Lines Matching full:fixes

27 # Bug Fixes and Other Changes
109 ## Bug Fixes and Other Changes
318 ## Bug Fixes and Other Changes
336 *   Fixes a code injection in `saved_model_cli` ([CVE-2022-29216](https://cve.mitre.org/cgi-bin/cve…
337 *   Fixes a missing validation which causes `TensorSummaryV2` to crash ([CVE-2022-29193](https://cv…
338 *   Fixes a missing validation which crashes `QuantizeAndDequantizeV4Grad` ([CVE-2022-29192](https:…
339 *   Fixes a missing validation which causes denial of service via `DeleteSessionTensor` ([CVE-2022-…
340 *   Fixes a missing validation which causes denial of service via `GetSessionTensor` ([CVE-2022-291…
341 *   Fixes a missing validation which causes denial of service via `StagePeek` ([CVE-2022-29195](htt…
342 *   Fixes a missing validation which causes denial of service via `UnsortedSegmentJoin` ([CVE-2022-…
343 *   Fixes a missing validation which causes denial of service via `LoadAndRemapMatrix` ([CVE-2022-2…
344 *   Fixes a missing validation which causes denial of service via `SparseTensorToCSRSparseMatrix` (…
345 *   Fixes a missing validation which causes denial of service via `LSTMBlockCell` ([CVE-2022-29200]…
346 *   Fixes a missing validation which causes denial of service via `Conv3DBackpropFilterV2` ([CVE-20…
347 *   Fixes a `CHECK` failure in depthwise ops via overflows ([CVE-2021-41197](https://cve.mitre.org/…
348 *   Fixes issues arising from undefined behavior stemming from users supplying invalid resource han…
349 *   Fixes a segfault due to missing support for quantized types ([CVE-2022-29205](https://cve.mitre…
350 *   Fixes a missing validation which results in undefined behavior in `SparseTensorDenseAdd` ([CVE-…
351 *   Fixes a missing validation which results in undefined behavior in `QuantizedConv2D` ([CVE-2022-…
352 *   Fixes an integer overflow in `SpaceToBatchND` ([CVE-2022-29203](https://cve.mitre.org/cgi-bin/c…
353 *   Fixes a segfault and OOB write due to incomplete validation in `EditDistance` ([CVE-2022-29208]…
354 *   Fixes a missing validation which causes denial of service via `Conv3DBackpropFilterV2` ([CVE-20…
355 *   Fixes a denial of service in `tf.ragged.constant` due to lack of validation ([CVE-2022-29202](h…
356 *   Fixes a segfault when `tf.histogram_fixed_width` is called with NaN values ([CVE-2022-29211](ht…
357 *   Fixes a core dump when loading TFLite models with quantization ([CVE-2022-29212](https://cve.mi…
358 *   Fixes crashes stemming from incomplete validation in signal ops ([CVE-2022-29213](https://cve.m…
359 *   Fixes a type confusion leading to `CHECK`-failure based denial of service ([CVE-2022-29209](htt…
360 *   Fixes a heap buffer overflow due to incorrect hash function ([CVE-2022-29210](https://cve.mitre…
372 This releases introduces several vulnerability fixes:
374 *   Fixes a code injection in `saved_model_cli` ([CVE-2022-29216](https://cve.mitre.org/cgi-bin/cve…
375 *   Fixes a missing validation which causes `TensorSummaryV2` to crash ([CVE-2022-29193](https://cv…
376 *   Fixes a missing validation which crashes `QuantizeAndDequantizeV4Grad` ([CVE-2022-29192](https:…
377 *   Fixes a missing validation which causes denial of service via `DeleteSessionTensor` ([CVE-2022-…
378 *   Fixes a missing validation which causes denial of service via `GetSessionTensor` ([CVE-2022-291…
379 *   Fixes a missing validation which causes denial of service via `StagePeek` ([CVE-2022-29195](htt…
380 *   Fixes a missing validation which causes denial of service via `UnsortedSegmentJoin` ([CVE-2022-…
381 *   Fixes a missing validation which causes denial of service via `LoadAndRemapMatrix` ([CVE-2022-2…
382 *   Fixes a missing validation which causes denial of service via `SparseTensorToCSRSparseMatrix` (…
383 *   Fixes a missing validation which causes denial of service via `LSTMBlockCell` ([CVE-2022-29200]…
384 *   Fixes a missing validation which causes denial of service via `Conv3DBackpropFilterV2` ([CVE-20…
385 *   Fixes a `CHECK` failure in depthwise ops via overflows ([CVE-2021-41197](https://cve.mitre.org/…
386 *   Fixes issues arising from undefined behavior stemming from users supplying invalid resource han…
387 *   Fixes a segfault due to missing support for quantized types ([CVE-2022-29205](https://cve.mitre…
388 *   Fixes a missing validation which results in undefined behavior in `SparseTensorDenseAdd` ([CVE-…
389 *   Fixes a missing validation which results in undefined behavior in `QuantizedConv2D` ([CVE-2022-…
390 *   Fixes an integer overflow in `SpaceToBatchND` ([CVE-2022-29203](https://cve.mitre.org/cgi-bin/c…
391 *   Fixes a segfault and OOB write due to incomplete validation in `EditDistance` ([CVE-2022-29208]…
392 *   Fixes a missing validation which causes denial of service via `Conv3DBackpropFilterV2` ([CVE-20…
393 *   Fixes a denial of service in `tf.ragged.constant` due to lack of validation ([CVE-2022-29202](h…
394 *   Fixes a segfault when `tf.histogram_fixed_width` is called with NaN values ([CVE-2022-29211](ht…
395 *   Fixes a core dump when loading TFLite models with quantization ([CVE-2022-29212](https://cve.mi…
396 *   Fixes crashes stemming from incomplete validation in signal ops ([CVE-2022-29213](https://cve.m…
397 *   Fixes a type confusion leading to `CHECK`-failure based denial of service ([CVE-2022-29209](htt…
398 *   Fixes a heap buffer overflow due to incorrect hash function ([CVE-2022-29210](https://cve.mitre…
403 This releases introduces several vulnerability fixes:
405 *   Fixes a code injection in `saved_model_cli` ([CVE-2022-29216](https://cve.mitre.org/cgi-bin/cve…
406 *   Fixes a missing validation which causes `TensorSummaryV2` to crash ([CVE-2022-29193](https://cv…
407 *   Fixes a missing validation which crashes `QuantizeAndDequantizeV4Grad` ([CVE-2022-29192](https:…
408 *   Fixes a missing validation which causes denial of service via `DeleteSessionTensor` ([CVE-2022-…
409 *   Fixes a missing validation which causes denial of service via `GetSessionTensor` ([CVE-2022-291…
410 *   Fixes a missing validation which causes denial of service via `StagePeek` ([CVE-2022-29195](htt…
411 *   Fixes a missing validation which causes denial of service via `UnsortedSegmentJoin` ([CVE-2022-…
412 *   Fixes a missing validation which causes denial of service via `LoadAndRemapMatrix` ([CVE-2022-2…
413 *   Fixes a missing validation which causes denial of service via `SparseTensorToCSRSparseMatrix` (…
414 *   Fixes a missing validation which causes denial of service via `LSTMBlockCell` ([CVE-2022-29200]…
415 *   Fixes a missing validation which causes denial of service via `Conv3DBackpropFilterV2` ([CVE-20…
416 *   Fixes a `CHECK` failure in depthwise ops via overflows ([CVE-2021-41197](https://cve.mitre.org/…
417 *   Fixes issues arising from undefined behavior stemming from users supplying invalid resource han…
418 *   Fixes a segfault due to missing support for quantized types ([CVE-2022-29205](https://cve.mitre…
419 *   Fixes a missing validation which results in undefined behavior in `SparseTensorDenseAdd` ([CVE-…
420 *   Fixes a missing validation which results in undefined behavior in `QuantizedConv2D` ([CVE-2022-…
421 *   Fixes an integer overflow in `SpaceToBatchND` ([CVE-2022-29203](https://cve.mitre.org/cgi-bin/c…
422 *   Fixes a segfault and OOB write due to incomplete validation in `EditDistance` ([CVE-2022-29208]…
423 *   Fixes a missing validation which causes denial of service via `Conv3DBackpropFilterV2` ([CVE-20…
424 *   Fixes a denial of service in `tf.ragged.constant` due to lack of validation ([CVE-2022-29202](h…
425 *   Fixes a segfault when `tf.histogram_fixed_width` is called with NaN values ([CVE-2022-29211](ht…
426 *   Fixes a core dump when loading TFLite models with quantization ([CVE-2022-29212](https://cve.mi…
427 *   Fixes crashes stemming from incomplete validation in signal ops ([CVE-2022-29213](https://cve.m…
428 *   Fixes a type confusion leading to `CHECK`-failure based denial of service ([CVE-2022-29209](htt…
434 This releases introduces several vulnerability fixes:
436 *   Fixes a code injection in `saved_model_cli` ([CVE-2022-29216](https://cve.mitre.org/cgi-bin/cve…
437 *   Fixes a missing validation which causes `TensorSummaryV2` to crash ([CVE-2022-29193](https://cv…
438 *   Fixes a missing validation which crashes `QuantizeAndDequantizeV4Grad` ([CVE-2022-29192](https:…
439 *   Fixes a missing validation which causes denial of service via `DeleteSessionTensor` ([CVE-2022-…
440 *   Fixes a missing validation which causes denial of service via `GetSessionTensor` ([CVE-2022-291…
441 *   Fixes a missing validation which causes denial of service via `StagePeek` ([CVE-2022-29195](htt…
442 *   Fixes a missing validation which causes denial of service via `UnsortedSegmentJoin` ([CVE-2022-…
443 *   Fixes a missing validation which causes denial of service via `LoadAndRemapMatrix` ([CVE-2022-2…
444 *   Fixes a missing validation which causes denial of service via `SparseTensorToCSRSparseMatrix` (…
445 *   Fixes a missing validation which causes denial of service via `LSTMBlockCell` ([CVE-2022-29200]…
446 *   Fixes a missing validation which causes denial of service via `Conv3DBackpropFilterV2` ([CVE-20…
447 *   Fixes a `CHECK` failure in depthwise ops via overflows ([CVE-2021-41197](https://cve.mitre.org/…
448 *   Fixes issues arising from undefined behavior stemming from users supplying invalid resource han…
449 *   Fixes a segfault due to missing support for quantized types ([CVE-2022-29205](https://cve.mitre…
450 *   Fixes a missing validation which results in undefined behavior in `SparseTensorDenseAdd` ([CVE-…
451 *   Fixes a missing validation which results in undefined behavior in `QuantizedConv2D` ([CVE-2022-…
452 *   Fixes an integer overflow in `SpaceToBatchND` ([CVE-2022-29203](https://cve.mitre.org/cgi-bin/c…
453 *   Fixes a segfault and OOB write due to incomplete validation in `EditDistance` ([CVE-2022-29208]…
454 *   Fixes a missing validation which causes denial of service via `Conv3DBackpropFilterV2` ([CVE-20…
455 *   Fixes a denial of service in `tf.ragged.constant` due to lack of validation ([CVE-2022-29202](h…
456 *   Fixes a segfault when `tf.histogram_fixed_width` is called with NaN values ([CVE-2022-29211](ht…
457 *   Fixes a core dump when loading TFLite models with quantization ([CVE-2022-29212](https://cve.mi…
458 *   Fixes crashes stemming from incomplete validation in signal ops ([CVE-2022-29213](https://cve.m…
459 *   Fixes a type confusion leading to `CHECK`-failure based denial of service ([CVE-2022-29209](htt…
509     deprecated. The "Bug Fixes and Other Changes" section lists more
517 ## Bug Fixes and Other Changes
663 *   Fixes a floating point division by 0 when executing convolution operators
665 *   Fixes a heap OOB read in shape inference for `ReverseSequence`
667 *   Fixes a heap OOB access in `Dequantize`
669 *   Fixes an integer overflow in shape inference for `Dequantize`
671 *   Fixes a heap OOB access in `FractionalAvgPoolGrad`
673 *   Fixes an overflow and divide by zero in `UnravelIndex`
675 *   Fixes a type confusion in shape inference for `ConcatV2`
677 *   Fixes an OOM in `ThreadPoolHandle`
679 *   Fixes an OOM due to integer overflow in `StringNGrams`
681 *   Fixes more issues caused by incomplete validation in boosted trees code
683 *   Fixes an integer overflows in most sparse component-wise ops
685 *   Fixes an integer overflows in `AddManySparseToTensorsMap`
687 *   Fixes a number of `CHECK`-failures in `MapStage`
689 *   Fixes a division by zero in `FractionalMaxPool`
691 *   Fixes a number of `CHECK`-fails when building invalid/overflowing tensor
694 *   Fixes an undefined behavior in `SparseTensorSliceDataset`
696 *   Fixes an assertion failure based denial of service via faulty bin count
699 *   Fixes a reference binding to null pointer in `QuantizedMaxPool`
701 *   Fixes an integer overflow leading to crash in `SparseCountSparseOutput`
703 *   Fixes a heap overflow in `SparseCountSparseOutput`
705 *   Fixes an FPE in `BiasAndClamp` in TFLite
707 *   Fixes an FPE in depthwise convolutions in TFLite
709 *   Fixes an integer overflow in TFLite array creation
711 *   Fixes an integer overflow in TFLite
713 *   Fixes a dangerous OOB write in TFLite
715 *   Fixes a vulnerability leading to read and write outside of bounds in TFLite
717 *   Fixes a set of vulnerabilities caused by using insecure temporary files
719 *   Fixes an integer overflow in Range resulting in undefined behavior and OOM
