[PATCH bpf-next 0/6] Consistent prefixes for libbpf interfaces

[PATCH bpf-next 0/6] Consistent prefixes for libbpf interfaces

[Andrey Ignatov]

This patch set renames a few interfaces in libbpf, mostly netlink related,
so that all symbols provided by the library have only three possible
prefixes:

% nm -D tools/lib/bpf/libbpf.so  | \
    awk '$2 == "T" {sub(/[_\(].*/, "", $3); if ($3) print $3}' | \
    sort | \
    uniq -c
     91 bpf
      8 btf
     14 libbpf

libbpf is used more and more outside kernel tree. That means the library
should follow good practices in library design and implementation to
play well with third party code that uses it.

One of such practices is to have a common prefix (or a few) for every
interface, function or data structure, library provides. It helps to
avoid name conflicts with other libraries and keeps API/ABI consistent.

Inconsistent names in libbpf already cause problems in real life. E.g.
an application can't use both libbpf and libnl due to conflicting
symbols (specifically nla_parse, nla_parse_nested and a few others).

Some of problematic global symbols are not part of ABI and can be
restricted from export with either visibility attribute/pragma or export
map (what is useful by itself and can be done in addition). That won't
solve the problem for those that are part of ABI though. Also export
restrictions would help only in DSO case. If third party application links
libbpf statically it won't help, and people do it (e.g. Facebook links
most of libraries statically, including libbpf).

libbpf already uses the following prefixes for its interfaces:
* bpf_ for bpf system call wrappers, program/map/elf-object
  abstractions and a few other things;
* btf_ for BTF related API;
* libbpf_ for everything else.

The patch adds libbpf_ prefix to interfaces that use none of mentioned
above prefixes and don't fit well into the first two categories.

Long term benefits of having common prefix should outweigh possible
inconvenience of changing API for those functions now.

Patches 2-4 add libbpf_ prefix to libbpf interfaces: separate patch per
header. Other patches are simple improvements in API.


Andrey Ignatov (6):
  libbpf: Move __dump_nlmsg_t from API to implementation
  libbpf: Consistent prefixes for interfaces in libbpf.h.
  libbpf: Consistent prefixes for interfaces in nlattr.h.
  libbpf: Consistent prefixes for interfaces in str_error.h.
  libbpf: Make include guards consistent
  libbpf: Use __u32 instead of u32 in bpf_program__load

 tools/bpf/bpftool/net.c            | 41 ++++++++++---------
 tools/bpf/bpftool/netlink_dumper.c | 32 ++++++++-------
 tools/lib/bpf/bpf.h                |  6 +--
 tools/lib/bpf/btf.h                |  6 +--
 tools/lib/bpf/libbpf.c             | 22 +++++-----
 tools/lib/bpf/libbpf.h             | 31 +++++++-------
 tools/lib/bpf/netlink.c            | 48 ++++++++++++----------
 tools/lib/bpf/nlattr.c             | 64 +++++++++++++++--------------
 tools/lib/bpf/nlattr.h             | 65 +++++++++++++++---------------
 tools/lib/bpf/str_error.c          |  2 +-
 tools/lib/bpf/str_error.h          |  8 ++--
 11 files changed, 171 insertions(+), 154 deletions(-)

-- 
2.17.1

[PATCH bpf-next 1/6] libbpf: Move __dump_nlmsg_t from API to implementation

[Andrey Ignatov]

This typedef is used only by implementation in netlink.c. Nothing uses
it in public API. Move it to netlink.c.

Signed-off-by: Andrey Ignatov <rdna@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
---
 tools/lib/bpf/libbpf.h  | 3 ---
 tools/lib/bpf/netlink.c | 3 +++
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 2ed24d3f80b3..8388be525388 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -304,11 +304,8 @@ int bpf_perf_event_read_simple(void *mem, unsigned long size,
 			       void **buf, size_t *buf_len,
 			       bpf_perf_event_print_t fn, void *priv);
 
-struct nlmsghdr;
 struct nlattr;
 typedef int (*dump_nlmsg_t)(void *cookie, void *msg, struct nlattr **tb);
-typedef int (*__dump_nlmsg_t)(struct nlmsghdr *nlmsg, dump_nlmsg_t,
-			      void *cookie);
 int bpf_netlink_open(unsigned int *nl_pid);
 int nl_get_link(int sock, unsigned int nl_pid, dump_nlmsg_t dump_link_nlmsg,
 		void *cookie);
diff --git a/tools/lib/bpf/netlink.c b/tools/lib/bpf/netlink.c
index fde1d7bf8199..da46d9358d9d 100644
--- a/tools/lib/bpf/netlink.c
+++ b/tools/lib/bpf/netlink.c
@@ -18,6 +18,9 @@
 #define SOL_NETLINK 270
 #endif
 
+typedef int (*__dump_nlmsg_t)(struct nlmsghdr *nlmsg, dump_nlmsg_t,
+			      void *cookie);
+
 int bpf_netlink_open(__u32 *nl_pid)
 {
 	struct sockaddr_nl sa;
-- 
2.17.1

[PATCH bpf-next 2/6] libbpf: Consistent prefixes for interfaces in libbpf.h.

[Andrey Ignatov]

libbpf is used more and more outside kernel tree. That means the library
should follow good practices in library design and implementation to
play well with third party code that uses it.

One of such practices is to have a common prefix (or a few) for every
interface, function or data structure, library provides. I helps to
avoid name conflicts with other libraries and keeps API consistent.

Inconsistent names in libbpf already cause problems in real life. E.g.
an application can't use both libbpf and libnl due to conflicting
symbols.

Having common prefix will help to fix current and avoid future problems.

libbpf already uses the following prefixes for its interfaces:
* bpf_ for bpf system call wrappers, program/map/elf-object
  abstractions and a few other things;
* btf_ for BTF related API;
* libbpf_ for everything else.

The patch adds libbpf_ prefix to functions and typedef in libbpf.h that
use none of mentioned above prefixes and doesn't fit well into the first
two categories.

Since affected part of API is used in bpftool, the patch applies
corresponding change to bpftool as well. Having it in a separate patch
will cause a state of tree where bpftool is broken what may not be a
good idea.

Signed-off-by: Andrey Ignatov <rdna@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
---
 tools/bpf/bpftool/net.c | 31 +++++++++++++++----------------
 tools/lib/bpf/libbpf.h  | 20 ++++++++++----------
 tools/lib/bpf/netlink.c | 37 ++++++++++++++++++++-----------------
 3 files changed, 45 insertions(+), 43 deletions(-)

diff --git a/tools/bpf/bpftool/net.c b/tools/bpf/bpftool/net.c
index ed205ee57655..ef83e8a08490 100644
--- a/tools/bpf/bpftool/net.c
+++ b/tools/bpf/bpftool/net.c
@@ -127,14 +127,14 @@ static int show_dev_tc_bpf(int sock, unsigned int nl_pid,
 	tcinfo.array_len = 0;
 
 	tcinfo.is_qdisc = false;
-	ret = nl_get_class(sock, nl_pid, dev->ifindex, dump_class_qdisc_nlmsg,
-			   &tcinfo);
+	ret = libbpf_nl_get_class(sock, nl_pid, dev->ifindex,
+				  dump_class_qdisc_nlmsg, &tcinfo);
 	if (ret)
 		goto out;
 
 	tcinfo.is_qdisc = true;
-	ret = nl_get_qdisc(sock, nl_pid, dev->ifindex, dump_class_qdisc_nlmsg,
-			   &tcinfo);
+	ret = libbpf_nl_get_qdisc(sock, nl_pid, dev->ifindex,
+				  dump_class_qdisc_nlmsg, &tcinfo);
 	if (ret)
 		goto out;
 
@@ -142,10 +142,9 @@ static int show_dev_tc_bpf(int sock, unsigned int nl_pid,
 	filter_info.ifindex = dev->ifindex;
 	for (i = 0; i < tcinfo.used_len; i++) {
 		filter_info.kind = tcinfo.handle_array[i].kind;
-		ret = nl_get_filter(sock, nl_pid, dev->ifindex,
-				    tcinfo.handle_array[i].handle,
-				    dump_filter_nlmsg,
-				    &filter_info);
+		ret = libbpf_nl_get_filter(sock, nl_pid, dev->ifindex,
+					   tcinfo.handle_array[i].handle,
+					   dump_filter_nlmsg, &filter_info);
 		if (ret)
 			goto out;
 	}
@@ -153,22 +152,22 @@ static int show_dev_tc_bpf(int sock, unsigned int nl_pid,
 	/* root, ingress and egress handle */
 	handle = TC_H_ROOT;
 	filter_info.kind = "root";
-	ret = nl_get_filter(sock, nl_pid, dev->ifindex, handle,
-			    dump_filter_nlmsg, &filter_info);
+	ret = libbpf_nl_get_filter(sock, nl_pid, dev->ifindex, handle,
+				   dump_filter_nlmsg, &filter_info);
 	if (ret)
 		goto out;
 
 	handle = TC_H_MAKE(TC_H_CLSACT, TC_H_MIN_INGRESS);
 	filter_info.kind = "clsact/ingress";
-	ret = nl_get_filter(sock, nl_pid, dev->ifindex, handle,
-			    dump_filter_nlmsg, &filter_info);
+	ret = libbpf_nl_get_filter(sock, nl_pid, dev->ifindex, handle,
+				   dump_filter_nlmsg, &filter_info);
 	if (ret)
 		goto out;
 
 	handle = TC_H_MAKE(TC_H_CLSACT, TC_H_MIN_EGRESS);
 	filter_info.kind = "clsact/egress";
-	ret = nl_get_filter(sock, nl_pid, dev->ifindex, handle,
-			    dump_filter_nlmsg, &filter_info);
+	ret = libbpf_nl_get_filter(sock, nl_pid, dev->ifindex, handle,
+				   dump_filter_nlmsg, &filter_info);
 	if (ret)
 		goto out;
 
