Commit d831ee84 authored by Eelco Chaudron's avatar Eelco Chaudron Committed by Alexei Starovoitov
Browse files

bpf: Add bpf_xdp_output() helper 



Introduce new helper that reuses existing xdp perf_event output
implementation, but can be called from raw_tracepoint programs
that receive 'struct xdp_buff *' as a tracepoint argument.

Signed-off-by: default avatarEelco Chaudron <echaudro@redhat.com>
Signed-off-by: default avatarAlexei Starovoitov <ast@kernel.org>
Acked-by: default avatarJohn Fastabend <john.fastabend@gmail.com>
Acked-by: default avatarToke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/bpf/158348514556.2239.11050972434793741444.stgit@xdp-tutorial
parent 4823b721

include/uapi/linux/bpf.h

+25 −1
Original line number Diff line number Diff line
@@ -2927,6 +2927,29 @@ union bpf_attr { 
 *

 *		**-ENOENT** if pidns does not exists for the current task.

 *

 * int bpf_xdp_output(void *ctx, struct bpf_map *map, u64 flags, void *data, u64 size)

 *	Description

 *		Write raw *data* blob into a special BPF perf event held by

 *		*map* of type **BPF_MAP_TYPE_PERF_EVENT_ARRAY**. This perf

 *		event must have the following attributes: **PERF_SAMPLE_RAW**

 *		as **sample_type**, **PERF_TYPE_SOFTWARE** as **type**, and

 *		**PERF_COUNT_SW_BPF_OUTPUT** as **config**.

 *

 *		The *flags* are used to indicate the index in *map* for which

 *		the value must be put, masked with **BPF_F_INDEX_MASK**.

 *		Alternatively, *flags* can be set to **BPF_F_CURRENT_CPU**

 *		to indicate that the index of the current CPU core should be

 *		used.

 *

 *		The value to write, of *size*, is passed through eBPF stack and

 *		pointed by *data*.

 *

 *		*ctx* is a pointer to in-kernel struct xdp_buff.

 *

 *		This helper is similar to **bpf_perf_eventoutput**\ () but

 *		restricted to raw_tracepoint bpf programs.

 *	Return

 *		0 on success, or a negative error in case of failure.

 */

#define __BPF_FUNC_MAPPER(FN)		\

	FN(unspec),			\
@@ -3049,7 +3072,8 @@ union bpf_attr { 
	FN(send_signal_thread),		\

	FN(jiffies64),			\

	FN(read_branch_records),	\

	FN(get_ns_current_pid_tgid),

	FN(get_ns_current_pid_tgid),	\

	FN(xdp_output),



/* integer value in 'imm' field of BPF_CALL instruction selects which helper

 * function eBPF program intends to call

kernel/bpf/verifier.c

+3 −1
Original line number Diff line number Diff line
@@ -3650,7 +3650,8 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env, 
		if (func_id != BPF_FUNC_perf_event_read &&

		    func_id != BPF_FUNC_perf_event_output &&

		    func_id != BPF_FUNC_skb_output &&

		    func_id != BPF_FUNC_perf_event_read_value)

		    func_id != BPF_FUNC_perf_event_read_value &&

		    func_id != BPF_FUNC_xdp_output)

			goto error;

		break;

	case BPF_MAP_TYPE_STACK_TRACE:
@@ -3740,6 +3741,7 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env, 
	case BPF_FUNC_perf_event_output:

	case BPF_FUNC_perf_event_read_value:

	case BPF_FUNC_skb_output:

	case BPF_FUNC_xdp_output:

		if (map->map_type != BPF_MAP_TYPE_PERF_EVENT_ARRAY)

			goto error;

		break;

kernel/trace/bpf_trace.c

+3 −0
Original line number Diff line number Diff line
@@ -1145,6 +1145,7 @@ static const struct bpf_func_proto bpf_perf_event_output_proto_raw_tp = { 
};



extern const struct bpf_func_proto bpf_skb_output_proto;

extern const struct bpf_func_proto bpf_xdp_output_proto;



