Pith. sign in

REVIEW 3 major objections 4 minor 86 references

Certified Side Channels

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read The format a private key is stored in can silently drop side-channel defenses and reveal the key.

desk verdict Key serialization formats can silently bypass SCA countermeasures in OpenSSL and mbedTLS—the attacks are real, but the abstract overstates cache-timing key recovery and real-world prevalence is not measured. read the letter →

arxiv 1909.01785 v2 pith:KA4EPKHQ submitted 2019-09-04 cs.CR

classification cs.CR
keywords side-channelanalysisprivatekeyformatsECDSAOpenSSLmbedTLSwNAFscalarmultiplicationlatticeattackstiming
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that the serialized format of a private key is itself a side-channel attack surface: two keys that are mathematically equivalent at the specification level can make a library execute very different arithmetic code, some of it unprotected. It demonstrates this in OpenSSL and mbedTLS with four distinct vulnerabilities, ranging from ECDSA keys carrying explicit curve parameters with a missing or zero cofactor, to DSA keys in PVK/MSBLOB form, to RSA keys whose optional parameters are absent. In the strongest case it recovers the long-term ECDSA private key of a signing server end to end, using only network response timing over Ethernet, and also using electromagnetic emanations. A sympathetic reader should care because the mitigations in individual primitives are bypassed without modifying any code or abusing invalid keys: the keys are standards-compliant and the libraries parse them normally.

What carries the argument

The load-bearing mechanism is runtime dispatch inside the key parser: the value of an optional ASN.1 field selects between a constant-time scalar-multiplication implementation and an older wNAF implementation. The specific switch is the cofactor in an explicit EC domain parameter block: OpenSSL's scalar-multiplication wrapper only takes the constant-time ladder when both the group order and the cofactor are nonzero, so a zero or omitted cofactor routes P-256 signing into ec_wNAF_mul, whose execution time and cache or EM profile vary with the secret nonce. In the RSA and DSA cases the same dispatch idea operates through key-format code paths that call variable-time bignum routines such as BN_mod_exp, BN_gcd, and BN_mod_inverse, and mbedTLS's binary GCD, without setting the library's constant-time flag.

What would settle it

Run the paper's 500,000-request timing attack against the same time-stamp server configured with an otherwise identical named-curve P-256 private key: if key recovery still succeeds at the reported rate, the claimed mechanism is wrong, and if it fails, the key format is confirmed as the load-bearing cause.

Watch

Extended reading notes

Core claim

The central discovery is that standards themselves, via optional fields in key encodings, determine which arithmetic routine runs on secret data. OpenSSL's ECDSA path normally uses a constant-time, architecture-optimized implementation for named curves, but a valid private key with explicit domain parameters falls to the generic method; if the optional cofactor is zero or omitted, the generic method skips its constant-time ladder and executes a wNAF double-and-add that leaks the nonce. The same pattern repeats elsewhere: PVK and MSBLOB DSA keys force recomputation of the public key with variable-time modular exponentiation; RSA key validation in OpenSSL recomputes secret components with variable-time GCD and inversion; and mbedTLS, when RSA parameters are missing, derives CRT values and the private exponent with the binary extended Euclidean algorithm. The paper completes the chain by turning the ECDSA leak into key recovery: a timing attack that filters short nonces from 500,000 signatures and solves a hidden-number-problem lattice with a 91 percent success rate, and an EM attack that recovers the key from 172 usable traces in about three minutes.

Load-bearing premise

The practical relevance of the attacks depends on standards-compliant but atypical key encodings appearing in real deployments; the paper itself concedes that OpenSSL's own command-line tool will not organically emit the vulnerable ECDSA key form and instead points to engines, wallets, and conversion workflows.

Editorial extensions

If this is right

  • A server that signs with a valid explicit-parameter EC key whose cofactor is zero or omitted can be fully key-recovered by a remote client merely measuring response times; the paper's Ethernet experiment succeeded in 91 of 100 attack instances.
  • Any workflow that loads DSA keys in PVK or MSBLOB form into OpenSSL re-exposes the cache-timing leak in modular exponentiation, since the public key is recomputed from the private key on every load.
  • Calling OpenSSL's RSA key-checking API on a legitimate RSA key is enough to leak bits of $p$ and $q$; the affected functions include both the direct check and the generic EVP_PKEY_check path, so any OpenSSL-linking application that validates keys inherits the exposure.
  • mbedTLS key loading is vulnerable even for completely standard, fully populated RSA keys, because the CRT-deduction path runs on every load; the paper groups all 256 possible PKCS#1 parameter combinations into four classes, with 32 of them hitting a leaking code path.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper, the same parser-dispatch pattern should be searchable by fuzzing optional fields in other serialization standards; a library that canonicalizes keys to internal named-curve representations at import time would close the entire class rather than patching one routine.
  • The timing attack's reliance on nonce-bitlength correlation suggests that other signing protocols that expose a network timing oracle, not just time-stamping, are viable candidates if they run on the same vulnerable parsing path.
  • If standards bodies made cofactor and CRT parameters mandatory, or required import-time validation against canonical values, the specification-level ambiguity this paper exploits would be largely eliminated at the source.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper argues that the format in which private keys are persisted, and in particular the presence or absence of optional parameters in standardized encodings, materially affects side-channel security. Using the Triggerflow execution-path tool, the authors identify four vulnerability classes in widely used libraries: (i) OpenSSL ECDSA with explicit curve parameters and a zero or omitted cofactor reaches the SCA-insecure wNAF scalar multiplication path (Section 3.1); (ii) OpenSSL DSA public-key derivation from PVK and MSBLOB key formats performs variable-time modular exponentiation (Section 3.2); (iii) OpenSSL RSA key validation exercises leaking primality, GCD, and inversion paths (Section 3.3); and (iv) mbedTLS RSA key loading computes missing parameters using leaking GCD and modular inversion routines (Section 3.4). The paper then reports two end-to-end ECDSA key recovery attacks on a time-stamping server using the Section 3.1 trigger: a remote wall-clock timing attack with 91% success over 100 runs, and an electromagnetic attack recovering the key in about three minutes from 172 usable traces. The authors responsibly disclosed the findings, obtained CVE-2019-1547, and contributed patches to OpenSSL and mbedTLS.

