From: Jann Horn jannh@google.com
[ Upstream commit 39bb67edcc582b3b386a9ec983da67fa8a10ec03 ]
The current code around TEE_IOCTL_PARAM_SIZE() is a bit wrong on 32-bit kernels: Multiplying a user-provided 32-bit value with the size of a structure can wrap around on such platforms.
Fix it by using saturating arithmetic for the size calculation.
This has no security consequences because, in all users of TEE_IOCTL_PARAM_SIZE(), the subsequent kcalloc() implicitly checks for wrapping.
Signed-off-by: Jann Horn jannh@google.com Signed-off-by: Jens Wiklander jens.wiklander@linaro.org Tested-by: Rouven Czerwinski rouven.czerwinski@linaro.org Signed-off-by: Sasha Levin sashal@kernel.org ---
**YES** This commit should be backported to stable kernel trees. Here's my extensive analysis: ## Vulnerability Analysis The commit fixes a real integer overflow vulnerability in the TEE (Trusted Execution Environment) subsystem on 32-bit kernels. The issue occurs in the `TEE_IOCTL_PARAM_SIZE()` macro defined as: ```c #define TEE_IOCTL_PARAM_SIZE(x) (sizeof(struct tee_param) * (x)) ``` Where `struct tee_ioctl_param` is 32 bytes (4 × 8-byte fields). On 32-bit systems, when a user provides a large `num_params` value, the multiplication `32 * num_params` can wrap around, potentially bypassing buffer length validation checks. ## Specific Vulnerable Code Locations The vulnerable pattern appears in 4 locations in `drivers/tee/tee_core.c`: 1. **Line 490**: `tee_ioctl_open_session()` - `sizeof(arg) + TEE_IOCTL_PARAM_SIZE(arg.num_params) != buf.buf_len` 2. **Line 568**: `tee_ioctl_invoke()` - `sizeof(arg) + TEE_IOCTL_PARAM_SIZE(arg.num_params) != buf.buf_len` 3. **Line 702**: `tee_ioctl_supp_recv()` - `sizeof(*uarg) + TEE_IOCTL_PARAM_SIZE(num_params) != buf.buf_len` 4. **Line 801**: `tee_ioctl_supp_send()` - `sizeof(*uarg) + TEE_IOCTL_PARAM_SIZE(num_params) > buf.buf_len` ## Attack Scenario On 32-bit systems, if `num_params = 0x08000000` (134,217,728), then: - `TEE_IOCTL_PARAM_SIZE(0x08000000) = 32 * 0x08000000 = 0x100000000` - This wraps to `0x00000000` on 32-bit systems - The size validation `sizeof(arg) + 0 != buf.buf_len` could be bypassed - Subsequent `kcalloc()` would allocate a huge buffer, but the size calculations would be wrong ## Why This Should Be Backported **Matches "YES" backport criteria from similar commits:** 1. **Security Fix**: Like Similar Commit #2 (rtl8188eu security hole) and #4 (gup_benchmark integer overflow), this prevents user-controlled integer overflow that could lead to memory corruption. 2. **Small, Contained Change**: The fix uses the kernel's standard `size_add()` and `size_mul()` helpers from `linux/overflow.h` - exactly like Similar Commit #3 which was noted as using `size_add()` for similar protection. 3. **Critical Subsystem**: TEE handles sensitive cryptographic operations and trusted applications, making security fixes here particularly important. 4. **Low Regression Risk**: The fix only adds overflow protection without changing functionality. The `size_mul()` and `size_add()` functions are well- tested kernel overflow helpers. 5. **User-Controllable Input**: The `num_params` value comes directly from user space through ioctl calls, making this a user-triggerable condition. 6. **Clear Security Impact**: While the commit states "no security consequences because kcalloc() implicitly checks for wrapping," this is defense-in-depth - the size validation should not rely on downstream allocation checks. The fix pattern exactly matches Similar Commit #4 (mm/gup_benchmark.c) which was backported for preventing integer overflow in size calculations with user-provided values. This commit follows the same security principle and uses the same overflow-safe arithmetic helpers that the kernel has standardized on for these scenarios.
drivers/tee/tee_core.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/drivers/tee/tee_core.c b/drivers/tee/tee_core.c index 98da206cd7615..a9a893bc19fa4 100644 --- a/drivers/tee/tee_core.c +++ b/drivers/tee/tee_core.c @@ -10,6 +10,7 @@ #include <linux/fs.h> #include <linux/idr.h> #include <linux/module.h> +#include <linux/overflow.h> #include <linux/slab.h> #include <linux/tee_drv.h> #include <linux/uaccess.h> @@ -19,7 +20,7 @@
#define TEE_NUM_DEVICES 32
-#define TEE_IOCTL_PARAM_SIZE(x) (sizeof(struct tee_param) * (x)) +#define TEE_IOCTL_PARAM_SIZE(x) (size_mul(sizeof(struct tee_param), (x)))
#define TEE_UUID_NS_NAME_SIZE 128
@@ -487,7 +488,7 @@ static int tee_ioctl_open_session(struct tee_context *ctx, if (copy_from_user(&arg, uarg, sizeof(arg))) return -EFAULT;
- if (sizeof(arg) + TEE_IOCTL_PARAM_SIZE(arg.num_params) != buf.buf_len) + if (size_add(sizeof(arg), TEE_IOCTL_PARAM_SIZE(arg.num_params)) != buf.buf_len) return -EINVAL;
if (arg.num_params) { @@ -565,7 +566,7 @@ static int tee_ioctl_invoke(struct tee_context *ctx, if (copy_from_user(&arg, uarg, sizeof(arg))) return -EFAULT;
- if (sizeof(arg) + TEE_IOCTL_PARAM_SIZE(arg.num_params) != buf.buf_len) + if (size_add(sizeof(arg), TEE_IOCTL_PARAM_SIZE(arg.num_params)) != buf.buf_len) return -EINVAL;
if (arg.num_params) { @@ -699,7 +700,7 @@ static int tee_ioctl_supp_recv(struct tee_context *ctx, if (get_user(num_params, &uarg->num_params)) return -EFAULT;
- if (sizeof(*uarg) + TEE_IOCTL_PARAM_SIZE(num_params) != buf.buf_len) + if (size_add(sizeof(*uarg), TEE_IOCTL_PARAM_SIZE(num_params)) != buf.buf_len) return -EINVAL;
params = kcalloc(num_params, sizeof(struct tee_param), GFP_KERNEL); @@ -798,7 +799,7 @@ static int tee_ioctl_supp_send(struct tee_context *ctx, get_user(num_params, &uarg->num_params)) return -EFAULT;
- if (sizeof(*uarg) + TEE_IOCTL_PARAM_SIZE(num_params) > buf.buf_len) + if (size_add(sizeof(*uarg), TEE_IOCTL_PARAM_SIZE(num_params)) > buf.buf_len) return -EINVAL;
params = kcalloc(num_params, sizeof(struct tee_param), GFP_KERNEL);
op-tee@lists.trustedfirmware.org