Merge branch 'writeable-bpf-tracepoints'

Matt Mullins says:

====================
This adds an opt-in interface for tracepoints to expose a writable context to
BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE programs that are attached, while
supporting read-only access from existing BPF_PROG_TYPE_RAW_TRACEPOINT
programs, as well as from non-BPF-based tracepoints.

The initial motivation is to support tracing that can be observed from the
remote end of an NBD socket, e.g. by adding flags to the struct nbd_request
header.  Earlier attempts included adding an NBD-specific tracepoint fd, but in
code review, I was recommended to implement it more generically -- as a result,
this patchset is far simpler than my initial try.

v4->v5:
  * rebased onto bpf-next/master and fixed merge conflicts
  * "tools: sync bpf.h" also syncs comments that have previously changed
    in bpf-next

v3->v4:
  * fixed a silly copy/paste typo in include/trace/events/bpf_test_run.h
    (_TRACE_NBD_H -> _TRACE_BPF_TEST_RUN_H)
  * fixed incorrect/misleading wording in patch 1's commit message,
    since the pointer cannot be directly dereferenced in a
    BPF_PROG_TYPE_RAW_TRACEPOINT
  * cleaned up the error message wording if the prog_tests fail
  * Addressed feedback from Yonghong
    * reject non-pointer-sized accesses to the buffer pointer
    * use sizeof(struct nbd_request) as one-byte-past-the-end in
      raw_tp_writable_reject_nbd_invalid.c
    * use BPF_MOV64_IMM instead of BPF_LD_IMM64

v2->v3:
  * Andrew addressed Josef's comments:
    * C-style commenting in nbd.c
    * Collapsed identical events into a single DECLARE_EVENT_CLASS.
      This saves about 2kB of kernel text

v1->v2:
  * add selftests
    * sync tools/include/uapi/linux/bpf.h
  * reject variable offset into the buffer
  * add string representation of PTR_TO_TP_BUFFER to reg_type_str
====================

Signed-off-by: Alexei Starovoitov <[email protected]>

Author: Alexei Starovoitov

Diff for: MAINTAINERS

@@ -10741,6 +10741,7 @@ L:	[email protected]
 L:	[email protected]
 F:	Documentation/blockdev/nbd.txt
 F:	drivers/block/nbd.c
+F:	include/trace/events/nbd.h
 F:	include/uapi/linux/nbd.h
 
 NETWORK DROP MONITOR

Diff for: drivers/block/nbd.c

@@ -44,6 +44,9 @@
 #include <linux/nbd-netlink.h>
 #include <net/genetlink.h>
 
+#define CREATE_TRACE_POINTS
+#include <trace/events/nbd.h>
+
 static DEFINE_IDR(nbd_index_idr);
 static DEFINE_MUTEX(nbd_index_mutex);
 static int nbd_total_devices = 0;
@@ -510,6 +513,10 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 	if (sent) {
 		if (sent >= sizeof(request)) {
 			skip = sent - sizeof(request);
+
+			/* initialize handle for tracing purposes */
+			handle = nbd_cmd_handle(cmd);
+
 			goto send_pages;
 		}
 		iov_iter_advance(&from, sent);
@@ -526,11 +533,14 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 	handle = nbd_cmd_handle(cmd);
 	memcpy(request.handle, &handle, sizeof(handle));
 
+	trace_nbd_send_request(&request, nbd->index, blk_mq_rq_from_pdu(cmd));
+
 	dev_dbg(nbd_to_dev(nbd), "request %p: sending control (%s@%llu,%uB)\n",
 		req, nbdcmd_to_ascii(type),
 		(unsigned long long)blk_rq_pos(req) << 9, blk_rq_bytes(req));
 	result = sock_xmit(nbd, index, 1, &from,
 			(type == NBD_CMD_WRITE) ? MSG_MORE : 0, &sent);
+	trace_nbd_header_sent(req, handle);
 	if (result <= 0) {
 		if (was_interrupted(result)) {
 			/* If we havne't sent anything we can just return BUSY,
@@ -603,6 +613,7 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 		bio = next;
 	}
 out:
+	trace_nbd_payload_sent(req, handle);
 	nsock->pending = NULL;
 	nsock->sent = 0;
 	return 0;
@@ -650,6 +661,7 @@ static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index)
 			tag, req);
 		return ERR_PTR(-ENOENT);
 	}
+	trace_nbd_header_received(req, handle);
 	cmd = blk_mq_rq_to_pdu(req);
 
 	mutex_lock(&cmd->lock);
@@ -703,6 +715,7 @@ static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index)
 		}
 	}
 out:
+	trace_nbd_payload_received(req, handle);
 	mutex_unlock(&cmd->lock);
 	return ret ? ERR_PTR(ret) : cmd;
 }

