From: Alexander Bluhm <alexander.bluhm@gmx.net>
Subject: Re: Push solock() down to sobind()
To: Vitaliy Makkoveev <mvs@openbsd.org>
Cc: tech@openbsd.org
Date: Tue, 9 Jan 2024 15:39:33 +0100

On Tue, Jan 09, 2024 at 04:17:27PM +0300, Vitaliy Makkoveev wrote:
> All existing sobind() calls are only wrapped by solock(). So push it
> down and convert to solock_shared() to make identical to sys_bind().
> solock_shared() makes senses because the most of these sockets are
> UDP sockets.

Changing to solock_shared() for NFS is risky.

I would prefer to do the solock_shared() conversion step by step
with proper review of the sub system.  Then pushing the lock down
to sobind() will be mechanical.

bluhm

> Index: sys/kern/uipc_socket.c
> ===================================================================
> RCS file: /cvs/src/sys/kern/uipc_socket.c,v
> retrieving revision 1.312
> diff -u -p -r1.312 uipc_socket.c
> --- sys/kern/uipc_socket.c	19 Dec 2023 21:34:22 -0000	1.312
> +++ sys/kern/uipc_socket.c	9 Jan 2024 13:08:15 -0000
> @@ -219,8 +219,13 @@ socreate(int dom, struct socket **aso, i
>  int
>  sobind(struct socket *so, struct mbuf *nam, struct proc *p)
>  {
> -	soassertlocked(so);
> -	return pru_bind(so, nam, p);
> +	int error;
> +
> +	solock_shared(so);
> +	error = pru_bind(so, nam, p);
> +	sounlock_shared(so);
> +
> +	return (error);
>  }
>  
>  int
> Index: sys/kern/uipc_syscalls.c
> ===================================================================
> RCS file: /cvs/src/sys/kern/uipc_syscalls.c,v
> retrieving revision 1.216
> diff -u -p -r1.216 uipc_syscalls.c
> --- sys/kern/uipc_syscalls.c	3 Jan 2024 11:07:04 -0000	1.216
> +++ sys/kern/uipc_syscalls.c	9 Jan 2024 13:08:15 -0000
> @@ -185,9 +185,7 @@ sys_bind(struct proc *p, void *v, regist
>  	if (KTRPOINT(p, KTR_STRUCT))
>  		ktrsockaddr(p, mtod(nam, caddr_t), SCARG(uap, namelen));
>  #endif
> -	solock_shared(so);
>  	error = sobind(so, nam, p);
> -	sounlock_shared(so);
>  	m_freem(nam);
>  out:
>  	FRELE(fp, p);
> Index: sys/net/bfd.c
> ===================================================================
> RCS file: /cvs/src/sys/net/bfd.c,v
> retrieving revision 1.80
> diff -u -p -r1.80 bfd.c
> --- sys/net/bfd.c	3 Aug 2023 09:49:08 -0000	1.80
> +++ sys/net/bfd.c	9 Jan 2024 13:08:15 -0000
> @@ -477,9 +477,7 @@ bfd_listener(struct bfd_config *bfd, uns
>  		break;
>  	}
>  
> -	solock(so);
>  	error = sobind(so, m, p);
> -	sounlock(so);
>  	if (error) {
>  		printf("%s: sobind error %d\n",
>  		    __func__, error);
> @@ -578,9 +576,7 @@ bfd_sender(struct bfd_config *bfd, unsig
>  		break;
>  	}
>  
> -	solock(so);
>  	error = sobind(so, m, p);
> -	sounlock(so);
>  	if (error) {
>  		printf("%s: sobind error %d\n",
>  		    __func__, error);
> Index: sys/net/if_pflow.c
> ===================================================================
> RCS file: /cvs/src/sys/net/if_pflow.c,v
