[net-next,v1,10/13] net: group skb_shinfo zerocopy related bits together.

Message ID	20210105220706.998374-11-jonathan.lemon@gmail.com (mailing list archive)
State	Superseded
Delegated to:	Netdev Maintainers
Headers	show Return-Path: <netdev-owner@kernel.org> From: Jonathan Lemon <jonathan.lemon@gmail.com> To: <netdev@vger.kernel.org>, <davem@davemloft.net>, <kuba@kernel.org>, <willemdebruijn.kernel@gmail.com>, <edumazet@google.com>, <dsahern@gmail.com> CC: <kernel-team@fb.com> Subject: [PATCH net-next v1 10/13] net: group skb_shinfo zerocopy related bits together. Date: Tue, 5 Jan 2021 14:07:03 -0800 Message-ID: <20210105220706.998374-11-jonathan.lemon@gmail.com> In-Reply-To: <20210105220706.998374-1-jonathan.lemon@gmail.com> References: <20210105220706.998374-1-jonathan.lemon@gmail.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8BIT Content-Type: text/plain Precedence: bulk
Series	Generic zcopy_* functions \| expand [net-next,v1,00/13] Generic zcopy_* functions [net-next,v1,01/13] skbuff: remove unused skb_zcopy_abort function [net-next,v1,02/13] skbuff: simplify sock_zerocopy_put [net-next,v1,03/13] skbuff: Push status and refcounts into sock_zerocopy_callback [net-next,v1,04/13] skbuff: replace sock_zerocopy_put() with skb_zcopy_put() [net-next,v1,05/13] skbuff: replace sock_zerocopy_get with skb_zcopy_get [net-next,v1,06/13] skbuff: Add skb parameter to the ubuf zerocopy callback [net-next,v1,07/13] skbuff: Call sock_zerocopy_put_abort from skb_zcopy_put_abort [net-next,v1,08/13] skbuff: Call skb_zcopy_clear() before unref'ing fragments [net-next,v1,09/13] skbuff: rename sock_zerocopy_* to msg_zerocopy_* [net-next,v1,10/13] net: group skb_shinfo zerocopy related bits together. [net-next,v1,11/13] skbuff: add flags to ubuf_info for ubuf setup [net-next,v1,12/13] tap/tun: add skb_zcopy_init() helper for initialization. [net-next,v1,13/13] skbuff: Rename skb_zcopy_{get\|put} to net_zcopy_{get\|put}

Context	Check	Description
netdev/cover_letter	success	Link
netdev/fixes_present	success	Link
netdev/patch_count	success	Link
netdev/tree_selection	success	Clearly marked for net-next
netdev/subject_prefix	success	Link
netdev/cc_maintainers	warning	23 maintainers not CCed: steffen.klassert@secunet.com daniel@iogearbox.net pabeni@redhat.com tglx@linutronix.de herbert@gondor.apana.org.au gnault@redhat.com bsd@fb.com kyk.segfault@gmail.com rdunlap@infradead.org viro@zeniv.linux.org.uk wei.liu@kernel.org willemb@google.com linmiaohe@huawei.com paul@xen.org elver@google.com kuznet@ms2.inr.ac.ru pablo@netfilter.org ast@kernel.org nogikh@google.com yoshfuji@linux-ipv6.org xen-devel@lists.xenproject.org jakub@cloudflare.com decui@microsoft.com
netdev/source_inline	success	Was 0 now: 0
netdev/verify_signedoff	success	Link
netdev/module_param	success	Was 0 now: 0
netdev/build_32bit	success	Errors and warnings before: 8866 this patch: 8866
netdev/kdoc	success	Errors and warnings before: 0 this patch: 0
netdev/verify_fixes	success	Link
netdev/checkpatch	warning	WARNING: From:/Signed-off-by: email address mismatch: 'From: Jonathan Lemon <bsd@fb.com>' != 'Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>'
netdev/build_allmodconfig_warn	success	Errors and warnings before: 9273 this patch: 9273
netdev/header_inline	success	Link
netdev/stable	success	Stable not CCed