@@ -196,7 +195,7 @@ static int do_show(int argc, char **argv)
 		usage();
 	}
 
-	sock = bpf_netlink_open(&nl_pid);
+	sock = libbpf_netlink_open(&nl_pid);
 	if (sock < 0) {
 		fprintf(stderr, "failed to open netlink sock\n");
 		return -1;
@@ -211,7 +210,7 @@ static int do_show(int argc, char **argv)
 		jsonw_start_array(json_wtr);
 	NET_START_OBJECT;
 	NET_START_ARRAY("xdp", "%s:\n");
-	ret = nl_get_link(sock, nl_pid, dump_link_nlmsg, &dev_array);
+	ret = libbpf_nl_get_link(sock, nl_pid, dump_link_nlmsg, &dev_array);
 	NET_END_ARRAY("\n");
 
 	if (!ret) {
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 8388be525388..710ff5724980 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -305,14 +305,14 @@ int bpf_perf_event_read_simple(void *mem, unsigned long size,
 			       bpf_perf_event_print_t fn, void *priv);
 
 struct nlattr;
-typedef int (*dump_nlmsg_t)(void *cookie, void *msg, struct nlattr **tb);
-int bpf_netlink_open(unsigned int *nl_pid);
-int nl_get_link(int sock, unsigned int nl_pid, dump_nlmsg_t dump_link_nlmsg,
-		void *cookie);
-int nl_get_class(int sock, unsigned int nl_pid, int ifindex,
-		 dump_nlmsg_t dump_class_nlmsg, void *cookie);
-int nl_get_qdisc(int sock, unsigned int nl_pid, int ifindex,
-		 dump_nlmsg_t dump_qdisc_nlmsg, void *cookie);
-int nl_get_filter(int sock, unsigned int nl_pid, int ifindex, int handle,
-		  dump_nlmsg_t dump_filter_nlmsg, void *cookie);
+typedef int (*libbpf_dump_nlmsg_t)(void *cookie, void *msg, struct nlattr **tb);
+int libbpf_netlink_open(unsigned int *nl_pid);
+int libbpf_nl_get_link(int sock, unsigned int nl_pid,
+		       libbpf_dump_nlmsg_t dump_link_nlmsg, void *cookie);
+int libbpf_nl_get_class(int sock, unsigned int nl_pid, int ifindex,
+			libbpf_dump_nlmsg_t dump_class_nlmsg, void *cookie);
+int libbpf_nl_get_qdisc(int sock, unsigned int nl_pid, int ifindex,
+			libbpf_dump_nlmsg_t dump_qdisc_nlmsg, void *cookie);
+int libbpf_nl_get_filter(int sock, unsigned int nl_pid, int ifindex, int handle,
+			 libbpf_dump_nlmsg_t dump_filter_nlmsg, void *cookie);
 #endif
diff --git a/tools/lib/bpf/netlink.c b/tools/lib/bpf/netlink.c
index da46d9358d9d..506bdfdbcab0 100644
--- a/tools/lib/bpf/netlink.c
+++ b/tools/lib/bpf/netlink.c
@@ -18,10 +18,10 @@
 #define SOL_NETLINK 270
 #endif
 
-typedef int (*__dump_nlmsg_t)(struct nlmsghdr *nlmsg, dump_nlmsg_t,
+typedef int (*__dump_nlmsg_t)(struct nlmsghdr *nlmsg, libbpf_dump_nlmsg_t,
 			      void *cookie);
 
-int bpf_netlink_open(__u32 *nl_pid)
+int libbpf_netlink_open(__u32 *nl_pid)
 {
 	struct sockaddr_nl sa;
 	socklen_t addrlen;
@@ -65,7 +65,7 @@ int bpf_netlink_open(__u32 *nl_pid)
 }
 
 static int bpf_netlink_recv(int sock, __u32 nl_pid, int seq,
-			    __dump_nlmsg_t _fn, dump_nlmsg_t fn,
+			    __dump_nlmsg_t _fn, libbpf_dump_nlmsg_t fn,
 			    void *cookie)
 {
 	bool multipart = true;
@@ -133,7 +133,7 @@ int bpf_set_link_xdp_fd(int ifindex, int fd, __u32 flags)
 	} req;
 	__u32 nl_pid;
 
-	sock = bpf_netlink_open(&nl_pid);
+	sock = libbpf_netlink_open(&nl_pid);
 	if (sock < 0)
 		return sock;
 
@@ -181,8 +181,8 @@ int bpf_set_link_xdp_fd(int ifindex, int fd, __u32 flags)
 	return ret;
 }
 
-static int __dump_link_nlmsg(struct nlmsghdr *nlh, dump_nlmsg_t dump_link_nlmsg,
-			     void *cookie)
+static int __dump_link_nlmsg(struct nlmsghdr *nlh,
+			     libbpf_dump_nlmsg_t dump_link_nlmsg, void *cookie)
 {
 	struct nlattr *tb[IFLA_MAX + 1], *attr;
 	struct ifinfomsg *ifi = NLMSG_DATA(nlh);
@@ -196,8 +196,8 @@ static int __dump_link_nlmsg(struct nlmsghdr *nlh, dump_nlmsg_t dump_link_nlmsg,
 	return dump_link_nlmsg(cookie, ifi, tb);
 }
 
-int nl_get_link(int sock, unsigned int nl_pid, dump_nlmsg_t dump_link_nlmsg,
-		void *cookie)
+int libbpf_nl_get_link(int sock, unsigned int nl_pid,
+		       libbpf_dump_nlmsg_t dump_link_nlmsg, void *cookie)
 {
 	struct {
 		struct nlmsghdr nlh;
@@ -219,7 +219,8 @@ int nl_get_link(int sock, unsigned int nl_pid, dump_nlmsg_t dump_link_nlmsg,
 }
 
 static int __dump_class_nlmsg(struct nlmsghdr *nlh,
-			      dump_nlmsg_t dump_class_nlmsg, void *cookie)
+			      libbpf_dump_nlmsg_t dump_class_nlmsg,
+			      void *cookie)
 {
 	struct nlattr *tb[TCA_MAX + 1], *attr;
 	struct tcmsg *t = NLMSG_DATA(nlh);
@@ -233,8 +234,8 @@ static int __dump_class_nlmsg(struct nlmsghdr *nlh,
 	return dump_class_nlmsg(cookie, t, tb);
 }
 
-int nl_get_class(int sock, unsigned int nl_pid, int ifindex,
-		 dump_nlmsg_t dump_class_nlmsg, void *cookie)
+int libbpf_nl_get_class(int sock, unsigned int nl_pid, int ifindex,
+			libbpf_dump_nlmsg_t dump_class_nlmsg, void *cookie)
 {
 	struct {
 		struct nlmsghdr nlh;
@@ -257,7 +258,8 @@ int nl_get_class(int sock, unsigned int nl_pid, int ifindex,
 }
 
 static int __dump_qdisc_nlmsg(struct nlmsghdr *nlh,
-			      dump_nlmsg_t dump_qdisc_nlmsg, void *cookie)
+			      libbpf_dump_nlmsg_t dump_qdisc_nlmsg,
+			      void *cookie)
 {
 	struct nlattr *tb[TCA_MAX + 1], *attr;
 	struct tcmsg *t = NLMSG_DATA(nlh);
@@ -271,8 +273,8 @@ static int __dump_qdisc_nlmsg(struct nlmsghdr *nlh,
 	return dump_qdisc_nlmsg(cookie, t, tb);
 }
 
