Re: [PATCH bpf-next v2] libbpf: fix compatibility for kernels without need_wakeup

["Jonathan Lemon" <jonathan.lemon@gmail.com>]

On 22 Oct 2019, at 1:25, Magnus Karlsson wrote:

> When the need_wakeup flag was added to AF_XDP, the format of the
> XDP_MMAP_OFFSETS getsockopt was extended. Code was added to the
> kernel to take care of compatibility issues arrising from running
> applications using any of the two formats. However, libbpf was
> not extended to take care of the case when the application/libbpf
> uses the new format but the kernel only supports the old
> format. This patch adds support in libbpf for parsing the old
> format, before the need_wakeup flag was added, and emulating a
> set of static need_wakeup flags that will always work for the
> application.
>
> v1 -> v2:
> * Rebased to bpf-next
> * Rewrote the code as the previous version made you blind
>
> Fixes: a4500432c2587cb2a ("libbpf: add support for need_wakeup flag in 
> AF_XDP part")
> Reported-by: Eloy Degen <degeneloy@gmail.com>
> Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
> ---
>  tools/lib/bpf/xsk.c | 79 
> +++++++++++++++++++++++++++++++++++++++++++++--------
>  1 file changed, 67 insertions(+), 12 deletions(-)
>
> diff --git a/tools/lib/bpf/xsk.c b/tools/lib/bpf/xsk.c
> index 7866500..aa16458 100644
> --- a/tools/lib/bpf/xsk.c
> +++ b/tools/lib/bpf/xsk.c
> @@ -73,6 +73,21 @@ struct xsk_nl_info {
>  	int fd;
>  };
>
> +/* Up until and including Linux 5.3 */
> +struct xdp_ring_offset_no_flags {
> +	__u64 producer;
> +	__u64 consumer;
> +	__u64 desc;
> +};
> +
> +/* Up until and including Linux 5.3 */
> +struct xdp_mmap_offsets_no_flags {
> +	struct xdp_ring_offset_no_flags rx;
> +	struct xdp_ring_offset_no_flags tx;
> +	struct xdp_ring_offset_no_flags fr;
> +	struct xdp_ring_offset_no_flags cr;
> +};

Why not name these with "_v1" instead of "_no_flags" in order to match
the same definitions in net/xdp/xsk.h?  I'd rather not have two 
identical
structures different names if it can be avoided.


> +
>  int xsk_umem__fd(const struct xsk_umem *umem)
>  {
>  	return umem ? umem->fd : -EINVAL;
> @@ -133,6 +148,54 @@ static int xsk_set_xdp_socket_config(struct 
> xsk_socket_config *cfg,
>  	return 0;
>  }
>
> +static bool xsk_mmap_offsets_has_flags(socklen_t optlen)
> +{
> +	return (optlen == sizeof(struct xdp_mmap_offsets)) ? true : false;
> +}
> +
> +static int xsk_get_mmap_offsets(int fd, struct xdp_mmap_offsets *off)
> +{
> +	struct xdp_mmap_offsets_no_flags off_no_flag;
> +	socklen_t optlen;
> +	int err;
> +
> +	optlen = sizeof(*off);
> +	err = getsockopt(fd, SOL_XDP, XDP_MMAP_OFFSETS, off, &optlen);
> +	if (err)
> +		return err;
> +
> +	if (xsk_mmap_offsets_has_flags(optlen))

I'd just use "if (optlen == sizeof(*off))" here.


> +		return 0;
> +
> +	/* getsockopt on a kernel <= 5.3 has no flags fields.
> +	 * Copy over the offsets to the correct places in the >=5.4 format
> +	 * and put the flags where they would have been on that kernel.
> +	 */

Would it be worthwhile adding a length check here?
Something like:

     if (optlen == sizeof(struct xdp_mmap_offsets_v1)
         return xsk_mmap_offset_v1(off);

     return -EINVAL;

Which makes it easy to revisit this later if the same
problem crops up again.