721 *   Fixes a vulnerability where missing validation causes `tf.sparse.split` to
724 *   Fixes a `CHECK`-fail when decoding resource handles from proto
726 *   Fixes a `CHECK`-fail with repeated `AttrDef`
728 *   Fixes a heap OOB write in Grappler
730 *   Fixes a `CHECK`-fail when decoding invalid tensors from proto
732 *   Fixes a null-dereference when specializing tensor type
734 *   Fixes a crash when type cannot be specialized
736 *   Fixes a heap OOB read/write in `SpecializeType`
738 *   Fixes an unitialized variable access in `AssignOp`
740 *   Fixes an integer overflow in `OpLevelCostEstimator::CalculateTensorSize`
742 *   Fixes an integer overflow in `OpLevelCostEstimator::CalculateOutputSize`
744 *   Fixes a null dereference in `GetInitOp`
746 *   Fixes a memory leak when a graph node is invalid
748 *   Fixes an abort caused by allocating a vector that is too large
750 *   Fixes multiple `CHECK`-failures during Grappler's `IsSimplifiableReshape`
752 *   Fixes multiple `CHECK`-failures during Grappler's `SafeToRemoveIdentity`
754 *   Fixes multiple `CHECK`-failures in `TensorByteSize`
756 *   Fixes multiple `CHECK`-failures in binary ops due to type confusion
758 *   Fixes a use after free in `DecodePng` kernel
760 *   Fixes a memory leak in decoding PNG images
762 *   Fixes multiple `CHECK`-fails in `function.cc`
764 *   Fixes multiple `CHECK`-fails due to attempting to build a reference tensor
766 *   Fixes an integer overflow in Grappler cost estimation of crop and resize
769 *   Fixes a null pointer dereference in Grappler's `IsConstant`
771 *   Fixes a `CHECK` failure in constant folding
773 *   Fixes a stack overflow due to self-recursive function in `GraphDef`
775 *   Fixes a heap OOB access in `RunForwardTypeInference`
777 *   Fixes a crash due to erroneous `StatusOr`
779 *   Fixes multiple crashes and heap OOB accesses in TFG dialect (MLIR)
781 *   Fixes a segfault in `simplifyBroadcast` (MLIR)
783 *   Fixes a null pointer dereference in `BuildXlaCompilationCache` (XLA)
814 This releases introduces several vulnerability fixes:
816 *   Fixes a floating point division by 0 when executing convolution operators
818 *   Fixes a heap OOB read in shape inference for `ReverseSequence`
820 *   Fixes a heap OOB access in `Dequantize`
822 *   Fixes an integer overflow in shape inference for `Dequantize`
824 *   Fixes a heap OOB access in `FractionalAvgPoolGrad`
826 *   Fixes an overflow and divide by zero in `UnravelIndex`
828 *   Fixes a type confusion in shape inference for `ConcatV2`
830 *   Fixes an OOM in `ThreadPoolHandle`
832 *   Fixes an OOM due to integer overflow in `StringNGrams`
834 *   Fixes more issues caused by incomplete validation in boosted trees code
836 *   Fixes an integer overflows in most sparse component-wise ops
838 *   Fixes an integer overflows in `AddManySparseToTensorsMap`
840 *   Fixes a number of `CHECK`-failures in `MapStage`
842 *   Fixes a division by zero in `FractionalMaxPool`
844 *   Fixes a number of `CHECK`-fails when building invalid/overflowing tensor
847 *   Fixes an undefined behavior in `SparseTensorSliceDataset`
849 *   Fixes an assertion failure based denial of service via faulty bin count
852 *   Fixes a reference binding to null pointer in `QuantizedMaxPool`
854 *   Fixes an integer overflow leading to crash in `SparseCountSparseOutput`
856 *   Fixes a heap overflow in `SparseCountSparseOutput`
858 *   Fixes an FPE in `BiasAndClamp` in TFLite
860 *   Fixes an FPE in depthwise convolutions in TFLite
862 *   Fixes an integer overflow in TFLite array creation
864 *   Fixes an integer overflow in TFLite
866 *   Fixes a dangerous OOB write in TFLite
868 *   Fixes a vulnerability leading to read and write outside of bounds in TFLite
870 *   Fixes a set of vulnerabilities caused by using insecure temporary files
872 *   Fixes an integer overflow in Range resulting in undefined behavior and OOM
874 *   Fixes a vulnerability where missing validation causes `tf.sparse.split` to
877 *   Fixes a `CHECK`-fail when decoding resource handles from proto
879 *   Fixes a `CHECK`-fail with repeated `AttrDef`
881 *   Fixes a heap OOB write in Grappler
883 *   Fixes a `CHECK`-fail when decoding invalid tensors from proto
885 *   Fixes a null-dereference when specializing tensor type
887 *   Fixes a crash when type cannot be specialized
889 *   Fixes a heap OOB read/write in `SpecializeType`
891 *   Fixes an uninitialized variable access in `AssignOp`
893 *   Fixes an integer overflow in `OpLevelCostEstimator::CalculateTensorSize`
895 *   Fixes an integer overflow in `OpLevelCostEstimator::CalculateOutputSize`
897 *   Fixes a null dereference in `GetInitOp`
899 *   Fixes a memory leak when a graph node is invalid
901 *   Fixes an abort caused by allocating a vector that is too large
903 *   Fixes multiple `CHECK`-failures during Grappler's `IsSimplifiableReshape`
905 *   Fixes multiple `CHECK`-failures during Grappler's `SafeToRemoveIdentity`
907 *   Fixes multiple `CHECK`-failures in `TensorByteSize`
909 *   Fixes multiple `CHECK`-failures in binary ops due to type confusion
911 *   Fixes a use after free in `DecodePng` kernel
913 *   Fixes a memory leak in decoding PNG images
915 *   Fixes multiple `CHECK`-fails in `function.cc`
917 *   Fixes multiple `CHECK`-fails due to attempting to build a reference tensor
919 *   Fixes an integer overflow in Grappler cost estimation of crop and resize
922 *   Fixes a null pointer dereference in Grappler's `IsConstant`
924 *   Fixes a `CHECK` failure in constant folding
926 *   Fixes a stack overflow due to self-recursive function in `GraphDef`
928 *   Fixes a crash due to erroneous `StatusOr`
930 *   Fixes multiple crashes and heap OOB accesses in TFG dialect (MLIR)
932 *   Fixes a null pointer dereference in `BuildXlaCompilationCache` (XLA)
939 This releases introduces several vulnerability fixes:
941 *   Fixes a floating point division by 0 when executing convolution operators
943 *   Fixes a heap OOB read in shape inference for `ReverseSequence`
945 *   Fixes a heap OOB access in `Dequantize`
947 *   Fixes an integer overflow in shape inference for `Dequantize`
949 *   Fixes a heap OOB access in `FractionalAvgPoolGrad`
951 *   Fixes an overflow and divide by zero in `UnravelIndex`
953 *   Fixes a type confusion in shape inference for `ConcatV2`
955 *   Fixes an OOM in `ThreadPoolHandle`
957 *   Fixes an OOM due to integer overflow in `StringNGrams`
959 *   Fixes more issues caused by incomplete validation in boosted trees code
961 *   Fixes an integer overflows in most sparse component-wise ops
963 *   Fixes an integer overflows in `AddManySparseToTensorsMap`
965 *   Fixes a number of `CHECK`-failures in `MapStage`
967 *   Fixes a division by zero in `FractionalMaxPool`
969 *   Fixes a number of `CHECK`-fails when building invalid/overflowing tensor
972 *   Fixes an undefined behavior in `SparseTensorSliceDataset`
974 *   Fixes an assertion failure based denial of service via faulty bin count
977 *   Fixes a reference binding to null pointer in `QuantizedMaxPool`
979 *   Fixes an integer overflow leading to crash in `SparseCountSparseOutput`
981 *   Fixes a heap overflow in `SparseCountSparseOutput`
983 *   Fixes an FPE in `BiasAndClamp` in TFLite
985 *   Fixes an FPE in depthwise convolutions in TFLite
987 *   Fixes an integer overflow in TFLite array creation
989 *   Fixes an integer overflow in TFLite
991 *   Fixes a dangerous OOB write in TFLite
993 *   Fixes a vulnerability leading to read and write outside of bounds in TFLite
995 *   Fixes a set of vulnerabilities caused by using insecure temporary files
997 *   Fixes an integer overflow in Range resulting in undefined behavior and OOM
999 *   Fixes a vulnerability where missing validation causes `tf.sparse.split` to
1002 *   Fixes a `CHECK`-fail when decoding resource handles from proto
1004 *   Fixes a `CHECK`-fail with repeated `AttrDef`
1006 *   Fixes a heap OOB write in Grappler
1008 *   Fixes a `CHECK`-fail when decoding invalid tensors from proto
1010 *   Fixes a null-dereference when specializing tensor type
1012 *   Fixes a crash when type cannot be specialized
1014 *   Fixes a heap OOB read/write in `SpecializeType`
1016 *   Fixes an unitialized variable access in `AssignOp`
1018 *   Fixes an integer overflow in `OpLevelCostEstimator::CalculateTensorSize`
1020 *   Fixes an integer overflow in `OpLevelCostEstimator::CalculateOutputSize`
1022 *   Fixes a null dereference in `GetInitOp`
1024 *   Fixes a memory leak when a graph node is invalid
1026 *   Fixes an abort caused by allocating a vector that is too large
1028 *   Fixes multiple `CHECK`-failures during Grappler's `IsSimplifiableReshape`
1030 *   Fixes multiple `CHECK`-failures during Grappler's `SafeToRemoveIdentity`
1032 *   Fixes multiple `CHECK`-failures in `TensorByteSize`
1034 *   Fixes multiple `CHECK`-failures in binary ops due to type confusion
1036 *   Fixes a use after free in `DecodePng` kernel
1038 *   Fixes a memory leak in decoding PNG images
1040 *   Fixes multiple `CHECK`-fails in `function.cc`
1042 *   Fixes multiple `CHECK`-fails due to attempting to build a reference tensor
1044 *   Fixes an integer overflow in Grappler cost estimation of crop and resize
1047 *   Fixes a null pointer dereference in Grappler's `IsConstant`
1049 *   Fixes a `CHECK` failure in constant folding
1051 *   Fixes a stack overflow due to self-recursive function in `GraphDef`
1053 *   Fixes a null pointer dereference in `BuildXlaCompilationCache` (XLA)
1060 This releases introduces several vulnerability fixes:
1062 *   Fixes a floating point division by 0 when executing convolution operators
1064 *   Fixes a heap OOB read in shape inference for `ReverseSequence`
1066 *   Fixes a heap OOB access in `Dequantize`
1068 *   Fixes an integer overflow in shape inference for `Dequantize`
1070 *   Fixes a heap OOB access in `FractionalAvgPoolGrad`
1072 *   Fixes an overflow and divide by zero in `UnravelIndex`
1074 *   Fixes a type confusion in shape inference for `ConcatV2`
1076 *   Fixes an OOM in `ThreadPoolHandle`
1078 *   Fixes an OOM due to integer overflow in `StringNGrams`
1080 *   Fixes more issues caused by incomplete validation in boosted trees code
1082 *   Fixes an integer overflows in most sparse component-wise ops
1084 *   Fixes an integer overflows in `AddManySparseToTensorsMap`
1086 *   Fixes a number of `CHECK`-failures in `MapStage`
1088 *   Fixes a division by zero in `FractionalMaxPool`
1090 *   Fixes a number of `CHECK`-fails when building invalid/overflowing tensor
1093 *   Fixes an undefined behavior in `SparseTensorSliceDataset`
1095 *   Fixes an assertion failure based denial of service via faulty bin count
1098 *   Fixes a reference binding to null pointer in `QuantizedMaxPool`
1100 *   Fixes an integer overflow leading to crash in `SparseCountSparseOutput`
1102 *   Fixes a heap overflow in `SparseCountSparseOutput`
1104 *   Fixes an FPE in `BiasAndClamp` in TFLite
1106 *   Fixes an FPE in depthwise convolutions in TFLite
1108 *   Fixes an integer overflow in TFLite array creation
1110 *   Fixes an integer overflow in TFLite
1112 *   Fixes a dangerous OOB write in TFLite
1114 *   Fixes a vulnerability leading to read and write outside of bounds in TFLite
1116 *   Fixes a set of vulnerabilities caused by using insecure temporary files
1118 *   Fixes an integer overflow in Range resulting in undefined behavior and OOM
1120 *   Fixes a vulnerability where missing validation causes `tf.sparse.split` to
1123 *   Fixes a `CHECK`-fail when decoding resource handles from proto
1125 *   Fixes a `CHECK`-fail with repeated `AttrDef`
1127 *   Fixes a heap OOB write in Grappler
1129 *   Fixes a `CHECK`-fail when decoding invalid tensors from proto
1131 *   Fixes an unitialized variable access in `AssignOp`
1133 *   Fixes an integer overflow in `OpLevelCostEstimator::CalculateTensorSize`
1135 *   Fixes an integer overflow in `OpLevelCostEstimator::CalculateOutputSize`
1137 *   Fixes a null dereference in `GetInitOp`
1139 *   Fixes a memory leak when a graph node is invalid
1141 *   Fixes an abort caused by allocating a vector that is too large
1143 *   Fixes multiple `CHECK`-failures during Grappler's `IsSimplifiableReshape`
1145 *   Fixes multiple `CHECK`-failures during Grappler's `SafeToRemoveIdentity`
1147 *   Fixes multiple `CHECK`-failures in `TensorByteSize`
1149 *   Fixes multiple `CHECK`-failures in binary ops due to type confusion
