| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <84cfe4d93cb5b02591f4bd921b828eb6f3e95faa.camel@linux.ibm.com>
Date: Thu, 12 Oct 2023 09:25:00 -0400
From: Mimi Zohar <zohar@...ux.ibm.com>
To: Roberto Sassu <roberto.sassu@...weicloud.com>,
viro@...iv.linux.org.uk, brauner@...nel.org,
chuck.lever@...cle.com, jlayton@...nel.org, neilb@...e.de,
kolga@...app.com, Dai.Ngo@...cle.com, tom@...pey.com,
dmitry.kasatkin@...il.com, paul@...l-moore.com, jmorris@...ei.org,
serge@...lyn.com, dhowells@...hat.com, jarkko@...nel.org,
stephen.smalley.work@...il.com, eparis@...isplace.org,
casey@...aufler-ca.com
Cc: linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
linux-nfs@...r.kernel.org, linux-integrity@...r.kernel.org,
linux-security-module@...r.kernel.org, keyrings@...r.kernel.org,
selinux@...r.kernel.org, Roberto Sassu <roberto.sassu@...wei.com>
Subject: Re: [PATCH v3 02/25] ima: Align ima_post_path_mknod() definition
with LSM infrastructure
On Thu, 2023-10-12 at 14:19 +0200, Roberto Sassu wrote:
> On Thu, 2023-10-12 at 07:42 -0400, Mimi Zohar wrote:
> > On Thu, 2023-10-12 at 09:29 +0200, Roberto Sassu wrote:
> > > On Wed, 2023-10-11 at 15:01 -0400, Mimi Zohar wrote:
> > > > On Wed, 2023-10-11 at 18:02 +0200, Roberto Sassu wrote:
> > > > > On Wed, 2023-10-11 at 10:38 -0400, Mimi Zohar wrote:
> > > > > > On Mon, 2023-09-04 at 15:33 +0200, Roberto Sassu wrote:
> > > > > > > From: Roberto Sassu <roberto.sassu@...wei.com>
> > > > > > >
> > > > > > > Change ima_post_path_mknod() definition, so that it can be registered as
> > > > > > > implementation of the path_post_mknod hook. Since LSMs see a umask-stripped
> > > > > > > mode from security_path_mknod(), pass the same to ima_post_path_mknod() as
> > > > > > > well.
> > > > > > > Also, make sure that ima_post_path_mknod() is executed only if
> > > > > > > (mode & S_IFMT) is equal to zero or S_IFREG.
> > > > > > >
> > > > > > > Add this check to take into account the different placement of the
> > > > > > > path_post_mknod hook (to be introduced) in do_mknodat().
> > > > > >
> > > > > > Move "(to be introduced)" to when it is first mentioned.
> > > > > >
> > > > > > > Since the new hook
> > > > > > > will be placed after the switch(), the check ensures that
> > > > > > > ima_post_path_mknod() is invoked as originally intended when it is
> > > > > > > registered as implementation of path_post_mknod.
> > > > > > >
> > > > > > > Signed-off-by: Roberto Sassu <roberto.sassu@...wei.com>
> > > > > > > ---
> > > > > > > fs/namei.c | 9 ++++++---
> > > > > > > include/linux/ima.h | 7 +++++--
> > > > > > > security/integrity/ima/ima_main.c | 10 +++++++++-
> > > > > > > 3 files changed, 20 insertions(+), 6 deletions(-)
> > > > > > >
> > > > > > > diff --git a/fs/namei.c b/fs/namei.c
> > > > > > > index e56ff39a79bc..c5e96f716f98 100644
> > > > > > > --- a/fs/namei.c
> > > > > > > +++ b/fs/namei.c
> > > > > > > @@ -4024,6 +4024,7 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
> > > > > > > struct path path;
> > > > > > > int error;
> > > > > > > unsigned int lookup_flags = 0;
> > > > > > > + umode_t mode_stripped;
> > > > > > >
> > > > > > > error = may_mknod(mode);
> > > > > > > if (error)
> > > > > > > @@ -4034,8 +4035,9 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
> > > > > > > if (IS_ERR(dentry))
> > > > > > > goto out1;
> > > > > > >
> > > > > > > - error = security_path_mknod(&path, dentry,
> > > > > > > - mode_strip_umask(path.dentry->d_inode, mode), dev);
> > > > > > > + mode_stripped = mode_strip_umask(path.dentry->d_inode, mode);
> > > > > > > +
> > > > > > > + error = security_path_mknod(&path, dentry, mode_stripped, dev);
> > > > > > > if (error)
> > > > > > > goto out2;
> > > > > > >
> > > > > > > @@ -4045,7 +4047,8 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
> > > > > > > error = vfs_create(idmap, path.dentry->d_inode,
> > > > > > > dentry, mode, true);
> > > > > > > if (!error)
> > > > > > > - ima_post_path_mknod(idmap, dentry);
> > > > > > > + ima_post_path_mknod(idmap, &path, dentry,
