Add EVP_CIPHER API for XAES-256-GCM with Key Commitment #2826
+314
−72
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -353,6 +353,15 @@ typedef struct { | |
| uint8_t k1[AES_BLOCK_SIZE]; | ||
| } XAES_256_GCM_CTX; | ||
|
|
||
| #define XAES_256_GCM_KEY_COMMIT_SIZE (AES_BLOCK_SIZE * 2) | ||
|
|
||
| typedef struct { | ||
| EVP_AES_GCM_CTX aes_gcm_ctx; | ||
| AES_KEY xaes_key; | ||
| uint8_t k1[AES_BLOCK_SIZE]; | ||
| uint8_t kc[XAES_256_GCM_KEY_COMMIT_SIZE]; | ||
| } XAES_256_GCM_KC_CTX; | ||
|
|
||
| static EVP_AES_GCM_CTX *aes_gcm_from_cipher_ctx(EVP_CIPHER_CTX *ctx) { | ||
| OPENSSL_STATIC_ASSERT( | ||
| alignof(EVP_AES_GCM_CTX) <= 16, | ||
|
|
@@ -388,6 +397,17 @@ static EVP_AES_GCM_CTX *aes_gcm_from_cipher_ctx(EVP_CIPHER_CTX *ctx) { | |
| assert((uintptr_t)ptr % 8 == 0); | ||
| ptr += (uintptr_t)ptr & 8; | ||
| return &((XAES_256_GCM_CTX *)ptr)->aes_gcm_ctx; | ||
| // XAES-256-GCM-KC | ||
| case NID_xaes_256_gcm_kc: | ||
| assert(ctx->cipher->ctx_size == sizeof(XAES_256_GCM_KC_CTX)); | ||
| ptr = ctx->cipher_data; | ||
| #if defined(OPENSSL_32_BIT) | ||
| assert((uintptr_t)ptr % 4 == 0); | ||
| ptr += (uintptr_t)ptr & 4; | ||
| #endif | ||
| assert((uintptr_t)ptr % 8 == 0); | ||
| ptr += (uintptr_t)ptr & 8; | ||
| return &((XAES_256_GCM_KC_CTX *)ptr)->aes_gcm_ctx; | ||
| default: | ||
| break; | ||
| } | ||
|
|
@@ -1781,7 +1801,6 @@ Extension to support nonce size less than 24 bytes: | |
| https://eprint.iacr.org/2025/758.pdf#page=24 | ||
| -----------------------------------------------------------------------*/ | ||
| #define XAES_256_GCM_KEY_LENGTH (AES_BLOCK_SIZE * 2) | ||
| #define XAES_256_GCM_MAX_NONCE_SIZE (AES_GCM_NONCE_LENGTH * 2) | ||
| #define XAES_256_GCM_MIN_NONCE_SIZE (20) | ||
| /* | ||
|
|
@@ -2033,3 +2052,125 @@ DEFINE_METHOD_FUNCTION(EVP_AEAD, EVP_aead_xaes_256_gcm) { | |
| out->seal_scatter = aead_xaes_256_gcm_seal_scatter; | ||
| out->open_gather = aead_xaes_256_gcm_open_gather; | ||
| } | ||
|
|
||
| // ------------------------------------------------------------------------------ | ||
| // ---------------- EVP_CIPHER XAES-256-GCM With Key Commitment ----------------- | ||
| // ----------- Reference: https://eprint.iacr.org/2025/758.pdf#page=6 ----------- | ||
| // ------------------------------------------------------------------------------ | ||
| static XAES_256_GCM_KC_CTX *xaes_256_gcm_kc_from_cipher_ctx(EVP_CIPHER_CTX *ctx) { | ||
| // alignment to be consistent with aes_gcm_from_cipher_ctx() | ||
| char *ptr = ctx->cipher_data; | ||
| #if defined(OPENSSL_32_BIT) | ||
| assert((uintptr_t)ptr % 4 == 0); | ||
| ptr += (uintptr_t)ptr & 4; | ||
| #endif | ||
| assert((uintptr_t)ptr % 8 == 0); | ||
| ptr += (uintptr_t)ptr & 8; | ||
|
Comment on lines
+2063
to
+2068
Contributor
There was a problem hiding this comment. I doubt we need these alignment requirements in addition to those of the gcm ctx. Can we try removing them here and on line 1858? Otherwise, do you know what would drive our need for them?
