All of lore.kernel.org
 help / color / mirror / Atom feed
From: fanchaoting <fanchaoting@cn.fujitsu.com>
To: "linux-nfs@vger.kernel.org" <linux-nfs@vger.kernel.org>
Subject: [PATCH] nfs4-acl-tools: when who's length is very big, nfs4_getacl core dump
Date: Wed, 26 Dec 2012 17:31:37 +0800	[thread overview]
Message-ID: <50DAC3F9.6070801@cn.fujitsu.com> (raw)

From: Fan Chaoting <fanchaoting@cn.fujitsu.com>

nfsv4 server can return an arbitrary
sized who's len(eg. wholen = 62343534343) in an FATTR4_WORD0_ACL request.
It can cause program core dump when call 'malloc((wholen + 1) * sizeof(char))'.

 This patch checked if who's len bigger than xattr_size when getfacl.

 This patch also fixed some code style.

Signed-off-by: Fan Chaoting <fanchaoting@cn.fujitsu.com>

the reproduce of this problem:
#####################################################################

1. change the nfs server's code.

diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index 731587c..5b12a6f 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -2163,6 +2163,13 @@ out_acl:
        }

        *attrlenp = htonl((char *)p - (char *)attrlenp - 4);
+
+       static int j;
+       j++;
+
+       if (j % 10 == 0 & bmval0 & FATTR4_WORD0_ACL) {
+               *attrlenp = 6000;
+       }
        *countp = p - buffer;
        status = nfs_ok;

@@ -3323,7 +3330,7 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo
        else
                iov = &rqstp->rq_res.head[0];
        iov->iov_len = ((char*)resp->p) - (char*)iov->iov_base;
-       BUG_ON(iov->iov_len > PAGE_SIZE);
+//     BUG_ON(iov->iov_len > PAGE_SIZE);
        if (nfsd4_has_session(cs)) {
                if (cs->status != nfserr_replay_cache) {
                        nfsd4_store_cache_entry(resp);

2. mount -t nfs4 -o acl NFSSERVERIP:/ /mnt
3. touch /mnt/test
4. nfs4_getfacl  /mnt/test               <--core dump

######################################################################

---
 libnfs4acl/acl_nfs4_xattr_load.c |    6 ++++++
 libnfs4acl/nfs4_acl_for_path.c   |    8 ++++----
 2 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/libnfs4acl/acl_nfs4_xattr_load.c b/libnfs4acl/acl_nfs4_xattr_load.c
index 089a139..ced1c95 100644
--- a/libnfs4acl/acl_nfs4_xattr_load.c
+++ b/libnfs4acl/acl_nfs4_xattr_load.c
@@ -139,6 +139,12 @@ struct nfs4_acl * acl_nfs4_xattr_load(char *xattr_v, int xattr_size, u32 is_dir)
 			goto err1;
 		}
 
+		/*wholen should less than xattr_size*/
+		if (wholen > xattr_size) {
+			errno = EINVAL;
+			goto err1;
+		}
+
 		who = (char *) malloc((wholen+1) * sizeof(char));
 		if (who == NULL) {
 			errno = ENOMEM;
diff --git a/libnfs4acl/nfs4_acl_for_path.c b/libnfs4acl/nfs4_acl_for_path.c
index 7461005..577dd1f 100644
--- a/libnfs4acl/nfs4_acl_for_path.c
+++ b/libnfs4acl/nfs4_acl_for_path.c
@@ -92,14 +92,14 @@ static int nfs4_getxattr(const char *path, void *value, size_t size)
 
 	res = getxattr(path, ACL_NFS4_XATTR, value, size);
 	if (res < -10000) {
-		fprintf(stderr,"An internal NFS server error code (%d) was returned; this should never happen.\n",res);
+		fprintf(stderr, "An internal NFS server error code (%d) was returned; this should never happen.\n", res);
 	} else if (res < 0) {
 		if (errno == ENOATTR)
-			fprintf(stderr,"Attribute not found on file.\n");
+			fprintf(stderr, "Attribute not found on file.\n");
 		else if (errno == EREMOTEIO)
-		    fprintf(stderr,"An NFS server error occurred.\n");
+		    fprintf(stderr, "An NFS server error occurred.\n");
 		else if (errno == EOPNOTSUPP)
-			fprintf(stderr,"Operation to request attribute not supported.\n");
+			fprintf(stderr, "Operation to request attribute not supported.\n");
 		else
 			perror("Failed getxattr operation");
 	}
-- 1.7.1 


             reply	other threads:[~2012-12-26  9:59 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-12-26  9:31 fanchaoting [this message]
  -- strict thread matches above, loose matches on Subject: below --
2013-01-11  9:52 [PATCH] nfs4-acl-tools: when who's length is very big, nfs4_getacl core dump fanchaoting
2012-12-26  9:23 fanchaoting

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=50DAC3F9.6070801@cn.fujitsu.com \
    --to=fanchaoting@cn.fujitsu.com \
    --cc=linux-nfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.