libbpf: keep FD_CLOEXEC flag when dup()'ing FD

Make sure to preserve and/or enforce FD_CLOEXEC flag on duped FDs.
Use dup3() with O_CLOEXEC flag for that.

Without this fix libbpf effectively clears FD_CLOEXEC flag on each of BPF
map/prog FD, which is definitely not the right or expected behavior.

Reported-by: Lennart Poettering <lennart@poettering.net>
Fixes: bc308d011ab8 ("libbpf: call dup2() syscall directly")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/r/20240529223239.504241-1-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h
index a0dcfb8..7e7e686 100644 (file)
--- a/tools/lib/bpf/libbpf_internal.h
+++ b/tools/lib/bpf/libbpf_internal.h
@@ -597,13 +597,9 @@ static inline int ensure_good_fd(int fd)
        return fd;
 }
 
-static inline int sys_dup2(int oldfd, int newfd)
+static inline int sys_dup3(int oldfd, int newfd, int flags)
 {
-#ifdef __NR_dup2
-       return syscall(__NR_dup2, oldfd, newfd);
-#else
-       return syscall(__NR_dup3, oldfd, newfd, 0);
-#endif
+       return syscall(__NR_dup3, oldfd, newfd, flags);
 }
 
 /* Point *fixed_fd* to the same file that *tmp_fd* points to.
@@ -614,7 +610,7 @@ static inline int reuse_fd(int fixed_fd, int tmp_fd)
 {
        int err;
 
-       err = sys_dup2(tmp_fd, fixed_fd);
+       err = sys_dup3(tmp_fd, fixed_fd, O_CLOEXEC);
        err = err < 0 ? -errno : 0;
        close(tmp_fd); /* clean up temporary FD */
        return err;