1151 *   Fixes a use after free in `DecodePng` kernel
1153 *   Fixes a memory leak in decoding PNG images
1155 *   Fixes multiple `CHECK`-fails in `function.cc`
1157 *   Fixes multiple `CHECK`-fails due to attempting to build a reference tensor
1159 *   Fixes an integer overflow in Grappler cost estimation of crop and resize
1162 *   Fixes a null pointer dereference in Grappler's `IsConstant`
1164 *   Fixes a `CHECK` failure in constant folding
1166 *   Fixes a stack overflow due to self-recursive function in `GraphDef`
1379 ## Bug Fixes and Other Changes
1488 *   Fixes a code injection issue in `saved_model_cli`
1490 *   Fixes a vulnerability due to use of uninitialized value in Tensorflow
1492 *   Fixes a heap OOB in `FusedBatchNorm` kernels
1494 *   Fixes an arbitrary memory read in `ImmutableConst`
1496 *   Fixes a heap OOB in `SparseBinCount`
1498 *   Fixes a heap OOB in `SparseFillEmptyRows`
1500 *   Fixes a segfault due to negative splits in `SplitV`
1502 *   Fixes segfaults and vulnerabilities caused by accesses to invalid memory
1505 *   Fixes a null pointer exception when `Exit` node is not preceded by `Enter`
1508 *   Fixes an integer division by 0 in `tf.raw_ops.AllToAll`
1510 *   Fixes a use after free and a memory leak in `CollectiveReduceV2`
1512 *   Fixes an undefined behavior via `nullptr` reference binding in sparse matrix
1515 *   Fixes a heap buffer overflow in `Transpose`
1519 *   Fixes a null pointer exception in `DeserializeSparse`
1521 *   Fixes an undefined behavior arising from reference binding to `nullptr` in
1524 *   Fixes a heap OOB read in `tf.ragged.cross`
1526 *   Fixes a heap OOB in shape inference for `QuantizeV2`
1528 *   Fixes a heap OOB read in all `tf.raw_ops.QuantizeAndDequantizeV*` ops
1530 *   Fixes an FPE in `ParallelConcat`
1532 *   Fixes FPE issues in convolutions with zero size filters
1534 *   Fixes a heap OOB read in `tf.raw_ops.SparseCountSparseOutput`
1536 *   Fixes vulnerabilities caused by incomplete validation in boosted trees code
1538 *   Fixes vulnerabilities caused by incomplete validation of shapes in multiple
1541 *   Fixes a segfault produced while copying constant resource tensor
1543 *   Fixes a vulnerability caused by unitialized access in
1546 *   Fixes several vulnerabilities and segfaults caused by missing validation
1549 *   Fixes an overflow producing a crash in `tf.range`
1551 *   Fixes an overflow producing a crash in `tf.image.resize` when size is large
1553 *   Fixes an overflow producing a crash in `tf.tile` when tiling tensor is large
1555 *   Fixes a vulnerability produced due to incomplete validation in
1558 *   Fixes multiple crashes due to overflow and `CHECK`-fail in ops with large
1561 *   Fixes a crash in `max_pool3d` when size argument is 0 or negative
1563 *   Fixes a crash in `tf.math.segment_*` operations
1601 Fixes an issue where `keras`, `tensorflow_estimator` and `tensorboard` were
1606 This release introduces several vulnerability fixes:
1608 *   Fixes a code injection issue in `saved_model_cli`
1610 *   Fixes a vulnerability due to use of uninitialized value in Tensorflow
1612 *   Fixes a heap OOB in `FusedBatchNorm` kernels
1614 *   Fixes an arbitrary memory read in `ImmutableConst`
1616 *   Fixes a heap OOB in `SparseBinCount`
1618 *   Fixes a heap OOB in `SparseFillEmptyRows`
1620 *   Fixes a segfault due to negative splits in `SplitV`
1622 *   Fixes segfaults and vulnerabilities caused by accesses to invalid memory
1625 *   Fixes a null pointer exception when `Exit` node is not preceded by `Enter`
1628 *   Fixes an integer division by 0 in `tf.raw_ops.AllToAll`
1630 *   Fixes a use after free and a memory leak in `CollectiveReduceV2`
1632 *   Fixes an undefined behavior via `nullptr` reference binding in sparse matrix
1635 *   Fixes a heap buffer overflow in `Transpose`
1639 *   Fixes a null pointer exception in `DeserializeSparse`
1641 *   Fixes an undefined behavior arising from reference binding to `nullptr` in
1644 *   Fixes a heap OOB read in `tf.ragged.cross`
1646 *   Fixes a heap OOB in shape inference for `QuantizeV2`
1648 *   Fixes a heap OOB read in all `tf.raw_ops.QuantizeAndDequantizeV*` ops
1650 *   Fixes an FPE in `ParallelConcat`
1652 *   Fixes FPE issues in convolutions with zero size filters
1654 *   Fixes a heap OOB read in `tf.raw_ops.SparseCountSparseOutput`
1656 *   Fixes vulnerabilities caused by incomplete validation in boosted trees code
1658 *   Fixes vulnerabilities caused by incomplete validation of shapes in multiple
1661 *   Fixes a segfault produced while copying constant resource tensor
1663 *   Fixes a vulnerability caused by unitialized access in
1666 *   Fixes several vulnerabilities and segfaults caused by missing validation
1669 *   Fixes an overflow producing a crash in `tf.range`
1671 *   Fixes an overflow producing a crash in `tf.image.resize` when size is large
1673 *   Fixes an overflow producing a crash in `tf.tile` when tiling tensor is large
1675 *   Fixes a vulnerability produced due to incomplete validation in
1678 *   Fixes multiple crashes due to overflow and `CHECK`-fail in ops with large
1681 *   Fixes a crash in `max_pool3d` when size argument is 0 or negative
1683 *   Fixes a crash in `tf.math.segment_*` operations
1825 ## Bug Fixes and Other Changes
1954 *   Fixes a heap out of bounds access in sparse reduction operations
1956 *   Fixes a floating point exception in `SparseDenseCwiseDiv`
1958 *   Fixes a null pointer dereference in `CompressElement`
1960 *   Fixes a null pointer dereference in `RaggedTensorToTensor`
1962 *   Fixes a null pointer dereference and a heap OOB read arising from operations
1965 *   Fixes an integer division by 0 in sparse reshaping
1967 *   Fixes a division by 0 in `ResourceScatterDiv`
1969 *   Fixes a heap OOB in `RaggedGather`
1971 *   Fixes a `std::abort` raised from `TensorListReserve`
1973 *   Fixes a null pointer dereference in `MatrixDiagPartOp`
1975 *   Fixes an integer overflow due to conversion to unsigned
1977 *   Fixes a bad allocation error in `StringNGrams` caused by integer conversion
1979 *   Fixes a null pointer dereference in `SparseTensorSliceDataset`
1981 *   Fixes an incorrect validation of `SaveV2` inputs
1983 *   Fixes a null pointer dereference in `UncompressElement`
1985 *   Fixes a segfault and a heap buffer overflow in
1988 *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
1990 *   Fixes a use after free in boosted trees creation
1992 *   Fixes a division by 0 in `ResourceGather`
1994 *   Fixes a heap OOB and a `CHECK` fail in `ResourceGather`
1996 *   Fixes a heap OOB in `ResourceScatterUpdate`
1998 *   Fixes an undefined behavior arising from reference binding to nullptr in
2001 *   Fixes an undefined behavior arising from reference binding to nullptr in
2004 *   Fixes an undefined behavior arising from reference binding to nullptr in
2007 *   Fixes an undefined behavior arising from reference binding to nullptr and
2010 *   Fixes a division by 0 in inplace operations
2012 *   Fixes a crash caused by integer conversion to unsigned
2014 *   Fixes an undefined behavior arising from reference binding to nullptr in
2017 *   Fixes a heap OOB in boosted trees
2019 *   Fixes vulnerabilities arising from incomplete validation in `QuantizeV2`
2021 *   Fixes vulnerabilities arising from incomplete validation in MKL
2024 *   Fixes an undefined behavior arising from reference binding to nullptr in
2027 *   Fixes an undefined behavior arising from reference binding to nullptr in
2030 *   Fixes an FPE in `tf.raw_ops.UnravelIndex`
2032 *   Fixes a crash in NMS ops caused by integer conversion to unsigned
2034 *   Fixes a heap OOB in `UpperBound` and `LowerBound`
2036 *   Fixes an undefined behavior arising from reference binding to nullptr in map
2039 *   Fixes a heap OOB in `SdcaOptimizerV2`
2041 *   Fixes a `CHECK`-fail in `MapStage`
2043 *   Fixes a vulnerability arising from incomplete validation in `MaxPoolGrad`
2045 *   Fixes an undefined behavior arising from reference binding to nullptr in
2048 *   Fixes a division by 0 in most convolution operators
2050 *   Fixes vulnerabilities arising from missing validation in shape inference for
2053 *   Fixes an arbitrary code execution due to YAML deserialization
2055 *   Fixes a heap OOB in nested `tf.map_fn` with `RaggedTensor`s
2057 *   Fixes a division by zero in TFLite
2059 *   Fixes an NPE in TFLite
2061 *   Fixes a vulnerability arising from use of unitialized value in TFLite
2063 *   Fixes an FPE in TFLite division operations
2065 *   Fixes an FPE in TFLite pooling operations
2067 *   Fixes an infinite loop in TFLite
2069 *   Fixes a heap OOB in TFLite
2071 *   Fixes a heap OOB in TFLite's `Gather*` implementations
2073 *   Fixes an undefined behavior arising from null pointer dereference in TFLite
2075 *   Fixes an undefined behavior arising from null pointer dereference in TFLite
2078 *   Fixes a FPE in LSH in TFLite
2080 *   Fixes a segfault on strings tensors with mismatched dimensions, arising in
2083 *   Fixes a use after free and a potential segfault in shape inference functions
2126 This release introduces several vulnerability fixes:
2128 *   Fixes a code injection issue in `saved_model_cli`
2130 *   Fixes a vulnerability due to use of uninitialized value in Tensorflow
2132 *   Fixes a heap OOB in `FusedBatchNorm` kernels
2134 *   Fixes an arbitrary memory read in `ImmutableConst`
2136 *   Fixes a heap OOB in `SparseBinCount`
2138 *   Fixes a heap OOB in `SparseFillEmptyRows`
2140 *   Fixes a segfault due to negative splits in `SplitV`
2142 *   Fixes segfaults and vulnerabilities caused by accesses to invalid memory
2145 *   Fixes a null pointer exception when `Exit` node is not preceded by `Enter`
2148 *   Fixes an integer division by 0 in `tf.raw_ops.AllToAll`
2150 *   Fixes an undefined behavior via `nullptr` reference binding in sparse matrix
2153 *   Fixes a heap buffer overflow in `Transpose`
2157 *   Fixes a null pointer exception in `DeserializeSparse`
2159 *   Fixes an undefined behavior arising from reference binding to `nullptr` in
2162 *   Fixes a heap OOB read in `tf.ragged.cross`
2164 *   Fixes a heap OOB read in all `tf.raw_ops.QuantizeAndDequantizeV*` ops
2166 *   Fixes an FPE in `ParallelConcat`
2168 *   Fixes FPE issues in convolutions with zero size filters
2170 *   Fixes a heap OOB read in `tf.raw_ops.SparseCountSparseOutput`
2172 *   Fixes vulnerabilities caused by incomplete validation in boosted trees code
2174 *   Fixes vulnerabilities caused by incomplete validation of shapes in multiple
2177 *   Fixes a segfault produced while copying constant resource tensor
2179 *   Fixes a vulnerability caused by unitialized access in
2182 *   Fixes several vulnerabilities and segfaults caused by missing validation
2185 *   Fixes an overflow producing a crash in `tf.range`
2187 *   Fixes an overflow producing a crash in `tf.image.resize` when size is large
2189 *   Fixes an overflow producing a crash in `tf.tile` when tiling tensor is large
2191 *   Fixes a vulnerability produced due to incomplete validation in
2194 *   Fixes multiple crashes due to overflow and `CHECK`-fail in ops with large
2197 *   Fixes a crash in `max_pool3d` when size argument is 0 or negative
2199 *   Fixes a crash in `tf.math.segment_*` operations
2211 This release introduces several vulnerability fixes:
2213 *   Fixes a heap out of bounds access in sparse reduction operations
2215 *   Fixes a floating point exception in `SparseDenseCwiseDiv`
2217 *   Fixes a null pointer dereference in `CompressElement`
2219 *   Fixes a null pointer dereference in `RaggedTensorToTensor`
2221 *   Fixes a null pointer dereference and a heap OOB read arising from operations
2224 *   Fixes an integer division by 0 in sparse reshaping
2226 *   Fixes a division by 0 in `ResourceScatterDiv`
2228 *   Fixes a heap OOB in `RaggedGather`
2230 *   Fixes a `std::abort` raised from `TensorListReserve`
2232 *   Fixes a null pointer dereference in `MatrixDiagPartOp`
2234 *   Fixes an integer overflow due to conversion to unsigned
2236 *   Fixes a bad allocation error in `StringNGrams` caused by integer conversion
2238 *   Fixes a null pointer dereference in `SparseTensorSliceDataset`
2240 *   Fixes an incorrect validation of `SaveV2` inputs
2242 *   Fixes a null pointer dereference in `UncompressElement`
2244 *   Fixes a segfault and a heap buffer overflow in
2247 *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
2249 *   Fixes a use after free in boosted trees creation