Diff for: include/linux/bpf.h

@@ -272,6 +272,7 @@ enum bpf_reg_type {
 	PTR_TO_SOCK_COMMON_OR_NULL, /* reg points to sock_common or NULL */
 	PTR_TO_TCP_SOCK,	 /* reg points to struct tcp_sock */
 	PTR_TO_TCP_SOCK_OR_NULL, /* reg points to struct tcp_sock or NULL */
+	PTR_TO_TP_BUFFER,	 /* reg points to a writable raw tp's buffer */
 };
 
 /* The information passed from prog-specific *_is_valid_access
@@ -361,6 +362,7 @@ struct bpf_prog_aux {
 	u32 used_map_cnt;
 	u32 max_ctx_offset;
 	u32 max_pkt_offset;
+	u32 max_tp_access;
 	u32 stack_depth;
 	u32 id;
 	u32 func_cnt; /* used by non-func prog as the number of func progs */

Diff for: include/linux/bpf_types.h

@@ -25,6 +25,7 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_KPROBE, kprobe)
 BPF_PROG_TYPE(BPF_PROG_TYPE_TRACEPOINT, tracepoint)
 BPF_PROG_TYPE(BPF_PROG_TYPE_PERF_EVENT, perf_event)
 BPF_PROG_TYPE(BPF_PROG_TYPE_RAW_TRACEPOINT, raw_tracepoint)
+BPF_PROG_TYPE(BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, raw_tracepoint_writable)
 #endif
 #ifdef CONFIG_CGROUP_BPF
 BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_DEVICE, cg_dev)

Diff for: include/linux/tracepoint-defs.h

@@ -45,6 +45,7 @@ struct bpf_raw_event_map {
 	struct tracepoint	*tp;
 	void			*bpf_func;
 	u32			num_args;
+	u32			writable_size;
 } __aligned(32);
 
 #endif

Diff for: include/trace/bpf_probe.h

@@ -69,8 +69,7 @@ __bpf_trace_##call(void *__data, proto)					\
  * to make sure that if the tracepoint handling changes, the
  * bpf probe will fail to compile unless it too is updated.
  */
-#undef DEFINE_EVENT
-#define DEFINE_EVENT(template, call, proto, args)			\
+#define __DEFINE_EVENT(template, call, proto, args, size)		\
 static inline void bpf_test_probe_##call(void)				\
 {									\
 	check_trace_callback_type_##call(__bpf_trace_##template);	\
@@ -81,12 +80,36 @@ __bpf_trace_tp_map_##call = {						\
 	.tp		= &__tracepoint_##call,				\
 	.bpf_func	= (void *)__bpf_trace_##template,		\
 	.num_args	= COUNT_ARGS(args),				\
+	.writable_size	= size,						\
 };
 
+#define FIRST(x, ...) x
+
+#undef DEFINE_EVENT_WRITABLE
+#define DEFINE_EVENT_WRITABLE(template, call, proto, args, size)	\
+static inline void bpf_test_buffer_##call(void)				\
+{									\
+	/* BUILD_BUG_ON() is ignored if the code is completely eliminated, but \
+	 * BUILD_BUG_ON_ZERO() uses a different mechanism that is not	\
+	 * dead-code-eliminated.					\
+	 */								\
+	FIRST(proto);							\
+	(void)BUILD_BUG_ON_ZERO(size != sizeof(*FIRST(args)));		\
+}									\
+__DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args), size)
+
+#undef DEFINE_EVENT
+#define DEFINE_EVENT(template, call, proto, args)			\
+	__DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args), 0)
 