Author
There was a problem hiding this comment. It will lead to many CI tests failed. I tried that. You can see the failed tests below. However, when I created and run with a docker, all tests still passed. |
||
| return (XAES_256_GCM_KC_CTX *)ptr; | ||
| } | ||
|
|
||
| static int xaes_256_gcm_CMAC_extract_key_commitment(AES_KEY *xaes_key, uint8_t *k1, | ||
| uint8_t *key_commitment, const uint8_t* nonce, const unsigned nonce_len) { | ||
| uint8_t W[AES_BLOCK_SIZE]; | ||
|
|
||
| uint8_t kc_prefix[4] = {0x58, 0x43, 0x4D, 0x54}; | ||
| uint8_t X1[AES_BLOCK_SIZE]; | ||
| OPENSSL_memcpy(W, kc_prefix, 4); | ||
| OPENSSL_memcpy(W + 4, nonce, AES_GCM_NONCE_LENGTH); | ||
|
|
||
| AES_encrypt(W, X1, xaes_key); | ||
| // For nonce size < 24 | ||
| // Reference: https://eprint.iacr.org/2025/758.pdf#page=24 | ||
| OPENSSL_memcpy(W, nonce + nonce_len - AES_GCM_NONCE_LENGTH, AES_GCM_NONCE_LENGTH); | ||
| W[AES_BLOCK_SIZE-4] = XAES_256_GCM_MAX_NONCE_SIZE - nonce_len; | ||
| W[AES_BLOCK_SIZE-3] = 0x01; | ||
| W[AES_BLOCK_SIZE-2] = 0x00; | ||
| W[AES_BLOCK_SIZE-1] = 0x01; | ||
|
|
||
| for (size_t i = 0; i < AES_BLOCK_SIZE; i++) { | ||
| W[i] ^= X1[i] ^ k1[i]; | ||
| } | ||
|
|
||
| AES_encrypt(W, key_commitment, xaes_key); | ||
| W[AES_BLOCK_SIZE-1] ^= 0x03; | ||
| AES_encrypt(W, key_commitment + AES_BLOCK_SIZE, xaes_key); | ||
|
|
||
| return 1; | ||
| } | ||
|
|
||
| static int xaes_256_gcm_init_key_commit(EVP_CIPHER_CTX *ctx, const uint8_t *key, | ||
| const uint8_t *iv, int enc) { | ||
| // Key length: 32 bytes | ||
| if (ctx->key_len != XAES_256_GCM_KEY_LENGTH) { | ||
| OPENSSL_PUT_ERROR(CIPHER, CIPHER_R_BAD_KEY_LENGTH); | ||
| return 0; | ||
| } | ||
|
|
||
| XAES_256_GCM_KC_CTX *xaes_ctx = xaes_256_gcm_kc_from_cipher_ctx(ctx); | ||
|
|
||
| // When main key is provided, initialize the xaes-256-gcm context | ||
| if(key != NULL) { | ||
| xaes_256_gcm_ctx_init(&xaes_ctx->xaes_key, xaes_ctx->k1, key); | ||
| } | ||
|
|
||
| // If iv is provided, even if main key is not, derive a subkey and its key commitment | ||
| /* Note: Unlike the MAC tag, key commitment only depends on the main key | ||
| * and iv, not on plaintext and aad, therefore we extract it in init when | ||
| * deriving subkey instead of at the end of encrytion like the MAC tag */ | ||
| if(iv != NULL) { | ||
| // Derive subkey | ||
| xaes_256_gcm_set_gcm_key(ctx, iv, enc); | ||
| // Extract key commitment | ||
| EVP_AES_GCM_CTX *gctx = &xaes_ctx->aes_gcm_ctx; | ||
| xaes_256_gcm_CMAC_extract_key_commitment(&xaes_ctx->xaes_key, xaes_ctx->k1, xaes_ctx->kc, iv, gctx->ivlen); | ||
| } | ||
|
|
||
| return 1; | ||
