_DMABUF, + + /* Force size to unsigned long to make the NET_IOV_ASSERTS below pass. + */ + NET_IOV_MAX =3D ULONG_MAX, +}; + struct net_iov { - unsigned long __unused_padding; + enum net_iov_type type; unsigned long pp_magic; struct page_pool *pp; struct net_iov_area *owner; diff --git a/net/core/devmem.c b/net/core/devmem.c index 7c6e0b5b6acb..69c160ad3ebd 100644 --- a/net/core/devmem.c +++ b/net/core/devmem.c @@ -32,7 +32,7 @@ static const struct memory_provider_ops dmabuf_devmem_ops; =20 bool net_is_devmem_iov(struct net_iov *niov) { - return niov->pp->mp_ops =3D=3D &dmabuf_devmem_ops; + return niov->type =3D=3D NET_IOV_DMABUF; } =20 static void net_devmem_dmabuf_free_chunk_owner(struct gen_pool *genpool, @@ -297,6 +297,7 @@ net_devmem_bind_dmabuf(struct net_device *dev, unsigned= int dmabuf_fd, =20 for (i =3D 0; i < owner->area.num_niovs; i++) { niov =3D &owner->area.niovs[i]; + niov->type =3D NET_IOV_DMABUF; niov->owner =3D &owner->area; page_pool_set_dma_addr_netmem(net_iov_to_netmem(niov), net_devmem_get_dma_addr(niov)); --=20 2.49.0.rc0.332.g42c0ae87b1-goog From nobody Sun Feb 8 02:41:22 2026 Received: from mail-pj1-f74.google.com (mail-pj1-f74.google.com [209.85.216.74]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id DFC2620E020 for ; Sat, 8 Mar 2025 21:40:50 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.74 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470052; cv=none; b=ZisQPC1yxOqBsnXaJkJIIrk60VHKMNm/6/4t7dknWIEW5h2DRVxYIf5/mZZJ4ROq2F2QMuHW3Ox88I/WoE9d0r4sguSj4mPDEPNETM8rqohUJhAvJwdQzRRvNZZ3HJPqhQSnXU4Xd3lBu5UjDjm7/mooOSmGMoqjOqcPqNcW8Rc= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470052; c=relaxed/simple; bh=7pQTGJiG+gkCckiu7VrEfaZJUM/GgV9bYtRCfpuAGVw=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=nXftwUOzkHuLRfH/mcz4IO3q3qD8YTphQvmSYSQgyAPKB3RzoHqG0JqH3Q1JeNb9PapO6FBVRO+LU5eBSjAvJoeJZLBNaV3BR9Wa2jOLm9sU1NpXxUUgiZHnvjsCpIDg6aUvoBcffXhXKvT6k3LfThKbz3oPI10C/9dEGzJMBus= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=Imy0V8nN; arc=none smtp.client-ip=209.85.216.74 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="Imy0V8nN" Received: by mail-pj1-f74.google.com with SMTP id 98e67ed59e1d1-2ff78bd3026so3989451a91.1 for ; Sat, 08 Mar 2025 13:40:50 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470050; x=1742074850; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=Uazn3Yur7EC+BlkBfnfCt5gd1JcQM6gP6YoaVvOEWgU=; b=Imy0V8nN5nGHdpiuYNrH9ia+JjrNGhnHfep+pjnJSZCsZ5MF1Z92ElyxgrMmbc+mYD A61kAt96y3fWt7K7FjcYnw/Hzl51cMSXp56wa8sLW/lGs/1JKilsRqw6QrTYxnx5y7uU pKbJpoh5uaNln+ySD6FRylaAlAZBbwXs6VpPF1oxoOOnR8pUZyaJei/i1HTZfuBZ751U HBwF3Jzorq8B+p3cmmmf6maRhjHgSxynb2agQ++hB+tZVBZlUzRZI0kkOANEorcs9xQq RDGDc2Vy/LhEE5Eb35uIqtG6WDukvVbMPpgV+m21iObKcBXhMkmT1wEemvVjpDSQ308N bGYg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470050; x=1742074850; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=Uazn3Yur7EC+BlkBfnfCt5gd1JcQM6gP6YoaVvOEWgU=; b=eDrb0xRtsddGBRpOCR2wfgELnJusOUh5STQonkTFBgQLwygxcsWHLe5oEoqOmqjhIM vZhu+bVZmMc6aEmiINJMi6mAJFGiS31NEbVtIunNnDKX/NfXUn7ltAdxjFCng20tdrLd g/ZGPr7sU4zm+UTBtNm1MHnT3UUFosDmKqNW3jbKV/htb6HaYokRQUMXYfQGgp7S0R85 uIUpbYzob8LhtpXBJyJLB2Tp+M/Dvqie5rF2aErhijjDeHOv1NygUroYuGhV+ldhsY7B 1bnCjHyR/tJylUYppLaPvPHW9B/o7K8bWQqTC+mywlXKHbVpdQQrC4O1GZaeivzSQjGd 6ukw== X-Forwarded-Encrypted: i=1; AJvYcCWMTVhUAbPcN5pKDtg7RnY7jJ+s19R5lWsm6PK4tg6AjlYj4roYHIDbJvXy78hVxc7UXnNae/kTq9tmqrc=@vger.kernel.org X-Gm-Message-State: AOJu0YwI4EXxG9uP187G+Yl6SeG3l22cJHu6h3bmi46A4mgi9ybjSi/4 1/bxj/RIWX5gUhieXobdUQF4xS0+9euBVG0Bu6lPf+d8DEiG6lpprljEaHLxTJ4k5UTa7p+HJiW nlBU4Wlll8NPM6DXjfWG3sg== X-Google-Smtp-Source: AGHT+IGXS8/G04rAqU1dTYOAUKMkFt2IW1nAEIfrdWh+khmoOhCAntkqTN5Oyi6ICyS0xlqCnZiOGrQcYYJWSHl5SA== X-Received: from pjbse15.prod.google.com ([2002:a17:90b:518f:b0:2fc:201d:6026]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90a:d605:b0:2ff:4a8d:74f8 with SMTP id 98e67ed59e1d1-2ff7ce61dbcmr12417520a91.6.1741470050325; Sat, 08 Mar 2025 13:40:50 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:38 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-3-almasrymina@google.com> Subject: [PATCH net-next v7 2/9] net: add get_netmem/put_netmem support From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Currently net_iovs support only pp ref counts, and do not support a page ref equivalent. This is fine for the RX path as net_iovs are used exclusively with the pp and only pp refcounting is needed there. The TX path however does not use pp ref counts, thus, support for get_page/put_page equivalent is needed for netmem. Support get_netmem/put_netmem. Check the type of the netmem before passing it to page or net_iov specific code to obtain a page ref equivalent. For dmabuf net_iovs, we obtain a ref on the underlying binding. This ensures the entire binding doesn't disappear until all the net_iovs have been put_netmem'ed. We do not need to track the refcount of individual dmabuf net_iovs as we don't allocate/free them from a pool similar to what the buddy allocator does for pages. This code is written to be extensible by other net_iov implementers. get_netmem/put_netmem will check the type of the netmem and route it to the correct helper: pages -> [get|put]_page() dmabuf net_iovs -> net_devmem_[get|put]_net_iov() new net_iovs -> new helpers Signed-off-by: Mina Almasry Acked-by: Stanislav Fomichev Tested-by: Lei Yang --- v5: https://lore.kernel.org/netdev/20250227041209.2031104-2-almasrymina@goo= gle.com/ - Updated to check that the net_iov is devmem before calling net_devmem_put_net_iov(). - Jakub requested that callers of __skb_frag_ref()/skb_page_unref be inspected to make sure that they generate / anticipate skbs with the correct pp_recycle and unreadable setting: skb_page_unref =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D - callers that are unreachable for unreadable skbs: gro_pull_from_frag0, skb_copy_ubufs, __pskb_pull_tail - callers that are reachable for unreadable skbs. These would only see rx unreadable skbs with pp_recycle set before this patchset and would drop a pp ref count. After this patchset they can see tx unreadable skbs with no pp attached and no pp_recycle set, and so now they will drop a net_iov ref via put_netmem: __pskb_trim, __pskb_trim_head, skb_release_data, skb_shift __skb_frag_ref =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D Before this patchset __skb_frag_ref would not do the right thing if it saw any unreadable skbs, either with pp_recycle set or not. Because it unconditionally tries to acquire a page ref, but with RX only support I can't reproduce calls to __skb_frag_ref even after enabling tc forwarding to TX. After this patchset __skb_frag_ref would obtain a page ref equivalent on dmabuf net_iovs, by obtaining a ref on the binding. Callers that are unreachable for unreadable skbs: - veth_xdp_get Callers that are reachable for unreadable skbs, and from code review they look specific to the TX path: - tcp_grow_skb, __skb_zcopy_downgrade_managed, __pskb_copy_fclone, pskb_expand_head, skb_zerocopy, skb_split, pksb_carve_inside_header, pskb_care_inside_nonlinear, tcp_clone_payload, skb_segment. Callers that are reachable for unreadable skbs, and from code review they look reachable in the RX path, although my testing never hit these paths. These are concerning. Maybe we should put this patch in net and cc stable? However, no drivers currently enable unreadable netmem, so fixing this in net-next is fine as well maybe: - skb_shift, skb_try_coalesce v2: - Add comment on top of refcount_t ref explaining the usage in the XT path. - Fix missing definition of net_devmem_dmabuf_binding_put in this patch. --- include/linux/skbuff_ref.h | 4 ++-- include/net/netmem.h | 3 +++ net/core/devmem.c | 10 ++++++++++ net/core/devmem.h | 20 ++++++++++++++++++++ net/core/skbuff.c | 30 ++++++++++++++++++++++++++++++ 5 files changed, 65 insertions(+), 2 deletions(-) diff --git a/include/linux/skbuff_ref.h b/include/linux/skbuff_ref.h index 0f3c58007488..9e49372ef1a0 100644 --- a/include/linux/skbuff_ref.h +++ b/include/linux/skbuff_ref.h @@ -17,7 +17,7 @@ */ static inline void __skb_frag_ref(skb_frag_t *frag) { - get_page(skb_frag_page(frag)); + get_netmem(skb_frag_netmem(frag)); } =20 /** @@ -40,7 +40,7 @@ static inline void skb_page_unref(netmem_ref netmem, bool= recycle) if (recycle && napi_pp_put_page(netmem)) return; #endif - put_page(netmem_to_page(netmem)); + put_netmem(netmem); } =20 /** diff --git a/include/net/netmem.h b/include/net/netmem.h index 16ef53ea713a..e8afe6b654aa 100644 --- a/include/net/netmem.h +++ b/include/net/netmem.h @@ -277,4 +277,7 @@ static inline unsigned long netmem_get_dma_addr(netmem_= ref netmem) return __netmem_clear_lsb(netmem)->dma_addr; } =20 +void get_netmem(netmem_ref netmem); +void put_netmem(netmem_ref netmem); + #endif /* _NET_NETMEM_H */ diff --git a/net/core/devmem.c b/net/core/devmem.c index 69c160ad3ebd..0cf3d189f06c 100644 --- a/net/core/devmem.c +++ b/net/core/devmem.c @@ -326,6 +326,16 @@ net_devmem_bind_dmabuf(struct net_device *dev, unsigne= d int dmabuf_fd, return ERR_PTR(err); } =20 +void net_devmem_get_net_iov(struct net_iov *niov) +{ + net_devmem_dmabuf_binding_get(net_devmem_iov_binding(niov)); +} + +void net_devmem_put_net_iov(struct net_iov *niov) +{ + net_devmem_dmabuf_binding_put(net_devmem_iov_binding(niov)); +} + /*** "Dmabuf devmem memory provider" ***/ =20 int mp_dmabuf_devmem_init(struct page_pool *pool) diff --git a/net/core/devmem.h b/net/core/devmem.h index 7fc158d52729..946f2e015746 100644 --- a/net/core/devmem.h +++ b/net/core/devmem.h @@ -29,6 +29,10 @@ struct net_devmem_dmabuf_binding { * The binding undos itself and unmaps the underlying dmabuf once all * those refs are dropped and the binding is no longer desired or in * use. + * + * net_devmem_get_net_iov() on dmabuf net_iovs will increment this + * reference, making sure that the binding remains alive until all the + * net_iovs are no longer used. */ refcount_t ref; =20 @@ -111,6 +115,9 @@ net_devmem_dmabuf_binding_put(struct net_devmem_dmabuf_= binding *binding) __net_devmem_dmabuf_binding_free(binding); } =20 +void net_devmem_get_net_iov(struct net_iov *niov); +void net_devmem_put_net_iov(struct net_iov *niov); + struct net_iov * net_devmem_alloc_dmabuf(struct net_devmem_dmabuf_binding *binding); void net_devmem_free_dmabuf(struct net_iov *ppiov); @@ -120,6 +127,19 @@ bool net_is_devmem_iov(struct net_iov *niov); #else struct net_devmem_dmabuf_binding; =20 +static inline void +net_devmem_dmabuf_binding_put(struct net_devmem_dmabuf_binding *binding) +{ +} + +static inline void net_devmem_get_net_iov(struct net_iov *niov) +{ +} + +static inline void net_devmem_put_net_iov(struct net_iov *niov) +{ +} + static inline void __net_devmem_dmabuf_binding_free(struct net_devmem_dmabuf_binding *binding) { diff --git a/net/core/skbuff.c b/net/core/skbuff.c index ab8acb737b93..ee2d1b769c13 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -89,6 +89,7 @@ #include =20 #include "dev.h" +#include "devmem.h" #include "netmem_priv.h" #include "sock_destructor.h" =20 @@ -7315,3 +7316,32 @@ bool csum_and_copy_from_iter_full(void *addr, size_t= bytes, return false; } EXPORT_SYMBOL(csum_and_copy_from_iter_full); + +void get_netmem(netmem_ref netmem) +{ + struct net_iov *niov; + + if (netmem_is_net_iov(netmem)) { + niov =3D netmem_to_net_iov(netmem); + if (net_is_devmem_iov(niov)) + net_devmem_get_net_iov(netmem_to_net_iov(netmem)); + return; + } + get_page(netmem_to_page(netmem)); +} +EXPORT_SYMBOL(get_netmem); + +void put_netmem(netmem_ref netmem) +{ + struct net_iov *niov; + + if (netmem_is_net_iov(netmem)) { + niov =3D netmem_to_net_iov(netmem); + if (net_is_devmem_iov(niov)) + net_devmem_put_net_iov(netmem_to_net_iov(netmem)); + return; + } + + put_page(netmem_to_page(netmem)); +} +EXPORT_SYMBOL(put_netmem); --=20 2.49.0.rc0.332.g42c0ae87b1-goog From nobody Sun Feb 8 02:41:22 2026 Received: from mail-pl1-f201.google.com (mail-pl1-f201.google.com [209.85.214.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 9F35F2135AB for ; Sat, 8 Mar 2025 21:40:52 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.214.