 #undef DEFINE_EVENT_PRINT
 #define DEFINE_EVENT_PRINT(template, name, proto, args, print)	\
 	DEFINE_EVENT(template, name, PARAMS(proto), PARAMS(args))
 
 #include TRACE_INCLUDE(TRACE_INCLUDE_FILE)
+
+#undef DEFINE_EVENT_WRITABLE
+#undef __DEFINE_EVENT
+#undef FIRST
+
 #endif /* CONFIG_BPF_EVENTS */

Diff for: include/trace/events/bpf_test_run.h

@@ -0,0 +1,50 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM bpf_test_run
+
+#if !defined(_TRACE_BPF_TEST_RUN_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_BPF_TEST_RUN_H
+
+#include <linux/tracepoint.h>
+
+DECLARE_EVENT_CLASS(bpf_test_finish,
+
+	TP_PROTO(int *err),
+
+	TP_ARGS(err),
+
+	TP_STRUCT__entry(
+		__field(int, err)
+	),
+
+	TP_fast_assign(
+		__entry->err = *err;
+	),
+
+	TP_printk("bpf_test_finish with err=%d", __entry->err)
+);
+
+#ifdef DEFINE_EVENT_WRITABLE
+#undef BPF_TEST_RUN_DEFINE_EVENT
+#define BPF_TEST_RUN_DEFINE_EVENT(template, call, proto, args, size)	\
+	DEFINE_EVENT_WRITABLE(template, call, PARAMS(proto),		\
+			      PARAMS(args), size)
+#else
+#undef BPF_TEST_RUN_DEFINE_EVENT
+#define BPF_TEST_RUN_DEFINE_EVENT(template, call, proto, args, size)	\
+	DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args))
+#endif
+
+BPF_TEST_RUN_DEFINE_EVENT(bpf_test_finish, bpf_test_finish,
+
+	TP_PROTO(int *err),
+
+	TP_ARGS(err),
+
+	sizeof(int)
+);
+
+#endif
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>

Diff for: include/trace/events/nbd.h

@@ -0,0 +1,107 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM nbd
+
+#if !defined(_TRACE_NBD_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_NBD_H
+
+#include <linux/tracepoint.h>
+
+DECLARE_EVENT_CLASS(nbd_transport_event,
+
+	TP_PROTO(struct request *req, u64 handle),
+
+	TP_ARGS(req, handle),
+
+	TP_STRUCT__entry(
+		__field(struct request *, req)
+		__field(u64, handle)
+	),
+
+	TP_fast_assign(
+		__entry->req = req;
+		__entry->handle = handle;
+	),
+
+	TP_printk(
+		"nbd transport event: request %p, handle 0x%016llx",
+		__entry->req,
+		__entry->handle
+	)
+);
+
+DEFINE_EVENT(nbd_transport_event, nbd_header_sent,
+
+	TP_PROTO(struct request *req, u64 handle),
+
+	TP_ARGS(req, handle)
+);
+
+DEFINE_EVENT(nbd_transport_event, nbd_payload_sent,
+
+	TP_PROTO(struct request *req, u64 handle),
+
+	TP_ARGS(req, handle)
+);
+
+DEFINE_EVENT(nbd_transport_event, nbd_header_received,
+
+	TP_PROTO(struct request *req, u64 handle),
+
+	TP_ARGS(req, handle)
+);
+
+DEFINE_EVENT(nbd_transport_event, nbd_payload_received,
+
+	TP_PROTO(struct request *req, u64 handle),
+
+	TP_ARGS(req, handle)
+);
+
+DECLARE_EVENT_CLASS(nbd_send_request,
+
+	TP_PROTO(struct nbd_request *nbd_request, int index,
+		 struct request *rq),
+
+	TP_ARGS(nbd_request, index, rq),
+
+	TP_STRUCT__entry(
+		__field(struct nbd_request *, nbd_request)
+		__field(u64, dev_index)
+		__field(struct request *, request)
+	),
+
+	TP_fast_assign(
+		__entry->nbd_request = 0;
+		__entry->dev_index = index;
+		__entry->request = rq;
+	),
+
+	TP_printk("nbd%lld: request %p", __entry->dev_index, __entry->request)
+);
+
+#ifdef DEFINE_EVENT_WRITABLE
+#undef NBD_DEFINE_EVENT
+#define NBD_DEFINE_EVENT(template, call, proto, args, size)		\
+	DEFINE_EVENT_WRITABLE(template, call, PARAMS(proto),		\
+			      PARAMS(args), size)
+#else
+#undef NBD_DEFINE_EVENT
+#define NBD_DEFINE_EVENT(template, call, proto, args, size)		\
+	DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args))
+#endif
+
+NBD_DEFINE_EVENT(nbd_send_request, nbd_send_request,
+
+	TP_PROTO(struct nbd_request *nbd_request, int index,
+		 struct request *rq),
+
+	TP_ARGS(nbd_request, index, rq),
+
+	sizeof(struct nbd_request)
+);
+
+#endif
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>