diff --git a/drivers/net/tap.c b/drivers/net/tap.c index 3f51f3766d18..f7a19d9b7c27 100644 --- a/drivers/net/tap.c +++ b/drivers/net/tap.c @@ -723,8 +723,7 @@ static ssize_t tap_get_user(struct tap_queue *q, void *msg_control, /* copy skb_ubuf_info for callback when skb has no error */ if (zerocopy) { skb_shinfo(skb)->destructor_arg = msg_control; - skb_shinfo(skb)->tx_flags |= SKBTX_DEV_ZEROCOPY; - skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG; + skb_shinfo(skb)->flags |= SKBFL_ZEROCOPY_FRAG; } else if (msg_control) { struct ubuf_info *uarg = msg_control; uarg->callback(NULL, uarg, false); diff --git a/drivers/net/tun.c b/drivers/net/tun.c index b64e2cc85751..dd9edbd72ae8 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -1815,8 +1815,7 @@ static ssize_t tun_get_user(struct tun_struct *tun, struct tun_file *tfile, /* copy skb_ubuf_info for callback when skb has no error */ if (zerocopy) { skb_shinfo(skb)->destructor_arg = msg_control; - skb_shinfo(skb)->tx_flags |= SKBTX_DEV_ZEROCOPY; - skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG; + skb_shinfo(skb)->flags |= SKBFL_ZEROCOPY_FRAG; } else if (msg_control) { struct ubuf_info *uarg = msg_control; uarg->callback(NULL, uarg, false); diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c index acb786d8b1d8..08b0e3d0b7eb 100644 --- a/drivers/net/xen-netback/interface.c +++ b/drivers/net/xen-netback/interface.c @@ -47,7 +47,7 @@ /* Number of bytes allowed on the internal guest Rx queue. */ #define XENVIF_RX_QUEUE_BYTES (XEN_NETIF_RX_RING_SIZE/2 * PAGE_SIZE) -/* This function is used to set SKBTX_DEV_ZEROCOPY as well as +/* This function is used to set SKBFL_ZEROCOPY_ENABLE as well as * increasing the inflight counter. We need to increase the inflight * counter because core driver calls into xenvif_zerocopy_callback * which calls xenvif_skb_zerocopy_complete. @@ -55,7 +55,7 @@ void xenvif_skb_zerocopy_prepare(struct xenvif_queue *queue, struct sk_buff *skb) { - skb_shinfo(skb)->tx_flags |= SKBTX_DEV_ZEROCOPY; + skb_shinfo(skb)->flags |= SKBFL_ZEROCOPY_ENABLE; atomic_inc(&queue->inflight_packets); } diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 3c34c75ab004..66fde9a7b851 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -430,28 +430,32 @@ enum { /* device driver is going to provide hardware time stamp */ SKBTX_IN_PROGRESS = 1 << 2, - /* device driver supports TX zero-copy buffers */ - SKBTX_DEV_ZEROCOPY = 1 << 3, - /* generate wifi status information (where possible) */ SKBTX_WIFI_STATUS = 1 << 4, - /* This indicates at least one fragment might be overwritten - * (as in vmsplice(), sendfile() ...) - * If we need to compute a TX checksum, we'll need to copy - * all frags to avoid possible bad checksum - */ - SKBTX_SHARED_FRAG = 1 << 5, - /* generate software time stamp when entering packet scheduling */ SKBTX_SCHED_TSTAMP = 1 << 6, }; -#define SKBTX_ZEROCOPY_FRAG (SKBTX_DEV_ZEROCOPY | SKBTX_SHARED_FRAG) #define SKBTX_ANY_SW_TSTAMP (SKBTX_SW_TSTAMP | \ SKBTX_SCHED_TSTAMP) #define SKBTX_ANY_TSTAMP (SKBTX_HW_TSTAMP | SKBTX_ANY_SW_TSTAMP) +/* Definitions for flags in struct skb_shared_info */ +enum { + /* use zcopy routines */ + SKBFL_ZEROCOPY_ENABLE = BIT(0), + + /* This indicates at least one fragment might be overwritten + * (as in vmsplice(), sendfile() ...) + * If we need to compute a TX checksum, we'll need to copy + * all frags to avoid possible bad checksum + */ + SKBFL_SHARED_FRAG = BIT(1), +}; + +#define SKBFL_ZEROCOPY_FRAG (SKBFL_ZEROCOPY_ENABLE | SKBFL_SHARED_FRAG) + /* * The callback notifies userspace to release buffers when skb DMA is done in * lower device, the skb last reference should be 0 when calling this. @@ -506,7 +510,7 @@ int skb_zerocopy_iter_stream(struct sock *sk, struct sk_buff *skb, * the end of the header data, ie. at skb->end. */ struct skb_shared_info { - __u8 __unused; + __u8 flags; __u8 meta_len; __u8 nr_frags; __u8 tx_flags; @@ -1433,7 +1437,7 @@ static inline struct skb_shared_hwtstamps *skb_hwtstamps(struct