> > > > > > > + mode_stripped, dev);
> > > > > > > break;
> > > > > > > case S_IFCHR: case S_IFBLK:
> > > > > > > error = vfs_mknod(idmap, path.dentry->d_inode,
> > > > > > > diff --git a/include/linux/ima.h b/include/linux/ima.h
> > > > > > > index 910a2f11a906..179ce52013b2 100644
> > > > > > > --- a/include/linux/ima.h
> > > > > > > +++ b/include/linux/ima.h
> > > > > > > @@ -32,7 +32,8 @@ extern int ima_read_file(struct file *file, enum kernel_read_file_id id,
> > > > > > > extern int ima_post_read_file(struct file *file, void *buf, loff_t size,
> > > > > > > enum kernel_read_file_id id);
> > > > > > > extern void ima_post_path_mknod(struct mnt_idmap *idmap,
> > > > > > > - struct dentry *dentry);
> > > > > > > + const struct path *dir, struct dentry *dentry,
> > > > > > > + umode_t mode, unsigned int dev);
> > > > > > > extern int ima_file_hash(struct file *file, char *buf, size_t buf_size);
> > > > > > > extern int ima_inode_hash(struct inode *inode, char *buf, size_t buf_size);
> > > > > > > extern void ima_kexec_cmdline(int kernel_fd, const void *buf, int size);
> > > > > > > @@ -114,7 +115,9 @@ static inline int ima_post_read_file(struct file *file, void *buf, loff_t size,
> > > > > > > }
> > > > > > >
> > > > > > > static inline void ima_post_path_mknod(struct mnt_idmap *idmap,
> > > > > > > - struct dentry *dentry)
> > > > > > > + const struct path *dir,
> > > > > > > + struct dentry *dentry,
> > > > > > > + umode_t mode, unsigned int dev)
> > > > > > > {
> > > > > > > return;
> > > > > > > }
> > > > > > > diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c
> > > > > > > index 365db0e43d7c..76eba92d7f10 100644
> > > > > > > --- a/security/integrity/ima/ima_main.c
> > > > > > > +++ b/security/integrity/ima/ima_main.c
> > > > > > > @@ -696,18 +696,26 @@ void ima_post_create_tmpfile(struct mnt_idmap *idmap,
> > > > > > > /**
> > > > > > > * ima_post_path_mknod - mark as a new inode
> > > > > > > * @idmap: idmap of the mount the inode was found from
> > > > > > > + * @dir: path structure of parent of the new file
> > > > > > > * @dentry: newly created dentry
> > > > > > > + * @mode: mode of the new file
> > > > > > > + * @dev: undecoded device number
> > > > > > > *
> > > > > > > * Mark files created via the mknodat syscall as new, so that the
> > > > > > > * file data can be written later.
> > > > > > > */
> > > > > > > void ima_post_path_mknod(struct mnt_idmap *idmap,
> > > > > > > - struct dentry *dentry)
> > > > > > > + const struct path *dir, struct dentry *dentry,
> > > > > > > + umode_t mode, unsigned int dev)
> > > > > > > {
> > > > > > > struct integrity_iint_cache *iint;
> > > > > > > struct inode *inode = dentry->d_inode;
> > > > > > > int must_appraise;
> > > > > > >
> > > > > > > + /* See do_mknodat(), IMA is executed for case 0: and case S_IFREG: */
> > > > > > > + if ((mode & S_IFMT) != 0 && (mode & S_IFMT) != S_IFREG)
> > > > > > > + return;
> > > > > > > +
> > > > > >
> > > > > > There's already a check below to make sure that this is a regular file.
> > > > > > Are both needed?
> > > > >
> > > > > You are right, I can remove the first check.
> > > >
> > > > The question then becomes why modify hook the arguments?
> > >
> > > We need to make sure that ima_post_path_mknod() has the same parameters
> > > as the LSM hook at the time we register it to the LSM infrastructure.
> >
> > I'm trying to understand why the pre hook parameters and the missing
> > IMA parameter are used, as opposed to just defining the new
> > post_path_mknod hook like IMA.
>
> As an empyrical rule, I pass the same parameters as the corresponding
> pre hook (plus idmap, in this case). This is similar to the
> inode_setxattr hook. But I can be wrong, if desired I can reduce.
The inode_setxattr hook change example is legitimate, as EVM includes
idmap, while IMA doesn't.
Unless there is a good reason for the additional parameters, I'm not
sure that adding them makes sense. Not modifying the parameter list
will reduce the size of this patch set.
--
thanks,
Mimi
Powered by blists - more mailing lists