-int nl_get_qdisc(int sock, unsigned int nl_pid, int ifindex,
-		 dump_nlmsg_t dump_qdisc_nlmsg, void *cookie)
+int libbpf_nl_get_qdisc(int sock, unsigned int nl_pid, int ifindex,
+			libbpf_dump_nlmsg_t dump_qdisc_nlmsg, void *cookie)
 {
 	struct {
 		struct nlmsghdr nlh;
@@ -295,7 +297,8 @@ int nl_get_qdisc(int sock, unsigned int nl_pid, int ifindex,
 }
 
 static int __dump_filter_nlmsg(struct nlmsghdr *nlh,
-			       dump_nlmsg_t dump_filter_nlmsg, void *cookie)
+			       libbpf_dump_nlmsg_t dump_filter_nlmsg,
+			       void *cookie)
 {
 	struct nlattr *tb[TCA_MAX + 1], *attr;
 	struct tcmsg *t = NLMSG_DATA(nlh);
@@ -309,8 +312,8 @@ static int __dump_filter_nlmsg(struct nlmsghdr *nlh,
 	return dump_filter_nlmsg(cookie, t, tb);
 }
 
-int nl_get_filter(int sock, unsigned int nl_pid, int ifindex, int handle,
-		  dump_nlmsg_t dump_filter_nlmsg, void *cookie)
+int libbpf_nl_get_filter(int sock, unsigned int nl_pid, int ifindex, int handle,
+			 libbpf_dump_nlmsg_t dump_filter_nlmsg, void *cookie)
 {
 	struct {
 		struct nlmsghdr nlh;
-- 
2.17.1

[PATCH bpf-next 3/6] libbpf: Consistent prefixes for interfaces in nlattr.h.

[Andrey Ignatov]

libbpf is used more and more outside kernel tree. That means the library
should follow good practices in library design and implementation to
play well with third party code that uses it.

One of such practices is to have a common prefix (or a few) for every
interface, function or data structure, library provides. I helps to
avoid name conflicts with other libraries and keeps API consistent.

Inconsistent names in libbpf already cause problems in real life. E.g.
an application can't use both libbpf and libnl due to conflicting
symbols.

Having common prefix will help to fix current and avoid future problems.

libbpf already uses the following prefixes for its interfaces:
* bpf_ for bpf system call wrappers, program/map/elf-object
  abstractions and a few other things;
* btf_ for BTF related API;
* libbpf_ for everything else.

The patch adds libbpf_ prefix to interfaces in nlattr.h that use none of
mentioned above prefixes and doesn't fit well into the first two
categories.

Since affected part of API is used in bpftool, the patch applies
corresponding change to bpftool as well. Having it in a separate patch
will cause a state of tree where bpftool is broken what may not be a
good idea.

Signed-off-by: Andrey Ignatov <rdna@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
---
 tools/bpf/bpftool/net.c            | 10 +++--
 tools/bpf/bpftool/netlink_dumper.c | 32 ++++++++-------
 tools/lib/bpf/netlink.c            | 10 ++---
 tools/lib/bpf/nlattr.c             | 64 ++++++++++++++++--------------
 tools/lib/bpf/nlattr.h             | 59 +++++++++++++--------------
 5 files changed, 94 insertions(+), 81 deletions(-)

diff --git a/tools/bpf/bpftool/net.c b/tools/bpf/bpftool/net.c
index ef83e8a08490..d441bb7035ca 100644
--- a/tools/bpf/bpftool/net.c
+++ b/tools/bpf/bpftool/net.c
@@ -69,7 +69,9 @@ static int dump_link_nlmsg(void *cookie, void *msg, struct nlattr **tb)
 	snprintf(netinfo->devices[netinfo->used_len].devname,
 		 sizeof(netinfo->devices[netinfo->used_len].devname),
 		 "%s",
-		 tb[IFLA_IFNAME] ? nla_getattr_str(tb[IFLA_IFNAME]) : "");
+		 tb[IFLA_IFNAME]
+			 ? libbpf_nla_getattr_str(tb[IFLA_IFNAME])
+			 : "");
 	netinfo->used_len++;
 
 	return do_xdp_dump(ifinfo, tb);
@@ -83,7 +85,7 @@ static int dump_class_qdisc_nlmsg(void *cookie, void *msg, struct nlattr **tb)
 	if (tcinfo->is_qdisc) {
 		/* skip clsact qdisc */
 		if (tb[TCA_KIND] &&
-		    strcmp(nla_data(tb[TCA_KIND]), "clsact") == 0)
+		    strcmp(libbpf_nla_data(tb[TCA_KIND]), "clsact") == 0)
 			return 0;
 		if (info->tcm_handle == 0)
 			return 0;
@@ -101,7 +103,9 @@ static int dump_class_qdisc_nlmsg(void *cookie, void *msg, struct nlattr **tb)
 	snprintf(tcinfo->handle_array[tcinfo->used_len].kind,
 		 sizeof(tcinfo->handle_array[tcinfo->used_len].kind),
 		 "%s",
-		 tb[TCA_KIND] ? nla_getattr_str(tb[TCA_KIND]) : "unknown");
+		 tb[TCA_KIND]
+			 ? libbpf_nla_getattr_str(tb[TCA_KIND])
+			 : "unknown");
 	tcinfo->used_len++;
 
 	return 0;
diff --git a/tools/bpf/bpftool/netlink_dumper.c b/tools/bpf/bpftool/netlink_dumper.c
index 6f5e9cc6836c..4e9f4531269f 100644
--- a/tools/bpf/bpftool/netlink_dumper.c
+++ b/tools/bpf/bpftool/netlink_dumper.c
@@ -21,7 +21,7 @@ static void xdp_dump_prog_id(struct nlattr **tb, int attr,
 	if (new_json_object)
 		NET_START_OBJECT
 	NET_DUMP_STR("mode", " %s", mode);
-	NET_DUMP_UINT("id", " id %u", nla_getattr_u32(tb[attr]))
+	NET_DUMP_UINT("id", " id %u", libbpf_nla_getattr_u32(tb[attr]))
 	if (new_json_object)
 		NET_END_OBJECT
 }
@@ -32,13 +32,13 @@ static int do_xdp_dump_one(struct nlattr *attr, unsigned int ifindex,
 	struct nlattr *tb[IFLA_XDP_MAX + 1];
 	unsigned char mode;
 
-	if (nla_parse_nested(tb, IFLA_XDP_MAX, attr, NULL) < 0)
+	if (libbpf_nla_parse_nested(tb, IFLA_XDP_MAX, attr, NULL) < 0)
 		return -1;
 
 	if (!tb[IFLA_XDP_ATTACHED])
 		return 0;
 
-	mode = nla_getattr_u8(tb[IFLA_XDP_ATTACHED]);
+	mode = libbpf_nla_getattr_u8(tb[IFLA_XDP_ATTACHED]);
 	if (mode == XDP_ATTACHED_NONE)
 		return 0;
 
@@ -75,14 +75,14 @@ int do_xdp_dump(struct ifinfomsg *ifinfo, struct nlattr **tb)
 		return 0;
 
 	return do_xdp_dump_one(tb[IFLA_XDP], ifinfo->ifi_index,
-			       nla_getattr_str(tb[IFLA_IFNAME]));
+			       libbpf_nla_getattr_str(tb[IFLA_IFNAME]));
 }
 
 static int do_bpf_dump_one_act(struct nlattr *attr)
 {
 	struct nlattr *tb[TCA_ACT_BPF_MAX + 1];
 
-	if (nla_parse_nested(tb, TCA_ACT_BPF_MAX, attr, NULL) < 0)
+	if (libbpf_nla_parse_nested(tb, TCA_ACT_BPF_MAX, attr, NULL) < 0)
 		return -LIBBPF_ERRNO__NLPARSE;
 
 	if (!tb[TCA_ACT_BPF_PARMS])
@@ -91,10 +91,10 @@ static int do_bpf_dump_one_act(struct nlattr *attr)
 	NET_START_OBJECT_NESTED2;
 	if (tb[TCA_ACT_BPF_NAME])
 		NET_DUMP_STR("name", "%s",
-			     nla_getattr_str(tb[TCA_ACT_BPF_NAME]));
+			     libbpf_nla_getattr_str(tb[TCA_ACT_BPF_NAME]));
 	if (tb[TCA_ACT_BPF_ID])
 		NET_DUMP_UINT("id", " id %u",
-			      nla_getattr_u32(tb[TCA_ACT_BPF_ID]));
+			      libbpf_nla_getattr_u32(tb[TCA_ACT_BPF_ID]));
 	NET_END_OBJECT_NESTED;
 	return 0;
 }
@@ -106,10 +106,11 @@ static int do_dump_one_act(struct nlattr *attr)
 	if (!attr)
 		return 0;
 
-	if (nla_parse_nested(tb, TCA_ACT_MAX, attr, NULL) < 0)
+	if (libbpf_nla_parse_nested(tb, TCA_ACT_MAX, attr, NULL) < 0)
 		return -LIBBPF_ERRNO__NLPARSE;
 
