Hi Greg,
this patch is applied in the wrong place (the wrong case) here in
sock_getsockopt(). The function seems to have changed in a number of
places after 5.4, apparently too much for an automatic(?!) merge.
On Sat, 2024-01-13 at 10:49 +0100, Greg Kroah-Hartman wrote:
> 5.4-stable review patch. If anyone has any objections, please let me
> know.
>
> ------------------
>
> From: Jörn-Thorben Hinz <jthinz@xxxxxxxxxxxxxxxxxxxx>
>
> [ Upstream commit 7f6ca95d16b96567ce4cf458a2790ff17fa620c3 ]
>
> Commit 9718475e6908 ("socket: Add SO_TIMESTAMPING_NEW") added the new
> socket option SO_TIMESTAMPING_NEW. Setting the option is handled in
> sk_setsockopt(), querying it was not handled in sk_getsockopt(),
> though.
>
> Following remarks on an earlier submission of this patch, keep the
> old
> behavior of getsockopt(SO_TIMESTAMPING_OLD) which returns the active
> flags even if they actually have been set through
> SO_TIMESTAMPING_NEW.
>
> The new getsockopt(SO_TIMESTAMPING_NEW) is stricter, returning flags
> only if they have been set through the same option.
>
> Fixes: 9718475e6908 ("socket: Add SO_TIMESTAMPING_NEW")
> Link:
> https://lore.kernel.org/lkml/20230703175048.151683-1-jthinz@xxxxxxxxxxxxxxxxxxxx/
> Link:
> https://lore.kernel.org/netdev/0d7cddc9-03fa-43db-a579-14f3e822615b@xxxxxxxxxxxxxxxx/
> Signed-off-by: Jörn-Thorben Hinz <jthinz@xxxxxxxxxxxxxxxxxxxx>
> Reviewed-by: Willem de Bruijn <willemb@xxxxxxxxxx>
> Signed-off-by: David S. Miller <davem@xxxxxxxxxxxxx>
> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
> ---
> net/core/sock.c | 11 +++++++++--
> 1 file changed, 9 insertions(+), 2 deletions(-)
>
> diff --git a/net/core/sock.c b/net/core/sock.c
> index 2c3c5df139345..a3ca522434a6e 100644
> --- a/net/core/sock.c
> +++ b/net/core/sock.c
> @@ -1309,9 +1309,16 @@ int sock_getsockopt(struct socket *sock, int
> level, int optname,
> break;
>
> case SO_LINGER:
> + case SO_TIMESTAMPING_NEW:
> lv = sizeof(v.ling);
> - v.ling.l_onoff = sock_flag(sk, SOCK_LINGER);
> - v.ling.l_linger = sk->sk_lingertime / HZ;
> + /* For the later-added case SO_TIMESTAMPING_NEW: Be
> strict about only
> + * returning the flags when they were set through the
> same option.
> + * Don't change the beviour for the old case
> SO_TIMESTAMPING_OLD.
> + */
> + if (optname == SO_TIMESTAMPING_OLD || sock_flag(sk,
> SOCK_TSTAMP_NEW)) {
> + v.ling.l_onoff = sock_flag(sk, SOCK_LINGER);
> + v.ling.l_linger = sk->sk_lingertime / HZ;
> + }
> break;
>
> case SO_BSDCOMPAT:
