OpenSSL have published a security advisory. Here's how it affects BoringSSL:
| CVE | Summary | Severity in OpenSSL | Impact to BoringSSL |
|---|---|---|---|
| CVE-2026-45447 | Heap Use-After-Free in the PKCS7_verify() Function | High | Not affected, impacted code was removed from BoringSSL in the initial fork |
| CVE-2026-34182 | CMS AuthEnvelopedData Processing May Accept Forged Messages | Moderate | Not affected, impacted code was removed from BoringSSL in the initial fork |
| CVE-2026-34183 | Unbounded Memory Growth in the QUIC PATH_CHALLENGE Handler | Moderate | Not affected, issue was introduced after fork |
| CVE-2026-35188 | Double-free When Checking OCSP Stapled Response | Moderate | Not affected, issue was introduced after fork |
| CVE-2026-42764 | NULL Pointer Dereference in QUIC Server Initial Packet Handling | Moderate | Not affected, issue was introduced after fork |
| CVE-2026-45445 | AES-OCB IV Ignored on EVP_Cipher() Path | Moderate | Not affected, issue was introduced after fork |
| CVE-2026-7383 | Possible Heap Buffer Overflow in ASN.1 Multibyte String Conversion | Low | Not affected, independently fixed in 2018 |
| CVE-2026-9076 | Out-of-Bounds Read in CMS Password-Based Decryption | Low | Not affected, impacted code was removed from BoringSSL in the initial fork |
| CVE-2026-34180 | Heap Buffer Over-read in ASN.1 Content Parsing | Low | Not affected, independently fixed in 2017. See discussion below. |
| CVE-2026-34181 | PKCS#12 Files with PBMAC1 Are Accepted with Short HMAC Keys | Low | Not affected, issue was introduced after fork |
| CVE-2026-42765 | NULL Dereference in Certificate Verification with OCSP Checking | Low | Not affected, issue was introduced after fork |
| CVE-2026-42766 | Possible NULL Dereference in Password-Based CMS Decryption | Low | Not affected, impacted code was removed from BoringSSL in the initial fork |
| CVE-2026-42767 | NULL Pointer Dereference in CRMF EncryptedValue Decryption | Low | Not affected, issue was introduced after fork |
| CVE-2026-42768 | Multi-RecipientInfo Bleichenbacher Oracle in CMS_decrypt() and PKCS7_decrypt() | Low | Not affected, impacted code was removed from BoringSSL in the initial fork |
| CVE-2026-42769 | Trust-Anchor Substitution via cert/issuer Typo in CMP rootCaKeyUpdate | Low | Not affected, issue was introduced after fork |
| CVE-2026-42770 | FFC-DH Peer Validation Uses Attacker-Supplied q | Low | Not affected, issue was introduced after fork |
| CVE-2026-42771 | Possible Out of Bounds Read in X509_VERIFY_PARAM_set1_email() | Low | Not affected, issue was introduced after fork |
| CVE-2026-45446 | Incorrect Tag Processing for Empty Messages in AES-GCM-SIV and AES-SIV modes | Low | Not affected, issue was introduced after fork |
This issue concerns the table-based crypto/asn1 module in OpenSSL. At the start of the project, we identified this module's parser as especially risky, particularly around mismatched use of integers, type-safety, and error-prone bounds-checking conventions. We wrote crypto/bytestring, a simpler ASN.1 and TLS parsing library, and have gradually replaced our parsing with the new library.
Although the crypto/asn1 parser is still in BoringSSL, in 2017 we capped the input size to mitigate its integer overflows, including CVE-2026-34180.
Over time, we've documented and trimmed the library, removed particularly complex parts of the parser, and fixed integer truncations at the API boundary. Building on this, we plan to reimplement the underlying table-processing logic with crypto/bytestring to further harden this code.