Vulnerabilities (CVE)

In Tensorflow before version 2.3.1, the `SparseCountSparseOutput` implementation does not validate that the input arguments form a valid sparse tensor. In particular, there is no validation that the `indices` tensor has the same shape as the `values` one. The values in these tensors are always accessed in parallel. Thus, a shape mismatch can result in accesses outside the bounds of heap allocated buffers. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in ... In Tensorflow before version 2.3.1, the `SparseCountSparseOutput` implementation does not validate that the input arguments form a valid sparse tensor. In particular, there is no validation that the `indices` tensor has the same shape as the `values` one. The values in these tensors are always accessed in parallel. Thus, a shape mismatch can result in accesses outside the bounds of heap allocated buffers. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.

Show More

In Tensorflow before version 2.3.1, the `SparseCountSparseOutput` implementation does not validate that the input arguments form a valid sparse tensor. In particular, there is no validation that the `indices` tensor has rank 2. This tensor must be a matrix because code assumes its elements are accessed as elements of a matrix. However, malicious users can pass in tensors of different rank, resulting in a `CHECK` assertion failure and a crash. This can be used to cause denial of service in servin ... In Tensorflow before version 2.3.1, the `SparseCountSparseOutput` implementation does not validate that the input arguments form a valid sparse tensor. In particular, there is no validation that the `indices` tensor has rank 2. This tensor must be a matrix because code assumes its elements are accessed as elements of a matrix. However, malicious users can pass in tensors of different rank, resulting in a `CHECK` assertion failure and a crash. This can be used to cause denial of service in serving installations, if users are allowed to control the components of the input sparse tensor. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.

Show More

In Tensorflow version 2.3.0, the `SparseCountSparseOutput` and `RaggedCountSparseOutput` implementations don't validate that the `weights` tensor has the same shape as the data. The check exists for `DenseCountSparseOutput`, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the hea ... In Tensorflow version 2.3.0, the `SparseCountSparseOutput` and `RaggedCountSparseOutput` implementations don't validate that the `weights` tensor has the same shape as the data. The check exists for `DenseCountSparseOutput`, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.

Show More

In Tensorflow before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, the implementation of `SparseFillEmptyRowsGrad` uses a double indexing pattern. It is possible for `reverse_index_map(i)` to be an index outside of bounds of `grad_values`, thus resulting in a heap buffer overflow. The issue is patched in commit 390611e0d45c5793c7066110af37c8514e6a6c54, and is released in TensorFlow versions 1.15.4, 2.0.3, 2.1.2, 2.2.1, or 2.3.1.

In Tensorflow before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, the `SparseFillEmptyRowsGrad` implementation has incomplete validation of the shapes of its arguments. Although `reverse_index_map_t` and `grad_values_t` are accessed in a similar pattern, only `reverse_index_map_t` is validated to be of proper shape. Hence, malicious users can pass a bad `grad_values_t` to trigger an assertion failure in `vec`, causing denial of service in serving installations. The issue is patched in commit ... In Tensorflow before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, the `SparseFillEmptyRowsGrad` implementation has incomplete validation of the shapes of its arguments. Although `reverse_index_map_t` and `grad_values_t` are accessed in a similar pattern, only `reverse_index_map_t` is validated to be of proper shape. Hence, malicious users can pass a bad `grad_values_t` to trigger an assertion failure in `vec`, causing denial of service in serving installations. The issue is patched in commit 390611e0d45c5793c7066110af37c8514e6a6c54, and is released in TensorFlow versions 1.15.4, 2.0.3, 2.1.2, 2.2.1, or 2.3.1."

Show More

In Tensorflow before versions 2.2.1 and 2.3.1, the implementation of `dlpack.to_dlpack` can be made to use uninitialized memory resulting in further memory corruption. This is because the pybind11 glue code assumes that the argument is a tensor. However, there is nothing stopping users from passing in a Python object instead of a tensor. The uninitialized memory address is due to a `reinterpret_cast` Since the `PyObject` is a Python object, not a TensorFlow Tensor, the cast to `EagerTensor` fail ... In Tensorflow before versions 2.2.1 and 2.3.1, the implementation of `dlpack.to_dlpack` can be made to use uninitialized memory resulting in further memory corruption. This is because the pybind11 glue code assumes that the argument is a tensor. However, there is nothing stopping users from passing in a Python object instead of a tensor. The uninitialized memory address is due to a `reinterpret_cast` Since the `PyObject` is a Python object, not a TensorFlow Tensor, the cast to `EagerTensor` fails. The issue is patched in commit 22e07fb204386768e5bcbea563641ea11f96ceb8 and is released in TensorFlow versions 2.2.1, or 2.3.1.