Significance. If the central claim is accepted, the paper makes a valuable contribution by moving side-channel analysis from isolated primitives to the key parsing and persistence layer, showing that standards-level optionality can silently switch a library onto an SCA-vulnerable code path. The end-to-end attacks use unmodified OpenSSL 1.1.1 and an unmodified TSA server, and the timing attack operates over a network with a threat model indistinguishable from a normal client, which makes the results concrete. The paper also ships a public dataset for the remote timing attack and integrates the new Triggerflow tests into OpenSSL's CI, which are concrete reproducibility and tooling strengths. The main weakness is that the 'real world' scope of the central claim rests on qualitative, anecdotal evidence about deployment prevalence rather than measurement; the technical vulnerability mechanisms themselves are credible and well supported.

major comments (3)
  1. [Section 3.1, 'Keys in the wild'; Section 5] The paper's central claim that key format 'plays a significant role in real world SCA security' depends on the real-world prevalence of the triggering key encodings, but this prevalence is asserted, not measured. The text concedes that the OpenSSL CLI will not organically emit the vulnerable ECDSA key form, and the cited instances (GOST engine, GOSTCoin, OpenSC, Windows driver signing) are anecdotal. No deployment survey, telemetry, or quantitative estimate is provided for any of these cases. If such keys are rare in practice, the demonstrated ECDSA attacks affect a much narrower population than the title and abstract suggest. The authors should either temper the 'real world' conclusion to 'can affect deployments that accept standards-parseable but atypical key encodings' or supply measurable evidence of prevalence.
  2. [Section 3.4, 'RSA: Bypass via Missing Parameters'] The section title and the surrounding framing imply that the mbedTLS vulnerability requires missing RSA parameters, but the paper itself states that the CRT-related leakage 'exists regardless of missing parameters in the private key' and that it executes on every private-key load. Only the private-exponent leakage in the 'CRT & d' group is specific to a key missing d. This is a scope-of-claim issue: the universal CRT leak is a separate, arguably more serious finding, and the format-specific contribution should be presented as the additional d-computation path rather than as a single 'bypass via missing parameters' vulnerability.
  3. [Abstract; Section 4] The abstract says the authors 'design and implement key recovery attacks utilizing signals ranging from electromagnetic (EM) emanations, to granular microarchitecture cache timings, to coarse traditional wall clock timings.' This implies that full key recovery is achieved with cache-timing signals as well as the other two. In the paper, full end-to-end key recovery is demonstrated for remote wall-clock timing (Section 4.1) and EM (Section 4.2); the microarchitecture cache-timing analysis in Sections 3.2 and 3.3 demonstrates leakage and partial exponent recovery but not completed key recovery for a full signature scheme. The abstract and introduction should be reworded to distinguish leakage analysis from completed key recovery.
minor comments (4)
  1. [Section 3.1] Calling a key with cofactor explicitly set to zero 'valid from a standards perspective' is too strong: SEC1 allows the cofactor to be optional, but if present it should be the correct positive integer. 'Standards-parseable' or 'accepted by the implementation' would be more accurate and would not weaken the attack.
  2. [Section 2.2, PVK key format] The PVK magic value is stated as '0xb0b5f11e2' with a footnote referring to 'bobsfile'; the standard PVK magic is 0xb0b5f11e. Please correct the value and remove the stray digit.
  3. [Global] There are several typos and awkward phrasings: 'trasmit' in Section 2.2, 'Impulsed' in Section 2.3, 'lattic errors' in Section 5, and the unexplained 'f_in matrix' label in Figure 2. These should be cleaned up.
  4. [Section 4.2, Signal analysis] The trace-quality criterion is described as a 'emphatically selected confidence interval' for the peak-to-RMS ratio. Please give the actual threshold or selection procedure, since this step affects the reported 422/500 trace retention rate and the final success rate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the paper's claims are supported by empirical demonstrations against unmodified external libraries.

full rationale

The paper is fundamentally empirical: it demonstrates concrete side-channel vulnerabilities in unmodified OpenSSL and mbedTLS, and completes end-to-end key recovery attacks. The central claim that key format affects SCA security is supported by direct measurement (Triggerflow path tracing, cache-timing traces, EM traces, and wall-clock timings) rather than by a derivation that assumes its own conclusion. Lattice parameters such as W, d, and j are tuned for attack efficiency from training samples and reported with success rates; this is standard attack engineering, not a fitted parameter renamed as a prediction. The paper's self-citations (Triggerflow [40], lattice formalization [61], and prior timing-attack methodology [22]) are either tooling descriptions or previously published peer-reviewed results with independent content; none serves as an unverified uniqueness theorem or smuggled ansatz that forces the central result. The 'Keys in the wild' discussion is anecdotal and the real-world prevalence of the triggering encodings is asserted rather than measured, but that is a correctness or impact concern, not circularity: the demonstrated attacks remain valid against the tested configurations regardless of how common those configurations are in practice. No equation in the paper reduces to its inputs by construction, and no prediction is equivalent to a fitted value. The derivation chain is therefore self-contained with respect to circularity, and the appropriate score is 0.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claims rest on the validity of the tested key encodings and the stability of the side-channel measurements. The free parameters listed are attack-tuning constants, not scientific parameters being predicted. No new physical or mathematical entities are introduced.