sk_buff *skb) static inline struct ubuf_info *skb_zcopy(struct sk_buff *skb) { - bool is_zcopy = skb && skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY; + bool is_zcopy = skb && skb_shinfo(skb)->flags & SKBFL_ZEROCOPY_ENABLE; return is_zcopy ? skb_uarg(skb) : NULL; } @@ -1452,14 +1456,14 @@ static inline void skb_zcopy_set(struct sk_buff *skb, struct ubuf_info *uarg, else skb_zcopy_get(uarg); skb_shinfo(skb)->destructor_arg = uarg; - skb_shinfo(skb)->tx_flags |= SKBTX_ZEROCOPY_FRAG; + skb_shinfo(skb)->flags |= SKBFL_ZEROCOPY_FRAG; } } static inline void skb_zcopy_set_nouarg(struct sk_buff *skb, void *val) { skb_shinfo(skb)->destructor_arg = (void *)((uintptr_t) val | 0x1UL); - skb_shinfo(skb)->tx_flags |= SKBTX_ZEROCOPY_FRAG; + skb_shinfo(skb)->flags |= SKBFL_ZEROCOPY_FRAG; } static inline bool skb_zcopy_is_nouarg(struct sk_buff *skb) @@ -1497,7 +1501,7 @@ static inline void skb_zcopy_clear(struct sk_buff *skb, bool zerocopy_success) if (!skb_zcopy_is_nouarg(skb)) uarg->callback(skb, uarg, zerocopy_success); - skb_shinfo(skb)->tx_flags &= ~SKBTX_ZEROCOPY_FRAG; + skb_shinfo(skb)->flags &= ~SKBFL_ZEROCOPY_FRAG; } } @@ -3323,7 +3327,7 @@ static inline int skb_linearize(struct sk_buff *skb) static inline bool skb_has_shared_frag(const struct sk_buff *skb) { return skb_is_nonlinear(skb) && - skb_shinfo(skb)->tx_flags & SKBTX_SHARED_FRAG; + skb_shinfo(skb)->flags & SKBFL_SHARED_FRAG; } /** diff --git a/net/core/skbuff.c b/net/core/skbuff.c index d520168accf9..124e4752afb6 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -1327,7 +1327,7 @@ static int skb_zerocopy_clone(struct sk_buff *nskb, struct sk_buff *orig, * @skb: the skb to modify * @gfp_mask: allocation priority * - * This must be called on SKBTX_DEV_ZEROCOPY skb. + * This must be called on skb with SKBFL_ZEROCOPY_ENABLE. * It will copy all frags into kernel and drop the reference * to userspace pages. * @@ -3264,8 +3264,7 @@ void skb_split(struct sk_buff *skb, struct sk_buff *skb1, const u32 len) { int pos = skb_headlen(skb); - skb_shinfo(skb1)->tx_flags |= skb_shinfo(skb)->tx_flags & - SKBTX_SHARED_FRAG; + skb_shinfo(skb1)->flags |= skb_shinfo(skb)->flags & SKBFL_SHARED_FRAG; skb_zerocopy_clone(skb1, skb, 0); if (len < pos) /* Split line is inside header. */ skb_split_inside_header(skb, skb1, len, pos); @@ -3954,8 +3953,8 @@ struct sk_buff *skb_segment(struct sk_buff *head_skb, skb_copy_from_linear_data_offset(head_skb, offset, skb_put(nskb, hsize), hsize); - skb_shinfo(nskb)->tx_flags |= skb_shinfo(head_skb)->tx_flags & - SKBTX_SHARED_FRAG; + skb_shinfo(nskb)->flags |= skb_shinfo(head_skb)->flags & + SKBFL_SHARED_FRAG; if (skb_orphan_frags(frag_skb, GFP_ATOMIC) || skb_zerocopy_clone(nskb, frag_skb, GFP_ATOMIC)) diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 2882d520f5b1..1954190b33c7 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1010,7 +1010,7 @@ struct sk_buff *tcp_build_frag(struct sock *sk, int size_goal, int flags, } if (!(flags & MSG_NO_SHARED_FRAGS)) - skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG; + skb_shinfo(skb)->flags |= SKBFL_SHARED_FRAG; skb->len += copy; skb->data_len += copy; diff --git a/net/kcm/kcmsock.c b/net/kcm/kcmsock.c index 56dad9565bc9..35a13681c337 100644 --- a/net/kcm/kcmsock.c +++ b/net/kcm/kcmsock.c @@ -786,7 +786,7 @@ static ssize_t kcm_sendpage(struct socket *sock, struct page *page, if (skb_can_coalesce(skb, i, page, offset)) { skb_frag_size_add(&skb_shinfo(skb)->frags[i - 1], size); - skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG; + skb_shinfo(skb)->flags |= SKBFL_SHARED_FRAG; goto coalesced; } @@ -834,7 +834,7 @@ static ssize_t kcm_sendpage(struct socket *sock, struct page *page, get_page(page); skb_fill_page_desc(skb, i, page, offset, size); - skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG; + skb_shinfo(skb)->flags |= SKBFL_SHARED_FRAG; coalesced: skb->len += size;

[net-next,v1,10/13] net: group skb_shinfo zerocopy related bits together.

Checks

Commit Message

Patch