Diff for: include/uapi/linux/bpf.h

@@ -168,6 +168,7 @@ enum bpf_prog_type {
 	BPF_PROG_TYPE_SK_REUSEPORT,
 	BPF_PROG_TYPE_FLOW_DISSECTOR,
 	BPF_PROG_TYPE_CGROUP_SYSCTL,
+	BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE,
 };
 
 enum bpf_attach_type {

Diff for: kernel/bpf/syscall.c

@@ -1789,12 +1789,16 @@ static int bpf_raw_tracepoint_open(const union bpf_attr *attr)
 	}
 	raw_tp->btp = btp;
 
-	prog = bpf_prog_get_type(attr->raw_tracepoint.prog_fd,
-				 BPF_PROG_TYPE_RAW_TRACEPOINT);
+	prog = bpf_prog_get(attr->raw_tracepoint.prog_fd);
 	if (IS_ERR(prog)) {
 		err = PTR_ERR(prog);
 		goto out_free_tp;
 	}
+	if (prog->type != BPF_PROG_TYPE_RAW_TRACEPOINT &&
+	    prog->type != BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE) {
+		err = -EINVAL;
+		goto out_put_prog;
+	}
 
 	err = bpf_probe_register(raw_tp->btp, prog);
 	if (err)

Diff for: kernel/bpf/verifier.c

@@ -405,6 +405,7 @@ static const char * const reg_type_str[] = {
 	[PTR_TO_SOCK_COMMON_OR_NULL] = "sock_common_or_null",
 	[PTR_TO_TCP_SOCK]	= "tcp_sock",
 	[PTR_TO_TCP_SOCK_OR_NULL] = "tcp_sock_or_null",
+	[PTR_TO_TP_BUFFER]	= "tp_buffer",
 };
 
 static char slot_type_char[] = {
@@ -1993,6 +1994,32 @@ static int check_ctx_reg(struct bpf_verifier_env *env,
 	return 0;
 }
 
+static int check_tp_buffer_access(struct bpf_verifier_env *env,
+				  const struct bpf_reg_state *reg,
+				  int regno, int off, int size)
+{
+	if (off < 0) {
+		verbose(env,
+			"R%d invalid tracepoint buffer access: off=%d, size=%d",
+			regno, off, size);
+		return -EACCES;
+	}
+	if (!tnum_is_const(reg->var_off) || reg->var_off.value) {
+		char tn_buf[48];
+
+		tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
+		verbose(env,
+			"R%d invalid variable buffer offset: off=%d, var_off=%s",
+			regno, off, tn_buf);
+		return -EACCES;
+	}
+	if (off + size > env->prog->aux->max_tp_access)
+		env->prog->aux->max_tp_access = off + size;
+
+	return 0;
+}
+
+
 /* truncate register to smaller size (in bytes)
  * must be called with size < BPF_REG_SIZE
  */
@@ -2137,6 +2164,10 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
 		err = check_sock_access(env, insn_idx, regno, off, size, t);
 		if (!err && value_regno >= 0)
 			mark_reg_unknown(env, regs, value_regno);
+	} else if (reg->type == PTR_TO_TP_BUFFER) {
+		err = check_tp_buffer_access(env, reg, regno, off, size);
+		if (!err && t == BPF_READ && value_regno >= 0)
+			mark_reg_unknown(env, regs, value_regno);
 	} else {
 		verbose(env, "R%d invalid mem access '%s'\n", regno,
 			reg_type_str[reg->type]);