VYPR
Unrated severityNVD Advisory· Published Nov 29, 2019· Updated Aug 6, 2024

CVE-2015-0837

CVE-2015-0837

Description

Libgcrypt before 1.6.3 and GnuPG before 1.4.19 expose secret RSA keys via a last-level cache side-channel attack on the mpi_powm function.

AI Insight

LLM-synthesized narrative grounded in this CVE's description and references.

Libgcrypt before 1.6.3 and GnuPG before 1.4.19 expose secret RSA keys via a last-level cache side-channel attack on the mpi_powm function.

Vulnerability

The mpi_powm function in Libgcrypt before version 1.6.3 and GnuPG before version 1.4.19 is vulnerable to a last-level cache (LLC) side-channel attack. The vulnerability arises because the modular exponentiation routine accesses a pre-computed table in a data-dependent pattern, allowing an attacker sharing the same CPU cache to infer secret key bits [1][2][3].

Exploitation

An attacker needs to be co-located on the same physical machine as the victim, able to run unprivileged code that can monitor LLC eviction sets or Prime+Probe timing differences. The core sequence involves performing a cache side-channel attack (e.g., LLC Prime+Probe) while the victim process executes mpi_powm on a secret RSA private key, then statistically analyzing the observed cache access patterns to recover key digits [3].

Impact

Successful exploitation allows the attacker to recover the RSA private key used during the modular exponentiation operation, leading to a complete breach of confidentiality for encrypted communications and digital signatures. The attack does not require any special privileges beyond the ability to execute code on the same hardware [3].

Mitigation

The fix is included in Libgcrypt 1.6.3 and GnuPG 1.4.19, both released on 2015-02-27 [1][2]. The update uses timing-invariant code paths to eliminate data-dependent cache accesses. Users of GnuPG 2.0 or 2.1 must update the shared Libgcrypt library to version 1.6.3; users of GnuPG classic can upgrade to 1.4.19 directly [1]. No workaround is available for unpatched installations.

AI Insight generated on May 26, 2026. Synthesized from this CVE's description and the cited reference URLs; citations are validated against the source bundle.

Affected products

18

Patches

0

No patches discovered yet.

Vulnerability mechanics

No source-code context for this CVE — mechanics is only generated when we can read the actual fix diff. Without that, the four sections (root cause, attack vector, affected code, fix) would be speculation rather than analysis.

References

5

News mentions

0

No linked articles in our index yet.