2251 *   Fixes a division by 0 in `ResourceGather`
2253 *   Fixes a heap OOB and a `CHECK` fail in `ResourceGather`
2255 *   Fixes a heap OOB in `ResourceScatterUpdate`
2257 *   Fixes an undefined behavior arising from reference binding to nullptr in
2260 *   Fixes an undefined behavior arising from reference binding to nullptr in
2263 *   Fixes an undefined behavior arising from reference binding to nullptr in
2266 *   Fixes an undefined behavior arising from reference binding to nullptr and
2269 *   Fixes a division by 0 in inplace operations
2271 *   Fixes a crash caused by integer conversion to unsigned
2273 *   Fixes an undefined behavior arising from reference binding to nullptr in
2276 *   Fixes a heap OOB in boosted trees
2278 *   Fixes vulnerabilities arising from incomplete validation in `QuantizeV2`
2280 *   Fixes vulnerabilities arising from incomplete validation in MKL
2283 *   Fixes an undefined behavior arising from reference binding to nullptr in
2286 *   Fixes an undefined behavior arising from reference binding to nullptr in
2289 *   Fixes an FPE in `tf.raw_ops.UnravelIndex`
2291 *   Fixes a crash in NMS ops caused by integer conversion to unsigned
2293 *   Fixes a heap OOB in `UpperBound` and `LowerBound`
2295 *   Fixes an undefined behavior arising from reference binding to nullptr in map
2298 *   Fixes a heap OOB in `SdcaOptimizerV2`
2300 *   Fixes a `CHECK`-fail in `MapStage`
2302 *   Fixes a vulnerability arising from incomplete validation in `MaxPoolGrad`
2304 *   Fixes an undefined behavior arising from reference binding to nullptr in
2307 *   Fixes a division by 0 in most convolution operators
2309 *   Fixes vulnerabilities arising from missing validation in shape inference for
2312 *   Fixes an arbitrary code execution due to YAML deserialization
2314 *   Fixes a heap OOB in nested `tf.map_fn` with `RaggedTensor`s
2316 *   Fixes a division by zero in TFLite
2318 *   Fixes an NPE in TFLite
2320 *   Fixes a vulnerability arising from use of unitialized value in TFLite
2322 *   Fixes an FPE in TFLite division operations
2324 *   Fixes an FPE in TFLite pooling operations
2326 *   Fixes an infinite loop in TFLite
2328 *   Fixes a heap OOB in TFLite
2330 *   Fixes a heap OOB in TFLite's `Gather*` implementations
2332 *   Fixes an undefined behavior arising from null pointer dereference in TFLite
2334 *   Fixes an undefined behavior arising from null pointer dereference in TFLite
2337 *   Fixes a FPE in LSH in TFLite
2339 *   Fixes a segfault on strings tensors with mismatched dimensions, arising in
2342 *   Fixes a use after free and a potential segfault in shape inference functions
2353 This release introduces several vulnerability fixes:
2355 *   Fixes a code injection issue in `saved_model_cli`
2357 *   Fixes a vulnerability due to use of uninitialized value in Tensorflow
2359 *   Fixes a heap OOB in `FusedBatchNorm` kernels
2361 *   Fixes an arbitrary memory read in `ImmutableConst`
2363 *   Fixes a heap OOB in `SparseBinCount`
2365 *   Fixes a heap OOB in `SparseFillEmptyRows`
2367 *   Fixes a segfault due to negative splits in `SplitV`
2369 *   Fixes segfaults and vulnerabilities caused by accesses to invalid memory
2372 *   Fixes a null pointer exception when `Exit` node is not preceded by `Enter`
2375 *   Fixes an integer division by 0 in `tf.raw_ops.AllToAll`
2377 *   Fixes an undefined behavior via `nullptr` reference binding in sparse matrix
2380 *   Fixes a heap buffer overflow in `Transpose`
2384 *   Fixes a null pointer exception in `DeserializeSparse`
2386 *   Fixes an undefined behavior arising from reference binding to `nullptr` in
2389 *   Fixes a heap OOB read in `tf.ragged.cross`
2391 *   Fixes a heap OOB read in all `tf.raw_ops.QuantizeAndDequantizeV*` ops
2393 *   Fixes an FPE in `ParallelConcat`
2395 *   Fixes FPE issues in convolutions with zero size filters
2397 *   Fixes a heap OOB read in `tf.raw_ops.SparseCountSparseOutput`
2399 *   Fixes vulnerabilities caused by incomplete validation in boosted trees code
2401 *   Fixes vulnerabilities caused by incomplete validation of shapes in multiple
2404 *   Fixes a segfault produced while copying constant resource tensor
2406 *   Fixes a vulnerability caused by unitialized access in
2409 *   Fixes several vulnerabilities and segfaults caused by missing validation
2412 *   Fixes an overflow producing a crash in `tf.range`
2414 *   Fixes an overflow producing a crash in `tf.image.resize` when size is large
2416 *   Fixes an overflow producing a crash in `tf.tile` when tiling tensor is large
2418 *   Fixes a vulnerability produced due to incomplete validation in
2421 *   Fixes multiple crashes due to overflow and `CHECK`-fail in ops with large
2424 *   Fixes a crash in `max_pool3d` when size argument is 0 or negative
2426 *   Fixes a crash in `tf.math.segment_*` operations
2438 This release introduces several vulnerability fixes:
2440 *   Fixes a heap out of bounds access in sparse reduction operations
2442 *   Fixes a floating point exception in `SparseDenseCwiseDiv`
2444 *   Fixes a null pointer dereference in `CompressElement`
2446 *   Fixes a null pointer dereference in `RaggedTensorToTensor`
2448 *   Fixes a null pointer dereference and a heap OOB read arising from operations
2451 *   Fixes an integer division by 0 in sparse reshaping
2453 *   Fixes a division by 0 in `ResourceScatterDiv`
2455 *   Fixes a heap OOB in `RaggedGather`
2457 *   Fixes a `std::abort` raised from `TensorListReserve`
2459 *   Fixes a null pointer dereference in `MatrixDiagPartOp`
2461 *   Fixes an integer overflow due to conversion to unsigned
2463 *   Fixes a bad allocation error in `StringNGrams` caused by integer conversion
2465 *   Fixes a null pointer dereference in `SparseTensorSliceDataset`
2467 *   Fixes an incorrect validation of `SaveV2` inputs
2469 *   Fixes a null pointer dereference in `UncompressElement`
2471 *   Fixes a segfault and a heap buffer overflow in
2474 *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
2476 *   Fixes a use after free in boosted trees creation
2478 *   Fixes a division by 0 in `ResourceGather`
2480 *   Fixes a heap OOB and a `CHECK` fail in `ResourceGather`
2482 *   Fixes a heap OOB in `ResourceScatterUpdate`
2484 *   Fixes an undefined behavior arising from reference binding to nullptr in
2487 *   Fixes an undefined behavior arising from reference binding to nullptr in
2490 *   Fixes an undefined behavior arising from reference binding to nullptr in
2493 *   Fixes an undefined behavior arising from reference binding to nullptr and
2496 *   Fixes a division by 0 in inplace operations
2498 *   Fixes a crash caused by integer conversion to unsigned
2500 *   Fixes an undefined behavior arising from reference binding to nullptr in
2503 *   Fixes a heap OOB in boosted trees
2505 *   Fixes vulnerabilities arising from incomplete validation in `QuantizeV2`
2507 *   Fixes vulnerabilities arising from incomplete validation in MKL
2510 *   Fixes an undefined behavior arising from reference binding to nullptr in
2513 *   Fixes an undefined behavior arising from reference binding to nullptr in
2516 *   Fixes an FPE in `tf.raw_ops.UnravelIndex`
2518 *   Fixes a crash in NMS ops caused by integer conversion to unsigned
2520 *   Fixes a heap OOB in `UpperBound` and `LowerBound`
2522 *   Fixes an undefined behavior arising from reference binding to nullptr in map
2525 *   Fixes a heap OOB in `SdcaOptimizerV2`
2527 *   Fixes a `CHECK`-fail in `MapStage`
2529 *   Fixes a vulnerability arising from incomplete validation in `MaxPoolGrad`
2531 *   Fixes an undefined behavior arising from reference binding to nullptr in
2534 *   Fixes a division by 0 in most convolution operators
2536 *   Fixes vulnerabilities arising from missing validation in shape inference for
2539 *   Fixes an arbitrary code execution due to YAML deserialization
2541 *   Fixes a heap OOB in nested `tf.map_fn` with `RaggedTensor`s
2543 *   Fixes a division by zero in TFLite
2545 *   Fixes an NPE in TFLite
2547 *   Fixes a vulnerability arising from use of unitialized value in TFLite
2549 *   Fixes an FPE in TFLite division operations
2551 *   Fixes an FPE in TFLite pooling operations
2553 *   Fixes an infinite loop in TFLite
2555 *   Fixes a heap OOB in TFLite
2557 *   Fixes a heap OOB in TFLite's `Gather*` implementations
2559 *   Fixes an undefined behavior arising from null pointer dereference in TFLite
2561 *   Fixes an undefined behavior arising from null pointer dereference in TFLite
2564 *   Fixes a FPE in LSH in TFLite
2566 *   Fixes a segfault on strings tensors with mismatched dimensions, arising in
2569 *   Fixes a use after free and a potential segfault in shape inference functions
2580 This release introduces several vulnerability fixes:
2582 *   Fixes a heap out of bounds access in sparse reduction operations
2584 *   Fixes a floating point exception in `SparseDenseCwiseDiv`
2586 *   Fixes a null pointer dereference in `CompressElement`
2588 *   Fixes a null pointer dereference in `RaggedTensorToTensor`
2590 *   Fixes a null pointer dereference and a heap OOB read arising from operations
2593 *   Fixes an integer division by 0 in sparse reshaping
2595 *   Fixes a division by 0 in `ResourceScatterDiv`
2597 *   Fixes a heap OOB in `RaggedGather`
2599 *   Fixes a `std::abort` raised from `TensorListReserve`
2601 *   Fixes a null pointer dereference in `MatrixDiagPartOp`
2603 *   Fixes an integer overflow due to conversion to unsigned
2605 *   Fixes a bad allocation error in `StringNGrams` caused by integer conversion
2607 *   Fixes a null pointer dereference in `SparseTensorSliceDataset`
2609 *   Fixes an incorrect validation of `SaveV2` inputs
2611 *   Fixes a null pointer dereference in `UncompressElement`
2613 *   Fixes a segfault and a heap buffer overflow in
2616 *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
2618 *   Fixes a use after free in boosted trees creation
2620 *   Fixes a division by 0 in `ResourceGather`
2622 *   Fixes a heap OOB and a `CHECK` fail in `ResourceGather`
2624 *   Fixes a heap OOB in `ResourceScatterUpdate`
2626 *   Fixes an undefined behavior arising from reference binding to nullptr in
2629 *   Fixes an undefined behavior arising from reference binding to nullptr in
2632 *   Fixes an undefined behavior arising from reference binding to nullptr in
2635 *   Fixes an undefined behavior arising from reference binding to nullptr and
2638 *   Fixes a division by 0 in inplace operations
2640 *   Fixes a crash caused by integer conversion to unsigned
2642 *   Fixes an undefined behavior arising from reference binding to nullptr in
2645 *   Fixes a heap OOB in boosted trees
2647 *   Fixes vulnerabilities arising from incomplete validation in `QuantizeV2`
2649 *   Fixes vulnerabilities arising from incomplete validation in MKL
2652 *   Fixes an undefined behavior arising from reference binding to nullptr in
2655 *   Fixes an undefined behavior arising from reference binding to nullptr in
2658 *   Fixes an FPE in `tf.raw_ops.UnravelIndex`
2660 *   Fixes a crash in NMS ops caused by integer conversion to unsigned
2662 *   Fixes a heap OOB in `UpperBound` and `LowerBound`
2664 *   Fixes an undefined behavior arising from reference binding to nullptr in map
2667 *   Fixes a heap OOB in `SdcaOptimizerV2`
2669 *   Fixes a `CHECK`-fail in `MapStage`
2671 *   Fixes a vulnerability arising from incomplete validation in `MaxPoolGrad`
2673 *   Fixes an undefined behavior arising from reference binding to nullptr in
2676 *   Fixes a division by 0 in most convolution operators
2678 *   Fixes vulnerabilities arising from missing validation in shape inference for
2681 *   Fixes an arbitrary code execution due to YAML deserialization
2683 *   Fixes a heap OOB in nested `tf.map_fn` with `RaggedTensor`s
2685 *   Fixes a division by zero in TFLite
2687 *   Fixes an NPE in TFLite
2689 *   Fixes a vulnerability arising from use of unitialized value in TFLite
2691 *   Fixes an FPE in TFLite division operations
2693 *   Fixes an FPE in TFLite pooling operations
2695 *   Fixes an infinite loop in TFLite
2697 *   Fixes a heap OOB in TFLite
2699 *   Fixes a heap OOB in TFLite's `Gather*` implementations
2701 *   Fixes an undefined behavior arising from null pointer dereference in TFLite
2703 *   Fixes an undefined behavior arising from null pointer dereference in TFLite
2706 *   Fixes a FPE in LSH in TFLite
2708 *   Fixes a segfault on strings tensors with mismatched dimensions, arising in
