| Message ID | 20190401103403.14130-1-omosnace@redhat.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [v3] kernfs: fix xattr name handling in LSM helpers | expand |
On Mon, Apr 1, 2019 at 6:34 AM Ondrej Mosnacek <omosnace@redhat.com> wrote: > The implementation of kernfs_security_xattr_*() helpers reuses the > kernfs_node_xattr_*() functions, which take the suffix of the xattr name > and extract full xattr name from it using xattr_full_name(). However, > this function relies on the fact that the suffix passed to xattr > handlers from VFS is always constructed from the full name by just > incerementing the pointer. This doesn't necessarily hold for the callers > of kernfs_security_xattr_*(), so their usage will easily lead to > out-of-bounds access. > > Fix this by moving the xattr name reconstruction to the VFS xattr > handlers and replacing the kernfs_security_xattr_*() helpers with more > general kernfs_xattr_*() helpers that take full xattr name and allow > accessing all kernfs node's xattrs. > > Reported-by: kernel test robot <rong.a.chen@intel.com> > Fixes: b230d5aba2d1 ("LSM: add new hook for kernfs node initialization") > Fixes: ec882da5cda9 ("selinux: implement the kernfs_init_security hook") > Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com> > --- > > v3: simplify kernfs xattr helpers as per Paul's suggestion > v2: just rebase to update diff context > > fs/kernfs/inode.c | 62 ++++++++++++++-------------------------- > include/linux/kernfs.h | 18 ++++++------ > security/selinux/hooks.c | 9 +++--- > 3 files changed, 33 insertions(+), 56 deletions(-) This is better, thanks. Merged into selinux/next.
On Tue, Apr 2, 2019 at 7:10 PM Paul Moore <paul@paul-moore.com> wrote: > On Mon, Apr 1, 2019 at 6:34 AM Ondrej Mosnacek <omosnace@redhat.com> wrote: > > The implementation of kernfs_security_xattr_*() helpers reuses the > > kernfs_node_xattr_*() functions, which take the suffix of the xattr name > > and extract full xattr name from it using xattr_full_name(). However, > > this function relies on the fact that the suffix passed to xattr > > handlers from VFS is always constructed from the full name by just > > incerementing the pointer. This doesn't necessarily hold for the callers > > of kernfs_security_xattr_*(), so their usage will easily lead to > > out-of-bounds access. > > > > Fix this by moving the xattr name reconstruction to the VFS xattr > > handlers and replacing the kernfs_security_xattr_*() helpers with more > > general kernfs_xattr_*() helpers that take full xattr name and allow > > accessing all kernfs node's xattrs. > > > > Reported-by: kernel test robot <rong.a.chen@intel.com> > > Fixes: b230d5aba2d1 ("LSM: add new hook for kernfs node initialization") > > Fixes: ec882da5cda9 ("selinux: implement the kernfs_init_security hook") > > Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com> > > --- > > > > v3: simplify kernfs xattr helpers as per Paul's suggestion > > v2: just rebase to update diff context > > > > fs/kernfs/inode.c | 62 ++++++++++++++-------------------------- > > include/linux/kernfs.h | 18 ++++++------ > > security/selinux/hooks.c | 9 +++--- > > 3 files changed, 33 insertions(+), 56 deletions(-) > > This is better, thanks. Merged into selinux/next. ... and I've just popped it off selinux/next. It looks like you need to export the kernfs functions. ld: security/selinux/hooks.o: in function `selinux_kernfs_init_security': /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux-5.1. 0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3397: undefined re ference to `kernfs_xattr_get' ld: /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux- 5.1.0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3408: undefine d reference to `kernfs_xattr_get' ld: /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux- 5.1.0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3441: undefine d reference to `kernfs_xattr_set' make: *** [Makefile:1033: vmlinux] Error 1
On Wed, Apr 3, 2019 at 3:24 AM Paul Moore <paul@paul-moore.com> wrote: > On Tue, Apr 2, 2019 at 7:10 PM Paul Moore <paul@paul-moore.com> wrote: > > On Mon, Apr 1, 2019 at 6:34 AM Ondrej Mosnacek <omosnace@redhat.com> wrote: > > > The implementation of kernfs_security_xattr_*() helpers reuses the > > > kernfs_node_xattr_*() functions, which take the suffix of the xattr name > > > and extract full xattr name from it using xattr_full_name(). However, > > > this function relies on the fact that the suffix passed to xattr > > > handlers from VFS is always constructed from the full name by just > > > incerementing the pointer. This doesn't necessarily hold for the callers > > > of kernfs_security_xattr_*(), so their usage will easily lead to > > > out-of-bounds access. > > > > > > Fix this by moving the xattr name reconstruction to the VFS xattr > > > handlers and replacing the kernfs_security_xattr_*() helpers with more > > > general kernfs_xattr_*() helpers that take full xattr name and allow > > > accessing all kernfs node's xattrs. > > > > > > Reported-by: kernel test robot <rong.a.chen@intel.com> > > > Fixes: b230d5aba2d1 ("LSM: add new hook for kernfs node initialization") > > > Fixes: ec882da5cda9 ("selinux: implement the kernfs_init_security