From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.jvdsn.com (smtp.jvdsn.com [129.153.194.31]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id C9D26372EF0; Mon, 20 Apr 2026 16:48:48 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=129.153.194.31 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1776703730; cv=none; b=PoPeDj5U6IPWd1IAWaUp5F4sUIgHRpVig5451+7CklfwQh2QRhzXrw/mhkXYTf/SK8sNVzXae6VxS8/Nl4r0gVq2ay6znBePjNyO03dxuDGbr4ry7YHa6wA+TohZ5rr1cud2jNiQZ1ykKlgJntrJIkqN+O34hbVW50xg4CP8N3s= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1776703730; c=relaxed/simple; bh=kcNBUdT1rAgPFcu1gMxQPN3qd7G83ytFuTKyNDWeDyM=; h=Message-ID:Date:MIME-Version:Subject:To:Cc:References:From: In-Reply-To:Content-Type; b=PRWz0cnV7PFdh7IthmDeRZiX1NZqarxQyfBbQlbh4dXZY/DuCLJZ4UuvpHwYghmM50klTGCNY0F+lbExiMKAdA9PShkPjpd9ekoWe2mTJuu8BlNjRw9KKxbO+2YUOL9na0UqWJla8lbK5fIEF+dbUniOaoZlFD2cJaPB6c43eVc= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=jvdsn.com; spf=pass smtp.mailfrom=jvdsn.com; dkim=pass (2048-bit key) header.d=jvdsn.com header.i=@jvdsn.com header.b=Mp92CaG0; arc=none smtp.client-ip=129.153.194.31 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=jvdsn.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=jvdsn.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=jvdsn.com header.i=@jvdsn.com header.b="Mp92CaG0" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=jvdsn.com; s=mail; t=1776703728; bh=kcNBUdT1rAgPFcu1gMxQPN3qd7G83ytFuTKyNDWeDyM=; h=Date:Subject:To:Cc:References:From:In-Reply-To; b=Mp92CaG088gDQZE4zp+wImm0FzCZB0mbLDI6k4yg9F13yV4mtY/Zv2ojTJbeAEiDu 7/2JAC3Oz+mO5BjgUr6vkMG5Z3LFf/21bbAT7TJ8iqwYpdLaxFbRZIgAoM8k2oUWk9 dKtXuQR4NkxkarQjw1OhGLRbub3PWnfbz+oOIed+lcKmSOQmM2DotF7jlydOjISDc+ cRyVxWnA/YD5mzsmxR2ZmLDPTqKi4w/JozxeyXjj70C+GNNFDu5Z2XgSwrnKb9VLmM m5FnOejYY0V1BKHTTUynFf4b9TUxSkNxzf4nV56MCtMSl9Y4y/P4smOFyKnOYE2qsh b+lx/gfs5LjIA== Message-ID: <9e13b506-576f-4753-96e4-9e12085627bc@jvdsn.com> Date: Mon, 20 Apr 2026 11:48:47 -0500 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Subject: Re: [PATCH 36/38] crypto: drbg - Remove redundant reseeding based on random.c state To: Eric Biggers , linux-crypto@vger.kernel.org, Herbert Xu Cc: linux-kernel@vger.kernel.org, Stephan Mueller , "Jason A . Donenfeld" References: <20260420063422.324906-1-ebiggers@kernel.org> <20260420063422.324906-37-ebiggers@kernel.org> From: Joachim Vandersmissen Content-Language: en-US In-Reply-To: <20260420063422.324906-37-ebiggers@kernel.org> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Hi Eric, On 4/20/26 1:34 AM, Eric Biggers wrote: > We're now incorporating 32 bytes from get_random_bytes() in the > additional input string on every request. The additional input string > is processed with a call to drbg_hmac_update(), which is exactly how the > seed is processed. Thus, in reality this is as good as a reseed. > > From the perspective of FIPS 140-3, it isn't as good as a reseed. But > it doesn't actually matter, because from FIPS's point of view > get_random_bytes() provides zero entropy anyway. > > Thus, neither the reseed with more get_random_bytes() every 300s, nor > the logic that reseeds more frequently before rng_is_initialized(), is > actually needed anymore. Remove it to simplify the code significantly. > > (Technically the use of get_random_bytes() in drbg_seed() itself could > be removed too. But it's safer to keep it there for now.) It's fair to say that the additional input is as good as a reseed (if FIPS is not considered), but then is there any reason to keep get_random_bytes() in drbg_seed()? You say it could be removed but it's safer to keep it there for now? In what way is it safer? The additional input is