2711 *   Fixes a use after free and a potential segfault in shape inference functions
2722 This release introduces several vulnerability fixes:
2724 *   Fixes a heap buffer overflow in `RaggedBinCount`
2726 *   Fixes a heap out of bounds write in `RaggedBinCount`
2728 *   Fixes a type confusion during tensor casts which leads to dereferencing null
2731 *   Fixes a reference binding to null pointer in `MatrixDiag*` ops
2733 *   Fixes a null pointer dereference via invalid Ragged Tensors
2735 *   Fixes a division by zero in `Conv3D`
2737 *   Fixes vulnerabilities where session operations in eager mode lead to null
2740 *   Fixes a `CHECK`-fail in `SparseCross` caused by type confusion
2742 *   Fixes a segfault in `SparseCountSparseOutput`
2744 *   Fixes a heap buffer overflow in `Conv3DBackprop*`
2746 *   Fixes a division by 0 in `Conv3DBackprop*`
2748 *   Fixes a `CHECK`-fail in `AddManySparseToTensorsMap`
2750 *   Fixes a division by 0 in `Conv2DBackpropFilter`
2752 *   Fixes a division by 0 in `Conv2DBackpropInput`
2754 *   Fixes a division by 0 in `Conv2D`
2756 *   Fixes a division by 0 in `QuantizedConv2D`
2758 *   Fixes a division by 0 in `QuantizedMul`
2760 *   Fixes vulnerabilities caused by invalid validation in
2763 *   Fixes a heap buffer overflow caused by rounding
2765 *   Fixes a `CHECK`-fail in `tf.raw_ops.EncodePng`
2767 *   Fixes a heap out of bounds read in `RaggedCross`
2769 *   Fixes a `CHECK`-fail in `DrawBoundingBoxes`
2771 *   Fixes a heap buffer overflow in `QuantizedMul`
2773 *   Fixes a `CHECK`-fail in `SparseConcat`
2775 *   Fixes a heap buffer overflow in `QuantizedResizeBilinear`
2777 *   Fixes a heap buffer overflow in `QuantizedReshape`
2779 *   Fixes a division by zero in `Conv2DBackpropFilter`
2781 *   Fixes a heap buffer overflow in `Conv2DBackpropFilter`
2783 *   Fixes a heap buffer overflow in `StringNGrams`
2785 *   Fixes a null pointer dereference in `StringNGrams`
2787 *   Fixes a `CHECK`-fail in `QuantizeAndDequantizeV4Grad`
2789 *   Fixes a `CHECK`-fail in `CTCGreedyDecoder`
2791 *   Fixes a heap buffer overflow in `SparseTensorToCSRSparseMatrix`
2793 *   Fixes a division by 0 in `QuantizedBiasAdd`
2795 *   Fixes a heap out of bounds in `QuantizedBatchNormWithGlobalNormalization`
2797 *   Fixes a division by 0 in `QuantizedBatchNormWithGlobalNormalization`
2799 *   Fixes a division by 0 in `QuantizedAdd`
2801 *   Fixes a division by 0 in `FractionalAvgPool`
2803 *   Fixes an OOB read in `MatrixTriangularSolve`
2805 *   Fixes a heap OOB in `QuantizeAndDequantizeV3`
2807 *   Fixes a `CHECK`-failure in `UnsortedSegmentJoin`
2809 *   Fixes a division by 0 in `DenseCountSparseOutput`
2811 *   Fixes a division by 0 in `FusedBatchNorm`
2813 *   Fixes a division by 0 in `SparseMatMul`
2815 *   Fixes a division by 0 in `Reverse`
2817 *   Fixes a heap buffer overflow in `SparseSplit`
2819 *   Fixes a heap OOB access in unicode ops
2821 *   Fixes a heap buffer overflow in `RaggedTensorToTensor`
2823 *   Fixes a `CHECK`-fail in `LoadAndRemapMatrix`
2825 *   Fixes a `CHECK`-fail in `tf.raw_ops.IRFFT`
2827 *   Fixes a `CHECK`-fail in `tf.raw_ops.RFFT`
2829 *   Fixes a null pointer dereference in `EditDistance`
2831 *   Fixes a null pointer dereference in `SparseFillEmptyRows`
2833 *   Fixes a heap OOB access in `Dilation2DBackpropInput`
2835 *   Fixes a reference binding to null in `ParameterizedTruncatedNormal`
2837 *   Fixes a set of vulnerabilities caused by lack of validation in
2840 *   Fixes a heap out of bounds read in `MaxPoolGradWithArgmax`
2842 *   Fixes a heap out of bounds read in `RequantizationRange`
2844 *   Fixes a memory corruption in `DrawBoundingBoxesV2`
2846 *   Fixes a reference binding to nullptr in `SdcaOptimizer`
2848 *   Fixes an overflow and a denial of service in `tf.raw_ops.ReverseSequence`
2850 *   Fixes a division by 0 in `MaxPoolGradWithArgmax`
2852 *   Fixes an undefined behavior in `MaxPool3DGradGrad`
2854 *   Fixes a heap buffer overflow in `MaxPool3DGradGrad`
2856 *   Fixes a heap buffer overflow in `AvgPool3DGrad`
2858 *   Fixes an undefined behavior and a `CHECK`-fail in `FractionalMaxPoolGrad`
2860 *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
2862 *   Fixes a heap buffer overflow in `MaxPoolGrad`
2864 *   Fixes a segfault in `CTCBeamSearchDecoder`
2866 *   Fixes a heap OOB read in `tf.raw_ops.Dequantize`
2868 *   Fixes a `CHECK`-fail due to integer overflow
2870 *   Fixes a heap buffer overflow and undefined behavior in `FusedBatchNorm`
2872 *   Fixes a division by zero in padding computation in TFLite
2874 *   Fixes a division by zero in optimized pooling implementations in TFLite
2876 *   Fixes a division by zero in TFLite's implementation of `SpaceToDepth`
2878 *   Fixes a division by zero in TFLite's implementation of `GatherNd`
2880 *   Fixes a division by zero in TFLite's implementation of `TransposeConv`
2882 *   Fixes a heap OOB read in TFLite's implementation of `Minimum` or `Maximum`
2884 *   Fixes a null pointer dereference in TFLite's `Reshape` operator
2886 *   Fixes a stack overflow due to looping TFLite subgraph
2888 *   Fixes a division by zero in TFLite's implementation of `DepthToSpace`
2890 *   Fixes a division by zero in TFLite's convolution code
2892 *   Fixes a division by zero in TFLite's implementation of `EmbeddingLookup`
2894 *   Fixes a division by zero in TFLite's implementation of `BatchToSpaceNd`
2896 *   Fixes a division by zero in TFLite's implementation of `SpaceToBatchNd`
2898 *   Fixes a division by zero in TFLite's implementation of `SVDF`
2900 *   Fixes a division by zero in TFLite's implementation of `Split`
2902 *   Fixes a division by zero in TFLite's implementation of `OneHot`
2904 *   Fixes a division by zero in TFLite's implementation of `DepthwiseConv`
2906 *   Fixes a division by zero in TFLite's implementation of hashtable lookup
2908 *   Fixes a integer overflow in TFLite concatentation
2910 *   Fixes a integer overflow in TFLite memory allocation
2912 *   Fixes a heap OOB write in TFLite
2914 *   Fixes a heap OOB read in TFLite
2916 *   Fixes a heap OOB and null pointer dereference in `RaggedTensorToTensor`
2918 *   Fixes vulnerabilities caused by incomplete validation in `SparseAdd`
2920 *   Fixes vulnerabilities caused by incomplete validation in
2923 *   Fixes vulnerabilities caused by incomplete validation in `SparseReshape`
2925 *   Fixes vulnerabilities caused by invalid validation in
2928 *   Fixes a heap buffer overflow in `BandedTriangularSolve`
2930 *   Fixes vulnerabilities caused by incomplete validation in
2933 *   Fixes an interpreter crash from vulnerabilities in `tf.io.decode_raw`
2935 *   Fixes a stack overflow in `ParseAttrValue` with nested tensors
2937 *   Fixes a null dereference in Grappler's `TrySimplify`
2939 *   Fixes a crash in `tf.transpose` with complex inputs
2941 *   Fixes a crash in `tf.strings.substr` due to `CHECK`-fail
2943 *   Fixes a segfault in `tf.raw_ops.SparseCountSparseOutput`
2945 *   Fixes a segfault in `tf.raw_ops.ImmutableConst`
2958 This release introduces several vulnerability fixes:
2960 *   Fixes a heap buffer overflow in `RaggedBinCount`
2962 *   Fixes a heap out of bounds write in `RaggedBinCount`
2964 *   Fixes a type confusion during tensor casts which leads to dereferencing null
2967 *   Fixes a reference binding to null pointer in `MatrixDiag*` ops
2969 *   Fixes a null pointer dereference via invalid Ragged Tensors
2971 *   Fixes a division by zero in `Conv3D`
2973 *   Fixes vulnerabilities where session operations in eager mode lead to null
2976 *   Fixes a `CHECK`-fail in `SparseCross` caused by type confusion
2978 *   Fixes a segfault in `SparseCountSparseOutput`
2980 *   Fixes a heap buffer overflow in `Conv3DBackprop*`
2982 *   Fixes a division by 0 in `Conv3DBackprop*`
2984 *   Fixes a `CHECK`-fail in `AddManySparseToTensorsMap`
2986 *   Fixes a division by 0 in `Conv2DBackpropFilter`
2988 *   Fixes a division by 0 in `Conv2DBackpropInput`
2990 *   Fixes a division by 0 in `Conv2D`
2992 *   Fixes a division by 0 in `QuantizedConv2D`
2994 *   Fixes a division by 0 in `QuantizedMul`
2996 *   Fixes vulnerabilities caused by invalid validation in
2999 *   Fixes a heap buffer overflow caused by rounding
3001 *   Fixes a `CHECK`-fail in `tf.raw_ops.EncodePng`
3003 *   Fixes a heap out of bounds read in `RaggedCross`
3005 *   Fixes a `CHECK`-fail in `DrawBoundingBoxes`
3007 *   Fixes a heap buffer overflow in `QuantizedMul`
3009 *   Fixes a `CHECK`-fail in `SparseConcat`
3011 *   Fixes a heap buffer overflow in `QuantizedResizeBilinear`
3013 *   Fixes a heap buffer overflow in `QuantizedReshape`
3015 *   Fixes a division by zero in `Conv2DBackpropFilter`
3017 *   Fixes a heap buffer overflow in `Conv2DBackpropFilter`
3019 *   Fixes a heap buffer overflow in `StringNGrams`
3021 *   Fixes a null pointer dereference in `StringNGrams`
3023 *   Fixes a `CHECK`-fail in `QuantizeAndDequantizeV4Grad`
3025 *   Fixes a `CHECK`-fail in `CTCGreedyDecoder`
3027 *   Fixes a heap buffer overflow in `SparseTensorToCSRSparseMatrix`
3029 *   Fixes a division by 0 in `QuantizedBiasAdd`
3031 *   Fixes a heap out of bounds in `QuantizedBatchNormWithGlobalNormalization`
3033 *   Fixes a division by 0 in `QuantizedBatchNormWithGlobalNormalization`
3035 *   Fixes a division by 0 in `QuantizedAdd`
3037 *   Fixes a division by 0 in `FractionalAvgPool`
3039 *   Fixes an OOB read in `MatrixTriangularSolve`
3041 *   Fixes a heap OOB in `QuantizeAndDequantizeV3`
3043 *   Fixes a `CHECK`-failure in `UnsortedSegmentJoin`
3045 *   Fixes a division by 0 in `DenseCountSparseOutput`
3047 *   Fixes a division by 0 in `FusedBatchNorm`
3049 *   Fixes a division by 0 in `SparseMatMul`
3051 *   Fixes a division by 0 in `Reverse`
3053 *   Fixes a heap buffer overflow in `SparseSplit`
3055 *   Fixes a heap OOB access in unicode ops
3057 *   Fixes a heap buffer overflow in `RaggedTensorToTensor`
3059 *   Fixes a `CHECK`-fail in `LoadAndRemapMatrix`
3061 *   Fixes a `CHECK`-fail in `tf.raw_ops.IRFFT`
3063 *   Fixes a `CHECK`-fail in `tf.raw_ops.RFFT`
3065 *   Fixes a null pointer dereference in `EditDistance`
3067 *   Fixes a null pointer dereference in `SparseFillEmptyRows`
3069 *   Fixes a heap OOB access in `Dilation2DBackpropInput`
3071 *   Fixes a reference binding to null in `ParameterizedTruncatedNormal`
3073 *   Fixes a set of vulnerabilities caused by lack of validation in
3076 *   Fixes a heap out of bounds read in `MaxPoolGradWithArgmax`
3078 *   Fixes a heap out of bounds read in `RequantizationRange`
3080 *   Fixes a memory corruption in `DrawBoundingBoxesV2`
3082 *   Fixes a reference binding to nullptr in `SdcaOptimizer`
3084 *   Fixes an overflow and a denial of service in `tf.raw_ops.ReverseSequence`
3086 *   Fixes a division by 0 in `MaxPoolGradWithArgmax`
3088 *   Fixes an undefined behavior in `MaxPool3DGradGrad`
3090 *   Fixes a heap buffer overflow in `MaxPool3DGradGrad`
3092 *   Fixes a heap buffer overflow in `AvgPool3DGrad`
3094 *   Fixes an undefined behavior and a `CHECK`-fail in `FractionalMaxPoolGrad`
3096 *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
3098 *   Fixes a heap buffer overflow in `MaxPoolGrad`
3100 *   Fixes a segfault in `CTCBeamSearchDecoder`
3102 *   Fixes a heap OOB read in `tf.raw_ops.Dequantize`
3104 *   Fixes a `CHECK`-fail due to integer overflow
3106 *   Fixes a heap buffer overflow and undefined behavior in `FusedBatchNorm`
3108 *   Fixes a division by zero in padding computation in TFLite
3110 *   Fixes a division by zero in optimized pooling implementations in TFLite
3112 *   Fixes a division by zero in TFLite's implementation of `SpaceToDepth`
3114 *   Fixes a division by zero in TFLite's implementation of `GatherNd`
3116 *   Fixes a division by zero in TFLite's implementation of `TransposeConv`
3118 *   Fixes a heap OOB read in TFLite's implementation of `Minimum` or `Maximum`
3120 *   Fixes a null pointer dereference in TFLite's `Reshape` operator
3122 *   Fixes a stack overflow due to looping TFLite subgraph
3124 *   Fixes a division by zero in TFLite's implementation of `DepthToSpace`
3126 *   Fixes a division by zero in TFLite's convolution code
3128 *   Fixes a division by zero in TFLite's implementation of `EmbeddingLookup`
3130 *   Fixes a division by zero in TFLite's implementation of `BatchToSpaceNd`
3132 *   Fixes a division by zero in TFLite's implementation of `SpaceToBatchNd`
3134 *   Fixes a division by zero in TFLite's implementation of `SVDF`
