net: Optimize xdp_do_flush() with bpf_net_context infos.

Every NIC driver utilizing XDP should invoke xdp_do_flush() after
processing all packages. With the introduction of the bpf_net_context
logic the flush lists (for dev, CPU-map and xsk) are lazy initialized
only if used. However xdp_do_flush() tries to flush all three of them so
all three lists are always initialized and the likely empty lists are
"iterated".
Without the usage of XDP but with CONFIG_DEBUG_NET the lists are also
initialized due to xdp_do_check_flushed().

Jakub suggest to utilize the hints in bpf_net_context and avoid invoking
the flush function. This will also avoiding initializing the lists which
are otherwise unused.

Introduce bpf_net_ctx_get_all_used_flush_lists() to return the
individual list if not-empty. Use the logic in xdp_do_flush() and
xdp_do_check_flushed(). Remove the not needed .*_check_flush().

Suggested-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

2 files changed, 31 insertions, 6 deletions

diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index a834f4b761bc..f5c6bc9093a6 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -2494,7 +2494,7 @@ struct sk_buff;
 struct bpf_dtab_netdev;
 struct bpf_cpu_map_entry;
 
-void __dev_flush(void);
+void __dev_flush(struct list_head *flush_list);
 int dev_xdp_enqueue(struct net_device *dev, struct xdp_frame *xdpf,
 		    struct net_device *dev_rx);
 int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_frame *xdpf,
@@ -2507,7 +2507,7 @@ int dev_map_redirect_multi(struct net_device *dev, struct sk_buff *skb,
 			   struct bpf_prog *xdp_prog, struct bpf_map *map,
 			   bool exclude_ingress);
 
-void __cpu_map_flush(void);
+void __cpu_map_flush(struct list_head *flush_list);
 int cpu_map_enqueue(struct bpf_cpu_map_entry *rcpu, struct xdp_frame *xdpf,
 		    struct net_device *dev_rx);
 int cpu_map_generic_redirect(struct bpf_cpu_map_entry *rcpu,
@@ -2644,8 +2644,6 @@ void bpf_dynptr_init(struct bpf_dynptr_kern *ptr, void *data,
 void bpf_dynptr_set_null(struct bpf_dynptr_kern *ptr);
 void bpf_dynptr_set_rdonly(struct bpf_dynptr_kern *ptr);
 
-bool dev_check_flush(void);
-bool cpu_map_check_flush(void);
 #else /* !CONFIG_BPF_SYSCALL */
 static inline struct bpf_prog *bpf_prog_get(u32 ufd)
 {
@@ -2738,7 +2736,7 @@ static inline struct bpf_token *bpf_token_get_from_fd(u32 ufd)
 	return ERR_PTR(-EOPNOTSUPP);
 }
 
-static inline void __dev_flush(void)
+static inline void __dev_flush(struct list_head *flush_list)
 {
 }
 
@@ -2784,7 +2782,7 @@ int dev_map_redirect_multi(struct net_device *dev, struct sk_buff *skb,
 	return 0;
 }
 
-static inline void __cpu_map_flush(void)
+static inline void __cpu_map_flush(struct list_head *flush_list)
 {
 }
 
diff --git a/include/linux/filter.h b/include/linux/filter.h
index c0349522de8f..02ddcfdf94c4 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -829,6 +829,33 @@ static inline struct list_head *bpf_net_ctx_get_xskmap_flush_list(void)
 	return &bpf_net_ctx->xskmap_map_flush_list;
 }
 
+static inline void bpf_net_ctx_get_all_used_flush_lists(struct list_head **lh_map,
+							struct list_head **lh_dev,
+							struct list_head **lh_xsk)
+{
+	struct bpf_net_context *bpf_net_ctx = bpf_net_ctx_get();
+	u32 kern_flags = bpf_net_ctx->ri.kern_flags;
+	struct list_head *lh;
+
+	*lh_map = *lh_dev = *lh_xsk = NULL;
+
+	if (!IS_ENABLED(CONFIG_BPF_SYSCALL))
+		return;
+
+	lh = &bpf_net_ctx->dev_map_flush_list;
+	if (kern_flags & BPF_RI_F_DEV_MAP_INIT && !list_empty(lh))
+		*lh_dev = lh;
+
+	lh = &bpf_net_ctx->cpu_map_flush_list;
+	if (kern_flags & BPF_RI_F_CPU_MAP_INIT && !list_empty(lh))
+		*lh_map = lh;
+
+	lh = &bpf_net_ctx->xskmap_map_flush_list;
+	if (IS_ENABLED(CONFIG_XDP_SOCKETS) &&
+	    kern_flags & BPF_RI_F_XSK_MAP_INIT && !list_empty(lh))
+		*lh_xsk = lh;
+}
+
 /* Compute the linear packet data range [data, data_end) which
  * will be accessed by various program types (cls_bpf, act_bpf,
  * lwt, ...). Subsystems allowing direct data access must (!)