hook") > > > Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com> > > > --- > > > > > > v3: simplify kernfs xattr helpers as per Paul's suggestion > > > v2: just rebase to update diff context > > > > > > fs/kernfs/inode.c | 62 ++++++++++++++-------------------------- > > > include/linux/kernfs.h | 18 ++++++------ > > > security/selinux/hooks.c | 9 +++--- > > > 3 files changed, 33 insertions(+), 56 deletions(-) > > > > This is better, thanks. Merged into selinux/next. > > ... and I've just popped it off selinux/next. It looks like you need > to export the kernfs functions. > > ld: security/selinux/hooks.o: in function `selinux_kernfs_init_security': > /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux-5.1. > 0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3397: undefined re > ference to `kernfs_xattr_get' > ld: /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux- > 5.1.0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3408: undefine > d reference to `kernfs_xattr_get' > ld: /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux- > 5.1.0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3441: undefine > d reference to `kernfs_xattr_set' > make: *** [Makefile:1033: vmlinux] Error 1 Hm, I *thought* I ran the build before posting... The problem is that the function names in kernfs.h and inode.c didn't match. EXPORT_SYMBOL_GPL() shouldn't be needed since both kernfs and SELinux are always built-in and there seems to be an existing trend in kernfs to not export functions unless necessary. I should have a fixed patch ready soon.
On Wed, Apr 3, 2019 at 3:25 AM Ondrej Mosnacek <omosnace@redhat.com> wrote: > On Wed, Apr 3, 2019 at 3:24 AM Paul Moore <paul@paul-moore.com> wrote: > > On Tue, Apr 2, 2019 at 7:10 PM Paul Moore <paul@paul-moore.com> wrote: > > > On Mon, Apr 1, 2019 at 6:34 AM Ondrej Mosnacek <omosnace@redhat.com> wrote: > > > > The implementation of kernfs_security_xattr_*() helpers reuses the > > > > kernfs_node_xattr_*() functions, which take the suffix of the xattr name > > > > and extract full xattr name from it using xattr_full_name(). However, > > > > this function relies on the fact that the suffix passed to xattr > > > > handlers from VFS is always constructed from the full name by just > > > > incerementing the pointer. This doesn't necessarily hold for the callers > > > > of kernfs_security_xattr_*(), so their usage will easily lead to > > > > out-of-bounds access. > > > > > > > > Fix this by moving the xattr name reconstruction to the VFS xattr > > > > handlers and replacing the kernfs_security_xattr_*() helpers with more > > > > general kernfs_xattr_*() helpers that take full xattr name and allow > > > > accessing all kernfs node's xattrs. > > > > > > > > Reported-by: kernel test robot <rong.a.chen@intel.com> > > > > Fixes: b230d5aba2d1 ("LSM: add new hook for kernfs node initialization") > > > > Fixes: ec882da5cda9 ("selinux: implement the kernfs_init_security hook") > > > > Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com> > > > > --- > > > > > > > > v3: simplify kernfs xattr helpers as per Paul's suggestion > > > > v2: just rebase to update diff context > > > > > > > > fs/kernfs/inode.c | 62 ++++++++++++++-------------------------- > > > > include/linux/kernfs.h | 18 ++++++------ > > > > security/selinux/hooks.c | 9 +++--- > > > > 3 files changed, 33 insertions(+), 56 deletions(-) > > > > > > This is better, thanks. Merged into selinux/next. > > > > ... and I've just popped it off selinux/next. It looks like you need > > to export the kernfs functions. > > > > ld: security/selinux/hooks.o: in function `selinux_kernfs_init_security': > > /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux-5.1. > > 0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3397: undefined re > > ference to `kernfs_xattr_get' > > ld: /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux- > > 5.1.0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3408: undefine > > d reference to `kernfs_xattr_get' > > ld: /pjob.data/scratch/t/server-build.lan-2100-uHT0/BUILD/kernel-5.0.fc31/linux- > > 5.1.0-0.rc3.git1.1.2.secnext.fc31.x86_64/security/selinux/hooks.c:3441: undefine > > d reference to `kernfs_xattr_set' > > make: *** [Makefile:1033: vmlinux] Error 1 > > Hm, I *thought* I ran the build before posting... Spoiler alert: you didn't. On my end I know I didn't build a full kernel to test that your patch built, I just built the fs/kernfs and security/selinux directories. I should add a verification step to help ensure stuff like this doesn't get through review, and you should *ALWAYS* make sure you can build and boot a kernel with your patches before submitting them; I don't care if you just changed the name of a variable, build a new kernel and boot it. > The problem is that > the function names in kernfs.h and inode.c didn't match. > EXPORT_SYMBOL_GPL() shouldn't be needed since both kernfs and SELinux > are always built-in and there seems to be an existing trend in kernfs > to not export functions unless necessary. I didn't look that closely at the failure, I just yanked your patch out of the tree. Historically most failures like this are usually caused by wrong/missing exports, so I just assumed that was the case.