-	if (tb[TCA_ACT_KIND] && strcmp(nla_data(tb[TCA_ACT_KIND]), "bpf") == 0)
+	if (tb[TCA_ACT_KIND] &&
+	    strcmp(libbpf_nla_data(tb[TCA_ACT_KIND]), "bpf") == 0)
 		return do_bpf_dump_one_act(tb[TCA_ACT_OPTIONS]);
 
 	return 0;
@@ -120,7 +121,7 @@ static int do_bpf_act_dump(struct nlattr *attr)
 	struct nlattr *tb[TCA_ACT_MAX_PRIO + 1];
 	int act, ret;
 
-	if (nla_parse_nested(tb, TCA_ACT_MAX_PRIO, attr, NULL) < 0)
+	if (libbpf_nla_parse_nested(tb, TCA_ACT_MAX_PRIO, attr, NULL) < 0)
 		return -LIBBPF_ERRNO__NLPARSE;
 
 	NET_START_ARRAY("act", " %s [");
@@ -139,13 +140,15 @@ static int do_bpf_filter_dump(struct nlattr *attr)
 	struct nlattr *tb[TCA_BPF_MAX + 1];
 	int ret;
 
-	if (nla_parse_nested(tb, TCA_BPF_MAX, attr, NULL) < 0)
+	if (libbpf_nla_parse_nested(tb, TCA_BPF_MAX, attr, NULL) < 0)
 		return -LIBBPF_ERRNO__NLPARSE;
 
 	if (tb[TCA_BPF_NAME])
-		NET_DUMP_STR("name", " %s", nla_getattr_str(tb[TCA_BPF_NAME]));
+		NET_DUMP_STR("name", " %s",
+			     libbpf_nla_getattr_str(tb[TCA_BPF_NAME]));
 	if (tb[TCA_BPF_ID])
-		NET_DUMP_UINT("id", " id %u", nla_getattr_u32(tb[TCA_BPF_ID]));
+		NET_DUMP_UINT("id", " id %u",
+			      libbpf_nla_getattr_u32(tb[TCA_BPF_ID]));
 	if (tb[TCA_BPF_ACT]) {
 		ret = do_bpf_act_dump(tb[TCA_BPF_ACT]);
 		if (ret)
@@ -160,7 +163,8 @@ int do_filter_dump(struct tcmsg *info, struct nlattr **tb, const char *kind,
 {
 	int ret = 0;
 
-	if (tb[TCA_OPTIONS] && strcmp(nla_data(tb[TCA_KIND]), "bpf") == 0) {
+	if (tb[TCA_OPTIONS] &&
+	    strcmp(libbpf_nla_data(tb[TCA_KIND]), "bpf") == 0) {
 		NET_START_OBJECT;
 		if (devname[0] != '\0')
 			NET_DUMP_STR("devname", "%s", devname);
diff --git a/tools/lib/bpf/netlink.c b/tools/lib/bpf/netlink.c
index 506bdfdbcab0..2d2edbbd8ae8 100644
--- a/tools/lib/bpf/netlink.c
+++ b/tools/lib/bpf/netlink.c
@@ -103,7 +103,7 @@ static int bpf_netlink_recv(int sock, __u32 nl_pid, int seq,
 				if (!err->error)
 					continue;
 				ret = err->error;
-				nla_dump_errormsg(nh);
+				libbpf_nla_dump_errormsg(nh);
 				goto done;
 			case NLMSG_DONE:
 				return 0;
@@ -190,7 +190,7 @@ static int __dump_link_nlmsg(struct nlmsghdr *nlh,
 
 	len = nlh->nlmsg_len - NLMSG_LENGTH(sizeof(*ifi));
 	attr = (struct nlattr *) ((void *) ifi + NLMSG_ALIGN(sizeof(*ifi)));
-	if (nla_parse(tb, IFLA_MAX, attr, len, NULL) != 0)
+	if (libbpf_nla_parse(tb, IFLA_MAX, attr, len, NULL) != 0)
 		return -LIBBPF_ERRNO__NLPARSE;
 
 	return dump_link_nlmsg(cookie, ifi, tb);
@@ -228,7 +228,7 @@ static int __dump_class_nlmsg(struct nlmsghdr *nlh,
 
 	len = nlh->nlmsg_len - NLMSG_LENGTH(sizeof(*t));
 	attr = (struct nlattr *) ((void *) t + NLMSG_ALIGN(sizeof(*t)));
-	if (nla_parse(tb, TCA_MAX, attr, len, NULL) != 0)
+	if (libbpf_nla_parse(tb, TCA_MAX, attr, len, NULL) != 0)
 		return -LIBBPF_ERRNO__NLPARSE;
 
 	return dump_class_nlmsg(cookie, t, tb);
@@ -267,7 +267,7 @@ static int __dump_qdisc_nlmsg(struct nlmsghdr *nlh,
 
 	len = nlh->nlmsg_len - NLMSG_LENGTH(sizeof(*t));
 	attr = (struct nlattr *) ((void *) t + NLMSG_ALIGN(sizeof(*t)));
-	if (nla_parse(tb, TCA_MAX, attr, len, NULL) != 0)
+	if (libbpf_nla_parse(tb, TCA_MAX, attr, len, NULL) != 0)
 		return -LIBBPF_ERRNO__NLPARSE;
 
 	return dump_qdisc_nlmsg(cookie, t, tb);
@@ -306,7 +306,7 @@ static int __dump_filter_nlmsg(struct nlmsghdr *nlh,
 
 	len = nlh->nlmsg_len - NLMSG_LENGTH(sizeof(*t));
 	attr = (struct nlattr *) ((void *) t + NLMSG_ALIGN(sizeof(*t)));
-	if (nla_parse(tb, TCA_MAX, attr, len, NULL) != 0)
+	if (libbpf_nla_parse(tb, TCA_MAX, attr, len, NULL) != 0)
 		return -LIBBPF_ERRNO__NLPARSE;
 
 	return dump_filter_nlmsg(cookie, t, tb);
diff --git a/tools/lib/bpf/nlattr.c b/tools/lib/bpf/nlattr.c
index 49f514119bdb..e52257a7367a 100644
--- a/tools/lib/bpf/nlattr.c
+++ b/tools/lib/bpf/nlattr.c
@@ -17,13 +17,13 @@
 #include <string.h>
 #include <stdio.h>
 
-static uint16_t nla_attr_minlen[NLA_TYPE_MAX+1] = {
-	[NLA_U8]	= sizeof(uint8_t),
-	[NLA_U16]	= sizeof(uint16_t),
-	[NLA_U32]	= sizeof(uint32_t),
-	[NLA_U64]	= sizeof(uint64_t),
-	[NLA_STRING]	= 1,
-	[NLA_FLAG]	= 0,
+static uint16_t nla_attr_minlen[LIBBPF_NLA_TYPE_MAX+1] = {
+	[LIBBPF_NLA_U8]		= sizeof(uint8_t),
+	[LIBBPF_NLA_U16]	= sizeof(uint16_t),
+	[LIBBPF_NLA_U32]	= sizeof(uint32_t),
+	[LIBBPF_NLA_U64]	= sizeof(uint64_t),
+	[LIBBPF_NLA_STRING]	= 1,
+	[LIBBPF_NLA_FLAG]	= 0,
 };
 
 static struct nlattr *nla_next(const struct nlattr *nla, int *remaining)
@@ -47,9 +47,9 @@ static int nla_type(const struct nlattr *nla)
 }
 
 static int validate_nla(struct nlattr *nla, int maxtype,
-			struct nla_policy *policy)
+			struct libbpf_nla_policy *policy)
 {
-	struct nla_policy *pt;
+	struct libbpf_nla_policy *pt;
 	unsigned int minlen = 0;
 	int type = nla_type(nla);
 
@@ -58,23 +58,24 @@ static int validate_nla(struct nlattr *nla, int maxtype,
 
 	pt = &policy[type];
 
-	if (pt->type > NLA_TYPE_MAX)
+	if (pt->type > LIBBPF_NLA_TYPE_MAX)
 		return 0;
 
 	if (pt->minlen)
 		minlen = pt->minlen;
-	else if (pt->type != NLA_UNSPEC)
+	else if (pt->type != LIBBPF_NLA_UNSPEC)
 		minlen = nla_attr_minlen[pt->type];
 
-	if (nla_len(nla) < minlen)
+	if (libbpf_nla_len(nla) < minlen)
 		return -1;
 
-	if (pt->maxlen && nla_len(nla) > pt->maxlen)
+	if (pt->maxlen && libbpf_nla_len(nla) > pt->maxlen)
 		return -1;
 
-	if (pt->type == NLA_STRING) {
-		char *data = nla_data(nla);
-		if (data[nla_len(nla) - 1] != '\0')
+	if (pt->type == LIBBPF_NLA_STRING) {
+		char *data = libbpf_nla_data(nla);
+
+		if (data[libbpf_nla_len(nla) - 1] != '\0')
 			return -1;
 	}
 