free parameters (4)
  • W (nonce MSB bound) = 2^7 for the timing attack
    Set from an empirical Gaussian fit of the nonce bitlength distribution (mean 247.80, s.d. 3.81) after filtering the 128 shortest traces. It is an attack tuning constant chosen to trade off lattice dimension and success rate, not a scientific claim.
  • d (lattice dimension) = 60 for timing attack, 120 for EM attack
    Number of signatures used to construct each lattice instance; chosen empirically to balance leakage information and computation cost.
  • j (parallel job count) = 55,000 for timing attack, 48 for EM attack
    Number of random d-subsets attempted to compensate for error probability in the collection phase; standard HNP lattice attack practice.
  • EM signal processing thresholds = not numerically specified (ratio, Euclidean distance thresholds)
    Experimental thresholds for trace filtering, end-trigger template matching, and Add-operation peak classification. The paper describes the method and qualitative bounds but does not publish exact values.
assumptions (4)
  • domain assumption The target OpenSSL and mbedTLS versions are representative of real-world deployments.
    The analyses target OpenSSL 1.1.1 and mbedTLS 2.18.1. The paper argues these are widely deployed but does not quantify market share; if deployments are different, the specific vulnerabilities may not transfer.
  • domain assumption The hardware side-channel behavior is stable enough for the described attacks.
    Section 3.2 notes cache sets remain constant across invocations, and the EM attack relies on template matching. Variability in cache or EM behavior would change the attack parameters and success rates.
  • standard math The lattice attack formulation from prior work applies to the nonce leakage.
    Section 2.4 builds on the hidden number problem formalization by Nguyen-Shparlinski and Benger et al.; the paper cites these as established results and adapts them to MSB leakage.
  • domain assumption Keys with explicit EC parameters and zero or omitted cofactor are valid per the relevant standards and accepted by OpenSSL.
    Section 3.1 states that all three constructed keys are valid from a standards perspective; this validity is the precondition for the countermeasure bypass.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Certified Side Channels." pith.science (2026). https://pith.science/paper/KA4EPKHQ

@misc{pith2026190901785,
  author       = {Pith},
  title        = {Pith review of: Certified Side Channels},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KA4EPKHQ}},
  note         = {Machine review of arXiv:1909.01785}
}
read the original abstract

We demonstrate that the format in which private keys are persisted impacts Side Channel Analysis (SCA) security. Surveying several widely deployed software libraries, we investigate the formats they support, how they parse these keys, and what runtime decisions they make. We uncover a combination of weaknesses and vulnerabilities, in extreme cases inducing completely disjoint multi-precision arithmetic stacks deep within the cryptosystem level for keys that otherwise seem logically equivalent. Exploiting these vulnerabilities, we design and implement key recovery attacks utilizing signals ranging from electromagnetic (EM) emanations, to granular microarchitecture cache timings, to coarse traditional wall clock timings.

Figures

Figures reproduced from arXiv: 1909.01785 by the authors.

Figure 1
Figure 1. New Triggerflow unit tests. OpenSSL); (iv) bypassing SCA countermeasures for RSA through key loading (Section 3.4, mbedTLS). 3.1 ECC: Bypass via Explicit Parameters From a standardization perspective, curve data for ECC key material gets persisted in one of two ways: either including the specific OID that points to a named curve with fixed pa￾rameters, or explicitly specifying the curve with ASN.1 syn￾tax. Mathemati… view at source ↗
Figure 2
Figure 2. L1 dcache trace showing distinctive access patterns [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Sequence of right-shifts and subtractions from a [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Direct correlation between wall-clock execution [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Experiment setup capturing EM traces using Pico [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Four different EM traces showing the last Add (A) [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

86 extracted references · 62 canonical work pages

  1. [1]

    FIPS PUB 140-2, National Institute of Standards and Technology, May 2001

    Security requirements for cryptographic modules. FIPS PUB 140-2, National Institute of Standards and Technology, May 2001. URL https://doi.org/10.6028/NIST.FIPS. 140-2

  2. [2]

    SEC 1, Standards for Efficient Cryptography Group, May 2009

    SEC 1: Elliptic Curve Cryptography. SEC 1, Standards for Efficient Cryptography Group, May 2009. URL http://www. secg.org/sec1-v2.pdf

  3. [3]

    Survey for performance & security problems of passive side-channel attacks countermeasures in ECC

    Rodrigo Abarzúa, Claudio Valencia Cordero, and Julio López Hernandez. Survey for performance & security problems of passive side-channel attacks countermeasures in ECC. IACR Cryptology ePrint Archive, 2019(10), 2019. URL https:// eprint.iacr.org/2019/010

  4. [4]

    New branch prediction vulnerabilities in OpenSSL and necessary software countermeasures

    Onur Acıiçmez, Shay Gueron, and Jean-Pierre Seifert. New branch prediction vulnerabilities in OpenSSL and necessary software countermeasures. In IMACC, volume 4887 of LNCS, pages 185–203. Springer, 2007. URL https://doi.org/10. 1007/978-3-540-77272-9_12

  5. [5]

    Rao, and Pankaj Rohatgi

    Dakshi Agrawal, Bruce Archambeault, Josyula R. Rao, and Pankaj Rohatgi. The EM side-channel(s). In CHES, volume 2523 of LNCS, pages 29–45. Springer, 2002. URL https: //doi.org/10.1007/3-540-36400-5_4

  6. [6]

    When one vulnerable primitive turns viral: Novel single-trace attacks on ECDSA and RSA

    Alejandro Cabrera Aldaya and Billy Bob Brumley. When one vulnerable primitive turns viral: Novel single-trace attacks on ECDSA and RSA. IACR Trans. Cryptogr. Hardw. Embed. Syst., 2020(2):196–221, 2020. URL https://doi.org/10. 13154/tches.v2020.i2.196-221

  7. [7]

    Cabrera Sarmiento, and Santiago Sánchez-Solano

    Alejandro Cabrera Aldaya, Alejandro J. Cabrera Sarmiento, and Santiago Sánchez-Solano. SPA vulnerabilities of the bi- nary extended Euclidean algorithm. J. Cryptographic Engi- neering, 7(4):273–285, 2017. URL https://doi.org/10. 1007/s13389-016-0135-4

  8. [8]

    Cabrera Sarmiento, and Santiago Sánchez-Solano

    Alejandro Cabrera Aldaya, Raudel Cuiman Márquez, Alejan- dro J. Cabrera Sarmiento, and Santiago Sánchez-Solano. Side- channel analysis of the modular inversion step in the RSA key generation algorithm. I. J. Circuit Theory and Applications, 45 (2):199–213, 2017. URL https://doi.org/10.1002/cta. 2283

Show all 86 references
  1. [9]

    Port contention for fun and profit

    Alejandro Cabrera Aldaya, Billy Bob Brumley, Sohaib ul Has- san, Cesar Pereida García, and Nicola Tuveri. Port contention for fun and profit. In IEEE S&P, pages 870–887. IEEE, 2019. URL https://doi.org/10.1109/SP.2019.00066

  2. [10]

    Cache-timing attacks on RSA key generation

    Alejandro Cabrera Aldaya, Cesar Pereida García, Luis Manuel Alvarez Tapia, and Billy Bob Brumley. Cache-timing attacks on RSA key generation. IACR Trans. Cryptogr. Hardw. Embed. Syst., 2019(4):213–242, 2019. URL https://doi.org/10. 13154/tches.v2019.i4.213-242

  3. [11]

    Falkner, Joop van de Pol, and Yuval Yarom

    Thomas Allan, Billy Bob Brumley, Katrina E. Falkner, Joop van de Pol, and Yuval Yarom. Amplifying side channels through performance degradation. In ACSAC, pages 422–

  4. [12]

    Horizontal collision correlation attack on elliptic curves – extended version

    Aurélie Bauer, Éliane Jaulmes, Emmanuel Prouff, Jean-René Reinhard, and Justine Wild. Horizontal collision correlation attack on elliptic curves – extended version. Cryptography and Communications, 7(1):91–119, 2015. URL https://doi. org/10.1007/s12095-014-0111-8

  5. [13]

    Side-channel analysis of Weierstrass and Koblitz curve ECDSA on Android smartphones

    Pierre Belgarric, Pierre-Alain Fouque, Gilles Macario-Rat, and Mehdi Tibouchi. Side-channel analysis of Weierstrass and Koblitz curve ECDSA on Android smartphones. In CT-RSA, volume 9610 of LNCS, pages 236–252. Springer, 2016. URL https://doi.org/10.1007/978-3-319-29485-8_14

  6. [14]

    Ooh Aah... Just a Little Bit

    Naomi Benger, Joop van de Pol, Nigel P. Smart, and Yuval Yarom. “Ooh Aah... Just a Little Bit”: A small amount of side channel can go a long way. In CHES, volume 8731 of LNCS, pages 75–92. Springer, 2014. URL https://doi.org/10. 1007/978-3-662-44709-3_5

  7. [15]

    Bernstein

    Daniel J. Bernstein. Cache-timing attacks on AES, 2005. URL http://cr.yp.to/papers.html#cachetiming

  8. [16]

    Bernstein and Bo-Yin Yang

    Daniel J. Bernstein and Bo-Yin Yang. Fast constant-time gcd computation and modular inversion. IACR Trans. Cryptogr. Hardw. Embed. Syst., 2019(3):340–398, 2019. URL https: //doi.org/10.13154/tches.v2019.i3.340-398

  9. [17]

    Bernstein, Joachim Breitner, Daniel Genkin, Leon Groot Bruinderink, Nadia Heninger, Tanja Lange, Chris- tine van Vredendaal, and Yuval Yarom

    Daniel J. Bernstein, Joachim Breitner, Daniel Genkin, Leon Groot Bruinderink, Nadia Heninger, Tanja Lange, Chris- tine van Vredendaal, and Yuval Yarom. Sliding right into disaster: Left-to-right sliding windows leak. In CHES, vol- ume 10529 of LNCS, pages 555–576. Springer, 20...

