csirt@kutim:~/threat-intel/cve/CVE-2021-37650

Katalog CVE

CVSS 7.8 · High EPSS 0,18%

Kerentanan Buffer Copy without Checking Size of Input ('Classic Buffer Overflow') pada google tensorflow. Skor CVSS 3.1 7.8 (High) menurut NVD.

Deskripsi (NVD)

TensorFlow is an end-to-end open source platform for machine learning. In affected versions the implementation for `tf.raw_ops.ExperimentalDatasetToTFRecord` and `tf.raw_ops.DatasetToTFRecord` can trigger heap buffer overflow and segmentation fault. The [implementation](https://github.com/tensorflow/tensorflow/blob/f24faa153ad31a4b51578f8181d3aaab77a1ddeb/tensorflow/core/kernels/data/experimental/to_tf_record_op.cc#L93-L102) assumes that all records in the dataset are of string type. However, there is no check for that, and the example given above uses numeric types. We have patched the issue in GitHub commit e0b6e58c328059829c3eb968136f17aa72b6c876. The fix will be included in TensorFlow 2.6.0. We will also cherrypick this commit on TensorFlow 2.5.1, TensorFlow 2.4.3, and TensorFlow 2.3.4, as these are also affected and still in supported range.

Rincian

Terbit di NVD
12 Agustus 2021
Terakhir diubah
17 Juni 2026
Status NVD
Modified
CVSS 3.1
7.8 (High), sumber NVD
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H · uraikan di kalkulator
Jenis kelemahan (CWE)
CWE-120: Buffer Copy without Checking Size of Input ('Classic Buffer Overflow')
CWE-787: Out-of-bounds Write
EPSS (30 hari)
0,18% peluang dieksploitasi; lebih tinggi dari 7,2% CVE lain

Produk dan versi terdampak

VendorProdukVersi terdampak
google tensorflow Versi 2.3.0 hingga sebelum 2.3.4
google tensorflow Versi 2.4.0 hingga sebelum 2.4.3
google tensorflow Versi 2.5.0
google tensorflow Versi 2.6.0 rc0
google tensorflow Versi 2.6.0 rc1
google tensorflow Versi 2.6.0 rc2

> Data dari konfigurasi CPE di NVD.

Referensi dan perbaikan

Sumber: NVD · CVE.org · EPSS. Halaman ini disusun otomatis dari data tersebut; pembaruan terakhir 5 Okt 2026, 20:10 WITA. Kalau sistem di Perangkat Daerah memakai produk ini dan butuh bantuan, laporkan lewat kanal pelaporan.