@@ -104,15 +105,15 @@ static inline int nlmsg_len(const struct nlmsghdr *nlh)
  * @see nla_validate
  * @return 0 on success or a negative error code.
  */
-int nla_parse(struct nlattr *tb[], int maxtype, struct nlattr *head, int len,
-	      struct nla_policy *policy)
+int libbpf_nla_parse(struct nlattr *tb[], int maxtype, struct nlattr *head,
+		     int len, struct libbpf_nla_policy *policy)
 {
 	struct nlattr *nla;
 	int rem, err;
 
 	memset(tb, 0, sizeof(struct nlattr *) * (maxtype + 1));
 
-	nla_for_each_attr(nla, head, len, rem) {
+	libbpf_nla_for_each_attr(nla, head, len, rem) {
 		int type = nla_type(nla);
 
 		if (type > maxtype)
@@ -144,23 +145,25 @@ int nla_parse(struct nlattr *tb[], int maxtype, struct nlattr *head, int len,
  * @arg policy          Attribute validation policy.
  *
  * Feeds the stream of attributes nested into the specified attribute
- * to nla_parse().
+ * to libbpf_nla_parse().
  *
- * @see nla_parse
+ * @see libbpf_nla_parse
  * @return 0 on success or a negative error code.
  */
-int nla_parse_nested(struct nlattr *tb[], int maxtype, struct nlattr *nla,
-		     struct nla_policy *policy)
+int libbpf_nla_parse_nested(struct nlattr *tb[], int maxtype,
+			    struct nlattr *nla,
+			    struct libbpf_nla_policy *policy)
 {
-	return nla_parse(tb, maxtype, nla_data(nla), nla_len(nla), policy);
+	return libbpf_nla_parse(tb, maxtype, libbpf_nla_data(nla),
+				libbpf_nla_len(nla), policy);
 }
 
 /* dump netlink extended ack error message */
-int nla_dump_errormsg(struct nlmsghdr *nlh)
+int libbpf_nla_dump_errormsg(struct nlmsghdr *nlh)
 {
-	struct nla_policy extack_policy[NLMSGERR_ATTR_MAX + 1] = {
-		[NLMSGERR_ATTR_MSG]	= { .type = NLA_STRING },
-		[NLMSGERR_ATTR_OFFS]	= { .type = NLA_U32 },
+	struct libbpf_nla_policy extack_policy[NLMSGERR_ATTR_MAX + 1] = {
+		[NLMSGERR_ATTR_MSG]	= { .type = LIBBPF_NLA_STRING },
+		[NLMSGERR_ATTR_OFFS]	= { .type = LIBBPF_NLA_U32 },
 	};
 	struct nlattr *tb[NLMSGERR_ATTR_MAX + 1], *attr;
 	struct nlmsgerr *err;
@@ -181,14 +184,15 @@ int nla_dump_errormsg(struct nlmsghdr *nlh)
 	attr = (struct nlattr *) ((void *) err + hlen);
 	alen = nlh->nlmsg_len - hlen;
 
-	if (nla_parse(tb, NLMSGERR_ATTR_MAX, attr, alen, extack_policy) != 0) {
+	if (libbpf_nla_parse(tb, NLMSGERR_ATTR_MAX, attr, alen,
+			     extack_policy) != 0) {
 		fprintf(stderr,
 			"Failed to parse extended error attributes\n");
 		return 0;
 	}
 
 	if (tb[NLMSGERR_ATTR_MSG])
-		errmsg = (char *) nla_data(tb[NLMSGERR_ATTR_MSG]);
+		errmsg = (char *) libbpf_nla_data(tb[NLMSGERR_ATTR_MSG]);
 
 	fprintf(stderr, "Kernel error message: %s\n", errmsg);
 
diff --git a/tools/lib/bpf/nlattr.h b/tools/lib/bpf/nlattr.h
index a6e2396bce7c..755a3312c87f 100644
--- a/tools/lib/bpf/nlattr.h
+++ b/tools/lib/bpf/nlattr.h
@@ -23,19 +23,19 @@
  * Standard attribute types to specify validation policy
  */
 enum {
-	NLA_UNSPEC,	/**< Unspecified type, binary data chunk */
-	NLA_U8,		/**< 8 bit integer */
-	NLA_U16,	/**< 16 bit integer */
-	NLA_U32,	/**< 32 bit integer */
-	NLA_U64,	/**< 64 bit integer */
-	NLA_STRING,	/**< NUL terminated character string */
-	NLA_FLAG,	/**< Flag */
-	NLA_MSECS,	/**< Micro seconds (64bit) */
-	NLA_NESTED,	/**< Nested attributes */
-	__NLA_TYPE_MAX,
+	LIBBPF_NLA_UNSPEC,	/**< Unspecified type, binary data chunk */
+	LIBBPF_NLA_U8,		/**< 8 bit integer */
+	LIBBPF_NLA_U16,		/**< 16 bit integer */
+	LIBBPF_NLA_U32,		/**< 32 bit integer */
+	LIBBPF_NLA_U64,		/**< 64 bit integer */
+	LIBBPF_NLA_STRING,	/**< NUL terminated character string */
+	LIBBPF_NLA_FLAG,	/**< Flag */
+	LIBBPF_NLA_MSECS,	/**< Micro seconds (64bit) */
+	LIBBPF_NLA_NESTED,	/**< Nested attributes */
+	__LIBBPF_NLA_TYPE_MAX,
 };
 
-#define NLA_TYPE_MAX (__NLA_TYPE_MAX - 1)
+#define LIBBPF_NLA_TYPE_MAX (__LIBBPF_NLA_TYPE_MAX - 1)
 
 /**
  * @ingroup attr
@@ -43,8 +43,8 @@ enum {
  *
  * See section @core_doc{core_attr_parse,Attribute Parsing} for more details.
  */
-struct nla_policy {
-	/** Type of attribute or NLA_UNSPEC */
+struct libbpf_nla_policy {
+	/** Type of attribute or LIBBPF_NLA_UNSPEC */
 	uint16_t	type;
 
 	/** Minimal length of payload required */
@@ -62,49 +62,50 @@ struct nla_policy {
  * @arg len	length of attribute stream
  * @arg rem	initialized to len, holds bytes currently remaining in stream
  */
-#define nla_for_each_attr(pos, head, len, rem) \
+#define libbpf_nla_for_each_attr(pos, head, len, rem) \
 	for (pos = head, rem = len; \
 	     nla_ok(pos, rem); \
 	     pos = nla_next(pos, &(rem)))
 
 /**
- * nla_data - head of payload
+ * libbpf_nla_data - head of payload
  * @nla: netlink attribute
  */
-static inline void *nla_data(const struct nlattr *nla)
+static inline void *libbpf_nla_data(const struct nlattr *nla)
 {
 	return (char *) nla + NLA_HDRLEN;
 }
 
-static inline uint8_t nla_getattr_u8(const struct nlattr *nla)
+static inline uint8_t libbpf_nla_getattr_u8(const struct nlattr *nla)
 {
-	return *(uint8_t *)nla_data(nla);
+	return *(uint8_t *)libbpf_nla_data(nla);
 }
 
-static inline uint32_t nla_getattr_u32(const struct nlattr *nla)
+static inline uint32_t libbpf_nla_getattr_u32(const struct nlattr *nla)
 {
-	return *(uint32_t *)nla_data(nla);
+	return *(uint32_t *)libbpf_nla_data(nla);
 }
 
-static inline const char *nla_getattr_str(const struct nlattr *nla)
+static inline const char *libbpf_nla_getattr_str(const struct nlattr *nla)
 {
-	return (const char *)nla_data(nla);
+	return (const char *)libbpf_nla_data(nla);
 }
 
 /**
- * nla_len - length of payload
+ * libbpf_nla_len - length of payload
  * @nla: netlink attribute
  */
-static inline int nla_len(const struct nlattr *nla)
+static inline int libbpf_nla_len(const struct nlattr *nla)
 {
 	return nla->nla_len - NLA_HDRLEN;
 }
 
-int nla_parse(struct nlattr *tb[], int maxtype, struct nlattr *head, int len,
-	      struct nla_policy *policy);
-int nla_parse_nested(struct nlattr *tb[], int maxtype, struct nlattr *nla,
-		     struct nla_policy *policy);
+int libbpf_nla_parse(struct nlattr *tb[], int maxtype, struct nlattr *head,
+		     int len, struct libbpf_nla_policy *policy);
+int libbpf_nla_parse_nested(struct nlattr *tb[], int maxtype,
+			    struct nlattr *nla,
+			    struct libbpf_nla_policy *policy);
 
-int nla_dump_errormsg(struct nlmsghdr *nlh);
+int libbpf_nla_dump_errormsg(struct nlmsghdr *nlh);
 
 #endif /* __NLATTR_H */
-- 
2.17.1

[PATCH bpf-next 4/6] libbpf: Consistent prefixes for interfaces in str_error.h.