  10. [18]

    Hardness of com- puting the most significant bits of secret keys in Diffie-Hellman and related schemes

    Dan Boneh and Ramarathnam Venkatesan. Hardness of com- puting the most significant bits of secret keys in Diffie-Hellman and related schemes. InCRYPTO, volume 1109 ofLNCS, pages 15 129–142. Springer, 1996. URL https://doi.org/10.1007/ 3-540-68697-5_11

  11. [19]

    Correlation power analysis with a leakage model

    Eric Brier, Christophe Clavier, and Francis Olivier. Correlation power analysis with a leakage model. In CHES, volume 3156 of LNCS, pages 16–29. Springer, 2004. URL https://doi. org/10.1007/978-3-540-28632-5_2

  12. [20]

    Faster software for fast endomor- phisms

    Billy Bob Brumley. Faster software for fast endomor- phisms. In COSADE, volume 9064 of LNCS, pages 127–

  13. [21]

    Billy Bob Brumley and Risto M. Hakala. Cache-timing tem- plate attacks. In ASIACRYPT, volume 5912 of LNCS, pages 667–684. Springer, 2009. URL https://doi.org/10.1007/ 978-3-642-10366-7_39

  14. [22]

    Remote timing attacks are still practical

    Billy Bob Brumley and Nicola Tuveri. Remote timing attacks are still practical. In ESORICS, volume 6879 of LNCS, pages 355–371. Springer, 2011. URL https://doi.org/10.1007/ 978-3-642-23822-2_20

  15. [23]

    Remote timing at- tacks are practical

    David Brumley and Dan Boneh. Remote timing at- tacks are practical. In USENIX Sec. USENIX As- sociation, 2003. URL https://www.usenix.org/ conference/12th-usenix-security-symposium/ remote-timing-attacks-are-practical

  16. [24]

    Fallout: Leaking data on meltdown-resistant CPUs

    Claudio Canella, Daniel Genkin, Lukas Giner, Daniel Gruss, Moritz Lipp, Marina Minkin, Daniel Moghimi, Frank Piessens, Michael Schwarz, Berk Sunar, Jo Van Bulck, and Yuval Yarom. Fallout: Leaking data on meltdown-resistant CPUs. In ACM CCS, pages 769–784. ACM, 2019. URL https:...

  17. [25]

    Rao, and Pankaj Rohatgi

    Suresh Chari, Josyula R. Rao, and Pankaj Rohatgi. Template at- tacks. In CHES, volume 2523 of LNCS, pages 13–28. Springer,

  18. [26]

    Horizontal correlation analy- sis on exponentiation

    Christophe Clavier, Benoit Feix, Georges Gagnerot, Mylène Roussellet, and Vincent Verneuil. Horizontal correlation analy- sis on exponentiation. In ICICS, volume 6476 of LNCS, pages 46–61. Springer, 2010. URL https://doi.org/10.1007/ 978-3-642-17650-0_5

  19. [27]

    A synthesis of side- channel attacks on elliptic curve cryptography in smart-cards

    Jean-Luc Danger, Sylvain Guilley, Philippe Hoogvorst, Cé- dric Murdica, and David Naccache. A synthesis of side- channel attacks on elliptic curve cryptography in smart-cards. J. Cryptographic Engineering , 3(4):241–265, 2013. URL https://doi.org/10.1007/s13389-013-0062-6

  20. [28]

    CacheAudit: A tool for the static analysis of cache side channels

    Goran Doychev, Boris Köpf, Laurent Mauborgne, and Jan Reineke. CacheAudit: A tool for the static analysis of cache side channels. ACM Trans. Inf. Syst. Secur., 18(1):4:1–4:32,

  21. [29]

    Dis- mantling real-world ECC with horizontal and vertical tem- plate attacks

    Margaux Dugardin, Louiza Papachristodoulou, Zakaria Najm, Lejla Batina, Jean-Luc Danger, and Sylvain Guilley. Dis- mantling real-world ECC with horizontal and vertical tem- plate attacks. In COSADE, volume 9689 of LNCS, pages 88–108. Springer, 2016. URL https://doi.org/10.1007...

  22. [31]

    Tempest: A signal problem

    Jeffrey Friedman. Tempest: A signal problem. NSA Cryptologic Spectrum , 2(3):26–30, 1972. URL https:// www.nsa.gov/Portals/70/documents/news-features/ declassified-documents/cryptologic-spectrum/ tempest.pdf

  23. [32]

    Nguyen, and Oded Regev

    Nicolas Gama, Phong Q. Nguyen, and Oded Regev. Lattice enumeration using extreme pruning. In EUROCRYPT, volume 6110 of LNCS, pages 257–278. Springer, 2010. URL https: //doi.org/10.1007/978-3-642-13190-5_13

  24. [33]

    A survey of microarchitectural timing attacks and countermea- sures on contemporary hardware

    Qian Ge, Yuval Yarom, David Cock, and Gernot Heiser. A survey of microarchitectural timing attacks and countermea- sures on contemporary hardware. J. Cryptographic Engi- neering, 8(1):1–27, 2018. URL https://doi.org/10.1007/ s13389-016-0141-6

  25. [34]

    RSA key extrac- tion via low-bandwidth acoustic cryptanalysis

    Daniel Genkin, Adi Shamir, and Eran Tromer. RSA key extrac- tion via low-bandwidth acoustic cryptanalysis. In CRYPTO, volume 8616 of LNCS, pages 444–461. Springer, 2014. URL https://doi.org/10.1007/978-3-662-44371-2_25

  26. [35]

    Stealing keys from PCs using a radio: Cheap elec- tromagnetic attacks on windowed exponentiation

    Daniel Genkin, Lev Pachmanov, Itamar Pipman, and Eran Tromer. Stealing keys from PCs using a radio: Cheap elec- tromagnetic attacks on windowed exponentiation. In CHES, volume 9293 of LNCS, pages 207–228. Springer, 2015. URL https://doi.org/10.1007/978-3-662-48324-4_11

  27. [36]

    Get your hands off my laptop: physical side-channel key-extraction at- tacks on PCs - extended version