3136 *   Fixes a division by zero in TFLite's implementation of `Split`
3138 *   Fixes a division by zero in TFLite's implementation of `OneHot`
3140 *   Fixes a division by zero in TFLite's implementation of `DepthwiseConv`
3142 *   Fixes a division by zero in TFLite's implementation of hashtable lookup
3144 *   Fixes a integer overflow in TFLite concatentation
3146 *   Fixes a integer overflow in TFLite memory allocation
3148 *   Fixes a heap OOB write in TFLite
3150 *   Fixes a heap OOB read in TFLite
3152 *   Fixes a heap OOB and null pointer dereference in `RaggedTensorToTensor`
3154 *   Fixes vulnerabilities caused by incomplete validation in `SparseAdd`
3156 *   Fixes vulnerabilities caused by incomplete validation in
3159 *   Fixes vulnerabilities caused by incomplete validation in `SparseReshape`
3161 *   Fixes vulnerabilities caused by invalid validation in
3164 *   Fixes a heap buffer overflow in `BandedTriangularSolve`
3166 *   Fixes vulnerabilities caused by incomplete validation in
3169 *   Fixes an interpreter crash from vulnerabilities in `tf.io.decode_raw`
3171 *   Fixes a stack overflow in `ParseAttrValue` with nested tensors
3173 *   Fixes a null dereference in Grappler's `TrySimplify`
3175 *   Fixes a crash in `tf.transpose` with complex inputs
3177 *   Fixes a crash in `tf.strings.substr` due to `CHECK`-fail
3179 *   Fixes a segfault in `tf.raw_ops.SparseCountSparseOutput`
3181 *   Fixes a segfault in `tf.raw_ops.ImmutableConst`
3194 This release introduces several vulnerability fixes:
3196 *   Fixes a heap buffer overflow in `RaggedBinCount`
3198 *   Fixes a heap out of bounds write in `RaggedBinCount`
3200 *   Fixes a type confusion during tensor casts which leads to dereferencing null
3203 *   Fixes a reference binding to null pointer in `MatrixDiag*` ops
3205 *   Fixes a null pointer dereference via invalid Ragged Tensors
3207 *   Fixes a division by zero in `Conv3D`
3209 *   Fixes vulnerabilities where session operations in eager mode lead to null
3212 *   Fixes a `CHECK`-fail in `SparseCross` caused by type confusion
3214 *   Fixes a segfault in `SparseCountSparseOutput`
3216 *   Fixes a heap buffer overflow in `Conv3DBackprop*`
3218 *   Fixes a division by 0 in `Conv3DBackprop*`
3220 *   Fixes a `CHECK`-fail in `AddManySparseToTensorsMap`
3222 *   Fixes a division by 0 in `Conv2DBackpropFilter`
3224 *   Fixes a division by 0 in `Conv2DBackpropInput`
3226 *   Fixes a division by 0 in `Conv2D`
3228 *   Fixes a division by 0 in `QuantizedConv2D`
3230 *   Fixes a division by 0 in `QuantizedMul`
3232 *   Fixes vulnerabilities caused by invalid validation in
3235 *   Fixes a heap buffer overflow caused by rounding
3237 *   Fixes a `CHECK`-fail in `tf.raw_ops.EncodePng`
3239 *   Fixes a heap out of bounds read in `RaggedCross`
3241 *   Fixes a `CHECK`-fail in `DrawBoundingBoxes`
3243 *   Fixes a heap buffer overflow in `QuantizedMul`
3245 *   Fixes a `CHECK`-fail in `SparseConcat`
3247 *   Fixes a heap buffer overflow in `QuantizedResizeBilinear`
3249 *   Fixes a heap buffer overflow in `QuantizedReshape`
3251 *   Fixes a division by zero in `Conv2DBackpropFilter`
3253 *   Fixes a heap buffer overflow in `Conv2DBackpropFilter`
3255 *   Fixes a heap buffer overflow in `StringNGrams`
3257 *   Fixes a null pointer dereference in `StringNGrams`
3259 *   Fixes a `CHECK`-fail in `QuantizeAndDequantizeV4Grad`
3261 *   Fixes a `CHECK`-fail in `CTCGreedyDecoder`
3263 *   Fixes a heap buffer overflow in `SparseTensorToCSRSparseMatrix`
3265 *   Fixes a division by 0 in `QuantizedBiasAdd`
3267 *   Fixes a heap out of bounds in `QuantizedBatchNormWithGlobalNormalization`
3269 *   Fixes a division by 0 in `QuantizedBatchNormWithGlobalNormalization`
3271 *   Fixes a division by 0 in `QuantizedAdd`
3273 *   Fixes a division by 0 in `FractionalAvgPool`
3275 *   Fixes an OOB read in `MatrixTriangularSolve`
3277 *   Fixes a heap OOB in `QuantizeAndDequantizeV3`
3279 *   Fixes a `CHECK`-failure in `UnsortedSegmentJoin`
3281 *   Fixes a division by 0 in `DenseCountSparseOutput`
3283 *   Fixes a division by 0 in `FusedBatchNorm`
3285 *   Fixes a division by 0 in `SparseMatMul`
3287 *   Fixes a division by 0 in `Reverse`
3289 *   Fixes a heap buffer overflow in `SparseSplit`
3291 *   Fixes a heap OOB access in unicode ops
3293 *   Fixes a heap buffer overflow in `RaggedTensorToTensor`
3295 *   Fixes a `CHECK`-fail in `LoadAndRemapMatrix`
3297 *   Fixes a `CHECK`-fail in `tf.raw_ops.IRFFT`
3299 *   Fixes a `CHECK`-fail in `tf.raw_ops.RFFT`
3301 *   Fixes a null pointer dereference in `EditDistance`
3303 *   Fixes a null pointer dereference in `SparseFillEmptyRows`
3305 *   Fixes a heap OOB access in `Dilation2DBackpropInput`
3307 *   Fixes a reference binding to null in `ParameterizedTruncatedNormal`
3309 *   Fixes a set of vulnerabilities caused by lack of validation in
3312 *   Fixes a heap out of bounds read in `MaxPoolGradWithArgmax`
3314 *   Fixes a heap out of bounds read in `RequantizationRange`
3316 *   Fixes a memory corruption in `DrawBoundingBoxesV2`
3318 *   Fixes a reference binding to nullptr in `SdcaOptimizer`
3320 *   Fixes an overflow and a denial of service in `tf.raw_ops.ReverseSequence`
3322 *   Fixes a division by 0 in `MaxPoolGradWithArgmax`
3324 *   Fixes an undefined behavior in `MaxPool3DGradGrad`
3326 *   Fixes a heap buffer overflow in `MaxPool3DGradGrad`
3328 *   Fixes a heap buffer overflow in `AvgPool3DGrad`
3330 *   Fixes an undefined behavior and a `CHECK`-fail in `FractionalMaxPoolGrad`
3332 *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
3334 *   Fixes a heap buffer overflow in `MaxPoolGrad`
3336 *   Fixes a segfault in `CTCBeamSearchDecoder`
3338 *   Fixes a heap OOB read in `tf.raw_ops.Dequantize`
3340 *   Fixes a `CHECK`-fail due to integer overflow
3342 *   Fixes a heap buffer overflow and undefined behavior in `FusedBatchNorm`
3344 *   Fixes a division by zero in padding computation in TFLite
3346 *   Fixes a division by zero in optimized pooling implementations in TFLite
3348 *   Fixes a division by zero in TFLite's implementation of `SpaceToDepth`
3350 *   Fixes a division by zero in TFLite's implementation of `GatherNd`
3352 *   Fixes a division by zero in TFLite's implementation of `TransposeConv`
3354 *   Fixes a heap OOB read in TFLite's implementation of `Minimum` or `Maximum`
3356 *   Fixes a null pointer dereference in TFLite's `Reshape` operator
3358 *   Fixes a stack overflow due to looping TFLite subgraph
3360 *   Fixes a division by zero in TFLite's implementation of `DepthToSpace`
3362 *   Fixes a division by zero in TFLite's convolution code
3364 *   Fixes a division by zero in TFLite's implementation of `EmbeddingLookup`
3366 *   Fixes a division by zero in TFLite's implementation of `BatchToSpaceNd`
3368 *   Fixes a division by zero in TFLite's implementation of `SpaceToBatchNd`
3370 *   Fixes a division by zero in TFLite's implementation of `SVDF`
3372 *   Fixes a division by zero in TFLite's implementation of `Split`
3374 *   Fixes a division by zero in TFLite's implementation of `OneHot`
3376 *   Fixes a division by zero in TFLite's implementation of `DepthwiseConv`
3378 *   Fixes a division by zero in TFLite's implementation of hashtable lookup
3380 *   Fixes a integer overflow in TFLite concatentation
3382 *   Fixes a integer overflow in TFLite memory allocation
3384 *   Fixes a heap OOB write in TFLite
3386 *   Fixes a heap OOB read in TFLite
3388 *   Fixes a heap OOB and null pointer dereference in `RaggedTensorToTensor`
3390 *   Fixes vulnerabilities caused by incomplete validation in `SparseAdd`
3392 *   Fixes vulnerabilities caused by incomplete validation in
3395 *   Fixes vulnerabilities caused by incomplete validation in `SparseReshape`
3397 *   Fixes vulnerabilities caused by invalid validation in
3400 *   Fixes a heap buffer overflow in `BandedTriangularSolve`
3402 *   Fixes vulnerabilities caused by incomplete validation in
3405 *   Fixes an interpreter crash from vulnerabilities in `tf.io.decode_raw`
3407 *   Fixes a stack overflow in `ParseAttrValue` with nested tensors
3409 *   Fixes a null dereference in Grappler's `TrySimplify`
3411 *   Fixes a crash in `tf.transpose` with complex inputs
3413 *   Fixes a crash in `tf.strings.substr` due to `CHECK`-fail
3415 *   Fixes a segfault in `tf.raw_ops.SparseCountSparseOutput`
3417 *   Fixes a segfault in `tf.raw_ops.ImmutableConst`
3430 This release introduces several vulnerability fixes:
3432 *   Fixes a heap buffer overflow in `RaggedBinCount`
3434 *   Fixes a heap out of bounds write in `RaggedBinCount`
3436 *   Fixes a type confusion during tensor casts which leads to dereferencing null
3439 *   Fixes a reference binding to null pointer in `MatrixDiag*` ops
3441 *   Fixes a null pointer dereference via invalid Ragged Tensors
3443 *   Fixes a division by zero in `Conv3D`
3445 *   Fixes vulnerabilities where session operations in eager mode lead to null
3448 *   Fixes a `CHECK`-fail in `SparseCross` caused by type confusion
3450 *   Fixes a segfault in `SparseCountSparseOutput`
3452 *   Fixes a heap buffer overflow in `Conv3DBackprop*`
3454 *   Fixes a division by 0 in `Conv3DBackprop*`
3456 *   Fixes a `CHECK`-fail in `AddManySparseToTensorsMap`
3458 *   Fixes a division by 0 in `Conv2DBackpropFilter`
3460 *   Fixes a division by 0 in `Conv2DBackpropInput`
3462 *   Fixes a division by 0 in `Conv2D`
3464 *   Fixes a division by 0 in `QuantizedConv2D`
3466 *   Fixes a division by 0 in `QuantizedMul`
3468 *   Fixes vulnerabilities caused by invalid validation in
3471 *   Fixes a heap buffer overflow caused by rounding
3473 *   Fixes a `CHECK`-fail in `tf.raw_ops.EncodePng`
3475 *   Fixes a heap out of bounds read in `RaggedCross`
3477 *   Fixes a `CHECK`-fail in `DrawBoundingBoxes`
3479 *   Fixes a heap buffer overflow in `QuantizedMul`
3481 *   Fixes a `CHECK`-fail in `SparseConcat`
3483 *   Fixes a heap buffer overflow in `QuantizedResizeBilinear`
3485 *   Fixes a heap buffer overflow in `QuantizedReshape`
3487 *   Fixes a division by zero in `Conv2DBackpropFilter`
3489 *   Fixes a heap buffer overflow in `Conv2DBackpropFilter`
3491 *   Fixes a heap buffer overflow in `StringNGrams`
3493 *   Fixes a null pointer dereference in `StringNGrams`
3495 *   Fixes a `CHECK`-fail in `QuantizeAndDequantizeV4Grad`
3497 *   Fixes a `CHECK`-fail in `CTCGreedyDecoder`
3499 *   Fixes a heap buffer overflow in `SparseTensorToCSRSparseMatrix`
3501 *   Fixes a division by 0 in `QuantizedBiasAdd`
3503 *   Fixes a heap out of bounds in `QuantizedBatchNormWithGlobalNormalization`
3505 *   Fixes a division by 0 in `QuantizedBatchNormWithGlobalNormalization`
3507 *   Fixes a division by 0 in `QuantizedAdd`
3509 *   Fixes a division by 0 in `FractionalAvgPool`
3511 *   Fixes an OOB read in `MatrixTriangularSolve`
3513 *   Fixes a heap OOB in `QuantizeAndDequantizeV3`
3515 *   Fixes a `CHECK`-failure in `UnsortedSegmentJoin`
3517 *   Fixes a division by 0 in `DenseCountSparseOutput`
3519 *   Fixes a division by 0 in `FusedBatchNorm`
3521 *   Fixes a division by 0 in `SparseMatMul`
3523 *   Fixes a division by 0 in `Reverse`
3525 *   Fixes a heap buffer overflow in `SparseSplit`
3527 *   Fixes a heap OOB access in unicode ops
3529 *   Fixes a heap buffer overflow in `RaggedTensorToTensor`
3531 *   Fixes a `CHECK`-fail in `LoadAndRemapMatrix`
3533 *   Fixes a `CHECK`-fail in `tf.raw_ops.IRFFT`
3535 *   Fixes a `CHECK`-fail in `tf.raw_ops.RFFT`
3537 *   Fixes a null pointer dereference in `EditDistance`
3539 *   Fixes a null pointer dereference in `SparseFillEmptyRows`
3541 *   Fixes a heap OOB access in `Dilation2DBackpropInput`
3543 *   Fixes a reference binding to null in `ParameterizedTruncatedNormal`
3545 *   Fixes a set of vulnerabilities caused by lack of validation in
3548 *   Fixes a heap out of bounds read in `MaxPoolGradWithArgmax`
3550 *   Fixes a heap out of bounds read in `RequantizationRange`
3552 *   Fixes a memory corruption in `DrawBoundingBoxesV2`
3554 *   Fixes a reference binding to nullptr in `SdcaOptimizer`
3556 *   Fixes an overflow and a denial of service in `tf.raw_ops.ReverseSequence`
3558 *   Fixes a division by 0 in `MaxPoolGradWithArgmax`
3560 *   Fixes an undefined behavior in `MaxPool3DGradGrad`
3562 *   Fixes a heap buffer overflow in `MaxPool3DGradGrad`
3564 *   Fixes a heap buffer overflow in `AvgPool3DGrad`
3566 *   Fixes an undefined behavior and a `CHECK`-fail in `FractionalMaxPoolGrad`
3568 *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