| } | ||
|
|
||
| static int xaes_256_gcm_key_commit_ctrl(EVP_CIPHER_CTX *ctx, int type, int arg, void *ptr) { | ||
|
|
||
| XAES_256_GCM_KC_CTX *xaes_ctx = xaes_256_gcm_kc_from_cipher_ctx(ctx); | ||
|
|
||
| switch(type) { | ||
|
Contributor
There was a problem hiding this comment. I think it's more prudent to check that ptr is not NULL. aws-lc/crypto/fipsmodule/cipher/cipher.c Line 185 in 88aec89
Author
There was a problem hiding this comment. Yes. I'll add checking ptr for two new added ctrl commands. |
||
| case EVP_CTRL_AEAD_GET_KC: | ||
| if(arg < XAES_256_GCM_KEY_COMMIT_SIZE) { | ||
| OPENSSL_PUT_ERROR(CIPHER, CIPHER_R_BUFFER_TOO_SMALL); | ||
| return 0; | ||
| } | ||
| OPENSSL_memcpy(ptr, xaes_ctx->kc, arg); | ||
| return 1; | ||
| /* Since the verification of key commitment (KC) can happen at any time, | ||
| * not only at the end of decryption like for MAC tag, we create a control | ||
| * command for KC verification, rather than a command to set KC, then | ||
| * verifying it at the end of decryption as for checking the MAC tag */ | ||
| case EVP_CTRL_AEAD_VERIFY_KC: | ||
| if(arg < XAES_256_GCM_KEY_COMMIT_SIZE) { | ||
| OPENSSL_PUT_ERROR(CIPHER, CIPHER_R_BUFFER_TOO_SMALL); | ||
| return 0; | ||
| } | ||
| if(OPENSSL_memcmp(xaes_ctx->kc, ptr, XAES_256_GCM_KEY_COMMIT_SIZE)) { | ||
| OPENSSL_PUT_ERROR(CIPHER, CIPHER_R_KEY_COMMITMENT_INVALID); | ||
| return 0; | ||
| } | ||
| return 1; | ||
| default: | ||
| return aes_gcm_ctrl(ctx, type, arg, ptr); | ||
| } | ||
| } | ||
|
|
||
| DEFINE_METHOD_FUNCTION(EVP_CIPHER, EVP_xaes_256_gcm_kc) { | ||
| OPENSSL_memset(out, 0, sizeof(EVP_CIPHER)); | ||
| out->nid = NID_xaes_256_gcm_kc; | ||
| out->block_size = 1; | ||
| out->key_len = XAES_256_GCM_KEY_LENGTH; | ||
| out->iv_len = XAES_256_GCM_MAX_NONCE_SIZE; | ||
| out->ctx_size = sizeof(XAES_256_GCM_KC_CTX); | ||
| out->flags = EVP_CIPH_GCM_MODE | EVP_CIPH_CUSTOM_IV | EVP_CIPH_CUSTOM_COPY | | ||
| EVP_CIPH_FLAG_CUSTOM_CIPHER | EVP_CIPH_ALWAYS_CALL_INIT | | ||
| EVP_CIPH_CTRL_INIT | EVP_CIPH_FLAG_AEAD_CIPHER | EVP_CIPH_FLAG_KC_CIPHER; | ||
| out->init = xaes_256_gcm_init_key_commit; | ||
| out->cipher = aes_gcm_cipher; | ||
| out->cleanup = aes_gcm_cleanup; | ||
| out->ctrl = xaes_256_gcm_key_commit_ctrl; | ||
| } | ||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can the negative tests that were added for XAES (e.g. shorter nonces) be modified to also test XAES-KC?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I made changes in
cipher_test.ccto test negative cases with invalid key/nonce, and shorter nonce for bothevp-xaes-256-gcmandevp-xaes-256-gcm-kc.