mixed into the HMAC_DRBG state prior to generating random bits, so should already provide sufficient assurance that the generated bits incorporate the output of get_random_bytes()? > > Signed-off-by: Eric Biggers > --- > crypto/drbg.c | 107 +++++--------------------------------------------- > 1 file changed, 9 insertions(+), 98 deletions(-) > > diff --git a/crypto/drbg.c b/crypto/drbg.c > index 7fd076ddc105..bab766026177 100644 > --- a/crypto/drbg.c > +++ b/crypto/drbg.c > @@ -91,22 +91,15 @@ > > #include > #include > #include > #include > -#include > #include > #include > #include > #include > > -enum drbg_seed_state { > - DRBG_SEED_STATE_UNSEEDED, > - DRBG_SEED_STATE_PARTIAL, /* Seeded with !rng_is_initialized() */ > - DRBG_SEED_STATE_FULL, > -}; > - > /* State length in bytes */ > #define DRBG_STATE_LEN SHA512_DIGEST_SIZE > > /* Security strength in bytes */ > #define DRBG_SEC_STRENGTH (SHA512_DIGEST_SIZE / 2) > @@ -135,13 +128,10 @@ struct drbg_state { > struct mutex drbg_mutex; /* lock around DRBG */ > u8 V[DRBG_STATE_LEN]; /* internal state -- 10.1.2.1 1a */ > struct hmac_sha512_key key; /* current key -- 10.1.2.1 1b */ > /* Number of RNG requests since last reseed -- 10.1.2.1 1c */ > size_t reseed_ctr; > - size_t reseed_threshold; > - enum drbg_seed_state seeded; /* DRBG fully seeded? */ > - unsigned long last_seed_time; > bool instantiated; > bool pr; /* Prediction resistance enabled? */ > struct crypto_rng *jent; > const u8 *test_entropy; > size_t test_entropylen; > @@ -237,76 +227,10 @@ static void drbg_hmac_generate(struct drbg_state *drbg, > drbg_hmac_update(drbg, addtl1, addtl1_len, addtl2, addtl2_len); > > memzero_explicit(addtl2, sizeof(addtl2)); > } > > -static inline void __drbg_seed(struct drbg_state *drbg, > - const u8 *seed1, size_t seed1_len, > - const u8 *seed2, size_t seed2_len, > - enum drbg_seed_state new_seed_state) > -{ > - drbg_hmac_update(drbg, seed1, seed1_len, seed2, seed2_len); > - > - drbg->seeded = new_seed_state; > - drbg->last_seed_time = jiffies; > - drbg->reseed_ctr = 1; > - > - switch (drbg->seeded) { > - case DRBG_SEED_STATE_UNSEEDED: > - /* Impossible, but handle it to silence compiler warnings. */ > - fallthrough; > - case DRBG_SEED_STATE_PARTIAL: > - /* > - * Require frequent reseeds until the seed source is > - * fully initialized. > - */ > - drbg->reseed_threshold = 50; > - break; > - > - case DRBG_SEED_STATE_FULL: > - /* > - * Seed source has become fully initialized, frequent > - * reseeds no longer required. > - */ > - drbg->reseed_threshold = DRBG_MAX_REQUESTS; > - break; > - } > -} > - > -static void drbg_seed_from_random(struct drbg_state *drbg) > - __must_hold(&drbg->drbg_mutex) > -{ > - u8 entropy[DRBG_SEC_STRENGTH]; > - > - get_random_bytes(entropy, DRBG_SEC_STRENGTH); > - > - __drbg_seed(drbg, entropy, DRBG_SEC_STRENGTH, NULL, 0, > - DRBG_SEED_STATE_FULL); > - > - memzero_explicit(entropy, DRBG_SEC_STRENGTH); > -} > - > -static bool drbg_nopr_reseed_interval_elapsed(struct drbg_state *drbg) > -{ > - unsigned long next_reseed; > - > - /* Don't ever reseed from get_random_bytes() in test mode. */ > - if (drbg->test_entropylen) > - return false; > - > - /* > - * Obtain fresh entropy for the nopr DRBGs after 300s have > - * elapsed in order to still achieve sort of partial > - * prediction resistance over the time domain at least. Note > - * that the period of 300s has been chosen to match the > - * CRNG_RESEED_INTERVAL of the get_random_bytes()' chacha > - * rngs. > - */ > - next_reseed = drbg->last_seed_time + 300 * HZ; > - return time_after(jiffies, next_reseed); > -} > - > /* > * Seeding or reseeding of the DRBG > * > * @drbg: DRBG state struct > * @pers: personalization / additional information buffer > @@ -323,11 +247,10 @@ static int drbg_seed(struct drbg_state *drbg, const u8 *pers, size_t pers_len, > { > int ret; > u8 entropy_buf[(32 + 16) * 2]; > size_t entropylen; > const u8 *entropy; > - enum drbg_seed_state new_seed_state = DRBG_SEED_STATE_FULL; > > /* 9.1 / 9.2 / 9.3.1 step 3 */ > if (pers_len > DRBG_MAX_ADDTL) { > pr_devel("DRBG: personalization string too long %zu\n", > pers_len); > @@ -353,13 +276,10 @@ static int drbg_seed(struct drbg_state *drbg, const u8 *pers, size_t pers_len, > else > entropylen = DRBG_SEC_STRENGTH; > BUG_ON(entropylen * 2 > sizeof(entropy_buf)); > > /* Get seed from in-kernel /dev/urandom */ > - if (!rng_is_initialized()) > - new_seed_state = DRBG_SEED_STATE_PARTIAL; > - > get_random_bytes(entropy_buf, entropylen); > > if (!drbg->jent) { > pr_devel("DRBG: (re)seeding with %zu bytes of entropy\n", > entropylen); > @@ -399,11 +319,12 @@ static int drbg_seed(struct drbg_state *drbg, const u8 *pers, size_t pers_len, > } > > if (pers_len) > pr_devel("DRBG: using personalization string\n"); > > - __drbg_seed(drbg, entropy, entropylen, pers, pers_len, new_seed_state); > + drbg_hmac_update(drbg, entropy, entropylen, pers, pers_len); > + drbg->reseed_ctr = 1; > ret = 0; > out: > memzero_explicit(entropy_buf, sizeof(entropy_buf)); > > return ret; > @@ -461,31 +382,25 @@ static int drbg_generate(struct drbg_state *drbg, > /* 9.3.1 step 5 is implicit with the chosen DRBG */ > > /* > * 9.3.1 step 6 and 9 supplemented by 9.3.2 step c is implemented > * here. The spec is a bit convoluted here, we make it simpler. > + * > + * We no longer try to detect when random.c has reseeded itself and call > + * drbg_seed() then too, since drbg_hmac_generate() adds bytes from > + * random.c to the additional input, which is a de facto reseed anyway. > */ > - if (drbg->reseed_threshold < drbg->reseed_ctr) > - drbg->seeded = DRBG_SEED_STATE_UNSEEDED; > - > - if (drbg->pr || drbg->seeded == DRBG_SEED_STATE_UNSEEDED) { > - pr_devel("DRBG: reseeding before generation (prediction " > - "resistance: %s, state %s)\n", > - str_true_false(drbg->pr), > - (drbg->seeded == DRBG_SEED_STATE_FULL ? > - "seeded" : "unseeded")); > + if (drbg->pr || drbg->reseed_ctr > DRBG_MAX_REQUESTS) { > + pr_devel("DRBG: reseeding before generation (prediction resistance: %s)\n", > + str_true_false(drbg->pr)); > /* 9.3.1 steps 7.1 through 7.3 */ > len = drbg_seed(drbg, addtl, addtl_len, true); > if (len) > goto err; > /* 9.3.1 step 7.4 */ > addtl = NULL; > addtl_len = 0; > - } else if (rng_is_initialized() && > - (drbg->seeded == DRBG_SEED_STATE_PARTIAL || > - drbg_nopr_reseed_interval_elapsed(drbg))) { > - drbg_seed_from_random(drbg); > } > > /* 9.3.1 step 8 and 10 */ > drbg_hmac_generate(drbg, buf, buflen, addtl, addtl_len); > > @@ -562,13 +477,10 @@ static int drbg_kcapi_seed(struct crypto_rng *tfm, > */ > > /* 9.1 step 4 is implicit in DRBG_SEC_STRENGTH */ > > drbg->pr = pr; > - drbg->seeded = DRBG_SEED_STATE_UNSEEDED; > - drbg->last_seed_time = 0; > - drbg->reseed_threshold = DRBG_MAX_REQUESTS; > memset(drbg->V, 1, DRBG_STATE_LEN); > hmac_sha512_preparekey(&drbg->key, initial_key, DRBG_STATE_LEN); > > /* Allocate jitterentropy_rng if not in test mode. */ > if (drbg->test_entropylen == 0) { > @@ -671,11 +583,10 @@ static inline int __init drbg_healthcheck_sanity(void) > if (!drbg) > return -ENOMEM; > > guard(mutex_init)(&drbg->drbg_mutex); > drbg->instantiated = true; > - drbg->reseed_threshold = DRBG_MAX_REQUESTS; > > /* > * if the following tests fail, it is likely that there is a buffer > * overflow as buf is much smaller than the requested or provided > * string lengths -- in case the error handling does not succeed