Skip to contentSkip to conversation

get_fs_type: Validate fs type string argument

PATCH Superseded Chandan Rajendra 1 participant 3 messages Jul 11, 2017 – Jul 24, 2017 VFS

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 get_fs_type: Validate fs type string argument

On ppc64, When a non-nul terminated string is passed as an argument to the mount(2) syscall, copy_mount_string() ends up allocating 64k (the PAGE_SIZE on ppc64) worth of space for holding the string in kernel's address space.

Later, in set_precision() (invoked indirectly by get_fs_type()), we end up assigning 65535 as the value to 'struct printf_spec'->precision member. This field has a width of 16 bits and hence a truncated version of the original value ends up being assigned. This causes the "WARN_ONCE(spec->precision != prec, "precision %d too large", prec)" statement inside set_precision() to be executed.

This commit fixes the bug by validating the length of the "filesystem type" argument passed to get_fs_type() function.

---
 fs/filesystems.c | 3 +++
 1 file changed, 3 insertions(+)
fs/filesystems.c +3 -0
index a920ad2..8e309d3 100644
@@ -275,6 +275,9 @@ struct file_system_type *get_fs_type(const char *name)
 	const char *dot = strchr(name, '.');
 	int len = dot ? dot - name : strlen(name);
 
+       if (len >= PATH_MAX)
+               return NULL;
+
 	fs = __get_fs_type(name, len);
 	if (!fs && (request_module("fs-%.*s", len, name) == 0)) {
 		fs = __get_fs_type(name, len);
Signature · 1 line
2.9.4
#1 lore raw

PATCH v2 1/1 get_fs_type: Validate fs type string argument

On ppc64, When a non-nul terminated string is passed as an argument to the mount(2) syscall, copy_mount_string() ends up allocating 64k (the PAGE_SIZE on ppc64) worth of space for holding the string in kernel's address space.

Later, in set_precision() (invoked indirectly by get_fs_type()), we end up assigning 65535 as the value to 'struct printf_spec'->precision member. This field has a width of 16 bits and also it is signed data type. Hence an invalid value ends up getting assigned. This causes the "WARN_ONCE(spec->precision != prec, "precision %d too large", prec)" statement inside set_precision() to be executed.

This commit fixes the bug by validating the length of the "filesystem type" argument passed to get_fs_type() function.

---
Changelog:
v1->v2: Fix commit message.

 fs/filesystems.c | 3 +++
 1 file changed, 3 insertions(+)
fs/filesystems.c +3 -0
index a920ad2..8e309d3 100644
@@ -275,6 +275,9 @@ struct file_system_type *get_fs_type(const char *name)
 	const char *dot = strchr(name, '.');
 	int len = dot ? dot - name : strlen(name);
 
+       if (len >= PATH_MAX)
+               return NULL;
+
 	fs = __get_fs_type(name, len);
 	if (!fs && (request_module("fs-%.*s", len, name) == 0)) {
 		fs = __get_fs_type(name, len);
Signature · 1 line
2.9.4
#2 lore raw
Quoted 31 lines · IST Chandan Rajendra

On ppc64, When a non-nul terminated string is passed as an argument to the mount(2) syscall, copy_mount_string() ends up allocating 64k (the PAGE_SIZE on ppc64) worth of space for holding the string in kernel's address space.

Later, in set_precision() (invoked indirectly by get_fs_type()), we end up assigning 65535 as the value to 'struct printf_spec'->precision member. This field has a width of 16 bits and also it is signed data type. Hence an invalid value ends up getting assigned. This causes the "WARN_ONCE(spec->precision != prec, "precision %d too large", prec)" statement inside set_precision() to be executed.

This commit fixes the bug by validating the length of the "filesystem type" argument passed to get_fs_type() function.

---
Changelog:
v1->v2: Fix commit message.

 fs/filesystems.c | 3 +++
 1 file changed, 3 insertions(+)
fs/filesystems.c +0 -0
index a920ad2..8e309d3 100644

fs/filesystems.c struct file_system_type *get_fs_type(const char *name) · line 283

Quoted 9 lines
fs/filesystems.c +3 -0
@@ -275,6 +275,9 @@ struct file_system_type *get_fs_type(const char *name)
 	const char *dot = strchr(name, '.');
 	int len = dot ? dot - name : strlen(name);
 
+       if (len >= PATH_MAX)
+               return NULL;
+
 	fs = __get_fs_type(name, len);
 	if (!fs && (request_module("fs-%.*s", len, name) == 0)) {
 		fs = __get_fs_type(name, len);

Hi Al Viro,

Can you please let me know if you have any review comments about this patch.

Signature · 1 line
chandan