Add wc_SignCert_cb API for external signing callbacks #9632

jackctj117 · 2026-01-08T22:22:09Z

This pull request adds support for signing certificates and CSRs using a user-provided callback function, enabling integration with external signing devices (such as TPMs or HSMs) without relying on the crypto callback infrastructure. This is particularly useful for FIPS-compliant applications and scenarios where offloading cryptographic operations is required. The changes include new API definitions, documentation, internal implementation, and tests for the callback-based signing mechanism.

New Callback-Based Certificate Signing API

Introduced the wc_SignCert_cb function and the wc_SignCertCb callback type, allowing certificates and CSRs to be signed via an external callback for flexible integration with devices like TPMs/HSMs. [1] [2] [3]

Internal Implementation

Added the internal MakeSignatureCb function to handle hashing, digest encoding, and invoking the user-provided signing callback, supporting both RSA and ECC key types.

Testing
Setup:
TPM simulator: swtpm running on port 2321
Built wolfSSL with: --enable-certgen --enable-certreq --enable-certext --enable-cryptocb
Built wolfTPM with: --enable-swtpm --enable-certgen --enable-debug
Tests Run:
Generated RSA and ECC test keys in TPM
Created CSRs using ./examples/csr/csr
Validated CSRs with openssl req -text -noout
Results:
wc_SignCert_cb compiled into wolfSSL
wolfTPM2_SignCertCb and CSR_MakeAndSign_Cb compiled into wolfTPM
Generated valid RSA (1228 bytes) and ECC (696 bytes) CSRs
CSRs verified successfully with OpenSSL

dgarske · 2026-01-14T17:54:51Z

Jenkins retest this please. History lost.

dgarske · 2026-01-14T17:56:51Z

wolfcrypt/src/asn.c

@@ -33861,6 +33805,103 @@ int wc_MakeCertReq(Cert* cert, byte* derBuffer, word32 derSz,
 #endif /* WOLFSSL_CERT_REQ */


+


I'd like this new feature to be wrapped in new build option to avoid code growth. Also some of this looks to be duplicate code now. Is there way way to implement such that the old code uses the callback path internally when not set. If so you can keep it always like it is.

…cation

Copilot

Pull request overview

This pull request adds support for signing certificates and CSRs using a user-provided callback function, enabling integration with external signing devices (TPMs/HSMs) without relying on the crypto callback infrastructure. This is particularly useful for FIPS-compliant applications where offloading cryptographic operations is not acceptable.

Changes:

Introduced new wc_SignCert_cb API and wc_SignCertCb callback type for external certificate/CSR signing
Refactored internal MakeSignature function to use new MakeSignatureCb internally for RSA and ECC, eliminating code duplication
Added configure option --enable-certsigncb to enable the feature

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 8 comments.

File	Description
wolfssl/wolfcrypt/asn_public.h	Added public API declarations for the callback-based certificate signing, including typedef for wc_SignCertCb and function declaration for wc_SignCert_cb
wolfcrypt/src/asn.c	Implemented internal MakeSignatureCb function and refactored MakeSignature to use callback path for RSA/ECC; added wc_SignCert_cb implementation
tests/api.c	Added test case test_wc_SignCert_cb with mock callback to verify the new API functionality
configure.ac	Added configuration option --enable-certsigncb to control compilation of the new feature

