Re: [PATCH net-next 4/7] ipv4: remove get_rttos

[Willem de Bruijn <willemdebruijn.kernel@gmail.com>]

Willem de Bruijn wrote:
> Willem de Bruijn wrote:
> > From: Willem de Bruijn <willemb@google.com>
> > 
> > Initialize the ip cookie tos field when initializing the cookie, in
> > ipcm_init_sk.
> > 
> > The existing code inverts the standard pattern for initializing cookie
> > fields. Default is to initialize the field from the sk, then possibly
> > overwrite that when parsing cmsgs (the unlikely case).
> > 
> > This field inverts that, setting the field to an illegal value and
> > after cmsg parsing checking whether the value is still illegal and
> > thus should be overridden.
> > 
> > Be careful to always apply mask INET_DSCP_MASK, as before.
> > 
> > Signed-off-by: Willem de Bruijn <willemb@google.com>
> > ---
> >  include/net/ip.h       | 11 +++--------
> >  net/ipv4/ip_sockglue.c |  4 ++--
> >  net/ipv4/ping.c        |  1 -
> >  net/ipv4/raw.c         |  1 -
> >  net/ipv4/udp.c         |  1 -
> >  5 files changed, 5 insertions(+), 13 deletions(-)
> > 
> > diff --git a/include/net/ip.h b/include/net/ip.h
> > index 6af16545b3e3..6819704e2642 100644
> > --- a/include/net/ip.h
> > +++ b/include/net/ip.h
> > @@ -92,7 +92,9 @@ static inline void ipcm_init(struct ipcm_cookie *ipcm)
> >  static inline void ipcm_init_sk(struct ipcm_cookie *ipcm,
> >  				const struct inet_sock *inet)
> >  {
> > -	ipcm_init(ipcm);
> > +	*ipcm = (struct ipcm_cookie) {
> > +		.tos = READ_ONCE(inet->tos) & INET_DSCP_MASK,
> > +	};
> >  
> >  	sockcm_init(&ipcm->sockc, &inet->sk);
> >  
> > @@ -256,13 +258,6 @@ static inline u8 ip_sendmsg_scope(const struct inet_sock *inet,
> >  	return RT_SCOPE_UNIVERSE;
> >  }
> >  
> > -static inline __u8 get_rttos(struct ipcm_cookie* ipc, struct inet_sock *inet)
> > -{
> > -	u8 dsfield = ipc->tos != -1 ? ipc->tos : READ_ONCE(inet->tos);
> > -
> > -	return dsfield & INET_DSCP_MASK;
> > -}
> > -
> >  /* datagram.c */
> >  int __ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len);
> >  int ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len);
> > diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c
> > index 6d9c5c20b1c4..98b1e4a8b72e 100644
> > --- a/net/ipv4/ip_sockglue.c
> > +++ b/net/ipv4/ip_sockglue.c
> > @@ -314,8 +314,8 @@ int ip_cmsg_send(struct sock *sk, struct msghdr *msg, struct ipcm_cookie *ipc,
> >  				return -EINVAL;
> >  			if (val < 0 || val > 255)
> >  				return -EINVAL;
> > -			ipc->tos = val;
> > -			ipc->sockc.priority = rt_tos2priority(ipc->tos);
> > +			ipc->sockc.priority = rt_tos2priority(val);
> > +			ipc->tos = val & INET_DSCP_MASK;
> >  			break;
> >  		case IP_PROTOCOL:
> >  			if (cmsg->cmsg_len != CMSG_LEN(sizeof(int)))
> > diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c
> > index 619ddc087957..0215885c6df5 100644
> > --- a/net/ipv4/ping.c
> > +++ b/net/ipv4/ping.c
> > @@ -768,7 +768,6 @@ static int ping_v4_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
> >  		}
> >  		faddr = ipc.opt->opt.faddr;
> >  	}
> > -	tos = get_rttos(&ipc, inet);
> 
> Here and elsewhere, subsequent code needs to use ipc.tos directly.

Actually I misunderstood the purpose of get_rttos.

It only masks the dsfield when passed to the routing layer, with
flowi4_init_output().

The other purpose of ipc->tos, to initialize iph->tos, takes the
unmasked version including ECN bits.