{"id":"CVE-2022-23593","aliases":["GHSA-gwcx-jrx4-92w2","BIT-tensorflow-2022-23593","PYSEC-2022-102","PYSEC-2022-157","PYSEC-2026-3193"],"title":"Segfault in `simplifyBroadcast` in Tensorflow","summary":"Segfault in `simplifyBroadcast` in Tensorflow","severity":"medium","cvss":5.9,"cvssVector":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H","vendor":"tensorflow","product":"tensorflow","ecosystem":"pip","affected":["tensorflow >= 2.8.0-rc0, < 2.8.0","tensorflow-cpu >= 2.8.0-rc0, < 2.8.0","tensorflow-gpu >= 2.8.0-rc0, < 2.8.0"],"patched":["tensorflow 2.8.0","tensorflow-cpu 2.8.0","tensorflow-gpu 2.8.0"],"published":"2022-02-09","updated":"2026-07-13","source":"OSV","sourceUrl":"https://osv.dev/vulnerability/GHSA-gwcx-jrx4-92w2","references":[{"url":"https://github.com/tensorflow/tensorflow/security/advisories/GHSA-gwcx-jrx4-92w2"},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2022-23593"},{"url":"https://github.com/tensorflow/tensorflow/commit/35f0fabb4c178253a964d7aabdbb15c6a398b69a"},{"url":"https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-cpu/PYSEC-2022-102.yaml"},{"url":"https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-gpu/PYSEC-2022-157.yaml"},{"url":"https://github.com/tensorflow/tensorflow"},{"url":"https://github.com/tensorflow/tensorflow/blob/274df9b02330b790aa8de1cee164b70f72b9b244/tensorflow/compiler/mlir/tfrt/jit/transforms/tf_cpurt_symbolic_shape_optimization.cc#L149-L205"}],"tags":["osv","pip"],"epss":0.0087,"epssPercentile":0.57285,"ingestedAt":"2026-07-13T18:57:58.633Z","slug":"CVE-2022-23593","body":"## Overview\n\n### Impact\nThe [`simplifyBroadcast` function in the MLIR-TFRT infrastructure in TensorFlow](https://github.com/tensorflow/tensorflow/blob/274df9b02330b790aa8de1cee164b70f72b9b244/tensorflow/compiler/mlir/tfrt/jit/transforms/tf_cpurt_symbolic_shape_optimization.cc#L149-L205) is vulnerable to a segfault (hence, denial of service), if called with scalar shapes.\n\n```cc \n  size_t maxRank = 0;\n  for (auto shape : llvm::enumerate(shapes)) {\n    auto found_shape = analysis.dimensionsForShapeTensor(shape.value());\n    if (!found_shape) return {};\n    shapes_found.push_back(*found_shape);\n    maxRank = std::max(maxRank, found_shape->size());\n  }   \n\n  SmallVector<const ShapeComponentAnalysis::SymbolicDimension*>\n      joined_dimensions(maxRank);\n```\n\nIf all shapes are scalar, then `maxRank` is 0, so we build an empty `SmallVector`.\n\n### Patches\nWe have patched the issue in GitHub commit [35f0fabb4c178253a964d7aabdbb15c6a398b69a](https://github.com/tensorflow/tensorflow/commit/35f0fabb4c178253a964d7aabdbb15c6a398b69a).\n\nThe fix will be included in TensorFlow 2.8.0. This is the only affected version.\n\n### For more information\nPlease consult [our security guide](https://github.com/tensorflow/tensorflow/blob/master/SECURITY.md) for more information regarding the security model and how to contact us with issues and questions.\n\n## Affected packages\n\n- `tensorflow >= 2.8.0-rc0, < 2.8.0`\n- `tensorflow-cpu >= 2.8.0-rc0, < 2.8.0`\n- `tensorflow-gpu >= 2.8.0-rc0, < 2.8.0`\n\n## Remediation\n\nUpgrade to a patched release:\n\n- `tensorflow 2.8.0`\n- `tensorflow-cpu 2.8.0`\n- `tensorflow-gpu 2.8.0`","depth":"sunlit","depthScore":33,"depthScoreParts":{"impact":32.5,"likelihood":0.2,"exploitation":0,"ransomware":0},"changes":[]}