> +	memcpy(&off_no_flag, off, sizeof(off_no_flag));
> +
> +	off->rx.producer = off_no_flag.rx.producer;
> +	off->rx.consumer = off_no_flag.rx.consumer;
> +	off->rx.desc = off_no_flag.rx.desc;
> +	off->rx.flags = off_no_flag.rx.consumer + sizeof(u32);
> +
> +	off->tx.producer = off_no_flag.tx.producer;
> +	off->tx.consumer = off_no_flag.tx.consumer;
> +	off->tx.desc = off_no_flag.tx.desc;
> +	off->tx.flags = off_no_flag.tx.consumer + sizeof(u32);
> +
> +	off->fr.producer = off_no_flag.fr.producer;
> +	off->fr.consumer = off_no_flag.fr.consumer;
> +	off->fr.desc = off_no_flag.fr.desc;
> +	off->fr.flags = off_no_flag.fr.consumer + sizeof(u32);
> +
> +	off->cr.producer = off_no_flag.cr.producer;
> +	off->cr.consumer = off_no_flag.cr.consumer;
> +	off->cr.desc = off_no_flag.cr.desc;
> +	off->cr.flags = off_no_flag.cr.consumer + sizeof(u32);


Then all this moves into a xsk_mmap_offset_v1() helper.

-- 
Jonathan

> +
> +	return 0;
> +}
> +
>  int xsk_umem__create_v0_0_4(struct xsk_umem **umem_ptr, void 
> *umem_area,
>  			    __u64 size, struct xsk_ring_prod *fill,
>  			    struct xsk_ring_cons *comp,
> @@ -141,7 +204,6 @@ int xsk_umem__create_v0_0_4(struct xsk_umem 
> **umem_ptr, void *umem_area,
>  	struct xdp_mmap_offsets off;
>  	struct xdp_umem_reg mr;
>  	struct xsk_umem *umem;
> -	socklen_t optlen;
>  	void *map;
>  	int err;
>
> @@ -190,8 +252,7 @@ int xsk_umem__create_v0_0_4(struct xsk_umem 
> **umem_ptr, void *umem_area,
>  		goto out_socket;
>  	}
>
> -	optlen = sizeof(off);
> -	err = getsockopt(umem->fd, SOL_XDP, XDP_MMAP_OFFSETS, &off, 
> &optlen);
> +	err = xsk_get_mmap_offsets(umem->fd, &off);
>  	if (err) {
>  		err = -errno;
>  		goto out_socket;
> @@ -492,7 +553,6 @@ int xsk_socket__create(struct xsk_socket 
> **xsk_ptr, const char *ifname,
>  	struct sockaddr_xdp sxdp = {};
>  	struct xdp_mmap_offsets off;
>  	struct xsk_socket *xsk;
> -	socklen_t optlen;
>  	int err;
>
>  	if (!umem || !xsk_ptr || !rx || !tx)
> @@ -551,8 +611,7 @@ int xsk_socket__create(struct xsk_socket 
> **xsk_ptr, const char *ifname,
>  		}
>  	}
>
> -	optlen = sizeof(off);
> -	err = getsockopt(xsk->fd, SOL_XDP, XDP_MMAP_OFFSETS, &off, &optlen);
> +	err = xsk_get_mmap_offsets(xsk->fd, &off);
>  	if (err) {
>  		err = -errno;
>  		goto out_socket;
> @@ -638,7 +697,6 @@ int xsk_socket__create(struct xsk_socket 
> **xsk_ptr, const char *ifname,
>  int xsk_umem__delete(struct xsk_umem *umem)
>  {
>  	struct xdp_mmap_offsets off;
> -	socklen_t optlen;
>  	int err;
>
>  	if (!umem)
> @@ -647,8 +705,7 @@ int xsk_umem__delete(struct xsk_umem *umem)
>  	if (umem->refcount)
>  		return -EBUSY;
>
> -	optlen = sizeof(off);
> -	err = getsockopt(umem->fd, SOL_XDP, XDP_MMAP_OFFSETS, &off, 
> &optlen);
> +	err = xsk_get_mmap_offsets(umem->fd, &off);
>  	if (!err) {
>  		munmap(umem->fill->ring - off.fr.desc,
>  		       off.fr.desc + umem->config.fill_size * sizeof(__u64));
> @@ -666,7 +723,6 @@ void xsk_socket__delete(struct xsk_socket *xsk)
>  {
>  	size_t desc_sz = sizeof(struct xdp_desc);
>  	struct xdp_mmap_offsets off;
> -	socklen_t optlen;
>  	int err;
>
>  	if (!xsk)
> @@ -677,8 +733,7 @@ void xsk_socket__delete(struct xsk_socket *xsk)
>  		close(xsk->prog_fd);
>  	}
>
> -	optlen = sizeof(off);
> -	err = getsockopt(xsk->fd, SOL_XDP, XDP_MMAP_OFFSETS, &off, &optlen);
> +	err = xsk_get_mmap_offsets(xsk->fd, &off);
>  	if (!err) {
>  		if (xsk->rx) {
>  			munmap(xsk->rx->ring - off.rx.desc,
> -- 
> 2.7.4