Linux CIFS filesystem development
 help / color / mirror / Atom feed
From: Frank Sorenson <sorenson@redhat.com>
To: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org, pc@manguebit.org, stable@vger.kernel.org
Subject: Re: [PATCH v3 11/11] smb: client: fix NameOffset and Next field validation in smb2_parse_contexts()
Date: Sat, 29 Aug 2026 11:26:08 -0500	[thread overview]
Message-ID: <666879b8-5117-4c41-ab1d-e50637e68a5d@redhat.com> (raw)
In-Reply-To: <CAKYAXd8_jvTz6DX-1OjkCu1UAe7qKKHKOyW7-yOqrqVtxC6dPQ@mail.gmail.com>


On 8/29/26 9:01 AM, Namjae Jeon wrote:
>>                  case 16:
>> -                       if (posix && !memcmp(name, smb3_create_tag_posix, 16))
>> +                       if (dlen && posix && !memcmp(name, smb3_create_tag_posix, 16))
> Could we check that dlen is large enough for the fixed POSIX fields
> before calling parse_posix_ctxt()? It reads 12 bytes even when
> DataLength is smaller.

so this needs to be:
                      if (dlen > 12 && posix && !memcmp(name, smb3_create_tag_posix, 16))

I think we'll need similar before the calls to the 
server->ops->parse_lease_buf:
                     if (dlen && !strncmp(name, 
SMB2_CREATE_REQUEST_LEASE, 4)) {                                   
*oplock = server->ops->parse_lease_buf(cc, epoch,                       
                                                    lease_key);

for smb2: smb2_parse_lease_buf(void *buf, __u16 *epoch, char *lease_key) 
{         struct create_lease *lc = (struct create_lease *)buf;         
*epoch = 0; /* not used */         if (lc->lcontext.LeaseFlags & 
SMB2_LEASE_FLAG_BREAK_IN_PROGRESS_LE)                 return 
SMB2_OPLOCK_LEVEL_NOCHANGE;         return 
le32_to_cpu(lc->lcontext.LeaseState);

/* See MS-SMB2 2.2.13.2.8 */ struct lease_context {         __u8 
LeaseKey[SMB2_LEASE_KEY_SIZE];         __le32 LeaseState;         __le32 
LeaseFlags;         __le64 LeaseDuration; } __packed;

for smb2, we only access through LeaseFlags (24 bytes), but the entire 
struct is 32

static __u8 smb3_parse_lease_buf(void *buf, __u16 *epoch, char 
*lease_key) {         struct create_lease_v2 *lc = (struct 
create_lease_v2 *)buf;         *epoch = le16_to_cpu(lc->lcontext.Epoch); 
         if (lc->lcontext.LeaseFlags & 
SMB2_LEASE_FLAG_BREAK_IN_PROGRESS_LE)                 return 
SMB2_OPLOCK_LEVEL_NOCHANGE;         if (lease_key)                 
memcpy(lease_key, &lc->lcontext.LeaseKey, SMB2_LEASE_KEY_SIZE);         
return le32_to_cpu(lc->lcontext.LeaseState);

/* See MS-SMB2 2.2.13.2.10 */ struct lease_context_v2 {         __u8 
LeaseKey[SMB2_LEASE_KEY_SIZE];         __le32 LeaseState;         __le32 
LeaseFlags;         __le64 LeaseDuration;         __u8 
ParentLeaseKey[SMB2_LEASE_KEY_SIZE];         __le16 Epoch;         
__le16 Reserved; } __packed; and for smb3, we'll need through Epoch, so

     SMB2_LEASE_KEY_SIZE + 4 + 4 + 8 + SMB2_LEASE_KEY_SIZE + 2 = 50,

with the entire struct being 52 bytes

and also before the call to parse_query_id_ctxt():
                           } else if (dlen > 16 && buf &&               
                        !strncmp(name, SMB2_CREATE_QUERY_ON_DISK_ID, 4)) 
{                                   parse_query_id_ctxt(cc, buf);
parse_query_id_ctxt(struct create_context *cc, struct smb2_file_all_info *buf)
{
         struct create_disk_id_rsp *pdisk_id = (struct create_disk_id_rsp *)cc;

         cifs_dbg(FYI, "parse query id context 0x%llx 0x%llx\n",
                 pdisk_id->DiskFileId, pdisk_id->VolumeId);
         buf->IndexNumber = pdisk_id->DiskFileId;

/* See MS-SMB2 2.2.14.2.9 */
struct create_disk_id_rsp {
         struct create_context_hdr ccontext;
         __u8   Name[8];
         __le64 DiskFileId;
         __le64 VolumeId;
         __u8  Reserved[16];
} __packed;


With varying sizes, how do we handle the check prior to the call to the version-specificserver->ops->parse_lease_buf ?


-- 
Frank Sorenson
sorenson@redhat.com
Principal Software Maintenance Engineer, filesystems
Red Hat


  reply	other threads:[~2026-08-29 16:26 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-26 15:31 [PATCH v3 00/11] smb: client: fix OOB reads and UAFs in SMB2/3 receive paths Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 01/11] smb: client: fix NextCommand bounds and aliasing UAF in receive_encrypted_standard() Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 02/11] smb: client: validate PDU length before smb2_get_data_area_len() struct access Frank Sorenson
2026-08-29 10:31   ` Namjae Jeon
2026-08-30  2:43     ` Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 03/11] smb: client: fix server->total_read not tracking sub-PDU size in receive_encrypted_standard() Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 04/11] smb: client: fix missing lower-bound check on DFS referral string offsets Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 05/11] smb: client: fix missing lower-bound on Next field in parse_server_interfaces() Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 06/11] smb: client: fix OOB struct field reads in move_smb2_ea_to_cifs() Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 07/11] smb: client: fix missing iov bounds check in parse_posix_sids() Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 08/11] smb: client: fix underflow in is_valid_oplock_break() notify offset check Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 09/11] smb: client: fix potential OOB read in smb3_enum_snapshots() Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 10/11] smb: client: fix incomplete bounds check on reparse buffer in cifs_query_reparse_point() Frank Sorenson
2026-08-26 15:31 ` [PATCH v3 11/11] smb: client: fix NameOffset and Next field validation in smb2_parse_contexts() Frank Sorenson
2026-08-29 14:01   ` Namjae Jeon
2026-08-29 16:26     ` Frank Sorenson [this message]
2026-08-29 16:52       ` Frank Sorenson
2026-08-30  1:29         ` Namjae Jeon
2026-08-30  1:40 ` [PATCH v3 00/11] smb: client: fix OOB reads and UAFs in SMB2/3 receive paths Namjae Jeon

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=666879b8-5117-4c41-ab1d-e50637e68a5d@redhat.com \
    --to=sorenson@redhat.com \
    --cc=linkinjeon@kernel.org \
    --cc=linux-cifs@vger.kernel.org \
    --cc=pc@manguebit.org \
    --cc=stable@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox