ŠVENDA, Petr, Antonín DUFKA, Milan BROŽ, Roman LACKO, Tomáš JAROŠ, Daniel ZAŤOVIČ and Josef POSPISIL. TPMScan: A wide-scale study of security-relevant properties of TPM 2.0 chips. Online. In IACR Transactions on Cryptographic Hardware and Embedded Systems. Bochum: Ruhr-University of Bochum, 2024, p. 714-734. ISSN 2569-2925. Available from: https://dx.doi.org/10.46586/tches.v2024.i2.714-734.
Other formats:   BibTeX LaTeX RIS
Basic information
Original name TPMScan: A wide-scale study of security-relevant properties of TPM 2.0 chips
Authors ŠVENDA, Petr (203 Czech Republic, guarantor, belonging to the institution), Antonín DUFKA (203 Czech Republic, belonging to the institution), Milan BROŽ (203 Czech Republic, belonging to the institution), Roman LACKO (703 Slovakia, belonging to the institution), Tomáš JAROŠ (703 Slovakia, belonging to the institution), Daniel ZAŤOVIČ (703 Slovakia) and Josef POSPISIL (203 Czech Republic).
Edition Bochum, IACR Transactions on Cryptographic Hardware and Embedded Systems, p. 714-734, 21 pp. 2024.
Publisher Ruhr-University of Bochum
Other information
Original language English
Type of outcome Proceedings paper
Field of Study 10200 1.2 Computer and information sciences
Country of publisher Germany
Confidentiality degree is not subject to a state or trade secret
Publication form electronic version available online
WWW URL
Organization unit Faculty of Informatics
ISSN 2569-2925
Doi http://dx.doi.org/10.46586/tches.v2024.i2.714-734
Keywords in English TPM; common criteria; fips140; RSA; ECDSA;
Tags International impact, Reviewed
Changed by Changed by: doc. RNDr. Petr Švenda, Ph.D., učo 4085. Changed: 16/4/2024 16:43.
Abstract
The Trusted Platform Module (TPM) is a widely deployed computer component that provides increased protection of key material during cryptographic operations, secure storage, and support for a secure boot with a remotely attestable state of the target machine. A systematic study of the TPM ecosystem, its cryptographic properties, and the orderliness of vulnerability mitigation is missing despite its pervasive deployment -- likely due to the black-box nature of the implementations. We collected metadata, RSA and ECC cryptographic keys, and performance characteristics from \numfws{} different TPM versions manufactured by \numvendors{} vendors, including recent Pluton-based iTPMs, to systematically analyze TPM implementations. Surprisingly, a high rate of changes with a detectable impact on generated secrets, the timing of cryptographic operations, and frequent off-chip generation of Endorsement Keys were observed. Our analysis of public artifacts for TPM-related products certified under Common Criteria (CC) and FIPS 140 showed relatively high popularity of TPMs but without explanation for these changes in cryptographic implementations. Despite TPMs being commonly certified to CC EAL4+, serious vulnerabilities like ROCA or TPM-Fail were discovered in the past. We found a range of additional unreported nonce leakages in ECDSA, ECSCHNORR, and ECDAA algorithms in dTPMs and fTPMs of three vendors. The most serious discovered leakage allows extraction of the private key of certain Intel's fTPM versions using only nine signatures with no need for any side-channel information, making the vulnerability retrospectively exploitable despite a subsequent firmware update. Unreported timing leakages were discovered in the implementations of ECC algorithms on multiple Nuvoton TPMs, and other previously reported leakages were confirmed. The analysis also unveiled incompleteness of vulnerability reporting and subsequent mitigation with missing clear information about the affected versions and inconsistent fixes.
Links
MUNI/A/1586/2023, interní kód MUName: Aplikovaný výzkum na FI: Forenzní aspekty kritických infrastruktur, aplikovaná kryptografie, kyberbezpečnostní cvičení, algoritmy plánování v logistice a pro zpracování dat z fyzikálních sensorů
Investor: Masaryk University, Applied research at FI: Forensic aspects of critical infrastructures, applied cryptography, cybersecurity trainings, scheduling algorithms logistics and algorithms for physical sensors
VJ02010010, research and development projectName: Nástroje pro verifikaci bezpečnosti kryptografických zařízení s využitím AI (Acronym: AI-SecTools)
Investor: Ministry of the Interior of the CR
PrintDisplayed: 27/7/2024 13:56