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470054; cv=none; b=V/4qwTYC7NGm2hTrq9V/SmXYNTEQtd1r35oP/nMKoV+70ZxlFb+sfdtyWp1xzvLPaT9mWy3EAtAjLfWYIpkfEEzCCVZ+4go3zbZ0OSD35zVWM60/DFNHpQLweAHz5+czeSlI3z3YacT/e77aicyZpYdq0yH1huX+8M1TzmqixZQ= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470054; c=relaxed/simple; bh=GgKXq6iE0gzrYkfJCs/+4j6Jqrx0/CCwuPwPGpNWSe8=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=hIJuuZtIpD7J71k9xwo+tMJKQL8jDA6qXPdMqXmIEcc/vZaeTtuAbtJzyTyEZkIJctSSPmNA40LxD4VxrRTvDmAL/wzL7XFU0jZeJYPOrFzDL/eJrH51XNeAusOHc490qJPLEKXdrao1W1T33bXba64SCeN1a22vMWUKBl8HClU= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=Z5v8mFbp; arc=none smtp.client-ip=209.85.214.201 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="Z5v8mFbp" Received: by mail-pl1-f201.google.com with SMTP id d9443c01a7336-223d86b4df0so85937265ad.3 for ; Sat, 08 Mar 2025 13:40:52 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470052; x=1742074852; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=SVSp6KkQQ3nBS0Irfd9WWWqAtqD5umJDSeIajOkZRLA=; b=Z5v8mFbp/kkxI2zVrvfAxZPvKp44b0YwOyE9Cbd1rbsqWXOfnzSbKRj3iJ+vPZsDZU KHfybUPXUktFlEPuK0+R3tVrYKkNUTqFaFUepHDMdSJWbD+7zkUhf97Reoaaef5PL/aS +JtOPhRahYme9wvfYVj2754avein22j4ki5r1MvULctlQz9DzKUISCTdk1fSh+mHjE+4 7e3szM3qLSFiC/l1GOAWbgNmy99rzwYlNRasaDVn/jemN98pBWUNQPKvVjBWLwlFfpvh bHLEi8gXug9x7+GVkqQUEQuoR+qnet9Jy/ICqDBpuamdGUoYxzm3KGk+F9dDtTWa2MTP pAzA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470052; x=1742074852; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=SVSp6KkQQ3nBS0Irfd9WWWqAtqD5umJDSeIajOkZRLA=; b=WNXY8jWc1iikdrkkMkow3zFyrTaLNFZkIvdA1MT9cKiaXrebCkZb1n9T58VbHzOXUM M8p+tDWnZtLnHrxgI8TUM3wyQJGtENej7NwtQUP54WH9pCrgt5YXZ9YGkkxenDJgoafO kLpYeimUsnWySZVvQ+fhuNxMaN0xQHDwczqTNw4cM8bxwQVS80IVxFzN4XR6gVnsJYov lnAGzhkZ1UUiuUAK0szfUN++7cO16p/GRkki3OvNMwloa9tp6DRPFGUfe+fXSk+Op/58 yU5Y5BJLadSbdkwJPOuhcN4735UEz152e8bvdDTKk4oWVeCTtzxM69mHbIKKTnJRh4+O gtIw== X-Forwarded-Encrypted: i=1; AJvYcCWhqjJCMOxHNwBpfEstKOtkAT+FGCaM8C5Fk4wCUCt38vK4HyP0q8WovZkUkXRkbM6ATQPRkp/uh4cFcGg=@vger.kernel.org X-Gm-Message-State: AOJu0Yxfv21/TQ7QM0z6CAyiOo1LmFbH1vdfmXVFUbsHjgokeXD2u29J /z97TI0VGlC2PfKexSZ79pQ80a0IIhBPt1bzpvvp6wXjYUyfqQaVcAvm6fyasi77R4TBJ4C/CWx Di6XNQrUw4U1WiDZcNcdfSg== X-Google-Smtp-Source: AGHT+IEJxGXdpquM70KYwrulbd57ZH/XyG67NkpllY5yK/S0TKLlEOlhc3K4p6RYqVYqjeOM/HYNoBDPLGSNEiADOw== X-Received: from plbml14.prod.google.com ([2002:a17:903:34ce:b0:223:6c8e:eaf1]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a17:902:ce8a:b0:220:faa2:c911 with SMTP id d9443c01a7336-2242888a696mr143314875ad.14.1741470052036; Sat, 08 Mar 2025 13:40:52 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:39 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-4-almasrymina@google.com> Subject: [PATCH net-next v7 3/9] net: devmem: TCP tx netlink api From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" From: Stanislav Fomichev Add bind-tx netlink call to attach dmabuf for TX; queue is not required, only ifindex and dmabuf fd for attachment. Signed-off-by: Stanislav Fomichev Signed-off-by: Mina Almasry Tested-by: Lei Yang --- v3: - Fix ynl-regen.sh error (Simon). --- Documentation/netlink/specs/netdev.yaml | 12 ++++++++++++ include/uapi/linux/netdev.h | 1 + net/core/netdev-genl-gen.c | 13 +++++++++++++ net/core/netdev-genl-gen.h | 1 + net/core/netdev-genl.c | 6 ++++++ tools/include/uapi/linux/netdev.h | 1 + 6 files changed, 34 insertions(+) diff --git a/Documentation/netlink/specs/netdev.yaml b/Documentation/netlin= k/specs/netdev.yaml index 36f1152bfac3..e560b05eb528 100644 --- a/Documentation/netlink/specs/netdev.yaml +++ b/Documentation/netlink/specs/netdev.yaml @@ -743,6 +743,18 @@ operations: - defer-hard-irqs - gro-flush-timeout - irq-suspend-timeout + - + name: bind-tx + doc: Bind dmabuf to netdev for TX + attribute-set: dmabuf + do: + request: + attributes: + - ifindex + - fd + reply: + attributes: + - id =20 kernel-family: headers: [ "linux/list.h"] diff --git a/include/uapi/linux/netdev.h b/include/uapi/linux/netdev.h index 7600bf62dbdf..7eb9571786b8 100644 --- a/include/uapi/linux/netdev.h +++ b/include/uapi/linux/netdev.h @@ -219,6 +219,7 @@ enum { NETDEV_CMD_QSTATS_GET, NETDEV_CMD_BIND_RX, NETDEV_CMD_NAPI_SET, + NETDEV_CMD_BIND_TX, =20 __NETDEV_CMD_MAX, NETDEV_CMD_MAX =3D (__NETDEV_CMD_MAX - 1) diff --git a/net/core/netdev-genl-gen.c b/net/core/netdev-genl-gen.c index 996ac6a449eb..f27608d6301c 100644 --- a/net/core/netdev-genl-gen.c +++ b/net/core/netdev-genl-gen.c @@ -99,6 +99,12 @@ static const struct nla_policy netdev_napi_set_nl_policy= [NETDEV_A_NAPI_IRQ_SUSPE [NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT] =3D { .type =3D NLA_UINT, }, }; =20 +/* NETDEV_CMD_BIND_TX - do */ +static const struct nla_policy netdev_bind_tx_nl_policy[NETDEV_A_DMABUF_FD= + 1] =3D { + [NETDEV_A_DMABUF_IFINDEX] =3D NLA_POLICY_MIN(NLA_U32, 1), + [NETDEV_A_DMABUF_FD] =3D { .type =3D NLA_U32, }, +}; + /* Ops table for netdev */ static const struct genl_split_ops netdev_nl_ops[] =3D { { @@ -190,6 +196,13 @@ static const struct genl_split_ops netdev_nl_ops[] =3D= { .maxattr =3D NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT, .flags =3D GENL_ADMIN_PERM | GENL_CMD_CAP_DO, }, + { + .cmd =3D NETDEV_CMD_BIND_TX, + .doit =3D netdev_nl_bind_tx_doit, + .policy =3D netdev_bind_tx_nl_policy, + .maxattr =3D NETDEV_A_DMABUF_FD, + .flags =3D GENL_CMD_CAP_DO, + }, }; =20 static const struct genl_multicast_group netdev_nl_mcgrps[] =3D { diff --git a/net/core/netdev-genl-gen.h b/net/core/netdev-genl-gen.h index e09dd7539ff2..c1fed66e92b9 100644 --- a/net/core/netdev-genl-gen.h +++ b/net/core/netdev-genl-gen.h @@ -34,6 +34,7 @@ int netdev_nl_qstats_get_dumpit(struct sk_buff *skb, struct netlink_callback *cb); int netdev_nl_bind_rx_doit(struct sk_buff *skb, struct genl_info *info); int netdev_nl_napi_set_doit(struct sk_buff *skb, struct genl_info *info); +int netdev_nl_bind_tx_doit(struct sk_buff *skb, struct genl_info *info); =20 enum { NETDEV_NLGRP_MGMT, diff --git a/net/core/netdev-genl.c b/net/core/netdev-genl.c index 2b774183d31c..6e5f2de4d947 100644 --- a/net/core/netdev-genl.c +++ b/net/core/netdev-genl.c @@ -931,6 +931,12 @@ int netdev_nl_bind_rx_doit(struct sk_buff *skb, struct= genl_info *info) return err; } =20 +/* stub */ +int netdev_nl_bind_tx_doit(struct sk_buff *skb, struct genl_info *info) +{ + return 0; +} + void netdev_nl_sock_priv_init(struct list_head *priv) { INIT_LIST_HEAD(priv); diff --git a/tools/include/uapi/linux/netdev.h b/tools/include/uapi/linux/n= etdev.h index 7600bf62dbdf..7eb9571786b8 100644 --- a/tools/include/uapi/linux/netdev.h +++ b/tools/include/uapi/linux/netdev.h @@ -219,6 +219,7 @@ enum { NETDEV_CMD_QSTATS_GET, NETDEV_CMD_BIND_RX, NETDEV_CMD_NAPI_SET, + NETDEV_CMD_BIND_TX, =20 __NETDEV_CMD_MAX, NETDEV_CMD_MAX =3D (__NETDEV_CMD_MAX - 1) --=20 2.49.0.rc0.332.g42c0ae87b1-goog From nobody Sun Feb 8 02:41:22 2026 Received: from mail-oa1-f73.google.com (mail-oa1-f73.google.com [209.85.160.73]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 9C62C215049 for ; Sat, 8 Mar 2025 21:40:54 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.160.73 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470058; cv=none; b=sazrmlcnLmE1afZb5zxIP+yhngqSWI4SrQMZP5Nzr6brPOpCK9GN4shK6gp/K+NbFAwgfSsEBI2QZZN/LX5pSuoBH5lKzDiX6LkdNBRX0JkxE+E7+8+VM297gpWMv1oLfg319M4ZCX+pfzz6+zXHwVAHxjZbedyZ7FUMea6eG14= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470058; c=relaxed/simple; bh=WPIVC4Re7yOqFf2/08Wl8WqS1gtP1WRrTzg2/LAljDk=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=fiIJUjarDpELBhNLht6Kg6H07oP4HmTPx0/YXJn8pw3dwMytrgJQjjcr0s4Oqg/JLHN4QQy4nJw98EJfMTnLSmhwkcq7ms6zYpKgn1fu11fSQ1eqCM6bLfdW1GuDBG6Iqf8qtU2+OG8sZKYNfMfHTMcFvPEmw8n7pp/eYfne9Rg= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=BppSZlYk; arc=none smtp.client-ip=209.85.160.73 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="BppSZlYk" Received: by mail-oa1-f73.google.com with SMTP id 586e51a60fabf-2c24c595a08so2243154fac.3 for ; Sat, 08 Mar 2025 13:40:54 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470054; x=1742074854; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=ndj53jSIP+x/kWUcZQo0fEgG5nfL08JNBgTFK122S/c=; b=BppSZlYk77g1WwPZMpEoTQMLJt+stmVdNMjHdjkQni3zsW7N4ehCyrMiqlgzyD/o1y MJt6DXL714y9jF/yJ5z/2FrtHGfk24UwajaVGnaYQffRcmTgQLdcEZs8KswIfBtA7etu qinQ4wbz0CxsW+kLzurCGc3Wjo/XE4/+wadC1pLb4ixWSSbUypmR8JhRMAAIWWzs2Bvh 2QMS2rnsxFrHPuMEg2EZh0NcBpng7CRhlo8jP0EUhw0JDckFGAjFlKXtah/yuVd62a9z qT3rfxc9xAQGqrGQHt4OI8y+nOHL4D+YqN1NUZUI/vxipSEOo7rnu+rx0gawSkp/K5Bf EKdQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470054; x=1742074854; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=ndj53jSIP+x/kWUcZQo0fEgG5nfL08JNBgTFK122S/c=; b=hNJcQkNmAU1USQNB6NWeLyZcXru9vqYRsS6x2uSPqW6VVmilyAO4bUjh/Ldh0bvZqa j+wQwq+lsCbf/8nFc0uIJtYD6hXUdE2dgAN8azIXMdHkgZZOiYMA3Ki2fIXThl5wjWMg KS+GG0XSsz1dN6pFvj683MKqfFdRQS4hcrnXXROzccj/qUS7hnKs9f6QarlVxLCn/tU0 BWW2OrsDRmX3be6BFO/5IfBTtFR+n3R0cyYV789EHgv8whO+9aUsp/G5ngKkwqs5b56Z JWpMbE7mmX8ifayIU1jNw+RFy7uE7RcFAukqHaksRMSxAwZGebBEtAY8ibXqv+o/okCf G4/A== X-Forwarded-Encrypted: i=1; AJvYcCW8JAp/tu2d4udGNnVayQuizy+zn+c0cSodrHUo6PcNbntC2EG1zZB1yas+r5ZhbZj+U7XQabssOEJ5YoQ=@vger.kernel.org X-Gm-Message-State: AOJu0YyQl67pPq+JGdvF6vBbk9Ftpf3w3/cK8qohpFfL0ftT6LPv+RYF wDvemgjKOlBKED0MPmwPB/wvhHV8qVW5t9ihxjrOWBR0esGKkiZqwBvGZO3gwJoM9ofuf/76UfX KMI/w5mF5KS5BDbLrac+vPg== X-Google-Smtp-Source: AGHT+IFaPY21IsOCKlYimnhxsj/Lv1F/e6YTN7dz4EZoR9Fw7n9v6+i05c8SEIDl0E3IX8fGnd79FV2q83q+R6CeXw== X-Received: from oabgc15.prod.google.com ([2002:a05:6870:678f:b0:2c1:5dcb:42ac]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6871:5226:b0:296:dd30:e90f with SMTP id 586e51a60fabf-2c26102d931mr4022050fac.16.1741470053735; Sat, 08 Mar 2025 13:40:53 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:40 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-5-almasrymina@google.com> Subject: [PATCH net-next v7 4/9] net: devmem: Implement TX path From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja , Kaiyuan Zhang Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Augment dmabuf binding to be able to handle TX. Additional to all the RX binding, we also create tx_vec needed for the TX path. Provide API for sendmsg to be able to send dmabufs bound to this device: - Provide a new dmabuf_tx_cmsg which includes the dmabuf to send from. - MSG_ZEROCOPY with SCM_DEVMEM_DMABUF cmsg indicates send from dma-buf. Devmem is uncopyable, so piggyback off the existing MSG_ZEROCOPY implementation, while disabling instances where MSG_ZEROCOPY falls back to copying. We additionally pipe the binding down to the new zerocopy_fill_skb_from_devmem which fills a TX skb with net_iov netmems instead of the traditional page netmems. We also special case skb_frag_dma_map to return the dma-address of these dmabuf net_iovs instead of attempting to map pages. The TX path may release the dmabuf in a context where we cannot wait. This happens when the user unbinds a TX dmabuf while there are still references to its netmems in the TX path. In that case, the netmems will be put_netmem'd from a context where we can't unmap the dmabuf, Resolve this by making __net_devmem_dmabuf_binding_free schedule_work'd. Based on work by Stanislav Fomichev . A lot of the meat of the implementation came from devmem TCP RFC v1[1], which included the TX path, but Stan did all the rebasing on top of netmem/net_iov. Cc: Stanislav Fomichev Signed-off-by: Kaiyuan Zhang Signed-off-by: Mina Almasry Acked-by: Stanislav Fomichev Tested-by: Lei Yang --- v6: - Retain behavior that MSG_FASTOPEN succeeds even if cmsg is invalid (Paolo). - Rework the freeing of tx_vec slightly to improve readability. Now it has its own err label (Paolo). - Squash making unbinding scheduled work (Paolo). - Add comment to clarify that net_iovs stuck in the transmit path hold a ref on the underlying dmabuf binding (David). - Fix the comment on how binding refcounting works on RX (the comment was not matching the existing code behavior). v5: - Return -EFAULT from zerocopy_fill_skb_from_devmem (Stan) - don't null check before kvfree (stan). v4: - Remove dmabuf_tx_cmsg definition and just use __u32 for the dma-buf id (Willem). - Check that iov_iter_type() is ITER_IOVEC in zerocopy_fill_skb_from_iter() (Pavel). - Fix binding->tx_vec not being freed on error paths (Paolo). - Make devmem patch mutually exclusive with msg->ubuf_info path (Pavel). - Check that MSG_ZEROCOPY and SOCK_ZEROCOPY are provided when sockc.dmabuf_id is provided. - Don't mm_account_pinned_pages() on devmem TX (Pavel). v3: - Use kvmalloc_array instead of kcalloc (Stan). - Fix unreachable code warning (Simon). v2: - Remove dmabuf_offset from the dmabuf cmsg. - Update zerocopy_fill_skb_from_devmem to interpret the iov_base/iter_iov_addr as the offset into the dmabuf to send from (Stan). - Remove the confusing binding->tx_iter which is not needed if we interpret the iov_base/iter_iov_addr as offset into the dmabuf (Stan). - Remove check for binding->sgt and binding->sgt->nents in dmabuf binding. - Simplify the calculation of binding->tx_vec. - Check in net_devmem_get_binding that the binding we're returning has ifindex matching the sending socket (Willem). --- include/linux/skbuff.h | 17 +++- include/net/sock.h | 1 + net/core/datagram.c | 48 ++++++++++- net/core/devmem.c | 105 ++++++++++++++++++++++-- net/core/devmem.h | 61 +++++++++++--- net/core/netdev-genl.c | 64 ++++++++++++++- net/core/skbuff.c | 18 ++-- net/core/sock.c | 6 ++ net/ipv4/ip_output.c | 3 +- net/ipv4/tcp.c | 50 ++++++++--- net/ipv6/ip6_output.c | 3 +- net/vmw_vsock/virtio_transport_common.c | 5 +- 12 files changed, 330 insertions(+), 51 deletions(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 14517e95a46c..67a7e069a9bf 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1707,13 +1707,16 @@ static inline void skb_set_end_offset(struct sk_buf= f *skb, unsigned int offset) extern const struct ubuf_info_ops msg_zerocopy_ubuf_ops; =20 struct ubuf_info *msg_zerocopy_realloc(struct sock *sk, size_t size, - struct ubuf_info *uarg); + struct ubuf_info *uarg, bool devmem); =20 void msg_zerocopy_put_abort(struct ubuf_info *uarg, bool have_uref); =20 +struct net_devmem_dmabuf_binding; + int __zerocopy_sg_from_iter(struct msghdr *msg, struct sock *sk, struct sk_buff *skb, struct iov_iter *from, - size_t length); + size_t length, + struct net_devmem_dmabuf_binding *binding); =20 int zerocopy_fill_skb_from_iter(struct sk_buff *skb, struct iov_iter *from, size_t length); @@ -1721,12 +1724,14 @@ int zerocopy_fill_skb_from_iter(struct sk_buff *skb, static inline int skb_zerocopy_iter_dgram(struct sk_buff *skb, struct msghdr *msg, int len) { - return __zerocopy_sg_from_iter(msg, skb->sk, skb, &msg->msg_iter, len); + return __zerocopy_sg_from_iter(msg, skb->sk, skb, &msg->msg_iter, len, + NULL); } =20 int skb_zerocopy_iter_stream(struct sock *sk, struct sk_buff *skb, struct msghdr *msg, int len, - struct ubuf_info *uarg); + struct ubuf_info *uarg, + struct net_devmem_dmabuf_binding *binding); =20 /* Internal */ #define skb_shinfo(SKB) ((struct skb_shared_info *)(skb_end_pointer(SKB))) @@ -3697,6 +3702,10 @@ static inline dma_addr_t __skb_frag_dma_map(struct d= evice *dev, size_t offset, size_t size, enum dma_data_direction dir) { + if (skb_frag_is_net_iov(frag)) { + return netmem_to_net_iov(frag->netmem)->dma_addr + offset + + frag->offset; + } return dma_map_page(dev, skb_frag_page(frag), skb_frag_off(frag) + offset, size, dir); } diff --git a/include/net/sock.h b/include/net/sock.h index 8daf1b3b12c6..59875bed75e7 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1816,6 +1816,7 @@ struct sockcm_cookie { u32 tsflags; u32 ts_opt_id; u32 priority; + u32 dmabuf_id; }; =20 static inline void sockcm_init(struct sockcm_cookie *sockc, diff --git a/net/core/datagram.c b/net/core/datagram.c index f0693707aece..09c74a1d836b 100644 --- a/net/core/datagram.c +++ b/net/core/datagram.c @@ -63,6 +63,8 @@ #include #include =20 +#include "devmem.h" + /* * Is a socket 'connection oriented' ? */ @@ -692,9 +694,49 @@ int zerocopy_fill_skb_from_iter(struct sk_buff *skb, return 0; } =20 +static int +zerocopy_fill_skb_from_devmem(struct sk_buff *skb, struct iov_iter *from, + int length, + struct net_devmem_dmabuf_binding *binding) +{ + int i =3D skb_shinfo(skb)->nr_frags; + size_t virt_addr, size, off; + struct net_iov *niov; + + /* Devmem filling works by taking an IOVEC from the user where the + * iov_addrs are interpreted as an offset in bytes into the dma-buf to + * send from. We do not support other iter types. + */ + if (iov_iter_type(from) !=3D ITER_IOVEC) + return -EFAULT; + + while (length && iov_iter_count(from)) { + if (i =3D=3D MAX_SKB_FRAGS) + return -EMSGSIZE; + + virt_addr =3D (size_t)iter_iov_addr(from); + niov =3D net_devmem_get_niov_at(binding, virt_addr, &off, &size); + if (!niov) + return -EFAULT; + + size =3D min_t(size_t, size, length); + size =3D min_t(size_t, size, iter_iov_len(from)); + + get_netmem(net_iov_to_netmem(niov)); + skb_add_rx_frag_netmem(skb, i, net_iov_to_netmem(niov), off, + size, PAGE_SIZE); + iov_iter_advance(from, size); + length -=3D size; + i++; + } + + return 0; +} + int __zerocopy_sg_from_iter(struct msghdr *msg, struct sock *sk, struct sk_buff *skb, struct iov_iter *from, - size_t length) + size_t length, + struct net_devmem_dmabuf_binding *binding) { unsigned long orig_size =3D skb->truesize; unsigned long truesize; @@ -702,6 +744,8 @@ int __zerocopy_sg_from_iter(struct msghdr *msg, struct = sock *sk, =20 if (msg && msg->msg_ubuf && msg->sg_from_iter) ret =3D msg->sg_from_iter(skb, from, length); + else if (unlikely(binding)) + ret =3D zerocopy_fill_skb_from_devmem(skb, from, length, binding); else ret =3D zerocopy_fill_skb_from_iter(skb, from, length); =20 @@ -735,7 +779,7 @@ int zerocopy_sg_from_iter(struct sk_buff *skb, struct i= ov_iter *from) if (skb_copy_datagram_from_iter(skb, 0, from, copy)) return -EFAULT; =20 - return __zerocopy_sg_from_iter(NULL, NULL, skb, from, ~0U); + return __zerocopy_sg_from_iter(NULL, NULL, skb, from, ~0U, NULL); } EXPORT_SYMBOL(zerocopy_sg_from_iter); =20 diff --git a/net/core/devmem.c b/net/core/devmem.c index 0cf3d189f06c..393e30d72dc8 100644 --- a/net/core/devmem.c +++ b/net/core/devmem.c @@ -17,6 +17,7 @@ #include #include #include +#include #include =20 #include "devmem.h" @@ -54,8 +55,10 @@ static dma_addr_t net_devmem_get_dma_addr(const struct n= et_iov *niov) ((dma_addr_t)net_iov_idx(niov) << PAGE_SHIFT); } =20 -void __net_devmem_dmabuf_binding_free(struct net_devmem_dmabuf_binding *bi= nding) +void __net_devmem_dmabuf_binding_free(struct work_struct *wq) { + struct net_devmem_dmabuf_binding *binding =3D container_of(wq, typeof(*bi= nding), unbind_w); + size_t size, avail; =20 gen_pool_for_each_chunk(binding->chunk_pool, @@ -73,8 +76,10 @@ void __net_devmem_dmabuf_binding_free(struct net_devmem_= dmabuf_binding *binding) dma_buf_detach(binding->dmabuf, binding->attachment); dma_buf_put(binding->dmabuf); xa_destroy(&binding->bound_rxqs); + kvfree(binding->tx_vec); kfree(binding); } +EXPORT_SYMBOL(__net_devmem_dmabuf_binding_free); =20 struct net_iov * net_devmem_alloc_dmabuf(struct net_devmem_dmabuf_binding *binding) @@ -119,6 +124,13 @@ void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf= _binding *binding) unsigned long xa_idx; unsigned int rxq_idx; =20 + xa_erase(&net_devmem_dmabuf_bindings, binding->id); + + /* Ensure no tx net_devmem_lookup_dmabuf() are in flight after the + * erase. + */ + synchronize_net(); + if (binding->list.next) list_del(&binding->list); =20 @@ -133,8 +145,6 @@ void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf_= binding *binding) WARN_ON(netdev_rx_queue_restart(binding->dev, rxq_idx)); } =20 - xa_erase(&net_devmem_dmabuf_bindings, binding->id); - net_devmem_dmabuf_binding_put(binding); } =20 @@ -197,8 +207,9 @@ int net_devmem_bind_dmabuf_to_queue(struct net_device *= dev, u32 rxq_idx, } =20 struct net_devmem_dmabuf_binding * -net_devmem_bind_dmabuf(struct net_device *dev, unsigned int dmabuf_fd, - struct netlink_ext_ack *extack) +net_devmem_bind_dmabuf(struct net_device *dev, + enum dma_data_direction direction, + unsigned int dmabuf_fd, struct netlink_ext_ack *extack) { struct net_devmem_dmabuf_binding *binding; static u32 id_alloc_next; @@ -241,22 +252,32 @@ net_devmem_bind_dmabuf(struct net_device *dev, unsign= ed int dmabuf_fd, } =20 binding->sgt =3D dma_buf_map_attachment_unlocked(binding->attachment, - DMA_FROM_DEVICE); + direction); if (IS_ERR(binding->sgt)) { err =3D PTR_ERR(binding->sgt); NL_SET_ERR_MSG(extack, "Failed to map dmabuf attachment"); goto err_detach; } =20 + if (direction =3D=3D DMA_TO_DEVICE) { + binding->tx_vec =3D kvmalloc_array(dmabuf->size / PAGE_SIZE, + sizeof(struct net_iov *), + GFP_KERNEL); + if (!binding->tx_vec) { + err =3D -ENOMEM; + goto err_unmap; + } + } + /* For simplicity we expect to make PAGE_SIZE allocations, but the * binding can be much more flexible than that. We may be able to * allocate MTU sized chunks here. Leave that for future work... */ - binding->chunk_pool =3D - gen_pool_create(PAGE_SHIFT, dev_to_node(&dev->dev)); + binding->chunk_pool =3D gen_pool_create(PAGE_SHIFT, + dev_to_node(&dev->dev)); if (!binding->chunk_pool) { err =3D -ENOMEM; - goto err_unmap; + goto err_tx_vec; } =20 virtual =3D 0; @@ -301,6 +322,8 @@ net_devmem_bind_dmabuf(struct net_device *dev, unsigned= int dmabuf_fd, niov->owner =3D &owner->area; page_pool_set_dma_addr_netmem(net_iov_to_netmem(niov), net_devmem_get_dma_addr(niov)); + if (direction =3D=3D DMA_TO_DEVICE) + binding->tx_vec[owner->area.base_virtual / PAGE_SIZE + i] =3D niov; } =20 virtual +=3D len; @@ -312,6 +335,8 @@ net_devmem_bind_dmabuf(struct net_device *dev, unsigned= int dmabuf_fd, gen_pool_for_each_chunk(binding->chunk_pool, net_devmem_dmabuf_free_chunk_owner, NULL); gen_pool_destroy(binding->chunk_pool); +err_tx_vec: + kvfree(binding->tx_vec); err_unmap: dma_buf_unmap_attachment_unlocked(binding->attachment, binding->sgt, DMA_FROM_DEVICE); @@ -326,6 +351,21 @@ net_devmem_bind_dmabuf(struct net_device *dev, unsigne= d int dmabuf_fd, return ERR_PTR(err); } =20 +struct net_devmem_dmabuf_binding *net_devmem_lookup_dmabuf(u32 id) +{ + struct net_devmem_dmabuf_binding *binding; + + rcu_read_lock(); + binding =3D xa_load(&net_devmem_dmabuf_bindings, id); + if (binding) { + if (!net_devmem_dmabuf_binding_get(binding)) + binding =3D NULL; + } + rcu_read_unlock(); + + return binding; +} + void net_devmem_get_net_iov(struct net_iov *niov) { net_devmem_dmabuf_binding_get(net_devmem_iov_binding(niov)); @@ -336,6 +376,53 @@ void net_devmem_put_net_iov(struct net_iov *niov) net_devmem_dmabuf_binding_put(net_devmem_iov_binding(niov)); } =20 +struct net_devmem_dmabuf_binding *net_devmem_get_binding(struct sock *sk, + unsigned int dmabuf_id) +{ + struct net_devmem_dmabuf_binding *binding; + struct dst_entry *dst =3D __sk_dst_get(sk); + int err =3D 0; + + binding =3D net_devmem_lookup_dmabuf(dmabuf_id); + if (!binding || !binding->tx_vec) { + err =3D -EINVAL; + goto out_err; + } + + /* The dma-addrs in this binding are only reachable to the corresponding + * net_device. + */ + if (!dst || !dst->dev || dst->dev->ifindex !