> retrieving revision 1.109
> diff -u -p -r1.109 if_pflow.c
> --- sys/net/if_pflow.c	23 Dec 2023 10:52:54 -0000	1.109
> +++ sys/net/if_pflow.c	9 Jan 2024 13:08:15 -0000
> @@ -467,9 +467,7 @@ pflow_set(struct pflow_softc *sc, struct
>  				memcpy(sa, sc->sc_flowsrc,
>  				    sc->sc_flowsrc->sa_len);
>  
> -				solock(so);
>  				error = sobind(so, m, p);
> -				sounlock(so);
>  				m_freem(m);
>  				if (error) {
>  					soclose(so, MSG_DONTWAIT);
> Index: sys/net/if_vxlan.c
> ===================================================================
> RCS file: /cvs/src/sys/net/if_vxlan.c,v
> retrieving revision 1.99
> diff -u -p -r1.99 if_vxlan.c
> --- sys/net/if_vxlan.c	23 Dec 2023 10:52:54 -0000	1.99
> +++ sys/net/if_vxlan.c	9 Jan 2024 13:08:15 -0000
> @@ -972,9 +972,7 @@ vxlan_tep_add_addr(struct vxlan_softc *s
>  		unhandled_af(vt->vt_af);
>  	}
>  
> -	solock(so);
>  	error = sobind(so, &m, curproc);
> -	sounlock(so);
>  	if (error != 0)
>  		goto close;
>  
> Index: sys/net/if_wg.c
> ===================================================================
> RCS file: /cvs/src/sys/net/if_wg.c,v
> retrieving revision 1.35
> diff -u -p -r1.35 if_wg.c
> --- sys/net/if_wg.c	1 Jan 2024 18:47:02 -0000	1.35
> +++ sys/net/if_wg.c	9 Jan 2024 13:08:15 -0000
> @@ -734,12 +734,11 @@ wg_socket_open(struct socket **so, int a
>  	sounlock(*so);
>  
>  	if ((ret = sosetopt(*so, SOL_SOCKET, SO_RTABLE, &mrtable)) == 0) {
> -		solock(*so);
>  		if ((ret = sobind(*so, &mhostnam, curproc)) == 0) {
> +			/* No lock required. Immutable for this socket. */
>  			*port = sotoinpcb(*so)->inp_lport;
>  			*rtable = sotoinpcb(*so)->inp_rtableid;
>  		}
> -		sounlock(*so);
>  	}
>  
>  	if (ret != 0)
> Index: sys/nfs/krpc_subr.c
> ===================================================================
> RCS file: /cvs/src/sys/nfs/krpc_subr.c,v
> retrieving revision 1.38
> diff -u -p -r1.38 krpc_subr.c
> --- sys/nfs/krpc_subr.c	3 Aug 2023 09:49:09 -0000	1.38
> +++ sys/nfs/krpc_subr.c	9 Jan 2024 13:08:15 -0000
> @@ -280,9 +280,7 @@ krpc_call(struct sockaddr_in *sa, u_int 
>  	sin->sin_family = AF_INET;
>  	sin->sin_addr.s_addr = INADDR_ANY;
>  	sin->sin_port = htons(0);
> -	solock(so);
>  	error = sobind(so, m, &proc0);
> -	sounlock(so);
>  	m_freem(m);
>  	if (error) {
>  		printf("bind failed\n");
> Index: sys/nfs/nfs_socket.c
> ===================================================================
> RCS file: /cvs/src/sys/nfs/nfs_socket.c,v
> retrieving revision 1.144
> diff -u -p -r1.144 nfs_socket.c
> --- sys/nfs/nfs_socket.c	3 Aug 2023 09:49:09 -0000	1.144
> +++ sys/nfs/nfs_socket.c	9 Jan 2024 13:08:15 -0000
> @@ -281,9 +281,7 @@ nfs_connect(struct nfsmount *nmp, struct
>  		sin->sin_family = AF_INET;
>  		sin->sin_addr.s_addr = INADDR_ANY;
>  		sin->sin_port = htons(0);
> -		solock(so);
>  		error = sobind(so, nam, &proc0);
> -		sounlock(so);
>  		if (error)
>  			goto bad;
>