Show More

In Tensorflow before versions 2.2.1 and 2.3.1, if a user passes a list of strings to `dlpack.to_dlpack` there is a memory leak following an expected validation failure. The issue occurs because the `status` argument during validation failures is not properly checked. Since each of the above methods can return an error status, the `status` value must be checked before continuing. The issue is patched in commit 22e07fb204386768e5bcbea563641ea11f96ceb8 and is released in TensorFlow versions 2.2.1, ... In Tensorflow before versions 2.2.1 and 2.3.1, if a user passes a list of strings to `dlpack.to_dlpack` there is a memory leak following an expected validation failure. The issue occurs because the `status` argument during validation failures is not properly checked. Since each of the above methods can return an error status, the `status` value must be checked before continuing. The issue is patched in commit 22e07fb204386768e5bcbea563641ea11f96ceb8 and is released in TensorFlow versions 2.2.1, or 2.3.1.

Show More

In Tensorflow before versions 2.2.1 and 2.3.1, if a user passes an invalid argument to `dlpack.to_dlpack` the expected validations will cause variables to bind to `nullptr` while setting a `status` variable to the error condition. However, this `status` argument is not properly checked. Hence, code following these methods will bind references to null pointers. This is undefined behavior and reported as an error if compiling with `-fsanitize=null`. The issue is patched in commit 22e07fb204386768e ... In Tensorflow before versions 2.2.1 and 2.3.1, if a user passes an invalid argument to `dlpack.to_dlpack` the expected validations will cause variables to bind to `nullptr` while setting a `status` variable to the error condition. However, this `status` argument is not properly checked. Hence, code following these methods will bind references to null pointers. This is undefined behavior and reported as an error if compiling with `-fsanitize=null`. The issue is patched in commit 22e07fb204386768e5bcbea563641ea11f96ceb8 and is released in TensorFlow versions 2.2.1, or 2.3.1.

Show More

In Tensorflow before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, the `tf.raw_ops.Switch` operation takes as input a tensor and a boolean and outputs two tensors. Depending on the boolean value, one of the tensors is exactly the input tensor whereas the other one should be an empty tensor. However, the eager runtime traverses all tensors in the output. Since only one of the tensors is defined, the other one is `nullptr`, hence we are binding a reference to `nullptr`. This is undefined behavio ... In Tensorflow before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, the `tf.raw_ops.Switch` operation takes as input a tensor and a boolean and outputs two tensors. Depending on the boolean value, one of the tensors is exactly the input tensor whereas the other one should be an empty tensor. However, the eager runtime traverses all tensors in the output. Since only one of the tensors is defined, the other one is `nullptr`, hence we are binding a reference to `nullptr`. This is undefined behavior and reported as an error if compiling with `-fsanitize=null`. In this case, this results in a segmentation fault The issue is patched in commit da8558533d925694483d2c136a9220d6d49d843c, and is released in TensorFlow versions 1.15.4, 2.0.3, 2.1.2, 2.2.1, or 2.3.1.

Show More

An issue was discovered on LG mobile devices with Android OS software before 2020-06-01. Local users can cause a denial of service because checking of the userdata partition is mishandled. The LG ID is LVE-SMP-200014 (June 2020).

An issue was discovered on LG mobile devices with Android OS 7.2, 8.0, 8.1, 9, and 10 (MTK chipsets). A dangerous AT command was made available even though it is unused. The LG ID is LVE-SMP-200010 (June 2020).

An issue was discovered on LG mobile devices with Android OS 9 and 10 (MTK chipsets). An AT command handler allows attackers to bypass intended access restrictions. The LG ID is LVE-SMP-200009 (June 2020).

An issue was discovered on LG mobile devices with Android OS 7.2, 8.0, 8.1, 9, and 10 (MTK chipsets). Code execution can occur via an MTK AT command handler buffer overflow. The LG ID is LVE-SMP-200008 (June 2020).

An issue was discovered on LG mobile devices with Android OS 7.2, 8.0, 8.1, 9, and 10 (MTK chipsets). Code execution can occur via a custom AT command handler buffer overflow. The LG ID is LVE-SMP-200007 (June 2020).

An issue was discovered on Samsung mobile devices with P(9.0) and Q(10.0) software. The DeX Lockscreen feature does not block access to Quick Panel and notifications. The Samsung ID is SVE-2020-17187 (June 2020).

An issue was discovered on Samsung mobile devices with Q(10.0) software. The Lockscreen feature does not block Quick Panel access to Music Share. The Samsung ID is SVE-2020-17145 (June 2020).

An issue was discovered on Samsung mobile devices with O(8.x), P(9.0), and Q(10.0) software. HWRResProvider allows path traversal for data exposure. The Samsung ID is SVE-2020-16954 (June 2020).

An issue was discovered on Samsung mobile devices with O(8.x) (with TEEGRIS) software. The Gatekeeper Trustlet allows a brute-force attack on user credentials. The Samsung ID is SVE-2020-16908 (June 2020).

An issue was discovered on Samsung mobile devices with O(8.x), P(9.0), and Q(10.0) (with TEEGRIS) software. Secure Folder does not properly restrict use of Android Debug Bridge (adb) for arbitrary installations. The Samsung ID is SVE-2020-17369 (June 2020).