3570 *   Fixes a heap buffer overflow in `MaxPoolGrad`
3572 *   Fixes a segfault in `CTCBeamSearchDecoder`
3574 *   Fixes a heap OOB read in `tf.raw_ops.Dequantize`
3576 *   Fixes a `CHECK`-fail due to integer overflow
3578 *   Fixes a heap buffer overflow and undefined behavior in `FusedBatchNorm`
3580 *   Fixes a division by zero in padding computation in TFLite
3582 *   Fixes a division by zero in optimized pooling implementations in TFLite
3584 *   Fixes a division by zero in TFLite's implementation of `SpaceToDepth`
3586 *   Fixes a division by zero in TFLite's implementation of `GatherNd`
3588 *   Fixes a division by zero in TFLite's implementation of `TransposeConv`
3590 *   Fixes a heap OOB read in TFLite's implementation of `Minimum` or `Maximum`
3592 *   Fixes a null pointer dereference in TFLite's `Reshape` operator
3594 *   Fixes a stack overflow due to looping TFLite subgraph
3596 *   Fixes a division by zero in TFLite's implementation of `DepthToSpace`
3598 *   Fixes a division by zero in TFLite's convolution code
3600 *   Fixes a division by zero in TFLite's implementation of `EmbeddingLookup`
3602 *   Fixes a division by zero in TFLite's implementation of `BatchToSpaceNd`
3604 *   Fixes a division by zero in TFLite's implementation of `SpaceToBatchNd`
3606 *   Fixes a division by zero in TFLite's implementation of `SVDF`
3608 *   Fixes a division by zero in TFLite's implementation of `Split`
3610 *   Fixes a division by zero in TFLite's implementation of `OneHot`
3612 *   Fixes a division by zero in TFLite's implementation of `DepthwiseConv`
3614 *   Fixes a division by zero in TFLite's implementation of hashtable lookup
3616 *   Fixes a integer overflow in TFLite concatentation
3618 *   Fixes a integer overflow in TFLite memory allocation
3620 *   Fixes a heap OOB write in TFLite
3622 *   Fixes a heap OOB read in TFLite
3624 *   Fixes a heap OOB and null pointer dereference in `RaggedTensorToTensor`
3626 *   Fixes vulnerabilities caused by incomplete validation in `SparseAdd`
3628 *   Fixes vulnerabilities caused by incomplete validation in
3631 *   Fixes vulnerabilities caused by incomplete validation in `SparseReshape`
3633 *   Fixes vulnerabilities caused by invalid validation in
3636 *   Fixes a heap buffer overflow in `BandedTriangularSolve`
3638 *   Fixes vulnerabilities caused by incomplete validation in
3641 *   Fixes an interpreter crash from vulnerabilities in `tf.io.decode_raw`
3643 *   Fixes a stack overflow in `ParseAttrValue` with nested tensors
3645 *   Fixes a null dereference in Grappler's `TrySimplify`
3647 *   Fixes a crash in `tf.transpose` with complex inputs
3649 *   Fixes a crash in `tf.strings.substr` due to `CHECK`-fail
3651 *   Fixes a segfault in `tf.raw_ops.SparseCountSparseOutput`
3653 *   Fixes a segfault in `tf.raw_ops.ImmutableConst`
3693         *   The new backend removes the redundant rescales and fixes some bugs
3742 ## Bug Fixes and Other Changes
3820         *   Additional tests and fixes for ADD and SUB operators.
3949     *   Fixes a heap buffer overflow in `RaggedBinCount`
3951     *   Fixes a heap out of bounds write in `RaggedBinCount`
3953     *   Fixes a type confusion during tensor casts which leads to dereferencing
3956     *   Fixes a reference binding to null pointer in `MatrixDiag*` ops
3958     *   Fixes a null pointer dereference via invalid Ragged Tensors
3960     *   Fixes a division by zero in `Conv3D`
3962     *   Fixes vulnerabilities where session operations in eager mode lead to
3965     *   Fixes a `CHECK`-fail in `SparseCross` caused by type confusion
3967     *   Fixes a segfault in `SparseCountSparseOutput`
3969     *   Fixes a heap buffer overflow in `Conv3DBackprop*`
3971     *   Fixes a division by 0 in `Conv3DBackprop*`
3973     *   Fixes a `CHECK`-fail in `AddManySparseToTensorsMap`
3975     *   Fixes a division by 0 in `Conv2DBackpropFilter`
3977     *   Fixes a division by 0 in `Conv2DBackpropInput`
3979     *   Fixes a division by 0 in `Conv2D`
3981     *   Fixes a division by 0 in `QuantizedConv2D`
3983     *   Fixes a division by 0 in `QuantizedMul`
3985     *   Fixes vulnerabilities caused by invalid validation in
3988     *   Fixes a heap buffer overflow caused by rounding
3990     *   Fixes a `CHECK`-fail in `tf.raw_ops.EncodePng`
3992     *   Fixes a heap out of bounds read in `RaggedCross`
3994     *   Fixes a `CHECK`-fail in `DrawBoundingBoxes`
3996     *   Fixes a heap buffer overflow in `QuantizedMul`
3998     *   Fixes a `CHECK`-fail in `SparseConcat`
4000     *   Fixes a heap buffer overflow in `QuantizedResizeBilinear`
4002     *   Fixes a heap buffer overflow in `QuantizedReshape`
4004     *   Fixes a division by zero in `Conv2DBackpropFilter`
4006     *   Fixes a heap buffer overflow in `Conv2DBackpropFilter`
4008     *   Fixes a heap buffer overflow in `StringNGrams`
4010     *   Fixes a null pointer dereference in `StringNGrams`
4012     *   Fixes a `CHECK`-fail in `QuantizeAndDequantizeV4Grad`
4014     *   Fixes a `CHECK`-fail in `CTCGreedyDecoder`
4016     *   Fixes a heap buffer overflow in `SparseTensorToCSRSparseMatrix`
4018     *   Fixes a division by 0 in `QuantizedBiasAdd`
4020     *   Fixes a heap out of bounds in
4023     *   Fixes a division by 0 in `QuantizedBatchNormWithGlobalNormalization`
4025     *   Fixes a division by 0 in `QuantizedAdd`
4027     *   Fixes a division by 0 in `FractionalAvgPool`
4029     *   Fixes an OOB read in `MatrixTriangularSolve`
4031     *   Fixes a heap OOB in `QuantizeAndDequantizeV3`
4033     *   Fixes a `CHECK`-failure in `UnsortedSegmentJoin`
4035     *   Fixes a division by 0 in `DenseCountSparseOutput`
4037     *   Fixes a division by 0 in `FusedBatchNorm`
4039     *   Fixes a division by 0 in `SparseMatMul`
4041     *   Fixes a division by 0 in `Reverse`
4043     *   Fixes a heap buffer overflow in `SparseSplit`
4045     *   Fixes a heap OOB access in unicode ops
4047     *   Fixes a heap buffer overflow in `RaggedTensorToTensor`
4049     *   Fixes a `CHECK`-fail in `LoadAndRemapMatrix`
4051     *   Fixes a `CHECK`-fail in `tf.raw_ops.IRFFT`
4053     *   Fixes a `CHECK`-fail in `tf.raw_ops.RFFT`
4055     *   Fixes a null pointer dereference in `EditDistance`
4057     *   Fixes a null pointer dereference in `SparseFillEmptyRows`
4059     *   Fixes a heap OOB access in `Dilation2DBackpropInput`
4061     *   Fixes a reference binding to null in `ParameterizedTruncatedNormal`
4063     *   Fixes a set of vulnerabilities caused by lack of validation in
4066     *   Fixes a heap out of bounds read in `MaxPoolGradWithArgmax`
4068     *   Fixes a heap out of bounds read in `RequantizationRange`
4070     *   Fixes a memory corruption in `DrawBoundingBoxesV2`
4072     *   Fixes a reference binding to nullptr in `SdcaOptimizer`
4074     *   Fixes an overflow and a denial of service in
4077     *   Fixes a division by 0 in `MaxPoolGradWithArgmax`
4079     *   Fixes an undefined behavior in `MaxPool3DGradGrad`
4081     *   Fixes a heap buffer overflow in `MaxPool3DGradGrad`
4083     *   Fixes a heap buffer overflow in `AvgPool3DGrad`
4085     *   Fixes an undefined behavior and a `CHECK`-fail in
4088     *   Fixes a heap buffer overflow in `FractionalAvgPoolGrad`
4090     *   Fixes a heap buffer overflow in `MaxPoolGrad`
4092     *   Fixes a segfault in `CTCBeamSearchDecoder`
4094     *   Fixes a heap OOB read in `tf.raw_ops.Dequantize`
4096     *   Fixes a `CHECK`-fail due to integer overflow
4098     *   Fixes a heap buffer overflow and undefined behavior in `FusedBatchNorm`
4100     *   Fixes a division by zero in padding computation in TFLite
4102     *   Fixes a division by zero in optimized pooling implementations in TFLite
4104     *   Fixes a division by zero in TFLite's implementation of `SpaceToDepth`
4106     *   Fixes a division by zero in TFLite's implementation of `GatherNd`
4108     *   Fixes a division by zero in TFLite's implementation of `TransposeConv`
4110     *   Fixes a heap OOB read in TFLite's implementation of `Minimum` or
4113     *   Fixes a null pointer dereference in TFLite's `Reshape` operator
4115     *   Fixes a stack overflow due to looping TFLite subgraph
4117     *   Fixes a division by zero in TFLite's implementation of `DepthToSpace`
4119     *   Fixes a division by zero in TFLite's convolution code
4121     *   Fixes a division by zero in TFLite's implementation of `EmbeddingLookup`
4123     *   Fixes a division by zero in TFLite's implementation of `BatchToSpaceNd`
4125     *   Fixes a division by zero in TFLite's implementation of `SpaceToBatchNd`
4127     *   Fixes a division by zero in TFLite's implementation of `SVDF`
4129     *   Fixes a division by zero in TFLite's implementation of `Split`
4131     *   Fixes a division by zero in TFLite's implementation of `OneHot`
4133     *   Fixes a division by zero in TFLite's implementation of `DepthwiseConv`
4135     *   Fixes a division by zero in TFLite's implementation of hashtable lookup
4137     *   Fixes a integer overflow in TFLite concatentation
4139     *   Fixes a integer overflow in TFLite memory allocation
4141     *   Fixes a heap OOB write in TFLite
4143     *   Fixes a heap OOB read in TFLite
4145     *   Fixes a heap OOB and null pointer dereference in `RaggedTensorToTensor`
4147     *   Fixes vulnerabilities caused by incomplete validation in `SparseAdd`
4149     *   Fixes vulnerabilities caused by incomplete validation in
4152     *   Fixes vulnerabilities caused by incomplete validation in `SparseReshape`
4154     *   Fixes vulnerabilities caused by invalid validation in
4157     *   Fixes a heap buffer overflow in `BandedTriangularSolve`
4159     *   Fixes vulnerabilities caused by incomplete validation in
4162     *   Fixes an interpreter crash from vulnerabilities in `tf.io.decode_raw`
4164     *   Fixes a stack overflow in `ParseAttrValue` with nested tensors
4166     *   Fixes a null dereference in Grappler's `TrySimplify`
4168     *   Fixes a crash in `tf.transpose` with complex inputs
4170     *   Fixes a crash in `tf.strings.substr` due to `CHECK`-fail
4172     *   Fixes a segfault in `tf.raw_ops.SparseCountSparseOutput`
4174     *   Fixes a segfault in `tf.raw_ops.ImmutableConst`
4250 ## Bug Fixes and Other Changes
4252 *   Fixes an access to unitialized memory in Eigen code
4254 *   Fixes a security vulnerability caused by lack of validation in
4257 *   Fixes a vulnerability caused by attempting to write to immutable memory
4260 *   Fixes a `CHECK`-fail in LSTM with zero-length input
4262 *   Fixes a security vulnerability caused by accessing heap data outside of
4278 ## Bug Fixes and Other Changes
4280 *   Fixes an access to unitialized memory in Eigen code
4282 *   Fixes a security vulnerability caused by lack of validation in
4285 *   Fixes a vulnerability caused by attempting to write to immutable memory
4288 *   Fixes a `CHECK`-fail in LSTM with zero-length input
4290 *   Fixes a security vulnerability caused by accessing heap data outside of
4306 ## Bug Fixes and Other Changes
4308 *   Fixes an access to unitialized memory in Eigen code
4310 *   Fixes a security vulnerability caused by lack of validation in
4313 *   Fixes a vulnerability caused by attempting to write to immutable memory
4316 *   Fixes a `CHECK`-fail in LSTM with zero-length input
4318 *   Fixes a security vulnerability caused by accessing heap data outside of
4336 ## Bug Fixes and Other Changes
4338 *   Fixes an access to unitialized memory in Eigen code
4340 *   Fixes a security vulnerability caused by lack of validation in
4343 *   Fixes a vulnerability caused by attempting to write to immutable memory
4346 *   Fixes a `CHECK`-fail in LSTM with zero-length input
4348 *   Fixes a security vulnerability caused by accessing heap data outside of
4365 ## Bug Fixes and Other Changes
4367 *   Fixes an access to unitialized memory in Eigen code
4369 *   Fixes a security vulnerability caused by lack of validation in
4372 *   Fixes a vulnerability caused by attempting to write to immutable memory
4375 *   Fixes a `CHECK`-fail in LSTM with zero-length input
4377 *   Fixes a security vulnerability caused by accessing heap data outside of
4403     major improvements involve handling peer failure and many bug fixes. Please
4607 ## Bug Fixes and Other Changes
4646     *   `tf.debugging.assert_shapes()` now works on `SparseTensor`s (Fixes
4734     major improvements involve handling peer failure and many bug fixes. Please
4740 *   Fixes various issues with saving a distributed model.