    Daniel Genkin, Itamar Pipman, and Eran Tromer. Get your hands off my laptop: physical side-channel key-extraction at- tacks on PCs - extended version. J. Cryptographic Engineer- ing, 5(2):95–112, 2015. URL https://doi.org/10.1007/ s13389-015-0100-7

  28. [37]

    ECDH key-extraction via low-bandwidth electro- magnetic attacks on PCs

    Daniel Genkin, Lev Pachmanov, Itamar Pipman, and Eran Tromer. ECDH key-extraction via low-bandwidth electro- magnetic attacks on PCs. In CT-RSA, volume 9610 of LNCS, pages 219–235. Springer, 2016. URL https://doi.org/10. 1007/978-3-319-29485-8_13

  29. [38]

    ECDSA key extraction from mobile de- vices via nonintrusive physical side channels

    Daniel Genkin, Lev Pachmanov, Itamar Pipman, Eran Tromer, and Yuval Yarom. ECDSA key extraction from mobile de- vices via nonintrusive physical side channels. In ACM CCS, pages 1626–1638. ACM, 2016. URL http://doi.acm.org/ 10.1145/2976749.2978353

  30. [39]

    Public- key cryptosystems from lattice reduction problems

    Oded Goldreich, Shafi Goldwasser, and Shai Halevi. Public- key cryptosystems from lattice reduction problems. In CRYPTO, volume 1294 of LNCS, pages 112–131. Springer,

  31. [40]

    Triggerflow: Regression testing by ad- vanced execution path inspection

    Iaroslav Gridin, Cesar Pereida García, Nicola Tuveri, and Billy Bob Brumley. Triggerflow: Regression testing by ad- vanced execution path inspection. In DIMVA, volume 11543 of LNCS, pages 330–350. Springer, 2019. URL https://doi. org/10.1007/978-3-030-22038-9_16

  32. [41]

    Cache template attacks: Automating attacks on in- clusive last-level caches

    Daniel Gruss, Raphael Spreitzer, and Stefan Mangard. Cache template attacks: Automating attacks on in- clusive last-level caches. In USENIX Sec. , pages 897–912. USENIX Association, 2015. URL https: 16 //www.usenix.org/conference/usenixsecurity15/ technical-sessions/presentation/gruss

  33. [42]

    Flush+flush: A fast and stealthy cache attack

    Daniel Gruss, Clémentine Maurice, Klaus Wagner, and Ste- fan Mangard. Flush+flush: A fast and stealthy cache attack. In DIMVA, volume 9721 of LNCS, pages 279–

  34. [43]

    Fast prime field elliptic- curve cryptography with 256-bit primes

    Shay Gueron and Vlad Krasnov. Fast prime field elliptic- curve cryptography with 256-bit primes. J. Cryptographic Engineering, 5(2):141–151, 2015. URL https://doi.org/ 10.1007/s13389-014-0090-x

  35. [44]

    Jason Hinek

    M. Jason Hinek. Cryptanalysis of RSA and its variants . Chapman & Hall/CRC Cryptography and Network Security. CRC Press, 2010. ISBN 978-1-4200-7518-2. URL https: //doi.org/10.1201/9781420075199

  36. [45]

    Textual encodings of PKIX, PKCS, and CMS structures

    Simon Josefsson and Sean Leonard. Textual encodings of PKIX, PKCS, and CMS structures. RFC 7468, RFC Editor, April 2015. URL https://datatracker.ietf.org/doc/ rfc7468/

  37. [46]

    Fast elliptic curve cryptography in OpenSSL

    Emilia Käsper. Fast elliptic curve cryptography in OpenSSL. In Financial Cryptography Workshops, volume 7126 of LNCS, pages 27–39. Springer, 2011. URL https://doi.org/10. 1007/978-3-642-29889-9_4

  38. [47]

    Attacking RSA-based sessions in SSL/TLS

    Vlastimil Klíma, Ondrej Pokorný, and Tomás Rosa. Attacking RSA-based sessions in SSL/TLS. In CHES, volume 2779 of LNCS, pages 426–440. Springer, 2003. URL https://doi. org/10.1007/978-3-540-45238-6_33

  39. [48]

    Paul C. Kocher. Timing attacks on implementations of Diffie- Hellman, RSA, DSS, and other systems. In CRYPTO, volume 1109 of LNCS, pages 104–113. Springer, 1996. URL https: //doi.org/10.1007/3-540-68697-5_9

  40. [49]

    Kocher, Joshua Jaffe, and Benjamin Jun

    Paul C. Kocher, Joshua Jaffe, and Benjamin Jun. Differential power analysis. In CRYPTO, volume 1666 of LNCS, pages 388–397. Springer, 1999. URL https://doi.org/10.1007/ 3-540-48405-1_25

  41. [50]

    SoC it to EM: ElectroMagnetic side-channel attacks on a com- plex System-on-Chip

    Jake Longo, Elke De Mulder, Dan Page, and Michael Tunstall. SoC it to EM: ElectroMagnetic side-channel attacks on a com- plex System-on-Chip. In CHES, volume 9293 of LNCS, pages 620–640. Springer, 2015. URL https://doi.org/10.1007/ 978-3-662-48324-4_31

  42. [51]

    ANSI X9.62- 2005: Public Key Cryptography For The Financial Services Industry: The Elliptic Curve Digital Signature Algorithm (ECDSA)

    Accredited Standards Committee X9 , editor. ANSI X9.62- 2005: Public Key Cryptography For The Financial Services Industry: The Elliptic Curve Digital Signature Algorithm (ECDSA). ANSI American National Standards Institute, 2005

  43. [52]

    Messerges, Ezzy A

    Thomas S. Messerges, Ezzy A. Dabbish, and Robert H. Sloan. Examining smart-card security under the threat of power anal- ysis attacks. IEEE Trans. Computers, 51(5):541–552, 2002. URL https://doi.org/10.1109/TC.2002.1004593

  44. [53]

    Revis- iting SSL/TLS implementations: New Bleichenbacher side channels and attacks

    Christopher Meyer, Juraj Somorovsky, Eugen Weiss, Jörg Schwenk, Sebastian Schinzel, and Erik Tews. Revis- iting SSL/TLS implementations: New Bleichenbacher side channels and attacks. In USENIX Sec. , pages 733–748. USENIX Association, 2014. URL https: //www.usenix.org/conferen...