=3D binding->dev->ifindex) { + err =3D -ENODEV; + goto out_err; + } + + return binding; + +out_err: + if (binding) + net_devmem_dmabuf_binding_put(binding); + + return ERR_PTR(err); +} + +struct net_iov * +net_devmem_get_niov_at(struct net_devmem_dmabuf_binding *binding, + size_t virt_addr, size_t *off, size_t *size) +{ + size_t idx; + + if (virt_addr >=3D binding->dmabuf->size) + return NULL; + + idx =3D virt_addr / PAGE_SIZE; + + *off =3D virt_addr % PAGE_SIZE; + *size =3D PAGE_SIZE - *off; + + return binding->tx_vec[idx]; +} + /*** "Dmabuf devmem memory provider" ***/ =20 int mp_dmabuf_devmem_init(struct page_pool *pool) diff --git a/net/core/devmem.h b/net/core/devmem.h index 946f2e015746..67168aae5e5b 100644 --- a/net/core/devmem.h +++ b/net/core/devmem.h @@ -23,8 +23,9 @@ struct net_devmem_dmabuf_binding { =20 /* The user holds a ref (via the netlink API) for as long as they want * the binding to remain alive. Each page pool using this binding holds - * a ref to keep the binding alive. Each allocated net_iov holds a - * ref. + * a ref to keep the binding alive. The page_pool does not release the + * ref until all the net_iovs allocated from this binding are released + * back to the page_pool. * * The binding undos itself and unmaps the underlying dmabuf once all * those refs are dropped and the binding is no longer desired or in @@ -32,7 +33,10 @@ struct net_devmem_dmabuf_binding { * * net_devmem_get_net_iov() on dmabuf net_iovs will increment this * reference, making sure that the binding remains alive until all the - * net_iovs are no longer used. + * net_iovs are no longer used. net_iovs allocated from this binding + * that are stuck in the TX path for any reason (such as awaiting + * retransmits) hold a reference to the binding until the skb holding + * them is freed. */ refcount_t ref; =20 @@ -48,6 +52,14 @@ struct net_devmem_dmabuf_binding { * active. */ u32 id; + + /* Array of net_iov pointers for this binding, sorted by virtual + * address. This array is convenient to map the virtual addresses to + * net_iovs in the TX path. + */ + struct net_iov **tx_vec; + + struct work_struct unbind_w; }; =20 #if defined(CONFIG_NET_DEVMEM) @@ -64,14 +76,17 @@ struct dmabuf_genpool_chunk_owner { dma_addr_t base_dma_addr; }; =20 -void __net_devmem_dmabuf_binding_free(struct net_devmem_dmabuf_binding *bi= nding); +void __net_devmem_dmabuf_binding_free(struct work_struct *wq); struct net_devmem_dmabuf_binding * -net_devmem_bind_dmabuf(struct net_device *dev, unsigned int dmabuf_fd, - struct netlink_ext_ack *extack); +net_devmem_bind_dmabuf(struct net_device *dev, + enum dma_data_direction direction, + unsigned int dmabuf_fd, struct netlink_ext_ack *extack); +struct net_devmem_dmabuf_binding *net_devmem_lookup_dmabuf(u32 id); void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf_binding *binding); int net_devmem_bind_dmabuf_to_queue(struct net_device *dev, u32 rxq_idx, struct net_devmem_dmabuf_binding *binding, struct netlink_ext_ack *extack); +void net_devmem_bind_tx_release(struct sock *sk); =20 static inline struct dmabuf_genpool_chunk_owner * net_devmem_iov_to_chunk_owner(const struct net_iov *niov) @@ -100,10 +115,10 @@ static inline unsigned long net_iov_virtual_addr(cons= t struct net_iov *niov) ((unsigned long)net_iov_idx(niov) << PAGE_SHIFT); } =20 -static inline void +static inline bool net_devmem_dmabuf_binding_get(struct net_devmem_dmabuf_binding *binding) { - refcount_inc(&binding->ref); + return refcount_inc_not_zero(&binding->ref); } =20 static inline void @@ -112,7 +127,8 @@ net_devmem_dmabuf_binding_put(struct net_devmem_dmabuf_= binding *binding) if (!refcount_dec_and_test(&binding->ref)) return; =20 - __net_devmem_dmabuf_binding_free(binding); + INIT_WORK(&binding->unbind_w, __net_devmem_dmabuf_binding_free); + schedule_work(&binding->unbind_w); } =20 void net_devmem_get_net_iov(struct net_iov *niov); @@ -123,6 +139,11 @@ net_devmem_alloc_dmabuf(struct net_devmem_dmabuf_bindi= ng *binding); void net_devmem_free_dmabuf(struct net_iov *ppiov); =20 bool net_is_devmem_iov(struct net_iov *niov); +struct net_devmem_dmabuf_binding * +net_devmem_get_binding(struct sock *sk, unsigned int dmabuf_id); +struct net_iov * +net_devmem_get_niov_at(struct net_devmem_dmabuf_binding *binding, size_t a= ddr, + size_t *off, size_t *size); =20 #else struct net_devmem_dmabuf_binding; @@ -140,18 +161,23 @@ static inline void net_devmem_put_net_iov(struct net_= iov *niov) { } =20 -static inline void -__net_devmem_dmabuf_binding_free(struct net_devmem_dmabuf_binding *binding) +static inline void __net_devmem_dmabuf_binding_free(struct work_struct *wq) { } =20 static inline struct net_devmem_dmabuf_binding * net_devmem_bind_dmabuf(struct net_device *dev, unsigned int dmabuf_fd, + enum dma_data_direction direction, struct netlink_ext_ack *extack) { return ERR_PTR(-EOPNOTSUPP); } =20 +static inline struct net_devmem_dmabuf_binding *net_devmem_lookup_dmabuf(u= 32 id) +{ + return NULL; +} + static inline void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf_binding *binding) { @@ -190,6 +216,19 @@ static inline bool net_is_devmem_iov(struct net_iov *n= iov) { return false; } + +static inline struct net_devmem_dmabuf_binding * +net_devmem_get_binding(struct sock *sk, unsigned int dmabuf_id) +{ + return ERR_PTR(-EOPNOTSUPP); +} + +static inline struct net_iov * +net_devmem_get_niov_at(struct net_devmem_dmabuf_binding *binding, size_t a= ddr, + size_t *off, size_t *size) +{ + return NULL; +} #endif =20 #endif /* _NET_DEVMEM_H */ diff --git a/net/core/netdev-genl.c b/net/core/netdev-genl.c index 6e5f2de4d947..6e7cd6a5c177 100644 --- a/net/core/netdev-genl.c +++ b/net/core/netdev-genl.c @@ -874,7 +874,8 @@ int netdev_nl_bind_rx_doit(struct sk_buff *skb, struct = genl_info *info) goto err_unlock; } =20 - binding =3D net_devmem_bind_dmabuf(netdev, dmabuf_fd, info->extack); + binding =3D net_devmem_bind_dmabuf(netdev, DMA_FROM_DEVICE, dmabuf_fd, + info->extack); if (IS_ERR(binding)) { err =3D PTR_ERR(binding); goto err_unlock; @@ -931,10 +932,67 @@ int netdev_nl_bind_rx_doit(struct sk_buff *skb, struc= t genl_info *info) return err; } =20 -/* stub */ int netdev_nl_bind_tx_doit(struct sk_buff *skb, struct genl_info *info) { - return 0; + struct net_devmem_dmabuf_binding *binding; + struct list_head *sock_binding_list; + struct net_device *netdev; + u32 ifindex, dmabuf_fd; + struct sk_buff *rsp; + int err =3D 0; + void *hdr; + + if (GENL_REQ_ATTR_CHECK(info, NETDEV_A_DEV_IFINDEX) || + GENL_REQ_ATTR_CHECK(info, NETDEV_A_DMABUF_FD)) + return -EINVAL; + + ifindex =3D nla_get_u32(info->attrs[NETDEV_A_DEV_IFINDEX]); + dmabuf_fd =3D nla_get_u32(info->attrs[NETDEV_A_DMABUF_FD]); + + sock_binding_list =3D genl_sk_priv_get(&netdev_nl_family, + NETLINK_CB(skb).sk); + if (IS_ERR(sock_binding_list)) + return PTR_ERR(sock_binding_list); + + rsp =3D genlmsg_new(GENLMSG_DEFAULT_SIZE, GFP_KERNEL); + if (!rsp) + return -ENOMEM; + + hdr =3D genlmsg_iput(rsp, info); + if (!hdr) { + err =3D -EMSGSIZE; + goto err_genlmsg_free; + } + + rtnl_lock(); + + netdev =3D __dev_get_by_index(genl_info_net(info), ifindex); + if (!netdev || !netif_device_present(netdev)) { + err =3D -ENODEV; + goto err_unlock; + } + + binding =3D net_devmem_bind_dmabuf(netdev, DMA_TO_DEVICE, dmabuf_fd, + info->extack); + if (IS_ERR(binding)) { + err =3D PTR_ERR(binding); + goto err_unlock; + } + + list_add(&binding->list, sock_binding_list); + + nla_put_u32(rsp, NETDEV_A_DMABUF_ID, binding->id); + genlmsg_end(rsp, hdr); + + rtnl_unlock(); + + return genlmsg_reply(rsp, info); + +err_unlock: + rtnl_unlock(); +err_genlmsg_free: + nlmsg_free(rsp); + return err; } =20 void netdev_nl_sock_priv_init(struct list_head *priv) diff --git a/net/core/skbuff.c b/net/core/skbuff.c index ee2d1b769c13..f911f8c6773f 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -1667,7 +1667,8 @@ void mm_unaccount_pinned_pages(struct mmpin *mmp) } EXPORT_SYMBOL_GPL(mm_unaccount_pinned_pages); =20 -static struct ubuf_info *msg_zerocopy_alloc(struct sock *sk, size_t size) +static struct ubuf_info *msg_zerocopy_alloc(struct sock *sk, size_t size, + bool devmem) { struct ubuf_info_msgzc *uarg; struct sk_buff *skb; @@ -1682,7 +1683,7 @@ static struct ubuf_info *msg_zerocopy_alloc(struct so= ck *sk, size_t size) uarg =3D (void *)skb->cb; uarg->mmp.user =3D NULL; =20 - if (mm_account_pinned_pages(&uarg->mmp, size)) { + if (likely(!devmem) && mm_account_pinned_pages(&uarg->mmp, size)) { kfree_skb(skb); return NULL; } @@ -1705,7 +1706,7 @@ static inline struct sk_buff *skb_from_uarg(struct ub= uf_info_msgzc *uarg) } =20 struct ubuf_info *msg_zerocopy_realloc(struct sock *sk, size_t size, - struct ubuf_info *uarg) + struct ubuf_info *uarg, bool devmem) { if (uarg) { struct ubuf_info_msgzc *uarg_zc; @@ -1735,7 +1736,8 @@ struct ubuf_info *msg_zerocopy_realloc(struct sock *s= k, size_t size, =20 next =3D (u32)atomic_read(&sk->sk_zckey); if ((u32)(uarg_zc->id + uarg_zc->len) =3D=3D next) { - if (mm_account_pinned_pages(&uarg_zc->mmp, size)) + if (likely(!devmem) && + mm_account_pinned_pages(&uarg_zc->mmp, size)) return NULL; uarg_zc->len++; uarg_zc->bytelen =3D bytelen; @@ -1750,7 +1752,7 @@ struct ubuf_info *msg_zerocopy_realloc(struct sock *s= k, size_t size, } =20 new_alloc: - return msg_zerocopy_alloc(sk, size); + return msg_zerocopy_alloc(sk, size, devmem); } EXPORT_SYMBOL_GPL(msg_zerocopy_realloc); =20 @@ -1854,7 +1856,8 @@ EXPORT_SYMBOL_GPL(msg_zerocopy_ubuf_ops); =20 int skb_zerocopy_iter_stream(struct sock *sk, struct sk_buff *skb, struct msghdr *msg, int len, - struct ubuf_info *uarg) + struct ubuf_info *uarg, + struct net_devmem_dmabuf_binding *binding) { int err, orig_len =3D skb->len; =20 @@ -1873,7 +1876,8 @@ int skb_zerocopy_iter_stream(struct sock *sk, struct = sk_buff *skb, return -EEXIST; } =20 - err =3D __zerocopy_sg_from_iter(msg, sk, skb, &msg->msg_iter, len); + err =3D __zerocopy_sg_from_iter(msg, sk, skb, &msg->msg_iter, len, + binding); if (err =3D=3D -EFAULT || (err =3D=3D -EMSGSIZE && skb->len =3D=3D orig_l= en)) { struct sock *save_sk =3D skb->sk; =20 diff --git a/net/core/sock.c b/net/core/sock.c index a0598518ce89..e1dae98b55ca 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3012,6 +3012,12 @@ int __sock_cmsg_send(struct sock *sk, struct cmsghdr= *cmsg, if (!sk_set_prio_allowed(sk, *(u32 *)CMSG_DATA(cmsg))) return -EPERM; sockc->priority =3D *(u32 *)CMSG_DATA(cmsg); + break; + case SCM_DEVMEM_DMABUF: + if (cmsg->cmsg_len !=3D CMSG_LEN(sizeof(u32))) + return -EINVAL; + sockc->dmabuf_id =3D *(u32 *)CMSG_DATA(cmsg); + break; default: return -EINVAL; diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index ea7a260bec8a..7d8a5f3fae9b 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -1015,7 +1015,8 @@ static int __ip_append_data(struct sock *sk, uarg =3D msg->msg_ubuf; } } else if (sock_flag(sk, SOCK_ZEROCOPY)) { - uarg =3D msg_zerocopy_realloc(sk, length, skb_zcopy(skb)); + uarg =3D msg_zerocopy_realloc(sk, length, skb_zcopy(skb), + false); if (!uarg) return -ENOBUFS; extra_uref =3D !skb_zcopy(skb); /* only ref on new uarg */ diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 46951e749308..8564b0e66bbd 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1059,6 +1059,7 @@ int tcp_sendmsg_fastopen(struct sock *sk, struct msgh= dr *msg, int *copied, =20 int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size) { + struct net_devmem_dmabuf_binding *binding =3D NULL; struct tcp_sock *tp =3D tcp_sk(sk); struct ubuf_info *uarg =3D NULL; struct sk_buff *skb; @@ -1066,11 +1067,24 @@ int tcp_sendmsg_locked(struct sock *sk, struct msgh= dr *msg, size_t size) int flags, err, copied =3D 0; int mss_now =3D 0, size_goal, copied_syn =3D 0; int process_backlog =3D 0; + bool sockc_valid =3D true; int zc =3D 0; long timeo; =20 flags =3D msg->msg_flags; =20 + sockc =3D (struct sockcm_cookie){ .tsflags =3D READ_ONCE(sk->sk_tsflags), + .dmabuf_id =3D 0 }; + if (msg->msg_controllen) { + err =3D sock_cmsg_send(sk, msg, &sockc); + if (unlikely(err)) + /* Don't return error until MSG_FASTOPEN has been + * processed; that may succeed even if the cmsg is + * invalid. + */ + sockc_valid =3D false; + } + if ((flags & MSG_ZEROCOPY) && size) { if (msg->msg_ubuf) { uarg =3D msg->msg_ubuf; @@ -1078,7 +1092,8 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr= *msg, size_t size) zc =3D MSG_ZEROCOPY; } else if (sock_flag(sk, SOCK_ZEROCOPY)) { skb =3D tcp_write_queue_tail(sk); - uarg =3D msg_zerocopy_realloc(sk, size, skb_zcopy(skb)); + uarg =3D msg_zerocopy_realloc(sk, size, skb_zcopy(skb), + sockc_valid && !!sockc.dmabuf_id); if (!uarg) { err =3D -ENOBUFS; goto out_err; @@ -1087,12 +1102,27 @@ int tcp_sendmsg_locked(struct sock *sk, struct msgh= dr *msg, size_t size) zc =3D MSG_ZEROCOPY; else uarg_to_msgzc(uarg)->zerocopy =3D 0; + + if (sockc_valid && sockc.dmabuf_id) { + binding =3D net_devmem_get_binding(sk, sockc.dmabuf_id); + if (IS_ERR(binding)) { + err =3D PTR_ERR(binding); + binding =3D NULL; + goto out_err; + } + } } } else if (unlikely(msg->msg_flags & MSG_SPLICE_PAGES) && size) { if (sk->sk_route_caps & NETIF_F_SG) zc =3D MSG_SPLICE_PAGES; } =20 + if (sockc_valid && sockc.dmabuf_id && + (!