4850 *   Fixes an undefined behavior causing a segfault in `tf.raw_ops.Switch`,
4852 *   Fixes three vulnerabilities in conversion to DLPack format
4856 *   Fixes two vulnerabilities in `SparseFillEmptyRowsGrad`
4859 *   Fixes several vulnerabilities in `RaggedCountSparseOutput` and
4867 *   Fixes an integer truncation vulnerability in code using the work sharder
4870 *   Fixes a format string vulnerability in `tf.strings.as_string`,
4872 *   Fixes segfault raised by calling session-only ops in eager mode,
4874 *   Fixes data leak and potential ASLR violation from `tf.raw_ops.StringNGrams`,
4876 *   Fixes segfaults caused by incomplete `SavedModel` validation,
4878 *   Fixes a data corruption due to a bug in negative indexing support in TFLite,
4880 *   Fixes a data corruption due to dimension mismatch in TFLite,
4882 *   Fixes several vulnerabilities in TFLite saved model format
4886 *   Fixes several vulnerabilities in TFLite implementation of segment sum
4890 *   Fixes a segfault in `tf.quantization.quantize_and_dequantize`,
4892 *   Fixes an undefined behavior float cast causing a crash,
4894 *   Fixes a lack of validation in `tf.raw_ops.DataFormatVecPermute` and
4898 *   Fixes a crash caused by writing to read only memory region
4900 *   Fixes a heap out of bounds access in filesystem globbing implementation
4961 ## Bug Fixes and Other Changes
4963 *   Fixes an undefined behavior causing a segfault in `tf.raw_ops.Switch`
4965 *   Fixes three vulnerabilities in conversion to DLPack format
4969 *   Fixes two vulnerabilities in `SparseFillEmptyRowsGrad`
4972 *   Fixes several vulnerabilities in `RaggedCountSparseOutput` and
4980 *   Fixes an integer truncation vulnerability in code using the work sharder API
4982 *   Fixes a format string vulnerability in `tf.strings.as_string`
4984 *   Fixes segfault raised by calling session-only ops in eager mode
4986 *   Fixes data leak and potential ASLR violation from `tf.raw_ops.StringNGrams`
4988 *   Fixes segfaults caused by incomplete `SavedModel` validation
4990 *   Fixes a data corruption due to a bug in negative indexing support in TFLite
4992 *   Fixes a data corruption due to dimension mismatch in TFLite
4994 *   Fixes several vulnerabilities in TFLite saved model format
4998 *   Fixes several vulnerabilities in TFLite implementation of segment sum
5004 *   Fixes deprecated usage of `collections` API
5010 ## Bug Fixes and Other Changes
5012 *   Fixes an undefined behavior causing a segfault in `tf.raw_ops.Switch`
5014 *   Fixes three vulnerabilities in conversion to DLPack format
5018 *   Fixes two vulnerabilities in `SparseFillEmptyRowsGrad`
5021 *   Fixes an integer truncation vulnerability in code using the work sharder API
5023 *   Fixes a format string vulnerability in `tf.strings.as_string`
5025 *   Fixes segfault raised by calling session-only ops in eager mode
5027 *   Fixes data leak and potential ASLR violation from `tf.raw_ops.StringNGrams`
5029 *   Fixes segfaults caused by incomplete `SavedModel` validation
5031 *   Fixes a data corruption due to a bug in negative indexing support in TFLite
5033 *   Fixes a data corruption due to dimension mismatch in TFLite
5035 *   Fixes several vulnerabilities in TFLite saved model format
5039 *   Fixes several vulnerabilities in TFLite implementation of segment sum
5054 *   Fixes deprecated usage of `collections` API
5060 ## Bug Fixes and Other Changes
5062 *   Fixes an undefined behavior causing a segfault in `tf.raw_ops.Switch`
5064 *   Fixes three vulnerabilities in conversion to DLPack format
5068 *   Fixes two vulnerabilities in `SparseFillEmptyRowsGrad`
5071 *   Fixes an integer truncation vulnerability in code using the work sharder API
5073 *   Fixes a format string vulnerability in `tf.strings.as_string`
5075 *   Fixes segfault raised by calling session-only ops in eager mode
5077 *   Fixes data leak and potential ASLR violation from `tf.raw_ops.StringNGrams`
5079 *   Fixes segfaults caused by incomplete `SavedModel` validation
5081 *   Fixes a data corruption due to a bug in negative indexing support in TFLite
5083 *   Fixes a data corruption due to dimension mismatch in TFLite
5085 *   Fixes several vulnerabilities in TFLite saved model format
5106 ## Bug Fixes and Other Changes
5108 *   Fixes an undefined behavior causing a segfault in `tf.raw_ops.Switch`
5110 *   Fixes three vulnerabilities in conversion to DLPack format
5114 *   Fixes two vulnerabilities in `SparseFillEmptyRowsGrad`
5117 *   Fixes an integer truncation vulnerability in code using the work sharder API
5119 *   Fixes a format string vulnerability in `tf.strings.as_string`
5121 *   Fixes segfault raised by calling session-only ops in eager mode
5123 *   Fixes data leak and potential ASLR violation from `tf.raw_ops.StringNGrams`
5125 *   Fixes segfaults caused by incomplete `SavedModel` validation
5127 *   Fixes a data corruption due to a bug in negative indexing support in TFLite
5129 *   Fixes a data corruption due to dimension mismatch in TFLite
5131 *   Fixes several vulnerabilities in TFLite saved model format
5151 ## Bug Fixes and Other Changes
5153 *   Fixes an undefined behavior causing a segfault in `tf.raw_ops.Switch`
5155 *   Fixes three vulnerabilities in conversion to DLPack format
5159 *   Fixes two vulnerabilities in `SparseFillEmptyRowsGrad`
5162 *   Fixes an integer truncation vulnerability in code using the work sharder API
5164 *   Fixes a format string vulnerability in `tf.strings.as_string`
5166 *   Fixes segfault raised by calling session-only ops in eager mode
5168 *   Fixes data leak and potential ASLR violation from `tf.raw_ops.StringNGrams`
5170 *   Fixes segfaults caused by incomplete `SavedModel` validation
5172 *   Fixes a data corruption due to a bug in negative indexing support in TFLite
5174 *   Fixes a data corruption due to dimension mismatch in TFLite
5176 *   Fixes several vulnerabilities in TFLite saved model format
5191 *   Fixes #41630 by including `max_seq_length` in CuDNN descriptor cache key
5286 ## Bug Fixes and Other Changes
5572 *   Fixes apiclient import.
5629 ## Bug Fixes and Other Changes
5648 *   Fixes a versioning bug which causes Keras layers from TF 1.x to be used
5653 ## Bug Fixes and Other Changes
5675 ## Bug Fixes and Other Changes
5812 ## Bug Fixes and Other Changes
6008 ## Bug Fixes and Other Changes
6010 *   Fixes a security vulnerability where converting a Python string to a
6025 ## Bug Fixes and Other Changes
6027 *   Fixes a security vulnerability where converting a Python string to a
6155 ## Bug Fixes and Other Changes
6158     *   Fixes concurrency issue with `tf.data.experimental.parallel_interleave`
6360 ## Bug Fixes and Other Changes
6368     *   Fixes critical bugs that help with `DenseFeatures` usability in TF2
6558 *   Fixes autocomplete for most TensorFlow API references by switching to use
6693 ## Bug Fixes and Other Changes
7206 ## Bug Fixes and Other Changes
7425 ## Bug Fixes and Other Changes
7434 ## Bug Fixes and Other Changes
7436 *   Fixes a potential security vulnerability where carefully crafted GIF images
7462 ## Bug Fixes and Other Changes
7685 ## Bug Fixes and Other Changes
7777 ## Bug Fixes and Other Changes
7888 ## Bug Fixes and Other Changes
7936 ## Bug Fixes and Other Changes
8103 ## Bug Fixes and Other Changes
8221 ## Bug Fixes and Other Changes
8238         tf.device(“/gpu:0”)`) (Fixes #14133)
8338 ## Bug Fixes and Other Changes
8441 ## Bug Fixes and Other Changes
8589 ## Bug Fixes and Other Changes
8609 *   Bug Fixes:
8734 ## Bug Fixes and Other Changes
8797 ## Bug Fixes and Other Changes
8951 ## Bug Fixes and Other Changes
8953 *   Fixes `strides` and `begin` dtype mismatch when slicing using int64 Tensor
8977 *   Correctness fixes for fft_length parameter to `tf.spectral.rfft` &
9055 ## Bug Fixes and Other Changes
9163 ## Bug Fixes and Other Changes
9312 ## Bug Fixes and Other Changes
9352 *   Multiple tfdbg bug fixes:
9356 *   Java Maven fixes for bugs with Windows installation.
9357 *   Backport fixes and improvements from external keras.
9397 ## Bug Fixes and Other Changes
9401 *   Google Cloud Storage fixes.
9502 ## Bug Fixes and Other Changes
9654 ## Bug Fixes and Other Changes
9726 ## Bug Fixes and Other Changes
9793 ## Bug Fixes and Other Changes
9809 *   Many documentation fixes
9849 ## Bug Fixes and Other Changes
9865 *   Many documentation fixes
9866 *   TensorBoard fixes: graphs with only one data point, Nan values, reload
9909 ## Bug Fixes and Other Changes
9927 *   Added tutorials, many documentation fixes.
9958 ## Bug Fixes and Other Changes
9969 *   Improvements and fixes to Docker image.
9984 ## Bug Fixes and Other Changes
9997 *   Documentation fixes and improvements
10084 ## Bug Fixes
10086 *   Lots of fixes to documentation and tutorials, many contributed by the