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-01-15T00:30:00Z

wolfcrypt/src/asn.c

+#if !defined(NO_RSA) && !defined(WOLFSSL_RSA_PUBLIC_ONLY) && !defined(WOLFSSL_RSA_VERIFY_ONLY)
+    if (keyType == RSA_TYPE && signCtx->rsaKey) {
+        /* For RSA, input is already encoded digest */
+        ret = wc_RsaSSL_Sign(in, inLen, out, *outLen,
+                             signCtx->rsaKey, signCtx->rng);
+        if (ret > 0) {
+            *outLen = (word32)ret;
+            ret = 0;
+        }
+    }
+#endif /* !NO_RSA && !WOLFSSL_RSA_PUBLIC_ONLY && !WOLFSSL_RSA_VERIFY_ONLY */
+
+#if defined(HAVE_ECC) && defined(HAVE_ECC_SIGN)
+    if (keyType == ECC_TYPE && signCtx->eccKey) {
+        /* For ECC, input is the raw hash */
+        ret = wc_ecc_sign_hash(in, inLen, out, outLen,
+                               signCtx->rng, signCtx->eccKey);
+    }
+#endif /* HAVE_ECC && HAVE_ECC_SIGN */
+
+#if defined(HAVE_ED25519) && defined(HAVE_ED25519_SIGN)
+    if (keyType == ED25519_TYPE && signCtx->ed25519Key) {
+        /* Ed25519 needs the original message, not hash */
+        /* Note: For Ed25519, 'in' should be the original message buffer */
+        /* This is a limitation of the refactoring - Ed25519 signs messages, not hashes */
+        ret = NOT_COMPILED_IN; /* Cannot support Ed25519 through callback path */
+    }
+#endif /* HAVE_ED25519 && HAVE_ED25519_SIGN */
+
+#if defined(HAVE_ED448) && defined(HAVE_ED448_SIGN)
+    if (keyType == ED448_TYPE && signCtx->ed448Key) {
+        /* Ed448 needs the original message, not hash */
+        ret = NOT_COMPILED_IN; /* Cannot support Ed448 through callback path */
+    }
+#endif /* HAVE_ED448 && HAVE_ED448_SIGN */
+
+#if defined(HAVE_FALCON)
+    if (keyType == FALCON_LEVEL1_TYPE || keyType == FALCON_LEVEL5_TYPE) {
+        if (signCtx->falconKey) {
+            /* Falcon needs the original message */
+            ret = NOT_COMPILED_IN; /* Cannot support Falcon through callback path */
+        }
+    }
+#endif /* HAVE_FALCON */
+
+#if defined(HAVE_DILITHIUM) && !defined(WOLFSSL_DILITHIUM_NO_SIGN)
+    if (keyType == DILITHIUM_LEVEL2_TYPE || keyType == DILITHIUM_LEVEL3_TYPE ||
+        keyType == DILITHIUM_LEVEL5_TYPE) {
+        if (signCtx->dilithiumKey) {
+            /* Dilithium needs the original message */
+            ret = NOT_COMPILED_IN; /* Cannot support Dilithium through callback path */
+        }
+    }
+#endif /* HAVE_DILITHIUM && !WOLFSSL_DILITHIUM_NO_SIGN */
+
+#if defined(HAVE_SPHINCS)
+    if (keyType == SPHINCS_FAST_LEVEL1_TYPE || keyType == SPHINCS_FAST_LEVEL3_TYPE ||
+        keyType == SPHINCS_FAST_LEVEL5_TYPE || keyType == SPHINCS_SMALL_LEVEL1_TYPE ||
+        keyType == SPHINCS_SMALL_LEVEL3_TYPE || keyType == SPHINCS_SMALL_LEVEL5_TYPE) {
+        if (signCtx->sphincsKey) {
+            /* Sphincs needs the original message */
+            ret = NOT_COMPILED_IN; /* Cannot support Sphincs through callback path */
+        }
+    }
+#endif /* HAVE_SPHINCS */


The InternalSignCb function uses multiple independent if statements instead of else-if chains. While this works correctly in practice because MakeSignature ensures only one key type is set, it's inconsistent with the coding pattern elsewhere and could cause issues if the logic changes. The function should use else-if to make the mutual exclusivity explicit and prevent potential future bugs.

Copilot · 2026-01-15T00:30:01Z

tests/api.c

+        word32 outSz = *outLen;
+
+        /* For RSA, input is pre-encoded digest, just sign it */
+        ret = wc_RsaSSL_Sign(in, inLen, out, outSz, rsaKey, NULL);


The test mock callback passes NULL as the RNG parameter to wc_RsaSSL_Sign, but it should pass signCtx->rng to be consistent with the internal implementation and to support RSA features that require RNG (such as blinding). The RNG is available in the context and should be used.

Copilot · 2026-01-15T00:30:01Z

tests/api.c

+static int test_wc_SignCert_cb(void)
+{
+    EXPECT_DECLS;
+#if defined(WOLFSSL_CERT_GEN) && defined(HAVE_ECC) && !defined(NO_ASN_TIME)
+    Cert cert;
+    byte der[FOURK_BUF];
+    int derSize = 0;
+    WC_RNG rng;
+    ecc_key key;
+    MockSignCtx signCtx;
+    int ret;
+
+    XMEMSET(&rng, 0, sizeof(WC_RNG));
+    XMEMSET(&key, 0, sizeof(ecc_key));
+    XMEMSET(&cert, 0, sizeof(Cert));
+    XMEMSET(&signCtx, 0, sizeof(MockSignCtx));
+
+    ExpectIntEQ(wc_InitRng(&rng), 0);
+    ExpectIntEQ(wc_ecc_init(&key), 0);
+    ExpectIntEQ(wc_ecc_make_key(&rng, 32, &key), 0);
+    ExpectIntEQ(wc_InitCert(&cert), 0);
+
+    (void)XSTRNCPY(cert.subject.country, "US", CTC_NAME_SIZE);
+    (void)XSTRNCPY(cert.subject.state, "state", CTC_NAME_SIZE);
+    (void)XSTRNCPY(cert.subject.locality, "locality", CTC_NAME_SIZE);
+    (void)XSTRNCPY(cert.subject.org, "org", CTC_NAME_SIZE);
+    (void)XSTRNCPY(cert.subject.unit, "unit", CTC_NAME_SIZE);
+    (void)XSTRNCPY(cert.subject.commonName, "www.example.com",
+        CTC_NAME_SIZE);
+    (void)XSTRNCPY(cert.subject.email, "test@example.com", CTC_NAME_SIZE);
+
+    cert.selfSigned = 1;
+    cert.isCA       = 0;
+    cert.sigType    = CTC_SHA256wECDSA;
+
+    /* Make cert body */
+    ExpectIntGT(wc_MakeCert(&cert, der, FOURK_BUF, NULL, &key, &rng), 0);
+
+    /* Setup signing context with key and RNG */
+    signCtx.key = &key;
+    signCtx.rng = &rng;
+
+    /* Sign using callback API */
+    ExpectIntGT(derSize = wc_SignCert_cb(cert.bodySz, cert.sigType, der,
+        FOURK_BUF, ECC_TYPE, mockSignCb, &signCtx, &rng), 0);


The test only exercises the ECC code path. The mockSignCb function includes RSA support, but test_wc_SignCert_cb only tests with ECC keys. Consider adding a separate test case or expanding this test to also verify RSA certificate signing via the callback to ensure complete test coverage.

Copilot · 2026-01-15T00:30:01Z

wolfcrypt/src/asn.c

+    word32 outLen;
+
+    (void)rng;
+    (void)heap;


The heap parameter is marked as unused with (void)heap on line 33871, but it is actually used on lines 33878 and 33898 for memory allocation. The (void)heap statement should be removed to avoid misleading readers and to prevent potential compiler warnings in some configurations.

Copilot · 2026-01-15T00:30:01Z

wolfcrypt/src/asn.c

+        else
+#endif /* !NO_RSA */
+        {
+            /* ECC/EdDSA: pass raw hash or message */


The comment says "ECC/EdDSA: pass raw hash or message" but EdDSA algorithms (Ed25519, Ed448) are explicitly not supported through the callback path as shown in InternalSignCb. The comment should be corrected to just say "ECC: pass raw hash" to avoid confusion.

tests/api.c

wolfssl/wolfcrypt/asn_public.h

Copilot

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 4 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-01-15T20:16:14Z

wolfcrypt/src/asn.c

+#ifndef WOLFSSL_CERT_SIGN_CB
+typedef int (*wc_SignCertCb)(const byte* in, word32 inLen,
+                             byte* out, word32* outLen,
+                             int sigAlgo, int keyType, void* ctx);
+#endif


The wc_SignCertCb typedef is duplicated here as a fallback when WOLFSSL_CERT_SIGN_CB is not defined. This creates potential for inconsistency if the public API typedef in asn_public.h is updated but this internal version is not. Consider removing this internal typedef and ensuring that code using this callback always includes the public header, or add a comment explaining why the duplication is necessary.

Copilot · 2026-01-15T20:16:14Z

wolfcrypt/src/asn.c

+#if (!defined(NO_RSA) && !defined(WOLFSSL_RSA_PUBLIC_ONLY) && !defined(WOLFSSL_RSA_VERIFY_ONLY)) || \
+    (defined(HAVE_ECC) && defined(HAVE_ECC_SIGN))
+    if (rsaKey || eccKey) {


The condition checks if RSA or ECC is enabled, but then the if statement assumes at least one key is non-NULL. If both RSA and ECC are disabled at compile time, the code after line 32084 (exit_ms) will still execute and may fall through to algorithms that are not supported. The logic should ensure that when RSA and ECC are both disabled, the function properly handles or rejects the request.