[Andrey Ignatov]

libbpf is used more and more outside kernel tree. That means the library
should follow good practices in library design and implementation to
play well with third party code that uses it.

One of such practices is to have a common prefix (or a few) for every
interface, function or data structure, library provides. I helps to
avoid name conflicts with other libraries and keeps API consistent.

Inconsistent names in libbpf already cause problems in real life. E.g.
an application can't use both libbpf and libnl due to conflicting
symbols.

Having common prefix will help to fix current and avoid future problems.

libbpf already uses the following prefixes for its interfaces:
* bpf_ for bpf system call wrappers, program/map/elf-object
  abstractions and a few other things;
* btf_ for BTF related API;
* libbpf_ for everything else.

The patch renames function in str_error.h to have libbpf_ prefix since it
misses one and doesn't fit well into the first two categories.

Signed-off-by: Andrey Ignatov <rdna@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
---
 tools/lib/bpf/libbpf.c    | 20 +++++++++++---------
 tools/lib/bpf/str_error.c |  2 +-
 tools/lib/bpf/str_error.h |  2 +-
 3 files changed, 13 insertions(+), 11 deletions(-)

diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index 9e68fd9fcfca..02888d36b805 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -470,7 +470,8 @@ static int bpf_object__elf_init(struct bpf_object *obj)
 		obj->efile.fd = open(obj->path, O_RDONLY);
 		if (obj->efile.fd < 0) {
 			char errmsg[STRERR_BUFSIZE];
-			char *cp = str_error(errno, errmsg, sizeof(errmsg));
+			char *cp = libbpf_strerror_r(errno, errmsg,
+						     sizeof(errmsg));
 
 			pr_warning("failed to open %s: %s\n", obj->path, cp);
 			return -errno;
@@ -811,7 +812,8 @@ static int bpf_object__elf_collect(struct bpf_object *obj)
 						      data->d_size, name, idx);
 			if (err) {
 				char errmsg[STRERR_BUFSIZE];
-				char *cp = str_error(-err, errmsg, sizeof(errmsg));
+				char *cp = libbpf_strerror_r(-err, errmsg,
+							     sizeof(errmsg));
 
 				pr_warning("failed to alloc program %s (%s): %s",
 					   name, obj->path, cp);
@@ -1140,7 +1142,7 @@ bpf_object__create_maps(struct bpf_object *obj)
 
 		*pfd = bpf_create_map_xattr(&create_attr);
 		if (*pfd < 0 && create_attr.btf_key_type_id) {
-			cp = str_error(errno, errmsg, sizeof(errmsg));
+			cp = libbpf_strerror_r(errno, errmsg, sizeof(errmsg));
 			pr_warning("Error in bpf_create_map_xattr(%s):%s(%d). Retrying without BTF.\n",
 				   map->name, cp, errno);
 			create_attr.btf_fd = 0;
@@ -1155,7 +1157,7 @@ bpf_object__create_maps(struct bpf_object *obj)
 			size_t j;
 
 			err = *pfd;
-			cp = str_error(errno, errmsg, sizeof(errmsg));
+			cp = libbpf_strerror_r(errno, errmsg, sizeof(errmsg));
 			pr_warning("failed to create map (name: '%s'): %s\n",
 				   map->name, cp);
 			for (j = 0; j < i; j++)
@@ -1339,7 +1341,7 @@ load_program(enum bpf_prog_type type, enum bpf_attach_type expected_attach_type,
 	}
 
 	ret = -LIBBPF_ERRNO__LOAD;
-	cp = str_error(errno, errmsg, sizeof(errmsg));
+	cp = libbpf_strerror_r(errno, errmsg, sizeof(errmsg));
 	pr_warning("load bpf program failed: %s\n", cp);
 
 	if (log_buf && log_buf[0] != '\0') {
@@ -1655,7 +1657,7 @@ static int check_path(const char *path)
 
 	dir = dirname(dname);
 	if (statfs(dir, &st_fs)) {
-		cp = str_error(errno, errmsg, sizeof(errmsg));
+		cp = libbpf_strerror_r(errno, errmsg, sizeof(errmsg));
 		pr_warning("failed to statfs %s: %s\n", dir, cp);
 		err = -errno;
 	}
@@ -1691,7 +1693,7 @@ int bpf_program__pin_instance(struct bpf_program *prog, const char *path,
 	}
 
 	if (bpf_obj_pin(prog->instances.fds[instance], path)) {
-		cp = str_error(errno, errmsg, sizeof(errmsg));
+		cp = libbpf_strerror_r(errno, errmsg, sizeof(errmsg));
 		pr_warning("failed to pin program: %s\n", cp);
 		return -errno;
 	}
@@ -1709,7 +1711,7 @@ static int make_dir(const char *path)
 		err = -errno;
 
 	if (err) {
-		cp = str_error(-err, errmsg, sizeof(errmsg));
+		cp = libbpf_strerror_r(-err, errmsg, sizeof(errmsg));
 		pr_warning("failed to mkdir %s: %s\n", path, cp);
 	}
 	return err;
@@ -1771,7 +1773,7 @@ int bpf_map__pin(struct bpf_map *map, const char *path)
 	}
 
 	if (bpf_obj_pin(map->fd, path)) {
-		cp = str_error(errno, errmsg, sizeof(errmsg));
+		cp = libbpf_strerror_r(errno, errmsg, sizeof(errmsg));
 		pr_warning("failed to pin map: %s\n", cp);
 		return -errno;
 	}
diff --git a/tools/lib/bpf/str_error.c b/tools/lib/bpf/str_error.c
index b8798114a357..3d211b642cb5 100644
--- a/tools/lib/bpf/str_error.c
+++ b/tools/lib/bpf/str_error.c
@@ -9,7 +9,7 @@
  * libc, while checking strerror_r() return to avoid having to check this in
  * all places calling it.
  */
-char *str_error(int err, char *dst, int len)
+char *libbpf_strerror_r(int err, char *dst, int len)
 {
 	int ret = strerror_r(err, dst, len);
 	if (ret)
diff --git a/tools/lib/bpf/str_error.h b/tools/lib/bpf/str_error.h
index 355b1db571d1..998eff7d6710 100644
--- a/tools/lib/bpf/str_error.h
+++ b/tools/lib/bpf/str_error.h
@@ -2,5 +2,5 @@
 #ifndef BPF_STR_ERROR
 #define BPF_STR_ERROR
 
-char *str_error(int err, char *dst, int len);
+char *libbpf_strerror_r(int err, char *dst, int len);
 #endif // BPF_STR_ERROR
-- 
2.17.1

[PATCH bpf-next 5/6] libbpf: Make include guards consistent

[Andrey Ignatov]

Rename include guards to have consistent names "__LIBBPF_<header_name>".

Signed-off-by: Andrey Ignatov <rdna@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
---
 tools/lib/bpf/bpf.h       | 6 +++---
 tools/lib/bpf/btf.h       | 6 +++---
 tools/lib/bpf/libbpf.h    | 6 +++---
 tools/lib/bpf/nlattr.h    | 6 +++---
 tools/lib/bpf/str_error.h | 6 +++---
 5 files changed, 15 insertions(+), 15 deletions(-)

diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h
index 6f38164b2618..4c78f61b7c71 100644
--- a/tools/lib/bpf/bpf.h
+++ b/tools/lib/bpf/bpf.h
@@ -20,8 +20,8 @@
  * You should have received a copy of the GNU Lesser General Public
  * License along with this program; if not,  see <http://www.gnu.org/licenses>
  */
-#ifndef __BPF_BPF_H
-#define __BPF_BPF_H
+#ifndef __LIBBPF_BPF_H
+#define __LIBBPF_BPF_H
 
 #include <linux/bpf.h>
 #include <stdbool.h>
@@ -111,4 +111,4 @@ int bpf_load_btf(void *btf, __u32 btf_size, char *log_buf, __u32 log_buf_size,
 int bpf_task_fd_query(int pid, int fd, __u32 flags, char *buf, __u32 *buf_len,
 		      __u32 *prog_id, __u32 *fd_type, __u64 *probe_offset,
 		      __u64 *probe_addr);
-#endif
+#endif /* __LIBBPF_BPF_H */
diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h
index 4897e0724d4e..d5d20682eeb6 100644
--- a/tools/lib/bpf/btf.h
+++ b/tools/lib/bpf/btf.h
@@ -1,8 +1,8 @@
 /* SPDX-License-Identifier: LGPL-2.1 */
 /* Copyright (c) 2018 Facebook */
 
-#ifndef __BPF_BTF_H
-#define __BPF_BTF_H
+#ifndef __LIBBPF_BTF_H
+#define __LIBBPF_BTF_H
 