(flags & MSG_ZEROCOPY) || !sock_flag(sk, SOCK_ZEROCOPY))) { + err =3D -EINVAL; + goto out_err; + } + if (unlikely(flags & MSG_FASTOPEN || inet_test_bit(DEFER_CONNECT, sk)) && !tp->repair) { @@ -1131,14 +1161,8 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghd= r *msg, size_t size) /* 'common' sending to sendq */ } =20 - sockc =3D (struct sockcm_cookie) { .tsflags =3D READ_ONCE(sk->sk_tsflags)= }; - if (msg->msg_controllen) { - err =3D sock_cmsg_send(sk, msg, &sockc); - if (unlikely(err)) { - err =3D -EINVAL; - goto out_err; - } - } + if (!sockc_valid) + goto out_err; =20 /* This should be in poll */ sk_clear_bit(SOCKWQ_ASYNC_NOSPACE, sk); @@ -1256,7 +1280,8 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr= *msg, size_t size) goto wait_for_space; } =20 - err =3D skb_zerocopy_iter_stream(sk, skb, msg, copy, uarg); + err =3D skb_zerocopy_iter_stream(sk, skb, msg, copy, uarg, + binding); if (err =3D=3D -EMSGSIZE || err =3D=3D -EEXIST) { tcp_mark_push(tp, skb); goto new_segment; @@ -1337,6 +1362,8 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr= *msg, size_t size) /* msg->msg_ubuf is pinned by the caller so we don't take extra refs */ if (uarg && !msg->msg_ubuf) net_zcopy_put(uarg); + if (binding) + net_devmem_dmabuf_binding_put(binding); return copied + copied_syn; =20 do_error: @@ -1354,6 +1381,9 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr= *msg, size_t size) sk->sk_write_space(sk); tcp_chrono_stop(sk, TCP_CHRONO_SNDBUF_LIMITED); } + if (binding) + net_devmem_dmabuf_binding_put(binding); + return err; } EXPORT_SYMBOL_GPL(tcp_sendmsg_locked); diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index d577bf2f3053..e9e752f08f87 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -1523,7 +1523,8 @@ static int __ip6_append_data(struct sock *sk, uarg =3D msg->msg_ubuf; } } else if (sock_flag(sk, SOCK_ZEROCOPY)) { - uarg =3D msg_zerocopy_realloc(sk, length, skb_zcopy(skb)); + uarg =3D msg_zerocopy_realloc(sk, length, skb_zcopy(skb), + false); if (!uarg) return -ENOBUFS; extra_uref =3D !skb_zcopy(skb); /* only ref on new uarg */ diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio= _transport_common.c index 7f7de6d88096..6e7b727c781c 100644 --- a/net/vmw_vsock/virtio_transport_common.c +++ b/net/vmw_vsock/virtio_transport_common.c @@ -87,7 +87,7 @@ static int virtio_transport_init_zcopy_skb(struct vsock_s= ock *vsk, =20 uarg =3D msg_zerocopy_realloc(sk_vsock(vsk), iter->count, - NULL); + NULL, false); if (!uarg) return -1; =20 @@ -107,8 +107,7 @@ static int virtio_transport_fill_skb(struct sk_buff *sk= b, { if (zcopy) return __zerocopy_sg_from_iter(info->msg, NULL, skb, - &info->msg->msg_iter, - len); + &info->msg->msg_iter, len, NULL); =20 return memcpy_from_msg(skb_put(skb, len), info->msg, len); } --=20 2.49.0.rc0.332.g42c0ae87b1-goog From nobody Sun Feb 8 02:41:22 2026 Received: from mail-pj1-f74.google.com (mail-pj1-f74.google.com [209.85.216.74]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 0AF0121519B for ; Sat, 8 Mar 2025 21:40:56 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.74 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470059; cv=none; b=FVDUDU5YPNX/h/RY9XSEtbJLFCET9k7U6JjS8SPQqXTCGhOsfaVsprzFxLeydqboZtbHXvntlJ3nlYXAX57wzSoiyQ4coHWkWXV+q1pMsy13O0jG6HlsRf/5Md4u2qCbZH7cOHpWI5Fe0e3mVC2LgDruFiHXUw8YdqAV2zmP7iA= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470059; c=relaxed/simple; bh=Waqs9wqvrVBOROjzMcSiYTjSpSVpd+3UCgCmMkkpDvo=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=jfdOWL7wMqCaL85h7QqSFVxnTV2TVpVhWiwq/nIJnalXLBBxYRF6wnoN/7xkP/UB7MxJeZVyOAZoRKxYiZA6G04INqzirSUeaYTNnqi8fPWlqQIJF8IaJl1oG/fUtyRjaUydPYBTEdUyIFO9OvwVEJyNFmC1N1Qvxix/nvMnfJg= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=HMk/spdp; arc=none smtp.client-ip=209.85.216.74 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="HMk/spdp" Received: by mail-pj1-f74.google.com with SMTP id 98e67ed59e1d1-2fecd603bb8so5192071a91.3 for ; Sat, 08 Mar 2025 13:40:56 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470055; x=1742074855; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=58V3xIMqcRJTwA4dVCAwUps2MRFlACWl18gn2/ylnO8=; b=HMk/spdp/J+DPl8rfC64B1CJE3PhdAx+kDBj2lf+8+rpPzx3CPE6knnwjkfp/kbi5l jsw5BHmClPe7E+m9m1/uKDJB0I03oKFRPIXLwXK+lG0hZ+r9pAKxSYXesyEM8CRpAffn +oH2MZuO8Mu/u9W1hQCQ9Emo9SdvsbtBaVJ9o8xAeR41uS5eEadXqzyTMGRjx5tCwxOX jvFZKbm047Qj3ZtnSLt2G1/ws9xhdoybHr9yKJ4t7P8Ii4iblBK9B444y0iJsbTutY+s 96KeqSmj7pabdKDtCglgtaSUz7cBj41NyJoi4W2ILcNaZ4ago61tftqD4Famdiazo83+ YeeQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470055; x=1742074855; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=58V3xIMqcRJTwA4dVCAwUps2MRFlACWl18gn2/ylnO8=; b=bq216Cl10JTg/kPTo+10+MTIh1P7eM8bCyqt8KEfALTDbT7EZW4wSCOG8BgM/4yCBQ l9Kkll+SjnJc3gpPk0foDB4WDOurAGMmOy9syUuclO5gTLyg2BxWSOWRjw/1Wt0LwE3f cJZyQSo3Btoijfq3KnRT6fACwpgkA9mTuuWfxasSoxzzbptC20mA4x/dWqyYTOakr5Qp dXL5iazOu+vRNPLEWrt6F6x4EWHvmCLPbDg+KkARxDXMAR5zeKqTnq4ZY0lafrUWtd26 wBzCoUyD0habFYvGzQ4gsAR2ge+gVNGCsBbbxReCB8VhdDG9TTn5Ij/oi10GLbQGeAHQ Z52g== X-Forwarded-Encrypted: i=1; AJvYcCVLJ15RmVltSBsiMuio+8C22dy5FpX6JI9Zyi0qnNat4K9QOHnkCe2vt3d3TUF0c05O3u+0+bPEFZ77XmM=@vger.kernel.org X-Gm-Message-State: AOJu0Yxxk+wvR9TMYWyb4RNXHEQJmpMQdMUtCvdlPG+PfkpsWgSgWzpl 7u0i8N+OQL6Ye4z1GqQn1Fjqic3Ify6NDmt2Z7jz8bf91f7aZUZCdqGYK6uovE8HsDitupg4/3i mI9iN7R357r9C4p4ML3k1TA== X-Google-Smtp-Source: AGHT+IFSb6U+tbU6T3ppbOqqdxqHzaRcKQ8NbHk2U0tmSzdSX0/NwuESjTr521WJ+R0cR10sB0m7/cPnT5827g+YgA== X-Received: from pgbck14.prod.google.com ([2002:a05:6a02:90e:b0:af2:681:e27a]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6a21:78a8:b0:1ee:d87c:8051 with SMTP id adf61e73a8af0-1f544af83d7mr15659224637.13.1741470055464; Sat, 08 Mar 2025 13:40:55 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:41 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-6-almasrymina@google.com> Subject: [PATCH net-next v7 5/9] net: add devmem TCP TX documentation From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Add documentation outlining the usage and details of the devmem TCP TX API. Signed-off-by: Mina Almasry Acked-by: Stanislav Fomichev Tested-by: Lei Yang --- v5: - Address comments from Stan and Bagas v4: - Mention SO_BINDTODEVICE is recommended (me/Pavel). v2: - Update documentation for iov_base is the dmabuf offset (Stan) --- Documentation/networking/devmem.rst | 150 +++++++++++++++++++++++++++- 1 file changed, 146 insertions(+), 4 deletions(-) diff --git a/Documentation/networking/devmem.rst b/Documentation/networking= /devmem.rst index d95363645331..1c476522d6f5 100644 --- a/Documentation/networking/devmem.rst +++ b/Documentation/networking/devmem.rst @@ -62,15 +62,15 @@ More Info https://lore.kernel.org/netdev/20240831004313.3713467-1-almasrymina@go= ogle.com/ =20 =20 -Interface -=3D=3D=3D=3D=3D=3D=3D=3D=3D +RX Interface +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =20 =20 Example ------- =20 -tools/testing/selftests/net/ncdevmem.c:do_server shows an example of setti= ng up -the RX path of this API. +./tools/testing/selftests/drivers/net/hw/ncdevmem:do_server shows an examp= le of +setting up the RX path of this API. =20 =20 NIC Setup @@ -235,6 +235,148 @@ can be less than the tokens provided by the user in c= ase of: (a) an internal kernel leak bug. (b) the user passed more than 1024 frags. =20 +TX Interface +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D + + +Example +------- + +./tools/testing/selftests/drivers/net/hw/ncdevmem:do_client shows an examp= le of +setting up the TX path of this API. + + +NIC Setup +--------- + +The user must bind a TX dmabuf to a given NIC using the netlink API:: + + struct netdev_bind_tx_req *req =3D NULL; + struct netdev_bind_tx_rsp *rsp =3D NULL; + struct ynl_error yerr; + + *ys =3D ynl_sock_create(&ynl_netdev_family, &yerr); + + req =3D netdev_bind_tx_req_alloc(); + netdev_bind_tx_req_set_ifindex(req, ifindex); + netdev_bind_tx_req_set_fd(req, dmabuf_fd); + + rsp =3D netdev_bind_tx(*ys, req); + + tx_dmabuf_id =3D rsp->id; + + +The netlink API returns a dmabuf_id: a unique ID that refers to this dmabuf +that has been bound. + +The user can unbind the dmabuf from the netdevice by closing the netlink s= ocket +that established the binding. We do this so that the binding is automatica= lly +unbound even if the userspace process crashes. + +Note that any reasonably well-behaved dmabuf from any exporter should work= with +devmem TCP, even if the dmabuf is not actually backed by devmem. An exampl= e of +this is udmabuf, which wraps user memory (non-devmem) in a dmabuf. + +Socket Setup +------------ + +The user application must use MSG_ZEROCOPY flag when sending devmem TCP. D= evmem +cannot be copied by the kernel, so the semantics of the devmem TX are simi= lar +to the semantics of MSG_ZEROCOPY:: + + setsockopt(socket_fd, SOL_SOCKET, SO_ZEROCOPY, &opt, sizeof(opt)); + +It is also recommended that the user binds the TX socket to the same inter= face +the dma-buf has been bound to via SO_BINDTODEVICE:: + + setsockopt(socket_fd, SOL_SOCKET, SO_BINDTODEVICE, ifname, strlen(ifname)= + 1); + + +Sending data +------------ + +Devmem data is sent using the SCM_DEVMEM_DMABUF cmsg. + +The user should create a msghdr where, + +* iov_base is set to the offset into the dmabuf to start sending from +* iov_len is set to the number of bytes to be sent from the dmabuf + +The user passes the dma-buf id to send from via the dmabuf_tx_cmsg.dmabuf_= id. + +The example below sends 1024 bytes from offset 100 into the dmabuf, and 20= 48 +from offset 2000 into the dmabuf. The dmabuf to send from is tx_dmabuf_id:: + + char ctrl_data[CMSG_SPACE(sizeof(struct dmabuf_tx_cmsg))]; + struct dmabuf_tx_cmsg ddmabuf; + struct msghdr msg =3D {}; + struct cmsghdr *cmsg; + struct iovec iov[2]; + + iov[0].iov_base =3D (void*)100; + iov[0].iov_len =3D 1024; + iov[1].iov_base =3D (void*)2000; + iov[1].iov_len =3D 2048; + + msg.msg_iov =3D iov; + msg.msg_iovlen =3D 2; + + msg.msg_control =3D ctrl_data; + msg.msg_controllen =3D sizeof(ctrl_data); + + cmsg =3D CMSG_FIRSTHDR(&msg); + cmsg->cmsg_level =3D SOL_SOCKET; + cmsg->cmsg_type =3D SCM_DEVMEM_DMABUF; + cmsg->cmsg_len =3D CMSG_LEN(sizeof(struct dmabuf_tx_cmsg)); + + ddmabuf.dmabuf_id =3D tx_dmabuf_id; + + *((struct dmabuf_tx_cmsg *)CMSG_DATA(cmsg)) =3D ddmabuf; + + sendmsg(socket_fd, &msg, MSG_ZEROCOPY); + + +Reusing TX dmabufs +------------------ + +Similar to MSG_ZEROCOPY with regular memory, the user should not modify the +contents of the dma-buf while a send operation is in progress. This is bec= ause +the kernel does not keep a copy of the dmabuf contents. Instead, the kernel +will pin and send data from the buffer available to the userspace. + +Just as in MSG_ZEROCOPY, the kernel notifies the userspace of send complet= ions +using MSG_ERRQUEUE:: + + int64_t tstop =3D gettimeofday_ms() + waittime_ms; + char control[CMSG_SPACE(100)] =3D {}; + struct sock_extended_err *serr; + struct msghdr msg =3D {}; + struct cmsghdr *cm; + int retries =3D 10; + __u32 hi, lo; + + msg.msg_control =3D control; + msg.msg_controllen =3D sizeof(control); + + while (gettimeofday_ms() < tstop) { + if (!do_poll(fd)) continue; + + ret =3D recvmsg(fd, &msg, MSG_ERRQUEUE); + + for (cm =3D CMSG_FIRSTHDR(&msg); cm; cm =3D CMSG_NXTHDR(&m= sg, cm)) { + serr =3D (void *)CMSG_DATA(cm); + + hi =3D serr->ee_data; + lo =3D serr->ee_info; + + fprintf(stdout, "tx complete [%d,%d]\n", lo, hi); + } + } + +After the associated sendmsg has been completed, the dmabuf can be reused = by +the userspace. + + Implementation & Caveats =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =20 --=20 2.49.0.rc0.332.g42c0ae87b1-goog From nobody Sun Feb 8 02:41:22 2026 Received: from mail-pj1-f73.google.com (mail-pj1-f73.google.com [209.85.216.73]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id DB6FE215784 for ; Sat, 8 Mar 2025 21:40:57 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.73 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470060; cv=none; b=rD4r1YDWPPgtpVz/hfM8Mvt25R/Uqh6ZZ3LScvoE7+n5F/NmfOFImP8MmcpgGu7oMBjoFlpBYesiBMHUT63dz+4i5W46EAoiA42v1j5DpgM6mIZ3dhMZyLFZlMerjo5N5EXlYEuU3zsnDv5Fpr3vAru7PfdtFfO3BcPgt5c4SOA= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470060; c=relaxed/simple; bh=EuanlU0YhzPMsh9Fv+D25HTrf6PC7vW/kt0OFwlSXtk=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=j5oD6xAa9ZIP7qN/+lQL/KSZZgW34QT6LGOs51iMkHPHwgdKlC9UrUTYRdsqZnkzGu1o9tETSRFJxVWNheqqkGxH58qfYyuwDsIQ1Ke7zcqlEypqpBGOhfY41wiRpiBCY/xG6ydPRdWBiFM0XHrf7ZsNkTNR3n4+RE2407J/EkI= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=XG/xgi4y; arc=none