Copilot · 2026-01-15T20:16:15Z

tests/api.c

+static int test_wc_SignCert_cb(void)
+{
+    EXPECT_DECLS;
+#if defined(WOLFSSL_CERT_GEN) && defined(HAVE_ECC) && !defined(NO_ASN_TIME)


The test only covers ECC key types, but the API and implementation support both RSA and ECC. Consider adding a test case for RSA to ensure the RSA code path (wc_RsaSSL_Sign in the callback) is also tested and works correctly with the new API.

Copilot · 2026-01-15T20:16:15Z

wolfcrypt/src/asn.c

+        /* Ed25519 needs the original message, not hash */
+        /* Note: For Ed25519, 'in' should be the original message buffer */
+        /* This is a limitation of the refactoring - Ed25519 signs messages, not hashes */
+        ret = NOT_COMPILED_IN; /* Cannot support Ed25519 through callback path */


The NOT_COMPILED_IN error code for Ed25519, Ed448, and post-quantum algorithms may be confusing to users since these algorithms ARE compiled in when their respective macros are defined. A more descriptive error code like NOT_SUPPORTED or a new error code like SIG_ALGO_NOT_SUPPORTED_BY_CB would better convey that the callback path doesn't support these signature types.

…upported algos

Add wc_SignCert_cb API for external signing callbacks

4aa26c1

jackctj117 self-assigned this Jan 8, 2026

jackctj117 marked this pull request as draft January 8, 2026 22:22

jackctj117 mentioned this pull request Jan 8, 2026

Add TPM support for wc_SignCert_cb callback API wolfSSL/wolfTPM#450

Open

jackctj117 added 3 commits January 8, 2026 15:57

removed idx, added to test matrix

2e3fa9b

Fix test_wc_SignCert_cb: pass RNG for ECC signing

7dbbf74

FIX: (word32) cast

8616922

jackctj117 marked this pull request as ready for review January 9, 2026 21:59

wolfSSL deleted a comment from devin-ai-integration bot Jan 9, 2026

jackctj117 requested a review from wolfSSL-Bot January 9, 2026 23:30

dgarske assigned dgarske, wolfSSL-Bot and lealem47 Jan 14, 2026

dgarske requested review from dgarske and lealem47 and removed request for wolfSSL-Bot January 14, 2026 16:59

dgarske requested changes Jan 14, 2026

View reviewed changes

dgarske unassigned dgarske and lealem47 Jan 14, 2026

jackctj117 added 3 commits January 14, 2026 14:43

Add --enable-certsigncb build option and eliminate cert signing dupli…

2d36f0d

…cation

Guard wc_SignCert_cb test with WOLFSSL_CERT_SIGN_CB to fix CI failures

c6dd3f1

Fix for when RSA is enabled

d84e1d1

lealem47 requested a review from Copilot January 14, 2026 23:45

Copilot started reviewing on behalf of lealem47 January 15, 2026 00:00 View session

Copilot AI reviewed Jan 15, 2026

View reviewed changes

Fix cert sign callback issues: heap usage, documentation, and test RNG

81df4a3

jackctj117 requested a review from Copilot January 15, 2026 20:12

Copilot started reviewing on behalf of jackctj117 January 15, 2026 20:13 View session

Copilot AI reviewed Jan 15, 2026

View reviewed changes

Address Copilot feedback: clarify typedef and fix error codes for uns…

90ce237

…upported algos

		@@ -33861,6 +33805,103 @@ int wc_MakeCertReq(Cert* cert, byte* derBuffer, word32 derSz,
		#endif /* WOLFSSL_CERT_REQ */

Add wc_SignCert_cb API for external signing callbacks #9632

Are you sure you want to change the base?

Add wc_SignCert_cb API for external signing callbacks #9632

Conversation

jackctj117 commented Jan 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dgarske commented Jan 14, 2026

Uh oh!

dgarske Jan 14, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

jackctj117 commented Jan 8, 2026 •

edited

Loading