 #include <linux/types.h>
 
@@ -23,4 +23,4 @@ int btf__resolve_type(const struct btf *btf, __u32 type_id);
 int btf__fd(const struct btf *btf);
 const char *btf__name_by_offset(const struct btf *btf, __u32 offset);
 
-#endif
+#endif /* __LIBBPF_BTF_H */
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 710ff5724980..28f83dd6022b 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -20,8 +20,8 @@
  * You should have received a copy of the GNU Lesser General Public
  * License along with this program; if not,  see <http://www.gnu.org/licenses>
  */
-#ifndef __BPF_LIBBPF_H
-#define __BPF_LIBBPF_H
+#ifndef __LIBBPF_LIBBPF_H
+#define __LIBBPF_LIBBPF_H
 
 #include <stdio.h>
 #include <stdint.h>
@@ -315,4 +315,4 @@ int libbpf_nl_get_qdisc(int sock, unsigned int nl_pid, int ifindex,
 			libbpf_dump_nlmsg_t dump_qdisc_nlmsg, void *cookie);
 int libbpf_nl_get_filter(int sock, unsigned int nl_pid, int ifindex, int handle,
 			 libbpf_dump_nlmsg_t dump_filter_nlmsg, void *cookie);
-#endif
+#endif /* __LIBBPF_LIBBPF_H */
diff --git a/tools/lib/bpf/nlattr.h b/tools/lib/bpf/nlattr.h
index 755a3312c87f..7198584a3040 100644
--- a/tools/lib/bpf/nlattr.h
+++ b/tools/lib/bpf/nlattr.h
@@ -11,8 +11,8 @@
  * Copyright (c) 2003-2013 Thomas Graf <tgraf@suug.ch>
  */
 
-#ifndef __NLATTR_H
-#define __NLATTR_H
+#ifndef __LIBBPF_NLATTR_H
+#define __LIBBPF_NLATTR_H
 
 #include <stdint.h>
 #include <linux/netlink.h>
@@ -108,4 +108,4 @@ int libbpf_nla_parse_nested(struct nlattr *tb[], int maxtype,
 
 int libbpf_nla_dump_errormsg(struct nlmsghdr *nlh);
 
-#endif /* __NLATTR_H */
+#endif /* __LIBBPF_NLATTR_H */
diff --git a/tools/lib/bpf/str_error.h b/tools/lib/bpf/str_error.h
index 998eff7d6710..b9157f5eebde 100644
--- a/tools/lib/bpf/str_error.h
+++ b/tools/lib/bpf/str_error.h
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: LGPL-2.1
-#ifndef BPF_STR_ERROR
-#define BPF_STR_ERROR
+#ifndef __LIBBPF_STR_ERROR_H
+#define __LIBBPF_STR_ERROR_H
 
 char *libbpf_strerror_r(int err, char *dst, int len);
-#endif // BPF_STR_ERROR
+#endif /* __LIBBPF_STR_ERROR_H */
-- 
2.17.1

[PATCH bpf-next 6/6] libbpf: Use __u32 instead of u32 in bpf_program__load

[Andrey Ignatov]

Make bpf_program__load consistent with other interfaces: use __u32
instead of u32. That in turn fixes build of samples:

In file included from ./samples/bpf/trace_output_user.c:21:0:
./tools/lib/bpf/libbpf.h:132:9: error: unknown type name ‘u32’
         u32 kern_version);
         ^

Fixes: commit 29cd77f41620d ("libbpf: Support loading individual progs")
Signed-off-by: Andrey Ignatov <rdna@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
---
 tools/lib/bpf/libbpf.c | 2 +-
 tools/lib/bpf/libbpf.h | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index 02888d36b805..85de1ebd4cb0 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -1379,7 +1379,7 @@ load_program(enum bpf_prog_type type, enum bpf_attach_type expected_attach_type,
 
 int
 bpf_program__load(struct bpf_program *prog,
-		  char *license, u32 kern_version)
+		  char *license, __u32 kern_version)
 {
 	int err = 0, fd, i;
 
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 28f83dd6022b..fbfc2aec0f0d 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -129,7 +129,7 @@ void bpf_program__set_ifindex(struct bpf_program *prog, __u32 ifindex);
 const char *bpf_program__title(struct bpf_program *prog, bool needs_copy);
 
 int bpf_program__load(struct bpf_program *prog, char *license,
-		      u32 kern_version);
+		      __u32 kern_version);
 int bpf_program__fd(struct bpf_program *prog);
 int bpf_program__pin_instance(struct bpf_program *prog, const char *path,
 			      int instance);
-- 
2.17.1

Re: [PATCH bpf-next 0/6] Consistent prefixes for libbpf interfaces

[Daniel Borkmann]

[ +Yonghong ]

On 10/04/2018 12:26 AM, Andrey Ignatov wrote:
> This patch set renames a few interfaces in libbpf, mostly netlink related,
> so that all symbols provided by the library have only three possible
> prefixes:
> 
> % nm -D tools/lib/bpf/libbpf.so  | \
>     awk '$2 == "T" {sub(/[_\(].*/, "", $3); if ($3) print $3}' | \
>     sort | \
>     uniq -c
>      91 bpf
>       8 btf
>      14 libbpf
> 
> libbpf is used more and more outside kernel tree. That means the library
> should follow good practices in library design and implementation to
> play well with third party code that uses it.
> 
> One of such practices is to have a common prefix (or a few) for every
> interface, function or data structure, library provides. It helps to
> avoid name conflicts with other libraries and keeps API/ABI consistent.
> 
> Inconsistent names in libbpf already cause problems in real life. E.g.
> an application can't use both libbpf and libnl due to conflicting
> symbols (specifically nla_parse, nla_parse_nested and a few others).
> 
> Some of problematic global symbols are not part of ABI and can be
> restricted from export with either visibility attribute/pragma or export
> map (what is useful by itself and can be done in addition). That won't
> solve the problem for those that are part of ABI though. Also export
> restrictions would help only in DSO case. If third party application links
> libbpf statically it won't help, and people do it (e.g. Facebook links
> most of libraries statically, including libbpf).
> 
> libbpf already uses the following prefixes for its interfaces:
> * bpf_ for bpf system call wrappers, program/map/elf-object
>   abstractions and a few other things;
> * btf_ for BTF related API;
> * libbpf_ for everything else.
> 
> The patch adds libbpf_ prefix to interfaces that use none of mentioned
> above prefixes and don't fit well into the first two categories.
> 
> Long term benefits of having common prefix should outweigh possible
> inconvenience of changing API for those functions now.
> 
> Patches 2-4 add libbpf_ prefix to libbpf interfaces: separate patch per
> header. Other patches are simple improvements in API.
> 
> 
> Andrey Ignatov (6):
>   libbpf: Move __dump_nlmsg_t from API to implementation
>   libbpf: Consistent prefixes for interfaces in libbpf.h.
>   libbpf: Consistent prefixes for interfaces in nlattr.h.
>   libbpf: Consistent prefixes for interfaces in str_error.h.
>   libbpf: Make include guards consistent
>   libbpf: Use __u32 instead of u32 in bpf_program__load
> 
>  tools/bpf/bpftool/net.c            | 41 ++++++++++---------
>  tools/bpf/bpftool/netlink_dumper.c | 32 ++++++++-------
>  tools/lib/bpf/bpf.h                |  6 +--
>  tools/lib/bpf/btf.h                |  6 +--
>  tools/lib/bpf/libbpf.c             | 22 +++++-----
>  tools/lib/bpf/libbpf.h             | 31 +++++++-------
>  tools/lib/bpf/netlink.c            | 48 ++++++++++++----------
>  tools/lib/bpf/nlattr.c             | 64 +++++++++++++++--------------
>  tools/lib/bpf/nlattr.h             | 65 +++++++++++++++---------------
>  tools/lib/bpf/str_error.c          |  2 +-
>  tools/lib/bpf/str_error.h          |  8 ++--
>  11 files changed, 171 insertions(+), 154 deletions(-)

Overall agree that this is needed, and I've therefore applied the
set, thanks for cleaning up, Andrey!

But, I would actually like to see this going one step further, in
particular, we should keep all the netlink related stuff inside
libbpf-/only/. Meaning, the goal of libbpf is not to provide yet
another set of netlink primitives but instead e.g. for the bpftool
dumper this should be abstracted away such that we pass in a callback
from bpftool side and have an iterator object which will then be
populated from inside the libbpf logic, meaning, bpftool shouldn't
even be aware of anything netlink there.