smtp.client-ip=209.85.216.73 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="XG/xgi4y" Received: by mail-pj1-f73.google.com with SMTP id 98e67ed59e1d1-2ff8340d547so2556957a91.2 for ; Sat, 08 Mar 2025 13:40:57 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470057; x=1742074857; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=RTUfOwUIRZvDZQegq2DBALS/SAbHr71GuK4ayNmHc88=; b=XG/xgi4y4fL7naALjEseVPvEJnxmpVv0pNSO99bmVCSwKyB5J6HApOheHJUcEAjX9t vv1TbdUQhxBEObVQ+KYiMgdPLtNnsJ3NVemP4CAZYV4y+wvMsQOpzEbHTYR0S5JqWRxD ahuYnXgQ1Z01pijvkZK4hCeQ1FtGVJ/bkrVBCMuB/pj6gY/Ij9934oYFpGbdLfpJJm9d 3157cjtuaGL9DXcACxMtPl94tK+OcRvHdFoLwaSHnRpSbLZAYyjCQCVBT2J78zYldD5z EYu+J8ut5Dw/ofn9Odf/v2iHHWXCUW4ZH7MRSLnPwHQAFYn/18XmKE/OXtNQxkM+A7LA yeww== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470057; x=1742074857; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=RTUfOwUIRZvDZQegq2DBALS/SAbHr71GuK4ayNmHc88=; b=NMmD6n3FM2gzr57JhuqA5fkaOqe6zTEbGwDFwwpBPcVZV4MI3Y+3YSdZjKq1uN38oP cMnHTh1hoSYwkp82rR/wshC30Q3ixKTqC4AkYis5+tFeflh4UjBnm6iFpTU5kjRlJY3O 3t+BWDmIYwfxbEHpcl9Q1ufJ2GfEsEb3MNG2X5wUw3jiT/YIRIH3zVZIA/TcqngRhe/d wKQmawGS9GnohIPq6r3kU925b+zcXiDVXzdX75kBTx0HigAbM2bk54hv2YoQrGKrCPZP T6eHaMq+Fdx+EFD9Yi11bWXWq0Pu5l0JVruIttbHOx/Pf5lPI5301Cb563E80zXTF4RG OusQ== X-Forwarded-Encrypted: i=1; AJvYcCXglmF0KkqmJxoCY9XYja4VZLka+4zNqsVdNNHNjnPos+94h96cfo1xCEVHUEWi1ynhzPYbwYAlzI2SeYQ=@vger.kernel.org X-Gm-Message-State: AOJu0Yw/nkfSI30S/gRLTbK6Xrl5C4sbvgVTiyvIeO+JyIJUzNfWUIqi IuJUFB0LUMpZC4g9MA8W7BuYsrFljB1EwOQXvOhsVtzYIUH9qop8WjQTsosCEluzG11Z/NaH4NQ eruJqZQe2mJa+Hsixp8VCkQ== X-Google-Smtp-Source: AGHT+IF+qJji92vVM7ENVK7QLbTkaQ8pyPqqTNc3UyFPjNmAEVOG0LmWKiODB/VGhC+7g3NuChwnLxDZaZOw+TV4OQ== X-Received: from pjbsv15.prod.google.com ([2002:a17:90b:538f:b0:2fa:a101:743]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90b:1c05:b0:2ee:a583:e616 with SMTP id 98e67ed59e1d1-2ff7ce6d543mr14055421a91.9.1741470057178; Sat, 08 Mar 2025 13:40:57 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:42 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-7-almasrymina@google.com> Subject: [PATCH net-next v7 6/9] net: enable driver support for netmem TX From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Drivers need to make sure not to pass netmem dma-addrs to the dma-mapping API in order to support netmem TX. Add helpers and netmem_dma_*() helpers that enables special handling of netmem dma-addrs that drivers can use. Document in netmem.rst what drivers need to do to support netmem TX. Signed-off-by: Mina Almasry Acked-by: Stanislav Fomichev Tested-by: Lei Yang --- v5: - Fix netmet TX documentation (Stan). v4: - New patch --- .../networking/net_cachelines/net_device.rst | 1 + Documentation/networking/netdev-features.rst | 5 ++++ Documentation/networking/netmem.rst | 23 +++++++++++++++++-- include/linux/netdevice.h | 2 ++ include/net/netmem.h | 20 ++++++++++++++++ 5 files changed, 49 insertions(+), 2 deletions(-) diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Docum= entation/networking/net_cachelines/net_device.rst index 6327e689e8a8..8c0334851b45 100644 --- a/Documentation/networking/net_cachelines/net_device.rst +++ b/Documentation/networking/net_cachelines/net_device.rst @@ -10,6 +10,7 @@ Type Name = fastpath_tx_acce =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D unsigned_long:32 priv_flags read_mostl= y __dev_queue_xmit(tx) unsigned_long:1 lltx read_mostl= y HARD_TX_LOCK,HARD_TX_TRYLOCK,HARD_TX_UNLOCK(t= x) +unsigned long:1 netmem_tx:1; read_mostly char name[16] struct netdev_name_node* name_node struct dev_ifalias* ifalias diff --git a/Documentation/networking/netdev-features.rst b/Documentation/n= etworking/netdev-features.rst index 5014f7cc1398..02bd7536fc0c 100644 --- a/Documentation/networking/netdev-features.rst +++ b/Documentation/networking/netdev-features.rst @@ -188,3 +188,8 @@ Redundancy) frames from one port to another in hardware. This should be set for devices which duplicate outgoing HSR (High-availabi= lity Seamless Redundancy) or PRP (Parallel Redundancy Protocol) tags automatica= lly frames in hardware. + +* netmem-tx + +This should be set for devices which support netmem TX. See +Documentation/networking/netmem.rst diff --git a/Documentation/networking/netmem.rst b/Documentation/networking= /netmem.rst index 7de21ddb5412..b63aded46337 100644 --- a/Documentation/networking/netmem.rst +++ b/Documentation/networking/netmem.rst @@ -19,8 +19,8 @@ Benefits of Netmem : * Simplified Development: Drivers interact with a consistent API, regardless of the underlying memory implementation. =20 -Driver Requirements -=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D +Driver RX Requirements +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =20 1. The driver must support page_pool. =20 @@ -77,3 +77,22 @@ Driver Requirements that purpose, but be mindful that some netmem types might have longer circulation times, such as when userspace holds a reference in zerocopy scenarios. + +Driver TX Requirements +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D + +1. The Driver must not pass the netmem dma_addr to any of the dma-mapping = APIs + directly. This is because netmem dma_addrs may come from a source like + dma-buf that is not compatible with the dma-mapping APIs. + + Helpers like netmem_dma_unmap_page_attrs() & netmem_dma_unmap_addr_set() + should be used in lieu of dma_unmap_page[_attrs](), dma_unmap_addr_set(= ). + The netmem variants will handle netmem dma_addrs correctly regardless o= f the + source, delegating to the dma-mapping APIs when appropriate. + + Not all dma-mapping APIs have netmem equivalents at the moment. If your + driver relies on a missing netmem API, feel free to add and propose to + netdev@, or reach out to the maintainers and/or almasrymina@google.com = for + help adding the netmem API. + +2. Driver should declare support by setting `netdev->netmem_tx =3D true` diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 9a297757df7e..ee8aa020eb3c 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1771,6 +1771,7 @@ enum netdev_reg_state { * @lltx: device supports lockless Tx. Deprecated for real HW * drivers. Mainly used by logical interfaces, such as * bonding and tunnels + * @netmem_tx: device support netmem_tx. * * @name: This is the first field of the "visible" part of this structure * (i.e. as seen by users in the "Space.c" file). It is the name @@ -2088,6 +2089,7 @@ struct net_device { struct_group(priv_flags_fast, unsigned long priv_flags:32; unsigned long lltx:1; + unsigned long netmem_tx:1; ); const struct net_device_ops *netdev_ops; const struct header_ops *header_ops; diff --git a/include/net/netmem.h b/include/net/netmem.h index e8afe6b654aa..068b6e1760dc 100644 --- a/include/net/netmem.h +++ b/include/net/netmem.h @@ -8,6 +8,7 @@ #ifndef _NET_NETMEM_H #define _NET_NETMEM_H =20 +#include #include #include =20 @@ -280,4 +281,23 @@ static inline unsigned long netmem_get_dma_addr(netmem= _ref netmem) void get_netmem(netmem_ref netmem); void put_netmem(netmem_ref netmem); =20 +#define netmem_dma_unmap_addr_set(NETMEM, PTR, ADDR_NAME, VAL) \ + do { \ + if (!netmem_is_net_iov(NETMEM)) \ + dma_unmap_addr_set(PTR, ADDR_NAME, VAL); \ + else \ + dma_unmap_addr_set(PTR, ADDR_NAME, 0); \ + } while (0) + +static inline void netmem_dma_unmap_page_attrs(struct device *dev, + dma_addr_t addr, size_t size, + enum dma_data_direction dir, + unsigned long attrs) +{ + if (!addr) + return; + + dma_unmap_page_attrs(dev, addr, size, dir, attrs); +} + #endif /* _NET_NETMEM_H */ --=20 2.49.0.rc0.332.g42c0ae87b1-goog From nobody Sun Feb 8 02:41:22 2026 Received: from mail-oo1-f73.google.com (mail-oo1-f73.google.com [209.85.161.73]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 121172163A5 for ; Sat, 8 Mar 2025 21:40:59 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.161.73 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470062; cv=none; b=lmWuUcmhkt+8eNWMJDXUhKI32erlUqwnm8nffuzJiPB2THinE+hfFiIOf3tfy4wcVYltKn6dDcTaPmxIs7WhXrdtIdFBc5QaSzApo7JXPO1OSAtkQAY82hMXoD4z8aqmTnLzGJPb6AjPUWSM+D3/mLNrzZ+OIIVtMwH2vxSLTR0= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470062; c=relaxed/simple; bh=gHHNZtQlI1ho4RJZpoBeZXSIeWU2nyDxuNd9wfJYYzs=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=R2+ebpmU8xbyNlY9pWYiZB+eWf8x7pl8/yGg7jebQaaVplSJTwcvrf5vrAfXyVgy7GymeKAUJkgDqGNGAymwJ1MleawDHr3UK+SEW0nsO/K78yzafSvn6MEMYTezSFB+/+CE04qdz00chSbxDtZV73WUWg1ySfKAsRms88e/KBw= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=yfozbKL7; arc=none smtp.client-ip=209.85.161.73 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="yfozbKL7" Received: by mail-oo1-f73.google.com with SMTP id 006d021491bc7-5fea43da180so2985960eaf.3 for ; Sat, 08 Mar 2025 13:40:59 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470059; x=1742074859; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=8MBZgO9/mMLxEqMacEbQemluc92d4gH7fDKPm4qz0ro=; b=yfozbKL7eN1ZxKNLGVCZ0T8OnLZQnN5qpz0YtFGm6BBF4limlekcg8TdoWS+IYQJmJ XhNaGhCeaj40GcgIktp9kdpoc4UwghS5Bqk0CznleyFcn28dTHtLg63cd24IdPyV/8Kg v/YuBQa2Ng6tPyIcj9kPjoMujaDPWQM28Y4zXGRcmMgZC5oLgvgG12ZgdxRh7kXK/XUk BXCXJV+tf5KztXhcv6ieDzodhZ606wMcqzCHcocOXvWogRbh6NTIDXDIyrJawvHYouiV vIDMA5FNeC7mHEizaClb5moqFYQ81vDcHE5BQi+Nv17hAcm1XYNrgrSt8wvXNzd/otjh gVWQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470059; x=1742074859; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=8MBZgO9/mMLxEqMacEbQemluc92d4gH7fDKPm4qz0ro=; b=KLtgaY3KESnN8pc12NUvjMcRKaoZIZqa+38gZlsnfHHgfMaIE5tsgKVyVhRYyVR0SP CE+wAkp/hAiHhIYn66G3b6GSR0N3bpVA8ySYA+ymsNPeWgvhjxTAHFfX9qQB+40v5ki8 EH1EK0SJBuz4zlqxncQx+Ev85+ivMTfYk3/h+0oP60iV9P/FsPYqjKtYBxBHNPnLZrCN 3Uai1IPM/55fcnNCiB9NOrQBmZ9MAHhqgwa/ObeGr8P6B75056O3nPuB9/9J5DMUQyo+ WZn3t9f5ZklGrTD3/hrU5C7r3oFLkiYjjF5sID6yqf0PtMNqTli8wnbIkj5N0Ea8sbd7 Gg6g== X-Forwarded-Encrypted: i=1; AJvYcCXVshSDKy3UgKbufaztFwDVGrExs1KWTq2UwX8qX7N5W1TTDDWDCAc9vwuAPFkhbcct/ndhBcwEApFvN40=@vger.kernel.org X-Gm-Message-State: AOJu0YyG5CN2JzpiUfrrTvsktXRyWfo956WTXZ5Fl03XIcPIcaphR5Ut ezKH1B2FDe79hvKbkmtR2nYd73RXS/CvuJb812SlszmxRRJwjzyUu0w0ag0ryP5hvZorxgZGdyv eWV6C8jUbuiDIbqhl0bFiIA== X-Google-Smtp-Source: AGHT+IGTSt/plByX/+92dhEDUet8TI73ntJDZTnO/NP8eEfo8nhvCGID+mTdHpAuMZUG8Bn86E6QbOxlIYhAKteIYA== X-Received: from oablw17.prod.google.com ([2002:a05:6870:8e11:b0:2c1:64e4:adb9]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6870:4d12:b0:2bc:66cc:1507 with SMTP id 586e51a60fabf-2c2610714b7mr4380461fac.12.1741470058886; Sat, 08 Mar 2025 13:40:58 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:43 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-8-almasrymina@google.com> Subject: [PATCH net-next v7 7/9] gve: add netmem TX support to GVE DQO-RDA mode From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Use netmem_dma_*() helpers in gve_tx_dqo.c DQO-RDA paths to enable netmem TX support in that mode. Declare support for netmem TX in GVE DQO-RDA mode. Signed-off-by: Mina Almasry Tested-by: Lei Yang --- v4: - New patch --- drivers/net/ethernet/google/gve/gve_main.c | 4 ++++ drivers/net/ethernet/google/gve/gve_tx_dqo.c | 8 +++++--- 2 files changed, 9 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/google/gve/gve_main.c b/drivers/net/ether= net/google/gve/gve_main.c index 6dcdcaf518f4..4e95aee1bb4c 100644 --- a/drivers/net/ethernet/google/gve/gve_main.c +++ b/drivers/net/ethernet/google/gve/gve_main.c @@ -2817,6 +2817,10 @@ static int gve_probe(struct pci_dev *pdev, const str= uct pci_device_id *ent) =20 dev_info(&pdev->dev, "GVE version %s\n", gve_version_str); dev_info(&pdev->dev, "GVE queue format %d\n", (int)priv->queue_format); + + if (!gve_is_gqi(priv) && !gve_is_qpl(priv)) + dev->netmem_tx =3D true; + gve_clear_probe_in_progress(priv); queue_work(priv->gve_wq, &priv->service_task); return 0; diff --git a/drivers/net/ethernet/google/gve/gve_tx_dqo.c b/drivers/net/eth= ernet/google/gve/gve_tx_dqo.c index 394debc62268..e74580dc7ebe 100644 --- a/drivers/net/ethernet/google/gve/gve_tx_dqo.c +++ b/drivers/net/ethernet/google/gve/gve_tx_dqo.c @@ -667,7 +667,8 @@ static int gve_tx_add_skb_no_copy_dqo(struct gve_tx_rin= g *tx, goto