| Message ID | 10735490.HC94hyceBX@wuerfel |
|---|---|
| State | Not Applicable, archived |
| Delegated to: | David Miller |
| Headers | show |
From: Arnd Bergmann <arnd@arndb.de> Date: Tue, 06 Oct 2015 22:52:46 +0200 > Commit 124613012db1 ("af_unix: Convert the unix_sk macro to an inline > function for type safety") was recently added to catch incorrect > uses of the unix_sk helper using compiler warnings. > > It has now caught one such case in lsm_audit.c. The code is technically > correct, but as it converts a const pointer to a non-const pointer, > the annotation got lost, which gcc now warns about. > > This patch avoids the warning by introducing an additional helper > that has const input and output, which makes the lsm_audit code build > cleanly again. > > Signed-off-by: Arnd Bergmann <arnd@arndb.de> > --- > I'm not entirely happy with this workaround myself, but could not come > up with a better one. You can make the argument unconditionally const, as Paul Moore has done in a separate patch submission. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thursday 08 October 2015 04:03:32 David Miller wrote: > From: Arnd Bergmann <arnd@arndb.de> > Date: Tue, 06 Oct 2015 22:52:46 +0200 > > > Commit 124613012db1 ("af_unix: Convert the unix_sk macro to an inline > > function for type safety") was recently added to catch incorrect > > uses of the unix_sk helper using compiler warnings. > > > > It has now caught one such case in lsm_audit.c. The code is technically > > correct, but as it converts a const pointer to a non-const pointer, > > the annotation got lost, which gcc now warns about. > > > > This patch avoids the warning by introducing an additional helper > > that has const input and output, which makes the lsm_audit code build > > cleanly again. > > > > Signed-off-by: Arnd Bergmann <arnd@arndb.de> > > --- > > I'm not entirely happy with this workaround myself, but could not come > > up with a better one. > > You can make the argument unconditionally const, as Paul Moore has done > in a separate patch submission. Ok, I see now how Paul's "audit: constify parts of common_audit_data and lsm_network_audit" patch caused the problem and is now gone from linux-next. That seems nicer indeed. Arnd -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/include/net/af_unix.h b/include/net/af_unix.h index cb1b9bbda332..1871b6436ee9 100644 --- a/include/net/af_unix.h +++ b/include/net/af_unix.h @@ -69,6 +69,11 @@ static inline struct unix_sock *unix_sk(struct sock *sk) return (struct unix_sock *)sk; } +static inline const struct unix_sock *unix_sk_const(const struct sock *sk) +{ + return (const struct unix_sock *)sk; +} + #define peer_wait peer_wq.wait long unix_inq_len(struct sock *sk); diff --git a/security/lsm_audit.c b/security/lsm_audit.c index 2deace208db2..cb07f1318a27 100644 --- a/security/lsm_audit.c +++ b/security/lsm_audit.c @@ -307,7 +307,7 @@ static void dump_common_audit_data(struct audit_buffer *ab, case LSM_AUDIT_DATA_NET: if (a->u.net->sk) { const struct sock *sk = a->u.net->sk; - struct unix_sock *u; + const struct unix_sock *u; int len = 0; char *p = NULL; @@ -337,7 +337,7 @@ static void dump_common_audit_data(struct audit_buffer *ab, } #endif case AF_UNIX: - u = unix_sk(sk); + u = unix_sk_const(sk); if (u->path.dentry) { audit_log_d_path(ab, " path=", &u->path); break;
Commit 124613012db1 ("af_unix: Convert the unix_sk macro to an inline function for type safety") was recently added to catch incorrect uses of the unix_sk helper using compiler warnings. It has now caught one such case in lsm_audit.c. The code is technically correct, but as it converts a const pointer to a non-const pointer, the annotation got lost, which gcc now warns about. This patch avoids the warning by introducing an additional helper that has const input and output, which makes the lsm_audit code build cleanly again. Signed-off-by: Arnd Bergmann <arnd@arndb.de> --- I'm not entirely happy with this workaround myself, but could not come up with a better one. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html