Thanks,
Daniel

Re: [PATCH bpf-next 0/6] Consistent prefixes for libbpf interfaces

[Yonghong Song]

On 10/4/18 7:22 AM, Daniel Borkmann wrote:
> [ +Yonghong ]
> 
> On 10/04/2018 12:26 AM, Andrey Ignatov wrote:
>> This patch set renames a few interfaces in libbpf, mostly netlink related,
>> so that all symbols provided by the library have only three possible
>> prefixes:
>>
>> % nm -D tools/lib/bpf/libbpf.so  | \
>>      awk '$2 == "T" {sub(/[_\(].*/, "", $3); if ($3) print $3}' | \
>>      sort | \
>>      uniq -c
>>       91 bpf
>>        8 btf
>>       14 libbpf
>>
>> libbpf is used more and more outside kernel tree. That means the library
>> should follow good practices in library design and implementation to
>> play well with third party code that uses it.
>>
>> One of such practices is to have a common prefix (or a few) for every
>> interface, function or data structure, library provides. It helps to
>> avoid name conflicts with other libraries and keeps API/ABI consistent.
>>
>> Inconsistent names in libbpf already cause problems in real life. E.g.
>> an application can't use both libbpf and libnl due to conflicting
>> symbols (specifically nla_parse, nla_parse_nested and a few others).
>>
>> Some of problematic global symbols are not part of ABI and can be
>> restricted from export with either visibility attribute/pragma or export
>> map (what is useful by itself and can be done in addition). That won't
>> solve the problem for those that are part of ABI though. Also export
>> restrictions would help only in DSO case. If third party application links
>> libbpf statically it won't help, and people do it (e.g. Facebook links
>> most of libraries statically, including libbpf).
>>
>> libbpf already uses the following prefixes for its interfaces:
>> * bpf_ for bpf system call wrappers, program/map/elf-object
>>    abstractions and a few other things;
>> * btf_ for BTF related API;
>> * libbpf_ for everything else.
>>
>> The patch adds libbpf_ prefix to interfaces that use none of mentioned
>> above prefixes and don't fit well into the first two categories.
>>
>> Long term benefits of having common prefix should outweigh possible
>> inconvenience of changing API for those functions now.
>>
>> Patches 2-4 add libbpf_ prefix to libbpf interfaces: separate patch per
>> header. Other patches are simple improvements in API.
>>
>>
>> Andrey Ignatov (6):
>>    libbpf: Move __dump_nlmsg_t from API to implementation
>>    libbpf: Consistent prefixes for interfaces in libbpf.h.
>>    libbpf: Consistent prefixes for interfaces in nlattr.h.
>>    libbpf: Consistent prefixes for interfaces in str_error.h.
>>    libbpf: Make include guards consistent
>>    libbpf: Use __u32 instead of u32 in bpf_program__load
>>
>>   tools/bpf/bpftool/net.c            | 41 ++++++++++---------
>>   tools/bpf/bpftool/netlink_dumper.c | 32 ++++++++-------
>>   tools/lib/bpf/bpf.h                |  6 +--
>>   tools/lib/bpf/btf.h                |  6 +--
>>   tools/lib/bpf/libbpf.c             | 22 +++++-----
>>   tools/lib/bpf/libbpf.h             | 31 +++++++-------
>>   tools/lib/bpf/netlink.c            | 48 ++++++++++++----------
>>   tools/lib/bpf/nlattr.c             | 64 +++++++++++++++--------------
>>   tools/lib/bpf/nlattr.h             | 65 +++++++++++++++---------------
>>   tools/lib/bpf/str_error.c          |  2 +-
>>   tools/lib/bpf/str_error.h          |  8 ++--
>>   11 files changed, 171 insertions(+), 154 deletions(-)
> 
> Overall agree that this is needed, and I've therefore applied the
> set, thanks for cleaning up, Andrey!
> 
> But, I would actually like to see this going one step further, in
> particular, we should keep all the netlink related stuff inside
> libbpf-/only/. Meaning, the goal of libbpf is not to provide yet
> another set of netlink primitives but instead e.g. for the bpftool
> dumper this should be abstracted away such that we pass in a callback
> from bpftool side and have an iterator object which will then be
> populated from inside the libbpf logic, meaning, bpftool shouldn't
> even be aware of anything netlink there.

Agreed. This indeed make sense, the user really only cares a few fields
like devname/id, attachment_type, prog_id, etc. I will take a look at
this later if nobody works on it.

> 
> Thanks,
> Daniel
> 

Re: [PATCH bpf-next 0/6] Consistent prefixes for libbpf interfaces

[Daniel Borkmann]

On 10/09/2018 08:43 AM, Yonghong Song wrote:
> On 10/4/18 7:22 AM, Daniel Borkmann wrote:
>> [ +Yonghong ]
>>
>> On 10/04/2018 12:26 AM, Andrey Ignatov wrote:
>>> This patch set renames a few interfaces in libbpf, mostly netlink related,
>>> so that all symbols provided by the library have only three possible
>>> prefixes:
>>>
>>> % nm -D tools/lib/bpf/libbpf.so  | \
>>>      awk '$2 == "T" {sub(/[_\(].*/, "", $3); if ($3) print $3}' | \
>>>      sort | \
>>>      uniq -c
>>>       91 bpf
>>>        8 btf
>>>       14 libbpf
>>>
>>> libbpf is used more and more outside kernel tree. That means the library
>>> should follow good practices in library design and implementation to
>>> play well with third party code that uses it.
>>>
>>> One of such practices is to have a common prefix (or a few) for every
>>> interface, function or data structure, library provides. It helps to
>>> avoid name conflicts with other libraries and keeps API/ABI consistent.
>>>
>>> Inconsistent names in libbpf already cause problems in real life. E.g.
>>> an application can't use both libbpf and libnl due to conflicting
>>> symbols (specifically nla_parse, nla_parse_nested and a few others).
>>>
>>> Some of problematic global symbols are not part of ABI and can be
>>> restricted from export with either visibility attribute/pragma or export
>>> map (what is useful by itself and can be done in addition). That won't
>>> solve the problem for those that are part of ABI though. Also export
>>> restrictions would help only in DSO case. If third party application links
>>> libbpf statically it won't help, and people do it (e.g. Facebook links
>>> most of libraries statically, including libbpf).
>>>
>>> libbpf already uses the following prefixes for its interfaces:
>>> * bpf_ for bpf system call wrappers, program/map/elf-object
>>>    abstractions and a few other things;
>>> * btf_ for BTF related API;
>>> * libbpf_ for everything else.
>>>
>>> The patch adds libbpf_ prefix to interfaces that use none of mentioned
>>> above prefixes and don't fit well into the first two categories.
>>>
>>> Long term benefits of having common prefix should outweigh possible
>>> inconvenience of changing API for those functions now.
>>>
>>> Patches 2-4 add libbpf_ prefix to libbpf interfaces: separate patch per
>>> header. Other patches are simple improvements in API.
>>>
>>>
>>> Andrey Ignatov (6):
>>>    libbpf: Move __dump_nlmsg_t from API to implementation
>>>    libbpf: Consistent prefixes for interfaces in libbpf.h.
>>>    libbpf: Consistent prefixes for interfaces in nlattr.h.
>>>    libbpf: Consistent prefixes for interfaces in str_error.h.
>>>    libbpf: Make include guards consistent
>>>    libbpf: Use __u32 instead of u32 in bpf_program__load
>>>
>>>   tools/bpf/bpftool/net.c            | 41 ++++++++++---------
>>>   tools/bpf/bpftool/netlink_dumper.c | 32 ++++++++-------
>>>   tools/lib/bpf/bpf.h                |  6 +--
>>>   tools/lib/bpf/btf.h                |  6 +--
>>>   tools/lib/bpf/libbpf.c             | 22 +++++-----
>>>   tools/lib/bpf/libbpf.h             | 31 +++++++-------
>>>   tools/lib/bpf/netlink.c            | 48 ++++++++++++----------
>>>   tools/lib/bpf/nlattr.c             | 64 +++++++++++++++--------------
>>>   tools/lib/bpf/nlattr.h             | 65 +++++++++++++++---------------
>>>   tools/lib/bpf/str_error.c          |  2 +-
>>>   tools/lib/bpf/str_error.h          |  8 ++--
>>>   11 files changed, 171 insertions(+), 154 deletions(-)
>>
>> Overall agree that this is needed, and I've therefore applied the
>> set, thanks for cleaning up, Andrey!
>>
>> But, I would actually like to see this going one step further, in
>> particular, we should keep all the netlink related stuff inside
>> libbpf-/only/. Meaning, the goal of libbpf is not to provide yet
>> another set of netlink primitives but instead e.g. for the bpftool
>> dumper this should be abstracted away such that we pass in a callback
>> from bpftool side and have an iterator object which will then be
>> populated from inside the libbpf logic, meaning, bpftool shouldn't
>> even be aware of anything netlink there.
> 
> Agreed. This indeed make sense, the user really only cares a few fields
> like devname/id, attachment_type, prog_id, etc. I will take a look at
> this later if nobody works on it.

Awesome, that would be great, thanks!