err; =20 dma_unmap_len_set(pkt, len[pkt->num_bufs], len); - dma_unmap_addr_set(pkt, dma[pkt->num_bufs], addr); + netmem_dma_unmap_addr_set(skb_frag_netmem(frag), pkt, + dma[pkt->num_bufs], addr); ++pkt->num_bufs; =20 gve_tx_fill_pkt_desc_dqo(tx, desc_idx, skb, len, addr, @@ -1045,8 +1046,9 @@ static void gve_unmap_packet(struct device *dev, dma_unmap_single(dev, dma_unmap_addr(pkt, dma[0]), dma_unmap_len(pkt, len[0]), DMA_TO_DEVICE); for (i =3D 1; i < pkt->num_bufs; i++) { - dma_unmap_page(dev, dma_unmap_addr(pkt, dma[i]), - dma_unmap_len(pkt, len[i]), DMA_TO_DEVICE); + netmem_dma_unmap_page_attrs(dev, dma_unmap_addr(pkt, dma[i]), + dma_unmap_len(pkt, len[i]), + DMA_TO_DEVICE, 0); } pkt->num_bufs =3D 0; } --=20 2.49.0.rc0.332.g42c0ae87b1-goog From nobody Sun Feb 8 02:41:22 2026 Received: from mail-pj1-f73.google.com (mail-pj1-f73.google.com [209.85.216.73]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 1AEBA216600 for ; Sat, 8 Mar 2025 21:41:00 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.73 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470063; cv=none; b=nFXL0R8peGTiPe4zgRR/2WwQSJONa4WLFJmceY31S+kOCBB4zv1h7CNbuNk1WUbp0YZLg4/72yE5PNU57r8xRBMSF37TnMFw3WS9J2hIQ3jscAb1sYzCfDIGUz9rNXLce4EHmPNSvC/WNuxwYkQwL5LXWT8EdkWPb8NKuPwhMko= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470063; c=relaxed/simple; bh=5hyZYxezWmmiBT1IswLrJGF/zzJLGbfExZqICWkHdIw=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=pikwU7OW43AGyvwbNC2eWfvVyEiXS8GEZUAP65GwsiVT0CRu/PNYTep0HD9lVk9bDdDpVyewqQe8jYKvm7X8jsqgSfUoLGB7Jkuil9CIPHVAdZiqDxnqZKAHQkS4iBxyZTMQop/1oDERAEZvFSiZgZiasVA4rRVP71hgp4emwdM= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=s4BBJbtT; arc=none smtp.client-ip=209.85.216.73 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="s4BBJbtT" Received: by mail-pj1-f73.google.com with SMTP id 98e67ed59e1d1-2ff53a4754aso8480568a91.2 for ; Sat, 08 Mar 2025 13:41:00 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470060; x=1742074860; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=8Wr/uTxLJd/vHEpArrWTAzyhW/+/hW24sD4ijoGNq1A=; b=s4BBJbtTwm6eQD36oIq9HpUkBHPKqYzkliXmWzfP98NKO6GQvFE89kNF3AudqWFO6S atYh4Spxz5F2R3ZuxVXOzokIF+yfytye6L8usPzMB1JlDU8GJMAhAgr5YFBNgz25KVFD FuEDt7UOKyE5301Dw0McpWPGmf7nudkViITkm3PlDMVRwzJS4MsMtmfCGaROAeZ+8fDK moy/v9x61N3pjwcnM0kXWO7s9ahIX8WExidY0+VWdBMxMxceYieBbeLzwpodxiWZXo1M PtTMvvJIUTyE3rF7okF8Pn3ymL2VYzkqdl+puP3XHNAcMmqCzHTFEcuzJiq01vSY2Gya i62g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470060; x=1742074860; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=8Wr/uTxLJd/vHEpArrWTAzyhW/+/hW24sD4ijoGNq1A=; b=egO0us41XKrxgD+l9EqTYvr/U6Tqh/69fJMfI6zSkMtZc8eLzgXR/EtnCMqt1FOgK4 nxE4PxGV4ccPImbRshvYfwcfzviA9HFWrprcyt0NWoN67J+VRIaTbPVw6WXdMAOtXJjh 93eKa1f/NvmvV8y0IIoTuhbYTtXOJj+56+ZHGzKjbkdpk8YY/CPwgVj3pJLnR+wqKuYQ e6Om+9LjLEFUqN8tB+nNUAsvywSqooR/WmBby2Fq6fwtqb6dO+3+kyc/8fZ2ajcZjk0e WKY8YJAhaDEf6mKZHKSz0O9tivO2JD0JaP/LQ5oo9zzl46rUu5E3q2AFpUvNp8PHzfjN E+tQ== X-Forwarded-Encrypted: i=1; AJvYcCU1nrdZlMwt1aJQMC3LfKlFbBMBf7BwkFlInkl/NqLIga6DoHVneCwuura6JmR0JGSpyh95tfclz2pm/LI=@vger.kernel.org X-Gm-Message-State: AOJu0YwgzfW+rNvJjafQP1/JQlrq5IcVERImdtOAdbKWHnADNZk+kOBN zBq7jkia9yG+Pbf/5YiOTrxJtm8ffp0sFKPD7mlpLcblixd/Fgq5SUXYPW7Kr+tLfVgp/DQhxnR /Noe3WYiNasoLjk7Zg2GqJg== X-Google-Smtp-Source: AGHT+IGyjeNA1taw2RdrwgYQli42mPxkFOYVzCwbtW6YJ7SSHLfeooUxpgj7xkNuM+/XP23C+mt2BUcMUhn7WAB/+A== X-Received: from pjbqn14.prod.google.com ([2002:a17:90b:3d4e:b0:2ea:9d23:79a0]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90b:1b05:b0:2ee:8427:4b02 with SMTP id 98e67ed59e1d1-2ff7cef76acmr12864048a91.28.1741470060429; Sat, 08 Mar 2025 13:41:00 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:44 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-9-almasrymina@google.com> Subject: [PATCH net-next v7 8/9] net: check for driver support in netmem TX From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" We should not enable netmem TX for drivers that don't declare support. Check for driver netmem TX support during devmem TX binding and fail if the driver does not have the functionality. Check for driver support in validate_xmit_skb as well. Signed-off-by: Mina Almasry Acked-by: Stanislav Fomichev Tested-by: Lei Yang --- v5: https://lore.kernel.org/netdev/20250227041209.2031104-8-almasrymina@goo= gle.com/ - Check that the dmabuf mappings belongs to the specific device the TX is being sent from (Jakub) v4: - New patch --- net/core/dev.c | 33 +++++++++++++++++++++++++++++++++ net/core/devmem.h | 6 ++++++ net/core/netdev-genl.c | 7 +++++++ 3 files changed, 46 insertions(+) diff --git a/net/core/dev.c b/net/core/dev.c index 1cb134ff7327..5553947123a0 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3868,10 +3868,43 @@ int skb_csum_hwoffload_help(struct sk_buff *skb, } EXPORT_SYMBOL(skb_csum_hwoffload_help); =20 +static struct sk_buff *validate_xmit_unreadable_skb(struct sk_buff *skb, + struct net_device *dev) +{ + struct skb_shared_info *shinfo; + struct net_iov *niov; + + if (likely(skb_frags_readable(skb))) + goto out; + + if (likely(!dev->netmem_tx)) + goto out_free; + + shinfo =3D skb_shinfo(skb); + + if (shinfo->nr_frags > 0) { + niov =3D netmem_to_net_iov(skb_frag_netmem(&shinfo->frags[0])); + if (net_is_devmem_iov(niov) && + net_devmem_iov_binding(niov)->dev !=3D dev) + goto out_free; + } + +out: + return skb; + +out_free: + kfree_skb(skb); + return NULL; +} + static struct sk_buff *validate_xmit_skb(struct sk_buff *skb, struct net_d= evice *dev, bool *again) { netdev_features_t features; =20 + skb =3D validate_xmit_unreadable_skb(skb, dev); + if (unlikely(!skb)) + goto out_null; + features =3D netif_skb_features(skb); skb =3D validate_xmit_vlan(skb, features); if (unlikely(!skb)) diff --git a/net/core/devmem.h b/net/core/devmem.h index 67168aae5e5b..919e6ed28fdc 100644 --- a/net/core/devmem.h +++ b/net/core/devmem.h @@ -229,6 +229,12 @@ net_devmem_get_niov_at(struct net_devmem_dmabuf_bindin= g *binding, size_t addr, { return NULL; } + +static inline struct net_devmem_dmabuf_binding * +net_devmem_iov_binding(const struct net_iov *niov) +{ + return NULL; +} #endif =20 #endif /* _NET_DEVMEM_H */ diff --git a/net/core/netdev-genl.c b/net/core/netdev-genl.c index 6e7cd6a5c177..6c5d62df0d65 100644 --- a/net/core/netdev-genl.c +++ b/net/core/netdev-genl.c @@ -972,6 +972,13 @@ int netdev_nl_bind_tx_doit(struct sk_buff *skb, struct= genl_info *info) goto err_unlock; } =20 + if (!netdev->netmem_tx) { + err =3D -EOPNOTSUPP; + NL_SET_ERR_MSG(info->extack, + "Driver does not support netmem TX"); + goto err_unlock; + } + binding =3D net_devmem_bind_dmabuf(netdev, DMA_TO_DEVICE, dmabuf_fd, info->extack); if (IS_ERR(binding)) { --=20 2.49.0.rc0.332.g42c0ae87b1-goog From nobody Sun Feb 8 02:41:22 2026 Received: from mail-pj1-f74.google.com (mail-pj1-f74.google.com [209.85.216.74]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id CAE18217656 for ; Sat, 8 Mar 2025 21:41:02 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.74 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470065; cv=none; b=DJ9Be605Rdidm9gogdIF1FRI0AjQI9LAaxgZoqFO3qUiTpP5VDdQ69MR0iNo0Xy8c/Iaju53/8PgsOSNHjoWv/m13SvHB7T8Qtq38CqyH/tSdQ+b1di/7SNYB7xGz+F8nIYtCnd1Uwdoe4r5qVEV3Jog8VIXxuSYWfbsfre2g3U= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741470065; c=relaxed/simple; bh=f3/4i8+4lq8xrFjfnctzLJZ2Vew1NUkXraPbpumger0=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=GSu/OSqo7Y/GIctil9x1ovM6zpzS9Z/2ycRyXIg4Kk0n17fFnGx87COfDGXLgVmJ/qC+SrP9HF5jNs4n8oOpiOSkZ5v1bmTOsZy4U/kfp0S9iNdCAlrITQgAPQ7u9u1JWi2gjXguYwowJIAUc7Eod7Nv3WiP9+a6qdI246pXC54= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=MIcWz81O; arc=none smtp.client-ip=209.85.216.74 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="MIcWz81O" Received: by mail-pj1-f74.google.com with SMTP id 98e67ed59e1d1-2fecd603bb8so5192121a91.3 for ; Sat, 08 Mar 2025 13:41:02 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1741470062; x=1742074862; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=yT18+yNp0G09gKREpoqxKnlnqPHvsdCe72nQqesZ4Kk=; b=MIcWz81OPAuwg3taZ4zAx5xSXVIKjAUVh9eqsXLbh28VSKvgiFsuJEdyssuC/6W8HW Z3TOpzfJzULXfwc/tKVk+Q9BGQ59VEUITNinwjhiHlBFRszW+2PsCGLpLbHrzx3lH0KO 2B+tao1AuqtbycVPjGPIdJuNfFDm/cDhLneflv/UZ3vtIOxYu8x1dmmLBDSX+HP0qmGg yq3avSDthOeKxIoi5ha0SDEMcQST05RQvRztkLM33Ml/ibixtmRyoN9sZtUHZ9AIw8F4 J2SYxJf0HO4spRZOkIbTm1pyiCwZAP8S9PJ2yt4wScweX3Pg/s05pcHrxC19BuRTMmYs PIag== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1741470062; x=1742074862; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=yT18+yNp0G09gKREpoqxKnlnqPHvsdCe72nQqesZ4Kk=; b=FTUSJxTz7LeS3folKGlvxnENakFjIU6Q6p3B1j37TyM+PhLQ5jumgyAdl7bz7CkRUb u6TDCH77NYsGdsiLeM3HRY9yFqYr9p+F2BKQg5BYyioZuvTdTMqwtdJ0lx2hLiXHIlAc 7wyeobwSstv+jwcApACBlxTx2TGo1yn7QEvXSgVW+e6Ri7yiUggT+kqJ4U1MxWoI9hCH hUMP2kYA3elrMsMByB9Kidl2MzE8a1kRUVdcck/b2rhqqXKgP2YAlPZTYbleZehmg251 /6o9z7726tKl8G9KDEvuyvWTiBzwKYMcocSxEtR2yFaaBBr88Fgm91qYuKF15FHG0Xhu KpYw== X-Forwarded-Encrypted: i=1; AJvYcCXx5CgetGJS5Rpq9z35cbihKSBRZKprGEYpQChl1DFt+i7H6vr33dTMnHUh79MobQBJhKIUqx00cH3+76Y=@vger.kernel.org X-Gm-Message-State: AOJu0Yx7hhGLpT0hdXB6OAHWDCnNQhYhZOS2gXWP2s3dpaxk0ltzNjfs 8SQojFkLipfiB9rKzypgUmVWSnOS4+MAy0026kggaAldogTWeux27lJrJpfusZpbmmu8uhDu3ye IPukeHkgJSCbN7sUxSkL+1g== X-Google-Smtp-Source: AGHT+IFE12Qlv3QvvPMel+MqZLR2aYyOinhSvgix1CfB9ElAh2jhVqjROglhCR1VOqSbHd0lBIP0/ve5Fqc6EhHNBQ== X-Received: from pgbcf2.prod.google.com ([2002:a05:6a02:842:b0:adf:4827:b70e]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6a21:497:b0:1ee:ced0:f0a4 with SMTP id adf61e73a8af0-1f544c377b2mr16751104637.34.1741470062099; Sat, 08 Mar 2025 13:41:02 -0800 (PST) Date: Sat, 8 Mar 2025 21:40:45 +0000 In-Reply-To: <20250308214045.1160445-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250308214045.1160445-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.rc0.332.g42c0ae87b1-goog Message-ID: <20250308214045.1160445-10-almasrymina@google.com> Subject: [PATCH net-next v7 9/9] selftests: ncdevmem: Implement devmem TCP TX From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux.dev, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , David Ahern , Neal Cardwell , Stefan Hajnoczi , Stefano Garzarella , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Shuah Khan , sdf@fomichev.me, asml.silence@gmail.com, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Add support for devmem TX in ncdevmem. This is a combination of the ncdevmem from the devmem TCP series RFCv1 which included the TX path, and work by Stan to include the netlink API and refactored on top of his generic memory_provider support. Signed-off-by: Mina Almasry Signed-off-by: Stanislav Fomichev Acked-by: Stanislav Fomichev Tested-by: Lei Yang --- v5: - Remove unnecassyr socat bindings (Stan). - Add exit_wait=3DTrue (Stan) - Remove unnecessary -c arg to ncdevmem in check_tx. v4: - Add TX test to devmem.py (Paolo). v3: - Update ncdevmem docs to run validation with RX-only and RX-with-TX. - Fix build warnings (Stan). - Make the validation expect new lines in the pattern so we can have the TX path behave like netcat (Stan). - Change ret to errno in error() calls (Stan). - Handle the case where client_ip is not provided (Stan). - Don't assume mid is <=3D 2000 (Stan). v2: - make errors a static variable so that we catch instances where there are less than 20 errors across different buffers. - Fix the issue where the seed is reset to 0 instead of its starting value 1. - Use 1000ULL instead of 1000 to guard against overflow (Willem). - Do not set POLLERR (Willem). - Update the test to use the new interface where iov_base is the dmabuf_offset. - Update the test to send 2 iov instead of 1, so we get some test coverage over sending multiple iovs at once. - Print the ifindex the test is using, useful for debugging issues where maybe the test may fail because the ifindex of the socket is different from the dmabuf binding. --- .../selftests/drivers/net/hw/devmem.py | 26 +- .../selftests/drivers/net/hw/ncdevmem.c | 300 +++++++++++++++++- 2 files changed, 311 insertions(+), 15 deletions(-) diff --git a/tools/testing/selftests/drivers/net/hw/devmem.py b/tools/testi= ng/selftests/drivers/net/hw/devmem.py index 3947e9157115..7fc686cf47a2 100755 --- a/tools/testing/selftests/drivers/net/hw/devmem.py +++ b/tools/testing/selftests/drivers/net/hw/devmem.py @@ -1,6 +1,7 @@ #!