On Thu, Mar 21, 2024 at 9:02 AM Arnaldo Carvalho de Melo
<acme@xxxxxxxxxx> wrote:
>
> Since its going directly to the syscall to avoid not having
> memfd_create() available in some systems, do the same for its
> MFD_CLOEXEC flags, defining it if not available.
>
> This fixes the build in those systems, noticed while building perf on a
> set of build containers.
>
> Fixes: 9fa5e1a180aa639f ("libbpf: Call memfd_create() syscall directly")
> Cc: Andrii Nakryiko <andrii@xxxxxxxxxx>
> Cc: Daniel Borkmann <daniel@xxxxxxxxxxxxx>
> Signed-off-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
> ---
> tools/lib/bpf/libbpf.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
> index efab29b8935bd9f7..635a38c6f947a34d 100644
> --- a/tools/lib/bpf/libbpf.c
> +++ b/tools/lib/bpf/libbpf.c
> @@ -1650,6 +1650,10 @@ static int sys_memfd_create(const char *name, unsigned flags)
> return syscall(__NR_memfd_create, name, flags);
> }
>
> +#ifndef MFD_CLOEXEC
> +#define MFD_CLOEXEC 0x0001U
> +#endif
> +
makes sense, and lgtm. Do you need it in the bpf or bpf-next tree the
most? your patch header doesn't specify this
> static int create_placeholder_fd(void)
> {
> int fd;
> --
> 2.44.0
>