  45. [54]

    TPM-FAIL: TPM meets timing and lattice attacks

    Daniel Moghimi, Berk Sunar, Thomas Eisenbarth, and Na- dia Heninger. TPM-FAIL: TPM meets timing and lattice attacks. In USENIX Sec. USENIX Associa- tion, 2020. URL https://www.usenix.org/conference/ usenixsecurity20/presentation/moghimi

  46. [55]

    PKCS #1: RSA cryptography specifications version 2.2

    Kathleen Moriarty, Burt Kaliski, Jakob Jonsson, and Andreas Rusch. PKCS #1: RSA cryptography specifications version 2.2. RFC 8017, RFC Editor, November 2016. URL https: //datatracker.ietf.org/doc/rfc8017/

  47. [56]

    Degenerate curve attacks: extending invalid curve attacks to Edwards curves and other models

    Samuel Neves and Mehdi Tibouchi. Degenerate curve attacks: extending invalid curve attacks to Edwards curves and other models. IET Information Security, 12(3):217–225, 2018. URL https://doi.org/10.1049/iet-ifs.2017.0075

  48. [57]

    Nguyen and Igor E

    Phong Q. Nguyen and Igor E. Shparlinski. The insecurity of the Digital Signature Algorithm with partially known nonces.J. Cryptology, 15(3):151–176, 2002. URL https://doi.org/ 10.1007/s00145-002-0021-3

  49. [58]

    Nguyen and Igor E

    Phong Q. Nguyen and Igor E. Shparlinski. The insecurity of the Elliptic Curve Digital Signature Algorithm with partially known nonces. Des. Codes Cryptogr., 30(2):201–217, 2003. URL https://doi.org/10.1023/A:1025436905711

  50. [59]

    Cache attacks and countermeasures: The case of AES

    Dag Arne Osvik, Adi Shamir, and Eran Tromer. Cache attacks and countermeasures: The case of AES. In CT-RSA, volume 3860 of LNCS, pages 1–20. Springer, 2006. URL https: //doi.org/10.1007/11605805_1

  51. [60]

    Cache missing for fun and profit

    Colin Percival. Cache missing for fun and profit. In BSD- Can, 2005. URL http://www.daemonology.net/papers/ cachemissing.pdf

  52. [61]

    Constant- time callees with variable-time callers

    Cesar Pereida García and Billy Bob Brumley. Constant- time callees with variable-time callers. In USENIX Sec., pages 83–98. USENIX Association, 2017. ISBN 978-1-931971-40-9. URL https://www.usenix.org/ conference/usenixsecurity17/technical-sessions/ presentation/garcia

  53. [62]

    Make sure DSA signing exponentiations really are constant- time

    Cesar Pereida García, Billy Bob Brumley, and Yuval Yarom. “Make sure DSA signing exponentiations really are constant- time”. In ACM CCS, pages 1639–1650. ACM, 2016. URL http://doi.acm.org/10.1145/2976749.2978420

  54. [63]

    CVE-2019-1547: research data and tooling

    Cesar Pereida García, Sohaib ul Hassan, Nicola Tuveri, Iaroslav Gridin, Alejandro Cabrera Aldaya, and Billy Bob Brumley. CVE-2019-1547: research data and tooling. Zenodo, April

  55. [64]

    Addi- tional cryptographic algorithms for use with GOST 28147-89, GOST R 34.10-94, GOST R 34.10-2001, and GOST R 34.11- 94 algorithms

    Vladimir Popov, Serguei Leontiev, and Igor Kurepkin. Addi- tional cryptographic algorithms for use with GOST 28147-89, GOST R 34.10-94, GOST R 34.10-2001, and GOST R 34.11- 94 algorithms. RFC 4357, RFC Editor, January 2006. URL https://datatracker.ietf.org/doc/rfc4357/

  56. [65]

    Power analysis attacks and countermeasures

    Thomas Popp, Stefan Mangard, and Elisabeth Oswald. Power analysis attacks and countermeasures. IEEE Design & Test of Computers, 24(6):535–543, 2007. URL https://doi.org/ 10.1109/MDT.2007.200. 17

  57. [66]

    Electromag- netic analysis (EMA): measures and counter-measures for smart cards

    Jean-Jacques Quisquater and David Samyde. Electromag- netic analysis (EMA): measures and counter-measures for smart cards. In E-smart, volume 2140 of LNCS, pages 200–

  58. [67]

    Michael O. Rabin. Probabilistic algorithm for testing primal- ity. J. Number Theory , 12(1):128–138, 1980. ISSN 0022- 314X. URL https://doi.org/10.1016/0022-314x(80) 90084-0

  59. [68]

    Breaking Ed25519 in WolfSSL

    Niels Samwel, Lejla Batina, Guido Bertoni, Joan Daemen, and Ruggero Susella. Breaking Ed25519 in WolfSSL. In CT-RSA, volume 10808 of LNCS, pages 1–20. Springer, 2018. URL https://doi.org/10.1007/978-3-319-76953-0_1

  60. [69]

    ITU-T X.690 Information technology – ASN.1 encoding rules: Specification of Basic Encoding Rules (BER), Canoni- cal Encoding Rules (CER) and Distinguished Encoding Rules (DER)

    ITU-T Telecommunication standardization sector of ITU, ed- itor. ITU-T X.690 Information technology – ASN.1 encoding rules: Specification of Basic Encoding Rules (BER), Canoni- cal Encoding Rules (CER) and Distinguished Encoding Rules (DER). ITU International Telecommunication ...