diff --git a/fs/kernfs/inode.c b/fs/kernfs/inode.c index 673ef598d97d..1e93fe8c67f4 100644 --- a/fs/kernfs/inode.c +++ b/fs/kernfs/inode.c @@ -288,63 +288,57 @@ int kernfs_iop_permission(struct inode *inode, int mask) return generic_permission(inode, mask); } -static int kernfs_node_xattr_get(const struct xattr_handler *handler, - struct kernfs_node *kn, const char *suffix, - void *value, size_t size) +int kernfs_node_xattr_get(struct kernfs_node *kn, const char *name, + void *value, size_t size) { - const char *name = xattr_full_name(handler, suffix); - struct kernfs_iattrs *attrs; - - attrs = kernfs_iattrs_noalloc(kn); + struct kernfs_iattrs *attrs = kernfs_iattrs_noalloc(kn); if (!attrs) return -ENODATA; return simple_xattr_get(&attrs->xattrs, name, value, size); } -static int kernfs_node_xattr_set(const struct xattr_handler *handler, - struct kernfs_node *kn, const char *suffix, - const void *value, size_t size, int flags) +int kernfs_node_xattr_set(struct kernfs_node *kn, const char *name, + const void *value, size_t size, int flags) { - const char *name = xattr_full_name(handler, suffix); - struct kernfs_iattrs *attrs; - - attrs = kernfs_iattrs(kn); + struct kernfs_iattrs *attrs = kernfs_iattrs(kn); if (!attrs) return -ENOMEM; return simple_xattr_set(&attrs->xattrs, name, value, size, flags); } -static int kernfs_xattr_get(const struct xattr_handler *handler, - struct dentry *unused, struct inode *inode, - const char *suffix, void *value, size_t size) +static int kernfs_vfs_xattr_get(const struct xattr_handler *handler, + struct dentry *unused, struct inode *inode, + const char *suffix, void *value, size_t size) { + const char *name = xattr_full_name(handler, suffix); struct kernfs_node *kn = inode->i_private; - return kernfs_node_xattr_get(handler, kn, suffix, value, size); + return kernfs_node_xattr_get(kn, name, value, size); } -static int kernfs_xattr_set(const struct xattr_handler *handler, - struct dentry *unused, struct inode *inode, - const char *suffix, const void *value, - size_t size, int flags) +static int kernfs_vfs_xattr_set(const struct xattr_handler *handler, + struct dentry *unused, struct inode *inode, + const char *suffix, const void *value, + size_t size, int flags) { + const char *name = xattr_full_name(handler, suffix); struct kernfs_node *kn = inode->i_private; - return kernfs_node_xattr_set(handler, kn, suffix, value, size, flags); + return kernfs_node_xattr_set(kn, name, value, size, flags); } static const struct xattr_handler kernfs_trusted_xattr_handler = { .prefix = XATTR_TRUSTED_PREFIX, - .get = kernfs_xattr_get, - .set = kernfs_xattr_set, + .get = kernfs_vfs_xattr_get, + .set = kernfs_vfs_xattr_set, }; static const struct xattr_handler kernfs_security_xattr_handler = { .prefix = XATTR_SECURITY_PREFIX, - .get = kernfs_xattr_get, - .set = kernfs_xattr_set, + .get = kernfs_vfs_xattr_get, + .set = kernfs_vfs_xattr_set, }; const struct xattr_handler *kernfs_xattr_handlers[] = { @@ -352,17 +346,3 @@ const struct xattr_handler *kernfs_xattr_handlers[] = { &kernfs_security_xattr_handler, NULL }; - -int kernfs_security_xattr_get(struct kernfs_node *kn, const char *suffix, - void *value, size_t size) -{ - return kernfs_node_xattr_get(&kernfs_security_xattr_handler, - kn, suffix, value, size); -} - -int kernfs_security_xattr_set(struct kernfs_node *kn, const char *suffix, - void *value, size_t size, int flags) -{ - return kernfs_node_xattr_set(&kernfs_security_xattr_handler, - kn, suffix, value, size, flags); -} diff --git a/include/linux/kernfs.h b/include/linux/kernfs.h index 39eea07c2900..b3f62e7bb996 100644 --- a/include/linux/kernfs.h +++ b/include/linux/kernfs.h @@ -371,10 +371,10 @@ __poll_t