BPF_CALL_3(bpf_get_stackid_raw_tp, struct bpf_raw_tracepoint_args *, args,

	   struct bpf_map *, map, u64, flags)
@@ -1220,6 +1221,8 @@ tracing_prog_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) 
#ifdef CONFIG_NET

	case BPF_FUNC_skb_output:

		return &bpf_skb_output_proto;

	case BPF_FUNC_xdp_output:

		return &bpf_xdp_output_proto;

#endif

	default:

		return raw_tp_prog_func_proto(func_id, prog);

net/core/filter.c

+15 −1
Original line number Diff line number Diff line
@@ -4061,7 +4061,8 @@ BPF_CALL_5(bpf_xdp_event_output, struct xdp_buff *, xdp, struct bpf_map *, map, 


	if (unlikely(flags & ~(BPF_F_CTXLEN_MASK | BPF_F_INDEX_MASK)))

		return -EINVAL;

	if (unlikely(xdp_size > (unsigned long)(xdp->data_end - xdp->data)))

	if (unlikely(!xdp ||

		     xdp_size > (unsigned long)(xdp->data_end - xdp->data)))

		return -EFAULT;



	return bpf_event_output(map, flags, meta, meta_size, xdp->data,
@@ -4079,6 +4080,19 @@ static const struct bpf_func_proto bpf_xdp_event_output_proto = { 
	.arg5_type	= ARG_CONST_SIZE_OR_ZERO,

};



static int bpf_xdp_output_btf_ids[5];

const struct bpf_func_proto bpf_xdp_output_proto = {

	.func		= bpf_xdp_event_output,

	.gpl_only	= true,

	.ret_type	= RET_INTEGER,

	.arg1_type	= ARG_PTR_TO_BTF_ID,

	.arg2_type	= ARG_CONST_MAP_PTR,

	.arg3_type	= ARG_ANYTHING,

	.arg4_type	= ARG_PTR_TO_MEM,

	.arg5_type	= ARG_CONST_SIZE_OR_ZERO,

	.btf_id		= bpf_xdp_output_btf_ids,

};



BPF_CALL_1(bpf_get_socket_cookie, struct sk_buff *, skb)

{

	return skb->sk ? sock_gen_cookie(skb->sk) : 0;

tools/include/uapi/linux/bpf.h

+25 −1
Original line number Diff line number Diff line
@@ -2927,6 +2927,29 @@ union bpf_attr { 
 *

 *		**-ENOENT** if pidns does not exists for the current task.

 *

 * int bpf_xdp_output(void *ctx, struct bpf_map *map, u64 flags, void *data, u64 size)

 *	Description

 *		Write raw *data* blob into a special BPF perf event held by

 *		*map* of type **BPF_MAP_TYPE_PERF_EVENT_ARRAY**. This perf

 *		event must have the following attributes: **PERF_SAMPLE_RAW**

 *		as **sample_type**, **PERF_TYPE_SOFTWARE** as **type**, and

 *		**PERF_COUNT_SW_BPF_OUTPUT** as **config**.

 *

 *		The *flags* are used to indicate the index in *map* for which

 *		the value must be put, masked with **BPF_F_INDEX_MASK**.

 *		Alternatively, *flags* can be set to **BPF_F_CURRENT_CPU**

 *		to indicate that the index of the current CPU core should be

 *		used.

 *

 *		The value to write, of *size*, is passed through eBPF stack and

 *		pointed by *data*.

 *

 *		*ctx* is a pointer to in-kernel struct xdp_buff.

 *

 *		This helper is similar to **bpf_perf_eventoutput**\ () but

 *		restricted to raw_tracepoint bpf programs.

 *	Return

 *		0 on success, or a negative error in case of failure.

 */

#define __BPF_FUNC_MAPPER(FN)		\

	FN(unspec),			\
@@ -3049,7 +3072,8 @@ union bpf_attr { 
	FN(send_signal_thread),		\

	FN(jiffies64),			\

	FN(read_branch_records),	\

	FN(get_ns_current_pid_tgid),

	FN(get_ns_current_pid_tgid),	\

	FN(xdp_output),



/* integer value in 'imm' field of BPF_CALL instruction selects which helper

 * function eBPF program intends to call

CRA Git | Maintained and supported by SUSTech CRA and CCSE