Re: [PATCH v3 bpf-next 5/8] libbpf: add API to get XDP/XSK supported features
From: Andrii Nakryiko <hidden>
Date: 2023-01-27 23:39:10
Also in:
bpf, intel-wired-lan
On Thu, Jan 26, 2023 at 4:59 AM Lorenzo Bianconi [off-list ref] wrote:
quoted hunk ↗ jump to hunk
Extend bpf_xdp_query routine in order to get XDP/XSK supported features of netdev over route netlink interface. Extend libbpf netlink implementation in order to support netlink_generic protocol. Co-developed-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Co-developed-by: Marek Majtyka <redacted> Signed-off-by: Marek Majtyka <redacted> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> --- tools/lib/bpf/libbpf.h | 3 +- tools/lib/bpf/netlink.c | 99 +++++++++++++++++++++++++++++++++++++++++ tools/lib/bpf/nlattr.h | 12 +++++ 3 files changed, 113 insertions(+), 1 deletion(-)diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index 898db26e42e9..29cb7040fa77 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h@@ -982,9 +982,10 @@ struct bpf_xdp_query_opts { __u32 hw_prog_id; /* output */ __u32 skb_prog_id; /* output */ __u8 attach_mode; /* output */ + __u64 fflags; /* output */ size_t :0; }; -#define bpf_xdp_query_opts__last_field attach_mode +#define bpf_xdp_query_opts__last_field fflags
is "fflags" an obvious name in this context? I'd expect "feature_flags", especially that there are already "flags". Is saving a few characters worth the confusion?
quoted hunk ↗ jump to hunk
LIBBPF_API int bpf_xdp_attach(int ifindex, int prog_fd, __u32 flags, const struct bpf_xdp_attach_opts *opts);diff --git a/tools/lib/bpf/netlink.c b/tools/lib/bpf/netlink.c index d2468a04a6c3..674e4d61e67e 100644 --- a/tools/lib/bpf/netlink.c +++ b/tools/lib/bpf/netlink.c@@ -9,6 +9,7 @@ #include <linux/if_ether.h> #include <linux/pkt_cls.h> #include <linux/rtnetlink.h> +#include <linux/netdev.h> #include <sys/socket.h> #include <errno.h> #include <time.h>@@ -39,6 +40,12 @@ struct xdp_id_md { int ifindex; __u32 flags; struct xdp_link_info info; + __u64 fflags; +}; + +struct xdp_features_md { + int ifindex; + __u64 flags; }; static int libbpf_netlink_open(__u32 *nl_pid, int proto)
[...]
quoted hunk ↗ jump to hunk
int bpf_xdp_query(int ifindex, int xdp_flags, struct bpf_xdp_query_opts *opts) { struct libbpf_nla_req req = {@@ -393,6 +460,38 @@ int bpf_xdp_query(int ifindex, int xdp_flags, struct bpf_xdp_query_opts *opts) OPTS_SET(opts, skb_prog_id, xdp_id.info.skb_prog_id); OPTS_SET(opts, attach_mode, xdp_id.info.attach_mode); + if (OPTS_HAS(opts, fflags)) {
maybe invert condition, return early, reduce nesting of the following code?
+ struct xdp_features_md md = {
+ .ifindex = ifindex,
+ };
+ __u16 id;
+
+ err = libbpf_netlink_resolve_genl_family_id("netdev",
+ sizeof("netdev"),
+ &id);nit: if it fits under 100 characters, let's leave it on a single line
+ if (err < 0)
+ return libbpf_err(err);
+
+ memset(&req, 0, sizeof(req));
+ req.nh.nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN);
+ req.nh.nlmsg_flags = NLM_F_REQUEST;
+ req.nh.nlmsg_type = id;
+ req.gnl.cmd = NETDEV_CMD_DEV_GET;
+ req.gnl.version = 2;
+
+ err = nlattr_add(&req, NETDEV_A_DEV_IFINDEX, &ifindex,
+ sizeof(ifindex));
+ if (err < 0)
+ return err;
+
+ err = libbpf_netlink_send_recv(&req, NETLINK_GENERIC,
+ parse_xdp_features, NULL, &md);
+ if (err)
+ return libbpf_err(err);
+
+ opts->fflags = md.flags;
+ }
+
return 0;
}[...]