Skip to contentSkip to conversation

ntfs3: validate split-point offset in indx_insert_into_buffer

PATCH In review Michael Bommarito 2 participants 2 messages Apr 17 – Apr 30

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ntfs3: validate split-point offset in indx_insert_into_buffer

indx_insert_into_buffer() computes

    used = used1 - to_copy - sp_size;
    memmove(de_t, Add2Ptr(sp, sp_size), used - le32_to_cpu(hdr1->de_off));

where sp and sp_size come from hdr_find_split(). hdr_find_split() walks entries by le16_to_cpu(e->size) without validating that each step stays within hdr->used or that the size field is at least sizeof(struct NTFS_DE). index_hdr_check(), the on-load gatekeeper, only validates header-level fields (used, total, de_off) and does not walk per-entry sizes.

A crafted NTFS image whose leaf INDEX_HDR reports used == total but contains one interior NTFS_DE with size = 0xFFF0 therefore passes validation, descends to indx_insert_into_buffer() through the ntfs_create() -> indx_insert_entry() path, and makes hdr_find_split() return an sp whose sp_size (0xFFF0) greatly exceeds the remaining bytes in the buffer. The u32 subtraction underflows and the memmove count becomes a near-4-GiB value, producing an out-of-bounds kernel write that corrupts adjacent allocations and panics the kernel.

Reproduced on 7.0.0-rc7 with UML + KASAN via a crafted image and a single 'touch' inside the mounted directory; crash site resolves to fs/ntfs3/index.c at the memmove. Trigger requires only local mount of an attacker-supplied filesystem image (USB, loopback, or removable media auto-mount).

Reject the split whenever the chosen sp plus its declared size already extends past hdr1->used. This is the minimal fix; it preserves the existing hdr_find_split() contract and relies on the same out: cleanup path as the pre-existing error returns.

A prior OOB read in the very same indx_insert_into_buffer() memmove was fixed in commit b8c44949044e ("fs/ntfs3: Fix OOB read in indx_insert_into_buffer") by tightening hdr_find_e(), but that fix does not cover the split-point size field path addressed here: sp is returned by hdr_find_split(), not hdr_find_e(), and the underflow is driven by sp->size rather than hdr->used exceeding hdr->total.

Fixes 82cae269cfa9 ("fs/ntfs3: Add initialization of super block")
Assisted-by Claude:claude-opus-4-7

---

- FYI, I have a larger refactor variant that migrates hdr_find_split() to the validated hdr_next_de() helper and closes the whole per-entry size-read class for that walker. Happy to send it as v2 if you prefer the wider change; otherwise this minimal guard is scoped to the actual memmove underflow site and is easier to backport.

 fs/ntfs3/index.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
fs/ntfs3/index.c +14 -0
index 2c43e7c27861..24add048b4b5 100644
@@ -1844,6 +1844,20 @@ indx_insert_into_buffer(struct ntfs_index *indx, struct ntfs_inode *ni,
 	memcpy(up_e, sp, sp_size);
 
 	used1 = le32_to_cpu(hdr1->used);
+
+	/*
+	 * hdr_find_split does not validate per-entry sizes, so a crafted
+	 * NTFS_DE whose le16 size field is out of range can place sp such
+	 * that (PtrOffset(hdr1, sp) + sp_size) exceeds used1. Without this
+	 * guard the u32 'used = used1 - to_copy - sp_size' underflows and
+	 * the subsequent memmove count becomes a near-4-GiB value,
+	 * triggering an out-of-bounds kernel write.
+	 */
+	if (PtrOffset(hdr1, sp) + sp_size > used1) {
+		err = -EINVAL;
+		goto out;
+	}
+
 	hdr1_saved = kmemdup(hdr1, used1, GFP_NOFS);
 	if (!hdr1_saved) {
 		err = -ENOMEM;
Signature · 1 line
2.53.0
#1 lore raw
Quoted 84 lines · Michael Bommarito

[You don't often get email from [email protected]. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]

indx_insert_into_buffer() computes

     used = used1 - to_copy - sp_size;
     memmove(de_t, Add2Ptr(sp, sp_size), used - le32_to_cpu(hdr1->de_off));

where sp and sp_size come from hdr_find_split(). hdr_find_split() walks entries by le16_to_cpu(e->size) without validating that each step stays within hdr->used or that the size field is at least sizeof(struct NTFS_DE). index_hdr_check(), the on-load gatekeeper, only validates header-level fields (used, total, de_off) and does not walk per-entry sizes.

A crafted NTFS image whose leaf INDEX_HDR reports used == total but contains one interior NTFS_DE with size = 0xFFF0 therefore passes validation, descends to indx_insert_into_buffer() through the ntfs_create() -> indx_insert_entry() path, and makes hdr_find_split() return an sp whose sp_size (0xFFF0) greatly exceeds the remaining bytes in the buffer. The u32 subtraction underflows and the memmove count becomes a near-4-GiB value, producing an out-of-bounds kernel write that corrupts adjacent allocations and panics the kernel.

Reproduced on 7.0.0-rc7 with UML + KASAN via a crafted image and a single 'touch' inside the mounted directory; crash site resolves to fs/ntfs3/index.c at the memmove. Trigger requires only local mount of an attacker-supplied filesystem image (USB, loopback, or removable media auto-mount).

Reject the split whenever the chosen sp plus its declared size already extends past hdr1->used. This is the minimal fix; it preserves the existing hdr_find_split() contract and relies on the same out: cleanup path as the pre-existing error returns.

A prior OOB read in the very same indx_insert_into_buffer() memmove was fixed in commit b8c44949044e ("fs/ntfs3: Fix OOB read in indx_insert_into_buffer") by tightening hdr_find_e(), but that fix does not cover the split-point size field path addressed here: sp is returned by hdr_find_split(), not hdr_find_e(), and the underflow is driven by sp->size rather than hdr->used exceeding hdr->total.

Fixes 82cae269cfa9 ("fs/ntfs3: Add initialization of super block")
Assisted-by Claude:claude-opus-4-7

---

- FYI, I have a larger refactor variant that migrates hdr_find_split() to the validated hdr_next_de() helper and closes the whole per-entry size-read class for that walker. Happy to send it as v2 if you prefer the wider change; otherwise this minimal guard is scoped to the actual memmove underflow site and is easier to backport.

  fs/ntfs3/index.c | 14 ++++++++++++++
  1 file changed, 14 insertions(+)
fs/ntfs3/index.c +14 -1
index 2c43e7c27861..24add048b4b5 100644
@@ -1844,6 +1844,20 @@ indx_insert_into_buffer(struct ntfs_index *indx, struct ntfs_inode *ni,
         memcpy(up_e, sp, sp_size);
 
         used1 = le32_to_cpu(hdr1->used);
+
+       /*
+        * hdr_find_split does not validate per-entry sizes, so a crafted
+        * NTFS_DE whose le16 size field is out of range can place sp such
+        * that (PtrOffset(hdr1, sp) + sp_size) exceeds used1. Without this
+        * guard the u32 'used = used1 - to_copy - sp_size' underflows and
+        * the subsequent memmove count becomes a near-4-GiB value,
+        * triggering an out-of-bounds kernel write.
+        */
+       if (PtrOffset(hdr1, sp) + sp_size > used1) {
+               err = -EINVAL;
+               goto out;
+       }
+
         hdr1_saved = kmemdup(hdr1, used1, GFP_NOFS);
         if (!hdr1_saved) {
                 err = -ENOMEM;
--
Quoted 2 lines

2.53.0

Hello,

Sorry for the delay.
Patch is applied. Thank you.

Regards,
Konstantin