Skip to contentSkip to conversation

cifs: fix crash in smb2_compound_op()/smb2_set_next_command()

PATCH In review Ronnie Sahlberg 1 participant 2 messages Jul 16, 2019 – Jul 17, 2019 1 message not in the archive

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 cifs: fix crash in smb2_compound_op()/smb2_set_next_command()

RHBZ: 1722704

In low memory situations the various SMB2_*_init() functions can fail to allocate a request PDU and thus leave the request iovector as NULL.

If we don't check the return code for failure we end up calling smb2_set_next_command() with a NULL iovector causing a crash when it tries to dereference it.

---
 fs/cifs/smb2inode.c | 12 ++++++++++++
 fs/cifs/smb2ops.c   | 11 ++++++++++-
 2 files changed, 22 insertions(+), 1 deletion(-)
fs/cifs/smb2inode.c +12 -0
index 278405d26c47..d8d9cdfa30b6 100644
@@ -120,6 +120,8 @@ smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon,
 				SMB2_O_INFO_FILE, 0,
 				sizeof(struct smb2_file_all_info) +
 					  PATH_MAX * 2, 0, NULL);
+		if (rc)
+			goto finished;
 		smb2_set_next_command(tcon, &rqst[num_rqst]);
 		smb2_set_related(&rqst[num_rqst++]);
 		trace_smb3_query_info_compound_enter(xid, ses->Suid, tcon->tid,
@@ -147,6 +149,8 @@ smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon,
 					COMPOUND_FID, current->tgid,
 					FILE_DISPOSITION_INFORMATION,
 					SMB2_O_INFO_FILE, 0, data, size);
+		if (rc)
+			goto finished;
 		smb2_set_next_command(tcon, &rqst[num_rqst]);
 		smb2_set_related(&rqst[num_rqst++]);
 		trace_smb3_rmdir_enter(xid, ses->Suid, tcon->tid, full_path);
@@ -163,6 +167,8 @@ smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon,
 					COMPOUND_FID, current->tgid,
 					FILE_END_OF_FILE_INFORMATION,
 					SMB2_O_INFO_FILE, 0, data, size);
+		if (rc)
+			goto finished;
 		smb2_set_next_command(tcon, &rqst[num_rqst]);
 		smb2_set_related(&rqst[num_rqst++]);
 		trace_smb3_set_eof_enter(xid, ses->Suid, tcon->tid, full_path);
@@ -180,6 +186,8 @@ smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon,
 					COMPOUND_FID, current->tgid,
 					FILE_BASIC_INFORMATION,
 					SMB2_O_INFO_FILE, 0, data, size);
+		if (rc)
+			goto finished;
 		smb2_set_next_command(tcon, &rqst[num_rqst]);
 		smb2_set_related(&rqst[num_rqst++]);
 		trace_smb3_set_info_compound_enter(xid, ses->Suid, tcon->tid,
@@ -206,6 +214,8 @@ smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon,
 					COMPOUND_FID, current->tgid,
 					FILE_RENAME_INFORMATION,
 					SMB2_O_INFO_FILE, 0, data, size);
+		if (rc)
+			goto finished;
 		smb2_set_next_command(tcon, &rqst[num_rqst]);
 		smb2_set_related(&rqst[num_rqst++]);
 		trace_smb3_rename_enter(xid, ses->Suid, tcon->tid, full_path);
@@ -231,6 +241,8 @@ smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon,
 					COMPOUND_FID, current->tgid,
 					FILE_LINK_INFORMATION,
 					SMB2_O_INFO_FILE, 0, data, size);
+		if (rc)
+			goto finished;
 		smb2_set_next_command(tcon, &rqst[num_rqst]);
 		smb2_set_related(&rqst[num_rqst++]);
 		trace_smb3_hardlink_enter(xid, ses->Suid, tcon->tid, full_path);
fs/cifs/smb2ops.c +10 -1
index 4b0b14946343..462890f4cb9c 100644
@@ -2027,6 +2027,10 @@ smb2_set_related(struct smb_rqst *rqst)
 	struct smb2_sync_hdr *shdr;
 
 	shdr = (struct smb2_sync_hdr *)(rqst->rq_iov[0].iov_base);
+	if (shdr == NULL) {
+		cifs_dbg(FYI, "shdr NULL in smb2_set_related\n");
+		return;
+	}
 	shdr->Flags |= SMB2_FLAGS_RELATED_OPERATIONS;
 }
 
@@ -2041,6 +2045,12 @@ smb2_set_next_command(struct cifs_tcon *tcon, struct smb_rqst *rqst)
 	unsigned long len = smb_rqst_len(server, rqst);
 	int i, num_padding;
 
+	shdr = (struct smb2_sync_hdr *)(rqst->rq_iov[0].iov_base);
+	if (shdr == NULL) {
+		cifs_dbg(FYI, "shdr NULL in smb2_set_next_command\n");
+		return;
+	}
+
 	/* SMB headers in a compound are 8 byte aligned. */
 