  61. [70]

    Degenerate fault attacks on elliptic curve parameters in OpenSSL

    Akira Takahashi and Mehdi Tibouchi. Degenerate fault attacks on elliptic curve parameters in OpenSSL. In EuroS&P, pages 371–386. IEEE, 2019. URL https://doi.org/10.1109/ EuroSP.2019.00035

  62. [71]

    Smart card security

    Michael Tunstall. Smart card security. In Smart Cards, To- kens, Security and Applications , pages 217–251. Springer, second edition, 2017. URL https://doi.org/10.1007/ 978-3-319-50500-8_9

  63. [72]

    Side-channel analysis of SM2: A late- stage featurization case study

    Nicola Tuveri, Sohaib ul Hassan, Cesar Pereida García, and Billy Bob Brumley. Side-channel analysis of SM2: A late- stage featurization case study. In ACSAC, pages 147–160. ACM, 2018. URL https://doi.org/10.1145/3274694. 3274725

  64. [73]

    Smart, and Yuval Yarom

    Joop van de Pol, Nigel P. Smart, and Yuval Yarom. Just a little bit more. In CT-RSA, volume 9048 of LNCS, pages 3–21. Springer, 2015. URL https://doi.org/10.1007/ 978-3-319-16715-2_1

  65. [74]

    RIDL: rogue in-flight data load

    Stephan van Schaik, Alyssa Milburn, Sebastian Österlund, Pietro Frigo, Giorgi Maisuradze, Kaveh Razavi, Herbert Bos, and Cristiano Giuffrida. RIDL: rogue in-flight data load. In IEEE S&P , pages 88–105. IEEE, 2019. URL https: //doi.org/10.1109/SP.2019.00087

  66. [75]

    CacheD: Identifying cache-based timing channels in production software

    Shuai Wang, Pei Wang, Xiao Liu, Danfeng Zhang, and Dinghao Wu. CacheD: Identifying cache-based timing channels in production software. In USENIX Sec. , pages 235–252. USENIX Association, 2017. URL https: //www.usenix.org/conference/usenixsecurity17/ technical-sessions/presenta...

  67. [76]

    Single trace attack against RSA key generation in Intel SGX SSL

    Samuel Weiser, Raphael Spreitzer, and Lukas Bodner. Single trace attack against RSA key generation in Intel SGX SSL. In AsiaCCS, pages 575–586. ACM, 2018. URL http://doi. acm.org/10.1145/3196494.3196524

  68. [77]

    DATA - differential ad- dress trace analysis: Finding address-based side-channels in binaries

    Samuel Weiser, Andreas Zankl, Raphael Spreitzer, Katja Miller, Stefan Mangard, and Georg Sigl. DATA - differential ad- dress trace analysis: Finding address-based side-channels in binaries. In USENIX Sec., pages 603–620. USENIX Associ- ation, 2018. URL https://www.usenix.org/c...

  69. [78]

    FLUSH+RELOAD: A high resolution, low noise, L3 cache side-channel attack

    Yuval Yarom and Katrina Falkner. FLUSH+RELOAD: A high resolution, low noise, L3 cache side-channel attack. In USENIX Sec. , pages 719–732. USENIX Asso- ciation, 2014. ISBN 978-1-931971-15-7. URL https: //www.usenix.org/conference/usenixsecurity14/ technical-sessions/presentation/yarom

  70. [79]

    Internet X.509 public key infrastructure time-stamp protocol (TSP)

    Robert Zuccherato, Patrick Cain, Carlisle Adams, and Denis Pinkas. Internet X.509 public key infrastructure time-stamp protocol (TSP). RFC 3161, RFC Editor, August 2001. URL https://datatracker.ietf.org/doc/rfc3161/. A mbedTLS vulnerable RSA keys Table 2: RSA keys that follow ...

  71. [140]

    URL https://doi.org/10.1007/ 978-3-319-21476-4_9

    Springer, 2015. URL https://doi.org/10.1007/ 978-3-319-21476-4_9

  72. [210]

    URL https://doi.org/10.1007/ 3-540-45418-7_17

    Springer, 2001. URL https://doi.org/10.1007/ 3-540-45418-7_17

  73. [299]

    URL https://doi.org/10.1007/ 978-3-319-40667-1_14

    Springer, 2016. URL https://doi.org/10.1007/ 978-3-319-40667-1_14

  74. [435]

    URL http://doi.acm.org/10.1145/ 2991079.2991084

    ACM, 2016. URL http://doi.acm.org/10.1145/ 2991079.2991084

  75. [1997]

    URL https://doi.org/10.1007/BFb0052231

  76. [2002]

    URL https://doi.org/10.1007/3-540-36400-5_ 3

  77. [2015]

    URL https://doi.org/10.1145/2756550

  78. [2020]

    URL https://doi.org/10.5281/zenodo.3736311

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.