nfsd: fix undefined behavior in nfsd4_layout_verify
authorAri Kauppi <ari@synopsys.com>
Fri, 5 May 2017 20:07:55 +0000 (16:07 -0400)
committerJ. Bruce Fields <bfields@redhat.com>
Tue, 9 May 2017 21:09:18 +0000 (17:09 -0400)
  UBSAN: Undefined behaviour in fs/nfsd/nfs4proc.c:1262:34
  shift exponent 128 is too large for 32-bit type 'int'

Depending on compiler+architecture, this may cause the check for
layout_type to succeed for overly large values (which seems to be the
case with amd64). The large value will be later used in de-referencing
nfsd4_layout_ops for function pointers.

Reported-by: Jani Tuovila <tuovila@synopsys.com>
Signed-off-by: Ari Kauppi <ari@synopsys.com>
[colin.king@canonical.com: use LAYOUT_TYPE_MAX instead of 32]
Cc: stable@vger.kernel.org
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
fs/nfsd/nfs4proc.c

index d86031b6ad79301c8ca0ceec9c8b991dd5db70f1..c453a1998e003d3e900407b266f1a15de5d5d94b 100644 (file)
@@ -1259,7 +1259,8 @@ nfsd4_layout_verify(struct svc_export *exp, unsigned int layout_type)
                return NULL;
        }
 
-       if (!(exp->ex_layout_types & (1 << layout_type))) {
+       if (layout_type >= LAYOUT_TYPE_MAX ||
+           !(exp->ex_layout_types & (1 << layout_type))) {
                dprintk("%s: layout type %d not supported\n",
                        __func__, layout_type);
                return NULL;