 	/* No padding needed */
@@ -2080,7 +2090,6 @@ smb2_set_next_command(struct cifs_tcon *tcon, struct smb_rqst *rqst)
 	}
 
  finished:
-	shdr = (struct smb2_sync_hdr *)(rqst->rq_iov[0].iov_base);
 	shdr->NextCommand = cpu_to_le32(len);
 }
 
Signature · 1 line
2.13.6

A message that is not in the archive (sent off-list or not yet received): [email protected]

#2 lore raw

----- Original Message -----

Quoted 84 lines

From: "Sasha Levin" <[email protected]>
To: "Sasha Levin" <[email protected]>, "Ronnie Sahlberg" <[email protected]>, "linux-cifs" <[email protected]>
Cc: "Steve French" <[email protected]>, "Stable" <[email protected]>, [email protected]
Sent: Tuesday, 16 July, 2019 11:27:10 AM
Subject: Re: [PATCH] cifs: fix crash in smb2_compound_op()/smb2_set_next_command()

Hi,

[This is an automated email]

This commit has been processed because it contains a -stable tag. The stable tag indicates that it's relevant for the following trees: all

The bot has tested the following trees: v5.2.1, v5.1.18, v4.19.59, v4.14.133, v4.9.185, v4.4.185.

v5.2.1: Build OK!
v5.1.18: Build OK!
v4.19.59: Failed to apply! Possible dependencies:
271b9c0c8007 ("smb3: Fix rmdir compounding regression to strict servers") c2e0fe3f5aae ("cifs: make rmdir() use compounding") c5a5f38f075c ("cifs: add a smb2_compound_op and change QUERY_INFO to use it")
dcbf91035709 ("cifs: change SMB2_OP_SET_INFO to use compounding") e77fe73c7e38 ("cifs: we can not use small padding iovs together with encryption")
f5b05d622a3e ("cifs: add IOCTL for QUERY_INFO passthrough to userspace") f733e3936da4 ("cifs: change mkdir to use a compound") f7bfe04bf0db ("cifs: change SMB2_OP_SET_EOF to use compounding")

v4.14.133: Failed to apply! Possible dependencies:
2e96467d9eb1 ("cifs: add pdu_size to the TCP_Server_Info structure") 3d4ef9a15343 ("smb3: fix redundant opens on root") 730928c8f4be ("cifs: update smb2_queryfs() to use compounding") 74dcf418fe34 ("CIFS: SMBD: Read correct returned data length for RDMA write (SMB read) I/O")
8ce79ec359ad ("cifs: update multiplex loop to handle compounded responses")
91cb74f5142c ("cifs: Change SMB2_open to return an iov for the error parameter")
93012bf98416 ("cifs: add server->vals->header_preamble_size") 9d874c36552a ("cifs: fix a buffer leak in smb2_query_symlink") c5a5f38f075c ("cifs: add a smb2_compound_op and change QUERY_INFO to use it")
f5b05d622a3e ("cifs: add IOCTL for QUERY_INFO passthrough to userspace")

v4.9.185: Failed to apply! Possible dependencies:
31473fc4f965 ("CIFS: Separate SMB2 header structure") 7fb8986e7449 ("CIFS: Add capability to transform requests before sending")
8ce79ec359ad ("cifs: update multiplex loop to handle compounded responses")
9bb17e0916a0 ("CIFS: Add transform header handling callbacks") b8f57ee8aad4 ("CIFS: Separate RFC1001 length processing for SMB2 read") da502f7df03d ("CIFS: Make SendReceive2() takes resp iov") ef65aaede23f ("smb2: Enforce sec= mount option") f5b05d622a3e ("cifs: add IOCTL for QUERY_INFO passthrough to userspace")

v4.4.185: Failed to apply! Possible dependencies:
141891f4727c ("SMB3: Add mount parameter to allow user to override max credits")
166cea4dc3a4 ("SMB2: Separate RawNTLMSSP authentication from SMB2_sess_setup")
16c568efff82 ("cifs: merge the hash calculation helpers") 275516cdcfa4 ("Print IP address of unresponsive server") 31473fc4f965 ("CIFS: Separate SMB2 header structure") 373512ec5c10 ("Prepare for encryption support (first part). Add decryption and encryption key generation. Thanks to Metze for helping with this.")
3baf1a7b9215 ("SMB2: Separate Kerberos authentication from SMB2_sess_setup")
7fb8986e7449 ("CIFS: Add capability to transform requests before sending")
834170c85978 ("Enable previous version support") 8ce79ec359ad ("cifs: update multiplex loop to handle compounded responses")
9bb17e0916a0 ("CIFS: Add transform header handling callbacks") adfeb3e00e8e ("cifs: Make echo interval tunable") da502f7df03d ("CIFS: Make SendReceive2() takes resp iov") ef65aaede23f ("smb2: Enforce sec= mount option") f5b05d622a3e ("cifs: add IOCTL for QUERY_INFO passthrough to userspace")

Quoted 3 lines

NOTE: The patch will not be queued to stable trees until it is upstream.

How should we proceed with this patch?

So it applies cleanly in v5.2.1 and v5.1.18.
I think it is sufficient to get into those two versions then. It is very hard to trigger this issue.

Quoted 4 lines

--
Thanks,
Sasha