/usr/bin/env python3 # SPDX-License-Identifier: GPL-2.0 =20 +from os import path from lib.py import ksft_run, ksft_exit from lib.py import ksft_eq, KsftSkipEx from lib.py import NetDrvEpEnv @@ -10,8 +11,7 @@ from lib.py import ksft_disruptive =20 def require_devmem(cfg): if not hasattr(cfg, "_devmem_probed"): - port =3D rand_port() - probe_command =3D f"./ncdevmem -f {cfg.ifname}" + probe_command =3D f"{cfg.bin_local} -f {cfg.ifname}" cfg._devmem_supported =3D cmd(probe_command, fail=3DFalse, shell= =3DTrue).ret =3D=3D 0 cfg._devmem_probed =3D True =20 @@ -25,7 +25,7 @@ def check_rx(cfg) -> None: require_devmem(cfg) =20 port =3D rand_port() - listen_cmd =3D f"./ncdevmem -l -f {cfg.ifname} -s {cfg.addr_v['6']} -p= {port}" + listen_cmd =3D f"{cfg.bin_local} -l -f {cfg.ifname} -s {cfg.addr_v['6'= ]} -p {port}" =20 with bkg(listen_cmd) as socat: wait_port_listen(port) @@ -34,9 +34,27 @@ def check_rx(cfg) -> None: ksft_eq(socat.stdout.strip(), "hello\nworld") =20 =20 +@ksft_disruptive +def check_tx(cfg) -> None: + cfg.require_ipver("6") + require_devmem(cfg) + + port =3D rand_port() + listen_cmd =3D f"socat -U - TCP6-LISTEN:{port}" + + with bkg(listen_cmd, exit_wait=3DTrue) as socat: + wait_port_listen(port) + cmd(f"echo -e \"hello\\nworld\"| {cfg.bin_remote} -f {cfg.ifname} = -s {cfg.addr_v['6']} -p {port}", host=3Dcfg.remote, shell=3DTrue) + + ksft_eq(socat.stdout.strip(), "hello\nworld") + + def main() -> None: with NetDrvEpEnv(__file__) as cfg: - ksft_run([check_rx], + cfg.bin_local =3D path.abspath(path.dirname(__file__) + "/ncdevmem= ") + cfg.bin_remote =3D cfg.remote.deploy(cfg.bin_local) + + ksft_run([check_rx, check_tx], args=3D(cfg, )) ksft_exit() =20 diff --git a/tools/testing/selftests/drivers/net/hw/ncdevmem.c b/tools/test= ing/selftests/drivers/net/hw/ncdevmem.c index 2bf14ac2b8c6..f801a1b3545f 100644 --- a/tools/testing/selftests/drivers/net/hw/ncdevmem.c +++ b/tools/testing/selftests/drivers/net/hw/ncdevmem.c @@ -9,22 +9,31 @@ * ncdevmem -s [-c ] -f eth1 -l -p 5201 * * On client: - * echo -n "hello\nworld" | nc -s 5201 -p 5201 + * echo -n "hello\nworld" | \ + * ncdevmem -s [-c ] -p 5201 -f eth1 * - * Test data validation: + * Note this is compatible with regular netcat. i.e. the sender or receive= r can + * be replaced with regular netcat to test the RX or TX path in isolation. + * + * Test data validation (devmem TCP on RX only): * * On server: * ncdevmem -s [-c ] -f eth1 -l -p 5201 -v 7 * * On client: * yes $(echo -e \\x01\\x02\\x03\\x04\\x05\\x06) | \ - * tr \\n \\0 | \ - * head -c 5G | \ + * head -c 1G | \ * nc 5201 -p 5201 * + * Test data validation (devmem TCP on RX and TX, validation happens on RX= ): * - * Note this is compatible with regular netcat. i.e. the sender or receive= r can - * be replaced with regular netcat to test the RX or TX path in isolation. + * On server: + * ncdevmem -s [-c ] -l -p 5201 -v 8 -f eth1 + * + * On client: + * yes $(echo -e \\x01\\x02\\x03\\x04\\x05\\x06\\x07) | \ + * head -c 1M | \ + * ncdevmem -s [-c ] -p 5201 -f eth1 */ #define _GNU_SOURCE #define __EXPORTED_HEADERS__ @@ -40,15 +49,18 @@ #include #include #include +#include =20 #include #include #include #include #include +#include =20 #include #include +#include #include #include #include @@ -79,6 +91,8 @@ static int num_queues =3D -1; static char *ifname; static unsigned int ifindex; static unsigned int dmabuf_id; +static uint32_t tx_dmabuf_id; +static int waittime_ms =3D 500; =20 struct memory_buffer { int fd; @@ -92,6 +106,8 @@ struct memory_buffer { struct memory_provider { struct memory_buffer *(*alloc)(size_t size); void (*free)(struct memory_buffer *ctx); + void (*memcpy_to_device)(struct memory_buffer *dst, size_t off, + void *src, int n); void (*memcpy_from_device)(void *dst, struct memory_buffer *src, size_t off, int n); }; @@ -152,6 +168,20 @@ static void udmabuf_free(struct memory_buffer *ctx) free(ctx); } =20 +static void udmabuf_memcpy_to_device(struct memory_buffer *dst, size_t off, + void *src, int n) +{ + struct dma_buf_sync sync =3D {}; + + sync.flags =3D DMA_BUF_SYNC_START | DMA_BUF_SYNC_WRITE; + ioctl(dst->fd, DMA_BUF_IOCTL_SYNC, &sync); + + memcpy(dst->buf_mem + off, src, n); + + sync.flags =3D DMA_BUF_SYNC_END | DMA_BUF_SYNC_WRITE; + ioctl(dst->fd, DMA_BUF_IOCTL_SYNC, &sync); +} + static void udmabuf_memcpy_from_device(void *dst, struct memory_buffer *sr= c, size_t off, int n) { @@ -169,6 +199,7 @@ static void udmabuf_memcpy_from_device(void *dst, struc= t memory_buffer *src, static struct memory_provider udmabuf_memory_provider =3D { .alloc =3D udmabuf_alloc, .free =3D udmabuf_free, + .memcpy_to_device =3D udmabuf_memcpy_to_device, .memcpy_from_device =3D udmabuf_memcpy_from_device, }; =20 @@ -187,14 +218,16 @@ void validate_buffer(void *line, size_t size) { static unsigned char seed =3D 1; unsigned char *ptr =3D line; - int errors =3D 0; + unsigned char expected; + static int errors; size_t i; =20 for (i =3D 0; i < size; i++) { - if (ptr[i] !=3D seed) { + expected =3D seed ? seed : '\n'; + if (ptr[i] !=3D expected) { fprintf(stderr, "Failed validation: expected=3D%u, actual=3D%u, index=3D%lu\n", - seed, ptr[i], i); + expected, ptr[i], i); errors++; if (errors > 20) error(1, 0, "validation failed."); @@ -393,6 +426,49 @@ static int bind_rx_queue(unsigned int ifindex, unsigne= d int dmabuf_fd, return -1; } =20 +static int bind_tx_queue(unsigned int ifindex, unsigned int dmabuf_fd, + struct ynl_sock **ys) +{ + struct netdev_bind_tx_req *req =3D NULL; + struct netdev_bind_tx_rsp *rsp =3D NULL; + struct ynl_error yerr; + + *ys =3D ynl_sock_create(&ynl_netdev_family, &yerr); + if (!*ys) { + fprintf(stderr, "YNL: %s\n", yerr.msg); + return -1; + } + + req =3D netdev_bind_tx_req_alloc(); + netdev_bind_tx_req_set_ifindex(req, ifindex); + netdev_bind_tx_req_set_fd(req, dmabuf_fd); + + rsp =3D netdev_bind_tx(*ys, req); + if (!rsp) { + perror("netdev_bind_tx"); + goto err_close; + } + + if (!rsp->_present.id) { + perror("id not present"); + goto err_close; + } + + fprintf(stderr, "got tx dmabuf id=3D%d\n", rsp->id); + tx_dmabuf_id =3D rsp->id; + + netdev_bind_tx_req_free(req); + netdev_bind_tx_rsp_free(rsp); + + return 0; + +err_close: + fprintf(stderr, "YNL failed: %s\n", (*ys)->err.msg); + netdev_bind_tx_req_free(req); + ynl_sock_destroy(*ys); + return -1; +} + static void enable_reuseaddr(int fd) { int opt =3D 1; @@ -431,7 +507,7 @@ static int parse_address(const char *str, int port, str= uct sockaddr_in6 *sin6) return 0; } =20 -int do_server(struct memory_buffer *mem) +static int do_server(struct memory_buffer *mem) { char ctrl_data[sizeof(int) * 20000]; struct netdev_queue_id *queues; @@ -685,6 +761,206 @@ void run_devmem_tests(void) provider->free(mem); } =20 +static uint64_t gettimeofday_ms(void) +{ + struct timeval tv; + + gettimeofday(&tv, NULL); + return (tv.tv_sec * 1000ULL) + (tv.tv_usec / 1000ULL); +} + +static int do_poll(int fd) +{ + struct pollfd pfd; + int ret; + + pfd.revents =3D 0; + pfd.fd =3D fd; + + ret =3D poll(&pfd, 1, waittime_ms); + if (ret =3D=3D -1) + error(1, errno, "poll"); + + return ret && (pfd.revents & POLLERR); +} + +static void wait_compl(int fd) +{ + int64_t tstop =3D gettimeofday_ms() + waittime_ms; + char control[CMSG_SPACE(100)] =3D {}; + struct sock_extended_err *serr; + struct msghdr msg =3D {}; + struct cmsghdr *cm; + __u32 hi, lo; + int ret; + + msg.msg_control =3D control; + msg.msg_controllen =3D sizeof(control); + + while (gettimeofday_ms() < tstop) { + if (!do_poll(fd)) + continue; + + ret =3D recvmsg(fd, &msg, MSG_ERRQUEUE); + if (ret < 0) { + if (errno =3D=3D EAGAIN) + continue; + error(1, errno, "recvmsg(MSG_ERRQUEUE)"); + return; + } + if (msg.msg_flags & MSG_CTRUNC) + error(1, 0, "MSG_CTRUNC\n"); + + for (cm =3D CMSG_FIRSTHDR(&msg); cm; cm =3D CMSG_NXTHDR(&msg, cm)) { + if (cm->cmsg_level !=3D SOL_IP && + cm->cmsg_level !=3D SOL_IPV6) + continue; + if (cm->cmsg_level =3D=3D SOL_IP && + cm->cmsg_type !=3D IP_RECVERR) + continue; + if (cm->cmsg_level =3D=3D SOL_IPV6 && + cm->cmsg_type !=3D IPV6_RECVERR) + continue; + + serr =3D (void *)CMSG_DATA(cm); + if (serr->ee_origin !=3D SO_EE_ORIGIN_ZEROCOPY) + error(1, 0, "wrong origin %u", serr->ee_origin); + if (serr->ee_errno !=3D 0) + error(1, 0, "wrong errno %d", serr->ee_errno); + + hi =3D serr->ee_data; + lo =3D serr->ee_info; + + fprintf(stderr, "tx complete [%d,%d]\n", lo, hi); + return; + } + } + + error(1, 0, "did not receive tx completion"); +} + +static int do_client(struct memory_buffer *mem) +{ + char ctrl_data[CMSG_SPACE(sizeof(__u32))]; + struct sockaddr_in6 server_sin; + struct sockaddr_in6 client_sin; + struct ynl_sock *ys =3D NULL; + struct msghdr msg =3D {}; + ssize_t line_size =3D 0; + struct cmsghdr *cmsg; + struct iovec iov[2]; + char *line =3D NULL; + unsigned long mid; + size_t len =3D 0; + int socket_fd; + __u32 ddmabuf; + int opt =3D 1; + int ret; + + ret =3D parse_address(server_ip, atoi(port), &server_sin); + if (ret < 0) + error(1, 0, "parse server address"); + + socket_fd =3D socket(AF_INET6, SOCK_STREAM, 0); + if (socket_fd < 0) + error(1, socket_fd, "create socket"); + + enable_reuseaddr(socket_fd); + + ret =3D setsockopt(socket_fd, SOL_SOCKET, SO_BINDTODEVICE, ifname, + strlen(ifname) + 1); + if (ret) + error(1, errno, "bindtodevice"); + + if (bind_tx_queue(ifindex, mem->fd, &ys)) + error(1, 0, "Failed to bind\n"); + + if (client_ip) { + ret =3D parse_address(client_ip, atoi(port), &client_sin); + if (ret < 0) + error(1, 0, "parse client address"); + + ret =3D bind(socket_fd, &client_sin, sizeof(client_sin)); + if (ret) + error(1, errno, "bind"); + } + + ret =3D setsockopt(socket_fd, SOL_SOCKET, SO_ZEROCOPY, &opt, sizeof(opt)); + if (ret) + error(1, errno, "set sock opt"); + + fprintf(stderr, "Connect to %s %d (via %s)\n", server_ip, + ntohs(server_sin.sin6_port), ifname); + + ret =3D connect(socket_fd, &server_sin, sizeof(server_sin)); + if (ret) + error(1, errno, "connect"); + + while (1) { + free(line); + line =3D NULL; + line_size =3D getline(&line, &len, stdin); + + if (line_size < 0) + break; + + mid =3D (line_size / 2) + 1; + + iov[0].iov_base =3D (void *)1; + iov[0].iov_len =3D mid; + iov[1].iov_base =3D (void *)(mid + 2); + iov[1].iov_len =3D line_size - mid; + + provider->memcpy_to_device(mem, (size_t)iov[0].iov_base, line, + iov[0].iov_len); + provider->memcpy_to_device(mem, (size_t)iov[1].iov_base, + line + iov[0].iov_len, + iov[1].iov_len); + + fprintf(stderr, + "read line_size=3D%ld iov[0].iov_base=3D%lu, iov[0].iov_len=3D%lu, iov[= 1].iov_base=3D%lu, iov[1].iov_len=3D%lu\n", + line_size, (unsigned long)iov[0].iov_base, + iov[0].iov_len, (unsigned long)iov[1].iov_base, + iov[1].iov_len); + + msg.msg_iov =3D iov; + msg.msg_iovlen =3D 2; + + msg.msg_control =3D ctrl_data; + msg.msg_controllen =3D sizeof(ctrl_data); + + cmsg =3D CMSG_FIRSTHDR(&msg); + cmsg->cmsg_level =3D SOL_SOCKET; + cmsg->cmsg_type =3D SCM_DEVMEM_DMABUF; + cmsg->cmsg_len =3D CMSG_LEN(sizeof(__u32)); + + ddmabuf =3D tx_dmabuf_id; + + *((__u32 *)CMSG_DATA(cmsg)) =3D ddmabuf; + + ret =3D sendmsg(socket_fd, &msg, MSG_ZEROCOPY); + if (ret < 0) + error(1, errno, "Failed sendmsg"); + + fprintf(stderr, "sendmsg_ret=3D%d\n", ret); + + if (ret !=3D line_size) + error(1, errno, "Did not send all bytes"); + + wait_compl(socket_fd); + } + + fprintf(stderr, "%s: tx ok\n", TEST_PREFIX); + + free(line); + close(socket_fd); + + if (ys) + ynl_sock_destroy(ys); + + return 0; +} + int main(int argc, char *argv[]) { struct memory_buffer *mem; @@ -728,6 +1004,8 @@ int main(int argc, char *argv[]) =20 ifindex =3D if_nametoindex(ifname); =20 + fprintf(stderr, "using ifindex=3D%u\n", ifindex); + if (!server_ip && !client_ip) { if (start_queue < 0 && num_queues < 0) { num_queues =3D rxq_num(ifindex); @@ -778,7 +1056,7 @@ int main(int argc, char *argv[]) error(1, 0, "Missing -p argument\n"); =20 mem =3D provider->alloc(getpagesize() * NUM_PAGES); - ret =3D is_server ? do_server(mem) : 1; + ret =3D is_server ? do_server(mem) : do_client(mem); provider->free(mem); =20 return ret; --=20 2.49.0.rc0.332.g42c0ae87b1-goog