An issue was discovered on Samsung mobile devices with O(8.x), P(9.0), and Q(10.0) software. The system area allows arbitrary file overwrites via a symlink attack. The Samsung ID is SVE-2020-17183 (June 2020).

An issue was discovered on Samsung mobile devices with Q(10.0) (with TEEGRIS on Exynos chipsets) software. The Widevine Trustlet allows arbitrary code execution because of memory disclosure, The Samsung IDs are SVE-2020-17117, SVE-2020-17118, SVE-2020-17119, and SVE-2020-17161 (June 2020).

An issue was discovered on Samsung mobile devices with O(8.x) and P(9.0) (Exynos 7570 chipsets) software. The Trustonic Kinibi component allows arbitrary memory mapping. The Samsung ID is SVE-2019-16665 (June 2020).

An issue was discovered on Samsung mobile devices with P(9.0) software. One UI HOME logging can leak information. The Samsung ID is SVE-2019-16382 (June 2020).

An issue was discovered on Samsung mobile devices with P(9.0) and Q(10.0) software. Attackers can disable the SEAndroid protection mechanism in the RKP. The Samsung ID is SVE-2019-15998 (June 2020).

An issue was discovered on LG mobile devices with Android OS 7.2, 8.0, 8.1, 9, and 10 software. A crafted application can obtain control of device input via the window system service. The LG ID is LVE-SMP-170011 (May 2020).

An issue was discovered on LG mobile devices with Android OS 7.2, 8.0, 8.1, 9, and 10 software. Arbitrary code execution can occur via the bootloader because of an EL1/EL3 coldboot vulnerability involving raw_resources. The LG ID is LVE-SMP-200006 (May 2020).

An issue was discovered on Samsung mobile devices with P(9.0) and Q(10.0) (with TEEGRIS) software. Attackers can determine user credentials via a brute-force attack against the Gatekeeper trustlet. The Samsung ID is SVE-2020-16908 (May 2020).

An issue was discovered on Samsung mobile devices with O(8.X), P(9.0), and Q(10.0) software. The Quram image codec library allows attackers to overwrite memory and execute arbitrary code via crafted JPEG data that is mishandled during decoding. The Samsung ID is SVE-2020-16943 (May 2020).

An issue was discovered on Samsung mobile devices with Q(10.0) software. Attackers can bypass Factory Reset Protection (FRP) via SPEN. The Samsung ID is SVE-2020-17019 (May 2020).

An issue was discovered on Samsung mobile devices with P(9.0) (Exynos chipsets) software. The S.LSI Wi-Fi drivers have a buffer overflow. The Samsung ID is SVE-2020-16906 (May 2020).

An issue was discovered on Samsung mobile devices with Q(10.0) software. Attackers can bypass the locked-state protection mechanism and designate a different preferred SIM card. The Samsung ID is SVE-2020-16594 (May 2020).

An issue was discovered on Samsung mobile devices with Q(10.0) (Exynos980 9630 and Exynos990 9830 chipsets) software. The Bootloader has a heap-based buffer overflow because of the mishandling of specific commands. The Samsung IDs are SVE-2020-16981, SVE-2020-16991 (May 2020).

An issue was discovered on Samsung mobile devices with O(8.X), P(9.0), and Q(10.0) (Exynos chipsets) software. Attackers can bypass the Secure Bootloader protection mechanism via a heap-based buffer overflow to execute arbitrary code. The Samsung ID is SVE-2020-16712 (May 2020).

An issue was discovered on Samsung mobile devices with Q(10.0) software. Attackers can bypass the locked-state protection mechanism and access clipboard content via USSD. The Samsung ID is SVE-2019-16556 (May 2020).

An issue was discovered on LG mobile devices with Android OS 8.0, 8.1, 9.0, and 10.0 (MTK chipsets) software. The MTK kernel does not properly implement exception handling, allowing an attacker to gain privileges. The LG ID is LVE-SMP-200001 (February 2020).

An issue was discovered on LG mobile devices with Android OS 8.0, 8.1, 9, and 10 software. Attackers can bypass Factory Reset Protection (FRP). The LG ID is LVE-SMP-200004 (March 2020).

An issue was discovered on LG mobile devices with Android OS 7.2, 8.0, 8.1, 9, and 10 software. A stack-based buffer overflow in the logging tool could allow an attacker to gain privileges. The LG ID is LVE-SMP-200005 (April 2020).

OPPO Android Phone with MTK chipset and Android 8.1/9/10/11 versions have an information leak vulnerability. The “adb shell getprop ro.vendor.aee.enforcing” or “adb shell getprop ro.vendor.aee.enforcing” return no.

An issue was discovered on Samsung mobile devices with P(9.0) and Q(10.0) software. Notification exposure occurs in Lockdown mode because of the Edge Lighting application. The Samsung ID is SVE-2020-16680 (April 2020).

An issue was discovered on Samsung mobile devices with Q(10.0) software. Information about application preview (in the Secure Folder) leaks on a locked device. The Samsung ID is SVE-2019-16463 (April 2020).