From: Jann Horn <jannh@xxxxxxxxxx>
commit bdebd6a2831b6fab69eb85cee74a8ba77f1a1cc2 upstream.
remap_vmalloc_range() has had various issues with the bounds checks it
promises to perform ("This function checks that addr is a valid
vmalloc'ed area, and that it is big enough to cover the vma") over time,
e.g.:
- not detecting pgoff<<PAGE_SHIFT overflow
- not detecting (pgoff<<PAGE_SHIFT)+usize overflow
- not checking whether addr and addr+(pgoff<<PAGE_SHIFT) are the same
vmalloc allocation
- comparing a potentially wildly out-of-bounds pointer with the end of
the vmalloc region
In particular, since commit fc9702273e2e ("bpf: Add mmap() support for
BPF_MAP_TYPE_ARRAY"), unprivileged users can cause kernel null pointer
dereferences by calling mmap() on a BPF map with a size that is bigger
than the distance from the start of the BPF map to the end of the
address space.
This could theoretically be used as a kernel ASLR bypass, by using
whether mmap() with a given offset oopses or returns an error code to
perform a binary search over the possible address range.
To allow remap_vmalloc_range_partial() to verify that addr and
addr+(pgoff<<PAGE_SHIFT) are in the same vmalloc region, pass the offset
to remap_vmalloc_range_partial() instead of adding it to the pointer in
remap_vmalloc_range().
In remap_vmalloc_range_partial(), fix the check against
get_vm_area_size() by using size comparisons instead of pointer
comparisons, and add checks for pgoff.
Fixes: 833423143c3a ("[PATCH] mm: introduce remap_vmalloc_range()")
Signed-off-by: Jann Horn <jannh@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Cc: stable@xxxxxxxxxxxxxxx
Cc: Alexei Starovoitov <ast@xxxxxxxxxx>
Cc: Daniel Borkmann <daniel@xxxxxxxxxxxxx>
Cc: Martin KaFai Lau <kafai@xxxxxx>
Cc: Song Liu <songliubraving@xxxxxx>
Cc: Yonghong Song <yhs@xxxxxx>
Cc: Andrii Nakryiko <andriin@xxxxxx>
Cc: John Fastabend <john.fastabend@xxxxxxxxx>
Cc: KP Singh <kpsingh@xxxxxxxxxxxx>
Link: http://lkml.kernel.org/r/20200415222312.236431-1-jannh@xxxxxxxxxx
Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
fs/proc/vmcore.c | 5 +++--
include/linux/vmalloc.h | 2 +-
mm/vmalloc.c | 16 +++++++++++++---
samples/vfio-mdev/mdpy.c | 2 +-
4 files changed, 18 insertions(+), 7 deletions(-)
--- a/fs/proc/vmcore.c
+++ b/fs/proc/vmcore.c
@@ -266,7 +266,8 @@ static int vmcoredd_mmap_dumps(struct vm
if (start < offset + dump->size) {
tsz = min(offset + (u64)dump->size - start, (u64)size);
buf = dump->buf + start - offset;
- if (remap_vmalloc_range_partial(vma, dst, buf, tsz)) {
+ if (remap_vmalloc_range_partial(vma, dst, buf, 0,
+ tsz)) {
ret = -EFAULT;
goto out_unlock;
}
@@ -624,7 +625,7 @@ static int mmap_vmcore(struct file *file
tsz = min(elfcorebuf_sz + elfnotes_sz - (size_t)start, size);
kaddr = elfnotes_buf + start - elfcorebuf_sz - vmcoredd_orig_sz;
if (remap_vmalloc_range_partial(vma, vma->vm_start + len,
- kaddr, tsz))
+ kaddr, 0, tsz))
goto fail;
size -= tsz;
--- a/include/linux/vmalloc.h
+++ b/include/linux/vmalloc.h
@@ -137,7 +137,7 @@ extern void vunmap(const void *addr);
extern int remap_vmalloc_range_partial(struct vm_area_struct *vma,
unsigned long uaddr, void *kaddr,
- unsigned long size);
+ unsigned long pgoff, unsigned long size);
extern int remap_vmalloc_range(struct vm_area_struct *vma, void *addr,
unsigned long pgoff);
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -34,6 +34,7 @@
#include <linux/llist.h>
#include <linux/bitops.h>
#include <linux/rbtree_augmented.h>
+#include <linux/overflow.h>
#include <linux/uaccess.h>
#include <asm/tlbflush.h>
@@ -3054,6 +3055,7 @@ finished:
* @vma: vma to cover
* @uaddr: target user address to start at
* @kaddr: virtual address of vmalloc kernel memory
+ * @pgoff: offset from @kaddr to start at
* @size: size of map area
*
* Returns: 0 for success, -Exxx on failure
@@ -3066,9 +3068,15 @@ finished:
* Similar to remap_pfn_range() (see mm/memory.c)
*/
int remap_vmalloc_range_partial(struct vm_area_struct *vma, unsigned long uaddr,
- void *kaddr, unsigned long size)
+ void *kaddr, unsigned long pgoff,
+ unsigned long size)
{
struct vm_struct *area;
+ unsigned long off;
+ unsigned long end_index;
+
+ if (check_shl_overflow(pgoff, PAGE_SHIFT, &off))
+ return -EINVAL;
size = PAGE_ALIGN(size);
@@ -3082,8 +3090,10 @@ int remap_vmalloc_range_partial(struct v
if (!(area->flags & (VM_USERMAP | VM_DMA_COHERENT)))
return -EINVAL;
- if (kaddr + size > area->addr + get_vm_area_size(area))
+ if (check_add_overflow(size, off, &end_index) ||
+ end_index > get_vm_area_size(area))
return -EINVAL;
+ kaddr += off;
do {
struct page *page = vmalloc_to_page(kaddr);
@@ -3122,7 +3132,7 @@ int remap_vmalloc_range(struct vm_area_s
unsigned long pgoff)
{
return remap_vmalloc_range_partial(vma, vma->vm_start,
- addr + (pgoff << PAGE_SHIFT),
+ addr, pgoff,
vma->vm_end - vma->vm_start);
}
EXPORT_SYMBOL(remap_vmalloc_range);
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -418,7 +418,7 @@ static int mdpy_mmap(struct mdev_device
return -EINVAL;
return remap_vmalloc_range_partial(vma, vma->vm_start,
- mdev_state->memblk,
+ mdev_state->memblk, 0,
vma->vm_end - vma->vm_start);
}
