On Mar 24, 2009, at 4:32 PM, J. Bruce Fields wrote:
On Wed, Mar 18, 2009 at 08:45:36PM -0400, Chuck Lever wrote:
The svc_addr_len() helper function can return a negative errno value,
but its return type is size_t, which is unsigned.
The RDMA transport code passes this return value directly to
memset(),
without checking first if it's negative. This could cause memset()
to
clobber a large piece of memory if svc_addr_len() has returned an
error.
I'd still like to understand when this can happen, to better
understand
how the error should be handled.
Also:
Signed-off-by: Chuck Lever <chuck.lever@xxxxxxxxxx>
---
include/linux/sunrpc/svc_xprt.h | 3 ++-
net/sunrpc/xprtrdma/svc_rdma_transport.c | 16 ++++++++++++++--
2 files changed, 16 insertions(+), 3 deletions(-)
diff --git a/include/linux/sunrpc/svc_xprt.h b/include/linux/sunrpc/
svc_xprt.h
index 0127dac..c2aa8cd 100644
--- a/include/linux/sunrpc/svc_xprt.h
+++ b/include/linux/sunrpc/svc_xprt.h
@@ -113,7 +113,7 @@ static inline unsigned short
svc_addr_port(struct sockaddr *sa)
return ret;
}
-static inline size_t svc_addr_len(struct sockaddr *sa)
+static inline int svc_addr_len(const struct sockaddr *sa)
{
switch (sa->sa_family) {
case AF_INET:
@@ -121,6 +121,7 @@ static inline size_t svc_addr_len(struct
sockaddr *sa)
case AF_INET6:
return sizeof(struct sockaddr_in6);
}
+
return -EAFNOSUPPORT;
}
It's clearly a bug to return -EAFNOSUPPORT in a size_t.
diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/
xprtrdma/svc_rdma_transport.c
index 3d810e7..d1ec6f9 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c
@@ -546,6 +546,7 @@ static void handle_connect_req(struct
rdma_cm_id *new_cma_id, size_t client_ird)
struct svcxprt_rdma *listen_xprt = new_cma_id->context;
struct svcxprt_rdma *newxprt;
struct sockaddr *sa;
+ int len;
/* Create a new transport */
newxprt = rdma_create_xprt(listen_xprt->sc_xprt.xpt_server, 0);
@@ -563,9 +564,20 @@ static void handle_connect_req(struct
rdma_cm_id *new_cma_id, size_t client_ird)
/* Set the local and remote addresses in the transport */
sa = (struct sockaddr *)&newxprt->sc_cm_id->route.addr.dst_addr;
- svc_xprt_set_remote(&newxprt->sc_xprt, sa, svc_addr_len(sa));
+ len = svc_addr_len(sa);
+ if (len < 0) {
+ dprintk("svcrdma: dst_addr has a bad address family\n");
+ return;
we're probably leaking something here.
I don't want to fix this until it's understood well enough to fix it
correctly.
Tom needs respond to this, but I think it would be safe to simply
BUG() here.
--b.
+ }
+ svc_xprt_set_remote(&newxprt->sc_xprt, sa, len);
+
sa = (struct sockaddr *)&newxprt->sc_cm_id->route.addr.src_addr;
- svc_xprt_set_local(&newxprt->sc_xprt, sa, svc_addr_len(sa));
+ len = svc_addr_len(sa);
+ if (len < 0) {
+ dprintk("svcrdma: src_addr has a bad address family\n");
+ return;
+ }
+ svc_xprt_set_local(&newxprt->sc_xprt, sa, len);
/*
* Enqueue the new transport on the accept queue of the listening
--
Chuck Lever
chuck[dot]lever[at]oracle[dot]com
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