kernfs_generic_poll(struct kernfs_open_file *of, struct poll_table_struct *pt); void kernfs_notify(struct kernfs_node *kn); -int kernfs_security_xattr_get(struct kernfs_node *kn, const char *suffix, - void *value, size_t size); -int kernfs_security_xattr_set(struct kernfs_node *kn, const char *suffix, - void *value, size_t size, int flags); +int kernfs_xattr_get(struct kernfs_node *kn, const char *name, + void *value, size_t size); +int kernfs_xattr_set(struct kernfs_node *kn, const char *name, + void *value, size_t size, int flags); const void *kernfs_super_ns(struct super_block *sb); int kernfs_get_tree(struct fs_context *fc); @@ -478,14 +478,12 @@ static inline int kernfs_setattr(struct kernfs_node *kn, static inline void kernfs_notify(struct kernfs_node *kn) { } -static inline int kernfs_security_xattr_get(struct kernfs_node *kn, - const char *suffix, void *value, - size_t size) +static inline int kernfs_xattr_get(struct kernfs_node *kn, const char *name, + void *value, size_t size) { return -ENOSYS; } -static inline int kernfs_security_xattr_set(struct kernfs_node *kn, - const char *suffix, void *value, - size_t size, int flags) +static inline int kernfs_xattr_set(struct kernfs_node *kn, const char *name, + void *value, size_t size, int flags) { return -ENOSYS; } static inline const void *kernfs_super_ns(struct super_block *sb) diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index b6e61524d68d..d5fdcb0d26fe 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -3394,7 +3394,7 @@ static int selinux_kernfs_init_security(struct kernfs_node *kn_dir, int rc; char *context; - rc = kernfs_security_xattr_get(kn_dir, XATTR_SELINUX_SUFFIX, NULL, 0); + rc = kernfs_xattr_get(kn_dir, XATTR_NAME_SELINUX, NULL, 0); if (rc == -ENODATA) return 0; else if (rc < 0) @@ -3405,8 +3405,7 @@ static int selinux_kernfs_init_security(struct kernfs_node *kn_dir, if (!context) return -ENOMEM; - rc = kernfs_security_xattr_get(kn_dir, XATTR_SELINUX_SUFFIX, context, - clen); + rc = kernfs_xattr_get(kn_dir, XATTR_NAME_SELINUX, context, clen); if (rc < 0) { kfree(context); return rc; @@ -3439,8 +3438,8 @@ static int selinux_kernfs_init_security(struct kernfs_node *kn_dir, if (rc) return rc; - rc = kernfs_security_xattr_set(kn, XATTR_SELINUX_SUFFIX, context, clen, - XATTR_CREATE); + rc = kernfs_xattr_set(kn, XATTR_NAME_SELINUX, context, clen, + XATTR_CREATE); kfree(context); return rc; }
The implementation of kernfs_security_xattr_*() helpers reuses the kernfs_node_xattr_*() functions, which take the suffix of the xattr name and extract full xattr name from it using xattr_full_name(). However, this function relies on the fact that the suffix passed to xattr handlers from VFS is always constructed from the full name by just incerementing the pointer. This doesn't necessarily hold for the callers of kernfs_security_xattr_*(), so their usage will easily lead to out-of-bounds access. Fix this by moving the xattr name reconstruction to the VFS xattr handlers and replacing the kernfs_security_xattr_*() helpers with more general kernfs_xattr_*() helpers that take full xattr name and allow accessing all kernfs node's xattrs. Reported-by: kernel test robot <rong.a.chen@intel.com> Fixes: b230d5aba2d1 ("LSM: add new hook for kernfs node initialization") Fixes: ec882da5cda9 ("selinux: implement the kernfs_init_security hook") Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com> --- v3: simplify kernfs xattr helpers as per Paul's suggestion v2: just rebase to update diff context fs/kernfs/inode.c | 62 ++++++++++++++-------------------------- include/linux/kernfs.h | 18 ++++++------ security/selinux/hooks.c | 9 +++--- 3 files changed, 33 insertions(+), 56 deletions(-)